Directorate General of Highways (Bina Marga)

and
Indonesia Highway Corporation (Jasa Marga)

FINAL REPORT

Highway Toll S t u d y
(TANO. 2762)
Asian Development B a n k

J U I Y 2,1997

Submitted by

Jan A. Johnson Harun al-Rasyid S. LUBIS

ALM Resources Ltd. Traffic Laboratory,
Sooke, B.C., C a n a d a Institute of Technology, B a n d u n g
 ABBREVIATION

Agency)

liina M'irga Directorate. of l + i g l i \ v ~ \[lcpirtmc~nt

,, ot I'uLiIic Work

I3K130U : Bcsar Kcuntung:,in I3iai.a Opcr,isi l<cmci~irci~in

(Vcliiclv operating cost s,i\,ings, including tiiiw sci\,ing)

BOI' : Build - Operate - Transtcr

BTO : Buici - Transfer - O p c r Ci k ~

B rs
' : Biro Pusat Statistik (Bui-c.,iLi tor Central Statistic)

CI'I : Consumer Price Incicx

: Economic I11te1~1al
Rat<! ~t I<ctLtrli

FIRR : Financial Intcrnal Rate of Return

Fcx : Foreign Exchange

1 IDM : Highway Design a n d Maintcixincc Model

IIHCM : Indonesia lHigliwa\. Capacit\r Mnn~ial

1DC : Interest during construction

IRMS : Interurban Road Management S\rsteiii

JM : PT'.Jasa Marga ( Indonesia Highway Corporation)

JV : joint Venture

KEPPRES : Keputusan Presiden (Presiclcntial Decree)

MCI : Mitra Pacific Consulinclo

MOF : Ministr~,of Finance

PCI : Pacific Consultant International

PCU : Passenger Car Unit

PPN : Pajak P e r t a ~ i i b ~ ~ hNil,ii

a n ( V ~ l u eA d d e d Tax)

PW : Department of Public \Voi-k

I? i 11ba 11s : liurc'iu tor I< &I], P.T J'is'i k4argCi

Rp. : Ru piali (I ncloncsi~inCUrrciici~)

:1 11' ' : Rcnclel, Palmer & 'frini tton

RUCM : R o a d User Cost Mocicl

-1r; 13 : '1' a1:

l t- t- IZCgLIlator\. l3o'lrLi

voc : Vehiclc Operating Cost

vo1 : Value of Tiine

EXECUTIVE SUMMARY

1. OVERVIEW
The Government of Indonesia and Jasa Marga ha\,e had almost 20 years of
experience with toll roads, starting with the opening of the jagorawi Toll R o d in
1978. In that time, under the orfi"l1imtion of J a w Marga, sc\'cn toll roads ,111Li
two bridges totaling 324.2kill ha\^ been constructed and operated. Sincc IWh,
BTO, BOT a nci joint opcra t i o i i pa r t iic rsli ips wit I1 outs i dc coni pa n ics t 111-0 LI g h lx)t 11
state enterprises a n d private consortiurns ha\,e been encouraged. Widening
avenues of participation enables a larger toll road pwgratiimc to be undcrt,ikcs.n
and mobilizes private domestic and foreign capital for invcstmcnt in the higher
volume links of tlie road svsteiii.

2. OBIECTIVE AND SCOPE

The objective of this Higliwav Toll Stuciv 11'1s been to recommend r i i i cippropri'itc
toll setting and tariff adjustment mechanism which ciiii be applied by Jasa M'11-p
and joint venture companies for existing aiid planned toll roads. Several ii1cic.x
formula approaches using different measures of inflation, including consuiiier
prices indices were examined and discussed with Government, PT Jasa Mar31
(Persero) ancl ADB. In acidition, options for simplifying atid streainlining of
institutional procedures for tariff cietcrminatioii and adjustment have been
proposed for consideration b\' Go\wmiient. The goals for cievclotiiii:; 1'
methociology and process for setting a n d revising rates included tlie folhviiig:
- Preserving atid extending the pliilosophv a n d intent of
Law NuIiibcr 13, 19SO On Roads to acicircss contemporarv I3CX
issues

- Full cost recovery

- Transpareiicv

- Consistencv 81 prcdictabilitv

- Provision of adequate private investor incentives

Achieving these goals clemaiids attaining tlie best balance among sometimes
competing objectives. This balance is required in order to allocate development
resources in the most optimal fashion ancl maxiinize the coiitribution to the I c ~ q -
term, over-arching objectives of national and regional development and
advancement of equiti?. These objectives, originally embodied in Law N L I I ~ T C I -
u, (reproduced in Appendix 1)include the following:
economic efficiencv
financial viability

I
 affordabilitv to users

The approach taken in this study was to first examine the behavior of traffic anci
revenues 011 existing toll roads and to review the histor\, o f toll tariffs and the
iiiipact o f tariff adjustments upon traffic and diversion. To accomplish this, a set
of case studies were selected i n consultation with jasa Marga which reprcscnted
a cross section of toll roads in !a\u. C;lsc studies included the following:

lass Miirzci Priva tc loin t Vcntu re

Urban
Mature: CaMUng-7-omallg Cawang-Ti. Priok
Developing: k
Jatingaleli-Kra~i\.a Pd. Pinang-Kp. Ranibut,in

I ntcru rba I 1
Mature:
Developing: I-'adalaraiig-Cileun\~i Tangcrang Barat-Mcrak

Analvsis included a r c \ % w of traffic volume a n d coinposition on these selected

toll roads anci utilized available origin/destination data to obtain profilcy on USCY
behavior before ancl after toll taritt changes. With the assistance and
cooperation of Jasa Marga, travel time a n d speed survevs were carried out a n d
VOC's cstiniated f o r toll and non-toll alternative routes. With this set ot
finclings, analysis of Jasa Mar31 traffic data, and past studies on demand
elasticity, estiniates of traffic diversion \\'ere made.

A re\.icw of the costs of toll roads- capital, operating, ancl maintenance, ~ v a s

carried out. This also included a comparison of the toll tariff for different \rehick
classes with respect to road maintenance cost assignment bv vehicle type.

To assess the impact of different toll levels upon economic anci financial returns,
a case study of a project under investment process was analyzed to assess the
consequences of different toll levels upon the satisfaction and trade-offs anicing
these yield objectives.

The economic rate of return (ERR) with and without toll diversion effects bvas
calculated ancl compared with the financial rate of return to the private joint
venture. Sensitivity tests on critical variables including construction cost o\w-
runs, opening year traffic, and traffic growth rates were carried out as well as
tests of toll rates and diversion effects on traffic and toll revenue vields.

A critique of risk allocation measures ancl current government policv regarding

land acquisition costs, toll tariffs, ancl revenues was undertaken. The assessment
took into account the connection between le\rels of risk, variabilitv of financial
return and the implications for tariff setting and cost recoverv.

The analyses of the case studies provided a fraiiicwork for the analvsis of tlie
toll tariff setting and adjustiiicnt mechanism. A toll tariff adjustment formula
has been proposed and options p u t forward for institutioiializing rate rcvicci.
and adjustment for existing and planned toll roads.

3. STUDY FINDINGS

3.1 Toll Road Traffic Volumes and Revenues

Traffic voluiiic has increased dramaticallv on the toll road system. In the
Jabotabck region, growth has been highest, For the Jakarta -Cikaiiipek, a\waglt
cvowtli lias been in tlie order of 77'X annuall\, since 1990, but as shown in thc
a
below table, lias started to moderate. For toll roads outside of the Jakart'i 'ire'>
of influence, growth has also iiiodcratcci from their initiallv high rates sliortl\.
after opening. This is notable on the Randung bvpass, Padalarang-Cileun\i '>lid
the section in Semarang, Jatingaleh-Krapva k.
Traffic Growth (ADT: % avg. chg/yr).
1990-'95 1993-95 __
1994-'95
Cawang-Tomang 16% 14% 13% 11%
Cawang-Tg. Priok 23% 17% 16% 9%
Jalingaleh - Krapyak 15% 18% 8% 8%
Jakarta-Cikampek 22% 19% 18% 13%
Padalarang -Cileunyi 34% 22% 17%
Dupak-Kebomas 40% 34%

The same phenomena can also be noted in revenues, but which have in gencral
grown faster than traffic. I n the case of Cawang-Tomang-Taiijung Priok toll
roads, this is explained largely by tariff increases in 1992 and 1995. For the other
cases, growth in Tariff Class 11 traffic was the major cause.

Revenue Growth (ADT: % avg. chg/yr).

1990-'95 1993-95
Cawang-Tomang 25% 24% 16% 16%
Cawang-Tg. Priok 26% 28% 19% 15%
Jalingaleh - Krapyak 20% 26% 8% 8%
Jakarta-Cikarnpek 31% 26% 22% 17%
Padalarang -Cileunyi 31% 25% 19%
3.2 Toll Tariffs, Elasticities, and Vehicle Operating Cost Savings
Toll roads organized by Jasa Marga tend to have lower tariffs per kn1,
particularly interurban toll roads. Padaleunyi, Cikampek, and Jagorawi charge
in the Rp. 80-90/km range, while the privatclv organized Jakarta-Merak toll road
tariff is Rp. 153/km. Tolls on Jakarta urban toll roads tcnci to Lx. the highest,
from Rp 250/km up to more than 550/ kin, the latter again for privatel\.
organized segments of the system.

Studies on travel elasticity with respect to tariffs on interurban toll roads l ~ i \ t -

indicated values in the order of -035 for Class I traffic and from -.OX to -0.33 f o r
Class 11. The cross section of toll road cases examined h o w c \ ~ showed
r that in
most cases, traffic tended after a few months to recover to its prwious growth
trend.

I n several studies for Indonesian toll roads diversion effects haw becn modeled
as a function of toll tariff divided bv travel time savings. Other approachc~s
~41icliamount to the same have been the construction of diversion cur\'es A S '1
function of total traffic volume in the corridor which in turn affects congestion
and user costs 011 the alternative, non-tolled road. Casc studies confirmcd t h c
hypothesis that once corridor traffic reaches 30,000 ADT, Class I traffic diversion
increases more rapidly, with Class I1 traffic following once total traffic reaches
-10,000 ADT.

Field surveys on case study roads measuring comparative distance and travel
time between toll and non-toll alternative roads indicated wide variations in
estimates of user savings compared to toll tariffs, ciepending upon the vchiclc
operating cost (VOC) model employed and the value of passenger time. A
comparison of the results of the new revised PCI VOC model commissioned by
Jasa Marga was made with the results from the original PCI moclel used for
BKBOK measurement. This comparison indicated that for several urban toll
roads the tariffs match or exceed user savings for what is considered a realistic
(Rp 7000/1ir/vehicle) valuation of time for users outside of Jakarta. If the lower
value of time (3281 Rp/hr) from the Indonesian Highway Capacity Manual
(IHCM) is applied with the new VOC model, only tariffs on the Cikampek ancl
Semarang case studies meet the 70% of user savings guideline applied by Bina
Marga.

The aspects of vehicle operating cost ancl value of time are discussed in detail i n
Chapter 4. A simple VOC model specification which avoids laborious data in its
implementation is preferred for the purpose of toll tariff adjustment. In the
meantime the choice for such a.model is the revised PCI model. However, it was
noted in meetings with Bina Marga that studies are on-going to validate the
revised PCI formula as well as examine further an updated realistic value of

4
time. In parallel with studies by the Institute of Road Engineering on calibration
of the HDM V O C model for the Interurban Road Management Systems (IRMS),
the Transport Road Research Laboratorv of tlie U.K. has just started studieh on
the value of time.

4. CIAWI-SUKABUMI TOLL ROAD PROIECT CASE STUDY

4.1 Economic Analysis

A n cconomic analysis was carried out on tltc poposeci toll r o a d project bct\v<-t’n
Ciawi and Sukabumi, part of the link between Bogor and Bandung, and ‘111
alternative to the congested Puncak route. Benefits were based on the p)tcntial
traffic which would divert to the new highwav facility, with and without tolls,
coinpared with continued use of the existing 11ighway. Benefits were nieasurcci
in terms of economic \ d ~ i c l eoperating cost and passenger time sa\rings
potentially gained on the new highway facilih,, coinpared with costs of usin;: tlw
existing road. Benefits to potential generated traffic along the corridor were a h 1
taken into account, am1 following standard practice, estimated a t one halt the-
unit benefits to normal growth traffic.

Estiiiiates of traffic diverting to the new highwav with and without tolls wcrc
based on the volume/capacity and speed- flow relationships from the I tiCh.1.
TIie nuimber of trips diverting to the new highway woulcl be however limited h\.
tlie tiumber of through trips. Local trips were assumed to remain largely on the
old road. The result of examining the project as a non-toll facility with its lower
capital cost would be an economic internal rate of return (EIRR) of 29.8%. Wlicn
domestic costs and time savings benefits are adjusted using standard conversion
factors’ and opportunity costs to reflect the real cost of non-tracleci components,
the effect is to increase the EIRR to 30.7%.

Organizing the Ciawi-Sukabumi highway as a toll facility would increase

construction and operation costs and reduce utilization because of the impact of
tolls upon diversion. This would have the effect of increasing the adjusted
economic cost from 372.1 billion Rp. to 415.6 billion. It was assumed that cost for
land acquisition would remain the same as for a non-toll facility.

The base case economic return for the project organized as a BOT toll road was
found to be 24.1%. Increasing costs by 10% would lower the EIRR to 22.8%. A
combined scenario of costs increased by 10% a i d traffic reduced by 10%would
reduce the EIRR to 21.8 %. An increase in toll levels would reduce traffic and

’Appendix 12. ADB Project Appraisal. Ciawi-Sukabumi Toll Road Project, May, 1997: SCF of0.943
applied on domestic, ion-traded components. and unskilled labour valued at 75% o f wase rate: Non-wol-h
and umkilled labour travel time saving similarly adjusted 011 benefit side.
consequently the volume of user benefits, which in turn reduces potenti'il
generated traffic. A 25% increase in tolls ~ ~ o u l decrease
cl diversion from thc
existing road, to 35% and 16%shares for Class 1 and Class 11 traffic after opening.

Conversely, decreasing the tariff level ~ o u l c increase

l diversion and the n u m l ~ c ~ r
of users benefiting from the new facility. Diversion has been estiniatccl t o
amount to 58% a n d 35% after opening for the two user groups. BY 2005, t h L w
shares coulcl climb to 65% and 39% of corridor traffic. The EIRR in this cnsc-
woulcl be 26%.

4.2 Financial Analysis

Costs, toll tariff schedules, and the structuring of financing tor thc financLiI
analysis have been based on tlie estiniates and proposals prepared b \ z PI' UuL>!i,i
Marga Utama. Total costs cif the project including assumed S'X inflation during
construction, and financing costs amount to SOS.5 billion R p . Excluding intercst
during construction, the estimated cost wciuld be 668.5 billion R p in currcmt
terms. In constant 1996 Rp. terms the financial cost woulJ be 556 billion Rp.

Several sensitivity tests were carried out to cleterminc the impact on the financiil
rate of return from variations in kev parameters. Applving the current Rp. cost
streams against traffic flows and tariffs assuined lw RPT/Bukaka, that is with
generated traffic starting at 25 % of normal traffic in opening year a n d increasing
to 30% bv 2005, the base financial internal rate of return (FIRR) o n the projc'ct
woul~lbe 20.9% over the concession period. With tlie Bukaka financial structure
as proposed, the return on equity would be 26.3%.

I n constant 1996 Rp. terms, this FIRR cash flow woulcl \field a project
. , return of
14%. This coinpares favourablv with a weighted real cost of capital i n the orclcr
of 7%.

A decrease in tariffs by 25% would increase Category I traffic by 20 to 25% and

Category 11 traffic by around 15%. This would increase benefits to tlie econoniv,
raising tlie EIRR to 26% from the base toll case return of 24%. From the private
investor point of view, the impact would be negative with lower toll road
revenues, decreasing the FIRR in constant terms for the project to the order of
12%from 14%.

Increasing tariffs by 25% would decrease diversion to tlie toll road, reducing
traffic by 12 to 15 %. This would have the impact of reducing the EIRR to the
economy to the order of 20%, but increase returns on tlie project to the order of
16%.
Given tliis situation wliere tlie imposition o f higher tariffs has the effect of tolling
- off traffic and reducing the benefits to the economy o f potential transport
savings, but perceived congestion relief and time savings remain high enough to
still produce increasing toll revenues, the main reference point for BOT tariff
setting becollies the financial return to the in\wtor. Tariff-traffic sensitivit\. tcssts
indicate that the proposed tariffs arc still l~clow the investor rcvenuc~-
maximizing level. If re\’ctitic maximizing bccanic the tariff-setting critcri‘i,
economic benefits would be reduc<d w i t h thc ISIRR to the ecoiioni\~f‘illiiig
below 20%.

Froin the investor’s point of view, the financial return to the project must ‘1s 1‘
minimum exceed the cost of capital. Based on the Bukaka financial plan w i t h h0
: 40 domestic : offshore debt a t interest rates of 20% a n d .ll%respccti\d\., tht,
weighted nominal average cost of debt is in the order of 17% when coiiiniitm~~nt
fees ancl charges arc included. Tlic forecast project rate of return in current
terms a t 20.9% yields a margin of 4 percentage points over this. A SIOWCY
clcvclopnicnt of generated traffic, not reaching 20% of normal growth traffic until
five years after opening w o u l d rcciucc the FIRR in current terms to 17.h%,o r in
constant, real terms, to tlie order of 11’%.

4.3 Conclusioiis
Return on investor equity, tlic other benchmark for finaiicial fcasibilitv, must ,it
least be comparable with returns available in otlicr sectors of comparable risk.
A margin of 5 percentage points over ‘1 5 to 10-vcar average doniestic cost ( i t
capital (nominal) at 20-21% could be taken as a reasonable long term liurcllc rate
of return. The forecast nominal 26% return on equitv for tlie financial base crlsc‘
fits within this zone of imwtor liurclle rates and is lower than what might be
acliievecl in otlier sectors.

A combination of adverse outcomes in the project such as a 10% cost over-run

combined with 10% less traffic than anticipated would reduce real returns to tlic
project to a real 11.7%. Given the cost of capital and tlie risks on costs ancl
revenues, tlie tariff levels as proposed represent a reasonable compromise in tlie
trade-off between economic returns to the national econoinv and returns to
private investment

5. RISK ALLOCATION AND MITIGATION

In contrast to some otlier countries offering toll road concessions, Indonesia does
not offer guarantees of opening year traffic, traffic growth, or minimum levels of
income from toll tariffs. Toll tariffs and rate adjustments are established by
Presidential decree and these are not officially set until tlie road is essentiallv
._
colilpletecl. Hence, risk coverage and distribution for prospective investors are
built into the terms of concession. These terms include the opening toll tariffs,
revenue sharing arrangements with Jasa Marga, the length of concession, and the
construction period. While rapid]\, growing traffic levels produce a huovalnt
atiiiosphere for revenue projections, risk balancing among the Government, as^
Marel, ancl the bidders has resulted in lengtliv and complicated negotiations.

5.1 Risk allocation/Miti~ation

For toll roads which undergo a bid process, lass Mare1 has adopted a numbcr of
nieasures to rcciuce the risk of default b \ r contractors before o r during tlw
construction process. These uwasures include pre-clualification of firms based ( i n
financial capability and construction expericnce, tencler document provisions,
and analvsis of cash flow and Ieiidcr commitments in tlic process of hid
evaluation.

Major sources of risk in cost o\'erruii and timc delay arc largel!s borne b\. th(,
bidder. While Jasa Marga guarantees that the government will acquire tlic right
of wav, the bidder funds land acquisition cost and bears the risk f o r cost over-
runs and delay. While negotiations around this p i n t include the extension o f
concession periods as compensation, considerable front-end risk still persists.
Room exists for the Government to assume more of tlie responsibility for land
acquisition, including exercise of powers to freeze land values along the riglit ot
w a y to curb speculative inflation of land prices.

As discussed in the section below, toll tariff risk has two elements; these arc
comprised of firstlv, the initial opening tariff and secondly the meclianism a n d
policy for future adjustments. The concern of investors is that adjustments will
not keep pace with debt service, operating, maintenance, and periodic
maintenance costs, particularly in an atmosphere of escalating prices,
deteriorating exchange rates, ancl significant portions of off-shore debt to repay.

Under current practice, the opening toll tariff is not officially accepted until thc
road is essentially complete. Jasa Marga can assure best efforts in processing
tlie proposed rate through officials channels up to the President's desk, but until
the KEPPRES is signed, the risk is with the investor. The contract with Jasa
Marga in the Authorization Agreement contains a proposed forinula for bi-
annual tariff escalation based on the consumer price index. However, risk
remains that adjustments inav not be approved in a future administration. This
uncertainty has perhaps reduced the number of potential bidders, and for those
accepting the risk, led to higher premiums to cover the perceived risk
6. TARIFF SETTING AND ADIUSTMENT MECHANISMS

The nature of toll road investment inakes this particular forin of BOT venture
more difficult to structure and finance than other forms of infrastructure o r
utilities. Toll road in\vstnicnt costs arc concentrated at the beginning ot tl1<>
project, pav-back periods are gcwerall\~long, anci rc\'cnucs arc more difficult to
predict than for other tiyes o f BOT investments. Unlike a power plant scr\.in:;
one bulk customer, toll road usage is the product of thousands of incii\.idual
customer decisions with the attendant wide variety of trip purposes and
willingness to pay, For these rcascins, rcvcnue-maxiiTiij5iiig tariffs anci higher
yields o n equitv are sought b\r investoi rates of return tend to be higher than
for other types of BOT projects in o r d e r to attract investors.

The cornerstone of a sound BOI policy is tlie establishment of 'i predictciblL*

svstein for setting and adjusting toll tariffs. The most successful approach to this
svstem is tlie incorporation in tlie bidding procedure anci autlioriz,ition
agreement of a foriiiula which rcsol\vs uncertaintv on tariffs anci lienci~
increases investor confidence on revenues. This resolution can be met through .i
tariff adjustment mechanism with cost indexing.

6.1 Current System

Under the original 1980 Icgislation anci later implementing regulations,' thl,
designation of a road section as a toll road anci toll tariffs by type of vehicle ' 1 ~ '
cieterniined by the President at tlie proposal of tlic Minister of Public Works. A
toll road can only be officially designated when it is essentiall\. coinpletc and
tolls cannot be stipulated until this forinality is carried out. As presentl\,
administered, this process requires a new Presidential Decree (KEPPRES) ever\'
time a rate is filed, inclucling opening of new interclianges on existing toll roads
where no change in vehicle/km rates may be entailed.

A lengthy and bureaucratic process is entailed to inove the tariff' setting and
adjustment proposal up to the Presiclent's desk. Starting with the investor
submitting the proposal to Jasa Marga for review, it is then passed to the
Ministry of Public Works. Here it is reviewed in detail again bv the
Subdirectorat Jalan To1 for conformance with BKOK guidelines. The proposal is
then submitted to the Minister of Public Works for approval. It then is passed
up to the Cabinet Secretariat for review, and if approved, will be submitted to
the President for decision.

' "Law # 13 of 1980 On Roacis" governing tlie administration of the road network system , a n d
Government Regulation #S, 1990 Regarding Toll Roads.
Tender documents issued by Jasa Marga in tlie 1995 bidding rounds stated that
initial toll tariffs must be proposed by investor with reference to other toll roads.
Guidelines also included the maximum tariff limit of 70% of BKBOK. The
general guidelines suggested that for classifying tariffs bv vehicle tvpe, Categor!,
IIA (medium, '-axle vehicles) tariffs shoulci be about 1.5 times Categorv I (light
vehicles) and Catcgorv IIB (heavy, >7 axles), about 2 times Categor\, I.

Be\wnd pledges of best efforts, toll road B 0 1 ' projects have no binding agrcemcmt
on tariffs until tlie facilitv lias been constructed and the President lias ' ~ p p r o \ . ~ d
tlic application. Similarly, there is no guarantcc that future applications tor taritt
l a systematic (11' regular basis. A s' L1'I S' C- U .S.S C1~
adjustments will be a p p r o \ ~ c on
previously financing for projects must be arrangcd i n a framework which cannot
guarantee a tariff which will necessarilv co\rer debt service costs. This clement
of uncertainty and risk must then be incluJcd a s a n implicit premium on the cost
of financing, particularlv for offshore funds.

Within the philosopliv of Law Number 13, it is necessary to develop ii set ot

aciministrative guidelines which will further interpret user savings for urban and
interurban toll roads. Increasingly as traffic builds in certain corridors,
congestion and delay costs will preciominate in the user savings function against
which tariffs are compared. Particularlv for urban toll road systems, congestion
charging by time of day should be introduced with lower rates for night anci oft-
peak users. Lower tariffs for night-time use of interurban toll roads should also
be considered for trucks anci buses wliicli are presently tolled-off sevc'ral
stretches of the toll road system. This would encourage better utilization of tlic>
toll road assets and reduce lieav\T traffic on tlie non-tolled public network.

The index formula and guidelines contained in tlie Authorization Agreement

represent a n approach which is already familiar to investors. Proposals to
further elaborate this formula in terms of framework for calculation anct
procedures for implementing are set out below. Finally a number of options
have been offered to formalize tlie process and provide an institutional context
for applying this tariff adjustment mechanism.

- Regional differences sliould be taken into account in the CPI

calculation as well as the user saving reference ceiling. Bureau of Statistics
publishes monthly consumer price indices for the 27 provincial capitals, anci
these sliould be used as a reference for regional CPI adjustment. This woulcl
reduce the dominance of Jakarta in the index, which would otherwise tencl to
distort increases.

. Because wholesale price indices on public works road and bridge

construction have gone up less on average than consumer prices over tlie last
vears, 7.1 % compared with 8.1% since 1993, routine and periodic niainteiianw
expense would not be expected to increase at the same rate as general inflation.
The B E index could be used for this coniponent of operating cost.

- operating and niaintenancc costs are a small percentage of total costs and
price indices could arguablv rcflcct onl\r the increases in that component.
Comparing these two elements o f local costs, operating and inaintenancc,
including periodic maintenance and toll equipment renewal against totcll
invcstincnt, on a present value o r capital r c c o \ u ~ basis,
' on-going, iion-tist*~i
opcrating and niaintenancc costs iiiiikc u p ahout XI'% of tiic total cost. I l o w ~ ~ ~ ~ ~ ~ r ,
short anci inecliuni term debt can be considered a \rariable cost as well iimi b\. this
argument should be partially indexed.

111 suniniarv, this proposed tariff adjustment formula with its compensation
factors anci user saving upper bounds would take the following form:

New Taritt (Tr) = Old Tariff * (1+a* [CPlr(iiew)-CPIr(ol~t)j/CPir(olJ)

where: a = proportion of lion-fixed operation costs to total costs,

CPlr = regional consumcr price index

Subject to:

N e w Tariff < [Vehicle operating cost savings bv corridor and time of

day + Passenger tinie savings bv region]

To handle cases where tariff adjustments would exceed the user saving
guideline, a monitoring process and institutional mechanism is needed. This
could continue to be carried out within the Ministry of Public Works
(Subdirectorat Jalan Tol) with Jasa Marga participation, or alternatively, lodged
with a Tariff Review Board outside the line ministrv .

6.2 Institutional and Procedural Options

Alternative institutional processes for toll-setting and adjustment have been sct
out for consideration. These options include the following:

Option I: Continue the present institutional system of determining the

opening tariffs, but simplify subsequent adjustments using the index formula
discussed above and delegating authority to the Minister of Public Works;

Option 11: President decides standard tariff adjustment formula for periodic
tariff adjustment and delegates approval over subsequent adjustments to
. .
Minister of Public Works; Review and monitoring mechanism within Ministry
of Public Works;

Option 111: President decides standard tariff adjustment formula for periodic
tariff adjustment and delegates approval over subsequent adjustments to
Minister of Public Works; review and nlonitoring mechanism placed with
Tariff Review Board outside line Ministry of Public Works, but reports to
Minister of Public Works.

The institutional options proposed for consideration should be compared on tlw

. o f three criteria:
basis
:

Simplification and standardization of procedures for tariff setting 'ind

adjustment;
Rcduction of risk premiums with more sccurit\. of acliieving toll tariffs ill; S L , ~
out i n Authorization Agreements;
Increased predictability of tariff adljustmcnts to cover operating, maintcn~incc~,
anci debt service costs.

Should tlie Government ciecide to procccci step-by-step in rcmodcling the,

institutional meclianism for tariff setting anci adjustnient, the options o u tlincd
above could be consicicred for transition steps.

7. Conclusions and Recommendations For Risk Mitigation and Tariff

Adiustment
Measures to iniprove the agreement tranicwork and regulatorv structure to
remove elements of uncertainty which are within Government control includi,
the following:
i) Make a consistent policy which limits investor exposure on land acquisition
cost overruns;
ii) Strengthen the long term planning function within Bina Mare1 tor
development of integrated urban ancl intercitv toll road svstems, anci make
available updated, comprehensive masterplans to guide land use a n d land
transport development;
iii) Incorporate in the tariff adjustment formula a mechanism which predictabl\.
reflects cost increases and institutionalizes a process for tariff review a n d
adjustment
iv) Simplify the tariff adjustment mechanism and delegate power to adjust rates
to tlie Minister of Public Works

Within this framework of general recommendations, several issues must be

addressed. These issues include both broader sectoral ancl macro-economic
concerns for medium and longer term planning as well as practical approachcs
for dealing with the current generation of BOT toll road projects.

I'
7.1 Sector Concerns
Sectoral concerns raised bv ADD include the need to increase transport efficient\..
Private sector provision of toll roads represents a means of ensuring cost
recoverv from users as well as r i i i instruiiicnt for funding improvements of thc
trunk network which w o u l d otherwise not be possiblc with limited domestic
funds o r soft international loans. A commercial approach to network
iniprovciiicnt in lieavil\~traidcd corridors allows inore rapid dcvclopmcnt tli'in
would otherwise occur a n d raises iiiorc tuii~isf r o m those users ablc to par than
would be available f r o m road L I S C charges
~ 'iiid taxes. Tolls arc also '111 cfficicmt
charge and redistributive tee on Taritt C,itcgorv I, higher income automotiilc
users wlio, pirticularlv in congested urb'in ~rciis,Ii'ive sufficicntlv high tinit,
values to be insensitive to toll tarifts.

Tliere are liowcvcr some caveats to this strict commercial view ot r o a d

development. Strong arguments exist for not abandoning tlie pliilosopli\. of
Law Number 13, in pirticular Articles 1.3 and 14 (see Appendix l), which states
that toll roads arc an alternativc to the p b l i c non-toll network, and must hcs
maintained as a choice to users bctwccn destinations and to serve local riccess
requirements. Bv their nature, high-standard toll roads must have a limitcd
number of interchanges and access points. Sccondlv, toll tariffs act a s r i n
adclitioiial charge on niovemcnt of goocis a n d people and the increased transport
cost will reduce tlie zone of infl~icnccand number of beneficiaries potentiall\,
affected by road clevelopmcnt. This will limit clcvclopment more than non-
limited access highways. The public benefits ot roacl development are reduced in
two ways; those users wlio are tolled-off either do not make intencled trips or
acid to the congestion on the public road. Increased transport costs are reflected
in passenger and freight tariffs and the economic return is recluced as shown in
the Ciawi-Sukabumi case study

This creates the need for a balanced network development plan. On one hand,
private investors must have assurance that a parallel, competitive high staiiciarcl
public freeway will not be constructed in tlie future to erode their revenue base.
On tlie other hand, local public roads must not be allowed to deteriorate to such
a level that local corridor users become captive to tlie toll road concession.

7.2 Toll Rate Adiustment Mechanisms

As discussed above, toll tariff acliustment needs to be institutionalized with a
predictable and transparent formula and a simplified, streamlined procedure for
administering. The current generation of BOT toll road projects has been
tendered and bid within tlie context of a proposed formula which reflects
inflation as measured by the national consumer price inde? (CPI). In this
context, bidders have proposed base and opening tariffs, anci these proposals
becoiiie part o f the Authorization Agreements.

Howc\,er, there are several arguments that the CPI indexing approach o\'cr-
rewards the concessionaire, in that in an\^ of the costs incurred b \ r the investor 'IK,
fixed in the mecliuiii and longer term. Operating and maintenance costs n i ~ l k c
u p less than half the annual costs, the major share o f costs being clebt scr\.icc.
The ADB supports the position that the indexing should be weighted o r
discounted to reflect onlv that portion of costs which increases with inflation.

Determining the variable portion of costs depends however on the timc hori;ron
taken. Given that tariff re\'ciiucs should cover debt service costs by the order ot
1.3 to 1.5 times? reliance u p o n the present domestic debt niarkct tor financing
iiieans heavy reliance on medium-term financing i n the order of 5 years. Hence,
over the medium term, a portion of debt scr\.icc anci the cost of capital 1wcoiiic.s
variable.

This then raises the question of off-shore financing with lower rates and longer
maturities available. For reasons of iiiacro-economic and balance of payments
stability as well as the long term cic\~clopniecitof the domestic capital iiiarkct,
local financing should be encouraged to the maximum extent. N o unneccssar\'
inducements should therefore be given to foreign borrowing. This then argucs
against inclusion of a foreign exchange acijustniciit factor in the tariff indexing
formula. Another reason raised bv the Govcrnmcnt for not further considering 1'
devaluation trigger in the adjustment is that Rupiah management policy anci its
gradual depreciation are alreadv reflected in domestic inflation rates, and adding
a foreign exchange factor in the forniula woulcl be double-counting.

Restricting then the adjustment formula to domestic price movement, economic

scenarios can be drawn for various cases to determine the proper form of
indexing. In an environment reminiscent o f Brazil or Mexico in tlie past ciec.idc3,
inflation accompanied by economic stagnation or shrinkage could lead to a
diminishing pool of customers able to afford tolls. On the other hand, much of
the success of the Jakarta urban toll roads can be attributed to per capita incomes
(and value of time) expanding two or more percentage points ahead of inflation.

Concerns with full CPI indexing also include the fact that high consumer
inflation in heavily weighted items such as food stuffs, housing, and clotliing has
limited relation to non-labour components of toll road costs. Escalating tariffs
could also lead to the perception of wind-fall profits by concessionaires and

See Appendix 9.8. I. describing tlie Omnibus A y e e i n e n t for Nova Scotia Highway 104 wliicli stipulates
coverage of I .2: I for the initial three years. and I.; tlirreafter.
public backlasi1, I t is I1oteci tl1ougIi that wit11 profit taxation am1 revenue sIlari11g
with Jasa M a p , part of these gains can be CaptUWd back.

Regional c p ~ ’ s\ . ~ ~ ~ U I reflect
CI inore accurately labour costs, than \.2wuld tlw
higher, Jakarta-clominatec11 iiational CPI; Public works indices used in Iiigliu~ay
contracts woul~lalso be lnorc appropriate for routine and periodic iiiaiiitenancL>
components of toll road operation.

Helice there are several rerisons to consider applying indexing to onl\f ‘I p r t 0 1

the annual cost stream of toll roads. This policy however should be dela!rcil until
the next generation o f BOT toll roads arc tendered. Imposing a discounted C1’1
formula on investors w1i0 have rcsponcleci to the proposed full CPI index
formula and may have proceeded through the signing of Authorization
Agrecnients woulcl open a legal and financial tangle. Applying the CPI
adjustment to the variable portion of costs woulci affect cash flows anticipated b\,
investors. To coinpensate and rcco\w their anticipated returns, opening taritts
would have to be renegotiated with the ol+xtive of making the present vduc o i
the revc‘iiuc stream the same as under the assumption of full CPI indexing.

Raising opening year tariffs would then decrease diversion ancl reduce the
iiurnlwr of beneficiaries and users froni the very beginning. For cxamplc,
discounting the CPI impact on tariffs by 0.9 or 0.8 would mean point reductions
in real financial returns of 0.8% to 1.6%. This woulci mean a demalid from
investors for compensating opening tariff increases by factors of 1.06 and 1.I1
respectively. This woulci reduce opening traffic bv the order of 0.9% and 2% and
economic returns 0.3 to 1%. If a narrower view of variable costs is taken wit11 <i
discount factor of 0.5 applied, the opening tariff demanded by investors woul‘i
increase by a factor of 1.3. I n the case of the Ciawi-Sukabumi project, the
adjusted economic return woulcl drop to below 20% from the base case 24%).
From a marketing ancl development strategy point of view, lower rates from
opening would encourage an e d i e r , larger customer base, as well as gain better
public acceptance. Most critical however in the current generation of projects is
the prospect of further negotiation delays and investor dissatisfactioli. Hence
this proposal for partial indexing should be delayed for later implementation
with a subsequent generation of tenders. The application of indexing 011 k’artial
costs should be examined further together with the question of time values using
consumer stated preference surveys.
 TABLE OF CONTENT

1. 1NTRODUCTION AND OVERVIEW .............................................................................. I

2 REVIEW O F TRAFFIC VOLUME, COMPOSITION Si T O L L REVENUE ............... 2

2.1 Traffic trends by signilicant origiii/destination pairs & vehicle types .......................
2.2 Es-post Coniparisons with li.nsibility study li)rccasts..................................................... 8
2.3 Summary & Conclusions .......... ........................................................... II
3. TOLL TAlllFF AND DIVERSION IMPACTS ............................................................ II
3.1. Current Rate and Past Rate Acijustiiicnt ........................................................... I I
. . . . . ,. ._
3.2 Impact of 1 ariff Increases 011 I rallic ............................................. ......... 18
3.3 Demand Elasticities and Diversion Analysis ...................... ................................ 21
4. KEVIEW OF VOC’S ,TIME SAVINGS : COMI’AKISON WITH TOLL I<AlE
LEVEL .................................................................................................................................... 26
4.1 Resiilts of Field Surveys on Toll niid Alternativc Roads ....
4.2 Review of VOC .......... ..................................................
4.3 Review of Time Saving ............... ................................................................... 31
4.4 Comparison of Tariff Vs. VOC Savings ........................................
4.5 Conclusion on VOC Saving ..................................... .:8
5. CAPITAL, OPERATING AND MAINTENANCE COSTS......................................... 39
5. I Comparison of Investment Costs for Different ~l‘ollRoads ..........................
5.2 Maintenance Costs. ....... ...................... ............
5.3 Comparison of Tariff & Maintenance Cost Responsibility By Vehicle Type:
Cikampek Toll Road Example ... ............................................. ............ .... 34
6. PROJECTED GROSS AND NET IlEVENUES FOR CASE STUDY TOLL IlOAOS46

7. CIAWI-SUKABUMI TOLL ROAD I’KOJECT CASE STUDY .................................. 48

7.1 Background .................................................................................................................... 48
7.2 Economic Analysis of Prqject As A Non-Toll Road ............
7.3 Impact Upon Econoiiiic Returns of BOT Toll Road lmplernentati
7.4 Financial Evaluation of the Ciawi-Sukabumi Toll Road Project .................................. 55
7.5 Foreign Exchange Impacts ....................................................... 58
7.6. Impact of Different Toll Tariff Policies On Financial Returns .................................... 63
8. REVIEW OF RISK ALLOCATION AND MITIGATION FRAMEWORK ..............64
8.1 Background: Present Concession Framework: ................................................
8.2 Current Framework for Allocation of Risk ......... ......
8.3 Regulatory Context ........................................................................ 74
9. TOLL TARIFF SETTING AND AD.IUSTMENT MECHANISMS ............................ 75
9.1 Background ..... ......................................................................
9.2 Current System. ..................................................................................... 76
9.3 Legal Framework and Philosophy. ...... ......................................... 76
9.4 BOT Toll Roads and Joint Ventures ....................................................... ......... 7 1
9.5 User Savings and Tariff Setting ................................................................... 7 s
9.6 Tarilf' Policy Precedents and Ad.justmcnt Mechanisms ................................................ SO
9.7 Jasa Marga Proposals F o r A Tarill' Acljiisliiiciit I:oriiiiila .......... ...................s I
9.8 Conclusions and Ohserwtioiis 011 '1'011 'I'ari11'A~i~~tistmeiit For11
9.9 Alternative Institutional Arrangeiiicnts For Processing 'l'oll Tariff I'ropos~iIs i\nd
Tariff Adj ustiiiriits ................... ..........................
9.10 Summary ................................... ...... .........
10. CONCLUSIONS AND IWCOMMENDATIONS FOR IUSK MITICATION AND
TARIFF ADJUSTMENT ...................................................................................................... XX

APPENDICES:

Appendix 1 : Law # 13, 7980 on Roads

Appendix 2.1 on Case Study Toll Ro,icIs,
: Annual Trend on Traffic and Revenue
1987- 1995
Appendix 4.1 : Percentage o f Toll Tariff to VOC Savings under Variow V O T
Assumptions
Appendix 4.2 : PCI VOC Moclel
Appendix 4.3 : Newly Revised VOC Moclel
Appendix 7.2 : Ciawi - Sukabuini Econoinic Analysis Table
Appendix 7.4 : Ciawi - Sukabuini Financial Analysis Table
Appendix 9.1 : International a nd Experience in Toll Tariff Setting and Adjustmcnt
Appendix 9.8.1 : Nova Scotia, Highway 104, Toll Structure
1. INTRODUCTION AND OVERVIEW

The Asian Development Bank commissioned this Highway Toll stud\. i n inid-
March, 1997. The Executing Agency for the stuciv has been the Directorate
General of Highways which along with PT Jasa Marga (Persero) has kwo\rictcci
overall guidance to the consultants in their work. Jasa Marga pro\rideci extcnsivc
office ancl logistical support. Through the liaison committee chaired b v the
Research Bureau with active participation bv Investnwnt De\dopment Di\-ision,
as well as Iiwcstment Analysis, Toll Collection, a n d L c p l Ditrisions, the st'lft of
Jasa Marga have been a vital s ~ u r c eof inforination and polic\r insight.

The objective of the study has been to assist the Government and Jasa Marga i n
reviewing the current toll setting forinula a n d to recommcnci a revised toll
setting methodology. This included development of a toll tariff adjustment
formula to ensure full cost recovery a i d transpirencv, anci prwiclc adcquatc
incentives to attract private sector investment.

The Government of Inclonesia anci Jasa Marga have Iiad alinost 20 \'ears of
experience with toll roads, starting with the opening of the Jagorawi Toll Road in
1978. In that time, uncler the organization of Jasa Marga, seven toll roads '1id
two bridges totaling 324.2km have been constructed a n d operated. Since .IWi,
two tvpes of partnerships with outsicle coinyanies through both statc entcrpisc~s
and private consortium have been encouraged Widening avenues of
participation enables a larger toll road prograinme to be undertaken and
inobilizes private domestic ancl foreign capital for investment in the highcr
volume links of the road system.

These public/private partnerships have been organized as "Build, Transfer,

Operate" (BTO) and " Build, Operate, Transfer" (BOT) types of structure. As ot
January, 1997,4 sections of toll road and the Tallo Lama Bridge (Ujung Panclang)
amounting to 140.3kin had been coiistructed and were being operated under BOT
arrangements. Another 10 sections amounting to 193kin are currently under
construction, two of these, Palimanan- Cirebon and Seinarang Seksi C organizcd
under modified turn-key (BTO) arrangement.

By the end of May, 1997, 31 projects were under consideration or starting

impleinentation. Jasa Marga had signed Authorization Agreements Ivitli 6
operators, and was in the process of negotiating agreements with 7 inore
potential partners. The other 17 under investinent process had been selected but
had not proceeded to the next stage of negotiation.

These projects represent a wide range of toll road projects, from urban to
interurban toll facilities. The urban network links include four proposals in the
Jakarta urban area and would amount to 60.4kin. I i ~ ~ l ~ in ~ lproposals
ed are
projects to complete links in the metropolitan ring road and connector s\rstem a:,
 well as adcl a combination road-rail transit corridor in the north-south axis o f the
city.

The interurban proposals amount to 905km of new toll road, of wliich 246.4 are in
West Java, ??6’.9kiii in Central Java, and 2S0.3km in East Java. A few projects
invited in the May, 1995 list havc bccn postponcct because of low traffic
expectations and lack o f bidder interest. These incluctcd the Medan-Binjai stretch
and Labuan-Cilegon-Bojonagara sections. Four previousl\, offered sections in tlw
Solo-Mojokerto corridor were also scaled back troiii 700k111 to 3S kin., anci the
Aloha-Waru-Tanjung Pcrak proposal has been reduced to 14.7km instead of tlw
originally planned 31.6km.

2 REVIEW OF TRAFFIC VOLUME, COMPOSITION &TOLL REVENUE

The case studies represent a cross section of toll roads in Java selected h\,
consultation with Jasa Marga in the carlv stage of the study. These comprise t o u r
jasa Marga ancl four private joint \witurc tinancect toll roads, covering urban ‘ l i d
interurban toll roads with mature a n d / o r dc\dopiiig traffic levels, a s s l i o w i i
below.
I’rivatr Joint Vcnturc

National statistics tell that traffic has grown 3 to 5 times in tlie last two decades.
TIie annual growth of vehicle ownership is in tlie range of 5 - 8 % for the average
province, ancl is higher in big cities. Car ownership levels are still at a very initial
stage at about 18 per 1000 persons nationally but this statistic is rather skewed
towards more developed areas. For example in 1995 DKI Jakarta’sfigure was 150
per 1000 persons, Java ,21, and Sumatera was 12 per 1000, all of which are still far
away from tlie international saturated level. With tlie plienoineiia of the national
car in recent years ancl trends towards cheaper car prices in tlie future, the
growth of car ownership is expected to be higher in tlie near future.

With regards to toll roads PT. Jasa Marga through tlie Division of Toll Collection
records the amount of traffic exiting and entering toll road sections as well as the
revenue over time. Questions on average daily traffic, monthly and/or yearl\,
variations of traffic am1 of re\renue can be exainined tliorouglily for any purpose.
For examples from monthly report of Jakarta-Cikainpek toll road , over tlie
period of 1991 to 1996 overall link average dailir traffic (ADT) increased by 32%)
annuallv for Class I, while Class HA increased a t 20% ancl, for Class 118 126"6.
Here, tlie 11B category represents mostly freight carriers which grew dramaticall\.
on certain links of tlie toll road, in this case towards tlic Cikampek end where a
d r y port and industrial estates arc located.

For all \diicle classes the Pd. Gccic Timur-Bekasi Barat link which is close* to
Jakarta is found to be the busiest amongst the nine links available on the toll
road. This can be explained ti\, the land use intensitv along that section, niostl\r
residential estates, which generates ancl attracts iiiorc traffic compared to other
sections.

Moreover, over tlie period of 1991 to 1996 o \ w a l l link traffic composition was in
tlie range of 62% to 93% for Class I, 5% to 2X?;3 for Class I I A and 1%to ~ I O ? ; ~tor
Class IIB. Tlie share of Class IIA and IIB is getting larger on the links approaching
Cikaiiipck area, indicating h i i d use changes arc still uncicr wav in tlie region.

The trends for tlie remaining case studies arc summarized in Table 2.1. It can be
seen that tlie growth of the link flow is iiiorc stable on mature toll r o d s ,
espc%iallv for Class I traffic. Tlie traffic growth range stabilitv f o r Class I I I3
vehicles is less, ancl tends to grow very rapidlv when the toll road is located iic'~r
an industrial estate or dry port, for examples in this case, are Jakarta-Cikampck
and Taiigerang Barat-Merak toll roads.

The average link traffic coinpositions arc within tlic range 80% to 88% for Class I ,
9%-14'%for Class IIA and, 2% to 1 7 % for Class IIB for all case studies.

Traffic trends ancl re\wiues can be examined b\r tlie iiuiiiber of aiiiiu'>l
transactions counted at tlie exit gates. Figures 2.7 ad 2.2 depict tlie aiiiiual traffic
trend ancl tlie corresponding revenues on all case studies toll roads. The growth
statistics are also summarized in Table 2.2, at least up to 1995, as 1996 statistics is
being audited when the study is carried out. For more detail tlie raw data on this
is attached in Appendix 2.1.

Over tlie period of 1989-1995, 1990-1995 ancl 1991-1995, amongst tlie interurban
case studies toll roads, Jakarta-Cikampek traffic grow, respectively, at 44%, 26%
aiid 22.5%. The other interurban case study toll road wliicli has growth rate
closer to Jakarta - Cikainpek is Dupak-Kebomas. These two toll roads were
identified as having mature traffic in tlie case studies. Others with developing
traffic show' a lower traffic growth over tlie period investigated. For instance,
Padalarang-Cileunyi and Tangeraiig Barat- Merak interurban toll roads grew at
18%and 6%, respectively.

Tlie urban toll roads of Cawang -Tomang and Cawang -Tg. Priok grew at 16%
and 23% over tlie period of 1991-1995. The other urban toll road under case
studies that has historical traffic data is Jatingaleli-Krapyak, while Pd. Piiiang -
Kp. Rambutan has just been operatcd only since 1996. On the Jatiiigalch -
Krapyak toll road traffic grew at 17%O\W the period of 1989-1995.

80

70

60

-
j,, 50
IT
W
'
C
40
.0
-_
.c- 30 ...x
20

10

1987 1988 1989 1990 1991 1992 1993 1994 1995

Figurc 2.1 Annual Traffic Trends

160
140
120
=
[L
100
80
-
._
0
._
-
5% 60
40
20
0

- 1987

Caw ang-Tomng
...o.. . Jakarta-Clkanpek
Dupak-Keboms
1988 1989

+Caw
1990 1991

ang-Tg. Riok
... x . . . Padalarang-Cileunyi
__ Tangerang Earat-Merak
1992

+
1993 1994

Jatingaleh - Krapyak
+W.Rnang - Kp.Rambulan
1995

I
Figure 2.2 Annual Revenue

Higliiiiq Toll Stud!, - A D 5 1997 4

 Table 2.1 Annual Link ADT Growth and Composition

Toll Road Period ADT Growth(%) Traffic Comoosition (%I

Class I 1IA llB Class I IIA IIB
n.a. n.a. n.a. n.a.
Cawang-Tg. Prick n.a. n.a. n.a. n.a.
Jatingaleh - Krapyakb) 1989-1996 29 4 1 la) 60a)
Jakarta-Cikampek 1991-1995 32,(22-42) 20,(11-56) 126,(44-418) 80,(62-93) 15,(5-28) 5,(1-10)
Padalarang -Cileunyi 1992-1996 25,(17-35) 24.(17-30) 68,(42-84) 88,(85-92) 10,(7-13) 2,(1-3)
Pd.Pinang - Kp.Rarnbutan n.a. n.a. n.a. n.a.
Dupak-Kebomas 1995-1996 43,(25-62) 24,(6-35) 15,(9-25) 80,(73-83) 9,(6-12) 11,(9-14)
1992-1996 42,(17-108) 26,(7.88) 90,(38,286) 80,(55-88) 14,(8-33) 6,(2-11)
lote: i'igures in the fable represent: mean, (range)
a) Period of 1992-1996
b, One section only

Table 2.2 Annual Traffic (Transaction) & Revenue

Cawang-Tomang 1988-1995 45,865.009 78,890.838 27 67

Cawang-Tomang 1991-1995 57,365,934 11 1,682,342 16 25
Cawang-Tg. Priok 1991-1995 18,876,070 37,440,886 23 26
Jatingaleh - Krapyak 1989-1995 2514,444 1,270,583 17 23
Jakarta-Cikarnpek 1988-1995 39,447,437 67,371,662 44,22.5' 72,31'
Padalarang -Cileunyi 1992-1995 11,917,583 15,583,838 18 15
Pd.Pinang - Kp.Rambutan na na n.a. n.a.
Dupak-Kebomas '1994-1995 6,753,013 6,480,428 40 112
1990-1995 12,676,749 36,817,408 6 56
Note * F o r period 1991-1995

5
Trends on annual revenues follow the traffic levels. However any historical trends
on revenues should be carefully considered as they are influenced bv the tariff
increases over time. The elasticity of demand due to tariff changes are the subject of
discussion in Section 3 later. In this section, w e outline the trends on rc\'cnucs,
regardless of the tariff levels. As can be seen from Figure 2.2, revenue incrcmcs
more rapidly over time on all toll roads, though traffic levels on few sections slo\,vl\.
increasccl o r even flattened o f f for soine periods, particularlv for the T'iiigc*i-'iiig
Barat - Merak toll roacl.

Some revenue statistics for the case studies toll roads arc sunimarixed in l-iihlc. 2.2,
showing the average annual re\vnuc in Rp. and its growth rate over tlic p*rioci
investigated. Some contrasts were evidenced amongst the toll roads, a fem, with low
annual traffic growth may have a significant increase on revenue. This excm~ilifies
the significant effect of tariff increases on revenue, such as on Tangcrang B'irat -
Merak.

With regards to urban toll road, tor example, over the period of 199~1to I')%
revenues froin Cawang-Tomang were three times the Cawang - Tg. Priok toll road,
though the two have almost the sanie revenue growth rate, i.e. 25% to 26'itn. -l'he
traffic on Cawang-Tomang is inore saturated than on Cawang-Tg. Priok, as c'in be
seen from the annual traffic growth rate in Table 2.2. Cawang - Tg.Priok is I3OT-
financed, while Cawang-Tomang is Jasa Marga-financed and had been operating
earlier. The two sections h a w now been merged uncicr one operator sincc 1990 \\,it11
75%:25% Joint Venture : PT. Jasa Marga revenue sharing, with a concession ( 1 1 111o1'c
than 30 year.

2.1 Traffic trends bv significant origin/destination pairs & vehicle types

Traffic by origin/clestinatioi.l was analvzeci from annual traffic reports provided ti\.
Jasa Marga. To see the historical trends, however, a fixed gate-to-gate o r
origin/destination setting is needed throughout the year, with preferabl\, a long
distance toll road in order to look at the vehicle types distribution. Also, this can
only be clone for a toll road that was built in one stage, with no further extensicin. In
this case Jakarta-Cikampek is a good example for illustration, ancl serves quite long
distance trips. Another long distance toll road such as Tangerang Barat-Mcrak
unfortunately has been gradual extended since 1988, ancl finished in 1996 u p to the
Merak toll gate. Therefore it was not taken as an example.

Figures 2.3 show trends of traffic bv destination on Jakarta-Cikampek toll r o a d For

Jakarta-Cikampek direction traffic from Pd.Gede Tiinur (PGT) to Cikarang grew
more rapidly than other 0 - D for all over the period at 1991-1995. Class 1 traffic grew
at 102%, class IIA at 15% and 135% for class IIB. Annual average growth re

6
39.84% for class 1, 6.42% for class HA and 60.42%for class IIB. Annual gro\\.th of
ffic for every 0 - D are summarised in Table 2.3.
Jakarta-Cikampek Direction

14000 .Class I. 1991 OClass I, 1995

12000 OClass IIA. 1991 DClass IIA. 1995

OClass IIB. 1991 OClass llB. 1995
10000
-i
f 8000
-
I

2 6000

4000

2000

Cika m pek J a ka rta Direction

5000 .Class I, 1991 OClass I, 1995

OClass IIA. 1991 OClass IIA. 1995
4000
OClass 116. 1991 OClass llB. 1995
-$ 3000
>
k
-
0
a 2000

1000

0
-
L P 1
1 3 Y
0 "
20,
m c
a
x m
2:
or

$$
g;
E 'L
m
-5
m .=
E
10
n ::
* 1 _mPF
:a
:5 .E
5
19 u Ou au"
om" Origin-Destination a

Figure 2.3 Trends of Traffic by Destination on Jakarta-CikampekToll Road

7
 Table 2.3 Annual Daily Traffic Growth

lakarta-Cikamoek Direction 1991-1995

I
From-To Aiiiiual Growth (%)
Class I * Class 1IA Class IIB
PGT-Bekasi Barat 37.60 -5.26 12.21
PGT-Bckasi Timur 4'1.35 4.64 10.71
PGT-Cibi tung 21 .99 4.26 11.1.:
PGT-Ci ka ra IIC 102.63 14.77 135.23
~ ~ ~ ~~~

PGT-Karawang Barat 26.51 11.93 87.50

PGT-Karawang Timur 14.40 1.34 90.63
PPGT-Cikampck 26.42 13.29 73.00
Averace 39.84 6-42 61) ;-I

From Cikampek to Jakarta on the other hand, traffic class I and IIA grew rapicilv on
Cikampek-Karawang Barat 0-D,while class IIB grew on Cikampek-Karawang
Timur. From Table 2.3, it can be sccn that class IIA (buses) may have a declining
growth on some 0 - D routes. This is perhaps because most buses changed their
normal routes to alternative routes.

2.2 Ex-post Comparisons with feasibility study forecasts

In many previous toll road feasibility studies, it was assumed that future traffic can
be estimated from the GDP growth per capita and the car ownership levels. For
simple interurban road networks, where at most two alternative road are available,
iiicluding the toll road, an empirical PCI style diversion curve approach was

8
frequentlv used in past feasibilitv studies', from which future traffic levels were
estimated. More discussion on diversion a n d / ~ rroad assignment studies will be
described in Section 3 below. Past traffic forecasts sometimes significantly de\.iate
from tlie actual figures, and for illustration it is interesting to compare the tr'itfic
forecasts niacie prior to construction with the actual traffic obser\vci during the
operation o f the toll roads. Three toll roads arc taken as examples for this purposes:
Jagorawi, Jakarta-Cikampek and Jakarta-Tangerang, all of which are interurban toll
roads.

Figure 2.5 depicts forecasts ancl actual traffic \zolume on the three toll roads. C 3 w
the period observed, traffic volume on Jagorawi toll road grew bv 12% annucillv,
28% on Jakarta-Cikampek and 23% on Jakarta-Tangerang. Tlie forecasts niii\' have
had good estimates for the initial periods, such as on Jagorawi, but deviatc ‘inti
uncierestiniate the traffic 5 to 6 years after. On Jakarta-Cikanipek, tlie initial period
traffic w a s overestiniatecl significantl\r, although later the actual growth was higher
than the forecast. The sanie thing happens on Jakarta-Merak, where forecastwas
quite match with the actual traffic in the initial w a r , and later tends to under
estimate tlie actual traffic.

Tlie cliscrepancv between forecast ancl actual traffic is caused b\r a number of factors.
This is a matter of the accuracy of tlie transport model, the quality of data and tlie
forecasting techniques employed in tlie studies to capture the future trends of
socioeconomic variables explaining toll road traffic behavior. Tlie fact that tlic actual
traffic tends to grow along tlie expected levels i n the beginning ancl deviates sonic
years later above the forecast levels would be a good subject for an ex-post
evaluation. This kind of study is generally lacking in any part of the globe. Land-use
control issues perhaps are a very important aspect behind this. Practice in Indonesia
has unfortunately been to put too much faith on the inaster plan style for both
regions and cities. Many revisions cannot be avoided soon after master plans are
finalized, since land use controls are difficult to enforce.

The model estiiiiate the proportion of trips using the toll road (P) based on the formulae:
P=K/((l.O+(a*(C/T)"b), where C/T the ratio of toll cost to travel time difference, a & b are
empirically derived from other choice set studies, and K is the upper diversion threshold. Tlie
formulae is also used to study the impact of tariff increases on traffic.
 Average Weighted ADT Average Weighted ADT
Average Weighted ADT
- - L - N N W W b - N " P V I ( D
LnovloLno - N W P O 1 0 0 0 0 0 0
x o
0 o0o o 0
0 o0o o ~ ~ 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0

1985 1983 7 1
G 1986 1984
lgE8 c
w 1989 1985
1987 p-'I
2
n c 1986
1988 m
E
R
1990 X 1987
W
c 1989 3 1988
? 1991
P 1990 0 1989
n
R
X
W
z 1991 1992 3
* 0
m 1991
1992 X
2 1993 1992
K 1993
1993
c 1994 1994
1994
0,
c
 2.3 Summary & Conclusions
Traffic volumes on toll roads and the alternative free road varies depending on the
intensity of the land use surrounding the road corridor. Toll roads located near a
city tend to have a peak h o u r spreading anci congestion phenomenon. I f the toll road
is rather congested tlic situation on tlic alternative tree road is bound to bc morc
congestcci.

For interurban toll roads, the situation now in Indonesia s h o w s that \,arious
inciustries prefer to place their factories, warehouses, housing anci super blocks on
the adjacent land along the toll road. With the availability o f an intcrcliange to exit
a n ~ i / o renter the toll road, the industries gain good accessibilit\, especially for their
distribution system in addition to having inore open space and clicapcr land away
from the city. With no strict land use control, as experienced by manv cic\~eloping
countries, tlie ribbon de\&qxiient along the interurban toll road has resulted in
severe access problems to get into the tit!,, with the link flow getting larger m c l
larger as the city is approached. Thc c a w of Jakarta-Cikaiiipck is a good cx'iniplc.
Increased commuting on the Jagorawi toll road leading to saturation in tie'ih h o u r s
is another example, not foreseen at the planning stage to occur so early.

The foregoing analysis tells how uncertaintv may affect a traffic forecast. With the
current practice on toll road investment, the risk on traffic forecast rests o11 the
investor side. In order to maintain revenue targets over tlie time horizon, in\x>stors
have to have a n accurate estimate of thc level of traffic. Since traffic fluctuates as the
road network anci land usc iicvelopment changes, i n the future at least within the
. current growing physical planning anci weak land use control practice, it is not cC1sy
to provide a good estimate of traffic, as shown in the above analvsis. More extensive
sensitivity analyses on revenue as traffic cliangcs would still be good practice in the
future, as well as the clevelopment of more robust traffic forecasting techniques.

3. TOLL TARIFF AND DIVERSION IMPACTS

3.1. Current Rate and Past Rate Adiustment

Current toll tariffs in Indonesia range quite widely in terms of charge per km.
Lowest rates are set on interurban toll roads in lower traffic, developing areas.
Table 3.1 summarises toll rate per kni for several interurban-urban toll roads.
Exainples for light vehicle tariffs include Tangerang Barat-Merak (Rp. 153/ kin) and
Semarang (Jatiiigaleh-Krapyak Rp. 59/kili). The next tier of interurban toll roads are
more developed and bear a n intermediate rate on longer distance or tl-rough traffic
ranging between RP. 86/km for tlie Paclaleunyi Toll Road, Rp. W/km on the
Cikampek, and RP. 8O/ kni on the Jagorawi. These longer distance exainples are
drawn from Toll facilities organized by Jasa Marga. Shorter trips, particularli, in
urban areas tend to cost more. On the Cikampek toll road, the 6 kin. trip from Pci.
Gecie Tiiiiur to Bekasi Berat costs R p . 25O/km.; a 16 km trip out to Cibitung w o u l d
drop to Rp. 156.3/km. For different class of vehicles the rule of thumb is tariff for
Class IIA is 1.5 times Class 1's tariff, and twice for Class IIB.

Privately organized toll facilities tend to h a w higher tariffs. Examples include the
Tangerang Barat-Merak toll road with tariffs for longer distance trips on tlw pri\,atc
section bevond Balajara Barat to Ciujung (20.6 km) and to Mcrak (59.6 km) cli,irgcd
at Rp.l-15.6 and l42.6/km rcspectivcl\r. On the Jasa Marga portion of the ro'id trom
Kebuii jeruk to Tangerang Barat, tariffs arc lower for the 23.2 kiii jouriicv, .lOi.S/ k m
for light vehicles. Similar contrasts can be found in Surabaya where the Jasa Marga
stretch between Waru and Sidoarjo of 11 km is cliargcci at 90.l/km. For the similar
length Tandes-Kcbomas 12 km stretch under private organization, the tariff is I<P.
248/km. Part of this facility is operated as an open svsteni and tariffs arc' not
directly comparable. For instancc to travel the -1.3 km from Tandcs to D u p k costs
Rp. 116.3/kni.

Urban toll roads in congested areas tend as should be cxpcctcci to charge liighcr
tariffs. On the Jakarta inner ring road svstcm, assuming a trip length of 12 km, taritfs
for this open system arc R p . 250/km for light vehicles. On the Dr. Ir. SeJi\wtmo 'id
Cengkareng facilities with tariffs of R p . 4000 and 2000, tolls could range between r\p.
250 to 500/km clepencling on trip length. The highest cost sections also inclucic. the
southern arc of the Outer Ring Road which is privately operated. An 8.8 km trik7
between Poiiciok Pinaiig to Lenten;: Agung costs R p . 284/klli. For the entire 1-1,s
km's open since late 1996, tiic tariff f r o m Ponciok Pinang to Kp. Raml>ut,ln is
currently Rp. 270/klii., but for 1.8 kiii between Fatiiiawati and Pondok Pin,lng, the
charge for light vehicles is Rp. 1000 or Rp. 556/kiii.

1 US$ = R p 2,400

12
3.1.1 Toll Tariffs and Inflation
For several of tlie toll roads organized and operated by Jasa Marga there Iias been no
acljustment of tariffs for sonic years. The last tariff increase on the Jagorawi,
Cikampek, Jakarta-Tangerang Barat, Scniarang, and Pacialeunyi toll roads was in
1Y92. Consumer price index increase in average 8.3 in the last 5 years. O \ w the
years the CPI increase plotted in Figure 3.1. With inflation as measurccl 131. the
Consumer Price Index, the real \ ~ a l u eo f thc tariffs lias fallen to 63% of its 1992 \.'llue
in constant terms.

Increase in toll rate (Rp/km) vary from place to $~cc. Jasa Mar31 financed toll r o d s
seem to have a rather stable rate in last 4 \uirs, while Joint Venture's rate increase
quite significantly in tlie last two years especially on Tomang-CaM,ang-Tg. Piok, 'ind
Pcl. Pinang-Kp. Ranibutan, see Figures 3.2 and 3.3,

1980 1981 1982 1983 1884 1985 1986 1987 1988 1989 1890 1991 1992 1893 1994 1995 1996

YEAR

Figure 3.1 Increase of CPI

13
 -
Toll Ratelkm Jasa Marga

12c
DlalmgaCh-Krapyak

1oc

8c

E
5 GO
z

40

119-30 1991 1992

Yea.
,993 1994 1995 19% ,997

Figure 3.2 Toll Ratefim - Jasa Marga

Toll Ralelkm .Joint Venture

300
0 Tomang-Cawang-Tg Priok
OPd Pioang-Kp Rambulan

1990 1991 1992 1995 19% 199i

Figure 3.3 Toll Ratefim -Joint Venture

14
in the case of tlie Kebun Jeruk- Tangerang Barat section o f Toll road, the avcragc
tariff weighted by vehicle class was Rp. 92/k111 i n 1988. Using this as a base \'ear,
tlie 1996 tariff would be equivalent to Rp. 39.5/kni in 1988 constant ternis, e\-cii after
the average increase in the order of 30% in 1992. Siniilarlv tor the Jakarta- Ci!anipck
toll road, the 1996 tariff in constant 1988 terms aiiiounted to R p . 36.8/km comp,ired
to the 1988 level of RP. 83/km. In 1992, the class-weiglitcci tariff was incrcascd by
one third to Rp. llO.-L/km; by 1996 this would represent RP. 69.8/km in l"? KP.
terms.

This decline of toll tariffs in constant R p . terms coupled with the incrcasc. i n rcal
income per capita explains much of the traffic increase seen on the toll road network.
Corresponcliiigly, the income effect for car users has been iiiorc dramatic, leading t o
high growth rates for this class.

The response o f traffic to toll tariffs is often iiiaskcd by the rate of traffic gro\vth;
economies with lower growth rates are less likcl!, to experience the upward shifts of
travel demand curves as car passenger time values increase in real terms faster than
toll rates increases.

I n developing economies, traffic growth is dominantly a function of eccinomic

growth. Elasticities of traffic with respect to GDP growth are generally high, \\.itli
values in the order o f 1.2 to 2 (Inctoncsia and Mexico) and from 1.5 to 2.2 (Mala\%
ancl Thailand). For that population of customers disposed to take advantage* of
significant time savings, the price sensitivity to tariff increases is relativelv low, with
price elasticities for automobile traffic is in the order of -0.1 to -0.2.

On the other hand, an economic crisis and downturn in activitv can Iia\ac a iiiorc
chilling effect on traffic than toll increases. The 1983 economic downturn in the
Indonesian economy appeared to have as great or greater an impact on traffic in the
Jagorawi corridor as the June 1983 tariff increases of 100 and 150 percent.'

3.1.2 Toll Tariffs: International Comparisons

Table 3.2 shows some international rates. Indonesian toll tariffs tend to be higher in
comparison with international tariffs, with few exceptions, short distance tariffs on
tlie Jagorawi, Jakarta Merak, ancl Cikainpek tend to be -higher than rates in Chile,
Philippines, Thailand, and the U.S. Longer distance tariffs on the Padaleunvi,
Jagorawi, and Cikampek would be comparable to Manila South Toll Road rates as
well as several of the US intercity toll roads. Through traffic, longer distance tariffs
on the Padaleunyi and Belmera are comparable with some rates in Brazil and in the
US., the latter exampled by the New Jersey turnpike. It should be noted that nianv

Technical Report No. 7, Technical Assistance and Training I n Toll Road Management for Pt. Jdsa
Marga (Pel-sero),New Jersey Turnpike Authority & Louis Bergel- International, IIIC., April, 1981

Htgimay Toll S t i d ! , - ADB 1947 15

of tlie U S . toll roads are older, mature facilities which were built or reconstructed
with cheaper money. In several cases these were subsidized with revenue bonds
which enjoyed treatment as tax shelters.

As tor urban toll roads, only the expeiisive Pcrito Moreno and 25 dc Ma\ro in Buc*nos
Aires at $US 0.20/km coinpire with Ccngkarcmg or the some o f the southern
sections of the JORR. Depending upon the trip length on tlic inner ring road,
CaMraiig-Tomang-Tg. Priok, costs/ kiii would be geiierallv comparable with peak
hour rates on tlie SRYl BOT projcct i n Southern California. Off-peak rates on this
new toll road in Orange County, California are halt or less than tariffs on the Jakarta
urban syste111.

16
able 3.2 Toll Tariffs :International Comparisons
Toll rateslkm (US$)
-
Car -
Bus

gentina (urban, 1994) (a)

Perito Moreno 8 25 de Mayo 0 200
azil (interurban, 1993) (b)
Rodovia dos Bandeirantes 0.022
Rodovia dos lmigrantes 0.018
Rodovia dos Trabalhadores 0.044
,ile (interurban. 1992) (b) weekday weekend weekday weekend weekdav weekend
Route 5 0017 0.034 0037 0.044 0.059 0.067

exico (c) 0.051 0.077 (per axle) 0.077 (per axle)

iilippines (d) 0.038

iailand (urban. 15 I (e) 3 to 5 axle vehicles , ixles

.t Stage Expressway, Bangkok 0.014 0.021 0.042

iited States (1991/92) (0 2 8 3 axle single units

New Jersey Turnpike 0.024 0.042 0.068
Pennsylvania Turnpike 0.024
Indiana Toll road 0.018
Delaware Turnpike 0.055
WEst Virginia Turnpike 0.026
East-West tollway, Illinois 0,017
North-South tollway, Illinois 0.036

& m
alifornia (1991) (g) 0.124 0.062
Orange County SR91 BOT project
(congestion pricing scheme)

'estern Europe
tercity (1989) (h) 0.013 0.02 0.02

etherlands: Randstad (Ams1erda.Rotterdam.the Hague,Utrecht)road pricing scheme

Cost per entrylpassage w o f f - p e a k
1.65 0.16
oruay: Stockholm Toll Ring (Dennis
ackage. 1991)
cost per entry. ring road 2.53 n/a 7.59

K: Cambridge, England (PAMELA Project as

'oposed for midPO's implementation
) $0.36/congestion unit, defined in terms of pre-specified
speed, delays, stops, say 3 minuteslkm

17
Table 3.2 (continued)
sources.

(a) Louis Berger International

(b) Louis Berger International
(c) FOA 8 Associados. (1995 @
6N.Pesos6US )
avg rates new 4-lane toll roads, 1987. Banobras study
0.022 0 042 0 05
rates from 1984 to 1988 generally followed the
Consumer Price Index
rates were raised in 1988189. ranging from SO 14
to .0321km for autos.
0 341 for buses and 0 0873 for trucks.
(d) Proposed tariffs. Manila South Tollway BOT Project.
Louis Berger International, 1993
(e) ALM Resources Ltd.
(f) Toll Financing. Dan Cupper. Editor,
Researched by Wilbur Smith Assoc. for
-
lnterational Bridoe. Tunnel. and
Turnpike Association
(g) Timothy Hau. "Congestion Charging Mechanisms for Roads." World Bank Working Paper, WPS 1071 Dec 1992
(h) World Bank Symposium. 1989. Earmarking 8 Road Funds

3.2 Impact of Tariff Increases on Traffic

Toll road user response toward changes o n tariffs levied can be studied at least from
two general approaches. Firstly, route choice decisions macle b!~ usei-s 1 ' s t,iritfs
change can be directly observed using time series or cross-sectional ciat'i ga thcrcct
from before and after a new tariff is implementecl. This gives an estimate of demand
elasticity from changes in tariff levels. In turn, an empirical model can also be
developed if a range of tariff levels can be tested in the toll road system. A ditacrsion
curve is one example, but can be used at most for binary case choice.

Secondly, a behavioral route choice response can also be investigated using a

network analysis tool. This will be more appropriate if the number of alternative
routes are numerous; in this case, the first approach can no longer be used.

The network based models are frequently used for toll road appraisal, not onlv to
give a rougli estimate of link flows and/or costs for economic evaluation, but also to
study tlie likely change of route choice distribution upon introduction of a set of
new tariffs.

The history of tariff increases is well documented. This coupled with the availability
of monthly traffic record provided by Jasa Marga was used as a source of data in
examining tlie impact of tariff increases on traffic. As cleinancl elasticities will be
discussed in the next section, this section concentrates on tariff impacts on iiiontlilv
traffic trends on the toll road.
c

For example, Figure 3.4 depicts the iiiipct o f tariff increase on several toll r o d s .
With tlic introduction of new tariff, traffic tended to decline a few months after and
then tended to catch u p with the previous trend. The same phenomena can be
observed on othcr toll roads when a n e w tariff has been introduced. This can be
explained by observing that travelers perceive the increased tariff as significant in
the beginning, and try to change their routing behavior. However after ‘1 5 to 6
month period, the traffic trend reco\wrs a n d follows the previous trend lietorc~the
n e w tariff was introduccci.

19
 Monthly ADT. Cawang-Tj.Priok TandesKebomas
3500 - 500
3000 .
. . ... :. ... .L.. .. .....:. ....
2 400
I

.c 2500 . . .. .:. . . ...I . _ _.: ... .. (Y

>
e $ 300
23
-
I- ,000
n
a
500
..... :. ... .:.... :..... :.-
.
-1994

d ' 9 9 5
-7998
;

: :
~

:._...1 .....:.. 1ltlC 96

-- 200
,o
0
E
e
L
100

0
.. -
-1995
,996
:
.. I . .. . .. ..'. . ... I, . . No tariff increase

Jatingale h-Kra pya k Ja karta-Cika mpe k

200 3000
1

' -1991

'
r

I- 500 I Tariff increase in July '92

--
+1993
1991
1992

01

Figure 3.6 Impact of tariff increase on traffic (Vehicle Class I)

20
3.3 Demand Elasticities and Diversion Analysis

3.3.1 Demand Elasticities

Given the impact of tariff increases discussed earlier, it is rather tricky to find an
estimate of elasticity. Traffic tends to increase over time, aiid the time reference
within which to choose a before a n d after situation will determine the elasticitv
level.

In later studies Dwijono (1997).’observed both arc elasticities ancl shrinkage ratios
on Jakarta-Cikampek. Arc elasticities and shrinkage ratios werc LletcrmincJ h s e c i
on traffic and toll rate changes from 1988 to 1993. The shrinkage ratio, Es, is ~iefinecl
as the ratio of the proportional change in quantitv anci proportional change in price:

Es=( (@2-Ql)/@l)/((P2-P1)/ P l )

aiicl the arc elasticity , Earc= (log Q2-log Ql)/(log P2- log P1)

Both measures were applied to different periods of analysis, one to four months
after toll increases, ancl to Class 1 ancl 2 traffic. Using average daily traffic flow
(monthly), the arc elasticity results after four months had mean values of -0.43and -
0.40 for class 1 ancl 2 traffic respectively. The shrinkage ratio method indicated
mean values after four months of -0.32 anci -0.26 respectivelv. Interesting shrinkage
ratio results also came from the distance coiiiparisoii. For the same four month
response period, medium distance, 20-40 km trips showed greater elasticity values
than shorter (<2Okm) or loliger (>40klii) trips.

Vehicle Type Distance

< 20 kin 20-40 kin >40 k m
Class I -0.35 -0.24 -0.4
Class I1 -0.29 -0.17 -0.3

Previous studies had indicated that demand for shorter trips is usually more elastic
than for long distance trips. However in the case of the Jakarta - Cikampek toll road,
fewer alternative routes were available for this medium trip length and customers
were more dependent or captive to the toll road. For the West Java interurban toll
roads the range was from -0.32 to 0.37 (+/- 0.11 Tangerang) for Class 1 ancl from -
0.26 (+/-0.14 Cikampek) , -0.35 (Jagorawi) to -0.33 (+/-0.2 Tangerang) for Class 2.
The study also pointed out that the shrinkage ratio elasticity results were in general

Dwijono, D (1997). Toll Elasticities oil the Jakarta-Cikampek Toll Road, paper presented a t IRF
World Meeting, Toronto
agreement with the "Curtin"' general rule of tliumb," that is 0.32 for Cars and 0.26
for Class 2 (trucks).

To find out the elasticities of clemand for the case studies toll roads some exercises
were conducted during tlic stud\, using tiionthlv ADT of the longest 0 - D trips on
each toll road as the quantity, and taking 2 months before the tariff introduced and
4 months after as the time reference. It was found that the elasticity varies bv vehicle
type and by toll road. Table 3.4 summarises the shrinkage ratios and arc elasticities
gathered from the exercise.

Table 3.4 Shrinkage Ratio and Arc Elasticities

VEHICLE CLASS I

' Curtin, J.F. (1968). Effect on Transit Riding, Highway Research Record 213: 8-18

H I ~ I I I TI dV l~Sttidy - AD5 1997 22

 Table 3.4 (continued)

VEHICLE CLASS IIA

I Openinq
. .
I Year
Urban Toll Roads
I
1 Cawang-Tomang

j i
1 I
Kb Jerui-Tangerang Brt
Kb Jeruk-Cw unq

Jatingaleh-Krapyak
, -2.16

1 j Tandes-Keboomas
II

I
Dec.'94

VEHICLE CLASS IIB

I Opening
I
II Shrinkage Ratio" I Arc Elasticitties II
(1 IUrban Toll Roads I
Year
I I
1 Cawang-Tomang July'92 Oct.'95
-2.i2
2 Cawang-Tanjung Priok July'92 Oct.'95
4.91
3 Pd.Pinang-KnRambutan
. . SeDt '95
Intercity Toll Roads

4 Jakarta-Cikampek Sept.'88 July'92

Pd.Gd.Timur-Cikampek 0.32
I I I
5 IJakarta-Merak I July'92 I Dec.'94
I Kb.Jeruk-Tangerang611. I I I

I L L
Note :Class 118 vehicle has been introduced since July '92, excep, Jakarta-Cikampek in May 1991.
Vehicle class IIA is more elastic than class IIB, and class I is the least affected bv tariff
increase.

Tlie foregoing analysis shows that toll road demand elasticity is inelastic, and in
sc)Ine circumstances it could even be liiglily inelastic. Although elasticit\, ma\'
inciicate user response towards tariff increase in the short term, the use of eListicit\.
analysis for a longer terni forecasting is rather limited.

The section on diversion analysis is devoted below to discuss the effect ot taritf
increase on the longer terni demand level.

3.3.2 Diversion Analysis and Traffic Assignment

Diversion of traffic d u e to change in supply attribute such as time o r tariff can he
studied using a ciiversioii curve or road network assignment model. Prior to using
anv o f these models to diversion 'analvsis, it is desireable to carr\' out model
calibration anci validation. Unfortunately, this is rarely done, as data ilia\' sometimes
be very limited and costly to collect.

In attempt to calibrate a diversion curve mociel Vanciebona & Hasanudciin ('IWh)'

analvzed the effects of toll increases anci network connectivity on traffic diversion
and utilization of the Jakarta urban system. Points raised i n this analysis included
tlie effects of congestion on alternative arterials and the relatively high proportion of
users captive to the tollwav because of tlie lack of competitive "free" altcrnati\w,
The 1992 substantial increase in tolls produced onlv a small reduction in the tollwav
share of traffic.

A n attraction factor method was applied of tlie form P=l/(l+a'(Ct/Ca)"b', where P

is the probability of choosing the tollway and Ct and Ca are perceived costs of
selecting tollway and arterial respectively. This type of model was also proposed by
Sweroacl in their 1990 studies. Estimations were better (higher R-squared values)
when operating costs were ignored anci only travel time saved and toll clicirges
taken into account. A high range of R-squared values was encountered showing tlie
difficulty in generalizing the applicability of attraction factor inodels.

A price demand elasticity model of the mixed log-linear form

P=(KTAa)eA(bT), where T is the toll charged and K, a, and b are parameters; these
were determined by Hasanudin in his 1994 paper. Trucks and buses showed a
better fit with this model than the logit and attraction factor models.

j Vandebona, U., Hasanuddin (1996). Traffic diversion models for toll road facilities: case study of

Jakrta.,Proc. of CODATU VII, New Dellii.

This paper further reinforced the view that urban route selection mainly relics on
tlie trade-offs between toll imposed and the possible travel time savings. Subject to
the caveats on estimating cieinanci elasticities, models based on this approach
provided the best Iiiatch with field data for buses and trucks. Also apparent in the
fiiidings was the greater sensitivitv to toll rates for short trips; for longer sections,
level of service factors were most important i n rout<>choices.

Several feasibility studies tor proposed n e w toll roads have used the "PCI" (PCicitic
Consultants International) diversion moctcl de\doped on the basis of 1989 suri'cy~s
in the Jakarta-Surabava corridors. Relving on out-of-pocket user cost and time
savings this model is of the form:
P= K/((1.0+ (a*(C/T)Ab] where P is the proportion of trips using the toll road.
C/T is the ratio of toll cost to t r a \ d time difference between lion-toll and toll road
routing. Travel time difference, T, is expressed in minutes. In the 1990 Bt)gor-
Banclung feasibility stud\, carried out b \ r JICA, a vehicle trip ciistribution iiicidcl \\''is
used which estimatccl attraction/gcneration of trips as a function of population 'ind
eiiiployment by zone, invcrscl\~proportioiial t o the distance between zoIies, m.itli
parameters cletermined b \ ~iiiultiple regression. Future traffic projections w e r c ~then
based on forecast growth in vehicle ownership. Traffic was assigned to links using
an algorithm which searched out the shortest route, calculated time traveled a d
then applied a diversion ratio curve to arrive at vehicle trips between zones.

Parameters, a and bp b\, vehicle t \ y tor the Bogor -Bandung study were cietermiiicd
as follows :

Vehicle type <l b

passenger car 2.855E-5 2.27684
Ligth truck 5.870E-4 1.74818
Heavy truck 6.078E-5 2.34860

K is tlie upper diversion thresholcl. 111 the Bandung study, K was set at unity,
implying a 100% diversion which would not happen in reality. For the Solo-Yogja
study6 K was set at 91%, but the source or logic was not given. Parameters a nncl b
used in the Solo-Yogja toll road study were 2.401 * 10A-5 aiid 2.0874 respectively -
apparently for all vehicle types.

In the 1995 Solo-Yoga study, traffic assignments were made with the TRANPLAN
model incorporating this diversion relationship and the resulting diversion
calculation suggested that a 20 minute time saving would produce a 35% diversion
to tlie toll road with an aggregate toll cost of Rp4000 (1994 RP).

'' Solo-Yogyakarta Toll Road Project, Planning and Feasibility Analysis, March 1995, BIEC
International in association with PT Bitia K ~ y (Persero),
i PT Seecons, and Pacific Consultants
International
For the purposes of estimating future toll levels and perceived costs, the Solo-Yogja
consultants assumed that absolute toll rates would increase by 8%/vear on average
in current terms, or 3.5 to 4.5% i n constant terms. With iiicomes (GPP/capita)
forecast to grow at between -i and 6% / w a r , the consultants assumed that motorist's
perception of out-of-pocket toll pavmcnts would deflate by 0.5 to 1.0%/ w a r . This
assumption seeins to implv that user costs would only be perceived in terins of the
increased value of time savings while fuel o r other operating cost components i n the
user calculus would remain constant. This is perhaps valid for Class 1, priv,itc* cars,
and Class 2, express &I intercit\. buses, b u t coulci be questioned for truck tr'ittic
where vehicle utilization and operating costs arc perceived bv operatcw as
significant.

The heart of diversion analvsis is tlie traffic modelling and forecasting. Using a
diversion curve approach to assess a longer term effect of supply's attribute change
on clemand level is viable for a simple network with only two alternatives. l'lic~oiilv
reservation, if anv, on using the technique is that the bchavioural consistency iiiCivbe
sacrificed, especiallv when an economic benefit is to be compared between 'with ,Ind
without' project. That is to sav the solutions obtained are not necessarily coinparable
with the equilibrium routing principle. When the network becomes inore complex,
tlie network based diversion model is more appropriate.

The level of diversion resulted from all models depends on tlie network (supply)
and ciemancl condition when it is being assessed. The level of tariff is represented in
the inociel through the network attribute, to which the assigninelit iiioclel will find
the solution for tlie routing principle assumed in tlie iiiodel.

4. REVIEW OF VOC'S ,TIME SAVINGS :COMPARISON WITH TOLL

RATE LEVEL

4.1 Results of Field Surveys on Toll and Alternative Roads

Fields surveys were carried on the second week of May 1997 for 10 days for all case
study toll roads. This is primarily conducted in order to obtain more reasonable data
input for VOC savings calculations. Data surveyed included distance, travel time
and speeds. Table 4.1 summarizes the survey results. It is worth noting that the
surveys coincided with tlie campaign period for 1997 general election. It was
necessary to carefully choose an appropriate time period to obtain a representative
sample, e.g. by doing the survey earlier in tlie morning.

26
4.2 Review of VOC

Table 4.1 Field Survey Results

Studies on vehicle operating cost (VOC) for Iiiclonesian conditions had bcen
initiated as part of tlie need to develop a moclel for economic evaluation o f road
niaintenance programmes. Several models have been developed, tor cxamplc,
GENMERRI in tlie 80's and RUCM' in 1992. The RUCM was partially derived from
tlie World Bank HDM-I11 vehicle operating cost model (VOCM), which then
superseded the VOC model of GFNMERRI. The vehicle operating cost model was
developed as part of an effort to establish a n Inter-urban Road Maintenance
Management Systems (IRMS) for local use in Indonesia.

I n general, vehicle operating cost comprises fuel consumption, lubricants, t\,res,

maintenance labour ancl parts, depreciation ancl interest components. I n estiniating
these costs, tlie above mentioned VOC models focus primarily upon road surface
irregularity (roughness) as tlie determinant factor of cost changes.

Hoff & Overgaard (1992). Road user cost model. Second technical advisory services on planning
and programming to Directorate General of Highways, Government of Indonesia

Hi,ylwm/ Toll Sfmhy - A D R 199; 27

The specifications for VOC models varv according to the purpose of model
development. The use of tlie model decides tlie levels of detail of mociel
specification. This mav range from very simple to medium and complicated models,
taking into account various variables that may explain cost changes. The
specification of VOC models for maintenance programming purposes falls into a
group o f complicated mociels, in which much data is required too be satisfxtoril\f
applied in the screening level analysis.

According to the legal basis for toll tariff setting, the section of Law no. 1.3, I‘JSOS
regarding toll roads, the toll rate plus tlic vehicle operating cost on tlic toll r o ~ d
shall not be higher than tlie VOC which users expericnce on the alternati\,c tree
road. If VOC savings which users gain froni trawling o n a toll road are used tor a
guide i n tlie cieterniination of toll rate, it is desireable to cievelop a quite siiiiplc. VOC
mociel such that the data collection and further updating of tlie toll rates can be done
without too much time and resources. However tlic model should be cxtcnsivel\~
calibrated based on field data, which for VOC means not ever\’ cost component can
be fully calibrated quicklv. The most obvious component is fuel, given user
perceptions and relative consumption rates. Other parameters may take more time
to observe changes and cost differences.

To study VOC on toll roads, PT. Jasa Marga, has been using a model c h d o p c c f in
1979 by Pacific Consultants International for a guide in setting toll rates. The model
was cievelopeci as part of a feasibility studv carried out for intra-urban toll r o < ~ ~inl s
Jakarta. The detail procedure is attached i n Appendix 4.2. I n the mociel each VOC
component is assumed to be best explained by the average running speed of the
veliicles. Irregularity of road surfaces, geometrics, and traffic conditions are not
explicitly taken into account i n factoring cost changes.

Currently, Jasa Marga has commissioned a consultant to review tlie PCI model and
now a modified model of VOC for toll roads is under review, tlie document of which
liass been released’. Tlx fuel consumption cost component is empirically de\:elopeci
using representative veliicles commonly using the toll roads by taking into account
traffic, road geometrics, and roughness effects. Other cost components are partially
adopted from an extensive review made on existing VOC models. The detail
procedure is attached in Appendix 4.3.

To illustrate the difference between the models Figures 4.1 and 4.2 depict the
amount of each vehicle operating cost component on two inter-citv toll roads

8The VOC components stated in the law also include the time value, apart from other conlnlon V O C
compoiien ts.
IT.Jasa Marga (1996). The calculation of vehicle operating cost. Final Report, Foundation for
Research & Industrial Affiliation - ITB. (in Bahasa Indonesia)
calculated based on the PCI and the newly revised moclel. Also shown in thc tigurcs
are the time values based on Jasa Marga's VOT. The corresponding VOC's for the
free (non-toll) roads are also shown in the figures, based 011speed obtained from thc
field surveys.

I
30 000

25 oon

20 000

-
-
n
E
VI
0
15000

0
t a ow

5 000

Figure 4.1 Vehicle Operating Cost ( Jakarta-Cikampek Inter-city toll road)

25 000

20 000

-
n
15000

+
In
0
v moo0

5 000

Figure 4.2 Vehicle Operating Cost ( Ciawi-Sukabumi)

The basic differences between the two models lie in the fuel, lubricant (oil) and
interest and time value components; other components remain the same as in
previous PCI model. Furthermore, in the new model interest cost was assumed to be
the same, whether or not users choose to travel via the toll road.

29
User savings are gained from the difference between the toll road cost compwients
and those of the non-toll road costs. Fuel savings from the two models were not
significantly different. The difference in interest cost savings is significant between
tlie two models, since interest cost was assumed to be the same for toll ancl iioii-t~dl
conditions in the newly developed model.

Lubricant consumption was found to be greater on the toll r o a d in the IICCV niodc4,
given the selected toll road sample and speed Ic\ds. Tlic model for lubric'int i n tlie
new iiioclel was adapted from the GENMERRI iiiociel which contains a roughness
correction.

Figure -4.3 ancl 4.4 coinpare tlie contribution of each saving component to the total
VOC saving including time betwccn the PCI ancl newly developed VOC iiiodc-1 on
Jakarta-Cikampek ancl Ciawi-Sukabumi toll roads tor vehicle Class 1. llic c'ilcuLition
is based on tlie reccnt Jasa Marga value of time.

The difference between the two VOC iiioclel lies in tlie interest component. 'The
newly developed model assumes that travellers will experience a marginal interest
saving, whether or not they travel via tlie toll road, whereas in the PCI tliere exists a
significant interest saving if travellers choose to travel via a toll road. Using Tlic I'CI
VOC model tlie amount of interest saving is 3 times the fuel saving for Jaharta-
Cikaiiipck toll road, am1 2.6 tiiiies for Ciawi-Sukabuiiii.

80

70

60

- 50

U 40
0)
c
OJ
5 30
2
0
a 20

10

0 m
FmI 7
-10
VOC component

Figure 4.3 Contribution of Each Saving Component to the Total VOC Saving on
Jakarta-Cikampek Toll Road

30
 DPCl
ONewModd

VOC component

Figure 4.4 Contribution of Each Saving Component to the Total VOC Saving on
Ciawi-Sukabumi

4.3 Review of Time Saving

With regards to value of tiiuc (vot), currcntl\r PT. Jasa Marga uses Rp. 12,2X7/hour
for class I, Rp. 18,534 /hour f o r class IIA and R p . 13,768 for class IIB, regardless of
the location of toll road. Tlic values wcrc derived bv assuming vehicle utilization as
the source of time saving, not the occupant's. This is in accordance with the original
PCI VOC model. I n past toll road feasibility studies, Iiowcver, various value of tinics
were assumed and derived either from an income approach or empirical traffic
diversion forinulae.

Table 4.1 summarises several values of time assumed in past toll road studies for
business travel (working time).

Table 4.1 Value of Times

3. 1
For example an income approach is used for Thc Northern Extension ot South-\Vest
Arc of Jakarta Interurban Toll Road by PCI (1Y89), while JICA(lY91) in Surab'iya-
Mojokerto used a diversion curve based calibration to have a proxy of value o f time.

The use of an equilibrium assignment model to derive the value of tinie in congested
Bangkok toll road network w a s demonstrated bv Motoinura & Laoha-unya"'. The
values of time derived from this model arc consicicrablv higher than those derived
from income data. Using a cli\usion curve to derive the value of time can '1s ~ ~ bc~ 1 1
made as JlCA did in the Suraba\,'i-Mojokerto studv. Such a diversion curve ~ " 1 1 1be
constructed from a stochastic based assignment with assumed driver perception of
cost of travel. Howe\wr, for a congested ncttvork it is better to model driver
behavior based on the cqu il ibri uiii assu nip ti on.

Many costs and benefits of transport projects, tor example, time savings, pain and
grief from accidents, anci environmental effects do not have a inarkct price. In this
case, a variety of approaches have been used to attenipt to establish what those
affected would be willing to pa\' for the benefits gained.

Approaches for value of time saving could be based from resource o r behavioural
values. The income method falls into the resource value approach, while other
transport choice behaviour mcthoils fall into the behavioural value approach.
Beliavioural approaches attempt to establish an explanatory function that includes
travel time anci costs as variables of trip making behaviour.

A preferred approach has been to try t o discover what travellers are willing to pav
to save time, either Lx 'revealed preference' o r by 'stated preference' methods1 1.
Unfortunately, there has not been a n v empirical value of time derived for Indonesia
conditions.

For the present study, various value of time are used in order to see how different
VOT assumption affect the V O C savings. Toll rates are then compared to the VOC
savings to indicate tariff as percentage of VOC saving (including time).

Figure 4.5 depicts the sensitivity of VOC saving towards time value assumption. A
range of value of time from Rp. 2,000 to Rp. 16,000 was tested, and compared with
the total VOC saving. It can be seen time savings over the VOT assumption account
for 20% to 70% in the PCI model and 36% to 88%in the new model.

"' Motoinura, Y., Laoha-unya, S. , Derivation of value of time and traffic demand curves in Bangkok,
Transportation Research Record 7395.
'I For example using stated preference in 1988 prices Dept. of Transport in UK introduced resource

values of time of 705.3 pence/hour for working time and 207.5 pence/hour for non-working tiine.
(100 pence = Rp. 4000,-)
 30000

25000

-d 20000

E
F 15000
._
>
:10000 r

5000

0
2000 4000 6000 8000 10000 12000 14000 16000
Value of Time (Rp.lhourlveh.)

Ja karta-Cikam pe k
30000

I
25000

20000
n n
a:dp 15000
._>
:10000
5000

0
2000 4000 6000 8000 10000 12000 14000 16000
Value of Time (Rp.lhour1veh.J

Cia w i-Su ka bum i

30000

25000

-d 20000

E
F 15000
._
>

L ktll
:10000
5000

0 2000 4000 6000

8000 10000 12000 14000 16000
Value of Time (Rp.lhourlveh.)

Figure 4.5 Contribution of the Time Saving to Total VOC Saving

7bii Sfi(di/ - ADB 1Y91

HI$I~IWI/ 33
Using UK’s value of time and income as a base line, a proxy of value of tiinc tor
Incionesia can be exercised as follow.

UK : (a) VOT (business) : 705 (peiice/hr/person)

(b) VOT per veli. Class 1 (2.5 occupaiic\r) : 70,530 (Rp/ hr/ veh)
(c) 1llcome/year/ person : USS;17,600
(assuming poundstr. 11,00O/vc,ir)

Indonesia :
(d) IIlcome/~ear/person : US$ 1200
= 0.0682 (c)
= 0.07 (1.)

( e )VOT per \re11 Class I . . . . .

Jasa Marga : Rp.12,287 = 0.1742 (b)
IHCM : R p . 3281 = 0.0465 (b)
Feasibility Studies
(JlCA, PCI) : R p . 7076 = 0.1003 (b)
Proxv VOT : R p . 5000 = 0.07 (b)

It can be verified that recent Jasa Marga practice of using VOT equals R p . 12,2S7/hr
for vehicle Class I appears to be rather high. In thc light of value of tiiiic studies, a
proxy VOT would be in the range of R p . 5000 to Rp. 7000/hr for vehicle Class I.
4.4 Comparison of Tariff Vs. VOC Savings
Using current basic price o f automotive components as input for the calculation,
Table 4.2 summarizes tariff as percentage of VOC savings uilcler various value of
t h e . Includecl in tlie table is Ciawi-Sukabumi. This required an initial tariff
assumption, so that the calculation of percentage of tariff can be made. Dctailcd
calculation are attached in Appendix 4.1.

Figure 4.6 and 4.7 depict tariff as pcrccntagc of V O C savings, using rcspcctivcl\.
recent Jasa Marga's VOT assumption (Rp.12,287/hr) and a moderate value. ot timc
of Rp. 7,076/hr.

Tariff for urban toll roads are higher than that of interurban toll road. I n all case
studies toll roads tariff are well below 70% VOC saving under Jasa Marga's VOT
assumption. Using VOT equals Rp. 7,076/ 111-most urban toll roads have tariff aho\~c
tlie 70% saving ceilings.

Tariff as %VOC Saving

140

s
0F U Model - New Model .. .... . 70%

67.92 70.58 69.48

70 . . _ . _ _ _ _ _.._
. .. ... . .

.I '5 p
ir
Location

Figure 4.6 Tariff as '%I VOC Saving Using JM's VOT Value (Rp.12,287/hr)

35
 140

96.71
-
100.17
Tariff as %VOC Saving

PCI Nbdel -New

105.99
Nbdel ....... 70%

98.09
8

........... ...........

Location

Figure 4.7 Tariff as '%IVOC

Saving Using VOT (Rp. 7076/hr)

Table 4.2 Tariff as '%#-age

of VOC saving using several value of time assumption
Vehicle class I

36
Table 4.2 (continued)
Vehicle class IIA

Table 4.2 (continued)

Vehicle class llB

Jasa Marga current practice seems to use a rather high value of time (Rp 12,287/hr
for Class I). Using this value, tariffs as a percentage of VOC savings were found in
the range of 17% to 56"; in interurban toll roads and, 9% to 55% for urban toll roads.
Most urban toll roads have tariff above 52% of VOC saving, except Jatingaleh-
Krapyak in which 9% was found. For interurban toll roads, Dupak-Kebomas has a
highest tariff (56%),whereas Jakarta-Cikampek being the lowest, i.e. 17% relative to
VOC savings.

Using IHCM (1995) assumed value of time, which is about a quarter of Jasa M i i r p ’ s ,
the tariffs as percentages of VOC savings arc now double all the pre\%ws mentioned
tariff. For example, for Class I tariff as ‘X, VOC saving on Jakarta-Cikainpck Lxurmcs
more or less 33%, Tangcrang Barat-Mcrak tariff change to 108% VOC sa\,ings, ‘ I n d
s o on.

Using the newly devclopeci nioclel to obtain the tariff as percentage of VOC saving,
it was fouiiJ that the ratios o f new model : PCI tariff is about 1.7 times if IHCM
value of time was used and 1.3 tinics if Jasa Marga’s value of time was used. For
example, using IHCM’s VOT tariff on Pci. Pinang-Kp. Rambutan is 99.7% VOC
saving under PCI mociel, and this change to 167% under the iiew mociel. Using J‘wa
Marga’s VOT this become 54.9%under PCI iiioclcl and 70.6% under tlie iicw i i i o d c +

Taiiff
_’ .
for vehicles Class IIA and l l B seem to be lower compared to tarrif Class I . All
tariff Class 11 vehicles are well below 100%VOC saving. The most expensive vehicle
Class II tariff were on Cawang-Tomang and I’d. Pinang-Kp. Rambutan toll roacis.

4.5 Conclusion on VOC Saving

The foregoing analysis s h o w s that the VOC and time value matters arc vet to be
settled. Several VOC models have been developed in tlie past ranging troni a simple
model based only on vehicle average speed to a complicated niociel taking into
account the irregularity of tlie road surface, traffic condition and road geometry. The
original PCI model is a simple model, in which each cost coinponelit is best
explained by average running speed. Otlier existing VOC models, such as HDM 111
based model are laborious to be used for tariff calculation, as the model specification
is rather complicated and data hungry.

In the future it is recommended to use a rather moderate model specification. One

alternative is to use the newly revised model previously discusseci.

Regarding the time value, it is strongly recommended to carry out an extensive

stated preference study to derive a proxy of value of time, not only for a toll road
routing choice, but also for other trip making choice behavior.

In the medium term, the VOC savings ceiling will still be as viable guideline for
tariff setting and adjustment. However, in the long term when total urban road
network, toll and non-toll, becomes more congested - partly due to the continuing
growth trend of car ownership plus failure to implement the land use master plan
and control, it is most likely that relative VOC savings will become smaller. If urban
toll and non-toll roads become equally congested tariffs will have to be raised to toll-
off traffic and restore levels of service. This prospect o f congestion charging will
necessitate a re-interpretation of the VOC saving ceiling.

In the medium term this can be resolved by re-interpreting the VOC costing
methodology and the tariff adjustment formulae mcchaiiism. In the longer r u n , '311
indexed tariff adjustment formulae could also iivxease tariffs beyond the 100% of
VOC ceiling. This situation is particularly truc tor toll road in urban area wl~erc~
rod
network perhrmaiicc would likcl\r be more congested in the future, espcciall\. i t 110
emphasis on urban transport strategv to favor public transport service put in place.
If this happens, as may also occur clepeiiciiiig on VOC costing methodology, the
legal framework governing the toll road operation will have to be re-interpreted.

5. CAPITAL, OPERATING AND MAINTENANCE COSTS

5.1 Coinparison of Investment Costs for Different Toll Roads

Examining initial construction costs for toll roads, initial investment costs for
intercity toll roads have increased by several multiples since the Jagorawi was first
built in 1978. The cost of this first toll road was 43.2 billion Rp in 1978, or about 1
billion/km. The first section of the Semarang toll road from Jatingaleh to Srondol
opened in 1983 ancl cost somewhat less, 4.73 billion o r 0.75 billion R p / kin. Somc- of
the lower cost of this can be attributed to the open toll system which required less
investment in toll plazas and access restriction. The cost of gates and cxI UI'11ment
remained comparable with other toll roads in terms of the 2.7% share of total cost.
The next section, Jatingaleh -Krapyak opened in 1987 ancl cost 1.75 billion Rp/ kin.
Cost of gates ancl toll macliinery was less, 1.6% of total cost, again because of the
open system ancl sharing of the gate at Jatingaleli.

The Cikainpek toll road opened in 19C9, costing 313 billion or 4.3 billion Rp/km.
Widening of the toll road from Cikampek to Cibitung has cost another 57 billion rp.
and also necessitated expansion of toll plazas. Including another 41.4 billion in
pavement overlay since 1993, total capital investment in the Cikampek has
amounted to 355 billion Rp. Cost of toll gates and equipment systems, including
renovations and improvements amounted to 2% of the total cost over this time o r 7.2
billion Rp.

The Banclung bypass, Padalarang- Cileunyi was built two years later and cost
slightly more per km., 4.7 Rp. billion or a total cost of 223 billion Rp. Cost of toll
gates ancl equipment, including new installations in 1995/96 amounted to 3.7 billion
or 2.8% of total investment.

39
For urban systems in Jakarta, costs are substantially higher. Including sections of
the Inner Ring Road, Harbour, and North-South connector, construction costs have
been in tlie order of 13 to 32 billion/km. According to the joint venture partner, PT
Citra Marga Nusapliala Persacla, costs of toll facilities and equipiieiit has ainountecl
to 15% to 20 % of total investment cost. While no information was iiiacle available
on toll gate and equipment costs for the south section of tlie Jakarta Outcr King
Road, the plan amount of 546.7 billion for 15km would work out to 37 billioii/km.

Using the consumer price index annual inflation rates, 1997 base costs Iia\~cbccw
cstiiiiatcd for the case study toll roads. The Seiiiarang bvpass sections rcniiiili thc
lowest cost, averaging 3.4 billion/kni. At current prices, tlie Cikampc,k ,ind
Paclaleunyi toll roads cost 8 and 7.5 billion/ k m to construct rcspectivel\r. TI><\ncbxt
iiiost expensive is tlie Tangerang Barat-Mcrak, estimated in 1997 ternis to cost 13.1
billion/ k m .
Table 5.1.1
Construction 8 Improvement Costs: Case Study Toll Roads
(bill. Rp) Est. In". in '97 CPI Inflation
RP.
Initial lnvestmenl Year Conslruction cost Land 8 Improvernents/Rehabilil Rp.bill.
ation
Toll Roads 8 Sections km's Operator total per km. Opened Roadway per km. Toll ROW Type ofwk. cost Yr Total per constr. impr.
facilities km

Jakarta-CikamDek 72 Jasa Marga 312.8 4.3 1989 256.5 579.0 8.0 1.851

Padalarang-Cileunyi 47.5 Jasa Marga 222.7 4.7 1991 355.1 7.5 1.594

Semaranq (a) 14.3 Jasa Marga 20.3 1.4 -

49.3 3.4
Krapyak-Jatingaleh 8 1987 14.02 1.75 Rehab Slope 1.63 1992 32.0 2.126 1.333
Jatingaleh-Srondol 6.3 1983 4.73 0 75 Widen 3.02 1995 17.3 2922 1.157

Surabaya (bJ
Dupak-Kebomas 8 PT. Margabumi 93/94
ext. to Manyar l/S 20.7 Matrarava 204.2 9 9 96 1726 83 142 1 7 4 35kmoverlay 1 1 9 1997 2371 11 5 1263 1
(6cm)

Tanqeranq Earat- Merak PT.Marga

-
Tangerang Earat Cilegon Et 69 Mandalasakti 651.2 9.4 92/95 902.3 13.1 1.386

Jakarta Urban System

Cawang-Tomang 16 Jasa Marga 267.8 16.7 1989 4956 31 0 1851

Cawang-Tg.Priok (c) 17 PT.Citra Marga 223.0 13.1 89/90 220 12.9 3 3893 229 1746
Nusaphala Persada

Tg Priok-AncoITm - PT.Citra Marga

Pluit (Hbr Rd) ( c ) 14 Nusaphala Persada 455.0 32.5 95/96 450 32.1 5 4869 348 107

Higliwav Toll Study - ADB 1997 41

~~ ~ ~ ~__
Table 5.1.1 (Continued)

JORR S: Pd. Pinang- PT. Marga

Kg, Rambutan (d) (secksi S) 14.83 Nurindo Eakti 546.7 36.9 95/96 5849 39.4 107
Total Project: Seksi S 8 E l : constr.
Ruas Pondok Pinang- Jagorawi-Cikunir (e) 1237.7 453
Seksi S 14.25 539.2 37.8 94-96 330.1 23.16 6.8 202.3 note: does not include vehicle cost @ 1.02 bill. or int. during COn5tr.. 53.3
Seksi €1 12.5 698.5 55.9 95-98 529.8 42.42 28 4 140.3 note does not include vehicles @ 3 bill or IDC.
129.6

Sources Jasa Marga Main Office except where noted

(a) Jasa Marga Semarang Branch Office
(b) PT Margabumi Matraraya
(c) PT Citra Marga Nusaphala Persada
(d) Summary reports on joint ventures & Pelita VI program estimates Jasa Marga
(e) Jasa Marga Investment Division

Highwav Toll Study - ADB 1997 4Ia

5.2 Maintenance Costs
Based on analvsis of the cietailccl 1995 and 1996 ]asa Marga Branch niaintenancc and
I

direct operating costs, two reprcscntati\w Jasa Marga cases chosen, the Cik'inipck
and Paclaleunyi cost in tlic order of 155 million and 130 million Rp./\.c>'>r to
maintain. This includes the cost of routine maintenance at 45 million/km/\.r plus
annualized cost of periodic ovcrla\~sin the order of 90 to 110 million per kni. The
Dupak-Keboiiias section of tlic Suraba\~,i toll road is estimated to run tiiorc
cxpensi\w than ADT niight indicate hccausc of the 11ca\~truck traffic carried 011 this
road.

While routine maintenance and o\,crla\: cost data for tlie Tangerang-Merak toll road
is only fragmentary, a higher niaintcnance cost in relation to ADT has also hui
estimated for this road, in the order of 150 iiiillion/kiii/vr.. Although the facility is
fairly new, evidence of cracking and concrete slab deterioration is already cvidcnt
on sections that were opeiicd onlv i n 1993 a n d 1995. Shoulders also are dctcrioratc'd
and moisture is probabl\r reaching into sub-lavers.

For Jakarta urban toll roads, routine maintcnancc costs are higher becausc of the
high traffic load and the higher cost of services. Including periodic maintcnancc,
annual costs are estimated to be in tlic order of 450 to 500 million Rp. /kin,

Operating costs for toll roads v a n ' both as a function of traffic and road length.
Costs include those of collecting tolls, maintaining toll plazas anci booths, as ~ ~ as l l
such aciclitional costs such as plaza lighting, direct admi~iistration,anci data
processing. Estimates have been based on Jasa Marga branch operating cost data
and information supplied by PT Citra Marga Nusaphala Persacia (Harbour Road
and Cawang-Tg. Priok) and PT Marga Buiiii Matraraya (Surabaya). Costs run in tlie
range of 4 to 7.5 inillion Rp/km/C)OO ADT for interurban roads such as Cikampek
a i d Padaleunyi, as well as the Surabava Dupak-Kebomas section which is under
private operation. This range is broadly comparable to the estimated operating cost
of the Ciawi-Sukabumi project, discussed further i n Chapter 7.

Sernarang unit costs are higher than might be expected given tlie open system but is
perhaps staffed higher coinpared to tlie short length and the lower traffic.
Tangerang Barat- Merak operating costs have been estimated b a d on the
operator's annual report on jont venture operations. Costs appear cornparatiid\,
high at 14.6 million Rp/kin/000 ADT, but this may include some indirct overheads
excluded in tlie direct operating costs examined for other toll roads.
Table 5.2.1
Roadway Maintenace B Toll Operation costs
(mill Rp)

Roadway rnaint cost Direct Operational Costs

Toll road Lengt 1995 1996 1995 AADT 1997 Est. Roadway maint costlkmlyr 1997 Est. Direct OP.
h Costs
total millRp total millRp total excl. rd. millRp mill Rp/km mill.Rplkrn/yr. mill Rplkm /000ADT/vr.
/km 8%/yr /km maint. /km 000 /000AADT & periodic * routine 8. periodic (Excl. roadway maint.)
esc.
(8%esc/ 83.4 Total 1.166 ='97 esc

Jkta-Cikampek 81 3123 39 3373.1 42 19033 15910 196 53 3.7 45.0 108 153.4 4.3
(a)
Padaleunyi (a) 46.8 1766 38 1907.3 41 9180 7414 158 27.5 5.8 44.0 89.2 133.3 6.7

Surabaya (b) 1996

Dupak-Kebomas 16.4 n/a 325 20 1936 1611 98 15.2 6.5 21.4 100.1 121.5 7.5
1995
Semarang (c) 14.3 284 20 306.72 21 2497 2213 155 14.2 10.9 23.2 70 92.7 12.7
1995
- Bt.-Merak
Tans. 34 903 27 975.2 29 7714 6811 200 16.0 12.5 40.3 116.8 157.0 14.6
(d)
(Ciujung. 1993194)
(2005 base yr. estimate by Bukaka, costs const.'97)
Ciawi-Sukabumi 53.5 9264 173 37.1 4.7 50.6 83.4 134.0 5.4
(e)

Jakarta Urban Toll roads @I .2 * intercity

Cawans- 30.3 9325 308 10071 332 15740 6415 212 134.9 1.6 359.0 100.08 459.0 1.8
Toman&
Cengkareng (9

Cawang-Tg.Priok 17 5038 296 4945 291 11997 6959 409 65 6.3 314.2 100.08 414.2 7.3
(9)
Table 5.2.1 (Continued)

JORR S: Pd. Pinang- under consl. (avg. Cawang)

Kg. Rambutan (h) 13.1 1823 1057 81 24 3.4 336.6 100.08 436.6 3.9
43
* periodic maintenance cost (Rp.mill 417 Ikrn) annualized over 5-yr. cycle at a Rp mill. 83.4/km/yr (4-lane)
from Bukaka est.. where ADT<25.000. 7-yr cycle assumed except for Tang-Merak 8 Surabaya
Heavy traffic. pvm't or sub-base conditions include escalation factors, ie., 1.2 (urban) 8 1.1 to 1.4 for Cikampek, Merak. Surabaya
66.7 mill Rplkmlyr also estimated for Sernarang case, based on 9.5 km overlay in 1994
(a) Jasa Marga annual report 8 Accounting Department
(b) Jasa Marga Branch Office records 8 PT Margabumi Matraraya
(c) Jasa Marga Accounting Dept. 8 Branch Oftice records
(d) Jasa marga Accounting Dept. 8 1995 summaru report on Joint venture operations
(e) PT Bukaka Marga Utarna
(r) Jas Marga Annual Report, AADT based on avg. Cawang flow to Jagorawi 8 Tomang
(9) PT Citra Marga Nusaphala Persada, ADT based on Jasa Marga chart. Cabang Cawang-Tj. Priok
(h) Jasa Maarga summary report on joint ventures with maintenace assumed at same share as Cawang-Tj.Priok, AADt from JM chart,
Cabang P. Pinang-L. Agung, Mar-Aug., 1996

4321
5.3 Comparison of Tariff & Maintenance Cost Responsibilitv B y Vehicle
Type: Cikampek Toll Road Example
A comparison of roadway maintenance costs and toll tariffs per km. Iias been carried
out for the case of the Cikampek toll road. Routine maintenaiice costs have been
estimated at 3.37 billion R p for 1996 or 45 niillion R p / kin. Periodic maintenance cost
has been taken a t 108.4 i n i l l i o n / km/vr (also shown in Table 6.2.2) and otltcr
maintenance costs such as spot i l ~ l t l r ~ ~ \ ~ e ~ l lurgent
e ~ l t s , maintcnancc., ~ n d
maintenance of structures estimate based on average expenditures for all jas‘i Mclrgil
branches for other roadwav- related maintenance.

These costs were then allocated to the three cost assignmcnt categories, vehiclc4ans,
equivalent stanclarcl axle-knis (ESA-kin’s), and passenger-car units-km’s (PCU-kms).
These allocations were based on attributions derived in preizious road cost
allocation studies for World Bank and Directorate of Land Transport. Avcrage
ESA’s by vehicle type are taken from Axle load surveys carried out for the l-leav~~
Loaclecl Road Improvement Project (HLRIP) b \ v PCI in 1995 but mav not reflect the
full extent of over-loacling, particularlv for medium ,-axle trucks which are ch,irgcd
as Class 1IA on toll roads.

Because much of the maintenance cost varies by vehicle-kin, and light Categorv I
vehicles predominate in the traffic flow, this group has the highest cost share. Based
on an average toll tariff of 100.8 R p / kin, maintenance cost shares for light vehicles
amount to about 5% of toll tariff. For Category IIA medium-sized vehicles, the
maintenance cost share of tariff at 170.5 Rp/vehicle varies between 4% for buses to
over 11%for 2-axle trucks. For over-loacled trucks which exceed the average IiLRIP
estimate of 2.08 equivalent stanclarcl axle load, the ratio of cost share to tariff woulcl
exceed that of heavy trucks which impose a maintenance cost in the order of 17% of
the toll tariff of 201.6 Rp/km.

The conclusions of this maintenance cost allocation analysis are that toll tariffs
adequately cover by seve:-al inultiples the maintenance cost shares imposed by
vehicles. Category I vehicles make up most of the flow and hence bear a large fleet
share because of the traffic volumes. For this group, maintenance cost responsibility
amounts to about 5% of tariff paid. With respect to heavy vehicles, the ratio of
vehicle maintenance cost responsibility to tariff is much higher, but still remains
below 20% on average. Given this, however, instances of gross overloading and
accompanying pavement wear and damage occur for individual trucks, particularly
in the medium, 2-axle truck group. Enforcement of vehicle weight regulations is the
only practical solution for this problem. For other cases where accelerated pavement
deterioration is occurring such as the segments of the Surabava Toll road which
serves heavy port traffic, underdesign of pavement is more the problem.

44
Table 5.3.1

Traffic Estimates and Charqinq Units: Inter-Urban Toll Roads (Cikampek Case Studyl
1996 Vehicle -kms Equivalent axles Pass car equivalents
Vehicle type Volumelday veh-kms/yr % per veh. ESA-kms per veh PCE-kms
(thous) mill. share (a) mill. (b) mill

cars 40.8 14.9 68% 0.0026 0.04 1 14.9

utility veh. 12.0 4.4 20% 0.0027 0.01 12 5.3
med. bus 0.4 0.1 1% 0.092 0.01 2 0.3
med. truck 4.5 1.7 8% 2.079 3.44 25 4.1
Hvy truck 1.9 0.7 3% 4.410 3.03 3 2.1
Lg. bus 0.8 0.3 1% 0.155 0.05 2.5 0.7

Total (c) 60.4 22.0 100% 6.6 27.4

(a)Equivalent standard axle (ESA) values taken from Table 3.6,

(West 8 Central Java) Final Engineering Report.
Heavy Loaded Road Improvement Project, January 1995, Pacific Consultants International
Report gives avg. for lg. buses, but super coach 2-axle units likely higher
(b) PCU factors taken from Final Engineering Report. HLRl Project. January, 1995.PCI
(c) Vehicleslday based on JM volumeldistance formula:
(daily volume each direction by section x section distance)lsum of section distances
(Ref: Cost Recovery for Roads in Developing Countries. Ian Heggie 8 Vincy Fon. World Bank. 1991)

Table 5.3.2
Road Maintenance Cost Allocation: Cikampek Toll Road
(1996 million Rp)

(Infl. factor 95-96: 1.08

Allocation of Costs According to
Total cost Costlkmlvr veh.-km's ESA-km's PCU-km's
1996 81
(a)
Routine maint. (b) 3373.06 41.64 37.478 4.164
Periodic maint. (c) 108.42 54.210 32.526 21.684
Spot Impr. (d) 5.00 2.500 2.500
Urgent (e) 0.48 0.483
Bridges (f) 2.56 0.769 0.769 1.025
Total Road Maintenance 158.11 95.440 39.959 22.709

Toll Facilities

Allocation of 1996 Estimated Maintenance Costs (excl. toll facilities1

Charge units (million) 22.05 6.579 27.360
Rp./veh-km 4.3291
Rp./ESA-km 6.0738
Rp/PCU-km 0.8300

Table 5.3.2 (continued)

(Reference: Heavy Loaded Road Improvement Project. PCI. 1995;
Watanada. T 8 W. D. Paterson et al. Highway Design 8 Maintenance Standards Series. WB.
1987
Road Cost Allocation: Road User Charge Study, (Perhubungan Darat)
(Halcrow fox, Nathan, Biec & Associates. 1983)
(a)Maintenace costs from Table A6.1 (Annex)
(b) 90% related to veh-km's; remainder lo heavy vehicles
(c) 50% related to veh-km's. 30% to heavy veh.. remainder to carriageway width (pcu's)
(d) 50% related to veh-km's and ESA's respectively
(e) 100% related to veh-km's
(0 30% related to veh-km, 8 ESAs, remainder to PCU's

Table 5.3.3
Routine 8 Periodic Cost Allocation by Vehicle Type Compared With Toll Tariff

Costlveh. mill. kmlyr Annual user cost share Toll Tariff

Comparison
w Tariff Costmarif
f
veh-km byveh,tvpe (mill.Rp) % Rdkm %
cars 5.17 14.9 77.060 49% 100.8 5.1%
utility veh. 5.34 4.4 23.374 15% 100.8 5.3%
med. bus 6.55 0.1 0.959 1% 170.5 3.8%
med. truck 19.03 1.7 31.503 20% 170.5 11.2%
Hvy truck 33.60 0.7 23.077 15% 201.6 16.7%
Lg. bus 7.35 03 2.1351% 170.5 4.3%

Total 22 158.108 100%

6. PROlECTED GROSS AND NET REVENUES FOR CASE STUDY TOLL

ROADS

Based on maintenance and operating costs developed in Chapter 5 and the prospects
for further traffic development, corridor congestion and growth trends discussed in
Chapter 2 and in the Interim Report, projections have been made of gross and net
revenues for the case study interurban toll roads.

For Jakarta -Cikampek, traffic growth rates have declined since the 23% registered in
1994 to a level of 14% to 15%. With the maturing of this facility it is forecast that
traffic would grow at 13%to the end of the decade, then decline to 12%. By 2005, a
medium growth forecast would indicate growth in the 9% range, declining to 8 % by
2010. Gross revenues based on trends for Cikainpek and Jagorawi would incrcase at
about 5%to 1.5%faster than ADT.
Jakarta Cikampek (72.5 km)
Traffic 8 Revenue: Medium Estimate

For the Paclalcunyi toll road, recent traffic grou4i has slowed from the rates of over
20 '%) in 1994 ancl iminecliately after opening. Traffic growth is forecast to slow to
16% up to 2000, and up to 2005, grow at 15%. Afterwards, growth would continue
in the order of 1176 to 1 2 %, declining to 7% to 8 % bv 2010. Gross revenues woulcl
crow about 0.75 to 1.5% faster than ADI' g i \ w the future prospect of more liea\,\r
0

traffic diverting to the toll road.

Padalarang-Cileunyi

For other case study interurban toll roads, growth estimates depend upon maturity
of development. In the case of Dupak-Kebomas, further dramatic grow in the order
of 20% is expected until 2000. By 2005 this is expected to decline t o 17%ancl then to
12% by 2010. For the case of Jatingaleh-Krap\.ak, growth is expected to be more
moderate, following the order of 8% seen in recent years. This is expectecl to
continue to 2000, and decline to 7% thereafter. More rapid development in the
Seiiiarang area and connection with the Trans-]a\ra Toll network could increase the
econoinic pace and traffic however.

For the Tangerang Barat-Merak toll road, the continued push of cie\dopmcnt west
of Jakarta a i d development of the port and industry in the Mcrak area coulcl be
cxpccteci to maintain traffic groivtli in the order of .l5% to 2000. F r o m then until
2010, growth is expected to continue a t a 12%r'itc.

7. CIAWI-SUKABUMI TOLL ROAD PROlECT CASE STUDY

7.1 Background
A nuiiibcr of studies have been carricci out on the Bogor-Bandung road corridor
starting with the 1990 JICA study, anci most recently 1996/97 reviews of the Ciawi -
Sukabumi Toll Road by Rcncicl Palmer & lritton ILtci. (RPT) for tlic lead investor, PI
Bukaka Marga Utama and bv Mitrapacific Consulinclo International (MCI) for the
Asian Dcvelopiiient Bank. The RPT review was largely a financial anali'sis with
traffic and revenue projections based on the in\,cstor-proposeci level of tariffs. The
MCI review submitted March, 1997, also carried out traffic forecasts as well 'is an
economic analysis of tlie project.

Both of these most recent studies reported favourable rates of return. The [ < P I
review reported a financial internal ratc of return of 21% for tlic base case m d a
return on equity of 26%. The MCI analysis found a n economic internal ratc of
return of 28% for the base case. This chapter draws some comparisons of traffic
forecasts and examines in iiiore detail the economic and fiiiaiicial returns of this
project. In particular, the economic analysis compares economic returns on tlie
project as a noii-toll public road and as a BOT toll road.

7.1.1 JICA 1990 Feasibility Study

I n 1990, a feasibility study on the Bogor-Bandung road development project was
carried out under the auspices of JICA. The studv concluded that construction of a
new road from Ciawi at the elid of the Jagorawi toll road through Cibaclek,
Sukabumi and Cianjur was feasible. It was planned that this new toll road ~ w ~ l d
extend to Citatah, connecting with the planned new Cikampek-Padalarang toll road.
The distance between Jakarta and Padalarang, (13 kin. from Bandung) bv this route
would be 176 kin after construction of the new highway. Based on traffic estimates
the study recommeiidecl staged construction, with an initial 2-lane, controlled access
toll road between Ciawi and Sukabumi by 1998. This 2-lane facility would be
extended to Citatah by 2005, anci the initial Ciawi-Sukaburni section would be
expanded 'to 4 lanes by 2010. Cost in constant 1989 Rp. for the first phase between
Ciawi and Sukabumi was estimated a t 241.2 billion (including land acquisition and
!
resettlement costs of 15.6 billion Rp. The third phase acidition of two aciditional
lanes was estimated to cost an additional 190 billion Rp. To construct a 4-lanc toll
road between Ciawi ancl Sukabumi was estiinated to total 430.8 billion R p . (US$
246.2 iiiillion at the 1989 1750 Rp. : $ rate). Using a 3% escalation rate 011 US5 costs
ancl a current rate of 2400 Rp./US$, this would represent ri 1997 cost ot 748 L ~ i l l i o n
Rp. , (720 billion Rp. excluding land acquisition). Using the Lilomcstic c(msu~ncr
price index ( 0 1 ) as an escalation factor, thc 1YY7 estimate would ha\^ h c c w 797
billion Rp. or 768 billion Rp. excluding land.

The Bukaka cost estimate for design and construction in constant 1 W 7 l<p.
(excluding inflation, VAT, interest during construction, and other financing costs)
amounts to 340 billion Rp. , less tlian the earlier JlCA estimate. Land costs on the
other hand reflect mucli more inflation, 88 billion in constant 1997 Rp., coinpired
with 30 billion Rp. if the 1989 estimate wcrc inflated by tlic CPI. Construction cost
estimates for the Bukaka proposal arc discussed in iiiorc detail below in scction
7.1.3.

Tlic 1990 JlCA studv also recoiiiiiiciidcd a nuinber of spot iiiiprweiiieiits on the
Puncak Pass road between Ciawi and Cianjur to case iiiiiiiecliatc traffic problcins.
These spot impro\~ementsincludeci widening and intersection iinproveimmts,
construction o f climbing lanes, pa\,iiig of slioulcicrs, and a number of measures to
improve safety. While this route is shorter to Cianjur, 50.3 kni compared to 80.5 kin
via Sukabumi, the terrain is more difticult and probleiiis of slope stahlit\, more
acute. Watershed protection and regional planning concerns also dictatc that
growth be ciirecteci towards the southern Cibaclak-Sukabumi corridor, awa\' fro111
tlie northern Puncak slopes.

The JICA study assumed that the Cikaiiipek-Pa~alaraligtoll road would also be
built ancl took the effect of this alternative corridor into account in network traffic
assignments. This alternative route from Jakarta through Cikampek to Bandung is
approximately 40 kin. shorter than tlie Bogor-Sukabumi-Ciaiijur route.

7.1.2 Traffic Estimates

Tlie 1990 JICA stuclv found froiii their traffic surveys carried out in 1989 that
corridor traffic was in tlie order of 9500 vehicles per day between Ciawi and
Sukabumi, declining to slightly more than 4000 ADT on the Sukabumi-Ciaiijur
stretch. Tlie JICA study forecast average annual growth in the order of 3.6% per
year for traffic, reaching by the year 2010 ADT's of around 20,000 for tlie study
section.

I n traffic studies carried out in cooperation with PT Bina Karya in 1996, the RPT
consultants recorded average daily traffic in tlie order of 20,000 vehicles per clay
between Ciawi ancl Cibaclak and over 17000 per day between Cibaclak and
Sukabumi. Comparing this with JICA 1989 survevs, traffic had grow^ at a n a\'ercige
anliual rate of 10.4% and 9 % on the two stretches of road respectivelv. Bina M q a
counts in 1994 and 1995 had also indicated higher growth rates in the micl-'90's, 13.5
% pel' year.

Traffic growth estimates for tlie RPT projections were based on estimates froin other
recent studies, with growth rates assumed to lie between that on arterial roads in
West Java, annual rates of 6.5%, and in the Jabotabek area with 15-20 ''it ,innudl
growth. Also quoted were estimates made bv U K Transport Research L'1bor~torv
for the Cikaiiipek-Pacialarang toll R o a d Stud\, at 6.8% tor passenger trattic ~ n c 7.2%
i
for freight traffic as well as PT Jasa Marga network analysis studies which Iiaw
taken an annual rate of 7%. According to tlic RPT text, growth rates used were .lO'j%,
9% , ancl 8% per year for 1996-2010 periocls rcspcctivelg for the Ciawi-Cicurug,
Cicurug-Cibadak, and Cibadak-Sukabumi sections and 6% per year tor the period
2011-2023 for all sections. To arrive at tlic \'ear 2000 base, Categorv 1 trattic wris
increased from 1996 by 11.2%, 14.7% and 7 % per vcar for the three road sc7ctions.
For the period 2000 to 2010, 8.4% annual growth was assumed for all \'eliiclc
categories for all scwtions.

In comparison, the MCI stud!, applied a network modeling approach usccl bv I T 1 in

other Java road studies. Zonal vcliicle trips were calculated using a regression
moclel with GRDP and population as inclcpcnclent variables. These trips wei~cthen
distributed using a gravity model ancl assigned to tlie network using an cquililviuiii
moclel. Diversion between tolled and non-tolled roads was simulated using tlie
moclel with the same diversion curves used by RPT which were previouslv gii'cii i n
the TRIP" study. These curves were in turn adopted from JlCA studies and sur\'cvs
on the Jagorawi,Jakarta -Cikampek, ancl Jakarta -Merak toll roads in 1988.

Comparing RPT ancl MCI traffic forecasts for total traffic assigned to tlie toll road,
estimates for the Ciawi-Cicurug were fairly close, RPT estimates for total traffic
being 4% higher than MCI fc,r the year 2001, but 8 to 12% lower than MCI for tlie
Cicurug-Cibadak and Cibaclak-Sukabumi stretches. For the 2010 projection, RPT's
estimates are 19% higher again for the first section, but for the sections beyond
Cicurug, 5% and 11%lower than MCI estimates. Average annual growth rates for
all sections applied by MCI are in the order of 10-11% for tariff class I ancl 11 up to
year 2010, and then decline bv one percentage point for the years 2010-2015, and by
the same amount again 2015-2020.

l 2 Update of Toll Road Investment Programme, Final report, April, 1990, I

TJasa Marga Persero (IURD
Loan) by Sweroad in association with INDEC & Associates.
Traffic Growth Parameters & Forecasts: Mitrapacific Consulindo International

-
1990 1995 2020
West Java (const. 1983 RP)
GRDP (bill Rp) 15.64 22.98 34.75 52.12 76.52 163.87
Avg. %chg/yr 8.0% 8.6% 8.4% 8.0% 7.9%
GRDP/capita (000 Rp) 442 582.8 800.1 1101.2 1497.2 2809.5
Avg. %chg/yr 5.7% 6.5% 6.6% 6.3% 6.5%

DKI Jakarta
GRDPIcapita (000 Rp) 1661.9 2270.3 3186.3 4513.7 6352.5 12936.7
Avg. %chg/yr 6.4% 7.0% 7.2% 7.1% 7 5%

Avg. %chg/yr MCI Proiected

West Java Vehicle fleet arowth -
90-95 2000-2010
cars 3.4% 9.8% 9.2%
buses 6.9% 5.2% 3.5%
trucks 2.5% 4.8% 6.2%

2005
Total Java TrafficGrowth (MCI trip generation analysis)
000 Trip ends 1150 1530 2020
Avg. %chg/yr 6.3% 5.9% 5.7%

The RPT study built up the separate cleiiients of traffic forecast to use tlic new toll
road. This potential toll road traffic would consist of diversion from the existing
corridor road, diversion from the Puncak route, and generated traffic a s s u i i i l d to
develop as a result of increased accessibility and growth in the route miic of
influence. Diversion from the existing corridor road was based on volumc/capicitv
and speed-flow relationships developed in the TRIP study and for the Iiidonesia
Highway Capacity Manuall? (IHCM) by Sweroad. With increasing traffic and
congestion, travel time savings would become more significant on tlie toll road and
diversion would increase as a result. In the early years after opening, diversion of
Gol. I vehicles to the toll road would range from 50% on the Ciawi-Cicurug section
to 46% for the Cicurug-Cibadak section. By 2005, this diversion was forecast to
increase to 57% on the first section, 54% on the second section, and 51% on tlie third
stretch, from Cibadak to Sukabumi. By 2020, this diversion would have increascci to
75% of corridor traffic for the first two sections. For Gol. I1 heavy vehicles, the
diversion curve applied was much flatter, assuming more of a concave shape, not
rising until total corridor traffic exceeds 40,000 ADT. Not until after 2005 would
diversion climb above 25% to 30% according to this model. By 2010, diversion of the

Indonesia Highway Capacity Manual, BINKCIT, E’eb. 1997, Sweroad in Association !<it11 11’ liina
Karya (Persel-o)
Gol. 11 group would have climbecl to 33% to 46%, with higher rates on tlic tirst
section from Ciawi.

The same diversion moclel was applied to forecast diversion from the Puiic'lk route
bv RPT. Travel time savings t o Cianjur were assumed to amount to 60 minutc-s via
a new toll road over tlie Puncak route, eve11 though tlic Puncak route is 30 k m
shorter. Bv 2005, RPT calculated the time saving to be 95 minutes and b\. 2010, 1-1-1
minutes. Based on this ancl the 63% share of survevecl Puiicak traffic 11eadcd t o
Ciaiijur and beyond, RPT forecast more than 10,000 ADI' to divert lx 201 0,
increasing to almost 18,000 ADT b\, 2020. The Cikatiit~e,ck-Pa~~'llaran:: toll i.oxi
~ ~ o u lalso
c l compete for this traffic.

The estimate of generated traffic brought about bv increased accessibilit\., economic

activity, and release of latent ileiiiand has been assumed bv RPT to be rehti\,elv
high, 25% from opening ycar, increasing to 30% of normal growth traffic b\. 2005.
These estimates were based on prospects for development in the Sukabumi region,
including the Pelabuhan Ratu clevclopment and increased automobile ownershiti
with the lower-taxed national car on tlie market.

While the MCI study briefly discussed generated traffic, no estimate ot its
~

magnitude or mention of its share in the total toll road traffic has been given i i i tlic
report. Redistributed trips were in theory considered in the trip distrihutioii model
employed by MCI. The study does state that induced traffic woulcl likelv not appcar
inimecliately with opening of the new toll road but woulcl clevelop after a few vears
of operation.

7.1.3 Cost Estimates

Bukaka has estimated the construction cost for the toll road project to amount to
429.8 billion Rp. The RPT review gives a total of 430 billion Rp. in 19% prices,
excluding supervision, land costs, finance charges, administration ancl VAT (PPN)
taxes. The TBV/Biiia Karya 1995 study estimate a construction cost of 498 billion
Rp., 16% more than the RPT Bukaka estimate. The RPT cited several areas of
potential cost saving; these included more efficient and standard designs for bridges
and box culverts, economies in earthworks, reduction in the initial thickness of
pavement to be compensated for with a heavier structural overlavs later a s heavy
traffic builds.

7.2 Economic Analysis of Project As A Non-Toll Road

The economic costs of the Ciawi-Sukabumi project as a non-tolled highwav include
the cost components of design, supervision, and construction, net of taxes, plus the
cost of land acquisition. Construction costs have been reduced to account for the
additional costs of a toll facility. These elements include toll barriers, service areas,
administrative offices, and toll apron lighting, fencing, signage, ancl ;arkin::. A
closed toll system also requires an estimated 10%additional cost in the provision of
interchanges, access ancl exit ramps. Also excluded is the capital and maintenance
cost for toll collection equipment as w e l l as aciditional persomiel costs for collc,ction
and administration. Routine and periodic maiiiteiiance costs for the ro'ild\v'i\. ‘ire
estimated to remain the w i l e as with a toll road although soinc reduction i n p \ u i
apron areas would be achiwcd.

On the benefit side, \diiclc operating cost s'i\,ings havc been counted tor tlic
reduction in distance with the n e w route between Ciawi-Sukabumi (4 k m t o r trips
passing the entire length) and tor sa\ings in reduced congestion cicl,i\.s \\,it11
increased operating speccis. Passenger time savings tor cars and buses Ii'i\,c uscd
the trwel time cstiinates based on a \ w a g c speeds of 80 kph and 60 k p h for Lol. 1
and II \~ehiclesrespectively. Economic time values 'ire as applied bv MCI, using the
jlCA methociologv as in the Cikampek-Cirebon 1990 stud\..

Because of the greater trawl distance ancl \vIiiclc operating cost o n thc Ci'iwi-
Sukabumi route for potential Puncak diversion, savings consist only of tinie, and
these are marginal. A more conservative view of generated traffic has been t'i ken
because of the high growth rates already contained in the normal growth projection.
This element is estimated at 25% of normal traffic at its highest, but following
customary practice, only half the benefits which accrue to normal traffic arc counted
for generated traffic.

Diversion to the new highway without tolls is based on the \,oluiiie/capacit\,, speed
flow relationships from the IHCM. The number of trips diverting to the new
highway is however limited by the number of through trips. Local trips were
assumed to remain largely on the old road.

The result of examining the project as a non-toll facility with its lower capital cost
would be an economic internal rate of return (EIRR) of 29.8%. WIien domestic costs
and time savings benefits are adjusted applying the staticlard conversion factor of
0.943 to domestic non-traded components ancl 75% valuation of lion-skilled labour,
the effect is to increase the EIRR to 30.7%.
(Economic Costs and Bencfits Adjusted for Non-Traded Domestic Componcnts)
Sensitivity Analysis: Ciawi-Sukabumi Road Proiect

IRR
-
Proiect Scenario Parameter Base Value Variation Prolect
Economic Analysis
As Public Non-Toll Road Works cost 372 1 30 7%
8 generated traffic +lo% 28 9%
@ 25% of base traffic -10% 32 7%

Traffic Volume +lo% 32.5%

-10% 28 7%

Combined increased cost &

decreased volume -10% 27.1%

7.3 Impact Upon Economic Returns of BOT Toll Road Implementation

Construction and operation o f the Ciawi-Sukabumi highway as a toll facilitv has the
economic impact of increasing construction anci operation costs and reducing
utilization because of the impact of tolls upon ciiversioii. Construction as a toll road
means increased costs for toll collection anci administration facilities as m’cII a s
provision of toll plazas, additional lane allowance on entrance anci exit ramps,
lighting and fencing at toll collection areas. This has the effect of increasing the
adjusted economic cost from 372.1 billion Rp. to 415.6 billion. It has been assumed
that area and cost for land acquisition would remain the same as for a iion-toll
facility.

The effect of tolls upon traffic diversion has been estimated based o n the same
IHCM -based diversion curves used by RPT in the 1996 analvsis. Results are
consistent with diversion observed for the intercitv case studies examined in
Chapter 5. As traffic increases in the corridor, service levels (volume/capacity )
deteriorate, travel time iiicreases and congestion worsens. Starting with opening
diversion of 47% ancl 25% for Gol. I and Gol. I1 categories, by 2005, diversion has
increasecl to 55% and 27% respectively. By 2010, diversion has reached 63% and
38%,and reaches the maximum assumed diversion of 75% and 65% before 2020.

The base case economic return (adjusted for domestic costs) for the project organized
as a BOT toll road is 24.1%. Increasing costs by 10%would lower the EIRR to 22.8%;
a cost reduction of 10% would increase the EIRR to 25.7%. Varying traffic r~olume
by plus and minus 10% has the effect of raising aiicl lowering the EIRR bv 1.6
percentage points. A combined scenario of costs increased bv 10% ancl traffic
reduced by 10%would reduce the EIRR to 21.3 %.

54
A n increase in toll levels reduces traffic ancl consequently the volume of user
benefits, which in turn reduces potential generated traffic. A 25% increase in tolls
would decrease diversion from the existing road, to 35% and 16%shares for Class I
and Class I1 traffic after opening. By 2010, diversion shares would only l
x 55% and
27% respectively for these two traffic groups. This would have the impact of
reducing the economic return to below 20%.

Conversely, decreasing the tariff level would increase diversion and the number of
users benefiting from the new facilitv. Diversion has been estimated to amount to
58% and 35% after opening for thc two user groups. By 2005, these shares could
cliiiib to 65% and 39% of corridor traffic. The maximum diversion of 75% and 70%
would be reached before 2020. The ElWW in this case would be 26.6%. This lower
toll scenario was also combined with an assumed higher response b!, gmcr,i ted
traffic, the same assumption used by RPT, that is induced traffic climbing to 3(l% of
normal growth traffic bv 2005. The effect of this woulil be to increase the Ell<[< to
26.9%. The effect of these sensitivitv tests is summarixcd below

(Aifjirstcrf Ecoiioriiic Costs Cf Tiriir 51wiiigs)

-
IRR
Proiect Scenario Parameter Base Value Variation

Economic Analysis
As BOT Toll Road Works cost 415.6 24.1 %
+l o% 22.8%
-10% 25.7%

Traffic Volume +lo% 25.7%

-10% 22.5%

Combined increased cost &

decreased volume - 10% 21.3%

Toll Levels +25% 19.9%

-25% 26.6%
8 generated traffic increase
to 30% of base traffic by 2005 Toll Levels -25% 26.9%

7.4 Financial Evaluation of the Ciawi-Sukabumi Toll Road Project

Costs, toll tariff schedules, and the structuring of financing for the financial analysis
have been based on the estimates and proposals prepared by PT Bukaka Marga
Utama. On a current Rp. basis, total costs of the project include an assumed 8%
inflation during construction. These with financing costs amount to 808.5 billion Rp.
Excluding interest during construction, the estimated cost would be 668.5 billion Rp.
Basic construction works cost excluding supervision at 3%, toll equipment, VAT ,
ancl land acquisition has been estimated by Bukaka at 429.8 billion Rp.

55
For tlie current Rp. analvsis using nominal interest rates, Bukaka’s planned financial
structure was followed with a debt to equit\r ratio of 75 : 25 ancl debt k i n g split
between local and foreign sources on a 60 : 40 basis . Interest costs for local loans
have been assumed at 20%; offsliorc loans have been assumed to cost 11“0. a
convertible bond issue with an interest rate of 12% lias also been inclucicd with
planned issue in 2001. According to the Bukaka financial moclcl, tlie total l o c ~ loan
l
amount including the domestic portion of interest during construction (IDC) \VOLIILI
total 405.2 billion Rp. TIie offshore loan amount plus the foreign component of IDC
would total 235.5 billion R p .

Toll tariffs have been forecast to increase b y 16.5% every two years with a n opwing
\war 2000 level in Rp./kni o f 172, 257, and 34.1 for Gol. I, Gol. IIA a n d IIB
respectively. Otlier income from concessions lias been estimated at 5% of tariff
revenue. Revenue sliaring with joint venture partner PT Jasa Marga would st,irt in
2006 a t 5% of gross revenues. In w a r 2014, this increases to 14% and continucs until
the end of the concession period in 2024.

Using the Bukaka financial nioctel, several sensitivity tests were carried o u t to
determine the impact on the nominal financial rate o f return from variations in key
parameters; Applying the basic cost streams against traffic flows and tariffs
assunieci by RPT/Bukaka, that is with generated traffic starting at 25 % of normal
traffic in opening year and increasing to 30% hv 2005, the base financial internal rate
of return (FIRR) on tlie project would be 20.9% o \ ~ tlie
r concession period. With tlic
Bukaka financial structure as proposed, the return on equity would be 26.3‘L.

A 10% increase in the base construction cost of 429.8 billion current Rp. would
reduce the project FIRR to 20.2% ancl equity FIRR to 25%. Conversely, a 10%
decrease in the base construction cost would increase these FIRR’s to 21.7% and
27.9% respectively. Traffic volume changes by plus or minus 10% shift tlie FIRR’s
for tlie project to 22.1% or 19.7%. tlie iinpact on tlie return on equity would be to
raise the FIRR to 28% or reduce it to 24%. A combination of a 10% increase in
construction cost and a 10% reduction in traffic reduces the FIRR‘s to 17.6% and
22.8% respectively. If generated traffic developecl in a more conservative fashion,
assuming only 10% induced traffic on opening and building to 20 % of nornial by
2005, the nominal FIRR‘s are reduced to 17.6%and 20.3 % on total project and equitv
respectively.

An increase in interest rates for domestic and offshore loans by one percentage point
has tlie impact of reducing the nominal FIRR on tlie total project finance bv 0.1% and
on equity, by 0.5%. Reducing interest costs would have a symmetrical effect, and a
reduction in interest costs by 10% (2 percentage points) would improve tlie nominal
FIRR’s on project and equity to 21.1% and 27.3% respectively.

56

I
Other tests included increases in land costs; a 20% increase in this component (base
estimate 100.6 billion Rp.) would reduce project and equity FIRRs to 20.5% a n d 25.6
%, This scenario was also tested with an extension of the concession period by five
years to 2028. The impact would be to increase the nominal FIRR on the total
project finance by 0.6% to 21.5% and on equitv FIRR, the impact is an incrcase of
0.4% to 26.7%. The results of thcsc sensitivitv tests on the FIRR are summarized
be1ow:

Table 7.4.1
Financial Analvsis (Current Rp., Bukaka Financial Model, Nominal Interest Rates:

Proiect Scenario Parameter Base Value Variation

Financial Analysis as BOT Toll Road ( b l
Total Investment cost (c) 808.5 20.9% 26.3%
Construction 429.8
works cost
+lo% 20.2% 25.0%
- 10% 21.7% 27.9%

Traffic Volume -10% 19.7% 24.1%

Generated traffic 10% to 14% of base, 2000-2004

20% after 17.6% 20.3%

Combined increased cost &

decreased volume 19.0% 22.8%

Interest rate on
Long Term Debt +5% 20.8% 25.8%
(local & offshore) -5% 21.0% 26.8%
-10% 21.1% 27.3%

Increased land +lo% 20.7% 25.9%

cost
+20% 20.5% 25.6%

5-yr. extension to concession 21.8% 27.0%

with increased land cost +lo% 21.5% 26.7%

(b) Generated traffic assumed at 25% of diverted traffic in 2000,increasing to 30% by 2005
(c) includes construction cost, supervision, VAT, land acquisition, project overheads,
interest during construction, and bank service charges for loans and guarantees

Higirrra!/ Toil Stud!, - A D 6 I997

A further financial analysis was carried out using real interest rates and constant
1996 costs and revenues. Following ADB procedure, price contingencic.; and
interest during construction was excluded as were depreciation and debt scwice
cash flows. With a 1996 RQ. cost o f 555.8 billion RQ., the financial internal rCitcof
return i n real ternis woulcl be 14Y:,. A 10% increase in costs would reciucc this to
13%;combining this with a 10%reduction in traffic would vielci an FIRR of 11.7'%,.
Results of this test are shown belo\\.:

Table 7.4.2
Financial Analysis (Constant 1996 Bill. Rp.1
Para me tecr Variation
Investment Cost 553.8
+lO%
-10%

Traffic Volume -1 OM 12.9<%,

Combined 11.7%
increased cost &
decreased volume

7.5 Foreign Exchange Impacts

Foreign exchange costs are entailed tor all highway investments; toll roads bccause
of their requirement for specialized collection and processing equipment reyuirc
additional investment with a high foreign component as well. Even though raw
materials are sourced locallv, foreign exchange costs are involved in extraction <ind
processing. Machinery for construction is either imported as complete picccs or
assembled locally with many imported components. This is the case for both dump
trucks and toil collection equipment although the value-added locallv mav \'arv
between 15% to 40% of the cost. Similarly for fuel, cement, and steel which are
manufactured domestically, the plant and equipment to produce and deliver these
basic goods has a high foreign component.

I n the 1990 JICA Feasibility Study on the Bogor-Bandung Project estimates were
made of the foreign exchange share for raw materials and inputs in construction.
These ranged from 100% for asphalt which is imported entirely to 40% for electricity
which is produced with perhaps local coal at power generating stations with
imported generators. A number of processed inputs such as fuel, cement, crushed
stone, and steel were estimated to have 60% foreign currency shares in their
delivered cost.
For the purposes of tlie Ciawi-Sukabumi project analvsis, construction costs have
bee11 analyzed according to foreign : local shares cierivcci for tlie 1996 Feasibility
Study of the Cirebon-Batang Toll r o a d carried out for Ministry of Public ~ ’ o r kas i d
tlie World Bank (DHV Consultants in association with PT Biiia K a r ~ am d PT
Hasfarni Dian Konsultan). Based on these shares anci the breakdown of construction
costs by activity prepared b\. Bukaka, it has been determined that foreign cxchange
costs woulcl coiiipose 53% of construction costs. Including project supervision ‘iiid
o\lerlicacls, design anci final engineering, and purchase of toll collection equipiiicnt,
foreign exchange costs w o u l d increase to a 55% share of total expeiiciiturc.

Given tlie financing mix proposed b \ r Bukaka, a quarter of the interest cost during
construction would be foreign. Based on the 1990 JlCA estimates quoted a h w e ,
routine a n c i periodic maintenance costs have 50% a n d 76% foreign exchange sIi.ii-cs.
Operation of the toll road is more labour intensive and toreign exchange costs ‘ire
only 20% of tlie total of this item, but the replacement of toll equipiiient on the
anticipated 8, 5, and 3 \’ear c \ d c s is foreign exchange intensive, an estimated 82%
eveii with local asseinbl\?anci installation. Taking periodic pawiiieiit o\~crIavsand
toll equipment replaceiiient on a n annualized basis, operations costs cstimcitcd to
total 15.3 billion Rp./\?r. would have a 5.5 billion R p . or Xi‘%,foreign exchange
coiiiponent. The breakdown construction contract and cost item is sIio\\m in
Table 7.5.7.

59
Table 7.5.1

Toll Road Construction & Operation: Foreiqn Exchanqe Shares by Component

Foreiqn Bukaka cost estimate (bill. Rp.]

Excha ne
Contract Item % share (a) -
For
g
&

Mobilization 85% (b) 2.2 19

Clear & grub 65% 3.3 21
Earth wks 78% 106.5 82 9
Sodding 12% 2.6 03
Culverts 44% 13.2 58
Bridges 44% 91.1 40 3
Interchanges 44% 30 13 2
Grade separations 44% 27.1 11 9
Select subgrade 55% 1.9 10
Drainage 26% 7.6 20
Related constr. 51% (C) 1.6 08
Subbase 75% 16 12 0
Asphalt treated base course (ATB) 42% 45.9 19 3
Binder 8 wearing course 42% 17.9 75
Roadside furnishings 13% (d) 13.3 17
Throughway toll barriers 45% (C) 20.4 92
Parking 8 svc. areas 47% (C) 14.4 68
Branch office 45% (C) 1.7 08
Landscaping 8%
- -
1.3 -
01
Subtotal: construction 53% 418.0 219 6

Misc. & general 51% (C) 11.8 6.0

Supervision @ 3% of constr. cost 50% 12.9 6.4
Design & final engineering 70% (e) 11.4 8.0

Toll collection equipment 82% (f) 21.4 17.5

Project overheads 65% 18.9 12.3

Total: construction & supervision 55% 494.4 269.9

Interest during constr. (IDC) (g) 139.9

Local 75% 104.7
Foreign 25% 35.2 35.2

Maintenance & Operations/yr (9)

Routine roadway maintenance 50% (e) 2.9 1.4
Periodic Maintenance 76% (el 0.1 0.1
Operations 20% 9.8 2.0
Equipment replacement 82% 2.6 2.1
Total annual operations cost 36% 15.3 5.5
Table 7.5.1 (Continued)

(a) O h Shares from Feasibility Study of Cirebon-BatangToll Road, (Public WorksNVorld Bank)
DHV Consultants in assoc with PT Bina Karya & PT Hasfarm Dian Konsultan, May, 1996
except where noted below
(b) Includes general item from Bukaka construction cost estimate
(c) Based on JICA estimates for Jkta ORR project.
Phase I Report, October, 1988, Pacific Consultants International & Nippon Koei Co. Ltd
in ASSOCwith PT SEECONS,
PT Nusantara Survey. PT Geode Pataka Alam
(d) Includes misc. fencing, guard rails. markings, signals, signage
(e) Based on Chapter 8 estimates and allocations, Feasibility Study on
Bogor-BandungRoad Project. JICA. Final Report. November. 1990
(fj Based on discussions with Jasa Marga 8 Bukaka. with some local assembly of foreign components
(g) Bukaka Cash flow projection. May. 1997

Examining foreign shares of equity investment a n d debt borrowings, tlic mix as

mcntioned above woulcl be 60 : 40 cloiiiestic equity to offshore and the sanic ratio
for clomestic and off sliorc debt. According to the 1997 Executivc Summary
prepared by PT Bukaka Marga Utania describing the project, total investment would
amount to US$ 345 million or 828 billion Rp (@ 2400 Rp/$). Of this total, 528.8
million would represent offshore equity and $109 million in offshore loans. I t a s
assumed land acquisition proceeds in 1997 and construction starts on sclicciulc in
1998, the peak expenditure and inflow of foreign excliange would occur in 1999 with
104.8 billion Rp. This would be composed of 23.4 billion Rp. in equity and 81.1
billion in offshore debt. The peak outflow of foreign exchange occurs in the last year
of foreign debt and IDC yay-back. According to the Bukaka financial plan, this
would occur in 2006 and be composed of 82.1 billion Rp. of loans and 14.4 t>illion
IDC. The inflow and outflow of foreign funds is shown in Table 7.5.2.
Table 7.5.2,
Foreiqn Exchanqe Flows (bill. Rp.)

CREDIT POSITION ~
1996- 1997
- - 1998
- - 1999
~ _ 2000
_ 2001
_ _2002
- -2003
- -2004 2005 2006 2007 2008
Loan
Disbursmentf(Repayment)
Local Investment Loan 0.0 9.3 63.3 105.5 99.0 -46.0 -4.5 -9.1 -13.6 -15.9 -20.4 -80.9 -86.8
Local IDC 0.0 1.2 10.6 32.1 60.9 -24.2 -1.6 -3.2 -4.7 -5.5 -7.1 -58.4 0.0
Offshore Investment 0.0 6.2 42.2 70.3 66.0 -2.7 -3.6 -7.1 -35.7 -53.6 -82.1 0.0 0.0
Loan
Offshore IDC 0.0 0.4 3.7 11.0 20.0 -3.2 -0.6 -1.3 -6.3 -9.4 -14.4 0.0 0.0

Total Equity inflow 168 100% 0.6 27.8 35.2 58.6 32.4 13.2 0.0 0.0 0.0
Domestic 101 60% 0.4 16.7 21.1 35.2 19.4 7.9
Offshore 67 40% 0.3 11.1 14.1 23.4 13.0 5.3

Net Foreign Exchange 0.3 17.7 60.0 104.8 99.0 -0.6 -4.2 -8.4 -42.0 -63.0 -96.6 0.0 0.0
Flow:
(Excluding future dividend throw-off on foreign
equity)

Source: Bukaka Financial Model (May,

1997)

62
7.6. Impact of Different Toll Tariff Policies O n Financial Returns
In the previous section 7.3, the impact upon ecoiioiiiic benefits to the economy was
examiiiecl for higher and lower tariff levels. for the financial analysis, further tests
were conducted to cietcriiiiiic the impact on traffic and revenues from clianges in
.'. levels plus and minus 25% of the tariffs proposed bv Bukaka. Based on the
taiitf
TRlP Upctate/lHCM diversic)n c u r w s used b \ r RPT, a ciccrcasc in tariffs ti\, 25%
wouILi increase Categorv I traffic bv 20 to 25% and Catcgor\. I1 traffic b\. 'irouiici
15%. This would increase benefits to the economv, raising the ElRR to 25.7'1;, from
the base toll case return of 23.6%. From the private investor point of vie\\,, the
iiiipact woulci be negative with lower toll road revenues, ciecreasing the FlrZrZ for the
project to 18%from 21% . Return on equity would similarly decline to 21 '%from
26%.

Increasing tariffs by 25% would decrease diversion to the toll road, reducing tr'iftic
by 12 to 15 %, or perhaps as much as 20%, taking the more elastic part 0 1 the
diversion curve. This would have the iinpact of reducing the ElRR to the econoiiiv
to below 2076, but increase iiivestnr returns on the project to the order of 23%.
Return on equity would increase to the order of 30% with tliese higher tariff Ie\~c*Is.

Table 7.6.1
Tariff Adiustment Sensitivity Analysis: Ciawi-Sukabumi Toll Road Proiect

Financial Analysis as BOT Toll Road

Traffic response
Parameter E
c
J
Lu
t
J
Base Case 20.9% 26.3%

Toll rate decreased by 25% 25% 15% 17.9% 21.2%

Toll rale increased by 25% -15% -12% 23.5% 31.1%

-20% -20% 22.5% 29.2%

Given this situation where tlie iinpositioii of higher tariffs has tlie effect of tolling -
off traffic and reducing the benefits to the ecoiioiiiy of potential transport savings,
but perceived congestion relief and time savings remain high enough to still
produce increasing toll revenues, the main reference point for BOT tariff setting
becomes the financial return to the investor. Tariff-traffic sensitivity tests indicate
that the proposed tariffs are still substantially below tlie investor revenue-
maximizing level. If revenue maximizing became the tariff-setting criteria, ecoiioiiiic
benefits would be even more reduced with the EIRR to the economy falling far
below 20%.
From the investor’s point of view, the financial return to the project must AS a
mininium exceed the cost of capital. Based on tlie Bukaka financial plan with 60 :
40 domestic : offshore debt at current interest rates of 20% and 11%respectivelv, the
weighted nominal average cost of debt is i n the order of 17%when commitmc~ntfees
and charges are included. The forecast project rate of return i n current terms ‘it
20.9% vielcis a margin of 4 percentage p i n t s o\’er this. A slower cie\dopmcnt of
generated traffic, not reaching 20% of normal growth traffic until fivc vears <ittc>r
opening would reduce the FIRR in current tcrnis to 17.6%,or i n constant, real t<mns,
to the order of 11%.

Return on investor equity, the other benchmark for financial fcasibilit\r, i i i u s t a t

least be comparable with returns available in other sectors of comparable risk. A
margin of 5 percentage points over a 5 to 10-vcar average clomestic, nominal cost of
capital at 20-21% could be taken as a reasonable long term hurclle current rate of
return. The forecast nominal 26% return on ccpitv for tlie financial base c‘isc tits
within this zone of investor Iiurclle rates and is lower than what might bcb a c l i i c ~ \ u i
in other sectors.

A combination of adverse outcomes i n the project such as ci 10% cost cnw-ruii

combined with 10% less traffic than anticipated would .reduce real returns to the
project to 11.7%. Given the cost of capital ancl the risks on costs and revenues, the
tariff levels as proposecl represent a reasonable comproiiiise i n the trade-off between
economic returns to the national economy and returns to private investnicnt

8. REVIEW OF RISK ALLOCATION AND MITIGATION FRAMEWORK

8.1 Backzround Present Concession Framework

Tlie 1980 Law #13 establishecl tlie framework for designation and operation of toll
roads, standards of service, and set an upper limit on the toll tariffs to be charged.
Tlie 1990 Government Regulation
. # 8, Regarciinp
. Toll Roads assigned part o r all of
the authority to Jasa Marga to organize toll roads and provided for organizing i n
cooperation with other parties. Jasa Marga has accumulated almost 20 \rears o f
experience in the design, construction and operation of toll roads. From this Jasa
Marga has evolved procedures ancl a framework for entering into BTO ancl BOT
joint ventures with other public and private companies for financing improvements
and implementing new toll road ventures.

The latest program of toll road investments planned for initiation in Repelita VI
amounted to over 1100 km. of new construction. In 1995/96,9 toll roads composed
of 19 separate project sections were proposed for competitive bid and joint venture
organization. Another 16 sections (15 road projects) were under negotiation as part
of a direct appointment process. As of Ma\> 1997, 32 projects were active in these
two groups. Six projects of the total coulcl be considered to have advanced into
impleinentatioii or reached a signed Authorization Agreement.

As discussed in the Inception Report for this studv, project risk elements incluclc. not
only traffic flows, revenues, and construction cost over-runs, but also cicla\. costs.
In addition to revenue concerns and the lack of guarantees on toll tariff Ic~\~cIs ‘It
project inception, a major current concern atiecting cost and construction &h\. is
lancl acquisition. In this chapter, the over,ill cn\rironmcntot risk is ciiscusscJ JS \\rcll
a s these specific concerns. General issues arc summarized in Table 8.7
Table 8.1
Toll Road Risk Assessment, Coveraqe and Allocation

Exposure 8 Protection
Type of Possible Event Controling Parties Project Lender
Risk CompanyIConsortium

Commercial: Proiect related

Concept Sponsor Independent reviews

costs Sponsor Fixed price. certain-date turnkey

liquidated damages for
non-performance

Quality of design. thoroughness of site iiivestigations

Tested technology 8 contractors experence
Contracts for wnstr , supply

Project supplies Supplier supply contracts

Market
general none (thousands Independent surveys to verify demand forecasts
of potential users)

Competitors various Quality of service, rates,

Gov. franchise 8 planning agreements

Commitment Gov't Equity commitments. quality of fofward planning 8

commitment to Infrastructure Master Plan
Sponsor Equity comitmenl. technical
8 credit strength
Contractual commitment to
cost
over-runs 8 operating criteria

Contractorloperator Sponsorlcontractoi Experience 8 performance, financial strength

Contractual arrangements:ie.. penalties. bonuses,
transparency
arms-length from sponsors, iiisurance

Pre-completion Risks
Change orders GovernmentISponsor Exposure depends on Contract
provisions
Delays due to
Proj. Company fault Sponsor Exposed partly exposed (insurance)
Force Majeure event none partly exposed (insurance)
Delayed site GovernmenVSponsor Exposedlbuy-out provision
possession
Contractor fault Contractor Liquidated damages
(% of contract price)
Inefficient Sponsorlcontractor exposed
implementation

Failure to complete Sponsorlcontractoi Foreclosure on assets JM pays debt service

liquidated damages from revenues, but does not
assume loans
Table 8.1 (continued)
ExDOSUre & Protection
Type of Possible Event Controling Parties Project Lender
Risk CompanylConsortium

Post-Completion Risks
Force Majeure
due to Gov't Government Buy-out mechanism
other none insurance

Technical defects Contractorisuppliers Performance guarantees, liquidated damages

Operationslmaint. contractorlsponsor exposedlcontractor damages

Commercial: Economic environment

Currencylinterest rate risk Government (partial) Hedging. sponsor guarantees
use of local financing
Gov't agreement to link tariff to debt service costs

Inflation Government (partial) Gov't agreement linking tariffs to inflation index

Country risk
i) Foreign exchange Government (partial) Guarantee of exch. availability
Revenues paid to
offshore escrow account
ii) Political Government (partial) Insurance. buy-out clauses
or none

iii) Legal Risk Government International arbitration. contractual law

iv) Force Majeure none insurance

Non-commercial: project specific

Regulatory Government Concession agreement conditions

Expropriation Government Previous record. concession terms

inclusion of local lenders, partners 8 sponsors, insurance

Stcte Enterprise Obligations Government Contracts, Guarantees of performance

Note: Because of the structure of the Joint Venture, Jasa Marga assumes Several roles, including that of sponsor.
State client and ultimate operator on behalf of the Government at the end of the concession period.

8.2 Current Framework for Allocation of Risk

The parties in the allocation of risk include the Government, Jasa Marga, atid the
project sponsors. The key concern of the private sponsors is the expected return on
their equity iiwestineiit. Depending upon the riskiness of the investment and the
constraints of debt, expected rates of return on total project cost vary between 17%
and 30%.1' Higher risks inean higher investor expectations 011 the rate of return;
reduction of risk and allocation of risk to those parties best able to control and
manage these risks is the major determinant in reducing the risk preiniums
demanciecl by financing entities. This then will translate into the required minimuin
toll tariff for financial feasibilit\r- and ultimatcl\, the economic returns of the project
to tlie national econoinv.

As discussed previouslv, toll roads are niorc risky \'cnturcs than soine othcr t!'iws of
BOT invcstments given the licav!' front-end invcstiiient anci early years of nc~pitivc
cash flow. Financial coinmitnwnts arc maclc in advance of traffic anci re\'ciiuc
streams being assured. Often, only far into the concession period will traffic huilcl
to sufficiently cover operating, maiiitcnancc ancl debt service costs. Achicvinf riii
adequate return and recouping the investment may take as long as 15 tci 20 \ ' ~ ' ~ i r s .
During this time span man\' unforeseen cvents and several series of economic ~ n c i
political cycles can occur.

Major,,
elements o f risk in BOT toll road projects include project clevelopnent, land
acquisition, construction cost, rcvcnues (as a function of traffic volunie and toll
tariff), and fund repatriation (currency fluctuation, exchange controls, off-shore debt
service, ancl repatriation of profit.) The broader aspects of political risk is t o a Lirge
degree out of the control of the parties involved in tlie agreeinent. These are not
discussed here beyond noting that Inclonesia's record of prudent fiscal iiianagcment
coupled with strong economic growth provides a inore reassuring environment than
many other countries competing for off-shore BOT funds.

In the most recent generation of planned BOT projects, tlie tendering process was
opened to domestic and international investors. Standard bid forms anci a svstematic
review and evaluation process were useci to increase confidence that the selection
process was conducted in a fair and even-handed manner. As mentioned above,
between December 1995 and June 1996, bids on 19 sections totaling 767 kilonieters
were due on six different dates to spread the load of preparation dnd evaluation.

Although the amount of data and the depth of feasibility and design information
varied widely among projects, Jasa Marga made the bidders responsible for analysis
and verification in the preparation of their bid. More extensive preparatory
analysis of design alternatives and studies into construction costs and traffic would
have made the bidding and evaluation process smoother and easier. Costs
estimates and bid elements varied widely as a result. Selection of winning bidders
was based on a number of criteria. These included the earliest opening \-ear,
construction cost, concession period, proposed toll tariffs, and discounted revenue
sharing to Jasa Marga. The weighting of engineering criteria in the selection in

Indonesia Road Sector Shtdy, March, 1996, World Bank

terms of adequacy of drainage, pavement strength or structural design is not clear,
but prequalification of bidders was intcnclecl to establish the competence o f the firms
involved.

The evaluation ancl extensive rounds of negotiation have taken much effort and time
on the part of both Jasa Marga staff and investors with so manv packages to be
consicierecl. While part of the elapsed 18 months since tlic start of this procc*sscan
be attributed to biclclers putting together consortia and financing packag:css, the
ilela!fs in reaching the final stage of signing Authorization Agreements h a w bccn
considerable.

Stanclard elements of Force Majcurc arc adclressecl i i i tlie proforma tcnilcr

Jocuiiicnts and Authorization Agreements. As noted below and in contrast to sonic
governments which self-insure at a lower cost, in tlie most recent generation of open
bidding, investors were asked to carry their own insurance on such risks. 111 the
event of default during project preparation o r after construction by either p t r t \ , in
the joint venture, Jasa Marga undertakes to make loan payments out 0 1 toll
revenues, but would not assume loans or debt obligations.

8.2.1 Project Preparation

The 1990 Government R e d a t i o n #8 RegardinK Toll Roads (Article 41 & 42)
stipulated that prefeasibilitv ancl land clearance expenses were to be borne ti\, tlic
Government. Feasibilitv, engineering, and final design expenses were to lw lxirnc
by Jasa Marga or by Jasa Marga in cooperation with the investor. Since construction
cost risks are borne bv tlie investor (and the contractor in case of fixed sun1
contracts), due diligence on the part of the investor and financial backers wo~ilcl
dictate that tliorough back-up investigations be carried out to assure themselves of
the best accuracy of construction cost and design. It is also in the interest of ]asa
Marga and the Government through the Ministry of Public Works to ensure that
designs are sound and evaluated by 3rd party expertise. This should be clone in a
manner as comparably thorough as should construction ancl maintenalice
supervision on the part of the ultimate owner of the asset.

Quality control provisions in Authorization Agreements for this latest, competitive

round will hopefully be more rigorously applied than in the preceding generation of
direct appointment /non-tender awards. Diligence in monitoring qualitv and
supervision of construction by Jasa Marga is in their best interest, both as joint
venture partner ancl ultimate operator of the asset. The Government as owner must
also take responsibility for ensuring qualitv.

8.2.2 Land Acquisition

Although as mentioned above, Government Regulation #8 stipulated that land
acquisition was the responsibility of Government, right of way is no longer
provided by Government because of budget constraints.'j While the Government
initially provided land for toll road projects, this policy changed for the 1995
generation of tenders with investors being given the responsibility for funding land
acquisition up to 15% to 20 % of construction cost. This policy was again altered in
February, 1997 to extend investor rcsponsibilitv beyond this liiiiit. While prescnt
policy is not clear, there apparentlv exists the possibilitv that over-runs Lwvond tlw
20% benchmark mav bc shared i n some fashion with the state enterprise pdrtnc'r,
Jasa Marga.

Since the land purchase and negotiations with property owners is carried out h\. the
Governmentl(,, without any control or participation by the in\wtor be\wnd th,it of
observer status, it woulcl be preferable that costs were borne bv the Govc.rnmcut.
Ideally, following the cievelopineiit of a toll road/public road masterplan the
Ministrv of Public Works, rights of w a y would be mapped o u t , and propcrtv
acquired well in advance of bidding and speculative plienomen,i. Gi\sen
Goveriiment eminent ciomain powers, values could be frozen and prop>rtics
purcliasecl based on current tax assessment values in these corridors. Instcad,
property acquisition in many cases does not start until bids are finalized o r
Agreements of Association signed. Insiders have then had an opportunity to
purchase tracts a i d the process of land price escalation is alreadv well ad\-;iiiccd.
The investor and the partner, Jasa Marga, hence incur an open-ended liabilitv which
in turn affects financing arrangements and the rate of return calculus. This has the
potential of raising contingencies of such a magnitude as to make the original bid

l 5 Article 41, Gov. Regulation No. 8/1990, land procurenlent is tlie obligation ot the governinail , i n J
ovonersliip remains with tlie government; hence tlie investor cannot use this asset as coll~itwil. I h c
to government limited funds, land acquisition costs provided by the investor and acquisition wsts
including accrued interest during tlie period are part of tlie investment costs borne by the iiivcstor.
The investor has to make available funds for land acquisition, but may appoint a represcntativc to
observe the acquisition process.
For the Sukabumi-Cianjur project, tlie original tender document provisions and conrlitions (Ikwisions
of September, 1995) stipulated tliat an irrevocable bank guarantee of 19% of construction costs be
submitted to Jasa Marga. In these conditioiis, Jasa Marga would have been responsible tor nny
shortfall of funds if land acquisition exceeded the estimate. If tlie land cost were to have been less,
the investor would only be responsible for funding tlie actual cost of land acquisition, b u t the
difference would be paid to Jasa Marga out of net revenue over some part of tlie operation period

A delay in tlie delivery of all or part of tlie land would be an event (as would force majeure) allowing
the investor to propose an extension of the Authorization Period.

In the case of proposed toll roads which are located in less developed arras, the amount of bank
guarantee to cover land acquisition cost appears flexible, in the case of the Medan Binjai road, 16% of
capital cost (from a previous 24%?). (March 13, 1996 addendum to Tender Document). this ,1lso
appears to have beeti the case in tlie Seiiiarang-Solo tender as well.

l6 KEPPRESS #55, 1993 sets out land acquisition procedures and Government committee structure.
obsolete, necessitating fresh negotiation cycles, anci prolonging the gestation period
for the investment.

Besides the optimum approach of resolving land acquisition costs prior to award of
concession, other alternatives are available. One policy pursued bv Chile is the
sharing of land acquisition cost over-runs betwwn Government and concessioiiiiire
o n a respective 80%/20% basis. If land acquisition and clearance become ‘1 C-JL~SC of
dela!., tlie Chilean Govcrnnwnt extends the concession period tor a n cqui\’,ilcmt,
compensatory period. It appears t l u t this pattern has been adopted to soinct ~ l c y , r c c ~
b\v Jasa Marga.

Another alternative w o u l ~ l be for Jasa Mare1 to pay tor land purcliasc from
refinancing proceeds of their existing toll road s\’stem and count this towarcis equitv
contribution and increased revenuc-sharing arrangements. This however C ~ L I I C I
complicate public share offering plans now on-going. Another option, einplo\td in
the United States, namely California and in states around the District of Columtiia,
has been to gain participatory contributions for toll road land requirements t r c i m
adjacent property clevelopers. This sometimes takes the form of Land clcinatioii. In
other worcls, those who benefit from high capacity inotorwavs ancl interchanges ,ilsn
help to py.

I n the event of delay in cielivcrv of all o r p a r t of the land, as well as with delays
caused by change in the scope o f work or events of force majeure, extension of the
Authorization (concession) period mav be proposed bv the investor. Such
extensions appear to be negotiable items as are tlie hanctliiig of land cost over-runs.
However, aclditional extensions on a 25 o r 30 year concession period offer a soft and
distant form of compensation for front-end delay costs. Normally with
circumstances more uncler control of the Government or implementing agency,
these delay costs would be borne by tliem iiicludiiig costs of idle equipment ancl
personnel as well as interest on loans during construction for the idle period.

8.2.3 Construction Cost Risk

The risks in construction costs are mainly those of over-runs anci are borne by the
investor and/or the contractor as part of a fixed sum contract. These can be reduced
through adequate prefeasibilitv, pre-design and competent final design to reduce to
the minimum cliange or variation orders. Execution of works by qualified
contractors under good supervision is the principle security available.

Unforeseen geotechnical conditions and other risks such as weather which are
beyond contractor control are covered bv standard clauses or stand-by financial
measures. These as well as insurance premiums against force majeure have been
made the responsibility of the investor as standard contract elements ancl included
as an insurance expense item. Since premiums for such insurance tend to be large,
usuallv it is cheaper however for tlie Government to self-insure against tliesc risks.
With the BOT investor insuring against these risks, costs are higher and m u s t
eveiituallv be reflected in the toll tariff. Delays wliich occur froin Government
action or inaction should also be covei.cd with adequate compensation for such
breaches. For this new generation of projects, these delays as well a s costs of
changes in the scope of work which a r c instructed bv the Government are t o be
reimbursed out o f reductions in toll revenue sharing to Jasa Marga and/or
extensions in the concession period.

8.2.4 Traffic Estimates

TIie bidder is responsible for traffic estimates, including base year traffic and
subsequent growth rates. Prelimiiiar\, feasibilitv studies have been carried out to
soine degree on most corridors b\' Biiia Marga. Generallv inore coiiiprelieiisivc a n d
detailed preparatorv feasibility studies bv sponsoring agencies have led to better
quality a i d inore competitive bidding. It remains, however, the proper
respoiisibility of the bidder to carr!. out their ow'n analysis of traffic potential and
diversion. While some countries such a s Malaizsia and Mexico in the earlier Ic5,iriiing
stages of BOT toll roads offered traffic guarantecs, later generations of 601' projects
have largely left this risk to the investor to be addressed as part of d u e diligence
procedures.

Provided the investor carries out their own traffic analysis in a competent manner,
the major risk which remains under government control is the future provision of
coiiipeting high stanciarci roads. As raised previously, prograinining r o a d
investment within the context of an overall masterplan which is open and available
to all bidders is the most cost-effective approach for attracting competitive bids and
increasing security of traffic attraction. Formulation of a masterplan would also
encourage inore coherent land use plaiining which in turn affects future traffic
genera tioii.

8.2.5 Toll Tariff Risk

This element of risk is discussed in tlie following Chapter , but it is important to
address this within the context of overall risk management aiicl allocation. Toll
tariff risk has two elements; these are comprised of firstly, the initial opening tariff
and secondly the mechaiiisin and policy for future adjustments. The concern of
investors is that adjustments will not keep pace with debt service, operating,
maintenance, and periodic mainteiiance costs, particularly in an atmosphere of
escalating prices, deteriorating exchange rates, and significant portions of off-shore
debt to repay.

Under current practice, the opening toll tariff is not officially accepted until the road
is essentially complete. Jasa Marga can assure best efforts in processing the
proposed rate through officials channels up to the President's desk, but until tlie
KEPPRES is signed, tlie risk is with the investor. Tlie contract with Jasa Marga in the
Authorization Agreement contains a proposed formula for bi-annual tariff esc'lla tion
based on tlie coiisumer price index. However, risk remains that adjustments mav
not be approved in a future aclministration. This uncertainty perhaps reduced the
number of potential biclilers, and for those accepting the risk, led to higher
premiums to cover tlic perceived risk.

8.2.6 Foreign Exchange Risk

Tlie issue of alleviating investor risk for inflation and currencv depreciation through
incorporating escalation factors in tariff tormulcis is aildrcssed in the next c1iaptc.r 011
toll tariffs. Mentioncd here 'ire the broader p l i c r implications for lnctoncsi,i in
private funding of infrastructure.

Because Indonesia lias maintainecl an open capital account for man\' . .vcars,
guarantees on foreign exchange transfers have not been an issue, and currencv
controls or limits on repatriation of profits have not been a concern. With a carefully
followed "managed float" exchange rate policr, Bank Indonesia lias Iiclcl a fairlv
steady "real" excliange rate with respect to the countrv's trading partners since tlic
devaluation in 1986. Interests rates for domestic loans have eased in recent \ ~ a r s ,
and domestic inflation measured bv tlie Consumer Price Index lias also declined
from more than 8 % in 1995 to 6.5 % in 1996. Tlie Minister of Finance announccct in
March of this \Tear that the inflation outlook for the current fiscal year was in the
order of 6%. However, maintaining high growth, low inflation performance is
difficult, even without the implications of increased integration with external capital
markets.

With the high cost of domestic financing and the practical and policy limits of
domestic funding capacity for infrastructure, at least part of tlie financing package
for the toll roads currently under investment process would come from off-shore.
External sourcing of funds also has limits, both for sliort-to-medium term flows
inward, and the longer term debt service claims outward. The current account and
balance of payments will continue to be a constraint on foreign debt obligations for
the foreseeable future. It is difficult to see how the entire portfolio of projects
currently in investment process, amounting to more than US$3 billion, can be
financed during the next five years and still fit - no matter how split- within tlie
Central Banks respective domestic and foreign infrastructure funding envelopes.
Even with a n implied 25% to 50% to be sourced from off-shore, the chore of
sterilizing the inflationary impact of funds injected into tlie economy during the
coiistruction phase will be significant. A later, more worrisome task will follow with
the demand for foreign exchange debt service as grace periods for off-sliore debt
expire.

73
8.3 Revulatorv Context
The necessity to develop and update the master plan for the public and toll r m c l
network has been mentioned earlier. The long term management polic\r anci
regulator\! overview of toll roads will also require tlie cievelopnient of polic\. for
longer term administration of tariffs. As mentioned i n the tarift chapter, as toll
roads mature and investors enter their periods of "enjoyment," public pressure
could be expected to build to cap tariff rates if tlicrc is a perception of "cxccss"
profits. This is a long term risk to the investor who expects rewards for bearing a
negative cash flow through tlie early . .\ ~ a r sof the investment, but it would be
prudent to address this early in the process. This aspect of monitoring and c>vcrsight
is acliiressecl in the following chapter on tariff setting anci adjustment.

Measures to improve the agreement framework and rcgulatory structure to r e n i o \ ~

elements of unccrtaintu which are within Governmcnt control include the follo\<ing:

i) Make a consistent policy which limits investor exposure on h n c i

acquisition cost overruns;
i i) Strengthen tlie long term planning function for ~levelopmcnt of
integratecl urban and intercity toll road systems, and make av,iilable
updated, compreliensive masterplans to guide land use and land
transport clevelopmciit;
iii) Incorporate in the tariff adjustment tormula a mechanism wliicli reflccts
cost increases and provides some devaluation protection for coverage of
off-shore debt service during the cie\dopmental pliase of investmcnt.
iv) Simplify the tariff adjustment mechanism anci delegate power to adjust
rates to the Minister of Public Works (See Chapter 9 on Tariff Setting &
Adjustment).
9 Strengthen the role of the Ministry of Public Works (Bina Marga) as the
Government body charged with balanced network clevelopment
including the following aspects:
a) Updating of tlie masterplan for toll road and alternative non-toll
road networks to ensure balanced network development in
keeping with the spirit of the 1980 Law On Roads and
Government Regulation ## 8, 1990 (Article I G.2, r e g d i i i g b n l m m f
de71elop,llrllt)
b) Strengthen implementation of the masterplan to ensure that the
service balance between toll road and public lion-toll alternit'ives
is established and maintained.

The inter-urban toll road network should serve longer distance, through traffic, but
the primary public highway system and local access network should continue to be
maintained according to Bina Marga's current policy of maintenance and betterment
programming. Adherence to the inasterplan should ensure that neither a service
inonopoly is allowed to clevelop on the part of the toll system, or con\~ersely,
parallel, high-capacity non-toll inotorwavs precipitously undercut the toll customcr
base.

9. TOLL TARIFF SETTING AND ADIUSTMENT MECHANISMS

9.1 Background
The nature of toll road investment makes this particular form of BOT vcnturcs iiiorc
difficult to structure ancl finance than other forins of infrastructure o r utilitics. Toll
road investment costs are concentrated a t the beginning of the project, pa\'-hack
perioLis are generally long, and revenues arc\ nwrc ciifficuit to predict t I m i for ot~icr
types o f BOT investments. Unlike a power plant serving one bulk customer, toll
road usage is tlie product of thousands of individual customer decisions with tlie
attendant wide varietv o f trip purpiscs and willingness to pav. For these reasons,
re\~eiiuc-maxiiiiiziiigtariffs and higher vielcls on equity are sought by invcsstors;
rates of return tend to be higher than for other types of BOT projects in cwder to
attract investors.

However, as pointed out by Bina Marga officials, payment of a road toll is 1' cash
transaction immediately felt in the pocket. I n contrast, telephone and electrical bills
have the perceptual aclvantage of a Lielaved impact on the consuiner. I -IC~IICC
governments and tlieir constituencies must be expected to continue to exert a n
interest in the tariff setting and acljustnient process.

As also discussed in Chapter 9, tariff setting and adjustment mechanisms are also
tlie first concern of concessionaires ancl their bankers. The setting of tariffs is
intimately connected with the structure ancl cost of financing. The cost preiniunis
set on risk or offsetting measures, mitigation ancl comfort to lenders become
ilicorporated in the tariff and rate of return calculus.

The cornerstone of a sound BOT policy is tlie establishment of a predictable svstein

for setting and adjusting toll tariffs. The most successful approach to this system is
the incorporation in the bidding procedure and authorization agreement of a
formula which resolves uncertainty on tariffs and hence increases investor
confidence on revenues. This resolution can be met through a tariff adjustment
mechanism with cost indexing.

75
9.2 Current Svstem
Uncler tlie original 1980 legislation and later implementing regulations,I: the
ciesignation of a road section as a toll road and toll tariffs by type of vehicle are
cieterminecl by the President a t the proposal of the Minister of Public Works. A toll
road can only be officiallv designated when it is cssentiall\~complete and tolls
cannot be stipulated until this formality is carried out. As presently aciniinistcwd,
this process requires a new Presidential Dccrcc (KEPPRES) everv tinac a rate is tiled,
including opening of new interchanges on existing toll roads where 110 clicln:;c in
vchicle/km rates inav be entailed.

A lengthy and bureaucratic process is entailed to niove the tariff setting and
adjustment proposal up to the President’s desk. Starting with the investor
submitting the proposal to Jasa Mar91 for review, it is then passed to the Ministn- o f
Public Works. Here it is reviewed in detail a3ain bv the Subdirectorat Jalan To1 for
conformance with BKOK guiilelincs. The proposal is then submitted to the Ministcr
of Public Works for approval. It then is passed up to the Cabinet Sccrct‘iri‘it for
review, and if approved, will be submitted to tlic President for decision.

A detailed comparison of user costs for the toll and non-toll alternative is carried
out bv both Jasa Marga ancl Bina Marga. This comparison takes into account
vehicle operating cost savings ( including time) or BKBOK over the entire length of
the particular toll road i n question and checks the level of tolls against toll road user
savings.

9.3 Legal Framework and Philosophv

When Law Number 13 of 1980 On Roads was enacted, one of the purposes o t the
legislation was to define the role of tlie road network system in serving balanced
interregional and national development. The Law defined the Iiicrarchv ancl
classification of roads according to their purpose and provided a framework in
which long term planning, construction ancl maintenance of roads woulcl be carried
out. The Law also provided for the organization of toll roads ancl the delegation of
competence to a State Corporation of Toll Roads (Jasa Marga).

At the time of tlie enactment of this law, earnings from oil exports were higher and
soft loans froin multilateral and country assistance programmes were more
plentiful. In more recent years the many competing demands for State funds for
development and the decrease in multilateral funds available has increased the need
to attract private investment for toll roads. From this has followed a policv for BOT
development and joint operation of toll roads. In Government Regulation Number
8, 1990 Regarding Toll Roads, Article 38 provided for Jasa Marga organizing toll

li “Law # 13 of 1980 On Roads” governing the administration of the road network system , and
Government Regulation #8,1990 Regarding Toll Roads.
roads in cooperation with clomestic ancl foreign companies. Devclopleot,
maintenance ancl operation however still remain the ultimate responsibility o t the
State Agency, Jasa Marga.

9.4 BOT Toll Roads and loint Ventures

As meiitionecl above, there h a w been several stages in the clcvclopmcnt ot toll rocid
financing in Inclonesia. During the initial stage, (197s - 19S3), toll r o d s IVLW
financed by state funds and foreign loans. This carried into thc second st‘>gc,( l W 3 -
1990) with “two-step” loans supplcnicntcJ ti\, bonds issued by J‘lsa M a r g i . l h i , tirst
mechanism involved soft foreign loans to the Government o f Indonesia being rc4ent
to Jasa Marga. Although foreign and multilateral loans for Jasa M a r p projects
continued into the 1990‘s, starting in 19x7, Jasa Marga has been building toll rocids
in co-operation anci joint venture with the private sector b y using BTO a n d 130T
svsteins. This has since cievclopcd from direct appointment of in\utors ti\. thc
Minister of Public Works to a competitive biciciing process started in April, I‘FJ5. In
the more recent invitations to tender anci subsequent joint venture ‘igrccmcnts,
modified terms have incluclcci such changes as revenue-sharing, anci the financing of
land acquisition by the investor as part of the total investment cost.

Tender documents issued by Jasa Marpi in the 1995 bidding rounds stated that
initial toll tariffs must be proposed by investor with reference to otlicr toll roads.
Guidelines also included the maximum tariff limit of 70% of BKBOK. The g<meral
guidelines suggested that for classifying tariffs bv vehicle t\rpe, Categon I I A
(medium, 2-axle vehicles) tariffs should be about 1.5 times Categon, 1 (light
vehicles) and Category IIB (heavy, >2axles), about 2 times Categorv 1.

Tender documents also provided that potential investors could propose a tariff
adjustment formula incorporating consumer price indexing, with tariff acijustmcnt
applications allowed every two or three years. Jasa Marga as joint venture partner
as well as the authorized State agency for organizing toll roads woul~lundertake to
process tlie tariff applications through the government regulatorv a d appro\:al
process, but cannot promise beyond its best efforts when or if applications might be
approved.

The investor proposal for opening tariffs on a new toll road can onlv be processed
when a “Development Report” on tlie road is submitted, the readiness of tlie toll
road has been checked by Jasa Marga, and tlie facility is determined to be
operational. After review by various departments within Jasa Marga involving
Research, Planning, Investment, Collection Management, Implementation, and Legal
Departments, a concept memorandum is forwarded to tlie Minister of Public
Works. This is reviewed internally with Bina Marga, more specifically, the

77
Subclirectorate of Toll Roads. This then follows the same process as proposdls tor
tariff adjustment to the Presiclent's desk for a p p r ~ ~ a l .

Hence, beyond pledges of best efforts, toll road BOT projects have no biiiding
agreement on tariffs until the facilih, has been constructed and the PrcsiJcnt has
approved the application. Similarly, there is no guarantee that future applic.,ltions
for tariff adjustments \\,ill be apprcrvcd on a sirstematic o r regular basis. As
discussed previouslv financing for projects must be arranged in A tramc~cvcrrhu,hich
cannot guarantee a tariff which will necessarilv cover debt service costs. 'I'his
element o f uncertaint\r and risk must then be included as a n implicit preniium c r n
the cost of financing, particularlv for offshore funds.

9.5 User Savings and Tariff Setting

The concept of user savings embodiccl in the 1980 Law On Roads prcn,icics a
reference point and an upper bound for toll tariff-setting. If savings which uscrs
gain from traveling on a toll road are used tor a guide i n the cleterniin,~tionof toll
rate, it is desirable to develop a simple VOC model such that the data collection ,Ind
further updating can be done without too much time and resources. Hotvetx~tlic
model sliould be extensi\.el\r calibrated based o n field data, which for VOC means
not every cost component can be fully calibrated quickly.

To study VOC on toll roads, PT. Jasa Marga, has been using a inoclel developed in
1979 by Pacific Consultant International as a guide in setting toll rates. The model
was developed as part of a feasibilitv stud\, carried o u t for intra-urban toll r o d s i n
Jakarta. In the model each VOC component is assumed to be best explained b\. the
average running speed of the vehicles. Irregularity of road surfaces, geometrics, ,ind
traffic conditions are not explicitly taken into account in factoring cost changes.

To provide a more appropriate model for calculating vehicle operating cost savings,
Jasa Marga commissioned a consultant to review the PCI model and a modified
model of VOC for toll roads has been releasedts. The fuel consumption cost
component is empirically cievelopeci using representative vehicles cominonl!. using
the toll roads by taking into account traffic, road geometrics, and roughness effects.
Other cost components are partially adopted from an extensive review made on
existing VOC models. The comparison between the two models has been discussed
previously in Chapter 4.

Another aspect of user savings is the value of time for passengers and private
vehicle drivers. There is a lack of empirical value of time studies conducted for
Indonesia conditions. As discussed in Chapter 4, past feasibility studies on toll

PT. Jasa Marga (1996). The calculation of vehicle operating cost. Final Report, Foulidation tor
Research & Industrial Affiliation - ITB. (in Bahasa Indonesia)

Highrun!, Toll St!t+ - AD8 1997 78

roads have assumed a wide range for values of time, derived frc)lii different
assumptioiis. Techniques such as stated preference are now developed and accepted
for deriving a value of time. Unfortunately, there has not been an\' empirical \f,llue
of time derived for Indonesian conditions.

Knowledge ancl experience on VOC modeling has been developing in both lass
Marga and Biiia Marga, but the specification of models used by Bina Marga such as
the RUCMI" for maintenance maiiagcmcnt planning and programming S C ~ Y '
different purpose. A standard model should be adopted for use in evaluating tdritl
proposals. This sliould allow simplificci application for both urban and intcr-urban
road analvsis. The new mocicl ilcvclopcd for Jasa Marga would satish. this
purpose. Similarly a standard approach for the valuation of user time needs t o be
developed including an adjustment for regional coiiciitions. The time value of R p
3,281 /hr/veli for Categor\r I vehicles developed for highway life cvcle anrllvsit; in
the Indonesian Highwa\r Capacitv Manual (IHCM) appears too low in relation to to
wage levels. The income approach as used for the Jakarta Intra-Urban Toll R o d
stud\, (Northern Extention of the Southwest Arc) yields a value in the orcicr of Rp
hoo(ito 8000,wliicli may be more reasonable.

Within the philosoph\r of Law Number 13, it is necessary to develop a sct of

aclmiiiistrative guicielines which will further interpret user savings for urban a n d
interurban toll roads. Increasingly as traffic builds in certain corridors, congestion
atid clelay costs will preclominate in the user savings function against which tariffs
are compared. Particularly for urban toll road systems, congestion cliarging by
time of clay should be introduced with lower rates for night and off-peak users.
Lower tariffs for night-time use of interurban toll roads slioulci also be considered
for trucks and buses which are prcsentlv tolled-off several stretches of the toll road
system. This wsould encourage better utilization of the toll road assets and reduce
heavy traffic on the non-tolled public network.

Consideration shoulcl also be given to expanding tariff categories and revision of

tariffs for Category IIA and IIB vehicles. Weigh-in-motion measuring ancl charging
equipment should be installed in the future with tariff schedules which reflect axle
loads. This should encourage more trucks and large buses with tandem axles
and/or empty back-hauls, but penalize over-loaded trucks, particularlv two-axle
units.

There are a number of implications of tliese tariff proposals for incentive off-peak
tariffs and peak-premium tariffs (congestion charging). One implication will be to
make toll collection more complex and necessitate updating of tariff display,

Hoff & Overgaard (1992). Road user cost model. Second technical advisory services on pldniiing
and prograinming to Directorate General of Highways, Government of Indonesia

Ht~Imw,To11 Stt(dy - ADii 199; 79

collection, ancl monitoring equipment. Another aspect is the need to make the user-
savings guicleline for tariff setting reflect relative travel costs by time period.
However, under the present framework ancl procedures for determining tariff
adjustments, this woulcl b c ~ o m can even more complcx and time consuming
process. With over a thousand kilometers of toll roads in operation anci scvclal
multiples of the present number of toll sections anci interchanges to be in opc~rclticm
bv 2001, the meclianism of approval will have to be simplified. The burdcw o f
decisions over tariff setting and adjustments will also havc to be ilelcgatcd in some
maniicr.

9.6 Tariff Policv Precedents and Adjustment Mechanisms

As in sc\wral countries, Incloncsia has adopted automatic tariff adjustment formulas
for other utilities such as electricity and telecoiiiiiiuiiicatioii~. These formulas take
into account changes in the cost of inputs and the weight of local and foreign debt
servicc costs. For instance, factors in the rate adjustment for the electricit\' tariff
include changes in the cost of fuel, cost of wholesale power purchasc, ilomcstic
inflation, anci the USS to Rp exchange rate. These factors arc weighted i n the
tormula bv coefficients reflecting the annual share in component cost.?" For thc
telecomiiiuiiicatioiis tariff formula, a weighted formula incorporating factors for
line and message unit (long distance and local) costs is used. For both power and
telecoiiimuiiicatioiis tariffs, approval authoritv rests with the responsible minister.

The first BOT project in Indonesia incorporating a cioiiiestic inflation anci exchange
rate adjustment niechanisni for automatic tariff adjustment was the Paiton Power
Plant. A similar type of formula was proposed for the Cikainpek-Padalaranf 'Toll
Road project by the concessionaire. The toll tariff adjustment proposed w a s a
function of the domestic inflation rate, the depreciation in the Rp to USS exchange
rate applied to the foreign equity and loan portion of total investment, and a delay
factor expressed in terms of months elapsed since the last tariff adjustment.?l This
proposed formula was proposed in the bid clocumcnt, but has never been accepted
bv Jasa Marga or the Government.

Such a n adjustment formula linking tariff increases to exchange rate fluctuations

anci the share of foreign investment (outstanding loans, bonds, and equitv share of

The percent change in electrical tariff = f(A *dF + B* dP + C' dl + D* dErJ where A, R, C, 1) are
annual coefficients weighting the relative cost components with dF being the ?; change i n cost fuel,
dP, the % change in the cost of purchased power, dL, the rate of domestic inflation, and dEr, the
average US$:Rp rate.
The new toll rate (TI) would equal the previous toll rate (Tp) times a function f(CP1, Rp, time) or
TI. = Tp * { Ir/Ip +[[ (Rr -Rp)/Ry] * Fc] + (0.025 * 11/12 J where Ir, Ip are inflation indices at time of
revision and past tariff setting, R is the exchange rate, Rp to US$, Fc is the foreign equity and offshore
loan portion to total investment, and 11is the number of nionths elapsed since the last adjustnlent.
total) llas been examined and considered in discussions with private investors.
Aclding this excliange rate factor to tlie tariff adjustment formula would have the
effect of adcling 0.5 to 1 additional percentage points to annual a\lcrage tariff
escalation during the first 10 years of operation. This is based o n a t\ppical
financing package of a 60 : 40 mix of domestic to foreign in\wstlnent, 20% equity
participation, and retirement over 8-10 \wars of foreign loans. At recent rates of R p .
depreciation of 3 % to 5%, this woulcl mean by the eighth year a tariff 1 0 % to 12%
higher than under the formula proposed bv Jasa Marga using onlv tlie cioinestic CPI
as an adjustment mechanism. The main concern from Governmcnt and user Lx)iiit o f
view is that under conditions of stagnant economic growth but increased inflation,
user savings woul~lnot increase in pice with tariffs. This could push tariff increases
wcll beyond the user savings ceiling, a n d lead to tolling o f f of more tr'ittic,
particularly for inter-urban toll roads and Class l l traffic.

Tests applying a pessimistic case scenario, that is a repeat of the 1986 43 'X
clevaluation coinciding with a peak vear for outstanding foreign loans, indicated a
16 % increase on that particular year's tariff compared with 8% annual increases
based on clomestic consuiner price indexing at the 1990's average annual ratc.

From the point of view of private investors and management of risk, addition of this
foreign excliange adjustment factor to the rate adjustment tormula would otter
additional security to offshore investors. However, as discussed in the section on
risk allocation and management, the two major investment risks are those on tlie
cost and delays in property acquisition and assurance of Governmental approval of
opening toll tariffs and their timely adjustment in line with maintenance and
operating costs.

Further and as raised by ADB, a long term policy objective is to encourage

development of a domestic boiicl market. This would encourage matching of
domestic debt with local currency revenues and broaden the market for debt issues.
1ncludi:ig a foreign exchange adjustment and protection mechanism would
encourage seeking off-shore financing and not contribute as inucli to deepening or
broadening of the domestic financial market. An additional concern is the effect on
the balance of payments, future claim on the current account, and the
accumulation of foreign currency obligations in the future.

9.7 Tasa Marga Proposals For A Tariff Adjustment Formula

Jasa Marga has studied this question of toll tariff setting and considered wavs to
make the process of tariff adjustment more simple and mechanical. I n accordance
with the 1980 Law On Roads and 1990 Regulation #8, Jasa Marga has suggested that
the initial tariffs for new toll roads or links be determined by tlie President together
with his determination of a rate adjustment formula for subsequent adjustments.
This initial stage would be implementecl by Presidential Decree as is the current
practice; subsequent adjustments woulcl be determined by the approved adjustment
formula under the authority of the Minister of Public Works. The period bctwccn
adjustments would be cletermineci by the Minister of Public Works in consultation
with the Presiclent.

The periodic acljustment mechanisni considered b\r Jasa Marga would be b a s ~ don
the percent change i n the Consumer Price Index (CPI). Subject to the new tarifl not
exceeding 70% of \diiclc operating cost saving (BKBOK), it would rcflcct thc
percent change in tlie CPI as illustrated below:

N e w Tariff (Tb) = Old Tariff * (1+[CPI(tiew)-CPl(~~ld)]/CPl(ol~l))

9.8 Conclusions and Observations on Toll Tariff Adiustnient Formulas

The CPI - indexed tariff acijustnient formula with its 70% BKBOK (including time
savings) limit has been proposed by Jasa Marga as a guide to potential investors in
the 1995 tender documents and instructions to bidders, subject to determination
according to prevailing laws and regulations." The proposed adjustment tormula
I,

is also contained as a "right to propose" every two years in the Authorization

Agreements (Article 14) which arc signed between Jasa Marga and toll road
investors. A certain element of unccrtaintv will persist until the toll road is g r a n t d
its Determination of Operation and approval of opening tariff. The plccigc of best
efforts of Jasa Marga to push through tariff adjustment proposals in the fraiiicivork
of the Authorization Agreenient is an intencled but limited form of comfort.

The user savings upper limit is set in Law Number 13. However, based on the
analysis in Chapter 4, the PCI inociel and certain approaches to time value inav be
overstating tlie user savings compared to results from the new VOC model
developed for Jasa Marga. Some tariffs appear to be 80% to over 100% of user
savings. Hence, pending resolution of this question and refining of the user savings
measurement methodology, the 70% guideline should perhaps be set aside.

The two year period between tariff adjustments set out in Jasa Marga Authorization
Agreements represents a good practical compromise between a one year and a three
year cycle This frequency of acljustments allows smaller increments of change and a
more gradual user adjustment than would a three year or longer cycle of
adjustment. On the other hand, annual changes would mean more frequent but
smaller changes. The shorter frequency however would be more costly in terms of
public information, printing new toll tickets, adjusting collection equipment and
changing sign boards. With new sections expected to open annually over the next
five to six years, the occurrence of tariff changes will become fairly randomly spaced
over time if an automatic formula approach is adopted.

82
Baseci on the above'analysis and c~iscussionswith various interests, the formukl and
guiclelines contained in the Authorization Agreement represetit an approach which
is already familiar to investors. A number of proposals to modify this formula in
terms of a framework for calculation and procedures for implementing have been
consicierecl. Tliesc are set out klow. Finall\# a number of options arc offered tor
ways to formalize the formula approach and provide a11 institutional context for
applving this tariff adjustment mcclianism.

Because of tlic danger that a consuilier price indexed formula could incrciwe rdtcs
faster tlian increases i n user savings, it is necessary to consider some adjustnlcnt
factors to tlie basic formula. Some additional considcratio~isfor adjusting the basic
CPI formula include:
- Regional differences could be taken into account in tlie CPl calcul'ilion
as wdl as the user saving reference ceiling. Bureau of Statistics publishes inonthl\~
consumer pricc indices for tlic 27 provincial capitals, and these shoulcl be used '1s a
reference for regional CPI adjustment. This would reduce tlie ciotninancc ot JaLirta
in the index, which would otherwise tend to distort increases. For example, ccwtral
and West Java have consistently had consumer price inflation 1%to 4% less than
Jakarta since 1990.

- Because wholesale pricc indices on public works road and bridge

construction have gone u p less on average tlian consumer prices over the last
wars, 7.1% compared with 8.1% since 2993, routine and periodic maintenance
expense woulcl not be expected to increase at the same rate as general inflation. The
BPS index could be used for this component of operating cost.

Comparison of Price indices for Consumer Goods

and Wholesale Price Indices For Public Works On Roads

Avq. % chq./yr
1990-'96 1993-'96
Consumer Price Index 8.4% 8.1%

Public Works:
Roads 8 bridges 7.2% 7.1%

- Operating and maintenance costs are a small percentage of total costs and
price indices could arguably only reflect the increases in variable components.
Comparing these two elements of local costs, operating and maintenance, including
periodic maintenance and toll equipment renewal against total investment, on a
present value or capital recovery basis, on-going, non-fixed costs make up about
20% of the total cost. Local debt service costs however are largely fixed for only the
short and mediuiii term and should by this argument be partially indexed. Short
anti meciium term debt would make-up a second factor for inclusion depending
upon maturities and temporal sensitiLrity.

- In addition to tlie above variable/ total cost factor there is the

consideration to reduce inflationary Eeedback effects on gooJs transport. This also
serves to protect consumer and user interests which ma\' not share to thc S J I I ~ C '
degree in the increase in per capita incomcs. Hence an "alplla" factor could be
applied to reduce the full CPI impact on tariff revision. Following the cx,1mplc. ot
some European concession toll formulas, this tactor would be set at a l c \ v l hcttwcen
0.5 and 0.9, in other words, the CPI impact would be reduced b!~between 3 1 % ciild
90%. It is intended that this "alpha" factor would obtain onlv on interurban toll
facilities, since in tlie future, off-peak tariffs M ~ O U I L address
~ the same group on urban
toll svsten1s.

Adjustment would under normal circumst,inccs bc carried o u t ever\' two years. 111
summary, this proposed tariff adjustmcnt formula with its compensation factors
and user saving upper bounds would take the following form:

New Tariff (Tr) = Old Tariff * {l+ a* [CPlr(iie~~)-CPIr(olc~))/CPIr(ol~~)~

Where a = proportion of non-fixed costs to total costs, and CPIr is the

appropriate regional consumer price index;

Subject to:

New Tariff < [Vehicle operating cost savings bv corridor and time of
clay + Passenger time savings by region]

To handle cases where tariff adjustments would exceed the user savings guideline, a
monitoring process and institutional tnechanism is required. As discussed in the
following Extion, this could be carried o u t within tlie Ministry of Public Works
(Subdirectorat Jalan Tol) with Jasa Marga participation, or alternativelv, lo~lgecl
with a Tariff Review Board outside the line ministry . This unit could also provide
an institutional mechanism for handling proposals from investors under Article 14.2
of the Authorization Agreement, that is "Proposals for Extraordinary Adjustment of
Toll Tariff" which takes into account events, for example Force Majeure, which may
have a material impact on the Business Plan of the investor.

As discussed in Appendix 9.1 to this chapter (International Experience In To11 Tariff

Setting and Adjustment), rate of return regulation of utilities and toll roads is
generally not favoured for a variety of administrative and practical reasons An
alternative guideline for monitoring would be the Corridor Traffic/Revenue re1ic.w
process applied in in some countries for BOT toll roads. (Sew Appendix 9.1 for a
ciescription of tlie review process for Nova Scotia Highwav 104 example.)

In this latter process, an over-sight committee perioclicallv reviews traffic l e \ d s and

growth in comparison with investor projections and business plans. I f traffic has
fallen far below expectations, and cash flow is insufficient to meet debt service and
operational obligations, this committee would assist i n renegotiating terms of tariff
and concession to forestall default. Convcrsclv, i t traffic anci rcvenuc's cxcc,.cd
business plan projections bv wide margins, tariff escalation according to the. index
forlnula coulcl be postponecl or suspended in tlie public interest. AItc*rii'itiw
institutional forms for review and monitoring are discussed further below' in tlw
elabnration of options for tariff adjustment mechanisms.

9.9 Alternative Institutional Arrangements For Processing Toll Tariff Proposals

And Tariff Adiustments
I n addition to tlie 464 kin. of toll roads presently i n operation, another 43 sections
arc presently under construction or investment process, If on average, tlicsc ‘ire
each opened in three stages, 135 Presiclciitial Decrees will have to be signed tor just
the new generation of toll roads coming on stream over the next four to five \ u i r s .
Requiring every tariff filing and adjustment to pass across tlie President's desk will
no longer be practical given tlie other burdens of office.

Given the situation outlined above of an urban ancl interurban toll road network
which will grow several-fold in extent and sets of toll tariffs which will e x p i n c i
even more in complexity, some alternative institutional processes for toll-setting
anci adjustment are set out for consideration below. The institutional options
proposed for consideration can be compared on the basis of three criteria:
- Simplification ancl standardization of procedures tor tariff setting and
adjustment;
- Reduction of risk premiums with inore security of achieving toll tariffs
as set out in Authorization Agreements;
- Increased predictability of tariff adjustments to cover operating,
maintenance, and debt service costs.

tariffs, but simplify subsequent adiustments using the index formula discussed
above and delegating authority to the Minister of Public Works:

A. Methodologv and Procedure for Tariff Setting for New BOT Toll Roads
i) Opening tariffs and inflation adjustment formula as agreed by Bid evaluation
committee, and approved by Minister of Public Works as part of investment process
and set out in Authorization Agreements;
ii) Together with Determination o f Operation, opening tariff passes tl1rougli
approval chain and is ciecicied by President together with adjustment formula;
iii) Result of tlie tariff calculatioli according to the decreed formula is clctermined
by tlie Minister of Public Works

B. Iiiiiilenientation Process & Institutional Consideratinns

i) Decision ancl approval for subscqucnt periodic tariff adjustments clc~.le+~tcd
b v the President to the Minister of Public Works
ii) Decision rendered after staff consultation with Cabinet SccrctarLit, Ministr!.
of Finance, Bappenas, Ministry of Communication (Land Transport Directorates)
iii) Technical staff within Public Works (Subclircctoratc of Toll Roads) ‘ind
Marpi responsible for
-updating ancl refining BKBOK model
-Calculating index forniula
iv) Review mechanism within Public Works with interagency/ministi.rial
committee to inonitor tariffs and corridor traffic volume and growth.

Option 11: President decides standard tariff adjustment formula for periodic
tariff adjustment and delegates approval over subsequent adjustments to Minister
of Public Works; Review and monitoring mechanism within Ministry of Public
Works

A. Metliociologv and Procedure for Tariff Settin): for New BOT Toll Roads
i) Opening tariffs ancl inflation adjustment formula as agreed bv Bid ev~ilucitioii
coininittee, anci approved by Minister of Public Works as part of investiiient process
anci set out in Authorization Agreements;
ii) President decides tariff adjustment formula for periodic tariff ailjustment,
ancl delegates approval over subsequent adjustinents to Minister of Public Works.
iii) Result of tlie tariff calculation according to the decreed formula is cietcriiiinecl
by tlie Minister of Public Works

B. Implementation Process & Institutional Considerations

i) Decision and approval for subsequent periodic tariff acljustiiients delcgatecl
by tlie President to tlie Minister of Public Works
ii) Decision rendered after staff consultation with Cabinet Secretariat, Ministry
of Finance, Bappenas, Ministry of Coniinunication (Land Transport Directorate)
iii) Technical staff within Public Works (Subdirectorate of Toll Roads) and Jasa
Marga responsible for
-updating and refining BKBOK model
-Calculating index formula
iv) Review mechanism witliin Public Works with interageiicv/ministerial
committee to monitor tariffs and corridor traffic volume ancl growth.

86
 Powers to designate toll road linkages ancl set tariffs would be deleg,llccl b\. the
President to the Minister of Public Works, \Tit11 final approval on iic)n-routine
acljustinents made in consultation with Ministerial committee incluciin:; C"il~inet
Sturetariat.

Option 111: President decides standard tariff adjustment formula for periodic
tariff adjustment and delegates approval over subsequent adjustments to Minister
of Public Works; review and monitoring mechanism placed with Tariff Ileview
Board outside line Ministry of Public Works, but reports to Minister of Public
Works.

A. Methoclologv anci Procedure for Tariff Setting and Adjustment

i) Opening tariffs and inflation adjustment formula as agreed tw Bid cvalu'ition
coininittee, anci approved bv Minister of Public \Vorks a s part of inwstincnt prcuxss
and set o u t in Authorization Agreements;
ii) President decides tariff adjustnient formula for periodic tariff actjustmcnt,
anct delegates approval ww subsequent adjustments to Minister of Public Works.

iii) Result of the tariff calculation according to the decreed formula is dctcmiincd
bv the Minister of Public Works

B. linpleinentation Process & Institutional Considerations

i) Decision and approval for subsequent periodic tariff acljustinents delepi ted
by the President to the Minister of Public Works
ii) Final approval on non-routine adjustments rendered in consultation with
inter-ministerial committee (Cabinet Secretariat, Ministry of Finance, Bappcnas,
Ministry of Communication (Land Transport Directorate)

iii) Technical staff within Public Works (Subdirectorate of Toll Roads) ancl Jasa
Marga responsible for
-updating and refining BKBOK reference inoclel
-Calculating index formula
iv) Review mechanism, tariff anct corridor traffic/revenue monitoring fuiiction
placed within Tariff Review Board established outside Bina Marga, but reporting to
Minister of' Public Works. Technical staff back-up could be provided bv Bilia
Marga and Jasa Marga.

The Tariff Review Board could consist of representatives of Ministry of Public

Works, other Ministries and Agencies such as Finance, Communications
(Directorate of Land Transport), Cabinet Secretariat, Bappenas, Jasa Marga, as well
as invited representation from user groups such as Yayasan Lembaga Konsumen
TOLL TARIFF SETTING AND ADJUSTMENT
Existing System

Investor

Proposed opening tariff & adj. formula set-out

in Authorization Agreement
& Approved by Minister of
Public Works
I
proposaI Proposal
Tariff Setting Tariff Adjustment

JM Review
1
JM Review

Review Review
Approval by Min. of PW Approval by Min. of PW

1
Cabinet Secretariat
1
Cabinet Secretariat
I

i
President
1
President

1
KEPPRES
1
KEPPRES
Toll Tariff Tariff Adjustment
TOLL TARIFF SETTING AND ADJUSTMENT
Option I

Investor

Proposed opening tariff & adj. formula set-out

in Authorization Agreement
& Approved by Minister of
Public Works

Proposal
+
Proposal
Tariff Setting Tariff Adjustment

1
JM Review
1
JM Review

1
Public Works Review
Approval by Min. of PW
I
Public Wo ks Review
Approval by Min. of PW

I
Cabinet Secretariat

1
President
1
extraordinary o r routine?

1
Interminist&iallAaency
Committee consultation
review & monitoring

1
- review 8, PWlJW tech. st
-updatehefine BKBOK
-calculation index formula

KEPPRES
- Opening Toll Tariff
- Formula (specific toll road):
for future Tariff Adjustment
- Framework for review & monitoring
- Delegation of Authoruty to
Min. Of Public Wks to
-Approve Formula calculation
& future adjustments
 TOLL TARIFF SETTING AND ADJUSTMENT
Option II

Investor

Proposed opening tariff 8 adj. formula set-out

in Authorization Agreement
8 Approved by Minister of
Public Works

4
(First cycle before KEPPRES)
4
[Subsequent Toll Road Ooeninqs 8 tariff cycles)
I
1
Proposal JM/Public Works: Review Proposal
Tariff Setting Reviewlcheck Proposal Tariff Adjustment
with Standard - opening Tariff
Tariff Adjustment formula

I1
-application of Standard
Tariff Adjustment formula 1
JMRe iew 1
Approval by Min. of PW
JM Review

1
Public Works Review
PublicWorks Review Approval by Min. of PW
Approval by Min. of PW

1 1 Extraordinary or routine?

+__ji
Cabinet Secretariat

1
President
Interministerial/Aqenc
onsultation
PW tech staff
custodian BKBOK

1
KEPPRES
1
Cabinet Secretariat
calculation index formula
Monitor tariffs
Monitor “revenue corridor
-Standard Formula: traffic volume 8 growth
for Tariff Adjustment
-Framework for review 8 monitoring
-Delegation of
Authority to
Min. Of Public Wks to
-Approve opening tariff
-Approve Formula calculation
8 future adjustments
 TOLL TARIFF SETTING AND ADJUSTMENT
Option 111

Investor
Proposed opening tariff & ad]. formula set-out
in Authorization Agreement
&Approved by Minister of
Public Works
I
+
(First cycle before KEPPRESI
+
(Subsequent Toll Road openinqs &tariff cvcles)
I
1
Prooosal
+
JM/Public Works: Review
+
Proposal
Tariff Setting Reviewlcheck Proposal Tariff Adjustment
with Standard
Tariff Adjustment formula
- opening Tariff
-application of Standard 1
JMRev wI1 Tariff Adiustment formula

1
Aooroval by Min. of PW
JM Review

1
Public Works Review
Approval by Min. of PW
Public V?orks Review
Approval by Min. of PW 1
1
Cabinet Secretariat d
Extraordiriarv or routine?

TRB supported by JW/PW tech staff

1
InterministeriaWAaency
Consultation with TRB custodian BKBOK
President

1 1
Cabinet Secretariat
calculation index formula
Monitor tariffs
Monitor "revenue corridor"
KEPPRES traffic volume 8 growth
- Standard Formula:
Tariff Adjustment
-Framework for review 8 monitoring
-Delegation of
Authority to
Min. Of Public Wks to
-Approve opening tariff
-Approve Formula calculation
& future adjustments
Indonesia, and inclustrv groups including truck, bus, and taxi companies 'llld

operator associations.

9.10 Summary
.,
The obiective is to establish a rcpAatory framework ancl guidelines
. which \vill
simplifv tariff review ancl processing. Routine tariff adjustments should be sc.ttlcd
mechanically by formula and routine tests. The State Executive would cic%\'otc
attention to exceptions and tariff filings outside the normal adjustment process.
The choice of option should also resolve the issue of approval of p r ~ p ~ s e 'opwing
cl
tariffs. This coulcl largely be accomplished upon signing of the Authori/.,ition
Agreement for BOT projects. Since the Minister of Public Works must appro\.c the
Concession License a i d Authorization Agreement, tariff proposals could lv y,i\,cn
Go\rernment approval in principle in the same step, subject to successful cotnplc~tioii
of the facilitv and its official clesignation on opening.

10. CONCLUSIONS AND RECOMMENDATIONS FOR RISK MITIGATION

AND TARIFF ADTUSTMENT

Measures to improve the agreement framework and regulatory structure to r e ~ n o \ ~ c

elements of uncertaintv which are within Government control include the following:
i) Make a consistent polic\r which limits investor exposure on land
acquisition cost overruns;
ii) Strengthen the long term planning function within Bina Mar31 tor
clevelopment of integrated urban ancl intercity toll road systems, and make a\,ailable
updated, comprehensive masterplans to guide land use and land transport
development;
iii) Incorporate in the tariff adjustment formula a mechanism which
predictably reflects cost increases and institutionalizes a process for tariff review and
adjustment
iv) Simplify the tariff adjustment mechanism and delegate power to adjust
rates to the Minister of Public Works

Within this framework of general recommendations, several issues must be

addressed. These issues include both broader sectoral and macro-ecoiioniic
concerns for medium and longer term planning as well as practical approaches for
dealing with the current generation of BOT toll road projects.

Sectoral concerns raised by ADB include the need to increase transport efficiency.
Private sector provision of toll roads represents a means of ensuring cost recovery
from users as well as an instrument for funding improvements of the trunk network
which would otherwise not be possible with limited domestic funds or soft
international loarrs. A commercial approach to network imprc>\’etnentin lica\rily
traveled corridors allows more rapid cievelopment than wouIci otlierwise occur and
raises more fun& from those users able to pay than would be available f r o m r o d
user charges and taxes. Tolls arc also an efficient charge and redistributiw t c on ~
Iff Category I, IiigIier income automobile users wlio, particularlv in coliscstcci
Ta1_‘
urba~iareas, have sufficiently high time values to Lie inscnsiti\.c to toll taritts.

There are however some caveats to this strict commercial vie\\. of r ( d

Jevelopiiient. Strong arguincnts exist for not abancioning the pliilosph!~ ot ldt>w
Number 13, in particular Articles 13 and 14 (see Appciiclix l), which states th‘it toll
roads are an alternative to tlie public non-toll network, ancl must Lie i n a i n t a i n c d <ISa
choice to users between destinations a i d to scnv local access requirements. l i v tlicir
nature, high-standard toll roads must ha\^ a limited number of interchanges ‘ind
access points. Secondly, toll tariffs act as a n additional charge 011 m~n~cmctit of
goods ancl people and the increased transport cost will reduce tlie zone of influence
ancl number of beneficiaries potentiallv affected bv road cievelopmcnt. This will
limit developnient more than non-limited access highways. The public benefits of
road clevelopment are reduced in two wa!’s; those users wlio are tolleci-oft cithcr d o
not make intended trips or add to the congestion on tlie public road. Incrcmeci
transport costs are reflected in passenger anci freight tariffs and tlic economic rcsturii
is reduced as shown in the Ciawi-Sukabumi case study.

This creates the need for a balanced network clevelopment plan. On oiic Iuiid,
private investors must have assurance tliat a parallel, competitive high standard
public freeway will not be constructed in tlie future to erode tlieir revenue base. 0 1 1
the other hand, local public roads must not be allowed to deteriorate to such a l c \ d
tliat local corridor users become captive to tlie toll road concession.

Toll Rate Adiustment Mechanisms

AS discussed above, toll tariff adjustment needs to be institutionalized with a
predictable and transparent formula anci a simplified, streamlinecl procedure for
aclministering. The current generation of BOT toll road projects has been tendered
and bid within tlie context of a proposed formula wliich reflects inflation as
measured by tlie national consumer price index (CPI) I n this context, bidders have
proposed base and opening tariffs, and these proposals become part of the
Authorization Agreement.

However, there are several arguments that the CPI indexing approach over-rewards
the concessionaire, in tliat many of tlie costs incurred by tlie investor are fixed in the
medium and longer term. Operating and maintenance costs make up Iess than half
the annual costs, tlie major share of costs being debt service. The ADB supports the
position that the iiidexing should be weighted or discounted to reflect that portion
of costs which increases with inflation.
Deterlnining tlie variable portion of costs clepe~ldshowever on the time horizon
taken. Given that tariff revenues should cover debt service costs by tlie order of 1.3
to 1.5 tiiiies,?? reliance upon the present domestic debt market for finalicing iiieans
heavy reliance on medium-term financing in the order of 5 years. Hence, ewer the
Iileciium term, a portion of debt ser\Zice and the cost of capital beconies variablc.

This then raises tlic question of off-shore financing with lower rates and longer
maturities available. For reasons of macro-economic and balance of pavmcvits
stabilitv as well as the long term cicvclopnient o f the domcstic capital market, local
financing slioulcl be encouraged to tlie iiiaximum extent. No unnecessary
inducements should therefore be given to foreign borrowing. This then argues
against inclusion of a foreign exchange adjustment factor in the tariff indexing
formula. Another reason raised by tlie Government for not further consicicring a
devaluation trigger in the adjustment is that Rupiah management policy and its
gradual depreciation arc alrcadv reflected in cioiiicstic inflation rates, and adding a
foreign exchange factor in the formula wouicl be ciou~lc-couiiting.

Restricting then tlie adjustiiient formula to domestic CPI movement, econoiiiic

scenarios can be drawn for various cases to cieterinine the proper form of indexing.
In an eiiviroimient reminiscent of Brazil or Mexico in the past clecacle, inflation
accompanied by economic stagnation o r shrinkage could lead to a diminishing p o l
of customers able to afford tolls. On the other hand, much of tlie success of the
Jakarta urban toll roads can be attributed to per capita incomes (and value of tiii-ie)
expanding two or more percentage points ahead of inflation.

Concerns with full CPI indexing also include the fact that high consumer iiifla tion in
heavily weighteci items such as food stuffs, housing, and clothing has limited
relation to non-labour coinponents of toll road costs. Escalating tariffs could also
lead to the perception of wind-fall profits by concessionaires and public backlash.
It is noted though that with profit taxation and revenue sliaring with Jasa Marga,
part of these gains can be captured back.

Regional CPI's would reflect more accurately labour costs, than would the higher,
Jakarta-dominated national CPI; Public works indices used in highway contracts
would also be more appropriate for routine and periodic maintenance components
of toll road operation.

Hence there are several reasons to consider applying indexing to only a part o f the
aiinual cost stream of toll roads. This policy however should be clelaveci until the

12 See Appendix 9.8.1, describing the Onunibus Agreement for Nova Scotia Highway 104 which
stipulates coverage of 1 . 2 1 for the initial three years, and 1.3 thereafter.

Hi$7wy T d i Sfid!, - A D 6 IYY; 90

 .

llext generation of BOT toll roads are tendered. Imposing a discounted CPI formula
on investors who have responded to the proposed full CPI index formula and may
I ~ a v eproceeded through the signing of Authorizatioll Agreements would open a
legal anti filiallcial taligle. Applying the CPI adjustment to the variable pWtiOi1 Of
cc)StSwoulj affectcash flows anticipated by investors. To compensate aiid I'CCOVCT
their anticipatecl returns, opening tariffs would have to be renegotiatcci witli the
objective of making the present value of the rc'veiiuc stream the saiiw as under the
assumption of full CPI indexing.

Raising opening year tariffs would then decrcasc diversion and reduce the number
of beneficiaries and users froin the ver'v kginning as shown in Table 10.1. For
example, discounting the CPI impact on tariffs by 0.9 o r 0.8 would mean point
reductions in real financial returns of 0.8% to 1.6%. This would mean a dcniand
from investors for compensating opciiing tariff increases of 1.(I6 and 1.I 1
respectively. This would reduce opening traffic by the order o f O.Y% and 2% and
economic returns 0.3 to 1%. If a narrower view of variable costs is taken with a
discount factor of 0.5 applied, the opening tariff demanded bv investors would
increase by a factor o f 1.3. In the case of the Ciawi-Sukabumi project, the adjusted
economic return woulcl drop to below 20% from the base case 24%.

Table 10.1
Impact Of Alternative Inflation Adiustment Factors On Financial Returns and Openinq Tariffs

Alternative CPI Adjustment/Discount ("Alpha" ) Factor Applications

Base Case
Full CPI Counted CPI Discounted
Alpha = 1.o 0.9 0.8 0.5

Parameter
Avg. tariff
escalation/yr 8.0% 7.2% 6.4% 4.0%

Financial Return
Current RD,
Project IRR 20.9% 20.1% 19.2% 16.5%
Equity IRR 26.3% 25.2% 23.7% 14.5%

Constant RD.
Project IRR
Before tax 15.9% 15.1% 14.3% 11.6%
After tax 14.0% 13.2% 12.2% 9.5%

Compensating Opening Tariff

Factor to increase Yr. 2000 tariffs
to restore FlRR to Investor
Expectation

91
Table 10.1 (Continued)

Opening Tariff Multiplier 1 .o 1.06 1.11 1.31

Impact on Traffic Diversion to Toll

Road
% decrease in Toll Road Utilization -0.9% -1.7% -21.1%

Impact on Economic Rate of Return

Resulting ElRR 24.1% 23.9% 23.2% 19.7%

Frcm a marketing and dcvclopment strategy p i n t of view, lower rates f r o m

opetiing woulcl encourage an earlier anci larger customer base, as well as gain better
public acceptance. Most critical Iiowever in the current generation of projects is the
prospect of further negotiation cielays anci iiivestor dissatisfaction. Tlic current
generation o f BOT toll road projects is taking a leiigtliv tinic to reach the
Authorization Agreement stage, with appreiitlv onlv six sections o u t o f thirt\'-one
projects in this package being signed as of Mav, 1997. Reacliing the point of actuall\,
starting works can often take longer.

While perhaps not ideal from a theoretical view of indexing, tlie current formula set
out in the Authorization Agreements constitutes a legal undertaking and should be
respected as such. Froiii a practical point of view, this element should be left intact,
b u t be subject to further review in tlie future as experience o n interurban toll roacl
traffic outside the Jabotabek region increases. As pointed out in the discussion of the
Ciawi-Sukabumi case study, achieving tlie growth rates and targets of generated
traffic may be problematic were econoiiiic growth to slow in the region. Hence the
proposal to apply CPI indexing to only a portion of the annual costs should be
clelayecl until the next generation of BOT projects. This should be exatninecl further
together with the question of time values and consumer stated preference surve)rs.

92
Appendices
 APPENDIX 1

LAW OF THE REPUBLIC

OF INDONESIA
NUMBER 13
OF THE YEAR 1980
ON R O A M

PT JAQAMARQA
CPERSEROI
 CONTENTS

I. PREAMBLE
2. CliAITERI :
3. Cl.lAPTER II : ROAD NETWORK .................. 4
Part One :
. The Role of Roads ...
Part two :
- Classification nf Roads
according to their Rvlcs 5
4. CHAMER Ill : ROAD PARTS ......................... (r
5. C H A F E R IV : AUTHORITY AND
COMPETENCE ....................... 6
6. CH,\PTER V : COMPETENCE 01:ROAD
DEVELOPMENT ..................... 7
Part Onc :
. Classification of Roads according
to Competence of Rvad
Developmenl ..__ I
Part Two :
- Competence uf heparalion of
long-tcrm General Plans.
Me&um-\erm Plans and
prognms. ond Competence of
Crealion of Mainknancc ........ 7
7, CIMMER VI : QUGhN1ZATlON O F I O L L
ROADS ................................... 10 L
Part One :
- Toll-mads ............................. 10
Part Iwo :
- Conditions lor Toll-roads ..
Part Three :
- Competence oforganization o l
Tvll-mads ...... II
Part Four :
Use orToll-roads ....
~
 I2 L A W OFTHE REPUDLIC OF INDONESISIA
8. CHAPIERVli : PROHIBITED ACTIONS ........
NUMDER 13 OFTHE YEAR 1980
.....
9 r H A m R W U : PENALPROVISIONS ........... I3 i

10. CHAPTER IX : TRANSITIONAL

. PROVlSION .......................... :. 14 ON
I1. CHAPTER X . : CONCLUSIONAL
. PROVISION ............................. 14 ROADS
ELIJCDATION : GENERAL ............................... 15
ARTICLE BY ARTICLE .......... . I 9 BY THE GRACE OF GOD ALivIIGHTY
THE PRESIDENT OFTHE REPUBLIC
OF INDONESIA,
Considering : a. that roads as one of the communica-
tion infrastructur substantially
constitute a major element in the
endeavour to elevate the life of the
nation to a higher plane and to pro-
mote the sense of oneness and tho
unity of the people towards the
national aim based upon the Pancasila,
a s laid down in the preamble
to the Constitution of 1945

b. that roads play an important part,

particularly in relation with t h e
realization of B balanced
interregional development and an even
distribution of the results of develop-
ment, as also in relation with the \
Stabilization of the notional defence and
security in the framework of the mani-
fest achievement of the aims of national
development;

c. 11131 in order to make roads fulfil their

role properly. the Government has the
right and the responsibility of road de-
velopment;

d. that to ensure the fulfilment of the role

of the road and the conceptional and

I
 .:, :;::. ..: 1 comprehensive occomplish~nentr d lhe WITH THE APPROVAL
.") ' development it is necessary to rcgulate OF THE HOUSE OF PEAOPLES'
all matters related to roads by Law: REPRESENTATIVES OFTHE'
REPUBLIC OF INDONESIA
In view of : I . Article 5 paragraph (I).article 20 para-
graph ( I ) and article 33 of thc HAS D E C I D E D :
Constitution of 1945: TI)
enact :
r.,':.!" THE LAW ON'ROADS
.>! '!I.~?:2:LawNumber5oftIte yem196l1ontlte
Basic Regulaliotl on Agrarian Priciples CIIAPTER I
(official Gazette of 1960. Nunlher 104.
. ,. . * I . .'' .Supplement tb the Official Cazettc GENERAL PROVISIONS
.! ~ . . . . No. 2043);
, Article I
. . 3. .Law Number 20 of [he year 1961 on
Cancellation of Rights to Lnnd and 10 In !his Law referred to with :
Inanimate Objects Existing Thcreon a. the State. the Slate d t h c Rcpublic of Indonesia.
(Official Gazette No. 288 of 1961.
I). tlic Cwqrnnrent- the State Apparaturc of the Unity State
Supplement to the Ollicinl Gazettc
No. 2224 ); 111 [he IZcpublic of Indonesia.
c. the Minister = the Minister responsible development of
4. Law Number 3 of the year 1965 on the road.
. '
Highway TrrOflic and Trnsporta<ion
d. the Regional Governments = the Regional Heads and
F (Official Gazette No. 25 of 1965.
the Regional Council Representative.
I Supplement to the Official Gazette
No. 2742); c. r o a d s = a unit or land-communication infrastructure
in any form whatsoever covering all the roadpans. in.
, . ,. 5. LAW Number 5 of the year 1974 on cluding conswclions and accesories required for traffic '
. ., Principles of Regional Government needs,
.. .(Official Gazette No. 38 of 1974,
1. public roads = roads destined for general traffic.
., . . I 1 "", " ' Supplement to the official Gazette
No. 3037): g special roads =mads other than referred to under 1,
i /
. . . t
11. toll roads= public roads for the use of which a toll charge
i. .,. . . is imposed.
. . ., . .. i. toll = a certain amount of money paid for the use of toll
roads.
, . . * ,,. j. road development= activities in the handling of matters
',,. ,!,. ... . ,, related to the network of roads, including determining
..
and concretizing of the targets of road development.

2 3
. .
CHAPTER I1 Part lko
Classification of Roads accordlng to their Roles
ROAD NETWORK
Part O n e Article 4.
The Role of Roads

Article 2. ( I ) Roods serving in pirticular transports characterized by

. . long-distance routes, high speed average and having an
efficiently limited number of accesses. are called Arte-
! ( I ) Roads play an importnnt role in the economic. political r i a w s c
. .! and sociocultural sectors and in the sector of security
and delence and roads are also used in the endeavour to (2) Roads serving collecting & distributing transports
achieve the greatest possible prosperity for the people. characterized by medium-distance routes. medium speed
average, and having a limited number of accesses. are
(2) Roads play an important part in the promotinn of all called Collector Roads.
Unit of Development Regions in the endeavour to reach
‘an ever more. spreading level of development between (3) Roads serving local transports characterized by short-
,{
the regions. distance routes low speed average. and having an un-
limited number of accesses. are called Local Roads.
(3) Roads from a unit ofthe road network system that binds
(4) The implementation ofthe provisions in paragraphs ( I ) ,
and connects the centres of growth with the territory ly-
( 2 ) and (3). are further stipulated by Government Regu.
ing in the influence sphere oftheir services in hierarchi-
cal relationship. lation.

Article 3.

( I ) The road network system playing a role in the distribu-

tive services for the development of all regions at no-
i
tional level, with all junctions of distributive services
growing into cities. forms a primary road network sys-
tem.
! (2) The road network system playing a role in the distribu-
!
tive services to the people in the cities, forms a second-
ary road network system.
( 3 ) The implementation of the provisions in paragraphs ( I )
and (2) are further stipulated by Government Regula-
tion.

4 5
 CHAPTER 111 ( 2 ) The competence referred to in paragraph ( I ) may be .
ROAD PARTS transferred either to legal entities or lo individuals, with
due observance of the public interests.
Article 5. (3) The conditions and procedures of delegation and/or
transfer as referrccl to in paragraphs ( I ) and (2) a n fur-
( I ) The mad-parts consis1 of the roed-way area. tllc right- ther stipulated by Government Regulation.
of-way area and the road control area.
CHAPTER V
( 2 ) The road-way area consists of the road-hed, the road- COMPETENCE OF ROAD DEVELOPMENT
side gutten and the safely thresholds.
(3) The right-of-way areaconsists of the road-way area and Part One
a certain strip outside the road-way area. Clmilication of Roads according to Competence
of Road Development
[4) The road-control area constitu~esof a strip of land nut-
side the right-of-way area, subjected to the control of Article 8.
the road administrator.
( 5 ) The implementailon of the provisions in paragraphs t I ), ( I ) Public roads, which me developed by the Minister. are
(2). (3) and (4) are further stipulated by Government classified as National Roads.
Regulation.
(2) Public roads. which are developed by the Regional
Government, are classified iu Regional Roads.
CHAPTER I V (3) Special roads. not developed according to the provisions
AUTHORITY AND COMPETENCE in paragraphs ( I ) and (2). are referred to in accordance
with the :
Article 6. - agencey.
legal entity.
-
~

individual L

(I) The authority over the roads rests with the State. concerned with their development.
(2) ‘lile authority of the State as referred to in paragraph (4) The implementation of the provisions in paragraphs . .
( I ) gives the Government the competence of road (I). (2) and (3) is further stipulated by Government
development. Regulation. I
i
Article 7. Part Two i
competence of Preparation of Long-ten General
(I) The compe\ence of the Government referred to i n Platu. Medium-termPI- and Rogrurrr,
article 6 paragraph (2) may be delegated or transferred and Compctence of Creation and Maintenance,
to government agencies both at central nnd at regional
level.
L

j
i

6 I
I

,. . ..
/
.
2 . .
 . . rial roads. collectorronds and local roads linked up with
Ariidi 9. a primary road network rests with the Ooyernment
12) The competence of the preparation of medium-term plun
( 1 ) Road development 3s referred to in article 6 paragnph and programming of the realization of arterial ronds,
(2) includes preparation of long-tern1 general plans, collector roads and local roads linked-up with a setwork
medium-term plan, programming, creation and meinte- of secondary roads i s transferred to the Regional Gov-
nance. ernments or delegated to officials or agencies either at
(2) Creation as referred to in paragraph ( I ) covers technical central or at regional level.
planning, construction, acceptance. hand-over and takc- (3) Tlie competence of preparation of ntediuni-term plan and
over. programming of the realization of special roads may he
transferred to :
Article 10.
- Regional Governments.
' he, competence of p p a r a t i o n of a long-term general
(1) T ~ lcgal entities.
plans for primary road networks rests with the Govern- - individuals,
ment
or delepaledlo officials or agencies either at central and
(2) The competence of prepscatton of a long-term general at regional level.
plans for secondary road nc' vorks is transfer red to the
Regional Governments ordLidgated toofficials or agen- (4)The implementation of the provisions in paragraphs 1 I ).
cies both at central and regional level. I?) and (3) is further stipulated by Government Regula-
tion.
(3) The competence of preparation of a longterm general.
plans for a network of special roads may be transferred Article 12.
to:
(I) The competence of technical planning and,conntructirin
- Regional Governments. 3nd that of maintenance of anerial roads. ccillector roads
- legal entities, or .and IiicaI roads linked-up with a network of primary
roads may be transferred to Regional Governments orL
- individuals, legal entities or delegated to officials or agencies either
or delegated to officials or agencies both at central and at central or at regional level.
at regional levet. 12) The competence of technical plonningand construction
and that of the maintenance of arterial roads, collector
(4) The implementation of the provisions in paragraphs roads and local roads linked-up with b network of scc-
(I). (2) and (3) is further stipulated by Government otidnry roads are transferred to Regional Governments
Regulation. or delegated to officials or agencies either at central in
at regional level.
Article 11. (3) Tlie competence technical planning and construction and
(1) The competence of the preparation of medium-term plan the coiiipetence of the maintenance of special roads are
and programming of therealization of networks of m e -

9
8
 delegated'to officials or to agencies at either central or Parto Two
regioiial instance, or transferred to: Conditions for Toll-roads

legal entities, Article 16.

- individuals.
(4) The competenceof acceptnnce hand-over and take-ovcr ( I ) A toll-road shall be of higher specification than the ex-
of arterial mads. collector roads and local roads linked isling public highway.
up with a network of primary roads rests with the Gov- (21 Atoll-road shall offer a higher reliability to the users
ernment. than the existing public highway.
( 5 ) Vie competence of acceptance hand-nuer and take-over (3) The lmplementation of the provisions in paragraphs ( I ) .
of arterial mads, secondary roads and local roads linked and (2) is further stipulated by Government Regulation.
up with a network of secondary roads is transferred to
the Regional Governments Part Three
(6) The implementation of the provision i n paragraphs ( I ) . Compelence of Organization of Toll-roads
(2). (3). (4) and ( 5 ) is further stipulated by Government
Regulation. Article 17.

CHAPTER VI ( I ) On the basis of the right to organize toll-roads as

ORGANIZATION OF TOLL-ROADS referred to in article 13. the Government delegates its
competence to a State Corporation of toll-mads
Part One
( 2 ) The Toll-road State Corporation referred to in paragraph
Toll-mads
( I ) is established according to the statutory regulations
Article 13. in force.
(3) The delegation ofcompetence to organize toll-mads does
The o w n h h i p and right of organizing toll-roads rests with not release the Government from its responsibility for
the Government. the roads. the organization of Which has been so i

delegated.
Article 14. ' ,
(4) The implementation of the provisions in paragraphs
(I) (2) and (3) is further stipulated by Government

I At a proposal of the Minister, the President determines a

road-section 85 a toll-road.

Arficle 15.
Regulation.

Part Four
Use of Toll-mads
Atoll-road is an alternative to an existing public highway.
Article 18.
(I)Toll-roads are destined only for road users with motor
vehicles against toll payment. '

10 II
 (2) The types of motorvehicles and the amounts of toll CIfAPTER V l l l
, charge referred to in paragraph ( I ) are determined by PENAL PROVISIONS
Presidential Decision.
Article 21.
(3) Use of toll-roads other than referred to in paragraph
(1) may be made possible with the approval of the
President. ( I ) Whosocver violates the provisions i n artlcle 20
paragraphs ( 1 ) and (2) is liable to imprisonment for
Article 19. maximally 3 (three) months or a fine of maximally
Rp. 150.000.-(one hundred and fifty thousand rupiah,).

( I ) Toll-road users arc obligated to comply with the statu-

.tory regulntions.on Highway Traffic and Transports. (2) Whosoever violates the provision in anicle 20 paragrapll
statutory regulations on Roads and other statutory regu- (4) is liable to imprisonment for maximally 7 (seven)
lation. days ora fine of maximally Rp. I5.000,- (fifteen thou-
sand rupiahs).
(2) TheToll-road State Corporation is obligated to compen-
sate lwscs incurred by loll-road users as a consequence (3) Whosoever violates the provision in anicle 2Oparagraph
of defaults in the organization of the toll-roads. (3) is liable to imprisonment for maximally 15 (fifteen)
years or a fine of maximally Rp. 50.COO.W,- (fifty
(3) The implementation of the provisions in paragraphs million rupiahs)
(1) and (2) is further stipulated by Government
Regulation. (4) Any propeny of the convicted person derived from or
intentionally used to commit the offence referred to in
paragraph (3) may be confiscated.
CHAPTER VII ( 5 ) The penal offence referred to in paragraph ( I ) or para-
PROHIBITED ACTIONS graph (2) constitutes a violation.
(6) The penal offence referred to in paragraph (3) consti-
Article 20. lutes a crime.
L

( I ) If is prohibited to commit actions within the right-of- Article 22.

way area and the cpnaol areaof any roads that may cause . .
an impediment to the role of such roads. i By Government Regulation may, as an implementation
(2) It is prohibited to execute any competences of road de- of this Law. be provided for threats of penalty of
velopment not in compliance with the statutory regula- maximally 3 (three) months imprisonment or of fines
tions in force. of maximally Rp. 150.000.(one hundred and fifty thou-
sand rupiahs).
(3) It is prohibited to operate any section of any road as toll
..> .
_~
I” road without Presidential Decisioni
(4) It is prohibit& to enter toll roads, except for ,011 road
users and officers in charge of the toll-roads

12 13
 . 'CHAPTERIX ELUCIDATION
TRANSITIONAL PROVISION OF
LAW OFTHE REPUDLIC OF INDONESIA
Article 23. NUhlllER 13 OFTHE YEAR 1980
CONCERNING
Existent statutory regulations not conirary to the provisions
ROADS
of this Law are declared still effective until amendecl or re-
arranged on the basis of this Law.

CHAPTER X G EN ERAL
CONCLUSIONAL PROVISION
I. R o a d s. as one of the communication infrnsstructures.
Article 24. B constitute essencialy a major element in the endeavour
to elevate the people's living-standard and to promote
This Law becomes effective as of the date of its pmnulpn- 9 the sense ofoneness and unity of the nation towards the
lion. In order that the public may take cognizance. hercof. achievement of the National Aims as intended to he re-
instruction is given to promulgate this Law by nnnouncc- alized through a sequence of comprehensive. well-di-
ment in the Official Cazette of the Republic of Indonesia. rected. integrated and continuous development plans.
Within such framework, roads play an important role in
Decided in Jakarta
realizing the targets of national development. such as
on the 29th of Desember 19RO.
evenly spreading of development and development re-
THE PRESIDENTOFTHE REPUBLIC
sults towards the creation of socinl justice for all PO-
OF INDONESIA. ple, considerable economic growth and sound and dy-
namic national stability. and in the long run the creation
sgd.
of B strong foundation for growth and development by
own efforts towards an Indonesian society that is pro-
gressive and just and prosperous bnsed upon the princi-
SOEHARTO
ples of the Pancasils philosophy. L

Promulgated in Jakarta 1 2. The growth and development of a national society and

, on the 29th of Desember 1980. Stale in panicular as also the growth and development
THE MINISTEWSTATE SECRETARY
OF THE REPUBLIC O F INDONESIA.
I of mankind in general, clearly require a rol of transpor-
lation services that supports the continuance of endeavors
of such society, and of mankind in general. Of the proc-
sgd. esses of the life of man towards the achievement of his
aims, his accumulative activities in undertaking are cat-
egorized ar the most important process, because they
SUDHARMONO SH produce what is identical to his principal aim. Thenfor
is needed a development in man's life whict is made
OFFICIALGAZETlEOFTHEREPUBLlC OF INDONE- possible when there is a pattern of efficiency in all
SIAOFTHE YEAR 1980 NUMBER 83. acitivities of his undertakings.

14 15

.. -
 people and pmducts will come into being., The flow of
goods is only made possible by distributive services, 4. Therehy we are aware that a balanced level 01 develop-
namely the inseparably integrated services of trade and ment between the regions is of considerable importance
transportation. Starting from the location of the nnttiral to guartinlee the targeted equalization o f growth as an
sources and ending with the last consumers. ' endeavour in concrelizing various development objec-
3. The scattered locations of both the natural sources and tives. The development level of a region ( i s . a territory
the last consumers demand that B pattern of efficiency within adtninisirative limits) will be influenced by the
' shall be followed in connecting the two. which is de-
Unit Development Unit involved.
picted by the forming of junctions of distributive serv- Going out from thc symptom that social systcms have
ices. From the economic point of view, it is these junc- the tendency to refuse the applicability of the law of
tions ofdistributive services that constitute the lootholds balance (as obvious from the symptom of evcr sharpen-
for the growth and development of cities. ing differences in the levels of regional development by
In such connection, each city will play a role in render; an uncontrolled course of the processes). so the devel-
opment of all Unit Development Region should in
ing services to the territory-lying in its sphere of influ-
principle i-e controlled. when achievement of a balanced
ence and they are all interconnected in a ceniin Iiiernr-
development is desired between the regions. Tne efforts
chical order. All cities together with their resp.
of controlling constitute substantially one of the balanc-
"hinterlands" will form with.the city of the highest hier-
ing steps in territorial development that may be laken
archical order, in the influence sphere of which they all
are. a territorial unit depicting an Unit Economic Re- either directly or indirectly. For instance. by way of
qiving an opportunity to seven1 Unit Devclopmenc Re-
gion. In this situation. man's whole life. including eco-
nomic-, political- and socia-cultural life and defence gion Of the small and economically weak category to
group together and so become larger and stronger.
and security. must bc completely grasped in the mecha-
nism of development. In such concept. a territory of 5. The process of grouping will bring an implication to the
which the development is controlled by the mechanism build-up of the supporting distribution system. Concur-
of development on the basis of the rules in economic rently with the understanding of territorial structure. the
life, with due regard to political-, sociocultural- and process of grouping is execuled by increasing the mar-
security-defence matters. represents Unit Development keting service capacity of one of'lhe hierarchically high-
Region. which will further be made the basis in formu- est ranking cities.
lating territorial development endeavours.
In the distribution system. the mad networksystem holds
In line with the understanding of the interconnection of and important part. because the increase of marketing
cities in hierarchical relationship, the road network will services is none other than the increjse of the density of
substantially show a certain structure. In such structure. distributive service. which demands expansion of the
the respective parts of the road network will play their communication infrastructure. such as the network ot
particular roles in accordance with the hierarchy within roads.
the Unit Development Region concerned.
In the framework of proper functioning of road network
As depictied above..the status of roads as one of the pans for the purpose of promotingthe economic growth
communication infrastructures is quite clear, as they tie in the hinterland. which constitutes an important
all cities together in hierarchical relationship and form a

16 17

. . .... .-. ,I-

 element for Ihe realization o f the spreading of dcvelop- Because the Government essentially discharges duties.
men1 and development results. i t is necessary to huild that according to reasonable criteria cannot be hdndlcd
roads of the specification "expressway". particularly ill hy tlie people themselves. so there for i n atranging lor
alre@y highly developed regions. With due regard to thercgulation there ofshall notonly b e c o v c ~ d t h e p r i n -
thesense of fairness. the road building ns shove referred cipal material constituting tlie identification of the main
toshall beaccomplished by the conswctiun of toll-roads. targets i n road development. but also emphasis on the
rights and responsibilities of both the Government and
6. I n the life ofanation, the position and the role of a r o d
the people and directions on the further regulation.
network system as above emphasised that funclanien-
tally a road network involves a vitnl nccessily o f the 8. With reference to the above set forth. the arrangenicnt
masses'and controls the formation of a regional Dcvcl- of the matters on roads must be immediately stabilized
opment Structure at National Level, especially such as in the form of a Law conserning the principal material
involves the realization of balanced development be- of the identification of the targets in road development.
tween the regions and the stabilization o f the national the stipulation of the rights and responsibilities and di-
security and defence in the framework of concretizing rections on the funher regulation.
the target of national development towards a just and
prosperous Indonesian society based on the Pancasila.
The.above elucidation shows the specific relevance - ARTICLE DY ARTICLE
i.e. thedirect relation - existing between the role of roads
Article 1.
and ths Regional Development Slructure. sothat thelink-
age oithe concept "spreading.of development" with the The definitions o f the terms in this article serve to pro-
said structure is' now quite clear. and also the aim vide for uniformity i n the understanding of this Law's
thereof- which i s specific. too-namely a halanced contents and of the implementary regulations thereof.
development between the regions, Thus: the realization
Letter a
and the expansion ofregional development and tlie bal-
Sufficiently clear
anced spreading thereoi and of the h i t s of develop-
ment between the regions. Considering the above set Letter b
out. i t is only proper that the Slate shall have the control Referred to with Government is the Government as
o f the road networks referred to in Law Number 5 of the year 1974 od
the Principles on the Regional Government.
7. The Government as the authority holder ofthe State hd-
ministration has the right of road development. As such Letter c
is h e case, each endeavour of road development must Sufficiently clear
be founded upon a spirit it of servitude to the Nation
Letter d
and the State.
Refemd to with Regional Government is the Re-
Road development as pnrt of the development o f com- gional Government as n f e m d to in Law Number
munication infrastructures involves both the elements 5 of the Year 1974 on the Principles on Regional
People and Government, so that the regulation thereof Government.
is directed to both the People and the Governmenf. I n
Letter e
such connection, a consensus i s required on the
Referred to with Roads i s the land communication
indentification o f the main targets in road development.

18 19
 infr&tructurcdestined for the traffic of vehicles. peo- Letter 1
ple and animals. Not included in this understanding l l i e determination of targcts includes the designine
are rail-roads, for instance railways. truck railroads of 3 long-term general plan. a medium-term plan
and cable-ways. and a program of the realization of the targets.
Referred to with thesentence pan"in iny form what- The realizaticrn of the targets includes the activities
soever" is hat the concept of road is not limited to in technical planning. construction and maintenance.
the conventional form of a road (on the land sur-
face) but includes roads spanning a large river, lake Article 2
or sea. below the land surface and under water ltun- Sufficiently clear.
nels), and above the land surface (Ily-over). Included
in road complementary constructions are construc- Article 3
tions that cannot be separated front the road. such Paragraph ( I )
as bridges, pontons, overpasses underpasses. park- Referred to with cities in this paragraph are those in
ing lots, c u l v e r ~buttresses
, and gutters/ sewers. In- connections with the road network, not the under-
cluded in road accessories are among others road standing of cities in Government administration.
beacons. traffic beacons. road signs. safety railings. Refened to with junction of distrihlive services is
right-of-way railings and right-of-way stakes. the junction as created by the application of effi-
ciency palterns to the flow of goods or people, and
Letter f becoming a foothold for the growth and develop-
Sufficiently clear. ment of a city pursuant to economic considerations.
Primary road networks arc closely connected to the
Letter g structure of Regional Development at,National
Referred to with Special Roads are the roads that Level. which according to Ihe elc.
are not destined for public traffic. such as irrigation
inspection roads. oil or gas pipe inspection roads. Paragraph (2)
plantation roads, mine roads, forest-or wood mads. Secondary road networks are closely connected to
private housing estate roads, roads required for State the structure of Regional Development of Cities.
securityldefence purposes. which according to the role of heir services consist
When a section of a special road pursuant to statu- ofArtery Roads, Collector Roads and L ofal Roads. '
tory regulations in force or by the proprietor IS Primary Artery Roads and Primary Collector Reads
opened to public traffic. then to such road section remain thorough fares although they into cities.
and such traffic the statutory regulations on Roads
and on Highway Traffic and Transponation s h d l Paragraph (3)
apply ! sufflciently Clear
Letter h
Sufficiently cleat Article 4
Paragraph (1)
Letter i Sufficiently Clear
Sufficiently clear
Paragraph (2)

i sufficiently Clear

20

~. . . .- . ..
 Paragraph 13) Article 7
Sufliciently clear In general. thecompetenceofroadenginccringrests
~aragraph(4) with the Government, bur next iherctd. as there are
sufficiently Clear particularities in the sectors of irrigation, estates.
mining, forestry.pons. defence and security, and oth-
Article 5 I ers. so in the execution of.road development it is .
necessary to duty observe the statutory regulations
Paragraph ( I ) !
I in the respective sectors.
sufliciently Clear !
Delegation andlor transfer of the cornpetencc a rc-
Paragraph (2) fened'toin this Article means that the right and au-
the roadway area is an area used for the contruction thority to execute the engineering are turned over to
of a road. coniisting of the roadbed. the roadside the agencies concerned.
guttecdsewen and the safety threshold (sill). rile transfer of competence to a legal entity or an
The madbed comprises the traffic Iane(s) with or individual is done with due regard to the interests .
wilhoutkaKcdentarcation strip or marking and the and the prosperity of the society in the surrounding 1'

road ghoulden. area. However, in doing so, for the reason that the i
'the safely thrcsholdlsill lies in the outermost part engineering and management of the roods consti-
of the roadway area and serves to protect the road tute the burden of the legal entity and individual as
construction. the transfer, the Government will not negleci its
interests.
Paragraph (3)
the right-of-way'arca is bounded by the boundary
marksof the right-of-way. Article 8
A certain strip of land oulside the roadway area but Pragraph ( I )
in the right-of-way area is meant to meet the condi- Public roads classified as National Roads, are called
tion of safCry space fw the UK of the road, amonq National roads.
others fortlie purpose of widening the roadway at J
Paragraph (2)
. , icttcr time.
1
Public roads classified as Regional Roads : '
Paragraph(4) ~

developed by a Provincial Government are

the Road Control Area is a swip of land lyng outside called Provincial roads
the Right-of-way Area, the use of which strip is un- ~

developed by a Subpmvincisl Government are

der the control of the road engineer. for the putposc called SubprovinciaVMunicipal Roads:
of guarding against obsawtion to the motordrives' ~

developed by a Village Government are called

sight and to the road constluction. in the case that Village Roads;
the space of the right-of-way is inadequate.
Paragraph (3)
Paragraph (5) Special Roads engineered by an Agency/ Legal
Sufliciently clear, Entity are a.0. called irrigation roads. estate (plan-
tation) roads. mining roads. forestry roads, private
Article 6 housing complex roads, porllharbor roads, and oth-
Sufliciently clear. ers.
I

22 23
... .. ~ ___.~ .,.
---+.Tr.l.~.
 Paragraph (4) Sufficiently clear.
Sufficiently clear. Paragraph (3)
Sufficiently clear.
Article 9. Paragraph (4)
Paragraph ( I ) ! Sufficiently clear.
Road engineering essentially comprises the Article 12.
dermination and the realization of the targets. !
Included in ihe determination of targets are the ac- Paragraph ( I )
tivities in designing long-term schemes, designing Sufficiently clear.
of inedium-term plans and programming of the re- Paragraph (2)
alization of the'plans. included in the realization of Sufficiently clear.
targets arethe activities in creation and maintenance.
Paragraph (3)
Paragraph (2 ) Sufficiently clear.
The concept of creation is not limited to the techni-
Paragraph (4)
cal designing and conslruction alone, but also in-
cludes acceptance. delivery and take -over. Sufficiently clejr.
An example of acceptance and delivery is, for in- Paragraph ( 5 )
stance, when the Government accepts the delivery Sufficiently clear.
of a Special Road from a legal entitylagency and
Paragraph (6)
then declares it to be a Public Road.
Sufficiently clear.
Article 10
Article 13.
' Paragraph ( I )
Suf ficien tly cl$ar. In view of the fact that a toll-road constitutes a part of
the network of public mads closely connected with the
Paragraph (2)
necessities of life of the great masses, it is only propCr
Sufficiently clear.
that tlie right of ownership and the right of development
Paragraph (3) of toll-roads are in the hands of the Government.
Sufficiently clear. Tlie administration of toll-roads COVCK all activities in
Paragraph (4) materializing the aims of their development and the ac-
: Sufficiently clear. tivities of h e i r operation.
Article 11. Such operational activities include the collection of toll-
charges, tlie regulation of the use of the toll-roads. the
Paragraph ( I )
taking of precautionary measures and other activities in
Sufficiently clear.
accordance with the purpose and objectives of the toll-
Paragraph (2)

24 25
 road undertaking. Referred to with "higherreliability" i s Constant serv-
Article 14. iceability and safety.
The proposal from the Minister to :termihe a road w c - Paragraph (3)
tion as toll-road shrll be hased tipon a mcdiuiii-tcriii
general planning and rheGovernircnt progrntn ti( irrnt-
ing a network ofmads. The construction 01tlic t a d 1 r d
ij Sufficiently clear.

Article 17.
shallbeexecutedwithoulimposin~abuiilenon tltcStntc. i
budget and the costs of operating a mritorvehiclc on tlic Paragraph ( 1)
i I n view of the fact that a toll-road is a public road
toll-road addcd with the toll-clime nitist he still I w c r ~

than the casts of aperaring a mnturveldcle on tlic cxi5t- -

he. open to public traffic and has particular char-
ing public highway. acteristics. it is only proper that the operation o f toll-
roads is executed by a State Corporate B d y .
Referred to with State Corporate Body i s what i s
Article 15.
referred to as such in Law Number 9 of the year
Referred to with "an nlternative" is. i l i a t bcsidcs tlic 1011- 1969 concerning the Forms of Slate Enterprises.
road. there must be another tlioroughfare with ilte m m c The particularity referred to above finds its cause in
place of origin and the same place oldestiiiation. so that the fact that the object of the enterprise constitutes
the road users w i l l be at liberty to choose rvliether or not o f roads. for the use of which a tollsharge is im-
to use the toll-road. I n the event that the alternative puh- posed.'however the purpose is here the levelling of
lic highway is not properly functioning the toll-road will the regional developmenu towards realization of a
automatically became apublic lhoroughfare without toll- balanced development between the regions, and
charge. evenly distributed.rasulls of development, besides
h e stabilization of the defence and security of Na-
Article 16. I tion and Slate.

Paragraph (I)
I Paragraph (2)
i Sufficiently clear.
Referred lo wilh "higher specification"of toll-roods
are the specification of free ways such as: I Pangraph (3) '
.
a. having no intersections on the same level as other Sufficiently clear.
roads;
b. having no direct accesses, except for controlled 1'
'Pmgraph(4) .
direct accesses: Sufficiently clear.
c. the costs of operating a motorvehicle on the toll-
road added with the toll-charge i s s t i l l lower than Article 18.
the costs of operating a motorvehicle on the ex- Paragraph ( 1 )
isting alternative public highway. l%eoperation Sufficiently clear.

i
casts af amatorvehicle include among others
the costs of fuel, lubrication. wear and tear and
time-value.
Paragraph (2)
Paragraph (2)
Sufficiently clear
Paragraph (3)

26 27
 Article 23. .
The purpose of this provision is to prevent r( fniv vnctiunt
in road matters.
Article 24,
Sufficiently clear.

30
 Referred to with"use other than referred to in para- Paragraph (4)
graph (I)'' includes, among others, the activities The reference of "entering a uAl-road" i? limited to
carried out for the national interest, delencc and se- pedrstrians who do not impede the role of thc r o d
curity activities on the loll-road, so that the toll-road and do not cause damage to the toll-road.
cannot be used tor normal motomzed traffic.
The approval of the President will be issued 31 the Article 21.
proposal of CRc Minister.
Paragraph f I )
Sumciently clear.
ArticIe 19.
Paragraph (2)
Paragraph ( I ) Sufficiently clear.
Sufficiently clear.
Paragraph (3j
P m g w h (2) The provisions in this' article an meant to give a
This matter has the purpose to give toll-mad users legal ground for criminal proceedings in road
the right to claim for and obtain compensation from operatiorfuse and constitute supplement to the pro-
cheToll-Road Operating State Corporation lor !asses visions in the Penal Code, Chapter VII Article I92
pmvenly caused by defaults of the said Toll-road and 193.
Corporation in administering the toll-road.
Paragraph (41
Paragraph (3) Sufficiently clex.
Sufficiently clear i Paragraph ( 5 )
Sufficiently clear.
Article 20
Paragraph (6)
Paraw& (1) This pcnni act is evaluated as a criminal offence br-
Obstruction of the road 's role may be caused by cause the consequences of such act may endanger .
misplaced use of Ihe road for instance, runningljog- the public safety or the safety of either people or
ging on the road, bicycling, or using the road as a goods.
-
place for roller-skating. skate b.oard riding on a 1
toll-road. or placing large stonedrocka, dumptng
sand, causing all SON of obstacles in the roadway Article 22.
area of public roads. To ensure the besl possible execulion of Ihe regulations
Paragraph (2) or measures which constitute the implementation of chis
Law, it is necessary 10 lay down penal sanctions as con-
Sufficiently clear.
tained in this article. Tttc penal sanctions only concern
Paragraph (3) matters Laving the character of trespasses.
Organizing any road section as a toll-road shall be
by Presidential decision on the basis of a proposal
from the Minister. so that organizing any road sec.
tion as a loll-road without Presidential Decision con-
stitutes a contravention of this Law,

29
 Appendix 2.1

^....

Tangerang Barat-Merak 1
Annual Growth
 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Vehicle speed on the toll road and alternative road IS taken from survey data
Time Value, Category I,Rp 3.281.25 lhourlveh.

Indonesia Highway Capacity Manual (IHCM)

- ~
 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Link Distance Travel Speed Tariff as % VOC Saving

No. Location PCI Model New Model Toll Tariff Tariff/km
Non-To1 Non-To1
To1 (km) To1 (kph) voc voc PCI Model New Model
(km)
(kph) Savinqs Savings

I I I I I I I I I I I I I
Vehicle speed on the toll road and alternatlve road IS taken from survey data
Time Value, Category I, Rp 12,287 /hour/veh

Jasa Marga, recent practise

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category I, Rp 1,341 lhourlveh

PCI (orlginal, 1979)

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category I, Rp 7,076 Ihourlveh

Jakarta lntra Urban Toll Road, Northern Extention of South-west Arc, 1989
 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

I I I I I I I I I I I I I
Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category I, Rp 8.880 Ihourlveh.

Surabaya-Mojokerto (JICA, 1391 )

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Link Distance Travel Speed Tariff as % VOC Saving

No. Location PCI Model New Model Toll Tariff Tariff/km
Non-To1 Non-To1
To1 (km)
(km)
To1 (kph) voc voc PCI Model New Model
(kph) Savinqs Savinqs

1 I I I I I I I I I I I I
Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category IiA, Rp 18,212 Ihourlveh

Indonesia Highway Capacity manual (IHCM)

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category HA, Rp 3,827 /hour/veh.

PCI (original.1979)
 Appendix 4.1

Link Distance Travel Speed Tariff as % VOC Saving

No. Location PCI Model New Model Toll Tariff Tarifflkm
Non-To1 Non-To1
To1 (km) To1 (kph) voc voc PCI Model New Model
(km)
(kph) Savinqs Savinqs

I I I I I I I I I I J
Vehicle speed on the toll road and alternative road IS taken from survey data
Time Value, Category HA, Rp 18,534 /hour/veh

Jasa Marga. recent practise

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Link Distance Travel Speed Tariff as % VOC Saving

No. Location PCI Model New Model Toll Tariff Tariff/km
Non-To1 Non-To1
To1 (km) To1 (kph) voc voc PCI Model New Model
(km)
‘kph) Savinqs Savinqs
t

1u uawi-uanjur, witn non-toll :

a. Alternative I via Sukabumi 80.5 84.6 67 27 48,301.35 31,710.69 16,020 199.00 33.17 50.52
b. Alternative II via Puncak 80.5 49.9 67 27 6,949.40 -840.70 16,020 199.00 230.52 -1905.49
I I I I I I i I I I I I I
Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category IIA, Rp 14,670 /hour/veh.

Jakarta lntra Urban Toll Road, Northern Extention of South-west Arc , 1989

~~~
 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

I I I I I I I I I I I I I
Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category IIA, Rp 7,960 /hour/veh.

Surabaya-Mojoketio (JICA, 1991)

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

I 1 I I I I I I I I
Vehicle speed on the tall road and alternative road is taken from survey data
Time Value, Category 116, Rp 4,971.2 /hour/veh.

Indonesia Highway Capacity Manual (IHCM)

-~
 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Link Distance Travel Speed Tariff as % VOC Saving

No. Location PCI Model New Model Toll Tariff Tarifflkrn
Non-To1 Non-To1
To1 (km) To1 (kph) voc voc PCI Model New Model
(km)
(kph) Savinqs Savinas

Time Value, Category 116, Rp 18,534 /hour/veh

Jasa Marga. recent practise

 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Time Value, Category IlB, Rp 3.152 lhourlveh

PCI (origmal. 1979)

_-- -_
 Appendix 4.1

Link Distance Travel Speed Tariff as % VOC Saving

No. Location PCI Model New Model Toll Tariff Tariff/km
Non-To1 Non-To1
To1 (km) To1 (kph) voc voc PCI Model hiew Model
(km)
('ph) Savinqs Savinqs

I I 1 I I I I I I I I I
Vehicle speed on the toll road and alternative road is taken from survey data
Time Value, Category IlE, Rp 3,659 /hour/veh

Jakarta lntra Urban Toll Road, Northern Extention of South-west Arc., 1989
 Appendix 4.1

Percentage of the Toll Tariff to the VOC Savings (with time value)

Time Value, Category 116, Rp 7.980 Ihourlveh.

Surabaya-Mojokerto (JICA, 1991)

 Appendix 4.2

PCI VEHICLE OPERATING COST MODEL

I. PCI Model (Toll Road)

Equation of Fuel Consumption :

Car : Y - 0.04376 S2 - 4.94078 S + 207.04840
Bus : Y -
- 0.14461 S2 - 16.10285 S + 636.50343
Truck : Y -- 0.13485 S2 - 15.12463 S + 592.60931
Y - Fuel consumption (litre11000km)
S - Running speed (kph)

Equation of Engine Oil Consumption :

Car : Y - 0.00029 S2 - 0.03134 S + 1.69613
Bus : Y - 0.00131 S2 - 0.15257 S + 8.30869
Truck : Y - 0.00118 S2 - 0.13770 S + 7.54073
Y - Oil engine consumption (litrell000km)

Equation of Tyre Wear :

Car : Y - 0.0008848 S + 0.0045333
Bus : Y - 0.0012356 S + 0.0065667
Truck : Y - 0.0015553 S + 0.0059333
Y -
- Wear of one tyre per 1000 krn

Equation of Maintenance Cost :

Maintenance Cost on Parts :
Car : Y -
- 0.0000064 S + 0.0005567
Bus : Y -
- 0.0000332 S + 0.0020891
Truck : Y -
- 0.0000191 S + 0.0015400
Y -
-
Maintenance parts, equated as depreciable
value of the vehicle, per 1000 km
Maintenance Hours of Labour :
Car : Y -
- 0.00362 S + 036267
Bus : Y -
- 0.0231 1 S + 1.97733
-
Truck : Y - 0.01511 S + 1.21200
Y - Hours of labour per 1000 km

Equation of Depreciation :
Car : Y -
- ll(2.5 S + 125)
Bus : Y -
- 1/(9.0 S + 450)
Truck : Y -
- ll(6.0 S + 300)
Y -
- Depreciation per 1000 km, equated as the A
'
depreciable value of the vehicle

Equation of Interest :
Car : Y -- 150/(500 S)
 Bus : Y - 150/(2571.42857 S)
Truck : Y - 150/(1714.28571 S )
Y -
- Interest per 1000 km, equated as the % depreciable
value of the vehicle

Equation of Insurance :
Car : Y - 38/(500 S)
Bus : Y - 6/(2571.42857 S)
Truck : Y - 61/(1714.28571 S)
Y - Insurance per 1000 km, equated as the new value of
the vehicle

Equation of Travelling Time :

Car : Y - .
Bus : Y - 1000 / s
Truck : Y - 1000 / s
Y -
- Travelling time per 1000 km

Average Number of Crew per Vehicle

Car : Driver 1
Bus : Driver 1
Conducter 1.7
Truck : Driver 1
Asistant 1

Overhead
Car : -
Bus : 10% of above sub total
Truck : 10% of above sub total

II. PCI Model (Non-Toll Road)

Equation of Fuel Consumption :

Car : Y - 0.05693 52 - 6.42593 S + 269.18567
Bus : Y - 0.21692 S2 - 24.15490 S + 954.78624
Truck : Y - 0.21557 S2 - 24.17699 S + 947.80862
Y - Fuel consumption (litre/lOOOkm)
S - Running speed (kph)

Equation of Engine Oil Consumption :

Car : Y - 0.00037 S2 - 0.04070 S + 2.20405
Bus : Y - 0.00209 S2 - 0.24413 S + 13.29445
Truck : Y - 0.00186 S2 - 0.22035 S + 12.06486
Y - Oil engine consumption (litre/lOOOkm)

Other items such as, equation for tyre wear, maintenance cost, depreciation,
interest, insurance, travelling time, and overhead are same as tollway.
 Appendix 4.3

NEW VEHICLE OPERATING COST MODEL

Fuel Consumption

There is a basic relationship between fuel consumption and speed out of road
geometry, roughness and traffic condition. The kind of consumption is the basic
fuel consumption, which is defined as fuel consumption on free flow condition, flat
gradient (0%) and road roughness relatively not affect the fuel consumption. The
basic fuel consumption model can be explained as follows.

Fuel consumption = basic fuel (I+( kk + kl + kr )) ...................... (1)

where :
basic fuel in liter/l OOOkm
kk = correction due to gradient
kl = correction due to traffic conditions
kr = correction due to road roughness

From our empirical modeling it has been found that the relationship between basic
fuel and speed, v (kph), can be expressed as :

Basic fuel for Category I = 0,0284 v2 - 3,0644 v + 141.68 . _. . . . . . (2)

Figure 1 shows the relationship between the basic fuel consumption and speed

For other vehicle categories, their basic fuel consumption are found from separate
experiments. Their basic relationships relative to the basic fuel consumption of
vehicle category I are given below.

Basic fuel Category IIA = 2.26533 ' Basic fuel Category I, and
Basic fuel Category llB = 2.90805 Basic fuel Category I

Table 3 summarized the correction value for each of the category

Correction due to (-) gradient (kk) < -5% - 0,337

-5% 5 g < 0% - 0,158
Correction due to (+) gradient (kk) 0% 5 g < 5% 0,400
2 5% 0,820
Correction due to traffic (ki) 0 5 v/c < 0,6 0,050
0,6 s vlc < 0,8 0.185
vlc 2 0.8 0.253
Correction due to road roughness (kr) < 3 mlkm 0,035
 IZ0
100 I
-E
x y = 0 0Z84x2 - 3 0644x + 141 68
x
0
80
\
R’ = 0 8089
.’t

10 20 30 40 50 60 70 80 90 100
Speed (kph)

Figure 1 Basic Fuel Curve for Vehicle Category I

Lubricant (Toll and Non-Toll)

The model has been adapted from GENMERRI, a VOC model developed for road
maintenance and programming. Table 4 shows the basic lubricant consumption in
literlkm against speed (kph). The values are subject to corrections due to road
roughness conditions, given in Table 5 .

Table 4 Basic Lubricant Consumption (literlkm)

Speed
(kmlhr) Category I
10 - 20 0.0032 0.0060 0.0049
20 - 30 0.0030 0.0057 0.0046
30 - 40 0.0028 0.0055 0.0044
40 - 50 0.0027 0.0054 0.0043
50 - 60 0.0027 0.0054 0.0043
60 - 70 0.0029 0.0055 0.0044
70 - 80 0.0031 0.0057 0.0046
80 - 90 0.0033 0.0060 0.0049
90 - 100 0.0035 0.0064 0.0053
100 - 110 0.0038 0.0070 0.0059
Table 5 Correction Factors for Lubricant Consumption for All Vehicle
Category

Roughness Correction Factor

< 3 mlkm 1.oo
2 3 rnlkm 1.50

Tyre

The equation has been adapted from PCI model. as summarized in Appendix 4.2

Vehicle Maintenance Cost

Maintenance cost comprises spare parts and wages for the maintenance labor. This also
adopted from the PCI model as given i n Appendix 4.2.

Depreciation

Depreciation is expressed given by the PCI model.

Interest

The modcl for interest is adopted from the HDM 111 model. where interest i n veh-hm. INT.
is expressed as a fraction ofthe proce of brand n e vehicle:
~

IN'I'=AINT/AKM

where:
AINT = 0.01 (AINVD)
AlNV = anntial interest of vehicle (I%)
AKM = avcrage annual mileage expressed i n hni

It is assumed that the interest would have a marginal effect 011 choice between traveling on
toll or non-toll road. Whether or not traveling via toll road the interest would be the wine.
therefore no saving obtained from interest component.

InsLlrilnce

Insurance is expressed given b y thc I'CI model

Value oftinie

'There is no hard evidence on valtic of time for lndoensia. IT1sttidy and Jasa Margo recciit
practice nssumed tliat vehicle utilization time savings is the basis for valuation. Mliicli i s
perhaps not that common.

Here user's time saving is asstimed as a soiirce for valite of time determination. Value5 of
time used i n previous feasibility studies were adjusted to retlect inflation and regional
variation. Regional variation can be approximated tising regional ininimtini wage laic
index supplied by BPS.
 APPENDIX 7 2

Table A 7.2.1
-
Ciawi SUkabumi:ECOnOmiC Analysis as Non-toll Road'
(Costs and Benefits Adiusted for non-traded components1
TraMc growth Scenario RPT Forecast on total corridor;
Economic Analysis as non-toll road (additional const~uction.maintenance8 operation costs 01 road as toll faciily excluded)
Generated Tramc assumed at 25% of normal traffic 8 50% of savings benefits counted
Normal growth traffic based on RPT Oct. '96 estimate Of Total corridor lraflic share
,- R0.I
lbill , .
Normal Traffic Growth:
Year Project Mainl. 8 OC Savings Time Savings Puncak Gen. Traffic Net Benefit
Costs operation. Total Total Total Total Diversion Benefils Stream
(a) (b)
1996 05 -0 5
1997 33 9 -33 9
1998 79 4 -79 4
1999 133 0 -1330
2000 99 5 0.7 17.6 10.4 0.0 28.0 3.5 -68 8
2001 25 8 2.0 34.5 22.2 0.0 56.7 7.1 36 0
2002 2.5 66.5 40.2 0.0 106.8 0.03 13.3 1177
2003 2.5 73.6 44.4 0.0 118.0 0.03 14.8 130 3
2004 2.6 81.9 49.4 0.0 131.3 0.03 16.4 145 2
2005 10 1 2.6 91.5 55.2 0.1 146.7 0.03 18.3 152 3
2006 95 2.6 101.9 61.5 0.1 163.5 0.04 20.4 171 7
2007 2.7 113.3 68.2 0.1 181.6 0.04 22.7 201 7
2008 2.8 125.5 75.5 0.1 201.1 0.04 25.1 223 5
2009 2.8 140.6 84.5 0.1 225.2 0.05 28.1 250 6
2010 41 3 3.2 156.0 93.6 0.1 249.7 0.05 31.2 236 6
201 1 95 3.2 168.4 101 1 0.1 269.6 0.05 33.7 290 6
2012 3.3 183.4 110.0 0.1 293.5 0.06 36.7 326 9
2013 38 9 3.8 197.3 1183 0.1 315.8 0.06 39.5 312 6
2014 3.9 214.6 128.7 0.1 343.4 0.06 42.9 382 5
2015 71 1 4.7 232.7 139.4 0.1 372.2 0.07 46.5 342 9
2016 97 4.8 249.6 149.4 0.1 399.2 0.07 49.9 434 7
2017 4.9 265.4 158 8 0.2 424.4 0.07 53.0 472 6
2018 50 280.8 168.0 0.2 448.9 0.08 56.1 500 1
2019 51 296.8 177.6 02 474 6 0.08 59 3 528 9
2020 10 1 5.2 314.2 188.1 0.2 502.5 0.09 62.8 550 1
2021 94 53 332.4 199 0 0.2 531.6 0.09 66.4 583 5
2022 5.4 351.7 210 5 02 562.4 0.10 70.3 627 5
2023 5.5 372.1 222 7 02 595.1 0.11 74 4 664 1
2024 5.6 393.7 235 6 02 629.5 0.11 78.7 702 7
2025 5.7 416.5 249 2 02 665 9 0.12 83.2 743 6
2026 5.8 440.6 263 7 03 704.5 0.00 88.1 786 8
2027 5.9 466 2 279 0 03 745.4 0.00 93.2 832 7
2028 5.9 493.2 295 1 03 788 6 0.00 98.6 881 3
2029 6.0 521.8 312 3 03 834 4 0.00 104.3 932 7
2030 6.1 552.1 330 4 03 882 8 0.00 110.3 987 0

ElRR (to 2023)= 30 7%

ElRR (to 2030); 30 8%
'Prolect Economic costs In Conslant 1996 Rp adjusted lo border prtces with SCF for "on-traded components
Overlays starting in 2005 and additional lanes in future years as per Bukaka plan shown !n cost stream
(a1 Bukaka May'97 estimate In constant'96 Rp
(bl (25% of normal IraHic with 50% of benefits counted
 APPENDIX 7.4

Table A 7.4.1
Ciawi - Sukabumi:Finanacial analysis as Toll Road
Traffic growth Scenario RPT Forecast
Normal growth traffic based on RPT estimate of Total corridor traffic share
Total Cost (bill Rp, Constant, 1996) 555.8

Year Project Maint. & Operating Cash Flow Income Net

costs operation Revenue Before Tx Cash flow
(a)
1996 0.6 0 0 -0.6
1997 38.6 0 -38.6 0.00 -38.6
1998 148.8 0 -148.8 0.00 -148.8
1999 187.1 0 -187.1 0.00 -187.1
2000 145.8 5.1 19.1 -131.9 0.00 -131.9
2001 35.4 10.0 50.9 5.5 1.03 4.5
2002 12.4 75.9 63.6 0.00 63.6
2003 12.6 78.3 65.7 0.00 65.7
2004 14.6 96.0 81.4 0.09 81.3
2005 11.3 15.2 99.3 72.9 7.94 65.0
2006 10.6 13.3 118.1 94.2 8.60 85.6
2007 18.9 120.6 101.8 16.77 85.0
2008 19.4 144.7 125.3 23.46 101.8
2009 16.1 148.7 132.6 33.24 99.3
2010 45.8 17.1 175.8 112.9 35.81 77.1
2011 10.6 15.5 175.7 149.7 39.90 109.8
2012 15.8 203.6 187.8 40.82 147.0
2013 43.2 17.1 203.3 143.0 46.79 96.2
2014 19.7 236.8 217.1 45.51 171.6
2015 79.0 27.7 236.7 130.0 54.13 75.9
2016 10.7 25.7 276.3 239.8 46.39 193.4
2017 20.8 276.1 255.3 58.34 197.0
2018 21.2 321.3 300.1 63.35 236.8
2019 23.4 314.1 290.7 75.88 214.8
2020 11.2 24.0 357.0 321.8 73.61 248.2
2021 10.5 22.4 346.7 313.9 82.02 231.9
2022 22.5 387.1 364.6 80.22 2843
2023 28.7 369.5 340.8 94.56 246.2
2024 23.5 352.7 329.2 88.21 241.0
2025 23.9 336.7 312.8 82.28 230.5
2026 25.5 321.4 295.9 76.75 219.2
2027 26.1 306.8 280.8 71.59 209.2
2028 25.3 292.9 267.6 66.78 200.8
2029 26.4 279.6 253.2 62.30 190.9
2030 28.6 266.9 238.3 58.11 180.2

FlRR (2023)= 15.9% 14.0%

FlRR (2030)= 16.5% 14.7%
'Project Financial costs in constant 1996 Rp. excludes price contingency & interest during construction
Depreciation and financingldebt service cash flow excluded
Overlays starting in 2005 and additional lanes in future years as per Bukaka plan shown in cost stream
(a) Bukaka/RPT Feb.'97 estimate in constant '96 Rp., includes toll equipment replacement & upgrading
Table A7.4.2

Toll Road Operatinq 8 maintenance costs

(bill. Rp)
Routine Maint. 8 Operatinq Costs Toll equipment replacement
Yr
- Adi. factor Bukaka Financial economic @ Bukaka Financial economic @
inn.@ inc. infl (base:199 0.9 Iinc. infl (base:199 0.9
8
- %
1997 1 1.080
1998 2 1.166
1999 3 1.260
2000 4 1.360 7.0 5.1 4.6
2001 5 1.469 14.7 10.0 9.0
2002 6 1.587 19.3 12.2 10.9 0.3 0.2 0.2
2003 7 1.714 21.2 12.4 11.1 0.4 0.2 0.2
2004 8 1.851 23.3 12.6 11.3 3.8 2.1 1.8
2005 9 1.999 25.7 12.9 11.6 4.6 2.3 2.1
2006 10 2.159 28.2 13.1 11.8 0.5 0.2 0.2
2007 11 2.332 31.0 13.3 12.0 13 5.6 5.0
2008 12 2.518 34.2 13.6 12.2 14.6 5.8 5.2
2009 13 2.720 37.6 13.8 12.4 6.2 2.3 2.1
2010 14 2.937 44.0 15.0 13.5 6.1 2.1 1.9
201 1 15 3.172 48.4 15.3 13.7 0.7 0.2 0.2
2012 16 3.426 53.3 15.6 14.0 0.7 0.2 0.2
2013 17 3.700 63.1 17.1 15.3 0 0.0 0.0
2014 18 3.996 69.4 17.4 15.6 9.2 2.3 21
2015 19 4.316 85.6 19.8 17.9 33.9 7.9 7.1
2016 20 4.661 94.1 20.2 18.2 25.9 5.6 5.0
2017 21 5.034 103.5 20.6 18.5 1.1 0.2 0.2
2018 22 5.437 113.9 21.0 18.9 1.2 0.2 0.2
2019 23 5.871 125.3 21.3 19.2 12.2 2.1 1.9
2020 24 6.341 137.8 21.7 19.6 14.6 2.3 2.1
2021 25 6.848 151.6 22.1 19.9 1.5 0.2 0.2
2022 26 7.396 166.7 22.5 20.3 0 0.0 0.0
2023 27 7.988 183.4 23.0 20.7 46.2 5.8 5.2
2024 28 8.627 201.7 23.4 21.0 1.1 0.1 0.1
2025 29 9.317 221.9 23.8 21.4 1.2 0.1 0.1
2026 30 10.063 244.1 24.3 21.8 12.2 1.2 1.1
2027 31 10.868 268.5 24.7 22.2 14.6 1.3 12
2028 32 11.737 295.4 25.2 22.6 1.5 0.1 0.1
2029 33 12.676 324.9 25.6 23.1 9.2 0.7 0.7
2030 34 13.690 357.4 26.1 23.5 33.9 2.5 2.2

Note: Bukaka assumes 10% inflation on OBM costs, with constant costs
in real terms; OBM assumed to increase with trafftc
Table A7.4.3

Cash Flow Proiection

II. rp.)
Toll Revenue Jasa Marqa Income Tax
Current R Const.'96 Rev. Share (30% EBT)
infl.@ (bill) % Bill.Rp Current R Const.'96
-8%
1.080
1.166
1.260
1.360 25.97 19.1
1.469 74.86 50.9 0.00
1.587 120.5 75.9 0.00
1.714 134.2 78.3 0.16 0.09
1.851 177.7 96.0 14.7 7.94
1.999 198.6 99.3 17.2 8.60
2.159 255 118.1 5% 5.91 36.2 16.77
2.332 281.3 120.6 5% 6.03 54.7 23.46
2.518 364.3 144.7 5% 7.23 83.7 33.24
2.720 404.3 148.7 5% 7.43 97.4 35.81
2.937 516.4 175.8 5% 8.79 117.2 39.90
3.172 557.5 175.7 5% 8.79 129.5 40.82
3.426 697.5 203.6 14% 28.50 160.3 46.79
3.700 752.1 203.3 14% 28.46 168.4 45.51
3.996 946.3 236.8 14% 33.15 216.3 54.13
4.316 1021.5 236.7 14% 33.14 200.2 46.39
4.661 1287.8 276.3 14% 38.68 271.9 58.34
5.034 1389.7 276.1 14% 38.65 318.9 63.35
5.437 1746.7 321.3 14% 44.98 412.5 75.88
5.871 1844.2 314.1 14% 43.97 432.2 73.61
6.341 2264.1 357.0 14% 49.99 520.1 82.02
6.848 2374.7 346.7 14% 48.54 549.4 80.22
7.396 2863.1 387.1 14% 54.19 699.4 94.56
7.988 2951.7 369.5 14% 51.73 704.6 88.21
8.627 3043.0 352.7 14% 49.38 709.8 82.28
9.317 3137.2 336.7 14% 47.14 715.1 76.75
10.063 3234.3 321.4 14% 45.00 720.4 71.59
10.868 3334.4 306.8 14% 42.95 725.8 66.78
11.737 3437.6 292.9 14% 41.00 731.2 62.30
12.676 3543.9 279.6 14% 39.14 736.6 58.11
13.690 3653.6 266.9 14% 37.36 742.1 54.21
r 2023) 1.03 1.0074

I
Table A 7.4.4b
Ciawi - Sukabumi:Finanacial analysis as Toll Road With costs reduced 10%
Traffic growth Scenario RPT Forecast
Normal ,orowth
. traffic based on RPT estimate of total corridor traffic share
Total Cost (bill Rp. Constant. 500 2

Year Project Maint. 8 Operatin Cash Flo Income Net

Costs operation. Revenue Before Ta Tax
. .Cash flow
(a)
1996 0.6 0 0 -0.6
1997 34.7 0 -34.7 0.0 -34.7
1998 133.9 0 -133.9 0.0 -133.9
1999 168.4 0 -168.4 0.0 -168.4
2000 131.2 5.1 19.1 -117.3 0.0 -117.3
2001 31.9 10.0 50.9 9.1 0.0 9.0
2002 12.4 75.9 63.6 0.0 63.6
2003 12.6 78.3 65.7 0.0 65.7
2004 14.6 96.0 81.4 0.1 81.3
2005 11.3 15.2 99.3 72.9 7.9 65.0
2006 10.6 13.3 118.1 94.2 8.6 85.6
2007 18.9 120.6 101.8 16.8 85.0
2008 19.4 144.7 125.3 23.5 101.8
2009 16.1 148.7 132.6 33.2 99.3
2010 45.8 17.1 175.8 112.9 35.8 77.1
201 1 10.6 15.5 175.7 149.7 39.9 109.8
2012 15.8 203.6 187.8 40.8 147.0
2013 43.2 17.1 203.3 143.0 46.8 96.2
2014 19.7 236.8 217.1 45.5 171.6
2015 79.0 27.7 236.7 130.0 54.1 75.9
2016 10.7 25.7 276.3 239.8 46.4 193.4
2017 20.8 276.1 255.3 58.3 197.0
2018 21.2 321.3 300.1 63.4 236.8
2019 23.4 314.1 290.7 75.9 214.8
2020 11.2 24.0 357.0 321.8 73.6 248.2
2021 10.5 22.4 346.7 313.9 82.0 231.9
2022 22.5 387.1 364.6 80.2 284.3
2023 28.7 369.5 340.8 94.6 246.2
2024 23.5 352.7 329.2 88.2 241.0
2025 23.9 336.7 312.8 82.3 230.5
2026 25.5 321.4 295.9 76.8 219.2
2027 26.1 306.8 280.8 71.6 209.2
2028 25.3 292.9 267.6 66.8 200.8
2029 26.4 279.6 253.2 62.3 190.9
2030 28.6 266.9 238.3 58.1 180.2

FlRR (2023)= 17.1% 15.2%

FlRR (2030)= 17.7% 15.8%
'Project Financial costs in constant 1996 Rp. excludes price contingency 8
interest during construction
Depreciation and financingldebt service cash flow excluded
Overlays starting in 2005 and additional lanes in future years as per Bukaka plan shown in cost stream
(a) Bukaka May.97 estimate in constant '96 Rp., includes toll equipment replacement 8 upgrading
Table A7.4.5
Ciawi Sukabumi: Cost Estimate (Bukaka estimate deflated to 1996 constant Rp.)
(Constant 1996 mill. Rp) infl. index 1 2 3 4 5
-
1.0 1.08 1.166 1.260 1.360 1.469
1996 Fin. costs Construction cash flow (1997 mill. Rp)
Construction Cost Estimates: 1996 1997 1998 1999 a 2001
Total 340181.5 4477.8 54226.7 131364.2 118845.5 31267.4

DeSian 10062.5 4477.8 5584.7

44% 56%
Construction 330119.0 48642.0 131364.2 118845.5 31267.4
Ciawi-Cicurug 95518.6 48642.0 46876.6
(%) 51% 49%
Cicurug-Cibadak 83719.3 63974.1 19745.1
76% 24%
Cibadak-W. Sukabumi 64598.3 20513.4 41153.6 2931.3
32% 64%
WSukabumi-E. Sukabumi 86282.9 57946.8 28336.1
Constr. incl. Proj. Supervision 340022.6 0 50101.2 135305.1 122410.9 32205.4
Subtotal: design & constr. 350085.0 4477.8 55685.9 135305.1 122410.9 32205.4
Toll &other equipment 16345.1 8486.9 7858.2
VAT @ 10% of total cost 36643.0 447.8 5568.6 14379.2 13026.9 3220.5
Total Constr. & Equip. costs 403073.1 4925.6 61254.5 158171.1 143296 0 35425.9

Land Acquisition 86413.6 27948.1 34504.5 23961.0

Bank guarantee for land Ac 24026.6 0 24026 0
Project Overheads 18368.0 632 5736 5000 5000 2500
Fin. Costs (Bk.fees @ 3 75% of 24026 0 24026

Total Proiect Cost excl IDC 556407 4 632 38610 0 148811 0 187132 4 145796 0 35425 9
 APPENDIX 9.1

INTERNATIONAL EXPERIENCE IN TOLL TARIFF SETTING AND

ADTUSTMENT

In many countries, the approval of tariff increases often becomes a political

matter with public pressure and elected representatives opposing increases and
objecting to fees which are obvious and immediately impact users. Rate
adjustments in Thailand are based on a pre-determined formula incorporating
the consumer price index. The Thai Parliament however sets rates on all toll
roads. Rate roll-backs in Thailand and Philippines led to collapsed or stalled
agreements with operating entities and were a factor in governments having to
offer more comfort to investors in subsequent agreements.

111 the more recent generation of BOT toll road concession agreements, - several
in countries such as Chile, Argentina, and Mexico with recent memories of
severe price inflation, indexed tariff formulas have become more common
features in pricing formulas. In Argentina, 1993 regulatory changes to encourage
more active concessionaire bidding linked tariffs to US$ exchange rates aiicl
inflation. The toll tariff can also be adjusted upwards to compensate for higher
than anticipated land acquisition costs.

In Chile, concession agreements similarly incorporate construction cost .’

escalation and cost of living indices. Government approval is required iii both
countries for subsequent tariff increases once roads are in operation. This
approval process requires a review of maintenance and operating cost increases
as well as a test of toll revenue coverage of debt service burdens.

In Mexico, recent concession and tariff policies reflect the troubled history of
earlier ventures in the 1980’s, and again in 1994/95 when optimistic traffic
estimates and investor/contractor eiithusiasm fell victim to economic
downturns. Economic crisis forced concessionaires and their bankers into
jeopardy and restructuring. Tariffs adjustments on new concession agreements
are pegged to construction price indices and for the foreign portion of debt, US$
exchange rates. Annual or bi-annual economy-wide compacts however are
intended as over-riding guidelines on general price, tariff, and wage increases.

Government Guarantees
I n the early stages of BOT contracts, several governments have offerecl
guarantees to investors in terms of tariffs, opening year traffic and traffic growth
With the increasing volume of private investment, fading memories of toll
corporation defaults, and acceptance by investors of toll roads as potentially
lucrative investments in high growth economies, these patterns of guarantees drc
becoming less prevalent. I n the mid to later 1980’s stages of the Nortl~-Soutli
Expressway programme, Malaysia offered guarantees on tariff levels and traffic
volume but in more recent programmes, the government has ceased offering
such guarantees, there being sufficient competition among investors. Opening
up the market to contractors less intimatelv involved with the ruling coalition
has also contributed to a more transparent approach less tinged with favoritism.

I n this process of moving towards a more open and competitive tendering

svstem, more risks have been accepted by the investor and financing costs have
probably been lower as a result. Under present policy, the government pledges
to review tariffs every two years and render prompt ciecisions on requests for
tariff increases. While government decisions on increases are not necessarily of
the magnitude requested by concessionaires, agreements cammonlv contain r\
coiisumer price index component.

In the Philippines, toll roads have been treated for regulatcry purposes as a
public utility with the Toll Regulatorv Board (TRB) having responsibility over
toll tariffs. The TRB has traditionally applied two tests, a “reasonable” rate of
return on equity to investors, and the impact of increases in debt service,
maintenance and operating costs occasioned through inflation or depreciation of
the peso. Tolls are to be adjusted according to official government indices as
defined through bidding documents and incorporated in the contract.

In the Philippine BOT law of 1990, the concessionaire is authorized to collect

”reasonable” tolls, but the TRB must approve the” fairness and equity” of the
charges. Under the Philippine BOT law, the rate of return criterion can also be
used to adjust toll tariffs to cornyensate the contractor for undue delays, caused
for example by land clearance. This act also provided that tolls should be
collected for only a fixed term, not to exceed 50 years.

To attract foreign funds and more competitive bids on recent toll road projects,
predetermined rate formulas are set out in concession bids and agreements.
For the Manila South Tollway, concession bids considered formulas with several
components. These components include indexing to consumer prices, foreign
exchange rates and international reference rates such as the London Interbank
Borrowing Rate (LIBOR). The weighting of these latter two components would
then vary over the concession period according to the foreign exchange share of
debt outstanding and adequate coverage of debt service.

Because of the thinness of the domestic long term funds market, domestic loans
were prominent only in the shorter term elements of the financing package.
This sourcing of funds hence has an impact on not only the cost of capital but
also the means for securing coverage of foreign currency debt repayment. To
secure more foreign fund participation in financing packages, the government
extended assurances that foreign payments would be protected by a
foreign/local currency adjustment factor in addition to inflation indexing in the
toll tariff rate formula.

Determining ”Reasonable Rates of Return”

Setting rates on the return on capital investecl was a traditional iiiethd ot
regulating utility and railwav tariffs in North America in pre-cleregulation davs.
While superficially attractive as a means of curbing “excessive ” profits, this
approach had two major drawbacks. Firstly, determining an appropriate rate
base means extensive involvement by the regulatory body into the affairs of the
operator in monitoring the efficient use of capital, the correct valuation of the
capital base, and the appropriateness of expenses and credits. I n turn
management has less incentive to control costs, and often a pattern has
developed of investing in uneconomic excess capacity to inflate the rate base.

To service this regulatory mandate has often required the build-up of an

extensive review and approval system ancl an accompanying bureaucratic
structure to carryout the sometimes elaborate procedures of tariff adjustment
review and approval. Common questions to be addressed in the more extensive
forms of rate of return regulation, particularly in the transportation industrv
include the following:
- what is the proper capital base on which the “reasonable” rate of return
(ROR) is to be calculated?
- what is the current and embedded cost of capital, equity and debt, to be
considered in the ROR?
- are new capital expenditures justified in terms of timing, cost, ancl scale?
- are maintenance and operating expenses justified and reasonable?

This extensive blanket of regulatory review which grew up in inany countries in

early to mid- century has fallen out of favour in the last two decades because the
accoinpanying burden of regulatory bureaucracy was found to lead tcl
ballooning administration costs on the part of both regulator and inclustrr.
Further a i d more damaging in rapidly changing economies was the stifling of
innovation and the rigidificatioii of organizational and operational structures.
Managerial incentives to control costs withered ancl tariffs in many cases becaluc
higher because of the inefficient cost structures they were based 011. The end
effect in many cases was the “capture” of regulators by the utility being
regulated and increased opportunities for corruption or “sweetheart” treatment.
Mature and Secure Revenue Streams
As the toll road investment matures, traffic and revenue streams will build a n d
the concessionaire enters the "period of eiijoyment. Investors start receiving
"

returns on equity, and higher cost debt has been paicl off or restructured to inore
attractive terms. With the emergence out of the initial period of .uncertaint!s,
revenues then give secure and comfortable margins of debt service coverage.

It is this period, usually in the second decade of a successful investment when

public ancl governmental concerns over excessive profits and "asset milking" arc
likely to rise. Looking ahead to this golden horizon, government regulators c;ln
build into concession agreements a review ancl reassessment mechanism to
moderate or cap escalation formulas originally contracted. Foreign exchange
adjustment factors in the tariff setting formula shouId decline in weight as debt is
retired. Operating and routine maintenance costs should also decline in
proportion to revenues as the investment matures. Sharing of profits is
provided through Jasa Marga joint venture agreements, both through diviclencis "'".,

and in more recent agreements, an increasing share of revenues after 10 years

with the build-up in traffic. As discussed previously with rate of return limits,
limiting or taxing away profit off-take can lead to either inefficient operations, or
"empire-building" to soak up the cash throw-off. Notable examples of the latter
include real estate ventures and office buildings funded by more than one bridge
and tunnel operating authority in the U.S.

This review device should be also be tied to the remaining life of the concession,
and the rehabilitation and reconstruction expenditures anticipated to renew the
economic life of the asset. To be avoided are restrictive legislation or regulatory
strictures on tariffs which may jeopardize the floating of new bond issues to
finance needed rehabilitation. This was a another problem encountered in the
financing of aging segments of toll roads in the U S . during the 1970's.
 APPENDIX 9.8.1

June 26,I997

Mr. Ion A. Johnson

Transport Economlrt
ALM Rasources Ltd.
P. 0,Box 8OY
6808 Paocoe Road
Sooke,6 C VQS IN0
Dear Mr.Johnson:

I 801writing In responae to your latter to Mr.Dori Stonehouse regcwrding the tall

structure for Highway 104 in Nova Scctia.

I have included a copy of tho relevant section of the Omnibus Agrebment

between the Provlnw and the Highway 104 Coi'poration which oullines the toll
adjustnmnt rnechonism.
The initial toll for the highway has been set at $3.00 per oar and $2.00 per wle
per truck, These rates were determined through a "Stated Preference Survey". In this
survey approximately 75% of cBr drivers etated they would be willing to pay the $3.00
and silghtly o w 50% of trucks the $2.00/erxle rate. The survey indicated slightly high
revenues could be achlaved with B higher toll but with the coritroversy surrounding
tolling the lower toll with a higher acceptance level was chosan.

Govnrnment provided Q $55,000,000 subsidy to the project; the remaining

$66,000,000was raiod througli toll revenue bonds. The financial model developed for
the rspeyment of the bonds takes into acoount tho roads anticipated revenue, costs
(rnalnternance,to\\collaction,bond repayment, etc.) and projected traffic growth. The
toll inorlase as llsted in section 5.l(b)(li) of the Omnibus Agregrnent are the toll
increased and the dates on which thoy will'occur. These increases are in line with the
anticipated Inflation rate. The rnechan!sm for increasing toll or foregoing an Increase
should the revenue or costs VGIY from the model are as given in the Omnibus
Agreement.

....12
 * 32 -
49, Bond and Notc Einanclag Documetlts Approved

The Province hereby approvsa of the Corporation entering into, executing, delivering and
perfonnidg 1s obligations under ewh oftbr Senior Bond Indenture, Junior Bond Indenture,
Note Indenture, Undenwiting Agreement, Consmiction Trust Agreement. Insurance Tmst
Agreement, and Mejor Maintenance Reserve Fund Agreement. The Province ackmwledgea
receipt of notice ofthe assignment to Montreal Trust Company of Canada, Montreal Trust
Coit\pany and The R-M ‘l’rust Co&any of thdProviace’sobllgations under this Omnibus
Agreement and agrees that each of such trustees shall be entitled to exercise each and wery
right of the Cnrporalian hereunder BS if it wcre a pmty hereto.

50. Initial Tall Rutos

Pursuant to Section lO(3) of the Highway 104 Wesfern Alfgnrnrnf Act, the Corporation
hereby sets thc initial Toll k t e s for the operation bn the Wcslern Alignment of the following
classes of vehicles:

(a) Trucks, Ruses, Sdlaol Buses, Farm Tractors, ‘rruckTractors, Trailers, Sami-
trailers, Pale Trailers, ury Vehicle not nrentioncd in (b) or (c):

SZ.00 per arlr

(b) Passenger c a m Taxicabs, vms other thnn those with oommtrcirclal licences,
motorcyclea:

$3.00 par vohfcle

(c) Recreation vc21lcles,motohomes (regardless of number of axles), ~ v etrailer

l
and pusenger car GOmbhatiOnS, travel trailer and two d e truck combinations:
$4.00 par vehicle

Ihc Province acknowledges that this Onmibus Agreement is an agreementpursuant to the

Act for the puqmses of Secthn lO(3) oftbe Act.
-.
-. -
 - 33 -
Tenns with initial capitals used in this Section 50 that are not defined in Schedule "A" have
rhe meaning attributed to them by the Mufor Vehicle Act (Nova Scotia) 8s now in effect.

s 1. nilKate Adjustment

(a) After h c Date of Acceptance, the CorpoMtion $hall calculate t!mDebt Strrice
Coverage Ratio no less fi'equently than qwterly. Such calculation shall be delivered to the
Bondholders' Represeiitative no later than 30 days after the elid of each calendar quarter.
 . 34 *

(b) The Toll RHtcs for each class of vehicle shall bc increased 8s follows:

(1) &!&t2aU~& On Janiiary 1 in each calendar year, beginning with the

second calendar yenr dler the Projected Completion Date, the Toll Rate
for each c l w of vehicle shall be increased by tha percentage inorewe,
if any, in the Concumer Price Index dwing the period fromthe later of
October 1, 1998 or the October 1st immediately preceding the last toll
adjustment under this Sc'Ftton, to the September 30th immediately
preceding the data of UIC detemination, provided that such increase
amounts to at \ a t 25yt.
Ifthe application of the foregoing formula would produce an iporcase
grcrrkr than 25$ but lcss than 506, the increase shall be 256; if the
applic#tion of such fonnul~would produce an increme of tmre than
SO$ but lcss than 756, the increase shall be 506, and so on.

(ii) - f& Whehcr or not here is any increase

under ParafCraph (i), the Toll Rate in respect of erch class of vehiales
shall be increwd by the following amounts:

-
On January 1,2001 506;
011 1,2004 - 256;
011 1,2006 - SO$:
J~IUWY
On January 1,2012 25$; -
On January 1,2017 - 508;
On January 1,2022 506. -
Notwithstanding the preceding sentence, if tlic Debt Service Coverage
Ratio Is greater than 2.3:1 in December of any year to and includiig
2004, or i~ W t e r than 1.9:l in any year thereafter, and thc
Noteholder's Rapreaentative so directs, the Carporallon will not
~ C E R E C tho 1'011 Rate h the next followlng year.

(iii) J $ " U & W , Each year aAer the

dotorinination of the Annual Corporation Budget, the Corpuration shall
formast tall revenues for the enwing yew (tawng account of any
Incroaaer resulting &om the operation of Pwgraphs (f) and (ii)) and
shall compare such forowst with thr ne@ for fun& to pay the amounts
required by such Annual Corporation Budgot, to make rwuired
 -35-
payments to the Major Maiatcnance Reserve Fund and the Debt Service
Reserve Account, to PRY inahring Bonds and Notes and to achieve the
Pro Forma Debt Service Coverage Ratio: to the extent that forecasted
toll revenuw, aAor exploring all reasonable options to reduce cxpc~ises~,
we ineuMcient la provide all required funds, the Toll Rate for each
class of motor vehicle shall be increased by such amount as may be
necessary to ensure that U w c is no deficit and to ensure that the Pro
Forma Debt Service Coverage Mi0 is ut leakt 1.2:1 during the b t 36
months after the Date of Acceptance md 1.3: I at any time thareafter;
all such increases shall be in units of 25 cents or integral multiples of
25 cents and if needed, any such iiicrease shall be rounded up to the
next h l g i multiple
~~ of25 cents.

(iv)-oilaR_ePorevoC_eclvre~- If the Debt Service Coverage

Ratio is less than 1.2:1 at any time dwhg the fitst 36 months following
the Date of Acceptance or less than 1.3: 1 at any time thereafter, then a1
any time within 12 months after receiving the Corporation's calculation
of the Debt Service Coveragc Ratio, the Bondholders'Representative
may, in its discretion, dmct the Corporation It) increase the Toll Rates
to an amount estimated by thc Bondholders' Representative to achieve
such Debt Swlco Coverage Ratio. The Corporation shall comply with
such dirwtion not later than the first day of the month which begins at
least 30 days after the receipt of tha direction er 011 suoh Later date as
thb Bondholders' Representative may direct. The dlrmtion may, but
need not, provide far tbe implementation of the Toll Rate increase on
a day on which a Toll Rate increase might otherwise takc p l w
pursuant to thitr Saction 5 1. Any such direction shall provide that tho
Toll Rates for each class of motor vehicle shall bc iricnaaed by a;\
equal amount. The Bondholders' Representative may not rrive more
than one direction pursuant to this Subsection (iv) in any 12 month
period. Prior to glvlng such a direction, the Bondholders'
Rc,presciktativoslrnll mok the recommendation of UIIndependent mfic
consultaut with wspect to the appropriate increase in Toll Rates and my
JirWion given by the Bondholders' Representative shall be consistent
with such rccommcndation, bur need not follow suoh muommendstion
precisely. 'he cost of such indopandent traffic consultant shall be
borne by the Corporation.
 - 36 -
(c) If the 'roll Rates haw been increucd pursuant to Subsection (iv) and within
12 month6 thereafitherc is H decline in the traffic on the road for any class of motor
vehicles or a decline 11 h e toll revenues derived therefrom, the Bondholders' hpresentative
may, 111 its discretion, direct the Cor&mtlol; to decreeso the Toll Rate for one or more classes
of vehicle. The Corporation shall comply with such direction on the ficst day of the first
month which begins at least 30 days after the rcceipt of such direction or on such later date
u the Bandholdad Representative niay direct. The Bondholders' Kepresenutivo shall only
i
direct that one or mare Toll Rates be decreased if an independent traffic cansultanc retained
by it advises that suoh a decrease will help maximize the revenues derived froom tolls. The
cost of such independent hafflo consultant shall be borne by the Cwporation.

(d) Notwithstanding any other provisions contained in this Part NU,if there is a
change in provincial low, regulation or ordinance or if any rule or order is issrrcd by any
agency or instrumentality of the Province which has the direct result of incrwing t$e
Operating Expenses of the Corporation to an amwmt in excess of 110% of the Opcratiilg
Expenses in tlrc current Annual Coyoratfo? Budget, the Bondholders' Representative may,
ifi addlilon to any other riglit to r q u h an adjustmcnl of Toll Rates, direct that the
Corporation increase h e 'I'd!Rates to the O X I C I ~iiecessary
~ to recover such increase in cmts
and the Corporfition shall comply with suclr request on the first day of the month wlrioh
begins at lewt 30 d ~ y safter the reccipt of the direction.

52. .Toll Collection

The Corporation shall collect either from the driver thereof In cash or by way of charging a
pre-establlshcd account using transponden or other devices, the appropriate toll depending
upon the clans ofvehicle as set out in Section 50, at the time such vehicle pmses through the
toll colleotfon fficilities. The Corporation's toll collection activities in oach calendar year
shall m e t ot emeed the relevant Operations Performance Criteria.

53. Enforcement of Laws

The Province will enforce laws and rcgulattions for the collection of tolls and LO cii~urot!wt
e f f d v r on attd after thc Date of Acceptance

Da I:s w 9 . 1