The Pew Research Center reports that “From 2000 to 2014 the share of adults living in middle-income households fell in 203 of the 229 U.S. metropolitan areas…” Pew previously reported that “Fully 49% of U.S. aggregate income went to upper-income households in 2014, up from 29% in 1970. The share accruing to middle-income households was 43% in 2014, down substantially from 62% in 1970.”
(Income is defined as money income received exclusive of certain money receipts, such as capital gains, before payments for such things as personal income taxes, Social Security, union dues and Medicare deductions.)
The shrinkage is most easily seen in an article by Tom Edsall, in which there is a graph showing that between 1970 and high + low middle classes shrank from 65% to 41%, upper income rose from 17% to 30% and lower income rose from 19% to 30%.
I cannot find in Pew’s analysis any mention of the rise in the immigrant population in the U.S. Between 1980 and 2010, the foreign born population in the country rose from 6.2% to 12.9%. The foreign-born share of the workforce from 6.7% to 16%. An explanation of the decline of the middle class can’t be sufficient without taking this demographic into account.
What Pew’s and everyone else’s analysis overlook, is that the foreign born workforce is distinctly more bi-modal than the native-born workforce. That is, this workforce has relatively more people in the lower and higher income segments and relatively less in the middle income brackets. This workforce has grown faster — much faster — than the native born workforce, especially the white workforce.
This more bi-modal profile subsides in the second generation; that is, children of workers in 1980, who today are in there 20s to 40s, but on the whole college graduation remains much lower. It stands to reason that the increase in the immigrant workforce, both first and second generation, contribute, perhaps significantly, to shrinking the size of the middle-income brackets.
I turn to a Bureau of Labor Statistics report on labor force in 2014. As if often the case, the use of median or average figures obscures the distribution. But the BLS report has some figures that are worth looking at.
In 2014, 48.3% of the foreign-born labor force was Hispanic, and 24.1% was Asian.
This first table is typically how foreign and native born workers are compared. A much higher percentage of native born workers are in higher status jobs:
Foreign born 30.7% 24.1%
Native born 39.8% 16.4%
The chart on gives you a clue as to the bimodal profile of immigrants.
Now look at the following table showing the distribution of employed persons by educational attainment. Hispanics and Asians account for 72.4% of employed persons. Combined, they present a strikingly bi-modal profile compared to white non-Hispanic native-born workers.
…………………………………………………< HS HS < BA BA plus
Hispanic foreign-born 42% 29% 15% 14%
Asian foreign-born 7% 18% 15% 61%
Hispanic + Asian foreign-born 30% 25% 15% 30%
White non-Hispanic native-born 3% 26% 29% 42%
The bipolar profile reflects the estimates that foreign workers comprise 75% of hired farm labor and 32% of computer programmers.
Now, multiple generations of immigrants have been carefully studied. Among Hispanics, about 36% are first generation and 34% are second generation, and 30% older than second generation (see Chart 1, here). (I cannot find data on multi-generational Asians.) This means that total “Hispanic native born” is about 55% second generation and 45% third or older generation. The table below shows that educational attainment of native-born Hispanics is much improved. So is it the case with native-born Asians. The bi-modal profile of the groups combined disappears, but educational status combined remains well below that of native-born whites.
…………………………………………………< HS HS < BA BA plus
Hispanic native-born 10% 31% 33% 25%
Asian native-born 3% 13% 21% 62%
Hispanic and Asian native-born 9% 28% 31% 31%
The foreign worker demographic has both grown tremendously and shifted internally in the past 30 to 40 years. A close look might show that the bipolar profile has increased, especially with the rapid growth of Asian immigrants. An explanation of the decline of the middle class can’t be sufficient without taking this demographic into account.