Measuring the Size of the Hidden Economy in Trinidad & Tobago, 1973-1999

Measuring the Size of the Hidden Economy
in Trinidad & Tobago, 1973-1999




Alain Maurin
Laboratoire d’Economie Appliquée au Développement
Université des Antilles et de la Guyane
Campus de Fouillole
Guadeloupe

Sandra Sookram
Economic Measurement Unit
Department of Economics
University of the West Indies
St. Augustine
Trinidad & Tobago

Patrick Kent Watson
Sir Arthur Lewis Institute of Economic & Social Studies
University of the West Indies
St. Augustine
Trinidad & Tobago

Tel: (868) 662-6965
Fax: (868) 645-6329
e-mail: pkwatson@fss.uwi.tt

A revised version of this paper is published in the International Economic Journal. 20(3): 321-341
(2006).



August 2004

---

ABSTRACT

In this paper, an attempt is made to measure the hidden economy of Trinidad & Tobago
using annual time series data covering the period 1970-1999, within the Structural
Cointegrating VAR (SCVAR) framework. Using a Tanzi-type currency demand
approach as a starting point, a multiple equation SCVAR model is estimated that contains
two long-run relationships linking the demand for currency with other variables. The
model is evaluated on the basis of its persistence profiles, its impulse responses and other
statistical criteria. It is solved using a Gauss-Siedel algorithm and establishes that the size
of the hidden economy rose from a low of about 14% of measured GDP in the early
1970s to a high of 36% in 1981, and is currently about 20% of measured GDP, with no
marked tendency to get larger in the near future. Hidden economic activity is also found
to be highly positively correlated with activity in the regular economy.

KEYWORDS: Caribbean, Trinidad & Tobago, Hidden Economy, Structural
Cointegrating VAR Models.

JEL CLASSIFICATION NUMBERS: C51, C52, E26, E60, O17, O54.

---

1. Introduction

In this paper, an attempt is made to measure the size of the hidden economy in
Trinidad & Tobago using annual time series data covering the period 1970-1999.
This is done within the Structural Cointegrating VAR (SCVAR) framework, applied
to a demand for currency function. The paper is the first in an ongoing research
project aimed at the identification and measurement of the hidden economy in certain
Caribbean countries. In fact it is intended that the specific case studied in this paper,
that of Trinidad & Tobago, serve as a template for future work on other countries of
the Caribbean.

Economists have long had an interest in the size of the hidden economy and, since
Cagan’s (1958) paper, voluminous studies1 about the size of the hidden economy have
been carried out for countries throughout the world. Gradually, interest has extended
to the Caribbean area. See in particular Faal (2003), Bennett (1995), Witter and
Kirton (1990) and Thomas (1989). To date, however, here has been no attempt to
measure the size of the hidden economy of Trinidad & Tobago, arguably the most
prosperous country in the Caribbean region and one where many conditions have
existed, and continue to exist, to ensure that hidden economic activity is widespread.
Two studies on the Trinidad and Tobago economy, Rampersad (1987) and Lloyd-
Evans and Potter (2002), yielded information on structural and other characteristics of
the informal sector, but no attempt was made in either to measure the size of the
hidden economy.


1 See Schneider and Enste (2000) for a fairly comprehensive survey of the literature.

---

What possible motivation can there be for determining the size of the hidden economy
in Caribbean countries like Trinidad & Tobago? Several general reasons are given in
the literature for conducting research in this area and perhaps the most widely cited,
which is applicable to the Trinidad & Tobago case, is the potential, in the absence of
information about the size of the hidden economy, for erroneous policy decisions
based on misleading statistical indicators. Indeed, economic policy measures may be
of a wrong magnitude or even in a wrong direction if they are based on such
indicators of the state of the economy. For example, the official unemployment rate
may be overstated if some of the officially unemployed are working in the hidden
economy. Similarly, the growth rate of real income may be understated if the hidden
economy is expanding more quickly than the “measured” economy, or the rate of
inflation may be overstated. Policy mistakes based on erroneous figures are
particularly costly for small developing countries like Trinidad and Tobago where
resources are very limited. Determining the size of the hidden economy in Trinidad
& Tobago and, indeed, in other countries of the Caribbean, is a worthy enterprise for
this reason alone.

There are many approaches to measuring the size of the hidden economy, perhaps as
many as there are definitions of the concept. See Schneider and Enste (2000). A
novel feature of this paper is that an estimate of the hidden economy of Trinidad and
Tobago is obtained using the Structural Cointegarting VAR approach. Indeed, this
paper represents the very first attempt to apply the SCVAR method to measure the
size of the hidden economy. SCVAR modeling is based on cointegration analysis,
and a major advantage of cointegration analysis, generally, is that it incorporates
information associated with both the long-run and short-run behaviour of economic


2

---

agents. The older studies, like those of Tanzi, did not take advantage of this and, as a
consequence, may have lost valuable information needed for estimating the size of the
hidden economy.

A particularly attractive feature of the SCVAR approach, as opposed to standard
cointegration or VAR analysis, is that it allows for the estimation of theory-consistent
long-run relationships between the variables in the system. The short-run dynamics
are freely estimated within a VAR framework. The model established may be
evaluated on the basis of its generalized impulse responses as well as on the
properties of its persistence profiles, another kind of impulse response that traces out
the response of error correction terms to a one-time shock in the vector of
disturbances, and may be interpreted as a measure of the speed of convergence toward
equilibrium. See Pesaran and Shin (1996), (1998), (2002) and Kilian (1999). These
generalized impulse responses and the persistence profiles are also extremely useful
in evaluating the impact of policy measures.

Using Tanzi’s (1980, 1982) currency demand approach as a starting point, we
estimate, using Eviews 5.0, a multiple equation SCVAR model containing two long-
run relationships linking the demand for currency with other variables and evaluate
this model on the basis of its persistence profiles, its impulse responses and other
statistical criteria. We solve this model using a Gauss-Siedel algorithm and determine
that the size of the hidden economy rose from a low of about 14% of measured GDP
in the early 1970s to a high of 36% in 1981, and is currently about 20% of measured
GDP, with no marked tendency to get larger in the near future.



3

---

The rest of the paper proceeds as follows. In the following section, we define what
we mean by the term “hidden economy” and present some reasons for the existence
and persistence of the hidden economy in Trinidad & Tobago. In Section 3, we
briefly review some empirical work done on the hidden economy in the English-
speaking Caribbean to date and in Section 4 we briefly discuss the currency demand
model that will form the basis of the SCVAR model. In section 5, we discuss the
fundamental features of SCVAR modeling. Section 6 is the central piece of the paper
where we set up the SCVAR model, estimate and evaluate it. In section 7, we use this
model to determine the size of production in the hidden economy over the period
1973-1999. In section 8 we draw some policy lessons and in section 9 we conclude
the paper.

2.
The hidden economy in Trinidad & Tobago: definition, existence and
persistence
The first obstacle in an exercise like this one is the definition of what constitutes the
hidden economy as well as an agreement about the appropriate terminology to be
employed. Like Faal (2003), we use the term “hidden economy” to refer to all
“unreported income which has contributed to value added according to the System of
National Accounts (1993)”, but which is not included in the official GDP measure
published in the National Income Accounts of Trinidad & Tobago, published by the
Central Statistical Office of that country. In so doing, we divide the total economy
into its hidden and measured, or regular, components.

Several authors use the term “hidden economy” to describe the phenomenon
discussed in this paper. These include the recent works of Giles (1999a, 1999b) and


4

---

Gadea and Serrano-Sanz (2002). However, there are many other concepts appearing
in the literature that are related, and sometimes equivalent, to what we will call in this
paper the hidden economy. These include terms like “shadow economy”, employed
by Helberger and Knepel (1988), “black economy”, employed by Pissarides and
Weber (1989), “underground economy”, employed by Faal (2003), Giles et al. (2002)
and Hill and Kabir (2000), and many other terms. Our definition incorporates activity
carried out in the so-called “informal economy”, or “informal sector” of the economy,
a concept which is widely used even in the Caribbean by statistical agencies, non
economists, or economists with special interests (like the labour market), to describe
economic activity that employs a handful of workers who earn low incomes, use
rudimentary equipment, and work outside the framework of laws and regulations.
The informal economy is perhaps better defined as the marginal (or even
marginalised) economy and it has long been of interest to Sociologists and Labour
Economists, if only for the effect that activity in this sector has on employment and
well-being. Examples of Caribbean studies of the informal economy include the work
on the Trinidad & Tobago economy by Rampersad (1987) and Lloyd-Evans and
Potter (2002), and on sub-sectors of the Jamaican economy by Smikle and Taylor
(1977) and LeFranc et al. (1989). In many of the countries of the Caribbean,
especially the smaller ones of the Eastern Caribbean, the distinction between the
hidden and the informal economies might be blurred in practice2.

Several reasons have been advanced for the existence and persistence of hidden
economic activity worldwide. See Schneider and Enste (2000), (2003) for a fairly

2 See Gërxhani (2003) for a very useful survey of studies done on the informal economy as well as for an appreciation of the
blurred lines between the concepts of “hidden economy” and “informal economy” even in the developed countries.


5

---

comprehensive review and analysis of such reasons. There is no doubt that many of
the major reasons advanced are valid for the Trinidad & Tobago case. Here we
consider the following:
• Increases in the tax burden;
• Intensity of government regulations;
• Perception of corruption;
• Discontent with quality of public services;
• Degree of ethnic fragmentation;
• Tax morale.
2.1
Increases in the Tax Burden Imposed on the Official Economy
The increase in the size of the hidden economy has been interpreted principally as a
reaction to the overburdening of firms and individuals by the apparatus of state,
principally through taxes. Most studies indeed show that the increase in the tax
contribution burden is one of the main causes for the increase of the hidden economy.
See for instance Schneider and Enste (2000), Cebula (1997), Johnson et al. (1998).
Figure 1 below shows the time path of the evolution of the Direct and Total Tax
burden (measured as a percentage of GDP) over the period 1970-1999, and if this
hypothesis is true, then the size of the hidden economy in Trinidad & Tobago should
reflect this pattern.
Figure 1
Evolution of Direct and Total Tax Rates in Trinidad & Tobago, 1970-1999


6

---

40
35
30
25
Direct Tax Rate
20
Total Tax Rate
15
10
5
1970
1975
1980
1985
1990
1995
Year
According to the hypothesis, individuals attempting to avoid paying taxes would turn
to the hidden economy, which would lead to a substantial loss in tax revenues. Even
if there should be an easing of the burden, the hidden economy may still persist. As
Spiro (1993, p. 255) points out, “once this habit is developed, it is unlikely that it will
be abandoned”, so that participants in the hidden economy are not likely to return to
the regulated economy, even in the long run.

At present there is no quantitative evidence on the existence of tax evasion in
Trinidad and Tobago but there is a considerable amount of anecdotal evidence.
Trinidad & Tobago has had its fair share of high taxes for most of the period covered
by the study. The Corporation tax rate stood at 45% until it was reduced in the 1990s
following the introduction of a 15% Value Added Tax (VAT), when it fell to 35%.
The marginal tax rate for individuals was even higher as Table 1 clearly shows:
Table 1
Marginal income tax rates in Trinidad & Tobago 1968-1999

Period
Income range
Marginal Income Tax
Rates


7

---

1968-1977 6,001-19,000
30-45%
19,001-60,000
50-60%
Over 60,000
70%
1978-1988 10,001-25,000
30-40%
25,001-60,000
45-60%
Over 60,000
70%
1990-1999 20,001-40,000
30-35%
Over 40,000
35-40%

The marginal tax rate for individuals was in the 1970s and 1980s as high as 70%, and
even after the VAT was introduced in 1990 the highest rate varied between 35-40%.
Details of the current taxation system in Trinidad & Tobago is outlined in Robinson-
Walters (2003). A summary of the system is shown in Appendix 2 to this paper.

2.2
Intensity of Government Regulations
An increase in the intensity of regulations3 tends to limit the freedom of choice for
individuals engaged in the official economy. Work by Johnson et al. (1997) predicts
that countries with more general regulation of their economies tend to have a higher
share of the hidden economy in the total GDP. Johnson et al. (1998) found that a one-
point increase in the Heritage Foundation’s regulation index, measured on a scale
going from 0 to 10, is associated with an 8.1 percentage point increase in the hidden
economy. The value of this index for Trinidad and Tobago has remained steadily
high over the years (averaging about 6.5) notwithstanding the “liberalization” of the
1990s and after. Those who publish the index have noted that “regulations and
bureaucratic red tape are burdensome” in Trinidad and Tobago.
An interesting and recent regulation, involving the regulation of minimum wages, was
introduced in Trinidad and Tobago in 1998. An econometric investigation by Strobl
and Walsh (2001) done after the implementation of this legislation showed that, while

3 This is usually measured by the number of laws and regulations, such as licenses requirements,
product market regulations and employment protection legislation.


8

---