A Short History of American Recessions, Depressions, and Panics - Part 3 (Updated 6-12-2020)
10/19/18 - I finally, after many years of procrastination, sent the manuscript of this series of hubs out for publishing as a book.
1/11/19 - I am PUBLISHED!!!! :-)
WHY WE DON'T WANT RECESSIONS
When Is A Recession Really Over
LIKE MOST THINGS IN LIFE ... IT DEPENDS. It depends on a lot of things, not the least of which is "what kind of recession is it?" You see, the dynamics of what starts a recession determines its longevity; which may or may not coincide with the "official" duration from the National Bureau of Economic Research (NBER).
Another complexity is "how do you measure recovery?" When do you actually know when the effect of downturn is truly over? There are, of course, many different measures where different ones are more appropriate than others at different times. For example, the NBER generally uses the switch from a declining GDP to one that is growing as their endpoint. But, is that necessarily a good measure for this recession or that recession? Again, it depends on the recession.
The Great Depression officially ended in March 1933, but who knew? I could argue it didn't end until WW II and be able to support it with statistics such employment which never did return to normal or that it took over 16 years for the number of banks to increase to pre-depression levels. There are other statistics as well such as the Wholesale Price Index, the Retail Index, the Cost of Living (now CPI) Index, GDP per capita, among others. Another example using these metrics was the 1865 depression. In this case, the depressions duration was 3 years, yet it took 7 years for the per capita GDP to return to pre-depression levels, 50 years for the Wholesale Price Index to recover, and 48 years for the Cost of Living Index to "bounce" back. So, take your pick.
The point is, of course, to make an informed assessment and not just a bumper sticker or throw-away line, you must consider multiple metrics of improvement, not just GDP. So. in Table 1 you will find several of these metrics for each recession. Note that not all recessions have all metrics because some metrics aren't available for the whole period.
MAJOR RECESSION RECOVERY PERIODS
CONTINUOUS DOWNTURN
| DEPTH (largest GPD/Capita Decline
| OFFICIAL DURATION (Yrs)
| RECOVERY - GDP*
| RECOVERY - WPI
| RECOVERY - COLI (CPI)
| Plus RECOVERY - RI
| Plus RECOVERY - # BANKS
|
---|---|---|---|---|---|---|---|
1796 - 1799
| 1.0%
| 3
| 2
| 17
| |||
1802 - 1804
| 1.4%
| 2
| 3
| 11
| 10
| ||
1807 - 1809
| 6%
| 3
| 3
| 2
| 1
| ||
1815 - 1821
| 3.8%
| 6
| 7
| 49
| 132
| ||
1828 - 1829
| 1.7%
| 1
| 3
| 24
| 9
| ||
1833 - 1842
| 3.8%
| 10
| 11
| 30
| 36
| ||
1853 - 1854
| 0.2%
| 2
| 2
| 1
| 1
| 1
| |
1857 - 1858
| 2.2%
| 2
| 3
| 6
| 5
| ||
1865 - 1867
| 7.6%
| 3
| 7
| 53
| 55
| 3
| 5
|
1873 - 1877
| 2.5%
| 5
| 4
| 44
| 53
| 7+
| 8
|
1882 - 1885
| 6.1%
| 3
| 5
| 34
| 43
| 1
| |
1893 - 1894
| 6.7%
| 2
| 2
| 7
| 15
| 8
| 1
|
1907
| 12.5%
| 1
| 9
| 2
| 5
| 1
| 1
|
1913 - 1914
| 9.4%
| 2
| 3
| 3
| 1
| ||
1920 - 1921
| 4.2%
| 2
| 3
| 29
| 27
| 1
| 23+
|
1929 - 1939
| 28.6%
| 4
| 10
| 13
| 14
| 16+
| |
1937 - 1938
| 4.4%
| 2
| 2
| 4
| 3
| ||
1960 - 1961
| 0.9%
| 2
| 3
| 5
| 1
| ||
1973 - 1975
| 2.6%
| 3
| 3
| 1
| 1
| ||
1979 - 1982
| 2.7%
| 4
| 4
| 1
| 1
| ||
2008 - 2009
| 3.6%
| 2
| 4
| 3
| 2
|
TABLE 1 - TIME IS IN MONTHS AND THE RECESSIONS THAT ARE BOLDED ARE FINANCIALLY-BASED, i.e. LACK OF FINANCIAL REGULATIONS, EASY CREDIT, SPECULATION, ASSET BUBBLES, THEN A TRIGGER.
* - GDP is actually GDP per Capita. Because population was growing so rapidly for the first 120 years of America's history, it distorts raw GDP, even when expressed in constant dollars. Dividing by population has the effect of normalizing GDP for population growth.
WPI is the Wholesale Price Index. The Wholesale price index refers to a mix of agricultural and industrial goods at various stages of production and distribution, including import duties. The Laspeyres formula is generally used. The index is calibrated to a given year, say 2005 = 100.
COLI, or Cost of Living Index, is a broad, theoretical measure of the effects of inflation on standard of living. The version used is a somewhat narrower facsimile, the Consumer Price Index - U (CPI-U), which measures price changes for all urban consumers. CPI-Us prior to 1913 are estimates made by the Federal Reserve Bank of Minneapolis.
RI, or Retail Index, is similar to the WPI in that it measures price changes at the retail level. Currently, the data availability is spotty.
# of Banks - One of the characteristics of strong economic downturns is the collapse of banks, especially before FDIC in the 1930s. A measure of the recessions severity is the percentage loss in banks coupled with a return to the previous level of banks.
Significance
ONE QUESTION THAT COMES TO MIND is whether there is a significant difference between the average durations of recessions based on financial reasons, e,g. 2008; or based on some other reason like the end of a war. To that end, let's look at a couple of averages from Table 1.
Financial Vs non-Financial Recessions
OFFICIAL DURATION
| PER CAPITA GDP DURATION
| WHOLESALE INDEX DURATION
| COST OF LIVING INDEX (CPI-U)
| |
---|---|---|---|---|
FINANCIALLY-BASED
| 3.7 years
| 5.5 years
| 13.5 years
| 10.9 years
|
NON-FINANCIALLY-BASED
| 2.5 years
| 3.2 years
| 19.1 years
| 30.6 years
|
Not Different at Alpha = 0.05
| Different at Alpha = 0.05
| Not Different at Alpha = 0.05
| Not Different at Alpha = 0.05
| |
Different at Alpha = 0.1
| Not Different at Alpha = 0.1
| Not Different at Alpha = 0.1
|
TABLE 2
The data set used consists of 10 financially-based recessions and 11 other type recessions; not particularly large. The hypothesis we are testing is whether there is a difference in means between the two sets of data. Using the two-tailed t-test, we see that, at the most commonly used level of significance of 5%, there is not a statistically significant difference between the two means when considering the "official" duration, the recovery time for the Wholesale Price Index, or the time it takes the CPI-U to begin to increase again. There is, however, a difference for "official" durations if we loosen our standard a bit to Alpha = 10%; but not for the WPI (the reason is the huge variances in durations).
There is, nevertheless, a significant difference at the 5% level between the mean recovery times (to reach pre-recession levels) for the metric, per Capita GDP.
The reason the WPI and COLI do not show up as significant (they might not anyway) is that the population is small and the variation of individual data points around the mean is very large. As such, I would have been very surprised if the null hypothesis had been disproved.
PERIOD
| TOTAL YEARS
| TOTAL RECESSIONS*
| RATE of TOTAL RECESSIONS
| FINANCIALLY-BASED RECESSIONS**
| RATE of FINANCIAL RECESSIONS
| Avg GDP Length of "OTHER" Recession
| Avg GDP Length of FINANCIAL Recession
| |
---|---|---|---|---|---|---|---|---|
CLASSICAL ECONOMICS
| 1796 - 1933, 2000 - 2009
| 146
| 37
| 1 Every 4 Years
| 11
| 1 Every 13.7 Years***
| 3.2 Years
| 6 Years
|
KEYNESIAN-BASED ECONOMICS
| 1933 - 2000
| 69
| 12
| 1 Every 5.8 Years
| 0
| Haven't Had One
| .3 Years
| 4 Years
|
Total Recessions Based on NBER
| 49
| 11
|
TABLE 3
* - Includes all recessions listed by the NBER, not just the ones analyzed previously
** - Based on the descriptions offered at http://en.wikipedia.org/wiki/List_of_recessions_in_the_United_States
*** - Excluding the Great 2008 Recession
It is a given that those who hold by Classical economic theory and the laissez-faire policy that is central to it believe very strongly it has never failed them. In fact, Table 3 should make it painfully obvious that Classical economics has always failed in the mid- to long-term. During the 146 years during which Classical economics was America's economic policy (save for maybe a totally of 10 years when the Whigs were in power and during the Teddy Roosevelt - Taft presidencies), serious, financially-based recessions, panics, or depressions were the norm at more than one per generation. Further, it should be clear, even though the Right denies this as well, there is a significant difference (at a 95% confidence level) between the lengths of financial recessions vs all others.
From the above, you can distill several things:
- Financially-based recessions are significantly longer than any other kind.
- Financially-based recessions are linked to Classical economics but not Keynesian (given there are no instances during the Keynesian period)
- Keynesian economics appears to lead to a much more stable economy over all.
This content is accurate and true to the best of the author’s knowledge and is not meant to substitute for formal and individualized advice from a qualified professional.
© 2015 Scott Belford