Wednesday, June 8, 2011

Follow-Up on TGS Graphs

. Wednesday, June 8, 2011

Daniel Davies and Alex Tabarrok objected to the graphs of GDP I included in this post on The Great Stagnation. Specifically, they didn't like the fact that the hypothetical lines I drew reflect constant linear growth rather than constant percentage growth. I.e., I didn't compound the growth when i drew those lines. They're right that the latter is a better measure of trend (it's what is used in almost all statistical analyses), so here's a new graph that takes that into consideration.

This graph shows the actual GDP per capita growth (circles) for the US, OECD, and entire world. The lines that begin in 1974 reflect what GDP per capita would look like if it had continued to grow at the 1960-1973 rate*. Note that a "Great Stagnation" hypothesis would expect significantly weaker growth post-1973, not the same amount and certainly not more. So the fact that the US was above the trend line until the early 2000s provides fairly strong evidence that if we're in a Great Stagnation it's more recent than Cowen argues, and doesn't correlate with stagnating median incomes all that well. In fact, the US does better than either the OECD or the globe, if "better" is defined as "closest to 1960-1973 trend", although the OECD trend line is quite a bit steeper**.

Anyway, just wanted to make sure I didn't leave the impression that my main point (about distribution) relies on faulty extrapolation.

*Specifically, I regressed a year counter on the log of GDP per capita (constant dollars, via WDI) from 1960-1973. The coefficient estimate represents the average growth in GDP per capita per year during that period. I then took that coefficient estimate and added it to 1973's GDP per capita to get 1974's predicted point, added the same constant to the predicted 1974 to get the predicted 1975, and so on.

**Of course the US is a big part of both OECD and world economies; if you removed the US from those groups the US would likely look still better in comparison. Although in the OECD's case, they added some countries during the series (e.g. Mexico, Slovakia) with lower per capita GDP than more established industrialized countries.


Follow-Up on TGS Graphs




Add to Technorati Favorites