Small-Business Participation in the Informal Sector of an Emerging Economy

Abstract

We investigate the characteristics of the owners of small businesses that participate in the informal sector of an emerging economy and their perception of the risk of detection by tax authorities while doing so. Data are gathered from a survey covering 1027 small businesses in Trinidad and Tobago. Results suggest that small business owners are motivated to participate in the informal sector when they believe that the risk of detection by the tax authorities is low and that government regulations are burdensome, but there is no evidence that the tax rate itself is an issue. Their perception of the risk of detection by the tax authority is determined largely by the time they spend and the income they earn in the formal sector.

Notes

1. This is similar to the definition of Rampersad (1987) who, in a study of Trinidad and Tobago, defines the informal sector as that section of the economy which is: (a) not recorded in the official statistics; (b) not legal; and (c) not taxed. She further sub-divides the sector into a ‘visible’ informal sector, an ‘invisible’ informal sector (that part of the informal sector which covers illicit and socially disvalued activities) and a domestic sector.

2. Various labels have been used to describe what, in this paper, we call the informal sector: the informal economy (Kim, 2005), the hidden economy (Maurin et al., 2006), the subterranean economy (Gutmann, 1977), the shadow economy (Schneider, 2005), the black economy (Pissarides and Weber, 1989).

3. The Central Statistical Office of Trinidad and Tobago defines a small enterprise as one that employs five or less persons.

4. Maurin et al. (2006) use the currency demand model to estimate the size of the informal sector in Trinidad and Tobago for the period 1973–1999.

5. Other studies carried out on Trinidad and Tobago include Rampersad (1987), Lloyd-Evans and Potter (2002) and Maurin et al. (2006).

6. Hoyman (1987) finds that women are likely to engage in informal activities at the same level as men, if not more, in some instances.

7. See also the work of Dillman (1978) for the advantages and disadvantages of the three survey interfaces: face-to-face, mail and telephone. He concludes that the face-to-face method is a more favourable method in a greater number of instances.

8. The frame includes firms of self-employed professionals, like doctors and lawyers who, although registered with the tax authority, have a greater potential than salaried workers to evade taxes (Pissarides and Weber, 1989; Lyssiotou et al., 2004).

9. Johnson et al. (1999) who use survey data from private manufacturing firms in Poland, Romania, Russia, the Slovak Republic and Ukraine, determine that small businesses have better opportunities for tax evasion, for example through the route of the informal sector, because there is less risk of detection (due to their size). Similar considerations influence sample selection in other informal sector studies (Bose, 1978; Deshpande, 1985).

10. The reference modality is the element that is not used in the estimation of the model.

11. The purpose of this normalisation is to identify the model's parameters. See Greene (2003) for a more detailed discussion of the estimation procedure.

12. Using the log likelihood of the full model (lnL₁) with the one from the constant only model (lnL₀), the likelihood ratio test statistic is computed as follows:

13. It is possible that some of the variables in this equation may be simultaneously determined (particularly income, participation and ‘hours spent in the regular economy’). The estimation technique used in this case will yield biased results and this should be borne in mind when interpreting the results shown in Table 2.

14. Giese and Hoffman (1999) employed the experiment method to investigate tax evasion and its determinants in Germany. Thirty-eight students, with equal amounts of males and females, participated in the experiment.

15. Gërxhani and Schram (2001) examine tax evasion in Albania and the Netherlands using a series of separate experiments on high school pupils, high school teachers, university students, university academic staff and university non-academic staff.

16. This variable is in index form as indicated in Table 1.

17. Anderson (1998) using survey data from Mongolia's informal sector during the transition period found that the average age of persons involved in this sector was 37 years.

18. It seems reasonable to suggest that a small business owner's decision to participate in the informal sector will not be a factor determining his/her perceived risk of participation, at least in the static framework used in this paper. The same caution about possible simultaneity bias in the case of the results shown in Table 2 also apply here, since some of the variables in this equation may be simultaneously determined. This should be borne in mind when interpreting the results shown in Table 3.

19. Several authors (see Long, 1997; Long and Freese, 2006) prefer the McKelvey-Zavoina to the McFadden R² on the grounds that it more closely approximates the R² obtained from regressions on the underlying latent variable.