Many have been talking about possible international contagion of the financial crisis in Turkey, and Peter Coy touched on the key issues in yesterday’s Bloomberg piece. Recent economic history and theory offer powerful lessons about contagion. Most important is that contagion isn’t automatic, but rather depends crucially on economic policy. And that lesson is fortunately showing up in virtually all the market commentary during the past few days

Consider in contrast the situation exactly twenty years ago on August 17, 1998 when Russia—in the middle of its own financial crisis—defaulted on its debt and the impact spread instantaneously around the world. The three charts on my bookmark, shown here, illustrate that experience. The emerging market bond index (EMBI) spread jumped in Asia, Africa, and Latin America at the time of the Russian dealt which is marked by the red vertical line.

Many argued at the time that this was the way the world markets had become in the globalized economy; policy makers at the IMF and elsewhere, it was argued, had to be ready with bailouts to deal with the inevitability that a financial crisis in one country would automatically lead to crises in many other countries.

This was the still a common view in January 2001 when I joined the U.S. Treasury to run the international division. A crisis was brewing in Argentina then, and many argued that we needed to be ready for bailouts of emerging markets around the world to deal with predicted contagion.

Argentina in 2001, like Russia in 1998, was in a financial crisis and it did default on its debt; however, unlike the Russia case, the international contagion effects were nowhere to be seen, as illustrated on the reverse “no contagion” side of the bookmark shown next.  The EMBI spread in Asia, Africa and Latin America was unaffected by the default.  

There were important policy differences between the two episodes. International policy was becoming more predictable and gradual so that the Argentinian default was widely anticipated, discounted and had little or no impact.  Other countries were better prepared, and, by this time, the notion that there were important policy differences that affected the likelihood of contagion became more widely discussed. One could find a great deal of difference between countries and thus find ways to discriminate between good policy and bad policy.  Kristin Forbes, then at MIT and the Treasury, was doing a lot of the research. There was also the new view that an orderly restructuring rather than a sudden default would lead to less contagion and would be possible if collective action clauses were introduced in sovereign debt.

In my view, it is encouraging that now the policy differences between countries are being examined carefully. Analysts are pointing to poor monetary and fiscal policy in Turkey as the source of the problem. They are creating heat maps, as shown below and drawn from the paper by Peter Coy that tries to show the differences in policy in different countries. Those most susceptible to contagion are countries where policy has been relatively poor in recent years such as Argentina and South Africa.

There is still a danger of poor policy responses in Turkey or elsewhere that could make the problem worse, and even lead to contagion.  Trade policy skirmishes could become a trade war and are already a source of uncertainty.  If Turkey resorted to capital controls to limit outflows, then investors might expect such controls in other countries, and get out while the getting is good, which could result in contagion. The fact that the IMF has developed and publicized the concept of capital flow management in recent years could hastened this contagion.

Some are blaming the normalization of policy in the United States, and suggesting that the policy be slowed or halted. The normalization by the Fed seems to put downward pressure on exchange rates in some emerging market countries. But it would be a mistake for the US and the world economy to change policy at this time.