Debt and Economic Growth

Yesterday Bob Marcin posted a thought-provoking chart and invited comment. I agreed to respond, but to provide some charts for comparison it is more convenient to do it in a post.

The chart purports to show that economic growth is the result of debt, and that over the last few decades it has required more and more debt to get a modest amount of growth.

My objections to this are pretty much textbook econonmics stuff. Bob asked if the approach was valid or perhaps too simplistic.

1. The creator of the chart has not demonstrated a real relationship. Normally you would do that by showing a tight correlation. Instead he shows a very sloppy correlation and asserts that the deviation is the result of an ever-growing need for debt. Basically, he is beginning by assuming the accuracy of his conclusion instead of looking for evidence.
2. To the extent that there is a relationship between GDP and debt, the causation probably runs in the other direction. While causal models are challenging to prove, it is likely that more borrowing occurs when times are good.
3. As Bob hinted, the model is indeed overly-simplified. The leading macro economic model has 600 variables. That is for many different purposes, but even a GDP model would have 20 or so. If we tried to do the job in a few variables, the first two we would choose would be unemployment and inflation.

Charts

To take an objective look at household debt, you should compare not to GDP but to household income. Let’s take a look at that.

As we can see, the correlation here is pretty tight. You can see that debt got ahead of income in the mid-80′s and during the last decade, but has now reversed.

It would also be interesting to compare household debt to assets. After all, if you are using debt to purchase an asset that is different from using it for consumption.

This type of chart is better on a log scale, since it permits more accurate comparison of time periods. Those who are trying to dramatize the growth of debt never do that, so I simply compared the two series.

As you can see, assets were reduced during the recession, but have now rebounded.

Problems

The biggest problem with this type of aggregate analysis is that the asset rebound is not distributed evenly through the population. The people who are unemployed or under-employed, or those whose home is their main asset, are still lagging.

These groups are not going to be helped by belt tightening or some giant reset. They need the opportunity for jobs and a rebound in the housing market. This is why both parties will (eventually) engage in policies that will try to stimulate the economy now, even though the need for long-term deficit solutions is obvious to all. Estimates suggest that the fiscal cliff could cause a decline in GDP of 4-5%. For those who think this is simply a Democrat viewpoint, you need to read more widely. Bob McTeer, the former Dallas Fed President and now at a conservative think tank, regularly writes in the terms I describe here. I see this as an issue of economic policy, not politics.

To summarize: Both here and in Europe, reducing debt is not something that can be solved by austerity alone.

And finally, I appreciate the opportunity for this discussion. WSAS would be a less interesting site if we all agreed about everything!