Income inequality refers to the distribution of income from all sources across all individuals or families or households. Incomes may be measured over a week or a month or a year or even longer, and before or after receipt of benefits and payment of taxes and housing costs. The choice between all these variations on income depends both on data availability and on the purpose of the analysis being performed.
There are a number of ways of measuring income inequality. One of the most straightforward is the P90/P10 ratio, which measures the gap between rich and poor as the ratio between incomes at two points in the whole distribution, chosen to represent ‘high’ and ‘low’ incomes respectively. It shows the ratio of the 90th income percentile (income level which only 10% of the population exceed) to the 10th percentile (the level which only 10% have less than). If this ratio goes up, the gap between rich and poor is growing.
The Gini coefficient (G) is more complicated and incorporates incomes at all levels of the distribution. G can take any value between 0, if all individuals had the same income, and 1 or 100% if a single individual received all the income in the economy. More generally, a G value of say 0.25 indicates that the income difference between any pair of individuals drawn at random is expected to be 2xG i.e. 50% of mean income. Rising G means that the gap between rich and poor is growing.
In the graph we show these two measures applied to disposable income in the United Kingdom – the amount of money that households have available for spending and saving after direct taxes (such as Income Tax and Council Tax) have been accounted for. It includes earnings from employment, private pensions and investments, as well as cash benefits provided by the state. Any given household income obviously provides a much higher standard of living for a single person than for a family of four. Household incomes are therefore adjusted or ‘equivalised’ using the modified OECD equivalence scale to produce a distribution of incomes across individuals. Equivalized income is household income divided by the number of ‘adult equivalents’. Children carry a smaller weight than adults when calculating the number of adult equivalents.
The two measures show the same general picture. There was a pronounced increase in income inequality in Britain in the 1980s. Since 1990 the two indicators have more or less flat-lined and so they are both still well above their 1977 levels.
Underreporting of high income households
These measures are based on incomes reported in household surveys. Two potential biases arise in recording incomes of the very rich: (a) rich households may not declare all their income sources when responding in surveys; (b) rich households may be under-sampled in the survey design. Rich households account for a small proportion of all households but account for a large proportion of total income. Under-sampling and income under-reporting will mean that inequality measures, especially the Gini, are under-estimated.
Income inequality measures in the UK from 2002 onwards have additionally been adjusted by the Office for National Statistics (ONS) to correct for these potential biases. The adjustment is based both on a detailed analysis of the rich households covered in the survey and additional information from tax data. The ONS find evidence to support the view that the main source of bias is that of under-reporting by survey respondents rather than under-coverage.
The adjusted Gini coefficients are expectedly higher from 2002 onwards. There has been a slight increase in income inequality since 2017. The Gini coefficient has increased from 33.4% in 2016 to 34.6% in 2019. According to the ONS, ‘this mainly reflects a fall in disposable income for the poorest 20% of people between FYE 2018 and FYE 2019.’