(12.) For example, for Greece, Matsaganis and Flevoto mou (2010) find that "farming or self-employment incomes are more likely to be underreported in tax returns: average underreporting rates for these two sources are 53% and 24%, respectively" (19).
Correcting for underreporting of self-employment income using our preferred correction factor reduces the measured agricultural productivity gap by 38%.
Motivated by empirical evidence, we focus on underreporting of self-employment income as a source of mismeasurement of value added.
Given the evidence indicating underreporting of self-employment income and the systematic positive relationship between mismeasurement and self-employed share of hours in agriculture, we evaluate the quantitative impact of underreporting of self-employment income on measured agricultural productivity gap.
We therefore conclude that mismeasurement of value added because of underreporting of self-employment income can account for a significant portion of the measured agricultural productivity gaps in Europe.
Section IV explores whether underreporting of self-employment income could explain the mismeasurement.
The first has been increasing reliance on household-based survey data, which are known to under-report investment-related and self-employment incomes relative to others.
Jenkins found much greater inequality during 1971 - 86 in both RDI and self-employment incomes than in employment income or private pensions (Table 8).
Households with high investment and self-employment incomes tend to be under-represented and, when represented, to under-report those two categories of income in particular.
In terms of levels, the shares of investment-related and self-employment incomes in household disposable income are markedly higher in the national accounts than in the FES.
Survey-based estimates of factor shares in household incomes suffer not only from the two problems identified here in Blue Book estimates but also from systematic under-reporting of investment and self-employment incomes. The latter problem has been seen to have distorted only marginally FES-based estimates of trends in factor shares during the first decade or so of the contemporary rise in inequality.
Rather, factor shares matter as compositional weights in the generation of overall inequality, given that the distributions of investment-related income and self-employment income are much more unequal than is that of employment income.