Abstract
This paper compares the prevalence and determinants of unreported employment in the three Baltic countries in 1998 and 2002 using a hitherto little used data set. The prevalence of unreported employment varies substantially across the three countries and across the two sampling years. Microeconometric estimations show that firm-related characteristics, such as sectoral activity, firm size and employment changes, are important determinants of unreported employment in all three countries. The impact of socio-demographic factors, such as gender, age and education, is generally less important and varies across countries and time. Only a small part of the changes in unreported employment between 1998 and 2002 can be accounted for by changes in firm-specific factors and socio-demographic characteristics. Exploratory calculations suggest that the gain for individuals undertaking unreported employment is modest or non-existent, in particular among individuals who engage regularly in such activities. This suggests that the many of the recipients of envelope wages may have few alternatives to accepting unreported employment.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
The WLB data set does not contain data on the reporting of the income from work undertaken on an own-account basis (self-employment).
OECD (2008) draws together information from a number of sources on informal employment in Poland, the Czech Republic, Slovakia and Hungary. In the Czech Republic, Slovakia and Hungary relatively few individuals are completely without formal income, while this is not the case in Poland.
See also the discussions of the difference between unreported work and unreported employment in Williams (2009d) and Commission (2007a, Sec. 4).
The data set also contains information as to whether the firm is foreign-owned, the marital status of the respondent and the size of the respondent's household, but only for 1998. However, approximately half of the observations are missing for these variables. When we added these variables in baseline estimations in Tables 3, 4 and 5 for 1998, most of the marginal effects became statistically significant, possibly because of the small sample size. Among the additional variables, the only significant effect was observed for the foreign ownership variable in Latvia, where employment by a foreign-owned company reduced the probability of unreported employment by 14%. Moreover, the use of the variables would rule out the pooling of data for the two survey years. We therefore decided against including the variables in the regressions presented in the section ‘Determinants of Unreported Employment’.
Indicating that respondents with a second job are somewhat more prone to receive envelope wage, cf. also the estimation results in the section ‘Determinants of Unreported Employment’.
Comparing the logit and probit models, the logit model predicts slightly higher probabilities for the lower and higher end of the distribution, while the probit model predicts slightly higher probabilities for the middle part of the distribution. We have examined the importance of using logit instead of probit in the main results presented in Table 6. For all three countries, the Hosmer and Lemeshow goodness of fit test does not reject the null hypothesis that there are no differences between the observed and predicted probabilities for logit or probit models. The estimated marginal effects are also qualitatively similar.
We have experimented with the inclusion of non-linearities of the age variable. In one estimation both a linear and a quadratic term was included, and in another estimation the logarithmic age was included. The results are, however, similar in the sense that the age variables do not attain statistical significance in the case of Estonia.
The shares of firms experiencing increasing or decreasing employment are relatively large. It is noticeable, however, that the region experienced rapid structural change in the period considered. Masso et al. (2006) found for the same period and countries that 25%–30% of all jobs were reallocated annually.
If both the age and the quadratic age are included, the quadratic age does not become statistically significant. The logarithmic age is statistically significant if it is included instead of the linear age, but the overall estimation results do not change to any noticeable extent.
The probability of unreported employment is 4% lower for a 60-year-old respondent than for a 25-year-old respondent in 1998, and 7% lower in 2002. As in the case of Latvia, the logarithmic age would also have attained statistical significance, but the overall estimation results would have changed little.
Another possibility is reverse causality, implying that firms that pay envelope wages to many of their employees are thriving and therefore expand their employment.
A joint Wald test cannot reject the hypothesis that the constraints implied by the use of occupational position and education instead of the dummies are valid.
The Stata command nldecompose for the Oaxaca-Blinder decomposition of nonlinear regression models is used for the decomposition, see Sinning et al. (2008).
The decomposition is based on the estimated coefficients; neither the coefficients nor the associated marginal effects are shown in this section.
The estimations in Table 6 for the two years merged found that the dummies for year 2002 were of substantial magnitude, broadly following the changes in the unconditional rates of unreported employment and highly significant in statistical terms. This suggests that changes in the constant term are a very important driver of the change because of changed coefficients.
Arias and Khamis (2008) employ a more sophisticated technique explicitly taking into account the endogenous selection between different employment categories. This approach, however, induces additional layers of complexity.
This effect is most pronounced in the calculations for 1998, but it is also present in some calculations for 2002.
On a scale between never justified (1) and always justified (10), the mean score was 3.15 for Estonia, 2.36 for Latvia and 3.77 for Lithuania. Only Latvians found tax evasion less justified than the EU25 average (score 2.57), while the Lithuanians were the most tolerant of tax evasion in the whole of EU25 (World Values Survey, 1999).
References
Abbott, P and Wallace, C . 2009: Patterns of participation in the formal and informal economies in the commonwealth of independent states. International Journal of Sociology 39 (2): 12–38.
Alm, J and Torgler, B . 2006: Culture differences and tax morale in the United States and in Europe. Journal of Economic Psychology 27 (2): 224–246.
Antila, J and Ylostalo, P . 1999: Working life barometer in the Baltic States 1999. Labour Policy Studies, No. 214, Ministry of Labour, Helsinki, Finland.
Antila, J and Ylostalo, P . 2003: Working life barometer in the Baltic States 2002. Labour Policy Studies, No. 247, Ministry of Labour, Helsinki, Finland.
Arias, O and Khamis, M . 2008: Comparative advantage, segmentation and informal earnings: A marginal treatment effects approach. IZA Discussion Paper, no. 3916.
Commission. 2007a: Undeclared work in the European Union. Special Eurobarometer, No. 284, European Commission.
Commission. 2007b: Taxation trends in the European Union. Data for the EU member states and Norway. Eurostat Statistical books, European Commission. Office for the Official Publications of the European Communities: Luxembourg.
Cowell, F . 1990: Cheating the government: The economics of evasion. MIT Press: Cambridge, MA.
Cramer, JS . 1999: Predictive performance of the binary logit model in unbalanced samples. The Statistician 48 (1): 85–94.
Danopoulos, CP and Znidaric, B . 2007: Informal economy, tax evasion and poverty in a democratic setting: Greece. Mediterranean Quarterly 18 (2): 67–84.
EKI. 2005: Varimajandus Eestis [The Estonian shadow economy]. Estonian Institute of Economic Research, http://www.ki.ee/publikatsioonid/valmis/Varimajandus_Eestis_2004(elanike_hinnangute_alusel).pdf.
Fabrizio, S and Mody, A . 2008: Breaking the impediments to budgetary reforms: Evidence from Europe. IMF Working Paper, WP/08/82, International Monetary Fund.
Kim, B-Y . 2005: Poverty and informal economy participation. Evidence from Romania. Economics of Transition 13 (1): 163–185.
Kriz, K, Meriküll, J, Paulus, A and Staehr, K . 2008: Why do individuals evade payroll and income taxation in Estonia? In: Pickhardt, M and Shinnick, E (eds). Shadow Economy, Corruption and Governance. Edward Elgar Publishing: Cheltenham, pp. 240–264.
Maloney, W . 2004: Informality revisited. World Development 32 (7): 1159–1178.
Masso, J, Eamets, R and Philips, K . 2006: Job flows and worker flows in the Baltic States: Labour reallocation and structural changes. In: Paas, T. and Eamets, R. (eds). Labor Market Flexibility, Flexicurity and Employment. Nova Science Publishers: New York.
Meriküll, J and Staehr, K . 2008: Unreported employment and tax evasion in mid-transition: Comparing developments and causes in the Baltic States. Working Paper of Eesti Pank, no. 6/2008.
OECD. 2000: Baltic States. A regional economic assessment. OECD Economic Surveys, Organisation for Economic Co-operation and Development: Paris.
OECD. 2008: OECD employment outlook. Organisation for Economic Co-operation and Development Ch. 2 (Declaring Work or Staying Underground: Informal Employment in Seven OECD Countries), http://browse.oecd.bookshop.org/oecd/pdfs/browseit.
Pissarides, C and Weber, G . 1989: An expenditure-based estimate of Britain's black economy. Journal of Public Economics 39 (1): 17–32.
Rodgers, P, Williams, C and Round, J . 2008: Workplace crime and the informal economy in Ukraine: Employee and employer perspectives. International Journal of Social Economics 35 (9): 666–678.
Round, J, Williams, C and Rodgers, P . 2008a: Corruption in the post-Soviet workplace: The experiences of recent graduates in contemporary Ukraine. Work, Employment & Society 22 (1): 149–166.
Round, J, Williams, C and Rodgers, P . 2008b: Everyday tactics and spaces of power: The role of informal economies in post-Soviet Ukraine. Social and Cultural Geography 9 (2): 171–185.
Schneider, F . 2002: The size and development of the shadow economies of 22 transition and 21 OECD countries. IZA Discussion Paper, No. 514.
Schneider, F and Enste, D . 2000: Shadow economies: Size, causes and consequences. Journal of Economic Literature 38 (1): 77–114.
Sedlenieks, K . 2003: Cash in an envelope: Corruption and tax avoidance as an economic strategy in contemporary Riga. In: Arnstberg, K-O and Boren, T (eds). Everyday Economy in Russia, Poland and Latvia. Almqvist & Wiksell: Stockholm, pp. 37–52.
Sinning, M, Hahn, M and Bauer, T . 2008: The blinder-oaxaca decomposition for nonlinear regression models. Stata Journal 8 (4): 480–492.
Smith, A and Stenning, A . 2006: Beyond household economies: Articulations and spaces of economic practice in postsocialism. Progress in Human Geography 30 (1): 1–14.
Tanzi, V . 1999: Uses and abuses of estimates of the underground economy. Economic Journal 109: F338–FF347.
Williams, C . 2008a: Envelope wages in central and Eastern Europe and the EU. Post-Communist Economies 20 (3): 363–376.
Williams, C . 2008b: Illegitimate wage practices in Eastern Europe: The case of ‘envelope wages’. Journal for East European Management Studies 13 (3): 65–83.
Williams, C . 2009a: Evaluating the extent and nature of ‘envelope wages’ in the European Union: A geographical analysis. European Spatial Research and Policy 16 (1): 116–129.
Williams, C . 2009b: Formal and informal employment in Europe: Beyond dualistic representations. European Urban and Regional Studies 16 (2): 147–159.
Williams, C . 2009c: The commonality of envelope wages in Eastern European economies. Eastern European Economics 47 (2): 37–52.
Williams, C . 2009d: The hidden economy in East-Central Europe: Lessons from a ten-nation survey. Problems of Post-Communism 56 (4): 15–28.
Williams, C . 2009e: The prevalence of envelope wages in the Baltic Sea region. Baltic Journal of Management 4 (3): 288–300.
WLB. 1998: Working Life Barometer dataset for 1998. WLB: Helsinki.
WLB. 2002: Working Life Barometer dataset for 2002. WLB: Helsinki.
World Bank. 2009: Governance matters 2009. Worldwide governance indicators, 1996–2008, http://info.worldbank.org/governance/wgi/index.asp.
World Values Survey. 1999: http://www.worldvaluessurvey.org/ (Online Data Analysis → 1999–2004 → Estonia [1999], Latvia [1999], Lithuania [1999] → 7. Religion and morale → Justification of social behaviours → Cheating on taxes (F116)).
Acknowledgements
The authors would like to thank two anonymous referees, David Grigorian and Martti Randveer for excellent comments. Pekka Ylöstalo from the Finnish Ministry of Labour kindly provided the Working Life Barometer data set and answered many questions. All errors remain the responsibility of the authors. The views expressed are those of the authors and do not necessarily represent the official views of Bank of Estonia.
Author information
Authors and Affiliations
Appendices
Appendix A
Ordered logit
Appendix B
Unordered explanatory variables
See Table B1.
Rights and permissions
About this article
Cite this article
Meriküll, J., Staehr, K. Unreported Employment and Envelope Wages in Mid-Transition: Comparing Developments and Causes in the Baltic Countries. Comp Econ Stud 52, 637–670 (2010). https://doi.org/10.1057/ces.2010.17
Published:
Issue Date:
DOI: https://doi.org/10.1057/ces.2010.17