Inter-generational Incomes: A Tempest Is Brewing

We are all aware of the admonition caveat emptor or buyer beware, but we should also add reader beware for studies that purport to deal with complex issues in facile ways. One such recent example is the McKinsey Global Institute's July 2016 Report "Poorer than Their Parents," that examines an important policy issue of inter-generational income patterns, but does not do the topic justice. In fact the report is methodologically flawed, while also miscasting its findings and missing the main policy issues.

The report claims that 65-70 percent of households in 25 countries included in the analysis, or 540-580 million people, experienced flat or falling market incomes between 2005 and 2014. This is in contrast to fewer than 10 million whose income remained flat or fell in the period 1993-2004. This is startling, especially if one doesn't dig into the reasoning employed in the report by MGI, usually a reliable source of policy relevant information. In reality these numbers are based on an examination of data for six countries, then extrapolated to a larger set of 25 countries, an approach no serious researcher would endorse. Moreover, the much more important finding, even with these data, is that disposable incomes were either flat or falling for only 20-25 percent of households, meaning that taxes and redistributive transfers dealt successfully with a full two-thirds of the market income decline. If correct, this is a major government policy victory.

But can we trust these data and findings emanating from them? I would have serious doubts. Consider two stark cases: the U. S. and Italy, two of the six countries where actual data were used. In the case of the U.S., the report claims that 81 percent of households had their incomes rebalanced to actually increase disposable income (MGI pp. 16,62). This finding does not square with usually reliable OECD data on the pre and post tax plus transfers income distribution measures for the U.S. and is at odds with other findings that median incomes have fallen and that the middle class (the larger central portion of the income distribution) has seen significant declines in real incomes (see Pew Foundation Study). Should we really believe that 98 percent of American households experienced disposable income gains in the 2005-2014 period? (MGI pp. 3,60).

In the case of Italy, the findings presented are internally inconsistent. On the one hand, we are told that all quintiles experienced large market as well as disposable income declines, while Italy's pre and post transfer Gini coefficients differ by a third (MGI, p. 59). The data on redistribution shows a huge difference in income inequality before and after taxes and transfers, yet the disposable incomes of all groups suffer? Perhaps the measures of income distribution used are inaccurate, possible since the data on the Gini coefficients differs markedly from those collected by the OECD.

The notion that offspring will fare less well than their parents is probably quite correct for many advanced economies, but the supporting evidence provided is unconvincing. As a first course of correction, one might distinguish between flat and falling incomes. Flat incomes for the current generation would be quite acceptable in light of many studies on faltering productivity, overly generous transfers to retirees, sharply rising life expectancies, and declining return to labor. To this list of potential threats to incomes of current workers must be added the risks associated with disruptive, labor-displacing technologies (see Brynjolfsson and McAfee, 2014).

Linking those whose incomes are likely to decline to weaker skills acquisition and poorer educational attainment is a logical link. On the other hand to indicate, as MGI does, that perceptions survey responses supported the report's findings is inaccurate since between 60-70 percent felt their incomes were not falling or didn't venture an opinion (MGI, p. 3). Again, the finding that among " those who felt incomes were falling" or the pessimists, one in three felt that their children's incomes would fall is a biased figure begging for better interpretation (MGI, p. 33). Finally the contributions to income changes that are derived from factors such as aggregate demand, demographics, labor market phenomenon and other factors are not based on rigorous country specific growth or income generation models and must also be hugely discounted.

Why bring this up? The reason is that there are significant inter-generational income issues at play that will affect the future distributions of income in many advanced economies, and increasingly in upper middle-income countries as well. This reality stems from declining working populations in some countries like Japan and Korea, increased longevity in most, high health care costs and competing claims on public coffers, and reduced growth prospects (see Spence et. al., 2015). There will be difficult choices to be made concerning public spending on health and pensions versus expenditures on growth-enhancing investment. It may well be that the current generation will end up poorer than the last (although these measures may need adjustment to take into consideration the shared economy, leisure-work tradeoffs etc.); however, the debate is not advanced with weak analysis and false claims. There is more serious work to be done. So, reader beware!


Brynjolfsson, E., & McAfee, A, (2014). The Second Machine age: Work, Progress, and Prosperity in a Time of Brilliant Technologies. New York, NY: Norton

McKinsey Global Institute (2015), "Poorer than Their Parents? Flat or Falling Incomes in Advanced Economies" McKinsey & Company

OECD (2015), "Redistribution of income", in Government at a Glance 2015, OECD Publishing. DOI:

Pew Research Center (2015), "The American Middle Class Is Losing Ground: No longer the majority and falling behind financially"

Spence, M, D. Leipziger, J. Manyika and R. Kanbur (2015), "Restarting the Global Economy: Harnessing the Forces of Economic Growth", Growth Dialogue Bellagio White Paper