So it turns out that the primary economic research paper cited by conservative politicians as the case for austerity in the face of the recent economic crisis is rife with errors. As is usually the case with macroeconomic news, John Cassidy has the most entertaining and informative summary of this story.
Essentially, two Harvard economists misrepresented the relationship between historical debt loads and economic growth because of three errors.
- They mislabeled a field in Excel so it read the wrong group of columns.
- They didn’t weight their averages, so 1 year of data from country A was treated as equally important as 20 years of data from country B.
- They omitted some years of data for some countries for reasons that aren’t clear to me but that seemed to help their case.
Thoughts that immediately come to mind:
- What is it with economists and Excel? The $6 billion(!) lost by JP Morgan in bad trades last year was due in part to a faulty risk assessment model caused by, you guessed it, a bad formula in Excel.
- Weighting the averages from each country is only a marginally better way to deal with those, what shall we call them, repeated measures. Surely economists know this?
- It appears that ecologists are, quantitatively speaking, leaps and bounds ahead of a field that is essentially the study of numbers.
Okay, okay, we can’t judge an entire field by this one terrible paper, and in reality macroeconomic data are just as messy and complex as ecological data. But lately I’ve been reading quite a bit about the role of mathematics in the ecological studies (thanks to E.O. Wilson, Jeremy Fox, and others) and while I am sometimes frustrated with how much time I spend thinking about analyses and how little time I spend thinking about the natural history, ecology, and conservation of birds, the statistical tools we use have been developed because they help us make better sense of a complex world. Given the dispiriting competition for jobs in my field, I’m starting to wonder if I shouldn’t go back to school and become an economist. JP Morgan can hire me for the low, low fee of $1 billion – I’d be worth it at twice the price.