Wednesday, April 17, 2013

Growth in a Time of Debt

A 2010 NBER working paper by Reinhart and Rogoff (also published in AER P&P) claimed that countries grow slower when they have high public debt to GDP ratios. Frank Jotzo pointed me to a blog which shows that there seem that the result is heavily influenced by a single year of -7.9% growth in New Zealand when the debt/GDP ratio was above 90% and there are also mistakes in the analysis. These issues are described in this new working paper. Without the New Zealand data point and correcting the mistake, the average growth rate in the 7 countries with debt ratios above 90% is 1.4% and with NZ is 0.3%. Still that is lower than the growth rate at the lower debt ratios and the highest growth rate is at the lowest debt ratio. Still, the negative correlation between debt to GDP ratio and growth rate does look real. But what is needed is probably a fixed effects regression of annual growth rates data on debt ratios or something like that. Also, Figure 2 in Reinhart and Rogoff presents medians as well as means of growth rates, which is more robust way of dealing with this data:

Table 1 in Reinhart and Rogoff shows data from a longer period and NZ now has positive growth at high debt ratios. Here the relationship looks more fragile though still hanging on. They also present similar data for developing countries where the relationship seems to be present too. So, while there are clearly problems with this paper I think the blog linked above is overly negative on the results. Reinhart and Rogoff have also responded to this criticism.

The real question though is about causality. Does high debt cause slow growth or vice versa?

Sunday, April 14, 2013

The End of Economic Growth

A few months ago, there was a debate about whether economic growth was coming to an end. Based on this graph it looks like economic growth ends around 2060:
The downtrend in growth rate is fairly linear, but of course it is just wild speculation to project it forward several decades. Trends can change direction. This is based on Congressional Budget Office estimates of potential GDP growth. Also, this is total GDP, not GDP per capita. It is much less controversial to argue that population growth will come to an end this century.

Tuesday, April 9, 2013

Stochastic Trend Featured in List of Savvy Maths and Stats Websites

You can see the other sites featured by following this link. There are some that I occasionally visit like Andrew Gelman's site and a lot of the others look interesting too.

Thursday, April 4, 2013

22 Year Citation Lag

I finally got a citation by someone other than myself to my first full length journal article, published in late 1990 just as I was starting my PhD! That's certainly an outlier in the citation distribution. The title of my "most failed journal article" now needs to pass to this one, which shows the importance of normalization.