Editor’s note: this post has been updated to reflect corrections to an error in the database used to calculate personal income and personal consumption expenditures for the 2011-12 time period. Originally, these numbers were flat. However, the correction shows a nominal increase in the range of 3 to 4%. We apologise for the error.
Just to put some meat on the bones around the cyclical downturn turning into recession, the personal income data paints a negative picture absent debt accumulation. Originally, the data were much worse than I anticipated but this was due to an error in the spreadsheets that I was using. Now I can see that the data are more benign and that recession is not imminent as ECRI says, but the data are building in that direction.
What we see in recessions is that personal income leads the cycle and when the growth in disposable personal income(real or nominal) turns down, spending also turns down. We are seeing that now as well. This usually triggers recession.
Looking at all of the US data available (since Jan 1929), what we see is this in terms of cyclical peaks from the pre-World War II era:
- In Jan 1930, year on year growth in personal income was already -3.7% when the first data point became available. So the US was clearly well into recession. There is no corresponding data on disposable personal income or personal consumption expenditures. The numbers troughed in July 1932 at -27.4%, year-on-year, foreshadowing a recovery that came in Mar 1933
- In Mar 1934, we saw a large mid-cycle personal income peak of 24.2% year-on-year. By Jan 1935, this had fallen to 5.0% before resuming again and peaking in Jun 1936 at 27.6%. The data in this time series is pretty choppy so I smooth it out when looking and rely instead on lagging six-month averages. These averages show growth peaking in May 1934 at 17.6%, troughing in May 1935 at 7.6% and then peaking yet again in November 1936. What’s clear here is that personal income growth can decline significantly and for up to a year without triggering recession. Recession came in Jun 1937.
- In Sep 1938, average personal income growth troughed. That was two months after recovery began. And it rose until Feb 1940, stuttered a bit and exploded when the US went into war preparations in 1941. Eventually, recession came in Mar 1945, preceded a full two years by a 31.6% peak in average personal income growth.
Rather than go on, let me give you a few examples from the end of the data series:
- In Sep 1973, a month before the Yom Kippur War, average personal income growth topped at 12.5%. Average disposal personal income growth topped in August and September 1973 at 13.4%. Recession was on the US by December. The interesting bit here was that the peak in personal consumption expenditures was in Mar-Jun 1973 at 11.3%. Consumer price inflation might have been a factor in restraining consumption expenditures prematurely.
- In Sep 1978, more than a year before recession began in Feb 1980, average personal income growth topped out at 12.8%. Average disposable personal income growth topped out in July at 12.5%. consumption expenditures topped in Jan 1979 at 12.3%. For me, this is a more classical pattern: income to retail sales to production and employment.
- In Jun 1989, just over a year before recession began, average personal income growth topped out at 8.4%. Because of tax hikes by Bush, average disposable personal income growth peaked even sooner, in Sep 1988. So (six-month trailing) average personal consumption expenditures followed the disposable income and peaked at 8.4% in Feb 1989, before personal income peaked. The key then is disposable income rather than total personal income. The 1990-91 recession proves it.
- In Sep 2000, average disposable personal income growth peaked. Recession followed by April 2001. Here average personal consumption expenditures peaked in growth in the Apr-Jul 2000 time frame at 8.1%. It’s hard to say why personal consumption expenditures’ peak preceded income growth but it could be the wealth effect at play.
- And finally, in the last recession, we saw a DPI peak in Jun-Aug 2006 at 7.1%. Personal consumption expenditures peaked at about the same time, a little earlier at 6.1% in may-Jun 2006
My conclusion from this is that personal income and personal consumption expenditures lead the cycle. This is retail sales is considered a leading forward indicator because it leads the cycle up or down. It goes from income to sales to production and employment. That’s the cycle order.
Here in this cycle, year-on-year six-month trailing average personal income growth peaked in Mar-Apr 2011 at 5.6%. Meanwhile the same smoothed second derivative number for DPI peaked in Dec 2010-Mar 2011 at 4.7%. That tells you two things. First, government stimulus has been a net detraction to personal income for over a year now. Second, the US economy has grown for the last year more robustly because of debt accumulation more than because of sustained personal income growth.
In fact, disposable personal income is now running at 2.6% in nominal terms, which is nearly flat in real terms. That will spell recession soon.
As Lakshman Achuthan of ECRI was saying last year, I don’t see how you get these kinds of numbers without eventually triggering a recession. Government likely does not have the ability to turn this cycle after it has turned down for so long. Yes, they did so via public works in 1934 and 1935, but there is nothing like that happening now. All signs indicate that the US is slipping into recession. Any negative economic shock like the looming US fiscal cliff will provide the push to make this cycle down obvious, just as 9/11 did and the Lehman crisis. But in each of these cases, the recession had already begun.