Only on a day like today does an over 1 percent decrease in industrial output move to third page. But this item (and this hilarious article h/t Economists View) reminded me to update the indicators used by the NBER BCDC are headed. Their trajectories are, in general, not too comforting.
Consider first the series that prompted the investigation. In Figure 1, I show (in logs) industrial production and manufacturing production; the former is the one considered by the NBER BCDC.
Figure 1: Log industrial production (blue), and log manufacturing production (NAICS definition), 2002=100. Line at industrial production peak; NBER recession dates shaded gray. Source: Federal Reserve Board via St. Louis Fed FRED II, accessed 15 September 2008.
The peak was in January 2008 for industrial production. Interestingly, the peak for manufacturing was earlier, in September 2007. I found this remarkable given the stress that has been made about strong export growth supporting US aggregate demand. I would have thought there would be a much less marked decline in manufacturing output. But apparently this is not the case.
Next consider nonfarm payroll employment. I’ve discussed the various aspects of the August employment situation release already , but it’s useful to recap what the peak in this series was: December 2007.
Figure 2: Log nonfarm payroll employment (blue). Dashed line indicates peak; NBER recession dates shaded gray. Source: Federal Reserve Board via St. Louis Fed FRED II, accessed 7 September 2008.
Now consider real personal income less transfers and real manufacturing and trade sales, the last two key series focused on by the NBER BCDC.
Figure 3: Log personal income less transfers in Ch.2000$ (blue). Real personal income calculated by subtracting off transfers from personal income, and deflating by the personal consumption expenditure deflator. Dashed line at peak; NBER recession dates shaded gray. Source: BEA GDP release of 29 August 2008, and Supplemental Table 2BU, and St. Louis Fed FRED II, accessed 7 September 2008, and author’s calculations.
Figure 4: Log manufacturing and trade sales in Ch.2000$ (blue). Dashed line at peak; NBER recession dates shaded gray. Source: BEA GDP release of 29 May, and Supplemental Table 2BU.
Both of the series peaked in October 2007, although it is conceivalble that manufacturing and trade sales have reversed course (the last observation is for June).
Now, not all the news is bad for the times-are-fine thesis. The month-old GDP index compiled by Macroeconomic Advisers indicates that June 2008 GDP was substantially up.
Figure 5: Monthly log GDP (Ch.2000$). NBER defined recession dates shaded gray. Source: Macroeconomic Advisers [xls], August 15, 2008 release.
Of course, June is in 2008Q2, and we’re now puzzling over 2008Q3. We’ll just have to wait for Macroeconomic Advisers’ newest estimate (which should come out any day now). e-forecasting‘s estimate for July and August month-on-month annualized growth averages out to 1.65%. The one point that I would observe is that just as official GDP is revised as the source data is revised, so too is the Macroeconomic Advisers series (and presumably the e-forecasting one as well). From the February Macroeconomic Advisers’ report on the monthly GDP series:
…As currently estimated, Monthly GDP in December 2007 was at an all-time high, suggesting that it’s unlikely that the economy slipped into recession at the end of last year. However, revisions to the underlying source data could overturn this observation.
I’ll let the reader decide what to make of these trends. For me, they merely confirm what Jim wrote a week and a half ago.
And if you remain skeptical that we are in, or about to enter, a recession, then three things to consider:
- The current financial crisis is in some sense the reflection of the failure of certain institutions to deleverage, or recapitalize, fast enough. This process is going to further reduce the availability of credit, thereby slowing down the economic activity further — a process laid out in this post.
- The decrease in the equity markets today, if persistent, will further constrain consumption, to the extent that wealth is a argument in the aggregate consumption function (4 cents on the dollar is the conventional wisdom regarding the MPC out of financial wealth).
- The losses to equity markets — and jamming up of credit markets — appears to be a trans-Atlantic phenomenon. Hence, the likelihood that rest-of-world economic activity can continue to sustain US aggregate demand seems ever smaller.
See also Brad Delong.
[Update, 9/16 9:12am Pacific]: Macroeconomic Advisers has released its GDP series, up to July. H/t Peter Summers.
Figure 6: Monthly log GDP (Ch.2000$), release of 16 Sep (blue) and of 15 Aug (red); MA forecast (teal +). NBER defined recession dates shaded gray. Source: Macroeconomic Advisers [xls].
[Update, 9/16 12:55pm Pacific]: e-forecasting’s Maria Simos has kindly provided me the data for e-forecasting’s monthly GDP series, extending up to August. It is plotted with the Macroeconomic Advisers’ series in Figure 7 below. The August figure was revised downward to -0.3%. This means the average growth rate over July and August has been revised down from 1.65% at annual rates to 0.05%.
Figure 7: Monthly log GDP (Ch.2000$), Macroeconomic Advisers release of 16 Sep (blue), MA forecast (teal +), and e-forecasting release of 16 Sep (maroon). Source: Macroeconomic Advisers [xls], and e-forecasting release of 16 Sep.
Interestingly, on 9/12, I noted the CBO estimated a -3% output gap as of 2009Q2, and my own quadratic trend estimate of potential, combined with WSJ forecast, implied a -4% output gap. Latest estimate (2018Q1 3rd release, CBO August 2018 estimate of potential) is a -6.8% negative output gap.
Thoughts by Jim on the original Paulson plan, here.