Social Cost-Benefit Analysis

Tamil Nadu Urban Sanitation Improvement Plans

A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation
Improvement Plans

Introduction
In recent decades, the world has made significant progress in enhancing sanitary
conditions. Sanitation conditions significantly improved between 2000 and 2017,
according to a 2019 report from the Joint Monitoring Programme (JMP), a
commission established by the World Bank and the United Nations to track global
progress in water, sanitation, and hygiene issues (WHO 2019). First, there has been a
sharp decline in the number of people who use open defecation, from 1.3 billion to
673 million. Second, from 56% to 74% of people, more people now have access to at
least a minimum degree of cleanliness.1. Notwithstanding these successes, the JMP
study also identified a number of challenges facing the attainment of the Sustainable
Development Goals (SDGs) pertaining to sanitation. First off, it is alarming that only
96 countries regularly gather statistics on the percentage of their population that has
access to adequately managed sanitation services, considering that SDG 6 calls for
universal access to such services by 2030. Second, some demographic groups have a
somewhat lopsided influence on the advancements made in sanitation access. More
specifically, 70% of people live in rural areas without access to basic sanitation,
suggesting that most of the earlier advancements came from urban areas.
Furthermore, people in the greatest wealth quintile in a given country have at least
twice as much access to basic sanitation as people in the poorest quintile. This is true
for all nations with disaggregated wealth data. The 2019 JMP's main takeaway in this
regard is that, despite the fact that there have been tremendous advancements over
the previous 17 years, more funding is still required to expand access to sanitation
for people that have been consistently underserved.

The study is an urban fecal sludge management (FSM) initiative run by the Indian
state of Tamil Nadu. According to the results, this FSM project has a net present
value (NPV) of $25.3 million and a benefit-cost ratio of 1.22. This can be translated
into the state benefiting economically and socially by about $1.22 for every dollar
invested in the project. Nonetheless, significant heterogeneity concealed by the
positive NPV was also disclosed by the CBA. First, a sensitivity study called a Monte
Carlo simulation revealed significant variability, with 30% of the trials yielding
negative net present values (NPVs). Thus, there is a non-trivial possibility that the
FSM project won't be financially feasible. Second, I discover that the two biggest
factors influencing the variability in NPV calculations are "baseline diarrhea
incidences" and "change to diarrhea incidence rates" utilizing tornado analysis. These
figures also show a deficiency of reliable data, underscoring the pressing need for
governments to improve data monitoring procedures. Finally, the FSM project is
divided into two autonomous phases, the largest of which serves densely populated
cities and has a higher net present value than the phase that concentrates on less
populous peri-urban areas. Therefore, in trying to reach out to underrepresented
people, there is a conflict between equality and efficiency that this case study brings
to light.

The case study's remaining sections are arranged as follows: The project and site's
history is covered in the first section. A succinct explanation of the CBA approach and
the related expenses and advantages of the FSM project come next. The report then
goes on to show the CBA results and includes sensitivity analysis. Finally, we wrap up
and go over the results.

Background
With a population of over 72 million people (ranked sixth in the nation) and a
geographic area of over 130,000 km2, Tamil Nadu is one of the largest states in
India.2. Tamil Nadu is a significant economic region in India, boasting the second-
highest GDP in the nation at over $260 billion, and home to numerous major cities
with a focus on technology services. Tamil Nadu has been playing catch-up when it
comes to access to better sanitation facilities, despite its economic might (see Table
1 for a review of the state's sanitation statistics over three data waves). With only
about 35% of its people having access to toilet facilities, Tamil Nadu placed 24th in
the nation in the 2001 census—one percentage point worse than the national
average. Tamil Nadu was placed 30th and 32nd, respectively, when the views of the
rural and urban people were considered individually. The percentage of the state's
population who have access to sanitation increased significantly to 48.3% according
to the 2011 census. Tamil Nadu's rating among all Indian states fell three spots to 27,
indicating that this state is growing at a slower rate than other states. According to
the most recent data from the National Family Health Survey's 2014–2015 edition,
Tamil Nadu has steadily improved since then. Currently, 52.2% of the state's
population has access to sanitation, outpacing other states in terms of rate of
increase as they rank 23rd in the nation. Tamil Nadu's urban population, which
currently stands at 19th place, is mostly responsible for this improvement.

The Tamil Nadu government outlined a thorough, two-part plan to enhance

sanitation services in its 12th 5-year plan, which ran from 2012 to 2017.3. First, the
Tamil Nadu government will launch the Nirmal Bharat Abhiyan Programme (a
continuation of the Total Sanitation Campaign) to build toilets throughout rural
communities utilizing a demand-driven method in order to continue expanding
access to restrooms in rural regions (Hueso and Bell 2013). In order to guarantee
user safety and increased adoption rates, the Tamil Nadu government will also build
gender-specific sanitary complexes in communities in addition to infrastructure.
Likewise, teachings on hygiene and sanitation will be added to school curricula to
instill in students the value of better sanitation from a young age.

Second, Tamil Nadu is giving improved sewage waste management a priority for its
urban areas. One factor is the urgent need to increase cities' capacity to safely
manage and dispose of sewage, since more urban families in Tamil Nadu have access
to toilets. In order to achieve this, the government will modernize current
infrastructure to boost sewage treatment capacity and construct sewage treatment
facilities to service 31 urban local bodies (ULBs).

State Investment Plan

In order to carry out urban sanitation as outlined in the 12th 5-year plan, the Tamil
Nadu government commissioned the Tamil Nadu Urban Sanitation Support
Programme (TNUSSP). In order to attain these goals, the TNUSSP created a State
Investment Plan (SIP), which outlines a five-phase plan for achieving the goals related
to urban sanitation (Table 2). The SIP specified the amount of capital and
infrastructure required, along with the number of households who would benefit at
each step. This case study aims to perform a CBA of the SIP's initial three phases. The
study focuses on these three phases for two key reasons. First, over these three
stages, almost 75% of the SIP's total beneficiary households will be contacted. It
follows that they are by far the most important component of the programme.
Secondly, only Phases 1 through 3 are covered in full in the current SIP paper.
Therefore, it is more practicable and policy-wise to perform a CBA of these three
phase

The general idea behind Phases 1 and 2 is to treat sewage using the STPs that are
currently in place. Sludge from septic tanks and toilets is now handled by many
outdated STPs that are ill-equipped to handle it; the sludge is either moved to other
locations or discharged untreated into rivers. Therefore, the goal of Phases 1 and 2 is
to modernize the pumping stations and decanting facilities that are now in place in
order to enable them to manage fecal sludge material from surrounding sources. In
turn, about 14 million people and over 86 cities, or ULBs, will profit throughout this
phase.

However, Phase 3's goal is to build new faecal sludge treatment plants (FSTPs) in
cities or ULBs that don't already have any. It is projected that 49 additional FSTPs will
be built, benefiting 110 communities and 3.7 million people in total.
Method

Cost Benefit Analysis

One of the most popular methods for determining a project's economic viability is
the CBA (Boardman, et al. 2017). The project will have positive economic benefits if
the net present value (NPV) is positive, and vice versa. A decision maker would
typically select the project with the highest net present value (NPV) when utilizing a
CBA to compare various projects, ensuring that the most economically efficient
option is selected (Boardman, et al. 2017).

The following are the measures that were done in this case study to conduct the
CBA:

1. Ascertain your position

2. List the advantages and costs of the project
3. Describe the procedures used to determine the costs and benefits.
4. Set parameters for the expenses and gains.
5. Decide on the right discount rate and project duration.
6. Determine the net present value.
7. Carry out a tornado analysis and Monte Carlo simulation.

Costs and Benefits of the Project

The underlying issue, the project's costs and benefits, the formulae that go along
with them, and the guidelines for determining their worth are the first four steps of a
CBA that are covered in detail in this subsection.

Standing
The CBA, which asks "Whose costs and benefits are in consideration?" is the question of
standing. It makes sense that Tamil Nadu has exclusive standing in this case study; that is,
only the expenses and gains borne by the state's citizens, enterprises, and government
would be taken into account.
Costs components
In the second step of the CBA, we list the project’s costs and benefits. We first focus on
the costs.

Capital costs
The building or capital expenses of new sanitation infrastructure, such as sewage
treatment facilities and pumping and decanting stations, make up the first cost
component. Phases 1 and 2 of the SIP call for the renovation of 91 pumping stations
and 55–85 decanting stations at an estimated cost of Rp35 million. Phase 3
construction expenses are significantly higher because it adds 49 additional STPs,
which will come with a price tag of approximately Rp2 billion.

Soft Components Cost

Other than capital costs, the successful introduction of sanitation infrastructure
requires support systems, such as technical assistance, and information and
education campaigns. According to the SIP, it is estimated that these soft
components account for 8.5% of total capital costs and consist of technical
assistance (2%), project management expenses (4%), and information and education
campaign expenditures (2.5%).

Sewage Trucking Costs

As the main objective of the SIP is to increase fecal sludge treatment capacity, it is
implicit that more waste will be sent for treatment and disposal. As such, we need to
include the costs for handling additional sewage disposal (Table 3 row 3). To do so,
we first determine the per-trip sewage trucking cost by relying on a survey conducted
by the TNUSSP on trucking operators (TNUSSP 2018). This consists of variable costs (e.g.,
fuel, truck maintenance) and fixed costs (e.g., cost of truck and pumping equipment). The
per-trip variable cost (𝑇𝑟𝑖𝑝𝑇𝑆𝐶𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒) is computed by dividing the annual variable cost
by number of trips taken by an average sewage truck in a year. On the other hand, the
per-trip fixed cost is computed by dividing the lifetime fixed costs of a sewage truck by
the product of its expected lifespan and average trips.
Second, we multiply the per-trip cost by the total number of additional disposal trips
(∆𝐷𝑖𝑠𝑝𝑜𝑠𝑎𝑙 × 𝐻𝐻𝑇𝑏𝑒𝑛𝑒). The former is determined by (i) the difference between
the ideal disposal rate per household-year (0.4, or once every 2.5 years) and the
current disposal rate (0.25, or once every 4 years) (TNUSSP 2016); and (ii) the total
number of household beneficiaries in each phase of the SIP

Operational and Maintenance Cost

Other than sewage trucking costs, the treatment plants also need to incur additional
operational and maintenance costs in treating the increased fecal sludge material.
The formula for the sewage treatment cost consists of the amount of additional fecal
sludge produced per year and the per- unit cost of treatment. The former is obtained
by multiplying the change in disposal rate by the number of household beneficiaries
and average tank size (∆𝐷𝑖𝑠𝑝𝑜𝑠𝑎𝑙 × 𝐻𝐻𝑇𝑏𝑒𝑛𝑒 × 𝑇𝑎𝑛𝑘). The latter (𝑇𝑟𝑒𝑎𝑡𝐶𝑜𝑠𝑡) was
obtained directly through correspondence with the TNUSPP (author of the SIP
document). There is a big distinction in fecal sludge treatment cost for STPs and
FSTPs. According to the TNUSPP, STPs can process and treat fecal sludge at a cheaper
rate ($0.04/m3) compared to FSTPs ($2.14/m3).
Benefit Components
We now focus on the benefits of the SIP. To this end, we are only considering the
health benefits of improved sanitation disposal in this case study, as these are the
most recognized and well-studied benefits of sanitation improvements (Haller,
Hutton, and Bartram 2007; Hutton and Haller 2004).

Mortality Benefits
The mortality benefits (Table 3, row 5) are determined
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛
) by three components: the
change in the disease incidence rate (∆𝐷𝐼); (ii) the baseline mortality rate
(𝐵𝐿𝑇𝑚𝑜𝑟𝑡); and (iii) the value of a statistical life (𝑉𝑆𝐿). To estimate the decrease in
disease incidence rate, we should ideally have information on the extent to which
water quality has improved as a result of the safer management of fecal sludge.
However, because baseline information on water quality in Tamil Nadu is unavailable
and there are no impact evaluation studies projecting potential improvements to
water quality, we are thus assuming that there is 25% reduction in mortality rates.7
Second, we attempt to retrieve the baseline mortality rate of diarrhea diseases from
official statistics collected by the Health and Family Welfare Department of Tamil
Nadu government.8 However, the statistics showed that even though there were
over 100,000 cases of diarrhea diseases, mortality remained at zero for several years.
As such, we conservatively estimate the mortality rate to be around 100 deaths per
year across the entire state.9 As the SIP focuses mainly on urban areas, we attribute
the baseline mortality rate to this specific segment of population by multiplying the
proportion of resident beneficiaries to the entire population in the state. To this end,
the baseline mortality rate from diarrhea is estimated at around 36 deaths per year.
Lastly, the value of a statistical life (VSL) in India is calculated by a conventional
benefits transfer formula where we multiply the VSL from the US Environmental
Protection Agency with the ratio of India’s per-capita gross domestic product to that
of the US, and finally raise this parameter to the income elasticity of VSL. In this
regard, the VSL in India is estimated at around $1.47 million.
Morbidity Benefits
The morbidity benefits are calculated in a similar manner as the mortality benefits
(Table 3, row 6), and are determined by three main parameters: (i) the change in the
disease incidence rate (∆𝐷𝐼), (ii) The baseline morbidity rate (𝐵𝐿𝑇𝑚𝑜𝑟𝑏); and (iii) the
cost of illness (𝐶𝑂𝐼). The baseline morbidity rate is estimated at around 165,000
cases for the entire state. This figure is obtained by averaging the annual diarrhea
cases computed by the Tamil Nadu government from 2010 to 2018. Similarly, we
apportion the baseline diarrhea cases to Phases 1 and 2, and Phase 3 separately, i.e.,
around 35,000 cases and 9,000 cases, respectively. Third, the COI is taken from a
review study, who found that the average cost of treating outpatient diarrhea cases
in developing countries is around $37. The COI consists of medical costs, the
opportunity cost of the time of the patient, and the opportunity cost of the time of
the caregiver.
Discount Rate and Project Timeframe
As projects typically last multiple years, it is necessary to discount the future streams of
costs and benefits to bring them to present terms. In this regard, the discount rate is set
at 6% per year. This is also the same rate used by the Asian Development Bank for
evaluating sanitation projects. As for the project timeframe, the SIP did not specify how
long the infrastructure is expected to last. As such, we assume the project will last for 40
years since other sources found that treatment plants in India typically last around 40–
50 years

Monte Carlo Simulation and Tornado Analysis

One of the main difficulties of conducting CBAs is that many parameters are not
deterministic, i.e., they take a range of values as opposed to one single value. As such,
we should not simply rely on a single set of parameter values to form definitive
conclusions about the project. A thorough way of conducting sensitivity analyses is a
Monte Carlo simulation where each parameter is given a statistical distribution (e.g.,
uniform, normal) and the moments or range of the distribution (e.g., we would
specify the mean and standard deviation for a normal distribution). The Monte Carlo
simulation is implemented by randomly drawing a value for each parameter based
on their assumed statistical distribution and computing an NPV for this set of
parameters. This exercise will be conducted 10,000 times, which in turn will yield
10,000 NPVs. These NPVs can then be summarized using a cumulative distribution
function. The purpose of the Monte Carlo simulation is to provide a realistic
distribution of NPVs for the project by taking into account the non-deterministic
nature of most parameters.

Results
We first present the main findings. Table 4 contains the present value of each major
component of the costs and benefits of Phases 1 and 2 and Phase 3. First, we see
that the overall NPV for all three phases is around $25.3 million. This means that we
expect the SIP to yield a social benefit of $25.3 million over the project’s lifetime of
40 years. Second, when examined by individual components, the largest contributor
to costs is sewage trucking, which accounts for around 43% of the total NPV cost. On
the other hand, despite the small number of avoided diarrhea deaths, mortality
benefits constitute the largest share of total NPV benefits (around 96%). This is due
to the relatively large value of a statistical life (VSL). Third, as we had computed costs
and benefits according to each phase of the SIP, we can examine their individual
NPVs. Toward this end, we see that the overall positive NPV was mainly driven by
Phases 1 and 2 as this delivered an NPV of $67 million. In contrast, Phase 3 yielded a
NPV of –$42 million. The main reason for the stark difference in NPV across the two
main phases is that construction cost is low in Phases 1 and 2 (around $470,000) and
much higher in Phase 3 (around $27 million). Similarly, sewage treatment costs are
much higher in Phase 3 despite serving fewer beneficiaries. This is because the
sewage is treated in FSTP rather than STP in Phase 3. Fourth, due to the difference in
scale between Phases 1 and 2 and Phase 3, we can also restate the earlier results in
terms of benefit-cost ratio for easier comparison.10 The benefit-cost ratio for Phases
1 and 2 is 2.62. This means that the state of Tamil Nadu can expect a return of $2.62
for every $1 invested. On the other hand, Phase 3 has a benefit- cost ratio of 0.4,
indicating a loss of $0.6 for every $1 invested.
Monte Carlo Simulation and Tornado Analysis
As mentioned earlier, the parameters used in CBAs are mostly non-deterministic. In
this regard, even though we had obtained an overall positive NPV using the mean
values of each parameter, it does not necessarily imply that the project would
invariably be positive regardless of what values the parameters take. One way to
incorporate the many different values that each parameter can take is via a Monte
Carlo simulation.

We ran 10,000 trials where a value for each parameter was picked randomly for each
trial. In turn, the NPVs from these 10,000 trials are summarized using a cumulative
distribution function in Figure 1. Around 46% of the trials resulted in negative NPV.
This is in clear contrast to the highly positive result we obtained if simply relying on
the mean value
We can also run separate Monte Carlo simulations for the individual phases to
uncover any heterogeneity between these two main phases. The NPVs for Phases 1
and 2 display a wide variance as they range from –$50 million to over $300 million.
More importantly, around 20% of the trials returned a negative NPV, suggesting that
even this part of the SIP is not entirely positive. For Phase 3, around 90% of the trials
are negative. Again, it should be highlighted that this implies a non-trivial portion of
the trials (10%) for Phase 3 returned positive NPV. In all, the Monte Carlo simulations
suggest that there is a wide range of values the NPVs could possibly take, and there
could be certain parameters highly influential in determining the outcomes.

As such, we further implement a tornado analysis to examine in greater detail which

parameters have outsized impacts on the NPVs. Figure 2 shows the outcomes of the
tornado analysis where the parameters with largest impacts are displayed in
descending order. The top two parameters with the largest explanatory power in the
variability of the NPVs are the baseline death rate and changes to diarrhea
incidences. For instance, the NPV at the 5th percentile value of the baseline death
rate is around –$80 million and is around $180 million at its 95th percentile value.
The range in NPV caused by changes to diarrhea incidence is similarly large. In
comparison, the next four most influential parameters (VSL, amount of sewage
disposal, income elasticity of VSL, and COI) have a much smaller impact on the NPV’s
range. The tornado analysis thus provides additional insights on which parameters
play the most important role in the CBA, and help us focus on better data collection
efforts for these parameters.
Discussions

This case study examined the economic

viability of a fecal sludge management
project using CBA. Specifically, we
examined an urban sanitation
improvement plan currently implemented
by the Tamil Nadu government. The main
objective of this plan is to upgrade and
construct fecal sludge treatment plants to
meet growing demand for sewage
management across cities in the state. We
analyzed the first three stages of a 5-phase
program, and found a positive NPV of $25.3
million. However, this highly positive NPV is
masked by heterogeneous outcomes to the
individual phases. Specifically, Phases 1 and
2 has an NPV of $67.4 million and Phase 3
delivers a negative NPV of –$42.1 million.
Further sensitivity analyses using a Monte
Carlo simulation confirm that Phases 1 and
2 outcomes tend toward being positive and
Phase 3 toward negative. Lastly, we also
implement a tornado analysis to examine
which parameters are most influential in
determining the NPVs. We find that health
measures such as the baseline diarrhea
mortality and improvement to diarrhea
incidences play outsized roles in NPVs
calculations.
There are several policy implications arising from the CBA results. First, it is not
surprising that Phases 1 and 2 and Phase 3 have contrasting outlooks. The former
leverages on existing infrastructure to benefit more households at much lower
construction costs. On the other hand, Phase 3 involves construction of costly FSTPs
that serve smaller population. In this regard, this result highlights the tradeoff
between efficiency and equity. Even though Phase 3 has a negative NPV, it does not
necessarily imply that the project should not be carried out. One way for policy
makers to justify projects that benefit smaller or more disadvantaged populations is
by using weights.

Second, the parameters highlighted as most influential in the tornado analysis are
statistics where it is difficult to obtain credible information. For instance, the
mortality rate from diarrhea diseases, while reported by the Tamil Nadu
government, is suspect as there are zero reported deaths for several years despite
still having over 100,000 cases of diarrhea annually. As such, for the purpose of this
case study, we needed to assume a range of diarrhea-related mortality. Similarly,
because there are no reliable studies on projected improvements to water quality
following the SIP, we need to assume a wide range of improvements from 5% to
50%. In this regard, this case study echoes the 2019 JMP report’s call for better
monitoring data so that policymakers and analysts can better evaluate projects
performance and viability.
 A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Introduction
The world has made great strides in improving sanitation conditions in recent decades. According to a
2019 report by the Joint Monitoring Program (JMP), a taskforce formed by the United Nations and the
World Bank to monitor global progress in water, sanitation, and hygiene issues, there were significant
improvements to sanitation conditions from 2000 to 2017 (WHO 2019). First, the population practicing
open defecation has fallen drastically from 1.3 billion to 673 million. Second, access to at least a basic
level of sanitation has increased from 56% to 74%.1 Despite these achievements, the JMP report also
highlighted several obstacles to achieving the sanitation-related Sustainable Development Goals (SDGs).
First, given that the objective of SDG 6 is universal access to safely managed sanitation by 2030, it is
concerning that only 96 countries actively collect data on the proportion of their population having access to
these services. Second, the improvements made to sanitation access are to some extent skewed by
certain segments of the population. Specifically, the 70% of the population lacking access to basic sanitation
are from rural areas, implying that the earlier improvements were mostly from cities. Moreover, among
countries with disaggregated wealth data, residents in the richest wealth quintile have at least twice as much
access to basic sanitation compared to the poorest group. In this regard, the overall message of the 2019
JMP is that while significant improvements have been made in the past 17 years, more investments are still
needed to further sanitation access for persistently underserved populations.

study is a city-wide fecal sludge management (FSM) project implemented in the state of Tamil
Nadu, India. The findings indicate that the net present value (NPV) of this FSM project is $25.3
million, and the corresponding the benefit-cost ratio is 1.22. This can be interpreted as, for every
dollar spent on the project, the state receives approximately $1.22 in economic and social benefit.
However, the CBA also revealed important heterogeneity masked by the positive NPV. First,
the Monte Carlo simulation (a type of sensitivity analysis) showed wide variability as 30% of the
trials returned negative NPVs. This means that the FSM project has a non-trivial chance of being
economically unviable. Second, using tornado analysis, I find that “baseline diarrhea incidences”
and “change to diarrhea incidence rates” are the two largest determinants of the variability in NPV
calculations. These are also statistics where there is a lack of credible information, thus highlighting
an urgent need for governments to implement better data monitoring processes. Lastly, the FSM
project consists of two independent phases where the one serving populous cities has a larger NPV
than the phase focusing on peri-urban areas with a lower population. Hence, this case study highlights
an inherent tension between efficiency and equity in reaching out to underserved populations.

The rest of the case study is organized as follows: The first section contains the background
to the project and site. This is followed by a brief description of the CBA methodology and the
associated costs and benefits of the FSM project. Next, the study presents the results of the CBA
and sensitivity analyses are contained. Lastly, we conclude and discuss the findings.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Background
State of Sanitation in Tamil Nadu
Tamil Nadu is one of the largest states in India, with a land size of over 130,000 km2 (ranked
10th in the country) and population of over 72 million (ranked 6th in the country).2 Tamil Nadu is
also economically important to India as it has the second-highest gross domestic product in the
country at around $260 billion and is home to major cities that specialize in technological services.
Despite Tamil Nadu’s economic prowess, the state has been playing catch-up in terms of access
to improved sanitation services (see Table 1 for a summary of Tamil Nadu’s sanitation records
across three data waves). According to the 2001 census, Tamil Nadu ranked 24th in the country,
with only around 35% of its population having access to toilet facilities, 1 percentage point below
the country’s average. When viewed separately by the rural and urban populations, Tamil Nadu
was ranked 30th and 32nd, respectively. The 2011 census showed a significant increase in the
sanitation access rate to 48.3% of the state’s population. However, this rate of increase is slower
than other states, as Tamil Nadu’s ranking among all Indian states dropped three places to 27th.
The latest statistics from the 2014–2015 edition of the National Family Health Survey showed that
Tamil Nadu has made steady improvements since then, as around 52.2% of the the population now
have access to sanitation, with the rate of increase now outpacing other states as they are now 23rd
in the country. However, this improvement is mostly driven by Tamil Nadu’s urban population,
which now ranks 19th; on the other hand, Tamil Nadu’s rural population dropped to 29th place.

2
For reference, Tamil Nadu has a similar land area to Greece.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Table 1: State of Sanitation in Tamil Nadu Relative

to Other States Across Survey Waves
Total (% of population Rural (% of population Urban (% of population
Survey Wave with access to toilets) with access to toilets) with access to toilets)
2000 Indian census 35.15 14.36 64.33
2010 Indian census 48.3 23.2 75.1
National Family 52.2 34 69.7
and Health Surveys
2014–2015
Source: Data obtained from https://visualize.data.gov.in

To improve sanitation services, the Tamil Nadu government set out a comprehensive, two-part
strategy in their 12th 5-year plan for the years 2012 to 2017.3 First, to continue increasing access
to toilets in rural areas, the Tamil Nadu government will embark on the Nirmal Bharat Abhiyan
program (continuation of the Total Sanitation Campaign) to construct toilets across rural villages
using a demand-driven approach (Hueso and Bell 2013). As infrastructure construction needs
to be paired with behavioral changes, the Tamil Nadu government will also construct gender-
specific sanitary complexes in villages to ensure safety for users and higher adoption levels.
Similarly, sanitation and hygiene lessons will be incorporated into school curricula to inculcate the
importance of improved sanitation from an early age.

Second, for the urban regions, Tamil Nadu is prioritizing improving sewage waste management.
One reason is that as more urban households in Tamil Nadu have access to toilets, there is an
urgent need to ramp up cities’ capacity to safely manage and dispose of sewage. To do so, the
government will build sewage treatment plants to serve 31 urban local bodies (ULBs) and upgrade
existing infrastructure to increase sewage treatment capacity.

State Investment Plan

The Government of Tamil Nadu commissioned the Tamil Nadu Urban Sanitation Support
Programme (TNUSSP) to implement urban sanitation as laid out in the 12th 5-year plan. To this
end, the TNUSSP drafted a State Investment Plan (SIP) where it is envisaged that the urban
sanitation objectives will be achieved in five phases (Table 2). The SIP detailed the infrastructure
and capital outlay needed, as well as the number of beneficiary households for each phase. The
purpose of this case study is to conduct a CBA of the first three phases of the SIP. There are two
main reasons why the study focuses on these three phases. First, of the total number of beneficiary
households in the SIP, around 75% will be reached in these three phases. Hence, it is obvious
that they form the most significant part of the program. Second, the current SIP document only
contains detailed information on Phases 1 to 3. As such, from a policy and practical viewpoint, it is
more feasible to conduct a CBA of these three phases.

3
http://www.spc.tn.gov.in/12plan_english/6.%20WATER_SUPPLY.pdf (accessed 1 July 2021).
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Table 2: Description of the Five Phases

of the Tamil Nadu Sanitation State Investment Plan
Details Phase 1 Phase 2 Phase 3 Phase 4 Phase 5 Total
No. of Sewage Treatment 41 34 49 78 157 359
Plant Locations
City Coverage
Corporations (excluding 8 3 11
Chennai)
Municipalities 26 30 51 17 124
Town Panchayats 35 53 59 111 270 528
Total Cities Covered 69 86 110 111 287 663

Household (HH) 2,408,835 1,339,048 948,335 454,145 1,225,200 6,375,563

Coverage
HH Covered by 741,487 195,131 936,618
Underground Sewage
Scheme
HH Covered by Fecal 1,667,348 1,143,917 948,335 454,145 1,225,200 5,438,945
Sludge Treatment Plant
No. of Persons/HH 3.86 3.82 3.89 3.87 3.87 19.31

Population Coverage 9,291,118 5,111,371 3,690,113 1,755,275 4,743,313 24,591,190

Cumulative Population 40% 60% 75% 81% 100%
Coverage
Source: The figures in the table were obtained from the TNUSSP report on State Investment Plan (https
://muzhusugadharam.co.in/wp-content/uploads/2020/10/TN-State-Investment-Plan_website.pdf). This case
study will only evaluate Phases 1, 2, and 3.

Broadly speaking, the plan for Phases 1 and 2 is to utilize current sewage treatment plants (STPs)
to treat sewage. At present, many of the existing STPs are ill-equipped to handle sludge from
toilets and septic tanks, with the sludge either dumped untreated into waterways or transported to
further locations. As such, the objective in Phases 1 and 2 is to upgrade existing pumping stations
and decanting facilities so that they can handle fecal sludge material from nearby sources. In turn,
around 86 cities or ULBs and over 14 million residents will benefit in this phase.

On the other hand, the purpose of Phase 3 is to construct new fecal sludge treatment plants (FSTPs)
in cities or ULBs without existing plants. It is anticipated that 49 new FSTPs will be constructed,
and they will collectively benefit 110 cities or 3.7 million residents.

4
It is obvious from the population ditterence that the cities reached in Phase 3 are smaller.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Method
Cost-Benefit Analysis
CBAs are one of the most widely used tools to assess the economic viability of individual projects
(Boardman, et al. 2017). A positive NPV means that the project will yield positive economic benefits,
and vice versa. Typically, when using a CBA to choose between different projects, a decisionmaker
will choose the project with the largest NPV, thus ensuring that the most economically efficient
choice is made (Boardman, et al. 2017).

The steps taken to conduct the CBA in this case study are as follows:

1. Determine standing
2. List out project’s costs and benefits
3. Detail the steps in calculating the costs and benefits
4. Parameterize the costs and benefits
5. Select appropriate discount rate and project timeframe
6. Compute a net present value
7. Conduct a Monte Carlo simulation and tornado analysis

Costs and Benefits of the Project

This subsection will describe in detail the first four steps of a CBA, i.e., the issue of standing, the
project’s costs and benefits, their accompanying formulas, and the parameters for computing
their value.

Standing
The issue of standing is the CBA, which refers to “whose costs and benefits are in consideration?”
In this case study, it is intuitive that standing is only given to Tamil Nadu, i.e., we will only consider
costs and benefits incurred by Tamil Nadu’s government, residents, and businesses.

Costs components
In the second step of the CBA, we list the project’s costs and benefits. We first focus on the costs.

Capital costs
The first cost component is the construction or capital costs of new sanitation infrastructure, i.e.,
sewage treatment plants, and pumping and decanting stations. According to the SIP, 91 pumping
stations and 55–85 decanting stations will be upgraded in Phases 1 and 2, costing around
Rp35 million. Construction costs are markedly higher for Phase 3 as it introduces 49 new STPs,
which will cost around Rp2 billion.5

5
There is no need to introduce formulas to compute capital costs as these estimates were provided in the SIP.

5
 A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Soft component costs

Other than capital costs, the successful introduction of sanitation infrastructure requires support
systems, such as technical assistance, and information and education campaigns. According to the
SIP, it is estimated that these soft components account for 8.5% of total capital costs and consist
of technical assistance (2%), project management expenses (4%), and information and education
campaign expenditures (2.5%).

Sewage trucking costs

As the main objective of the SIP is to increase fecal sludge treatment capacity, it is implicit that
more waste will be sent for treatment and disposal. As such, we need to include the costs for
handling additional sewage disposal (Table 3 row 3). To do so, we first determine the per-trip
sewage trucking cost by relying on a survey conducted by the TNUSSP on trucking operators
(TNUSSP 2018). This consists of variable costs (e.g., fuel, truck maintenance) and fixed costs (e.g.,
cost of truck and pumping equipment). The per-trip variable cost (𝑇𝑟𝑖𝑝𝑇𝑆𝐶𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒) is computed by
dividing
𝐴𝑛𝑛𝑢𝑎𝑙𝑇𝑆𝐶the annual variable cost by number of trips taken by an average sewage truck in a year
𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒
( ). On the other hand, the per-trip fixed cost (𝑇𝑟𝑖𝑝𝑇𝑆𝐶 ) is computed by dividing
𝑇𝑟𝑖𝑝𝑇𝑦𝑒𝑎𝑟 𝑓𝑖𝑥𝑒𝑑

the lifetime fixed costs of a sewage truck by the product of its expected lifespan and average trips
𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒𝑆𝐶𝑓𝑖𝑥𝑒𝑑
per year ( ). In all, the sewage trucking cost is estimated be around $2.55/trip.
𝑇𝑟𝑖𝑝𝑦𝑒𝑎𝑟 × 𝐿𝑖𝑓𝑒𝑡𝑖𝑚𝑒

Table 3: List of Costs and Benefit Items and Accompanying Formulas

1 Capital/construction costs Estimates provided in the State Investment Plan

document by TNUSSP (CapCost)
2 Soft component costs SoftCom% × (CapCost)
3 ∆𝐷𝑖𝑠𝑝𝑜𝑠𝑎𝑙 𝐷 𝐻𝐻_𝑏𝑒𝑛𝑒 𝐷 𝑏𝑇𝑟𝑖𝑝_𝑆𝐶𝑓𝑖𝑥𝑒𝑑 + 𝑇𝑟𝑖𝑝_𝑆𝐶𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒)
trucking cost
where 𝑇𝑟𝑖𝑝_𝑆𝐶 = and

𝑇𝑟𝑖𝑝_𝑦𝑒𝑎𝑟

4 Increased sewage ∆Disposal × HH_bene × Tank × TreatCost

management cost
Benefit Components Formula
5 Reduced mortality ∆𝐷𝐼 𝐷 𝐴𝑡𝑡_𝑚𝑜𝑟𝑡 𝐷 𝑉𝑆𝐿𝐼𝑛𝑑𝑖𝑎
from diarrheal disease 𝑉𝑆𝐿 𝐺𝐷𝑃𝐼𝑛𝑑𝑖𝑎 𝗌
where 𝐼𝑛𝑑𝑖𝑎 =( ) ∙ 𝑉𝑆𝐿𝑈𝑆 and
𝐺𝐷𝑃𝑈𝑆

𝐼𝑛𝑑_𝑏𝑒𝑛𝑒
𝐴𝑡𝑡𝑚𝑜𝑟𝑡 =
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛

6 Reduced morbidity ∆𝐷𝐼 𝐷 𝐴𝑡𝑡_𝑚𝑜𝑟𝑏 𝐷 𝐶𝑂𝐼

from diarrheal disease
𝐼𝑛𝑑_𝑏𝑒𝑛𝑒
where 𝐴𝑡𝑡 = 𝐷 𝐵𝐿_𝑚𝑜𝑟𝑏
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛

6 Source: Author.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Second, we multiply the per-trip cost by the total number of additional disposal trips
(∆𝐷𝑖𝑠𝑝𝑜𝑠𝑎𝑙 × 𝐻𝐻𝑇𝑏𝑒𝑛𝑒). The former is determined by (i) the difference between the ideal
disposal rate per household-year (0.4, or once every 2.5 years) and the current disposal
rate (0.25, or once every 4 years) (TNUSSP 2016); and (ii) the total number of household
beneficiaries in each phase of the SIP.

Operation and maintenance costs

Other than sewage trucking costs, the treatment plants also need to incur additional operational
and maintenance costs in treating the increased fecal sludge material. The formula for the sewage
treatment cost consists of the amount of additional fecal sludge produced per year and the per-
unit cost of treatment. The former is obtained by multiplying the change in disposal rate by the
number of household beneficiaries and average tank size (∆𝐷𝑖𝑠𝑝𝑜𝑠𝑎𝑙 × 𝐻𝐻𝑇𝑏𝑒𝑛𝑒 × 𝑇𝑎𝑛𝑘). The latter
(𝑇𝑟𝑒𝑎𝑡𝐶𝑜𝑠𝑡) was obtained directly through correspondence with the TNUSPP (author of the SIP
document). There is a big distinction in fecal sludge treatment cost for STPs and FSTPs. According
to the TNUSPP, STPs can process and treat fecal sludge at a cheaper rate ($0.04/m3) compared
to FSTPs ($2.14/m3).

Benefit components
We now focus on the benefits of the SIP. To this end, we are only considering the health benefits of
improved sanitation disposal in this case study, as these are the most recognized and well-studied
benefits of sanitation improvements (Haller, Hutton, and Bartram 2007; Hutton and Haller
2004).6

Mortality benefits
The mortality benefits (Table 3, row 5) are determined by three components: (i) the change in
the disease incidence rate (∆𝐷𝐼); (ii) the baseline mortality rate (𝐵𝐿𝑇𝑚𝑜𝑟𝑡); and (iii) the value of
a statistical life (𝑉𝑆𝐿). To estimate the decrease in disease incidence rate, we should ideally have
information on the extent to which water quality has improved as a result of the safer management
of fecal sludge. However, because baseline information on water quality in Tamil Nadu is unavailable
and there are no impact evaluation studies projecting potential improvements to water quality, we
are thus assuming that there is 25% reduction in mortality rates.7 Second, we attempt to retrieve
the baseline mortality rate of diarrhea diseases from official statistics collected by the Health and
Family Welfare Department of Tamil Nadu government.8 However, the statistics showed that
even though there were over 100,000 cases of diarrhea diseases, mortality remained at zero for
several years. As such, we conservatively estimate the mortality rate to be around 100 deaths per
year across the entire state.9 As the SIP focuses mainly on urban areas, we attribute the baseline
mortality rate to this specific segment of population by multiplying the proportion of resident

6
There are other possible benefits of proper sanitation disposal and management such as time-savings and economic
growth (e.g., see Van Minh and Hung [2011]). We are not considering time savings here as the assumption is that
households already have access to improved sanitation facilities. Estimation of economic activity benefits is highly
contextual and requires highly detailed data for computation.
7
If there are suitable data on water quality, we could use dose-response function to estimate changes to diarrhea
incidences.
8
https://www.tnhealth.org/dph/dphdb.php (accessed 10 December 2020).
9
This assumption will be relaxed as we conduct a Monte Carlo simulation to allow greater variance in the parameters.
7
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

𝐼𝑛𝑑𝑇𝑏𝑒𝑛𝑒
beneficiaries to the entire population in the state ( ). To this end, the baseline mortality
𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛
rate from diarrhea is estimated at around 36 deaths per year. Lastly, the value of a statistical life
(VSL) in India is calculated by a conventional benefits transfer formula where we multiply the VSL
from the US Environmental Protection Agency with the ratio of India’s per-capita gross domestic
product to that of the US, and finally raise this parameter to the income elasticity of VSL (Viscusi
and Aldy, 2003). In this regard, the VSL in India is estimated at around $1.47 million.

Morbidity benefits
The morbidity benefits are calculated in a similar manner as the mortality benefits (Table 3, row 6),
and are determined by three main parameters: (i) the change in the disease incidence rate (∆𝐷𝐼);
(ii) the baseline morbidity rate (𝐵𝐿𝑇𝑚𝑜𝑟𝑏); and (iii) the cost of illness (𝐶𝑂𝐼). The baseline morbidity
rate is estimated at around 165,000 cases for the entire state. This figure is obtained by averaging
the annual diarrhea cases computed by the Tamil Nadu government from 2010 to 2018. Similarly,
we apportion the baseline diarrhea cases to Phases 1 and 2, and Phase 3 separately, i.e., around
35,000 cases and 9,000 cases, respectively. Third, the COI is taken from a review study conducted
by Baral et al. (2020), who found that the average cost of treating outpatient diarrhea cases in
developing countries is around $37. The COI consists of medical costs, the opportunity cost of the
time of the patient, and the opportunity cost of the time of the caregiver.

Discount Rate and Project Timeframe

As projects typically last multiple years, it is necessary to discount the future streams of costs and
benefits to bring them to present terms. In this regard, the discount rate is set at 6% per year. This
is also the same rate used by the Asian Development Bank for evaluating sanitation projects. As
for the project timeframe, the SIP did not specify how long the infrastructure is expected to last. As
such, we assume the project will last for 40 years since other sources found that treatment plants
in India typically last around 40–50 years (Raghuvanshi, et al. 2017).

Monte Carlo Simulation and Tornado Analysis

One of the main difficulties of conducting CBAs is that many parameters are not deterministic,
i.e., they take a range of values as opposed to one single value. As such, we should not simply rely
on a single set of parameter values to form definitive conclusions about the project. A thorough
way of conducting sensitivity analyses is a Monte Carlo simulation where each parameter is given
a statistical distribution (e.g., uniform, normal) and the moments or range of the distribution
(e.g., we would specify the mean and standard deviation for a normal distribution). The Monte
Carlo simulation is implemented by randomly drawing a value for each parameter based on
their assumed statistical distribution and computing an NPV for this set of parameters. This
exercise will be conducted 10,000 times, which in turn will yield 10,000 NPVs. These NPVs can
then be summarized using a cumulative distribution function. The purpose of the Monte Carlo
simulation is to provide a realistic distribution of NPVs for the project by taking into account the
non-deterministic nature of most parameters.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

To provide more insights into the policy, I will also conduct a tornado analysis. The purpose of this
analysis is to examine which parameters have the greatest effect on the NPVs, and is implemented
in the following steps. First, we fix all parameters at their average value, except for one. Second, we
recover two NPVs using the average value for all parameters and the two extreme values (typically
at the 5th and 95th percentile levels) of the unfixed parameter. Third, the range of the NPVs
between these two extreme values is the impact of this parameter. Fourth, we repeat this exercise
for all other parameters, and then rank the parameters in terms of their range.

Results
We first present the main findings. Table 4 contains the present value of each major component
of the costs and benefits of Phases 1 and 2 and Phase 3. First, we see that the overall NPV for all
three phases is around $25.3 million. This means that we expect the SIP to yield a social benefit
of $25.3 million over the project’s lifetime of 40 years. Second, when examined by individual
components, the largest contributor to costs is sewage trucking, which accounts for around 43%
of the total NPV cost. On the other hand, despite the small number of avoided diarrhea deaths,
mortality benefits constitute the largest share of total NPV benefits (around 96%). This is due to
the relatively large value of a statistical life (VSL). Third, as we had computed costs and benefits
according to each phase of the SIP, we can examine their individual NPVs. Toward this end, we
see that the overall positive NPV was mainly driven by Phases 1 and 2 as this delivered an NPV
of $67 million. In contrast, Phase 3 yielded a NPV of –$42 million. The main reason for the stark
difference in NPV across the two main phases is that construction cost is low in Phases 1 and 2
(around $470,000) and much higher in Phase 3 (around $27 million). Similarly, sewage treatment
costs are much higher in Phase 3 despite serving fewer beneficiaries. This is because the sewage is

Table 4: List of Parameters and Their Statistical Range

Parameter Description Mean Range Source
SoftCom% Soft component cost as 8.5% 6.5–10.5 TNSP SIP
a proportion of capital
costs
COI Outpatient cost of $36.56
illness for a single bout
of diarrhea
VSLUS Value of statistical life $9.7 million 6.8 million Environmental
for United States to 12.6 million Protection Agency
GDPIndia GDP per capita for $2,104 1,894 to 2,314 World Bank statistics
India (2019) database
GDPUS GDP per capita for $65,118 58,606 to 71,630
United States (2019)

continued on next page

9
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Table 4 continued
Parameter Description Mean Range Source
BL_mort Baseline diarrhea 100 deaths/year 0 to 200 Health and Family
mortality rate for Welfare Department,
Tamil Nadu Government of Tamil
BL_morb Baseline diarrhea 165,000 cases/ 135,000 Nadu
morbidity rate for year to 400,000
Tamil Nadu Author’s estimate

∆Disposal Change to household 0.15 times/year 0.1 to 0.3 TNSP Baseline studies
sewage disposal rate (2016)
Tank Average size of 7.6 m3 3 to 15
septic tank
Lifetime_SCfixed Lifetime sewage $30,405 24,324 TNPP Desludging
trucking fixed costs to 36,486 operators in
Lifetime Lifetime of sewage 15 years 10 to 20 Periyanicken-
truck Palayam and
Narasimhanaicken-
Annual_SCvariable Annual sewage trucking $29,190 26,270 Palayam (2018)
costs to 32,108
HH_bene Number of household 3,747,883 NA State Investment Plan
beneficiaries in document prepared by
Phases 1 & 2 TNPP
Number of household 948,335
beneficiaries in Phase 3
Ind_bene Number of individual 14,402,489
beneficiaries in
Phases 1 & 2
Number of individual 3,690,113
beneficiaries in Phase 3
TreatCost Sewage treatment costs $0.04/m3 0.035 to 0.05 Direct correspondence
(for STP) with TNPP
$2.14/m3 1.7 to 2.57
(for FSTP)
Discount rate 7% 5 to 9
Source: Author.

treated in FSTP rather than STP in Phase 3. Fourth, due to the difference in scale between Phases
1 and 2 and Phase 3, we can also restate the earlier results in terms of benefit-cost ratio for easier
comparison.10 The benefit-cost ratio for Phases 1 and 2 is 2.62. This means that the state of Tamil
Nadu can expect a return of $2.62 for every $1 invested. On the other hand, Phase 3 has a benefit-
cost ratio of 0.4, indicating a loss of $0.6 for every $1 invested.

10
The benefit-cost ratio is computed by dividing total benefits by total costs.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Monte Carlo Simulation and Tornado Analysis

As mentioned earlier, the parameters used in CBAs are mostly non-deterministic. In this regard,
even though we had obtained an overall positive NPV using the mean values of each parameter, it
does not necessarily imply that the project would invariably be positive regardless of what values
the parameters take. One way to incorporate the many different values that each parameter can
take is via a Monte Carlo simulation.

We ran 10,000 trials where a value for each parameter was picked randomly for each trial. In turn,
the NPVs from these 10,000 trials are summarized using a cumulative distribution function in
Figure 1. Around 46% of the trials resulted in negative NPV. This is in clear contrast to the highly
positive result we obtained if simply relying on the mean value.

Figure 1: Cumulative Distribution Functions

of Net Present Values from Monte Carlo Simulations

1.0
0.9

0.8
0.7
0.6

0.5
0.4

0.3
0.2

0.1

1,000
($ million)
NPV (Overall) NPV (Phase 1 & 2) NPV (Phase 3)

Author.

We can also run separate Monte Carlo simulations for the individual phases to uncover any
heterogeneity between these two main phases. The NPVs for Phases 1 and 2 display a wide
variance as they range from –$50 million to over $300 million. More importantly, around 20% of
the trials returned a negative NPV, suggesting that even this part of the SIP is not entirely positive.
For Phase 3, around 90% of the trials are negative. Again, it should be highlighted that this implies
a non-trivial portion of the trials (10%) for Phase 3 returned positive NPV. In all, the Monte Carlo
simulations suggest that there is a wide range of values the NPVs could possibly take, and there
could be certain parameters highly influential in determining the outcomes.
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

As such, we further implement a tornado analysis to examine in greater detail which parameters
have outsized impacts on the NPVs. Figure 2 shows the outcomes of the tornado analysis where
the parameters with largest impacts are displayed in descending order. The top two parameters
with the largest explanatory power in the variability of the NPVs are the baseline death rate and
changes to diarrhea incidences. For instance, the NPV at the 5th percentile value of the baseline
death rate is around –$80 million and is around $180 million at its 95th percentile value. The range
in NPV caused by changes to diarrhea incidence is similarly large. In comparison, the next four
most influential parameters (VSL, amount of sewage disposal, income elasticity of VSL, and COI)
have a much smaller impact on the NPV’s range. The tornado analysis thus provides additional
insights on which parameters play the most important role in the CBA, and help us focus on better
data collection efforts for these parameters.

Figure 2: Tornado Analysis of Parameters Used in the Computation of Overall NPV

100 100
Baseline death rate of diarrhea 180.00

Changes to diarrhea incidences 9.50 45.50

VSL for United States ($)

Amount of sewage per disposal (m3) 13.80 4.20

lncome elasticity of VSL 0.59 0.51

Outpatient COl of a single case of diarrhea ($) 18.42 131.35

Discount rate (%)

Change to sewage disposal per year per household 0.19 0.11

(for Phase 2 & 3 residents)

Source: Author.

Discussions
This case study examined the economic viability of a fecal sludge management project using CBA.
Specifically, we examined an urban sanitation improvement plan currently implemented by the
Tamil Nadu government. The main objective of this plan is to upgrade and construct fecal sludge
treatment plants to meet growing demand for sewage management across cities in the state. We
analyzed the first three stages of a 5-phase program, and found a positive NPV of $25.3 million.
However, this highly positive NPV is masked by heterogenous outcomes to the individual phases.
Specifically, Phases 1 and 2 has an NPV of $67.4 million and Phase 3 delivers a negative NPV of
–$42.1 million. Further sensitivity analyses using a Monte Carlo simulation confirm that Phases 1
and 2 outcomes tend toward being positive and Phase 3 toward negative. Lastly, we also implement
a tornado analysis to examine which parameters are most influential in determining the NPVs. We
A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

find that health measures such as the baseline diarrhea

mortality and improvement to diarrhea incidences play
outsized roles in NPVs calculations.

There are several policy implications arising from the

CBA results. First, it is not surprising that Phases 1 and
2 and Phase 3 have contrasting outlooks. The former
leverages on existing infrastructure to benefit more
households at much lower construction costs. On the
other hand, Phase 3 involves construction of costly
FSTPs that serve smaller population. In this regard,
this result highlights the tradeoff between efficiency
and equity. Even though Phase 3 has a negative NPV,
it does not necessarily imply that the project should
not be carried out. One way for policy makers to justify
projects that benefit smaller or more disadvantaged
populations is by using weights.

Second, the parameters highlighted as most influential

in the tornado analysis are statistics where it is difficult
to obtain credible information. For instance, the
mortality rate from diarrhea diseases, while reported
by the Tamil Nadu government, is suspect as there
are zero reported deaths for several years despite still
having over 100,000 cases of diarrhea annually. As
such, for the purpose of this case study, we needed to
assume a range of diarrhea-related mortality. Similarly,
because there are no reliable studies on projected
improvements to water quality following the SIP, we
need to assume a wide range of improvements from 5%
to 50%. In this regard, this case study echoes the 2019
JMP report’s call for better monitoring data so that
policymakers and analysts can better evaluate projects
performance and viability.

Third, it should also be highlighted that we have only

included health-related benefits in this study. While sanitation investments are pursued from the
motivation of reducing diarrhea and waterborne diseases, there is also growing interests in how
they may provide economic benefits. For instance, a recent sewage disposal project in Dumaguete,
Philippines was purported to enhance economic activities from waterways and tourism around
the town. While it is possible for CBA to incorporate changes to economic activities, the data
requirements and economic modeling are much more intensive, which are beyond the scope of
this case study.
 A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

References
Baral, R., J. Nonvignon, F. Debellut, S. A. Agyemang, A. Clark, and C. Pecenka. 2020. Cost of
Illness for Childhood Diarrhea in Low- and Middle-Income Countries: A Systematic Review
of Evidence and Modelled Estimates. BMC Public Health 20: 1–13.
Boardman, A. E., D. H. Greenberg, A. R. Vining, D. L. Weimer. 2017. Cost-benefit Analysis: Concepts
and Practice. Cambridge: Cambridge University Press.

Haller, L., G. Hutton, and J. Bartram. 2007. Estimating the Costs and Health Benefits of Water and
Sanitation Improvements at Global Level. Journal of Water and Health 5(4): 467–80.

Hueso, A., and B. Bell. 2013. An Untold Story of Policy Failure: The Total Sanitation Campaign in
India. Water Policy 15(6): 1001–1017.
Hutton, G., and L. Haller. 2004. Evaluation of the Costs and Benefits of Water and Sanitation
Improvements at the Global Level. https://apps.who.int/iris/bitstream/handle/10665/68568
/WHO_SDE_WSH_04.04.pdf (accessed 23 June 2021).

Raghuvanshi, S., V. Bhakar, C. Sowmya, and K. Sangwan. 2017. Waste Water Treatment Plant Life
Cycle Assessment: Treatment Process to Reuse Water. Procedia CIRP, 61: 761–66.

TNUSSP. 2016. Baseline Studies: Tiruchirappalli, Periyanaicken-Palayam, Narasimhanaicken-

Palayam. https://muzhusugadharam.co.in/baseline-studies-tiruchirappalli/ (accessed
23 June 2021).
TNUSSP. 2018. Desludging Operators in Periyanaicken-palayam and Narasimhanaicken-palayam: An
Overview. https://muzhusugadharam.co.in/desludging-operators-in-periyanaicken-palayam
-and-narasimhanaicken-palayam-an-overview/ (accessed 23 June 2021).

Van Minh, H., and N. V. Hung. 2011. Economic Aspects of Sanitation in Developing Countries.
Environmental Health Insights: 5(EHI): S8199.

Viscusi, W. K., and J. E. Aldy. 2003. The Value of a Statistical Life: A Critical Review of Market
Estimates throughout the World. Journal of Risk and Uncertainty 27(1): 5–76.

World Health Organization (WHO). 2019. Progress on Household Drinking Water, Sanitation and
Hygiene 2000-2017: Special Focus on Inequalities. Geneva: World Health Organization.

14
 A Cost-Benefit Analysis of Tamil Nadu Urban Sanitation Improvement Plans

Study Questions
1. The two most influential parameters on the net present value calculations are baseline diarrhea
mortality and improvement to diarrhea incidence rates. In this case study, we had to assume
the values of these parameters as there is a lack of information. Suppose you are provided with
some resources, think of a monitoring and evaluation strategy that would collect these data/
information in the most efficient and accurate manner.
2. Phases 1 and 2 reach out to larger populations in big cities using existing infrastructure. On
the other hand, the beneficiaries in Phase 3 are in smaller cities, and the construction of new
infrastructure is needed. In this regard, Phase 3 will always fare relatively poorer in a CBA.
One way to justify investments in projects such as Phase 3 is to assign larger weights to the
beneficiaries’ benefits. Think about how we can compute weights objectively.
3. We did not include the economic benefits of improved sanitation treatment and disposal in this
case study. On the other hand, there are several possible benefits, such as increased tourism,
increase in housing value, and more government taxes being collected. Think about how we can
calculate these benefits and include them in the CBA. Specifically, work out the formulas and
parameters as shown in Tables 3 and 4.

Asian Development Bank Institute

ADBI, located in Tokyo, is the think tank of the Asian Development ADBI encourages printing or copying information exclusively for personal
Bank (ADB). Its mission is to identify effective development strategies and noncommercial use with proper acknowledgment of ADBI. Users
and improve development management in ADB’s developing member are restricted from reselling, redistributing, or creating derivative works
countries. for commercial purposes without the express, written consent of ADBI.

The ADBI Development Case Study Series presents public policy Asian Development Bank Institute
initiatives for achieving the Sustainable Development Goals in Asia Kasumigaseki Building 8F
and the Pacific. It provides a timely resource for discussion-based 3-2-5 Kasumigaseki, Chiyoda-ku
learning among policy makers in ADBI’s capacity-building and training Tokyo 100-6008
programs. Japan
Tel: +813 3593 5500
www.adbi.org