Debt and economic growth

What is the impact of higher debt on economic growth? This is not an easy to answer question. Some analysts will have you believe that debt, and here I am referring to government debt, is either bad for growth or good for growth. These thoughts about debt can be summarised under mainly three ideologies:

·         Barro-Ricardian: Indifferent between taxes and government debt; households only care about  present discounted value taxes

·         Keynesian: Debt can be welfare enhancing increasing both current and future consumption

·         Neoclassical: Debt reduction has positive effects on investment through lower interest rates

My personal view is that debt on growth incorporates all three views with some views being stronger under certain conditions. Economists often use a solvency rule for analysing whether government debt is sustainable. This solvency rule inherently assumes that debt cannot be rolled over and that there is a point at which any change in government finance decisions cannot curb the growth in debt. In other words you will perpetually borrow more funds to finance maturing debt or you will simply have to default.

There are good reasons why government debt can be bad. Funds used for financing higher debt service costs could have been used to do something more productive. Higher debt, if perceived unsustainable, can lead to massive capital outflows - thus a reduction in foreign investment. Often higher interest rates are required to attract additional financing which increases debt service costs even more.

But, debt can also be good. Especially if government borrowed funds to be more productive - i.e. there is a future return for the borrowed funds.

We can then see that debt has a direct impact on growth only in the way expectations are formed. Debt itself  does not necessarily add or subtract to growth since debt is only an accumulation of public finances (public finances directly affect growth). These expectations create investor fears and confidence, which leads investors to either to purchase government bonds or to sell them or not to purchase new bonds.

Simple figures or simple statistical tests often inform people's views on debt. Take as an example a simple scatter plot of South African debt and economic growth since 1980 - It shows that the level of debt is associated with positive economic growth (if you plot a trend line through the bubbles). On average it would seem that higher debt is associated with increasing GDP growth. But using such simple figures would be a very misleading representation of the effects of debt on growth. Some authors (just read about the recent Reinhart and Rogoff controversy) suggest that such figures, accounting for many cross countries and a long enough history, show us meaningful results. It must be added that these attacks ignore a lot of the other good research (and more serious research for that matter) that Reinhart and Rogoff have done. The point that I am trying to make is that we cannot simply infer relationships between any two variables without knowing whether other common factors influence debt and GDP while also controlling for possible feedback loops between debt and GDP -  debt, interest rates and economic growth are endogenous - that means that they influence each other: Higher debt accumulated through government investment programs can increase economic growth while higher economic growth makes it easier to finance debt. Higher debt levels are associated with higher debt service costs which are linked to a risk premium (given that debt increases), but at the same time higher interest rates increases debt (this is if monetary policy has interest rates rising). Thus one needs to properly account for the future path of all these variables in a simultaneous setup to determine whether debt is truly sustainable. This is exactly why looking at the aggregate debt level is wrong - one needs to analyse the composition of debt and all its contingent liabilities (as an example if Eskom defaults then government guarantees will finance the default).

Figure 1: Government debt and economic growth

Or you can use more fancy statistical techniques that tell you at which point debt will have a negative impact on growth - smells like a Laffer curve:

Figure 2: Simple growth regression with debt and debt square

If one were to believe the results of Figure 2 it would suggest that a debt to GDP ratio higher than 42% would lead to less economic growth.

Unfortunately the past is not always a good predictor of the future (and these figures and econometric tests all depend on data collected from the past). If that were the case then we would be able to predict recessions! So using simple Figures and statistics are not that useful in determining debt sustainable.

While debt sustainability is certainly part of solvency calculations, it is the role of expectations that are important: Let's say you want to invest your hard earned cash in government bonds - most likely you will look for a good  and safe return (remember you are a risk averse investor) with a degree of certainty that you will get your principal and the interest back when the bond reaches some date (bond maturity). Your views on good returns are informed by the monetary policy interest rate and a certain risk premium (the risk premium is some kind of incentive that the government pays you for purchasing a bond - this is linked to the probability of default). The higher the probability that government will default on its debt obligations, the higher the risk premium (here measured as simply the yield on government bonds). And here is the tricky part - what actually informs the probability of default? 

Rating agencies use various probability models to tell us how safe our debt is...but just what makes one country's debt more sustainable than another? Why can Japan have debt well over 100% to GDP, have very small GDP growth numbers while other emerging market countries run a risk when debt exceeds 60% of GDP. Are the Japanese better at paying back the principle and interest rate than South Africa? Both countries finance the majority of their debt domestically and thus avoids currency risks and sudden stops in capital flows. Both countries print their own money, but South Africa has higher inflation which decreases the real value of debt over time while Japan has long periods of negative inflation which increases the real value of debt over time. Once again it would seem like expectations is the main character  in making debt sustainable or unsustainable. 

While this blog entry does not advocate that any country ramp up debt, it does question the views of the market on debt sustainability. It is obvious that if governments ramp up debt with no expected return tax payers will finance the debt and receive nothing in return...when debt runs away at crazy interest rates that are required to finance debt service costs then people might have to make crazy adjustments by paying higher tax rates up until the point where the country is forced to default. But at which point is debt unsustainable? I don't think rating agencies have this right either. Like any good financial investor, a future bond holder should analyse the reasons for why debt is accumulated - ff it is simply to finance transfers or higher government wages then there is a good possibility that government won't make a return. Unless higher wages translate into higher savings which are used for investment purposes.

In essence, the issue regarding the effect of debt on growth seems to be somewhat stupid since we cannot truly measure expectations - we just see it after the fact (higher credit spreads, capital outflows etc.). And we cannot understand expectations completely because they are even different for countries with similar economic conditions. Perhaps the closest we can get is that people follow sometimes silly rating agencies - and these guys might influence expectations. But you can already see some kind of decoupling of expectations to ratings: after the recent SA bond downgrade yields and credit default spreads hardly increased.

And SA has actually done a good job - a majority of debt has been accumulated to finance investment projects. It is up to the government and us to tell us what the expected returns of these projects are. But for now I would not worry about SA debt being unsustainable or whether it is hurting economic growth ;0)