Reducing evasion seems to be a priority for all revenue agencies, especially now when public finances are stretched. Self-employment income suffers from greater underreporting than employment income. Having observed a rapid increase in the number of self-employed individuals in the recent past, the question about how much the self-employed underreport is of upmost relevance.

Research to be presented at the Royal Economic Society''s annual conference in Brighton in March 2016 investigates the extent of income underreporting by the self-employed in Great Britain for the period 2010-12 and the characteristics of the individuals most associated with non-compliance with the aim of answering not only ''How Much'' but also ''Who''.

The researchers – Ana Cinta Cabral, Christos Kotsogiannis and Gareth Myles – find that self-employment income is underreported on average by 19.3%, which translates into a lower bound estimate of unreported taxable income of 1.6% of GDP.

Data on expenditures and income recorded in the Living Costs and Food Survey for the period 2010-2012 are used. The method exploits the discrepancy between the income and expenditure patterns of the self-employed and the employed to infer non-compliance.

In a nutshell, in the survey the self-employed report higher levels of expenditure than the employed but lower levels of income. Given that employment income is subject to withholding taxes (PAYE) and self-employment income lacks of third party reporting, income reported by the employed is recognised to be accurately reported while self-employment income can be misreported.

Therefore, the relationship between expenditure and income for the employed can be classified as correct and from it the amount of income the self-employed truly hold in order to sustain the level of expenditure they claim can be inferred.

This estimation is undertaken using different items of expenditure (food, utilities and a basket of nondurables) with similar results: self-employment income needs to be multiplied on average by 1.24 to obtain true income.

The study then proceeds to assess how the discrepancy between expenditure and income for the two groups maps into underreporting and successfully discards other alternative potential explanations such as different preferences between the two groups or different way of funding current expenditure. The critical assessment of the link of the estimate to underreporting gives more credibility to these estimates.

The researchers investigate further which characteristics of the individuals are mostly associated to noncompliance using not only characteristics that typically appear on tax returns but also a wider range of socioeconomic characteristics (''Who?'').

They find that men are less compliant than women and that individuals become more compliant as they age. Regarding occupations, blue-collar households seem to underreport more than white-collar. Self-employed operating in partnerships are found to underreport more than own account self-employed and the presence of more than one self-employed in the household seems to give more opportunity to evade.

Overall, this study responds to a need for a reassessment of the extent of income underreporting by the self-employed as it is becoming an increasingly common occupational choice in the UK.

Not only does it provide an estimate using the expenditure-based method but it critically assesses how this estimate links to underreporting. It moves one step further from simply estimating the extent of underreporting to analysing the patterns of noncompliance to be able to inform tax policy choices.

The findings of this study are useful to inform the revenue agency about the individuals to whom compliance activities should be directed in order to increase their impact given that resources are limited.

Self-Employment Underreporting in Great Britain: Who and How much?