Once again, I find insightful observations (with statistical underpinnings) on Professor Mark Perry’s “Carpe Diem” blog. Here’s a common sense explanation of at least part of the recent years’ decline in U.S. household income.
Not surprisingly, as a greater percentage of people retire and/or go on “disability” (an increasingly abused government/court perk), the average/median household income declines.
Some demographic trends that might explain the stagnation and decline in US household income
The two charts above show some interesting patterns in several data series over time and provide some possible demographic explanations for the stagnation and decline in real median US household income since around the year 2000. Here are some observations and comments:
1. The top chart above shows that the decline in real median household income in the US from the peak of $57,843 in 1999 to $53,627 last year (in 2014 dollars, data here in Table H-12) was accompanied by a gradual increase in the share of US households with no earners, which increased from less than 20% in 1999 to 24% in 2014. A regression analysis over the 1999-2014 period shows that every one percent increase in the share of US households with no earners is associated with a decrease in median annual household income of slightly more than $1,200. (Note: The R-squared of that regression was 89.3% and the t-statistic for the independent variable was -10.83.) Over that same period, the share of US households with two or more earners declined from slightly more than 45% in 1999 to 39.2% in 2014, which is another demographic factor that could explain the decline in median household income over the last decade.
2. What would explain the fact that the share of US households with no earners has steadily risen over the last 15 years to an all-time of 24% last year? The bottom chart above provides one explanation: the rising share of the US adult population represented by: a) retired workers and b) disabled workers. According to Social Security Administration data, the number of retired workers plus the number of disabled workers remained stable at about an 18% share of the US adult population (18 years and over) between 1993 and about 2000 before gradually rising to an all-time high of 21.3% by 2014.
For example, in the ten-year period between 2004 and 2014, the number of retired workers increased by 27% (from 33 million to nearly 42 million) and the number of disabled workers increased by 37.5% (from 7.95 million to nearly 11 million). Because the adult population increased by less than 11% during that decade, the share of the US adult population represented by retired and disabled workers increased from 18.3% to 21.3% between 2004 and 2014 and was by far the largest increase over any previous 10-year period in the SSI data back to 1970. As can be seen in the bottom chart above, the rise in the share of retired and disabled workers over the last decade accompanied the decline in US median household income over that period. A regression analysis shows that every 1% increase in the share of the US adult population represented by retired and disabled workers is associated with a $1,200 decline in median household income.
Bottom Line: Perhaps the stagnation and decline in US household income that gets so much media and political attention isn’t necessarily the result of the usual negative factors that get cited so frequently: stagnating wages, reduced economic and employment opportunities for the average, middle-class American, the increased share of rising income or wealth going to the top X%, the hollowing out of the middle class, the claims that the middle class is shrinking/losing ground/disappearing/declining, etc.
Rather, perhaps there’s a less-nefarious, demographic-driven reason that household incomes have been stagnating/declining in recent years — the increase in the share of US households with no earners, which is largely driven by the aging US population and the increasing number of retired workers, and to a lesser extent by the increasing number and share of disabled workers. Finally, there’s been nearly a six percentage point decline in the share of US households with two or more earners since 1999, which could be another demographic change that has contributed to a decline in median household income.
With some of the significant changes outlined above in important demographic factors that have taken place over the last decade or so: an increase in the share of US households with no earners, a decrease in the share of US households with two or more earners, an increase in the number and share of US adults who are retired or disabled, and a gradual decline in the average household size (from 2.67 to 2.54 over the last 20 years), along with the devastating effects of the Great Recession on the US economy and household incomes, it maybe would actually be a surprise if there hadn’t been a decline in median household income in recent years.