This paper employs micro-level data to uncover the factors that determine
individuals' decision to engage in payroll and/or income tax evasion in Estonia. Using
cross-section data and modern microeconometric methods, we estimate the marginal
effects on tax evasion of different individual characteristics. To increase the reliability
of our results, we use data from four different sources: (1) Self-reported tax evasion
data from surveys conducted by the Estonian Institute for Economic Research. (2)
The Estonian tax board's compliance data based on the results of auditing corporate
bodies. (3) Household survey data on income, saving and consumption. (4) Data from
labour force surveys where respondents self-report otherwise unregistered income.
We generally find that men evade taxes more frequently than women. Persons with
low registered income generally engage in tax evasion more frequently than highincome
earners. Surprisingly, tax evasion appears to be equally frequent among
different age groups. We interpret the latter result as a consequence of the Estonian
transition process, where especially older persons have experienced problems getting
jobs in the formal part of the economy. From firm’s side of factors, the large number
of workers in the company lowers the probability of evasion. Our results provide
inputs for discussion of the auditing and fine system in Estonia, and also to the overall
tax system.