What is the long term impact of economic and political crises on poor countries? Valerie Cerra and Sweta Saxena take a shot at this question in the latest issue of the AER.
They look at how output responds to different sorts of crises. Consider this graph, which shows the percentage decline in output (vertical axis) in the years after a currency crisis (where time is on the horizontal axis):
The output decline is large and persistent, even after a decade.
Compare this to a dual currency and banking crisis:
The pattern looks the same, but the decline is three times as great.
Here is where it gets interesting: output also declines after civil wars, but the rebound is rapid and substantial. This is the pattern of output after wars break out:
Output rebounds after civil war, but not after currency crises.
The rebound from civil war makes sense in a simple growth framework: capital is destroyed, the economy is out of equilibrium, and investment brings us back to equilibrium after the conflict ends. We don’t get all the way, however, perhaps because wars persist, or human capital takes longer to re-accumulate. Or perhaps there is some kind of permanent institutional decline thaat makes investment less attractive.
The authors also look at another type of ‘political crisis’, mainly a descent into autocracy. This too causes a large output decline. Here is what happens to out put when you have a dual political crisis: civil war and an increase in authoritarianism:
The output decline is three times as deep, and very persistent. My guess is that the combination of institutional decline and conflict means that capital and investment decline permanently.
What is surprising to me is that the output decline after economic crises are similarly large and persistent. Do they cause a permanent decline in the quality of the investment climate? A rise in the price of capital? A hostile political climate?
It’s too bad the authors do not venture an opinion. They also do not chart the impact on output growth rates–information that could help clarify the nature of the impact.
Neoclassical growth models have specific predictions about level versus growth effects of a crisis or event. Charting these dual dynamics, and trying to unpack the causal mechanism, would make for a terrific dissertation project.
One Response
I’m just floating this out there, but couldn’t there be selection bias issues? Countries may have civil wars and the like for reasons unrelated to the strength of their economy. If, however, output before a currency/banking crisis was unsustainable, and thus the crisis occurred, then output would fall and not recover. It reminds me of South Korea – everything was fine and dandy, but when the crisis kicked off, all of a sudden people realized that the underlying system wasn’t as sound as they thought it was and that much of the previous growth had been built on sand. Reachieving that growth (that was previously based on ignorance) would be rather difficult.