Household income equality: changing families or changing pay-cheques?

Household income equality: changing families or changing pay-cheques?

You are unlikely to have come across a discussion of the US economy recently without a reference to income inequality and wage stagnation. It is widely accepted that the Gini Index, a measure of the distribution of income in the economy, has increased markedly from the 1960s; that is, incomes have become more unequal. There has also been documented median wage stagnation. Real median wages in America peaked in 1999 at $57 843 in 2014-equivalent dollars.[1] Such data are often used as definitive insight into the labour market and the incomes of American citizens. However, the most commonly used data are household numbers. Unfortunately, as household numbers conflate both the composition of households and economic trends into a single data point, such an indicator ultimately gives clear information about neither.

In fact, change in income data in the US is wholly fitting with changes evident in household composition and demography. Downward pressure on the median wage is created by an ageing population and an increase in part-time work. Moreover, changes in household structure also have a significant effect on household income inequality.

Interestingly, it is possible for the median household income to decrease even with rising wages. As the population ages, workers leave the workforce and retire. Before they retire, they are near their highest earning potential, having accrued experience through years in the workforce. However, their post retirement income can be expected to be significantly lower. As this demographic cohort moves beyond their peak earning potential and start retiring, it is expected that wages would decrease. Additionally, those aged 20-24 years old, who represent another large cohort of the population, will have recently entered or are soon to enter the workforce at their lowest earning potential – likely below the median wage.

There has also been a notable uptick in single-income households. According to the American Enterprise Institute, there were 52% more households in the US in 2013 than 1980, yet the population had only increased 39%. Per capita income is spread more thinly among households, and so median household incomes decline even if individual incomes remained constant or increased over the period. Multiple-income households have concurrently fallen from 45% in 1999 to under 40% in 2016.[2] Just over 50% of adults currently co-reside with a spouse, down from 70% in 1967, whilst the number of adults living alone has increased from 8% to 14%.[3] It is no surprise that households with fewer income earners put downward pressure on median income.

In addition, Americans are working less. In recent history, average weekly hours worked peaked along with median income in 1999. Since then it has fallen by 8.8%, while in the same period, median household income has fallen by 7.2%.[4] Considering the drop in median household income in this way asks a very different set of questions to wage stagnation. While the drop could be a case of underemployment, it may also be a sign of shifting preference around work-life balance and increasing availability of desirable part time work.

These trends combine to make household income a less reliable indicator of real wages in the economy. It conflates social and demographic trends with labour market behaviour in a way that hides causality. This is not to say that median wages have not stagnated, but that the extent of the stagnation is likely to be overstated when households are used as the unit of analysis.

Furthermore, societal change has similarly impacted the Gini Index. A 2014 study by Greenwood, Guner, Kocharkov and Santos[5] found that nearly all of the increase in income inequality is the result of a phenomenon called assortative mating, where people create household units with a spouse similar to themselves. It found an increase in the Gini Index from 33 in 1960 to 43 in 2005. Although the study notes that the change in wages across individuals is “the primary driver of this increase in income inequality,” the effects have been significantly amplified by concurrent increases in female labour force participation and assortative mating. Two married doctors earn significantly more than a single waiter, for instance. Increased opportunities for women also mean more female doctors, and through assortative mating, they are likely to marry male doctors. Accordingly, the prevalence of high earning dual-income households increases.

Intuitively, if these high earners were spread more evenly through households, then it would result in a more even spread of income across households. Additionally, it would pull up the median wage. In 2005, couples who both had graduate degrees earned 119% above the median wage. If such individuals had not formed a household together, remained single or married low or non-earners, both of their households would still likely be above the median income level. By controlling for marriage preferences by randomising marriages in the population, the study found that the Gini Index had barely moved, from 33 to 34.[6]

There is little doubt that inequality is worsening in America. Technological change is hitting the middle income routine and administrative jobs hardest, leaving only high skilled, highly paid jobs and lower paid, manual and service jobs in their wake. But it is important to recognise the limits of commonly used measures of inequality. Evidently, using households as the unit of analysis can obscure changes caused by the labour market by conflating it with demographic and social change.


Image: Flickr

[1] U.S. Bureau of the Census (2014). Real Median Household Income in the United States [MEHOINUSA672N]. FRED, Federal Reserve Bank of St. Louis. Retrieved July 2, 2016 from

[2] Perry, M. J. (2016). Update: How changing household composition, household work hours, and retirement explain median household income. AEIdeas. Retrieved from

[3] U.S. Bureau of the Census (2015). Current Population Survey, Annual Social and Economic Supplement, 1967-2015. Retrieved from

[4] Perry, M. J. (2016). Update: How changing household composition, household work hours, and retirement explain median household income. AEIdeas. Retrieved from

[5] Greenwood, J. Guner, N. Kocharkov, G. & Santos, C. (2014). Marry Your Like: Assortative Mating and Income Inequality. American Economic Review, 104(5), 348-53.

[6] Greenwood, J. Guner, N. Kocharkov, G. & Santos, C. (2014). Marry Your Like: Assortative Mating and Income Inequality. American Economic Review, 104(5), 348-53.