April 10, 2016

On Econometrics and Forecasting

After a series of articles on Keynesian economics, libertarian philosophy and the role of  government in our society, it is time to shift gear. There are so many important issues to talk about in modern economics, political economy and the practice of economic theory. One issue that often pops up is the role of econometrics in economics. Here is an article I wrote in 2014 on a now-defunct blog. Enjoy!
The Nobel Memorial Prize in Economics was awarded for the first time in 1969. The inaugural laureates were Ragnar Frisch and Jan Tinbergen, two pioneers in econometrics. The motivation was that econometrics provides economists with excellent forecasting tools.
Over the decades econometrics has evolved with several more econometricians earning the Nobel Memorial Prize. Today, econometrics is so highly regarded that even reputable Congressmen use it as an argument for their policy agenda. The latest example is Paul Ryan (R-WI) who, according to the Washington Examiner, recently expressed his appreciation for econometrics as an exact science:
econometrics has evolved to a place where with conservative econometric modeling we can get much much more closer to the truth and reality based upon experience
Congressman Ryan was making a point about the effects of tax cuts and the importance of including supply-side effects on economic behavior and, in the end, the federal budget. If supply-side effects are included, Ryan argued, the macroeconomic case for tax cuts looks much better as tax revenue will increase when lower taxes stimulate growth.
The Congressman is correct in that tax cuts are good for economic growth. There is ample evidence in the economics literature to prove this, based on econometrics as well as other methodologies. However, Paul Ryan should be very careful in relying too much on econometricians to motivate tax cuts. 
Let us return to Jan Tinbergen. Back in the 1930s when Tinbergen was rising to stardom in the economics profession, he attended a conference at King’s College in Cambridge, UK, to talk about econometrics. During his lecture he mentioned that he had estimated the increase in British imports when consumption increased, and found that his number was the same as British economist John Maynard Keynes had estimated years earlier without using econometrics.
Keynes, one of the attendees, then complimented Tinbergen, with the politely insulting comment that “I am glad you found the right number”. Keynes, like other heterodox economists such as Friedrich Hayek, was no fan of econometrics. When another econometrician pointed out that econometrics allows your forecasts to be exactly right, Keynes responded that such forecasting is a two way street, famously finishing his argument with “it is better to be approximately right than exactly wrong.”
These are wise words from a wise man. Throughout the decades, countless econometricians have been proven exactly wrong. In an evaluation for the National Tax Journal of the Reagan era tax cuts, Daniel Suits and Ronald Fisher explain:[i]
In 1980, supply-siders predicted that their tax reduction program would balance the federal budget by the year 1984. The outcome was a record deficit, but the forecast had been embraced by enough people to enact the proposal, and none of the skeptics could have proven, in any objective sense, that the projection was in error at the time it was made.
In 2010 British economist Paul Ormerod explained that the forecasting problem is much more general than that:[ii]
[The] track record of economic forecasting of the overall economy, even just 1 year ahead, is poor. … In early 2008, very few economists predicted what was about to happen, and almost all published forecasts envisaged positive economic growth during 2009. … The track record of predicting recessions, when the risks to companies from overall economic environment are at their highest, is exceptionally bad.
An econometric forecast in 2007 that the U.S. economy would continue to grow along its then-current path would have miscalculated the budget deficit by up to 289 percent: a continuation of the budget trend from 2004-2007 would have pointed to a deficit of $118-125bn for 2008.
The deficit turned out to be $458.6bn.
Congressman Ryan does a respectable job in promoting growth-stimulating tax cuts. His advocacy of econometrics, or “dynamic scoring” as it is called in Congressional parlance, is an honest effort at leveling the playing field in budgetary forecasting. The problem for Ryan is that once he has been proven exactly wrong, his opponents will use his exactly wrong calculations to dismiss future tax cuts.
Paul Ryan need not get into the muddy waters of econometrics to prove that tax cuts work. The literature of economics and public policy is filled to the brim with proof that lower taxes lead to more growth, lower unemployment and higher levels of prosperity for the general public.
I fear though that his fascination with dynamic scoring and econometrics is driven by a desire to show that tax cuts are good for the federal budget, more than anything else. If this is the case, we need to take the debate over tax cuts, economic growth and supply-side economics to an entirely new level.

[i] Suits, D B and Fisher, R C: A Balanced Budget Constitutional Amendment: Economic Complexities and Uncertainties; National Tax Journal, Vol. 38, No. 4 (December, 1985) pp. 467-477.
[ii] Ormerod, P: Risk, Recessions and the Resilience of the Capitalist Economies; Risk Management, Vol. 12, No. 1 (February, 2010) pp. 83-99.

No comments:

Post a Comment