Saturday, July 26, 2014

Trends in real per capita income

 - by New Deal democrat

In the last several weeks, I have done extensive research on trends in real, inflation adjusted Income per capita over the last 20 years, starting with 1993,  Translated from economics into regular english, what I have been looking at is how well each individual American, and also each individual *working* American fared in terms of income over that period.

Real personal income minus transfer payments (Social Security, etc.) is one of the 4 generally accepted markers for whether the economy is in a recession vs. an expansion.  Calculating per capita real personal income is easy, courtesy of the St. Louis FRED, and here it is (blue), and I've additionally shown personal income divided per capita only by the civilian labor force rather than by the population as a whole (red):

You can easily see that real per capita personal income rose sharply in the tech boom of the 1990s (up 24.2% from 1993 to 2000), stalled briefly during and after the 2001 recession (but never dipping below the 2000 level), and then rose more slowly  through 2007 (up 11.2% since 2000) before falling during the Great Recession, and then rising again to a new high in 2012 (up 12.2% since 2000).

In general, the civilian labor force tracked population growth until 2008, and since then has grown much more slowly than population as a whole, chiefly because of the onslaught of Boomers retiring.

In any event, whether we adjust by population as a whole, or by just the labor force, per capita real personal income has generally grown throughout the last 20 years.

Although it is not nearly as broad a measure as the Bureau of Economic Statistics' "personal income" metric, another way in which per capita income might be looked at is as real per capita adjusted gross income, i.e., the income reported to the IRS on tax forms.

While this is not kept graphically by either the IRS or the St. Louis FRED, I have created the following table showing real adjusted gross income from 1993 through 2011 (the last year available at the IRS site), adjusted per capita both by population, and by the size of the civilian labor force (+those not in the labor force but who want a job now).  The second measure is not perfect, but does make a reasonable approximation of taking into account the wave of Boomer retirements (all figures in $Trillions):

Year Pop adjusted CLF adjusted Year Pop adjusted CLF adjusted
1993 4.8134.3562001 6.008 6.014
1994 4.8644.4862002 5.6825.699
1995 5.0154.726 20035.676 5.680
1996 5.2145.0102004 5.950 5.974
19975.5205.534 2005 6.2316.345
1998 5.8545.721 2006 6.528
1999 6.1236.0482007 6.708 6.685
2000 6.3656.36520086.3226.295
avg 1993-2000 5.4715.258 avg 2001-08 6.138 6.148

2009 5.6285.625

2010 5.8245.887

2011 5.8235.911

avg 2004-11 61266.152

Note that I have averaged each 8 year period of 1993 - 2000, and also 2001 - 2008, and separately 2004-2011.  The bottom line is that real per capita adjusted gross income was higher in both of the 8 year period calculations since 2000, than in the 8 year period ending in 2000.

Still, by virtually any measure, incomes since 2000 have generally been below that year.  That is partly because of demographics, since labor force participation starts to decline slightly after age 55 and more significantly after age 60, as shown in this graph (h/t Doug Short):

Additionally, the year 2000 marked the lowest point in the last 40 years for the unemployment rate, as shown in this graph of the unemployment rate on an annualized basis:

A smaller percentage of the population has been earning wages and salaries since the turn of the Millennium, and so median and average income has suffered.

Several weeks ago, David Cay Johnston wrote a column in which he used 2000 as a benchmark, and cumulated average and aggregate per capita adjusted gross income since then through 2012.  I criticized the column, and he objected to the criticisms.

As I read the column originally, the main point of the article appeared to be that Americans are having a tough time because wages have stagnated, and the discussion of income and the Bush tax cuts in the second part of the article were evidence offered in support. 

I have subsequently been advised, if I understand correctly, that the first part of the article (including the discussion of wages) was only introductory, and the main point of the article was:
1.  Bush promised that his tax cuts would deliver prosperity for all.
2.  Bush promised that the prosperity for all would include rising incomes, even including demographic changes that were well known.
3. Therefore it is appropriate to measure Bush's promise by using the measure of income, and declines in reported income due to demographic changes do not matter, since Bush promised they would not matter, (and possibly agreed that the appropriate yardstick was a comparison with 2000).
4. Looking at income as reported to the IRS, not only was there not prosperity for all, but the average taxpayer LOST income.
5. Therefore, by Bush's own yardstick, his tax cuts were a failure and different policy choices should be made.

If this is all true, then my criticism of the article as to conflating wages with income, and not adjusting for demographics, is moot, as they are not relevant to that yardstick.   Thus, while the writer did not adjust for Boomer retirements, had he done so, the outcome, as indicated above, is the same.  

Viewing the matter not by any unique Bush yardstick, but as a matter of general economic trends, it seems clear upon my reading that most readers did in fact confuse wages and income. In such discussions, a clarifying statement that to the effect that wages have stagnated, but partly due to demographics (retiring Boomers) and partly due to increased unemployment, income as reported by taxpayers has declined, would be easy to include.  I think it is a reasonable contemplation that the confusion evident among the comments I read, was a probable result of the omission of such a clarifying statement. Regardless, if the purpose of the article is as I have described it in 5 points above, then the issue of wages and demographics are simply not applicable in terms of an examination of what is said to be Bush's own yardstick. 

But to summarize in terms of economic data vs. a particular political yardstick:

1. The trend in real per capita personal income has been rising over the last 20 years, including since 2000.
2. The trend in real per capita adjusted gross income reported to the IRS has also risen generally in the last 20 years, but in much more variable fashion, with extended periods between old and new peaks.
3. The year 2000 marked the peak of the 1990s tech boom.  Only as measured by that specific year, or to a much smaller degree, using 1999 as the benchmark, does a subsequent cumulative decline appear.  As a matter of general economic discussion, whether measuring subsequent multi-year periods by reference to that specific year is appropriate or not, I leave to the reader.