This blogger isn’t a fan of the oft-made analogy between the US economy now and the Japanese economy of the early 1990s — an analogy normally used to analyse the possibility of the US facing a similar Lost Decade.
There are sufficient differences between the two cases — especially those related to Japan’s decades-long export-led growth model — that we suspect the Japanese lost decade had as much to do with the country’s unique circumstances as with factors shared in common with the US now.
But, well, that obviously doesn’t mean the US can’t suffer a decade of stagnancy for its own reasons.
Justin Wolfers, drawing on the work of economist Jeremy Nailewaik (pdf), says we’re already halfway there.
But Wolfers looks at a different version of GDP, one that relies on the sum of all income rather than spending. Nailewaik’s paper refers to this income-based measure as GDP(I), or just GDI (Gross Domestic Income).
But this measure actually peaked in the fourth quarter of 2006, roughly a year before GDP started falling. And it remains, both in aggregate terms and per capita, below that 2006 level even now:
Why use the income- rather than spending-based measure?
As Wolfers writes, in theory the two are supposed to be equal, each measuring the same thing — the nation’s output — using equally valid conceptual approaches. But they use different sources of data, and each has problems specific to its methodology.
We won’t pretend to understand the nuances of the two measurements, though Nailewaik explains some of the difficulties in arriving at both in his paper.
The important issue is to know which is a better, more meaningful reflection of the economy. Nailewaik explains why he sides with GDI near the end of his paper:
Considerable evidence suggests that the growth rate of GDP(I) better captures the business cycle fluctuations in true output growth than does the growth rate of GDP(E). For the initial growth rate estimates, the revisions evidence over the past 15 years, the correlations with other business cycle indicators, and the recent behavior of the estimates around cyclical turning points all point to this conclusion. For the latest estimates that have passed through their cycle of revisions, careful consideration of the nature of the source data, statistical analysis of the information added by the revisions, and statistical tests, as well as informal comparisons with other business cycle indicators, again all suggest that GDP(I) growth is better than GDP(E) growth at tracking fluctuations in true output growth.
Relatedly, one other intriguing finding from the paper is that GDI also shows a more pronounced, less smooth business cycle going back a few decades than is captured by GDP.
Perhaps the Great Moderation was a bit less moderate than we thought.
We’re halfway to a lost decade – Freakonomics