The relevant passage from his story here on the recent debate about whether we've had, face, or are in a depression makes a real hash of it:
But search for a precise definition of economic depression and you'll be hard pressed to find anything more specific than it's more severe than typical business cycle recessions, tends to cross multiple countries and lasts much longer.
Anecdotal rules of thumb -- cited in The Economist magazine and elsewhere -- center on a peak to trough drop in real gross domestic product of more than 10 percent or recessions lasting more than three years.
On that measure, the 1929-1933 Great Depression in the United States qualifies with a 27 percent loss of GDP and a peak unemployment rate of some 25 percent. The shorter 1937 and 1945 downturns qualify on the GDP measure alone too.
"Hard pressed"? The most useful rule of thumb learned way back in my childhood is not even mentioned: back-to-back years with GDP declines, on the analogy of recessions, which are back-to-back quarters with GDP declines. String out a recession long enough with annual GDP failing to surpass a previous high and you have a depression.
People may have to disagree about such definitions, but not about the data behind the theory.
The GDP decline of the 1929 depression is not correctly represented by the writer. Nominal GDP in 1929 was $103.6 billion, falling to its nadir in 1933 to $56.4 billion, a 45.56 percent drop, not 27 percent as the author states. It took until 1941 to surpass 1929 GDP.
Nor did GDP decline from 1937 to 1938 by more than 10 percent. It declined by 6.3 percent, from $91.9 billion to $86.1 billion. But GDP in 1939 exceeded that achieved in 1937, technically not a depression within a depression because there weren't back-to-back years of GDP decline.
And the GDP decline between 1945 and 1946 was a measly 0.36 percent, falling to $222.2 billion from $223 billion. The $1.9 billion decline between 1948 and 1949 was only 0.71 percent.
Missing from the story are the real 10 percent or greater depressions in the 20th century apart from The Great one: the depression of 1907-1911, when nominal GDP fell by 11.1 percent; and the depression of 1920-1925, when GDP fell almost 17 percent. Prohibition, dontchaknow. The roaring '20s were really a lot shorter than ten years.
If the 2008-2009 depression will compare to anything, it will be to 1937-1938's 6.3 percent decline, or to 1913-1916 when GDP fell 6.6 percent. The problem is the numbers are still fluid. The numbers from the Bureau of Economic Analysis still show a nominal decline in one year only, 2009, of 1.8 percent from 2008, despite reports of larger nominal declines in 2008 from 2007 and in 2009 from 2008 in the neighborhood of 3.8 percent.
If it's pretty clear we've had at most only a very small depression, we're technically out of it in 2010 due to government spending. It's equally clear, however, that current GDP is so anemic in the aftermath that we may well repeat the episode.
