False Negatives: Earnings Underreporting, Tax Overreporting in Surveys Worldwide

Household survey incomes are subject to misreporting and measurement issues biasing the static and dynamic assessment of inequality and poverty. Non-positive incomes are particularly problematic as they represent extreme statistics in income distributions, are incompatible with sustainable consumption streams, and cannot be squared with households’ observed behaviors and other socio-economic outcomes. We find that, in high income countries, the main source of extremely low disposable incomes is unduly high reported tax and social security withholdings. In transitional economies between the upper-middle- and high-income status, the main sources are negative self-employment income, followed by negative capital income and high tax liabilities. Lower down, among middle- and low-income countries, negative self-employment incomes play a leading role. Meanwhile, households with negative incomes appear to be typically as well off as, or better off than other households in terms of material wellbeing, both according to the short-term and long-term indicators. By contrast, zero-income households appear to be materially poor. We surmise that zero or small negative incomes correspond predominantly to chronically deprived households who temporarily fall into material poverty, while large negatives correspond to chronically well-off households under-reporting earnings, or writing off capital losses or tax assessments from surrounding years. Hence, ‘tax overreporting’ may explain low incomes – especially the zeros – in high income countries, while ‘earnings underreporting’ plays a greater role in upper-middle- and lower-income countries. Underestimation of rental values among homeowners is one specific challenge.