Battle of the Professors!

Location Date: 
May 7, 2013

Without a doubt everyone in the investment business has a copy of the 2009 best seller “This Time is Different:  Eight Centuries of Financial Folly.”  In 2010, the authors, Harvard professors Ken Rogoff and Carmen Reinhart (“RR”), published a study entitled “Growth in a Time of Debt”1. It was this study however that proved to be a major influence on public policy in the US and Europe. 

The conclusion of the paper was rather unambiguous in that GDP growth slowed as the level of debt rose.  However, if you took the mean, instead of the median, you actually had a negative GDP result when debt to GDP levels reached 90%. RR was careful to point out that the median was a much better statistic and a more conservative interpretation of the data. Unfortunately, many on the conservative right side (both political and media) used the study to promote their views that austerity was crucial, and took the outrageous position that 90% was a kind of Maginot line or fiscal cliff level not to be crossed or the national economy would collapse – pure sensationalism that was never espoused by RR.

The summary below shows how RR’s study is presented across different sample periods and analysis to demonstrate the correlation between growth and debt levels.

Source: Bloomberg


The paper, which was not peer reviewed,came under severe scrutiny in April when a UMass grad student Thomas Herndon and two UMass professors, Ash and Pollin, (collectively “HAP”) released a paperthat found a coding error in the excel spreadsheet provided to them by RR.  Instead of averaging across 20 countries they only used 15; after adjusting the data the result was virtually the same.  HAP also stated that data from Canada,Australia and New Zealand (all have had relatively high debt and high growth experiences) should have been included. Fair point, but the data was not available at the time of the study.  After including this new data,the growth rate improves by about 1% as shown in the table above.  Finally, HAP thought that the methodological approach was incorrect - but we can leave that issue for now.

The uproar caused by the HAP paper was embraced by those on the left (both political and media) as a refutation of austerity policies around the world. Again, this too is a nonsensical position. It should be noted though that the left does have better comedic prowess - if you have not seen the Colbert Report's take on the issue you have to watch this video (click here to view now) - extremely funny!

So what have we learned from all of this messy debate?

“There are lies, damn lies and statistics.”3  Since the dawn of statistical analysis, people have used stats to reinforce their claims often with little regard to the appropriateness of their application to an issue.  A corollary is “liars figure and figures lie.”  One has to be on guard and thoroughly review the statistical analysis before accepting anything in social/economic/political sciences as proof positive.  Even then, empirical findings are better thought of as guides and not ‘right’ or ‘wrong.’

Heuristics and biases are everywhere in this debate. As a huge fan of Kahneman’s behavioural economics work, it is not surprising to find many commentators using the mental short cuts; including confirmation bias i.e. using information that conveniently conforms to a predisposed viewpoint.  In the RR case, we have to get rid of the idea that the 90% ratio of debt to GDP is of any material consequence because it is not!  It is far too complicated an issue to be reduced to one single number but that is exactly what people have done.  As Lawrence Summers penned in a recent Op-Ed in the FT;  “Even if a threshold existed, why should it be the same in countries with and without their own currency, with very different financial systems, cultures, degrees of opinion....” 4

Correlation does not mean causation. It is safe to say that there is a negative correlation between GDP growth and the level of debt to GDP. But the more important issue is whether high debt levels cause slower growth or whether slow growth cause higher debt levels? The evidence here can support both sides of the argument. Confusing correlation with causation is a classic behavioural bias.  It should be noted that RR went out of their way to not imply causation. As a side note, I find it interesting that the Maastricht Treaty of 1992 imposed debt to GDP levels of 60% on each European country - either that was fatally flawed from the outset or the number needs to be revised to reflect the new information.

So what should we make of all this debate? 

Clearly the quality of the debate is guttural but then that makes for good media and politics. I do not think it takes a PhD in economics to understand that no government can spend beyond their means forever. Taking the high road, it does seem that programs on both side of the debate should be implemented in an effective and efficient way in order to be pro growth and fiscally responsible. In fact RR wrote on FT Op-Ed on May 1 entitled “Austerity is not the only answer to a debt problem.” 5   On the government spending/pump priming side, programs need to create both employment but also high value infrastructure assets that have long term benefits for economic activity.  The CBO notes that the “multiplier” on the 2009 Obama fiscal program was 1  i.e. it was just a transfer of wealth;governments have to do a better job of issuing ‘good’ debt and not wasting taxpayer contributions on ‘pork’.  RR also believe that certain ‘non transparent’ forms of financial repression will emerge, such as, governments cramming debt into “domestic pension funds, insurance companies and banks.” 6  Sound familiar?  On the structural side, the government needs to put in more programs to encourage investment and innovation.  They critically need to reduce the tax on labour not to mention simplifying the tax code itself.  I am not holding my breath – it would appear that the politicians will continue to rely on the central banks to “kick the can” a bit farther down the road before the important but difficult decisions are made.

Jim McGovern

NBER Working Paper Series, “Growth in aTime of Debt”, Working Paper 15639, Carmen M. Reinhart, Kenneth S. Rogoff, January 2010 www.nber.orp/papers/w15639

2 PERI University of Massachusetts Amherst,“Does High Public Debt Consistently Stiffle Economic Growth? A Critique of Reinhart and Rogoff”, Thomas Herndon, Michael Ash, Robert Pollin, April 15,2013, Working Paper Number 322

Quote attributed to Mark Twain

“The buck does not stop with Reinhart and Rogoff”, Financial Times, Lawrence Summers, May 5, 2013

5“Austerity is not the only answer to a debt problem”, Financial Times op-ed, Carmen Reinhart and Ken Rogoff, May 1, 2013