Recent news from the Greek debt restructuring is cause to continue thinking hard about potential causal mechanisms between debt and growth.
In my last post reflecting on Reinhart and Rogoff, my conclusion was that we need to take a closer look at case studies. Time series data are too sparse to help establish causality, and this is the thorny question for policy. Even correlations, when they are so sparse, are not as informative as a more in depth understanding of the individual cases.
Today I’m going to look at the Belgian debt experience. After a (not so) sleepy decade hovering around a 50%, the debt to GDP ratio rose dramatically during the 1980–82 recession. By 1984, Belgium’s debt to GDP ratio rose above 90% and its “high debt episode” began.
The 1982 recession was very severe. It was Europe-wide, and the proximate cause was the 1981 oil shock. The 1981 annual report from the National Bank of Belgium linked the struggles faced by the manufacturing to the inability to pass through the oil price shock onto consumers.
As a result, unemployment in Belgium rose to 13%. Government spending as a fraction of the economy went up in response. Much of this was driven by increases in unemployment expenditures†. By the end of the year the government deficit was at a whopping. 16 percent of GDP. Interest expenses exploded and in response Belgium started a long process of fiscal consolidation.
But let’s consider Belgium’s next door neighbor — the Netherlands. The recession also hit the Netherlands hard. Real wages (unit labor costs) in both countries had been rising throughout the 1970’s. The recession only made the competitiveness problem worse and by 1982, unemployment in the Netherlands rose to 12%. And again, government spending rose .
These two countries were similar along other lines as well. Both were “Full Democracies” as determined by the Polity IV index. Both were similarly rich, around 17,000 USD per capita. Both were small, open economies in Northern Europe. After the recession, both countries adopted highly corporatist approaches to economic policy in which the government played a large role in wage setting.
But they did differ in one important respect —(as you might have guessed,) Belgium had a much higher debt to GDP ratio. Below I plot the evolution of debt to GDP ratios and yearly growth rates using the HAP version of the Reinhart and Rogoff dataset. While the debt ratios followed similar curves, debt in Belgium exploded to much higher levels than in the Netherlands. Yet growth in the two countries was almost identical.
By comparing Belgium to the very similar Netherlands, I’m trying to control for a variety of variables that we might think are important for growth, such as wealth, openness to trade, quality of governance, geography, and industrial structure††. If the effect of debt on growth is indeed large, we can then attribute any remaining gap in growth to the very different levels of debt. This is similar to Acemoglu and Robinson’s approach when they try to isolate the effect of good economic institutions by looking at places across a border that developed side by side, except for institutional quirks on one side and not the other.
But the problem is that there’s not much of a growth gap. What does this tell us about the likely effect of debt on growth?
- The effect can be random. It cannot be that extended periods of high debt to GDP automatically cause low growth. But if the effect is random, this makes it even harder to make inferences about debt and growth.
- Being lucky with industrial structure at key points in history likely matters more for growth than debt. The fact that the Netherlands and Belgium had similar economic structures was the likely reason why their growth rates were so correlated, and why debt barely mattered. This means that if you’re worried about growth in, say, the US and the UK†††, be more worried about what industries they’re competitive in and not their debt burdens.
† Page 15 of the NBB document states “The extent of the gap which has developed between public expenditure and revenue in Belgium during the past few years is exceptional. This can undoubtedly be explained to some extent by the severity of the recession and the level of unemployment”
†† When reading the NBB documents the crash in manufacturing seemed to play a big role. Documents on the Netherlands suggest that energy (the natural gas boom) may have played a larger role in the Dutch economy, but the problems facing manufacturing were also commonly brought up in the texts. Also, the remarkable correlation between the two countries’ growth rates suggests that they were buffered by common external shocks, which, given the openness of the economies, suggests that they had similar industrial compositions.
††† I purposely omit the elephants in the room — Greece and Italy—because I believe there are other problems that arise when you don’t have an independent monetary policy. I believe that lies at the core of their problems, not simply debt.
Crosslinked from medium.