How reliable are income data collected with a single question?

Income is an important correlate for numerous phenomena in the social sciences. But many surveys collect data with just a single question covering all forms of income. This raises questions over the reliability of the data that are collected. Issues of reliability are heightened when individuals are asked about the household total rather than own income alone. We argue that the large literature on measuring incomes has not devoted enough attention to 'single-question' surveys. We investigate the reliability of single-question data by using the UK Office for National Statistics's Omnibus survey and the British Social Attitudes survey as examples. We compare the distributions of income in these surveys-individual income in the Omnibus and household income in the British Social Attitudes survey-with those in two larger UK surveys that measure income in much greater detail. Distributions compare less well for household income than for individual income. Disaggregation by gender proves fruitful in much of the analysis. We also establish levels of item non-response to the income question in single-question surveys from a wide range of countries. Copyright (c) 2010 Royal Statistical Society.