THE UNDERGROUND ECONOMY IN THE CARIBBEAN COUNTRIES

MAGAZZINO COSIMO¹*, BUSCEMI ANTONINO², HAGOS YALLWE ALEM^3
1School Of Politicalsciences, Roma Tre University, Italian Economic Association (SIE) Royal Economic Society (RES)
²University Of Rome Tor- Vergata, Department Of Economics, Law and Institutions
³University Of Rome Tor- Vergata, Department Of Economics, Law and Institutions
* Corresponding Author : cmgazzino@uniroma3.it

Received : 14-11-2011     Accepted : 03-12-2011     Published : 05-12-2011
Volume : 2     Issue : 2       Pages : 124 - 132
Int J Econ Bus Model 2.2 (2011):124-132

Conflict of Interest : None declared

The purpose of this study is to analyze the underground economy in the Caribbean coutries for the period 1970-2009, using annual data with a panel approach, including a set of social and institutional variables (the degree of urbanization, the Index of Globalization, the Index of Participation and the Index of Competition). After a brief introduction, a survey of the economic literature on this issue is shown, before estimating the determinants of the underground economy. The higher tax burden, the higher the tax evasion and higher currency holdings relative to M2. Some notes on the policy implications of this analysis conclude the paper.

Underground economy, Caribbean countries, tax burden, governance.

O17, D78, H11, H26.

The underground economy is a widespread phenomenon which might involve a lot of economic sectors. It consists of various types of activities, ranging from domestic work to registered businesses that underestimate their sales and overestimate their expenditures. Since informal economy data are not reflected in the national statistics, trying to estimate its size and value could be particularly troublesome. In fact, if the informal sector is not included in the official macro-indicators, these will be downward biased, precluding international comparisons, given the inaccurate data collection procedure. As Feige (1990) stressed, the size and growth of unreported income and the implied tax gap affect the size of government deficits, government debt and tax reform policies.
Recognizing its policy relevance, researches on the black economy started in the Fifties. Kaldor (1956) and Cagan (1958) marked the early beginnings of this kind of research. Yet, there are still some scepticism on measuring its size and evolution.
The aim of this paper is to analyze the underground economy in the Caribbean area, studying the effects of such macro variables on currency demand in a panel framework, from 1970 to 2009. In fact, few studies on this issue have been devoted to a homogeneous group of countries. Moreover, as stated by Greenidge et al. (2009) in the Caribbean very little research has been done on the topic of the informal economy.
The outline of the paper is organized as follows. The succeeding section provides clarifying concepts in order to define the underground economy. While Section 3 provides data and methodological issue, Section 4 presents the empirical results. The last section winds up with concluding remarks and some policy implications.

In the literature, several terms are commonly used to define the unmeasured economy: informal, hidden, underground, invisible, shadow, unofficial, unrecorded, parallel, clandestine and subterraneous. To a large extent, these differences in terminology and definitions reflect differences in research objectives (Gërxhani, 1999; Schneider and Enste, 2000). In their taxonomy, Mirus and Smith (1997) distinguish informality by whether activities are legal or illegal, and by whether they involve monetary or non-monetary barter transactions. While Dreyden and College (1996) classify the informal economy into three categories: (1) “pure” tax evasion, (2) the irregular economy, and (3) illegal activities.
Generally, individuals are motivated to participate in the underground economy either to avoid the payment of taxes or to prevent the loss of any government assistance (Bajada, 1999). But another two motives could be represented by: a) inflation, since the rise in the cost of living can increase the real tax burden and so provide a possible motive for working in the underground economy (O’Higgins, 1985); b) friends and relatives: in fact, another motive for participating in the underground economy is an individual’s knowledge that friends and/or relatives are working in the subterranean economy (Cowell, 1990).
It is sometimes suggested that participation in the underground economy can offer an escape from economic hardship in the official economy. As Henry (1981) also suggests, it may also offer individuals access to goods and services that may not be available in the official economy because of inadequate demand.
Any economic activity that is not computed in the statistics of the national income is considered as a part of the hidden economy. Table 1 below shows some of the various ways, both legal and illegal, in which people can participate in the underground economy.
Castells and Portes (1989) present several hypotheses to explain the growth of the undergroun sector. They suggest the expansion of the informal economy is:
• Part of the process of economic restructuring following the structural crisis of the 1970s.
• A reaction against the state’s regulation of the economy.
• The result of increasing international competition.
• The process of industrialization in many developing countries.
• The result of poverty in which millions of people subject to harsh living conditions are forced to accept any solution to their misery.
Measuring the informal economy is not an easy task. Depending on how and what one chooses, there are about fourteen methods which have been used to estimate the size of the underground economy (Georgiou, 2007). These approaches can be broadly dividend in to two major approaches: direct and indirect.
Direct (or micro) approaches include micro-surveys of the informal economy, tax audits and other compliance methods.
Vuletin (2008) classified indirect approaches as : (i) the discrepancy between national expenditure and income statistics; (ii) the discrepancy between the official and actual labour force; (iii) the “electricity consumption” approach by Kauffman and Kaliberda (1996); (iv) the “transaction” approach by Feige (1979); (v) the “currency demand” approach by Cagan (1958); and (vi) the “Multiple Indicators Multiple Causes” (MIMIC) approach of Frey and Weck-Hanneman (1984). The transaction and currency demand approach are together called “monetary” approach.

Gutmann (1977) provided the first published estimate of the size of the underground economy in the United States and his work subsequently generated considerable interest.
CBA (1980) and Carter (1984) have made contributions to discussing illicit economic behaviour in Australia, while Bajada (1999) showed that the underground economy has been growing rapidly particularly since 1994. Moreover, on the basis of Granger causality tests, he concluded that disturbances in the official economy, which affect the level of income, dominate the effects of the same disturbance generated by rising levels of unemployment.
A research on the Bulgarian shadow economy using the physical input method indicated that the relative proportion of the shadow economy from the official economy has fluctuated between 25% and 37% of GDP over the 1989-1998 period, with a declining trend since 1996 in the relative proportion of the shadow economy (AA.VV., 2001).
Dell’Anno (2003) using the Italian shadow economy pinpointed two relevant results: (i) the relationship between underground economy and growth rate of GDP is negative; according to Frey and Weck-Hanneman (1984), on the contrary to the MIMIC estimated in New Zealand (Giles, 1999a) and Canada (Giles and Tedds, 2002); (ii) the growth rate of currency outside of banks is not relevant as indicator of the Italian shadow economy.
Faal (2003) demonstrated the existence of a large underground economy in Guyana, which averaged about 40 percent of GDP in the Seventies, before increasing sharply to an average of 76 percent of official GDP during the Eighties, reflecting excessive Government regulation and inward-looking economic policies. During the Nineties the size of underground economy fell, even though at an average of 47 percent.
Chaudhuri et al. (2005) estimate the size of the hidden economy using state level data from India over the period 1974 to 1996. On average, the size has grown from 13.1 percent to 26.3 percent showing a growth rate of 3 percent per annum. They also have shown that an increased growth of per capita newspaper circulation helps to curb the growth in the size of the shadow economy activities.
Maurin et al. (2006) measured the underground economy of Trinidad & Tobago using annual time series data covering the period 1970-1999, within the Structural Cointegrating VAR (SCVAR) framework. The size of the underground 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.
Dell’Anno et al. (2007) offered estimations of the evolution of the underground economy in three Mediterranean countries (France, Spain and Greece). The results confirm that unemployment, the fiscal burden and self-employment are the main causes of the underground economy in these countries, and confirm that an inverse relationship exists between the official GDP growth rate and that of the unofficial economy.
The size of the informal economy and the relative contribution of each underlying factor in 32 Latin American and Caribbean countries in the early 2000s have been provided by Vuletin (2008). He found that a burdensome tax system, rigid labor markets, higher inflation, and dominance of the agriculture sector are the key factors in determining the informal economy, representing altogether around 79 percent of the informal economy variance.
The estimates provided by Greenidge et al. (2009) – concerning the Barbadian informal sector for the period 1972-2007 – indicate that the underground sector is quite large and has grown over time to about one-third the size of the official economy.
Arby et al. (2010) hypothesized that in a country like Pakistan, tax is not the only factor that induces the economic agents to be in informal but the lack of education also keeps them away from a formal and recordable way of economic activity. The results indicate a negative coefficient of the variable for education. The ARDL approach shows that the underground economy increased from less than 30 percent in 1960s to 33 percent in 1990s and then declined to 23 percent on average in 2000s. While the electricity consumption approach shows that unmeasured economy increased from about 5 percent of total size of the economy in 1970s to 29 percent in 1990s and then declined to 27 percent in 2000s. Finally, the MIMIC model shows that the informal economy has been around 30 percent.
The estimates for Ethiopia due to Asaminew (2010) revealed that the informal economy was widespread between 1977-1991, since the country was troubled by civil war and policy instability. During this period, the shadow economy reached an average size of 41.5% of the recorded economy. After the Nineties reform packages, the size of the informal economy declined to 30%.
Biswas et al. (2010) studied the effects of the shadow economy on pollution levels as well as on corruption levels in the public administration. Using panel data covering the period 1999-2005 for more than 100 countries, the empirical findings confirm that the relationship between the shadow economy and the levels of pollution are dependent on the levels of corruption.
The nexus between religion and shadow economy has been investigated by Heinemann and Schneider (2010), for 162 countries around the world. The empirical findings indicate that countries with close ties between state and religion have a lower share of economic activity in the informal economy. This is in line with the view that religion uses its normative influence to protect state interests if there is a close alliance between both.
The size of the underground economy in Japan was empirically tested by Kanao and Hamori (2010). The results show that (i) the size of the underground GDP peaked in the early 1990s but has been declining since; (ii) the underground economy reached its peak in 1992, approximating 25% of nominal GDP.
Schneider et al. (2010) presented estimations of the underground economies for 162 countries, over 1999 to 2007. According to their findings, the weighted average size of the underground economy (as a percentage of “official” GDP) in Sub-Saharan Africa is 37.6%, in Europe and Central Asia 36.4% and in high-income OECD countries 13.4%. Moreover, econometric analyses show how an increased burden of taxation (direct and indirect ones), combined with (labour market) regulations and the quality of public goods and services as well as the state of the ‘official’ economy are the driving forces of the shadow economy.
Kholodilin and Thieben (2011) studied the extent of the shadow economy in 38 OECD member states over the period 1991-2007 is investigated. The ranking of countries is consistent with that of other studies: among the industrial countries above average are, for instance, Greece, Italy and Portugal. Below average are some Scandinavian countries and the USA. On the other hand, Eastern European countries have relatively high underground economies between 4% and 6% of official GDP, while Turkey has an estimated underground economy of somewhat below 10% of GDP.

The monetary approach of estimating underground economy was developed on the Cagan’s (1958) proposition that higher tax rates induce people to use currency for transactions to avoid tax reporting. The pioneering works in this area include Gutmann (1977), Feige (1979), and Tanzi (1983). This approach consists of specifying a demand for currency equation to be used to derive the effect of a change in the tax level on that demand. To estimate our model, which includes a set of social and institutional variables, we can summarize this in the following specification for currency demand:
C/M2 = f(Y, r, p, R, G, E, F, OT, U, IG, IP, IC) (1)
where C/M2 = the ratio between currency and money supply (M2), Y = real per capita income, r = the real interest rate, p = the inflation rate, R = the real Government revenue, G = the real Government expenditure, E = the private consumption expenditure, F = the domestic credit to private sector, OT = the openness to trade, U = the degree of urbanization, IG = the Index of Globalization, IP = the Index of Participation, IC = the Index of Competition, (see the [Table-1] for more details).
In using the regression approach, we find evidence of persistence in the dependent variable, while the explanatory variables are endogenous. When such econometric problems exist, the traditional panel data estimators (Pooled OLS, Fixed Effects and Random Effects models) don’t yield consistent estimates. The GMM dynamic panel data methods, however, can simultaneously deal with the problem of persistence and endogeneity. We, therefore, focus our discussion of results to this estimator as argued below. The use of GMM dynamic panel data model requires that the error term of the equation in levels be white noise. This is tested by checking the absence of second order autocorrelation in the differenced equation. Building on the work of Anderson and Hsiao (1981), Arellano and Bond (1991), Arellano and Bover (1995) and Blundell and Bond (1998) developed a system estimator (GMM-SYS) that uses additional moment conditions.
We constructed a balanced panel for twelve Caribbean countries: Antigua and Barbuda, Bahamas, Barbados, Dominica, Dominican Republic, Grenada, Haiti, Jamaica, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, and Trinidad and Tobago. In [Table-2] the variables of the model are summed up. All series contain yearly data, with the specific source.
The competition variable portrays the electoral success of smaller parties, that is, the percentage of votes gained by the smaller parties in parliamentary and/or presidential elections. The variable is calculated by subtracting from 100 the percentage of votes won by the largest party (the party which wins most votes) in parliamentary elections or by the party of the successful candidate in presidential elections. The variable thus theoretically ranges from 0 (only one party received 100 % of votes) to 100 (each voter cast a vote for a distinct party).

In this section we present and discuss the results of our econometric analyses. [Table-3] reports the descriptive statistics for the variables used in the estimation. There is a wide cross-country variation in the currency ratio for Caribbean countries ranging from 10 to 77 percent.
A simple correlation coefficient between two variables indicates a linear relationship between the variables without controlling for the effect of other potential explanatory variables. To gauge the separate effect of each explanatory variable after controlling for the effects of other variables, we further investigate the relationship using regression analyses.
The [Table-5] contains the empirical findings of our regression analyses. The dynamic specification of empirical models leads the LSDV estimator to be inconsistent and unbiased when T is fixed and N goes to infinity (Verbeek, 2008). In this case, instrumental variable estimators are used to solve this econometric issue.
The first two columns represent the Pooled OLS and Within Groups results, respectively. In the third column are shown the estimates for the Anderson-Hsiao First-Differenced 2SLS estimator. With this instrumental variable regression, the dependent variable is specified in first-differences. We treat the year dummies and the country-specific dummies as (strictly) exogenous, and also included in the instrument set. Moreover, we use as instruments also the first and second lag of the dependent and explanatory variables. In fact, we treat as endogenous the lag dependent variable in first-differences. The fourth column contains the results for the Arellano-Bond First-Differenced GMM, using the finite-sample correction proposed by Windmeijer (2005). We include only the second and third lags of the level of our variables, as well as the year and country dummies, in the instrument set. Only the equations in first-differences are used in estimation. In the last column the GMM-System estimates are presented. We use a 2SLS as the one-step estimator. Moreover, the lagged levels of C/M2, Y, r, p, R, G, E and F dated from t-3 to t-4 are the instruments for the equations in first-differences; while the lagged first-differences of these variables dated t-2 (only) are the instruments for the equations in levels. As instruments for the equations in levels only, the year and country dummies have been chosen.
For the specification (1), the Hansen’s (1982) J statistic chi-squared test is reported. A significant test statistic indicates that the instruments may not be valid; here, we don’t reject the null hypothesis. Then we perform a test to determine whether endogenous regressors in the model are in fact exogenous (the C difference-in-Sargan statistic is reported): since the test statistic is not significant, the variables being tested must be treated as exogenous. The specification test pertains to testing the presence or absence of second order autocorrelation. The result of the Arellano and Bond (1991) test reports an autocorrelation statistic with a P-Value greater than 0.05, implying that the null is not rejected at 5 percent. We, therefore, conclude that there is no evidence for second order autocorrelation.
Most of the coefficients have the expected signs. The coefficient of interest rate is negative, reflecting the opportunity cost on holding currency; while the coefficient on real per capita income is positive, reflecting the fact that economic growth acceleration would arise a corresponding increase in the use of currency, following increases in the aggregate demand. Moreover, the lagged value of the currency ratio is highly significant, which implies that the dependent variable is persistent.
The higher tax burden, the higher the tax evasion and higher currency holdings relative to M2, other things being equal. Given these results on Government revenue, it can be said that countries with high fiscal burden experience rampant tax evasion; if high tax rates can be backed by strong enforcement capabilities, however, tax evasion would be small (Embaye, 2007).
In the last equation, we report the results of robustness checks on the coefficient estimates of the variables of interest. The robustness check is undertaken by including the openness to trade, the urbanization degree, the Index of Participation and the Index of Competition as additional conditioning variables. The results of including these covariates in our estimation demonstrate that they are significant. In fact, this inclusion increases the robustness of the significance of the dependent variable.

The purpose of this paper is to analyze the underground economy in twelve Caribbean countries, studying the effects of such macro variables on currency demand in a panel context. To this extent, we have studied the determinants of currency/M2 ratio in a model which includes a set of social and institutional variables (the degree of urbanization, the Index of Globalization, the Index of Participation and the Index of Competition) using annual data for the period 1970-2009. According to our econometric results, the higher tax burden, the higher the tax evasion and higher currency holdings relative to M2, other things being equal. Given these results on Government revenue, it can be said that countries with high fiscal burden experience rampant tax evasion; if high tax rates can be backed by strong enforcement capabilities, however, tax evasion would decrease. So, a significant portion of the underground economy can be converted to the recorded economy through optimal tax policy.
As stated in Bajada (2009), “it is imperative, therefore, that policy makers need to seriously consider the rules, regulations, taxes and welfare benefits that may have ignited enthusiasm for underground activity while launching initiatives to combat growing areas of noncompliance”.
The main outcomes of this paper, in accordance with the conclusions founded by Schneider (1997), Schneider and Enste (2000), Dell’Anno (2003), Forte and Magazzino (2010) and Magazzino (2008, 2009, 2010, 2011) are that a rising of taxes and social security contributions, combined with increasing State regulatory activities, […] are the major driving forces behind the size and growth of the shadow economy” (Schneider and Enste, 2000).

[1] AA.VV. (2001) The Shadow Economy in Bulgaria, Harvard University – Agency for Economic Analysis and Forecasting – Institute for Market Economics, April.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[2] Anderson T.W., Hsiao C. (1981) Journal of the American Statistical Association, 76, 598-606.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[3] Arby M.F., Malik M.J., Hanif M.N. (2010) The Size of Informal Economy in Pakistan, State Bank of Pakistan Working Paper Series, May, 33.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[4] Arellano M., Bond S. (1991) Review of Economic Studies, 58, 277-297.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[5] Arellano M., Bover O. (1995) Journal of Econometrics, 68, 29-51.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[6] Asaminew E. (2010) The Underground Economy and Tax Evasion in Ethiopia: Implications for Tax Policy, October, typewritten.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[7] Bajada C. (1999) The Economic Record, 75, 231, December, 369-384.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[8] Baltagi B.H. (2000) Nonstationary Panels, Cointegrating Panels and Dynamic Panels, New York: Elsevier.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[9] Biswas A.K., Farzanegan M.R., Thum M., (2010) Pollution, Shadow Economy and Corruption: Theory and Evidence, typewritten.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[10] Blundell R., Bond S. (1998) Journal of Econometrics, 87, 115-143.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[11] Cagan P. (1958) Journal of Political Economy, 66, 3, 302-328.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[12] Carter M. (1984) Economic Record, 60 (170), 209-221.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[13] Portes A., Castells M., Benton L.A., (1989), The Informal Economy: Studies in Advanced and Less Developed Countries, John Hopkins University Press.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[14] Commercial Bank of Australia (CBA) (1980) The Underground Economy in Australia, CEA Economic Review, September.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[15] Cowell F.A. (1990) Cheating the Government, Cambridge: MIT Press.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[16] Dell’Anno R. (2003) Estimating the Shadow Economy in Italy: a Structural Equation Approach, Department of Economics – University of Aarhus Working Paper, 2003-07.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[17] Dell’Anno R., Gómez-Antonio M., Alañon-Pardo A. (2007) Empirical Economics, 33:197.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[18] Dreyden A., College W. (1996) Beating the System, in Pozo, S. (ed.), Exploring the Underground Economy: Studies of Illegal and Unreported Activity, Kalamazoo, Michigan: W.E. Upjohn Institute for Employment Research.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[19] Embaye A. (2010) Underground economy estimates for non-OECD countries using currency demand method, 1984-2005, MPRA Paper, 20308, January.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[20] Faal E. (2003) Currency Demand, the Underground Economy, and Tax Evasion: The Case of Guyana, IMF Working Paper, January, WP/03/7.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[21] Feige E.L. (1979) How Big is the Irregular Economy?, Challenge, 22, 1, 5-13.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[22] Feige E.L. (1994) Public Finance, 49 (Supplement), 119-136.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[23] Forte F., Magazzino C. (2011) Journal of Analytical and Institutional Economics, XXVIII, 3, 295-321.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[24] Frey, B., Weck-Hanneman, H., (1984), The Hidden Economy as an “Unobservable” variable, European Economic Review, 26, 1, 33-53.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[25] Georgiou G.M. (2007) Measuring the Size of the Informal Economy: A Critical Review, Central Bank of Cyprus – Euro System, Working Paper Series, 2007-1.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[26] Gërxhani K. (1999) Informal Sector in Developed and Less Developed Countries: A Literature Survey, Tinbergen Institute and University of Amsterdam.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[27] Giles E.A. (1999) Modeling the hidden Economy in the Tax-gap in New Zealand, Department of Economics – University of Victoria Working Paper, Canada.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[28] Greenidge K., Holder H., Mayers S. (2009) Business, Finance and Economics in Emerging Economies, 4, 1.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[29] Gutmann P.M. (1977) Financial Analyst Journal, 26-34.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[30] Hansen L.P. (1982) Econometrica, 50, 1029-1054.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[31] Henry S. (1981) Can I Have it in Cash?-A Study of Informal Institutions and Unorthodox Ways of Doing Things, London: Astragal Books.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[32] Kaldor N. (1956) Indian tax reform: report of a survey, Ministry of Finance – Government of India.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[33] Kaliberda A., Kaufmann D. (1996) Integrating the Unofficial Economy into the Dynamics of Post-Socialist Economies: A Framework of Analysis and Evidence, The World Bank Policy Research Working Paper Series, 1691.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[34] Kanao K., Hamori S. (2010) The Size of the Underground Economy in Japan, MPRA Paper, March, 21562.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[35] Kholodilin K., Thieben U. (2011) The shadow economy in OECD countries: Panel-data evidence, Deutsches Institut für Wirtschaftsforschung Discussion Papers, 1122, May.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[36] Magazzino C. (2008) Modelli interpretativi della dinamica della spesa pubblica e “curva di Armey”: il caso italiano, 1862-2001, Notizie di Politeia, XXIV, 92, 45-60.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[37] Magazzino C. (2009) Economia Pubblica, 5-6, 133-156.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[38] Magazzino C. (2010) Il Risparmio Review, LVIII, 1, January-March, 5-33.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[39] Magazzino C. (2011) The nexus between public expenditure and inflation in the Mediterranean countries, Theoretical and Practical Research in Economic Fields.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[40] Maurin A., Sookram S., Watson P.K., (2006) International Economic Journal, 20(3), 321-341.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[41] Mirus R., Smith R. (1997) Canada’s Underground Economy: Measurement and Implications, in Lippert, O., Walker, M., (eds.), The Underground Economy: Global Evidence of its Size and Impact.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[42] O’Higgins M. (1985) The Relationship Between the Formal and Hidden economies: An Exploratory Analysis for Four Countries, in Gaertner, W., Wenig, A., (1985), The Economics of the Shadow Economy – Proceedings of the International Conference on the Economics of the Shadow Economy, Germany: Springer-Verlag.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[43] Sargan J.D. (1958) Econometrica, 26(3), 393-415.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[44] Sargan J.D. (1988) Testing for Misspecification after Estimation Using Instrumental Variables, in Maasoumi, E., (ed.), Contributions to Econometrics: John Denis Sargan, Cambridge: Cambridge University Press.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[45] Schneider F. (1997) Economic Affairs, 17, 3, 42-48.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[46] Schneider F., Enste D. (2000) Journal of Economic Literature, 38, 77-114.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[47] Schneider F., Buehn A., Montenegro C.E., (2010) International Economic Journal, 24, 4, December, 443-461.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[48] Tanzi V. (1983) The Underground Economy in the United States: Annual Estimates, 1930-1980, IMF Staff Papers, 33, 283-305.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[49] Vuletin G. (2008) Measuring the Informal Economy in Latin America and the Caribbean, IMF Working Paper, WP/08/102.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[50] Windmeijer F. (2005) Journal of Econometrics, 126, 25-51.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

	Fig. 1- Currency/M2 ratio and Government revenue in the Caribbean countries in 2000. Source: our elaborations.
	Table 1- Types of underground economic activities. Source: Mirus and Smith (1997).
	Table 2- List of the variables. Source: our elaborations.
	Table 3- Exploratory data analysis. Sources: WB, IMF, Vanhanen, Dreher – KOF and PWT.
	Table 4- Correlation matrix. Notes: Bonferroni adjustment applied. Sources: WB, IMF, Vanhanen, Dreher – KOF and PWT.
	Table 5- Panel data estimations. Notes: (1) POLS, (2) Within Groups, (3) IV (GMM-Dif), (4) and (5) GMM-System; a: Adjusted R2, b: R2 within. Number of groups = 12. Year dummies are included in all models. Asymptotic standard errors in parentheses. For the diagnostic tests P-Values are reported. Sources: WB, IMF, Vanhanen, Dreher – KOF and PWT