1

THE IMPACT OF REGULATION ON ECONOMIC GROWTH IN

THE CARIBBEAN: A PANEL DATA INVESTIGATION

Nlandu Mamingi*, Nichelle Yearwood** and Tracy Maynard**

*The University of the West Indies, Cave Hill Campus, P. O. Box 64,Bridgetown,

BB11000, Barbados. Email: [email protected]. Tel: (246) 417 4278.

**Central Bank of Barbados, Barbados.

Abstract
The paper uses a panel data fixed effects model based on a sample of 14 Caribbean
countries over the period 2004-2012 in order to investigate the impact of the
aggregated as well as disaggregated levels of business regulation on the economic
growth performance in the region. The panel data estimation results indicate that in
most cases a heavy regulatory burden is a drag on economies. This also holds for
the most part when the disaggregated measures of regulation are considered.
Indeed, there seems to be an inverse relationship between time taken to start a
business and economic growth. Similarly, a large tax burden negatively affects
output, and more regulations as it relates to trading across borders depress economic
growth. The study also reasserts that a positive relationship exists between good
governance and economic performance. A useful byproduct of the empirical
investigation is the index formulated from the World Bank Ease of Doing Business
index which allows for the disaggregated analysis of regulatory reform in the
region.

Keywords: Regulation, Economic Growth, Caribbean, Fixed effects model

JEL classification: O54, C43, L51.

 2

I INTRODUCTION1

Unquestionably, the regulatory environment of an economy is an important

determinant of economic growth. While in some situations it can be a key element

of growth promotion, in other circumstances it can become a binding constraint. As

such, it is important to identify and analyse those regulatory elements that are

essential in influencing the growth dynamic of an economy. This process must

ensure that regulatory objectives are reviewed in order to determine whether they

have been effectively met and have resulted in positive spillovers on growth.

Similarly, it is important to investigate whether regulation has instead adversely

affected growth due to inefficiencies in its implementation or due to uneven welfare

effects.

An optimistic outlook subscribes to the view that regulation is put in place

by local authorities to address market imperfections that may negatively impact

citizens (Guasch and Hahn 1999). For example, negative externalities produced

during the course of market activity (e.g., pollution) can be abated by regulatory

stringency. Additionally, a regulatory environment that protects workers and

investors via labour and property laws can be a major consideration for firms

(Tannenwald 1997).

1
This paper heavily borrows from Yearwood (2014). We acknowledge Simon
Naitram’s contribution and the comments made by the participants at the 34th
Annual Review Seminar of the Central Bank of Barbados on an earlier version of
the paper. We also thank the referees for their useful comments. All remaining
errors are our own.
 3

Few scholars, however, believe in the benevolent government argument.

Instead, regulation is seen as a tool used by regulatory bodies and incumbent firms

to capture economic rents by limiting competition (Stigler 1971). By imposing

significant barriers to entry, those firms that can potentially add to the productive

capacity of the economy are restricted, and those that are able to enter the market

may not perform optimally due to resources being deviated to address regulatory

restrictions (Busse and Groizard 2008). Additionally, a regulatory environment rife

with bribery and corruption may discourage potential investors from considering

such countries for investment.

The Caribbean has been subjected to poor regulatory performance rankings

by various organizations. While there has been some improvement in regulatory

reform over the recent past, it does not appear to have had a substantial impact on

economic growth. This paper attempts to quantify and analyse the impact of

business regulation on Caribbean economies. Using a sample of 14 countries over

the period of 2004-2012, an unbalanced fixed effects model is implemented in order

to gauge the impact of regulation on the macroeconomic performance of the region

at aggregated as well as disaggregated levels. Because of the short length of the

time series and challenges related to the differing levels of data availability for the

14 territories, it is beyond the scope of this paper to incorporate all measures of

regulation2.

2
Contrary to the common belief, enlarging the number of countries is not necessary
a panacea.
 4

In order to provide a holistic picture of the regulatory framework,

specifically as it relates to the business environment, this paper uses data from the

World Bank Ease of Doing Business database. The subcomponents that we find

most pertinent to the policy environment include regulations that deal with starting

a business, registering property, trading across borders, protecting investors,

enforcing contracts and paying taxes.

This paper makes two empirical contributions to the literature. First, this

study constructs a regulatory index which is specific and the first of this kind for

the Caribbean. Second, to the best of our knowledge, this is the first study that

quantitatively investigates the relationship between economic growth (GDP per

capita) and business regulation in aggregated as well as disaggregated ways within

the Caribbean.

Estimation results indicate that regulations that affect starting a business,

paying taxes, and trading across borders significantly and negatively impact

economic growth (real GDP per capita). However, the analysis also highlights that

the regulatory burden as it relates to enforcing contracts significantly and positively

affects real GDP per capita. We also investigate whether government effectiveness

is a contributing factor to growth, and find that it is significantly and positively

related to real GDP per capita growth.

The paper proceeds as follows. Section II succinctly reviews the literature

on regulation. Section III outlines the methodology. Section IV presents the

 5

estimation results and their interpretations. Section V contains concluding remarks

and policy recommendations.

II LITERATURE REVIEW

Countries adopt regulation to ensure equitable distribution of economic rents,

efficient productivity in the market and to temper any negative externalities that

may occur due to economic activities (Guasch and Hahn 1999). By removing

certain market failures and improving economic efficiency, regulation can

positively impact growth. On the other hand, regulation creates substantial

compliance costs, and undesirable market distortions which negatively impact

economic growth. Therefore, the impact of regulation in the literature is

inconclusive. Investigations involving regulation are usually centered on either

aggregated regulation or disaggregated regulation. The aggregated- economy wide

form of - regulation focuses on the business environment and the cost of doing

business in an economy, while the disaggregated measures of regulation

concentrate on the drivers of economic growth which include labour productivity,

the product market and environmental regulations.

Gorgens et al. (2003), Swanson (2008), Dawson and Seater (2013), Frye

and Shleifer (1997), etc. found that regulation negatively impacts economic growth.

Gorgens et al. (2003), who investigated the functional form of the relationship

between regulation and growth using a semi-parametric regression, acknowledged

that countries with a higher level of regulation are more likely to have an adverse

impact on growth. Swanson (2008) assessed how environmental regulation

interacts with the development process. He noted that environmental regulation

 6

which entails restricting industrial access to environmental resources (air, water,

eco-systems) in order to provide some of these resources to other sectors of society

was found to have an unfavorable impact on economic growth.

In an attempt to model regulation, Dawson and Seater (2013) used a

macroeconomic construct of regulation that captured its level, growth rate and

transition dynamic effects and investigated how these impact macroeconomic

output in the United States. The authors concluded that regulation reduced the

aggregate growth rate by about two percentage points over the sample period, 1949-

2005. Frye and Shleifer (1997) determined that the small businesses in Moscow,

Russia, were less productive due to the high level of regulation and corruption that

existed, which in turn translated to lower economic growth. Regulation in the form

of labour policies relating to the hiring and firing of persons can restrict investment

in research and development due to high compliance costs and restrictions on

labour allocation that reduces productivity (Bassanini and Ernst 2002). Di Tella and

MacCulloch (2005) also noted that labour policies negatively impact employment

rates and hence growth. In contrast, in the Caribbean context, Downes et al. (2004)

found that labour market legislation was not a significant determinant of

employment.

Busse and Groizard (2008) observed that those countries that had high

amounts of regulation had less positive effects from foreign direct investment thus

leading to unfavourable economic growth. According to the study, regulatory

barriers prevented useful productive foreign technology being assimilated into the

local economy. Mamingi et al. (2008), Dasgupta et al. (2006), Dasgupta et al.
 7

(2001), Lanoie et al. (1998), Konar and Cohen (1997), Pargal and Wheeler (1996)

and Hamilton (1995) also investigated the relationship between informal

institutions and economies.

In contrast, Tannenwald (1997) emphasised that countries with sound

regulations — ensuring protection of property rights, safe working conditions and

efficient institutions — may attract workers and investment which contributes

positively to growth. Environmental regulation can promote productivity in firms

that were operating at subpar levels before policy implementation while labour

regulation that provides incentives to workers through autonomy in the work place

and tenure can increase labour productivity and hence economic activity (Storm

and Naastepad 2007).

Djankov et al. (2006), using a sample of 135 countries of differing levels of

development and sourcing from the World Bank Ease of Doing Business Index,

found that those countries with “less burdensome regulation” grew faster than those

with restrictive policies. In a related study, Messaoud and Teheni (2014) examined

the robustness of the relationship between business regulations and economic

growth using a sample of 162 countries over the period 2007-2011. The results

showed that most regulation indices are positively correlated to the average growth

rate. Haidar (2009) investigated the relationship between investor protection and

economic growth using a new measure of legal protection of minority shareholders

against expropriation by corporate insiders—termed Investor Protection Index—

for more than 170 countries around the world. The author found that the level of

investor protection matters for cross country differences in economic growth:

 8

countries with strong protection tend to grow faster than those with poor investor

protection.

A major shortcoming of some of the studies on regulation, for example,

Estache and Wren-Lewis (2009), Gorgens et al. (2005) and Petreski (2014), is the

underlying assumption that the institutions in the developed world work the same

way as those in developing countries. In reality, regulatory reform may be less

effective in lesser developed economies due to major structural inefficiencies found

within the regulatory networks. Like many developing economies, Caribbean

countries are plagued by underdeveloped institutions. Indeed, the Caribbean

countries lack a sufficiently good environment for business as a result of their weak

institutions (IDB 2009). The latter IDB report indicates that in the Caribbean legal

systems are costly and outdated, regulation is burdensome and taxes are

discriminatory. Additionally, there is a shortage of persons with the requisite skill

to work within the regulatory sector (Downes and Husbands 2003). These

inefficiencies make it difficult for regulatory reform to have its intended impact.

Downes et al. (2004) underlined the importance of the labour market regulation for

economic growth and specially ascertained that it was imperative that the Caribbean

countries of Jamaica, Barbados and Trinidad and Tobago correct the inefficiencies

in the labour market system before they could benefit from the institutional reform

found in countries with higher levels of development.

Empirically, there is the issue of scarcity of regulatory data that needs to be

underlined as this limits the ability to capture static and dynamic effects of

regulation on economic growth (Busse and Groizard 2008). Additionally, most

 9

studies use qualitative data often amalgamated from business surveys and other

indicators and composited into indices to reflect regulatory quality and stringency

(Nicoletti et al. 2000, etc.). Most of those data do not pass the scientific rigor test.

Not surprisingly, many studies have adopted the use of the World Bank Doing

Business indicators which are more objective (SAGPA 2010; Djankov et al. 2000

for studies including developing nations). At the very least, two lessons emerge

from the literature review. First, the impact of regulation on output is really an

empirical matter. Second, there is no thorough analysis of the impact of regulation

in the large sense on output for the Caribbean.

III METHODOLOGY

This section aims at developing the methodology to quantify the impact of

regulation on the macroeconomic performance in the Caribbean as well as provide

a description of the data used for such an enterprise. A growth model is proposed

and comprises regulatory indicators as well as control variables for 14 Caribbean

countries over the time period 2004 – 2012. The countries of interest are: Antigua

and Barbuda, the Bahamas, Belize, Barbados, Dominica, Dominican Republic,

Grenada, Guyana, Jamaica, St. Kitts and Nevis, St. Lucia, St. Vincent and the

Grenadines, Suriname and Trinidad and Tobago. Panel data is useful as it boosts

the sample size and helps to offset the constraints of limited data. Separate growth

regressions were run in order to gauge not only the relationship between

disaggregated measures of regulation and growth, but also the overall regulatory

burden and the region’s economic performance.

 10

3.1 Data

Data used to compile the key regulatory variables were sourced from the

World Bank Ease of Doing Business index. This World Bank index ranks countries

from 1 to 189 and is constructed by aggregating the percentile ranks of countries

based on 10 sub-indices. This databank was chosen in lieu of the many other

regulatory indicators available because of its objectivity as well as the availability

of a disaggregated component that could provide insight for policy purposes.

Additionally, finding a relevant regulatory proxy that provided sufficient data for

all 14 countries posed a challenge. Though the rankings provided by the index were

insufficient for panel data modelling (time span of 2 years for most Caribbean

territories), the underlying data used to construct the index seemed more promising

since it was available over a longer period. In light of this, an alternative index was

constructed, using information from the database for the relevant regulatory

subcomponents used in the analysis.

Of the 10 sub-indices that comprise the World Bank Ease of Doing Business

index, the 6 that seemed the most relevant and simultaneously contained the most

observations were utilized. These subcomponents included starting a business,

registering property, protecting investors, paying taxes, trading across borders and

enforcing contracts.

The Starting a Business measure essentially seeks to capture the investment

climate for new firms. This includes the number of procedures as well as the time

and cost required to meet government requirements to operate a business. Entry

 11

regulation can have either a positive or negative relationship with growth. If the

public interest theory holds, then a positive relationship is expected; on the

contrary, if the capture theory is applicable, then a negative relationship is likely.

Registering property encapsulates the security of the property laws of a

country. This measure includes regulations pertaining to procedures necessary for

an investor to buy and transfer the property title from the seller (World Bank

2014e). The transfer of property rights is important for investor incentive since it

ensures that investment returns are allocated correctly. Elias (2012) noted that the

Doing Business report ranked Trinidad and Tobago almost last in this sub-category

while Holden and Holden (2005) asserted that Jamaica’s Torrens system was

expensive and slow, making the registering process unattainable by the average

citizen. Burdensome regulation can deter firms from pertinent investment

opportunities.

The Protecting investors’ proxy aims to capture the strength of protection

provided to minority shareholders (World Bank 2014a). These regulatory

requirements (securities regulations, civil procedures, etc.) are important for stock

market development of an economy which provides efficient credit for firms to

invest (Djankov et al. 2008). A lack of investor protection is seen as negatively

related to economic growth.

The “Paying Taxes” proxy measures the tax contributions as well as the tax

burden that firms face. These include profit, property, labour taxes, etc. (World

Bank 2014b). Taxes can act as a disincentive for firms which minimize investment
 12

and income (Djankov et al. 2008). An initial investigation into the tax environment

suggests that though the governments of the respective countries place some

emphasis on taxing firms, the tax administration burden is not high relative to other

small open economies, for example, Mauritius. Figure 1 depicts the percentage of

firms in the respective countries that views the tax rates and tax administration as a

major obstacle to the operation of their firms. It is clear that a larger percent of firms

find the tax rates to be burdensome. This is not surprising given the high tax to

profit ratios found within the territories.

Trading across borders captures the procedures, cost and time needed to

export and import a cargo of goods via ship (World Bank 2014c). Small open

economies like those found in the Caribbean, are heavily reliant on foreign

investment and technology to bolster their economies, and so, the level of

bureaucracy and stringency of the regulatory process can hamper the business

environment for firms.

[ INSERT FIG. 1 ABOUT HERE]

According to World Bank survey, firms in Jamaica and Suriname take the

longest to clear customs (13 days). Firms in Guyana also have a lengthy process

where, on average, they spend 12 days to complete customs procedures. These

numbers compare unfavourably to other comparator countries. For example, in

Mauritius it takes 9 days for the average firm to clear customs.

Enforcing contracts is a measure of the effectiveness of the legal framework

of an economy (World Bank 2014d). Court systems generally aim at ensuring fair
 13

judgments by considering errors and possible corruption which can result in many

procedures. The World Bank survey indicates that less than 5% of firms in

Dominica describe the court system as presenting a major obstacle to the operation

of their business. An outlier to this result is Suriname where 44% of the firms

interviewed expressed that the court system was at least a major obstacle for

business. According to an IDB (2009) report, it takes the court on average three to

four years to deliver verdicts due largely to issues highlighted above as well as a

sense of ambiguity on the balance of power between the arms of the government.

Concerning judicial independence, the Caribbean has performed relatively

well. The survey results indicate that in spite of the slow judicial process present in

most Commonwealth countries, the majority of firms are confident in the efficiency

and reliability of the legal system to render fair judgments. However, too many

protocols can be cumbersome and costly and so deter potential investors, causing a

negative correlation with income (Djankov et al. 2003). In spite of cumbersome

and lengthy judicial procedures, it can be argued that the length of time can ensure

fair judgements and so can positively affect economic growth.

We follow the methodology devised by Loayza et al. (2004) to create an

index for the subcomponents of regulation (starting a business, protecting investors,

etc.) in an effort to glean the regulatory performance of the respective Caribbean

countries in relation to the world. In order to do this, the index is constructed using

the values for the subcomponent for each country for each year under study.

Additionally, the minimum and maximum values for each regulatory

subcomponent for each respective year for the world are also used. The formula for
 14

the index is presented below and ranges from 0 to 1 where values closer to one

indicate heavier regulatory burdens. The index is derived from the following

expressions

xij  min[ X w ]
(1)
max[ X w ]  min[ X w ]

and

max[ X w ]  xij
(2)
max[ X w ]  min[ X w ]

where higher values of X in formula (1) indicate larger regulatory burden; lower

values of X in formula (2) indicate larger regulatory burden; i stands for country

and j represents the year; x denotes the value of each subcomponent for each year
ij

for the respective country; min[Xw] and max[Xw] represent the min and max value

of each subcomponent for all countries in the world for every respective year.

A simple average was then taken to create the aggregate index value

attributed to each Caribbean country where higher values connote heavier

regulatory burden.

An initial look at the data reveals that the disaggregated indices discussed

above are closely comparable to the doing business rankings. Data was used for

2014 as it ranked all those countries chosen in our sample (see Table 1).
 15

[ INSERT TABLE 1 ABOUT HERE]

Government effectiveness was chosen from among the six governance

indicators provided by World Wide Governance indicators, as it most reflects the

institutional component important in developing countries. Government

effectiveness captures perceptions on the quality of public policy, government

processes as well as the independence of the civil servants from political pressure

(Jalilian et al. 2006; World Bank 2013). A positive sign is expected.

Real GDP per capita has been found suitable to capture economic growth

here rather than its growth counterpart singularly because regulation tends to

remain constant over long periods of time. In addition to the regulatory indicators,

GDP per capita is also dependent on various control variables included in the

regression analysis. These comprise foreign direct investment (FDI) measured as

net foreign direct investment inflows as a percentage of GDP, the investment ratio

captured by gross capital formation as well as population growth. These indicators

were all sourced from the World Bank development indicators database in an

attempt to reduce data inconsistencies. In addition, natural disasters variable

captured as a dummy variable with 1 if the event occurs in a particular year and 0

otherwise was also of interest. The variable was sourced from em.dat.

3.2 Methodology

The paper uses a panel data regression of the type

 16

yit    X it    i  eit (3)

where y is the dependent variable, i=1,2,3,…,N denotes the cross section index,

t=1,2,3,…,T represents the time index, α denotes the overall constant, X is the

matrix of explanatory variables, β is the vector of slope coefficients,  i represents

the unobservable individual-specific effect and eit stands for the usual stochastic

disturbance term.

At the very least, two static models can be derived from model (3): fixed

effects (FE) model and random effects (RE) model. Indeed, if  i is fixed or

correlated with all the variables in X, then model (3) qualifies as a fixed effects

model, which can be written as

yit   i  X it   eit (4)

where the overall constant has been eliminated. On the contrary if  i is random

like the usual error term, that is, uncorrelated with all the variables in X, then model

(3) becomes a random effects model and can be written as follows

y it    X it   vit (5)

where the new error term consists of two components:  i and eit , that is,

vit   i  eit (6)

Model (4) is estimated by pooled least squares methods. The fixed effects

model is called within model if it uses time-demeaned variables, in which case the
 17

individual effects are eliminated. Model (5) is the random effects model or

variance component model. It is estimated by generalized least squares (GLS)

method to deal with issues of autocorrelation and heteroscedasticity.

The choice between the two types of models can be done at several levels

although the Hausman specification test seems to have the edge over other means

of choice decision. In final analysis, because of the plausible correlation between

the unit specific effects and the explanatory variables, “fixed effects model is

almost always more convincing than random effects model for policy analysis using

aggregated data.” (Woolridge 2006, 498).

The Hausman specification test is used to test the null hypothesis of random

effects versus the alternative hypothesis of fixed effects. It is a chi-squared test

whose rejection of the null hypothesis means that FE is adequate.

The first set of models used in this paper deals with the aggregated measure

of regulation as explained above. Thus, we have the following

LGDPit   i   1 ARit   2 GE it   3 POPG it   4 LDI it   5 LFDI it   6 DU it  eit (7)

where LGDP is logged real GDP per capita,  i is fixed country specific effects,

AR denotes the aggregate regulatory measure, GE captures government

effectiveness, PopG is population growth, LDI is logged domestic private

investment as a percentage of GDP, LFDI is logged foreign direct investment as a

percentage of GDP, DU is dummy variable capturing the incidence of natural

 18

disasters (1 if natural disasters occur in a given year in a given country and 0

otherwise) and e is the regular error term assumed to be well-behaved.

Model (7) is the aggregated form of our fixed effects model. The

corresponding random effects model is given by

LGDPit     1 ARit   2 GE it   3 POPG it   4 LDI it   5 LFDI it   6 DU it  vit (8)

where vit   i  eit is the composite error term and other variables are defined as

above.

The second set of models is concerned with disaggregated measures of

regulation. The first model that we call disaggregated fixed effects model is as

follows

LGDPit   i   1 SBit   2 RPit   3 PI it   4 EC it   5TBit 

(9)
 7 GE it   8 PopG it   9 LDI it   10 LFDI it   11 DU it  eit

where SB connotes the regulation measure for starting a business measure, RP

represents the regulation measure for registering property, PI stands for the

regulation measure for protecting investors, PT is a measure for paying taxes, EC

captures a measure for enforcing contracts, TB is a measure for trading across

borders and GE captures government effectiveness. All other variables are defined

as in model (5), and e denotes the error term assumed to be well-behaved.

The second model is the analogue disaggregated random effects model

given by
 19

LGDPit    1 SBit   2 RPit   3 PI it   4 ECit   5TBit 

 7 GEit   8 PopGit   9 LDI it  10 LFDI it  11 DU it  vit (10)

where variables are defined as in model (9), and v denotes the composite error term

as above.

An issue that has not been addressed is that of endogeneity. It is possible

that wealthier countries have better institutions and more effective regulatory

systems. This gives rise to the issue of endogeneity. Endogeneity is understood as

significant correlation(s) between right hand side explanatory variable(s) and the

error term. As such the presence of endogeneity results in biasedness and

inconsistency of the ordinary least squares estimators. Most studies approach this

issue by applying two stage least squares (2SLS) method or IV Hausman-Taylor

estimation procedure. In this study, endogeneity is tested before contemplating

2SLS.

This is illustrated with model (7) for which the aggregate regulation

measure is suspected to be endogenous. In the first stage, the reduced form is

obtained by regressing each endogenous variable on the exogenous or

predetermined variables and retrieving the residuals from this auxiliary regression.

Residual1 here denotes the residuals from the auxiliary aggregate regulation

equation. In the second stage, model (7) augmented by residual1 is run. The t test

statistic related to the parameter associated to residual1 provides the test for

endogeneity (Woolridge 2006, 532-533). A rejection of the null hypothesis

(variable is exogenous) means there is endogeneity and appropriate method of

estimation such as 2SLS in fixed effects form should be used. Note that if there are
 20

more than one right-hand side endogenous variable then an F-test is of interest to

decide on endogeneity.

IV MODEL ESTIMATION, RESULTS AND INTERPRETATIONS

The aggregated model gives insight on the overall effect regulation has on growth.

Note that here an aggregate regulatory measure constructed from the World Bank

Ease of Doing Business database is of interest along with control variables.

Table 2 contains the results of the regression estimation of model (7) and

model (8). The Hausman test statistic reveals that the null hypothesis of random

effects is rejected in favor of fixed effects model at the 5% and 10% levels of

significance. Moreover, Table 3 results indicate through the behaviour of residual1,

obtained as explained above, that the aggregated measure of regulation is not an

endogenous variable. Indeed, the value of the associated t statistic leads to the non

rejection of the null hypothesis (the associated parameter value is null). Thus, there

is no need for formal 2SLS. We consequently interpret the fixed effects results of

Table 2. The fixed effects model results reveal a significant negative relationship

between the aggregate regulatory measure and logged real GDP per capita.

[INSERT TABLE 2 ABOUT HERE]

[INSERT TABLE 3 ABOUT HERE]

 21

A 1% increase in the regulatory burden leads real GDP per capita to

3
decrease by 0.726% . This confirms the conclusion reached by Djankov et al.

(2006) which asserts that countries with higher regulatory burden do not perform

economically as well as those that do not. The result also matches the conclusion

by IDB (2009) that indicates that the regulatory burden in the Caribbean can be a

major constraint on growth. Government effectiveness, private domestic

investment and foreign direct investment positively and significantly impact

economic growth. For example, a 1% improvement in government effectiveness

yields a 0.04% increase in real GDP per capita. The apparent counter-intuitive

result generated by the positive impact of natural disasters can be easily explained

in the context of short run.

At the outset we point out that since the disaggregated model has been

derived from the aggregated model, we do not redo all tests (e.g., endogeneity test).

That said, in lieu of aggregate regulatory index, Table 4 uses the following

regulatory indices: starting a business, registering property, protecting investors,

paying taxes, enforcing contracts and trading across borders. The Hausman test

statistic indicates that the fixed effects model is statistically the appropriate one.

3
Elasticity is computed as  X j where X j is the mean of the variable of interest.
For Log-log relationship, elasticity is rather straightforward. The marginal effect
is  Y j where Y j is the mean of the dependent variable.
 22

The model passes the overall chi-square test as the size of the p-value indicates.

Regulatory indices measuring starting a business, paying taxes, enforcing contracts

as well as trading across borders were found to all significantly impact real GDP

per capita. As per a priori expectations, all disaggregated measurements with the

exception of the registering property and enforcing contracts regulatory proxies

were negatively correlated with growth.

As mentioned before, cumbersome regulation negatively impacts the

economic performance of an economy 4 . Time and cost factors largely impact

investor decisions. The negative correlation between starting a business and income

levels within the Caribbean context emphasizes this point. For most Caribbean

territories the number of procedures to start a business is not as cumbersome and

as much of a constraint when compared to the time and cost to complete these

procedures. It took 694 days to complete all the necessary requirements to start a

business in Suriname for the year 2013. This represents the worst scenario case in

the region and is way above the world minimum of 5 days. Holden and Holden

(2005) noted that registering a business in Jamaica is unreasonable and expensive,

so much so that it has forced businessmen to operate within the informal economy,

emphasizing the similarly drawn conclusions by Djankov et al. (2002).

The paying taxes measure indicates that the tax burden is also a constraint

within the Caribbean. To corroborate, from Table 4, we conclude that a 1% increase

4
Because of a strong negative relationship between the two variables registered
under random effects models, the lack of significance found in fixed effects is
sidestepped.
 23

in the tax burden is related to a 0.10% decrease in real GDP per capita. Note that as

seen above the tax burden is associated more so with the high tax rates

characteristics of the region rather than administrative issues.

[INSERT TABLE 4 ABOUT HERE]

Regulation associated with trade can also prove to be detrimental to growth.

Excessive procedures and licenses can mean higher producer costs. This form of

regulatory burden seems to be particularly important in determining

macroeconomic performance in the Caribbean given the large coefficient and high

level of significance of the trading across borders proxy. In fact, when considering

the elasticity, a 1% increase in trading across borders regulation is associated with

a .21% decrease in income. This seems to be in keeping with Djankov et al. (2010)

who assert that delays in exports can deter trade and by extension income levels

across countries. The results match those from the enterprise survey. Indeed,

according to the latter, businessmen within the region identified customs and trade

regulation as major obstacles in the operation of their businesses. A reduction of

these regulations may therefore benefit growth.

The enforcing contracts regulatory measure is positively and significantly

correlated with growth. The result implicitly shows that a 1% increase in

regulations relating to contract enforcement can be linked to a 0.50% increase in

real GDP per capita. This seems to confirm the opinion of most Caribbean

businessmen in that the court systems are fair and uncorrupted and so create a good

investment climate.
 24

Though a lack of investor protection is seen as negatively correlated with

real GDP per capita, the results indicate that this regulatory measure is not

significant in the Caribbean context. Government effectiveness impact does not

show up with fixed effects model but is present with random effects model. Because

of its strong presence in the aggregated regulatory model we fully acknowledge its

positive presence. This corroborates the argument that a well-functioning

institution leads to increased economic welfare and thereby productivity growth.

V CONCLUSION AND POLICY RECOMMENDATIONS

This paper examines the relationship between regulation and economic growth in

the Caribbean over the period 2004-2012. The paper derives the appropriate

regulation indices to study the extent to which aggregated and disaggregated

regulations are binding constraints to economic growth in the Caribbean context.

The panel data estimation results indicate that in most cases a heavy

regulatory burden is a drag on the economies. This also holds for the most part

when the disaggregated measures of regulation are considered. Indeed, there seems

to be an inverse relationship between time taken to start a business and economic

growth. Similarly, a large tax burden negatively affects output, and also more

regulations on trading across borders depress economic growth. The study also

reasserts that a positive relationship exists between good governance and economic

performance.

Care must be taken to not only address regulatory inefficiencies but also

institutional weaknesses of these economies. For instance, while the regression

 25

estimations detail that an increase in regulation concerned with enforcing contracts

positively affects growth, red tape and inefficiencies that deter a speedy and fair

trial may have adverse effects on viable investments and firms. Providing sufficient

funding or increasing salaries in occupations within legal institutions may

encourage or can lead to quicker verdicts and perhaps a fairer distribution of access.

In reducing the regulatory burden associated with starting a business,

possible policy prescriptions include decentralizing regulatory bodies in an effort

to make them more accessible. Additionally, online applications and processing can

reduce time and cost constraints (Holden and Holden 2005). As such, documents

can be scanned or mailed, reducing travel costs.

As noted above, the tax burden is largely as a result of high tax rates and

not necessarily cumbersome administration. Reducing the tax burden entails

addressing inefficiencies in the tax system rather than a direct cut in taxes as

Caribbean countries have limited fiscal space. Reducing taxes as well as revisiting

the economic agents that are taxed may be a better alternative. For example, many

Caribbean economies offer exorbitant concessions and act as tax havens for

offshore firms. Reducing these concessions (including those that supply an

exemption from corporate tax), while reducing the high tax rates of domestic firms

may enhance investment activity.

References

Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2004. “Institutions

as the Fundamental Cause of Long-Run Growth.” NBER Working Paper

 26

Series 10481.

Bassanini, Andrea, and Ekkhehardt Ernst. 2002. “Labour Market Institutions,

Product and Market Regulation, and Innovation: Cross-Country Evidence.”

OECD Economics Department Working Papers, No 316. doi:

http://dx.doi.org/10.1787/002243151077.

Busse, Matthias, and José L. Groizard. 2008. “Foreign Direct Investment,

Regulations and Growth.” The World Economy doi: 10.1111/j.14679701.

2008.01106.x.

Dasgupta, Susmita, Jong Ho Hong, Benoit Laplante, and Nlandu Mamingi.

2006. “Disclosure of Environmental Violations and Stock Market in the

Republic of Korea.” Ecological Economics 58, no. 4: 759-777.

Dasgupta, Susmita Benoit Laplante, and Nlandu Mamingi, 2001. “Pollution and

Capital Market in Developing Countries.” Journal of Environmental

Economics and Management 42, no. 3: 310-335.

Dawson, W. John, and John J. Seater. 2013. “Federal Regulation and Aggregate

Economic Growth.’ Journal of Economic Growth 18, no. 2: 133-177. doi:

10.1007/s10887-013-9088-y.

De Soto, Hernando. 1990. The Other Path. New York: Harper and Row.

Di Tella, Rafael, and Robert MacCulloch 2005. “The Consequences of Labor

Market Flexibility: Panel Evidence based on Survey Data.” European

Economic Review 49: 1225-1259. doi: 10.1016/j.euroecorev.2003.11.002.

 27

Djankov, Simeon, Rafael La Porta, R., Florencio Lopez-de-Silanes, and Andrei

Shleifer. 2008. “The Law and Economics of Self-dealing.” Journal of

Financial Economics 88: 430-465. Doi:10.1016/j.jfineco.2007.02.007.

Djankov, Simeon, Caralee McLiesh, and Rita M. Ramalho. 2006.

“Regulation and Growth.” The World Bank, mimeo.

Djankov, Simeon, Rafael La Porta, R., Florencio Lopez-de-Silanes, and Andrei

Shleifer, A. 2003. “Courts.” Quarterly Journal of Economics 118, no. 2:

453-517.

Djankov, Simeon, Rafael La Porta, R., Florencio Lopez-de-Silanes, and Andrei

Shleifer, A. 2002. “The Regulation of Entry.” Quarterly Journal of

Economics CXVII, no. 1: 1-37.

Djankov, Simeon, Rafael La Porta, Florencio Lopez-de-Silanes, and Andrei

Shleifer. 2000. “The Regulation of Entry.” NBER Working Paper 7892.

Downes, Andrew, and Husbands, A. 2003. “Human Resource Systems for

Regulatory Institutions: An Imperative for the Caribbean.” In 1st Annual

Conference of Utility Regulators in the Caribbean. Trinidad & Tobago.

Downes, S. Andrew. 2009. “Flexible Labour Markets, Workers’ Protection and

Active Labour Market Policies in the Caribbean.” In Macroeconomía del

Desarrollo. Santiago, Chile: ECLAC.

Downes, Andrew, Nlandu Mamingi and Rose-Marie B. Antoine 2004. “Labor

Market Regulation and Employment in the Caribbean.” In Law and

 28

Employment: Lessons from Latin American and the Caribbean, edited by

James J. Heckman and Carmen Pagés. University of Chicago Press.

Estache, Antonio, and Liam Wren-Lewis. 2009. “Toward a Theory of Regulation

for Developing Countries: Following Jean-Jacques Laffont’s Lead.”

Journal Economic Literature 47, no. 3: 729-770.

http:www.aeaweb.org/articles.php?doi=10.1257/jel.47.3.729.

Féres, José, and Arnaud Reynaud. 2012. “Assessing the Impact of Formal and

Informal Regulations on Environmental and Economic Performance of

Brazilian Manufacturing Firms.” Environmental and Resource Economics

52, no. 1: 65-85.

Frye, Tymothy, and Andrei Shleifer. 1997. “The Invisible Hand and the

Grabbing Hand.” American Economic Review 87, no. 2: 354-358.

Gørgens, Tue, Martin Paldam, and Allan Würtz. 2005. “Growth, Income and

Regulation: a Non-Linear Approach Centre for Applied Microeconomics

(CAM).” Department of Economics, University of Copenhagen.

Gorgens,Tue,Martin Paldam, and Allan Wurtz. 2003. “How does Public Regulation

affect Growth? University of Aarhus Department of Economics Working

Paper, Working Paper no. 2003-14.

Guasch, J. Luis, and Robert W. Hahn. 1999. “The Costs and Benefits of Regulation:

Implications for Developing Countries.” The World Bank Research

Observer 14, no. 1: 137-158.

 29

Haidar, Jamal Ibrahim. 2009. “Investor Protection and Economic Growth.”

Economic Letters 103: 1-4.

Hamilton, T. James. 1995. “Pollution as News: Media and Stock Market Reaction

to the Toxics Release Inventory Data.” Journal of Environmental

Economics and Management 28: 98-113.

Harriss-White, Barbara. 1997. “Informal Economic Order: Shadow States, Private

Status States, States of Last Resort and Spinning States: A Speculative

Discussion Based on S. Asian Case Material.” QEH Working Paper

Series , Working Paper no. 6.

Holden, Paul, and Sarah Holden. 2005. “Jamaica: A Private Sector Assessment.”

Inter-American Development Bank. Mimeo.

IDB. 2009. “Institutions and the Legal Framework for Business Development in

the Caribbean.” Private Sector Development Discussion Papers, No. 3.

Jalilian, Hossein, Colin Kirkpatrick, and David Parker. 2006. “The Impact of

Regulation on Economic Growth in Developing Countries: A Cross-

Country Analysis.” Mimeo.

Laffont, Jean Jacques, and Jean Tirole. 1991. “The Politics of Government

Decision-Making: A Theory of Regulatory Capture.” The Quarterly

Journal of Economics 106, no. 4: 1089-1127.

Lanoie, Paul, Benoit Laplante, and Maurice Roy. 1998. “Can Capital Markets

Create Incentives for Pollution Control?” Ecological Economics 26, no. 31-
 30

41.

Loayza, Norman V., Anna Maria Oviedo, and Luis Servén, 2005. “The Impact of

Regulation Growth and Informality: Cross-Country Evidence.” World

Bank Policy Research Working Paper Series 3623.

Mamingi, Nlandu, Susmita Dasgupta, Jong Ho Hong, and Benoit Laplante. 2008.

“Understanding Firms’ Environmental Performance: Does News Matter?”

Environmental Economics and Policy Studies 9, no. 2: 67-79.

Messaoud, Boudhiaf, and Zribi El Ghak Teheni. 2014. “Business Regulations and

Economic Growth: What can be explained?’’ International Strategic

Management Review 2: 69-78.

Nicoletti, Giuseppe, and Stefano Scarpetta. 2003. “Regulation, Productivity and

Growth:OECD Evidence.”OECD Economics Department Working Papers

no. 347. doi: http://dx.doi.org/10.1787/078677503357.

Pargal, Sheoli, and David Wheeler. 1996. “Informal Regulation of Industrial

Pollution in Developing Countries: Evidence from Indonesia.” Journal of

Political Economy 104, no. 6: 1314-1327.

Petreski, Marjan. 2014. Regulatory Environment and Development Outcomes:

Empirical Evidence from Transition Economies.” Journal of Economics –

Obsahy Ročníkov 62, no. 3: 225-248.

Pigou, C. Arthur. 1932. “The Economics of Welfare (4th Edition).” London:

Macmillan and Co.

 31

Posner, A. Richard. 1974. “Theories of Economic Regulation.” The Bell Journal

of Economics and Management Science 5, no. 2: 335-358.

SAGPA (Swedish Agency for Growth Policy Analysis). 2010. “The Economic

Effects of the Regulatory Burden.” Report 2010/14.

Schneider, Friedrich, and Dominik H. Enste. 2000. “Shadow Economies: Size,

Causes, and Consequences.” Journal of Economic Literature XXXVIII,

no. 1:77-114. doi:10.1257/jel.38.1.77

Seibel, Dieter Hans. 2001. “Mainstreaming Informal Financial Institutions.”

Journal of Developmental Entrepreneurship 6, no. 1: 83-95.

Stigler, George. 1971. “The Theory of Economic Regulation.” Bell Journal of

Economics and Management Science 2 (spring): 3-21.

Storm, Servaas, and C.W.M. Naastepad. 2007. “Why Labour Market Regulation

May Payoff: Worker Motivation, Co-ordination and Productivity Growth.”

Economic and Labour Market Paper 2007/ 4: International Labour

Organization.

Swanson, Timothy. 2008. “Economic Growth and Environmental Regulation: A

Discussion of International Experiences.” In ADB Workshop on “A

Macro-Environmental Strategy for China.” Beijing.

Tannenwald, Robert. 1997. “State Regulatory Policy and Economic Development.”

New England Economic Review March/April issue: 83-108.

Woolridge, Marc Jeffrey. 2006. Introductory Econometrics: A Modern Approach.”

 32

4th edition, South-Western.

World Bank. 2013. Government Effectiveness. [cited August 22nd, 2014.

Available from http://info.worldbank.org/governance/wgi/index.aspx#doc

World Bank. 2014a. Protecting Investors Methodology [cited August 23rd 2014].

http://www.doingbusiness.org/methodology/protecting-investors

World Bank. 2014b, Paying Taxes Methodology [cited August 23rd 2014].

http://www.doingbusiness.org/methodology/paying-taxes

World Bank. 2014c. Trading Across Borders Methodology [cited August 23rd

2014]. http://www.doingbusiness.org/methodology/trading-across-borders

World Bank. 2014d. Enforcing Contracts Methodology [cited August 23rd

2014]. http://www.doingbusiness.org/methodology/enforcing-contracts.

World Bank. 2014e.Registering Property Methodology[cited August 23rd 2014].

http://www.doingbusiness.org/methodology/registering-property

Yearwood, Nichelle. 2014. Regulation and Economic Growth in the Caribbean.

M.Sc. Research Paper in Business and Financial Economics, The University

of The West Indies, Cave Hill Campus.

 33

Appendix

Table 1: Disaggregated Regulation Indices from 2014 data.

Registering Property Index Ease of Doing
Business rank
Guyana 0.231538087 2
Dominican Republic 0.236622121 1
Jamaica 0.264900459 4
Belize 0.275452595 3
Dominica 0.288729233 9
Barbados 0.294309745 8
Grenada 0.298164113 5
Antigua and Barbuda 0.299618598 7
St Lucia 0.305472929 6
St Vincent and the Grenadines 0.319421506 10
Trinidad and Tobago 0.338865746 11
St Kitts and Nevis 0.340411925 12
Suriname 0.36148413 13
Bahamas 0.377706128 14

Protecting Investors Index Ease of Doing

Business rank
Trinidad and Tobago 0.336207 2
Antigua and Barbuda 0.372701 1
Dominica 0.372701 5
Grenada 0.372701 7
St Kitts and Nevis 0.372701 5
St Lucia 0.372701 7
St Vincent and the Grenadines 0.372701 3
Guyana 0.476437 6
Jamaica 0.476437 3
Dominican Republic 0.510057 4
Bahamas 0.543678 7
Belize 0.580172 8
Barbados 0.717529 10
Suriname 0.821264 9
 34

Paying taxes Index Ease of Doing

Business rank
Dominican Republic 0.112385 5
Bahamas 0.126 1
Belize 0.175142 2
Suriname 0.175415 4
St Lucia 0.185105 3
Barbados 0.191044 6
Grenada 0.193809 9
St Vincent and the Grenadines 0.211099 7
Dominica 0.215221 8
Guyana 0.217744 11
Trinidad and Tobago 0.227306 10
St Kitts and Nevis 0.239335 12
Jamaica 0.251375 13
Antigua and Barbuda 0.329579 14

Trading across borders Index Ease of Doing

Business rank
Barbados 0.122570702 2
Dominican Republic 0.130794841 1
Grenada 0.138830716 4
St Vincent and the Grenadines0.139161067 3
Bahamas 0.158003462 5
St Kitts and Nevis 0.159193992 6
Antigua and Barbuda 0.184266807 10
Trinidad and Tobago 0.185146681 7
Guyana 0.185368742 8
Belize 0.187603571 11
Dominica 0.192799793 9
St Lucia 0.220461092 14
Jamaica 0.221192518 13
Suriname 0.223337495 12
Enforcing Contracts Index Ease of doing
business rank
Dominican Republic 0.276863 2
Guyana 0.290156 1
Antigua and Barbuda 0.31449 3
Jamaica 0.337806 6
St Vincent and the Grenadines 0.339181 4
St Kitts and Nevis 0.377951 5
Bahamas 0.392368 7
St Lucia 0.424426 9
Grenada 0.426109 8
Dominica 0.431567 10
Barbados 0.460185 11
Belize 0.508156 12
Trinidad and Tobago 0.527604 13
Suriname 0.634442 14

Source:Data from the WB Doing Business File.

Note: Index built from formulas (1) and (2).
 35

Table 2: Panel data estimation results for aggregated regulatory model (7) and

model (8): The Caribbean, 2004-2012.

Dependent variable: logged GDP per capita (2005 US)

Random Effects Fixed Effects

Variables Coefficient Std Coefficient Std

C 8.992* .213

WB aggregate regulatory -2.496* .442 -2.575* .450

Government effectiveness .169* .064 .131** .067

population growth -.005 .060 -.011 .062

logged domestic investment .048*** .034 .050*** .034

logged foreign direct investment .119* .021 .118* .026

Natural Disasters Dummy .092* .019 .096* .020

Hausman p-value of chi-squared

statistic=14.34 (19)=0.0000
(p-value=0.026)

Notes: Fixed effects model (7) in within form and random effects model (8) are of

interest; *, **, *** mean statistically significant at the 1%, 5% and 10% levels,

respectively. Std stands for standard error. Hausman statistic tests for random

effects vs fixed effects. P-value of chi-squared tests for the overall fit of the fixed

effects model.
 36

Table 3: 2SLS fixed effects for testing for endogeneity of WB aggregate

for aggregated regulatory Model (7): The Caribbean, 2004-2012.

Dependent variable: logged GDP per capita (2005 US)

Variables Coefficient Std

WB aggregate -3.218* .673

Government effectiveness .148** .067

population growth .007 .063

logged domestic investment .050*** .034

logged foreign direct investment .120* .026

Natural Disasters Dummy .103* .020

Residual1 1.154 .903

p-value of  20
2
 0.000

Note: see Table 2. Residual1 represents the residuals from the regression of WB

aggregate on all exogenous or predetermined variables of the model, which are the

lagged aggregated regulatory measure and the five other explanatory variables of

model (7) plus a constant term. Std stands for standard error.
 37

Table 4. Panel data estimation results: the disaggregated regulatory models

(9) and (10): the Caribbean, 2004-2012.

Dependent variable: Logged GDP per capita (2005 US)

Random Effects Fixed Effects

Variables Coefficient Std Coefficient Std

C 7.730* .306

WB Starting a business -.936** .435 -.580 .495

WB Registering Property .309 .315 .156 .323

WB Protecting Investors -.204 .272 -.179 .295

WB Paying Taxes -.672** .317 -.719** .344

WB Enforcing Contracts 2.109* .644 1.282** .753

WB Trading across Borders -1.138* .403 -1.097* .408

Government Effectiveness .122** .074 .024 .081

Population growth .052 .051 .052 .054

Logged Domestic Investment .078* .040 .073* .031

Logged Foreign Direct .113* .026 .119* .026

investment
Natural Disasters dummy .034 .028 .021 .029

Hausman statistic p-value of

(11)=39.96; p-value  242  0.000
=0.000
Note: variables are explained in the text. WB: World Bank. See also Note to Table

2. *, **, and ***:significant at the 1%, 5%, and 10% levels, respectively.