Policy coordination often receives the same type of response as St. Augustine gave chastity: “Lord, grant me chastity and continence, but not yet.” A new volume from the IMF, edited by Atish R. Ghosh and Mahvash S. Qureshi, includes the papers from a 2015 symposium devoted to this subject. Policymakers in an open economy who take each other’s actions into account should be able to reach higher levels of welfare than they would working in isolation.  But actually engaging in coordination turns out to be harder–and less common– than many may think.

Jeffrey Frankel of Harvard’s Kennedy School of Government uses game theory to illustrate the circumstances that hamper coordination. One factor may be a fundamental divergence in how different policymakers view a situation. Many analysts on this side of the Atlantic, for example, use the “locomotive game” to show that Germany should engage in expansionary fiscal policies that would raise output for all nations. But (most) German policymakers have different views of the external impact of deficit spending. In the case of the Eurozone, a deficit in one country increases the probability that it will need a bailout by the other members of the monetary union. Only rules such as those of the Stability and Growth Pact that limit deficit expenditures can eliminate the moral hazard that would otherwise lead to widespread defaults.

Charles Engel of the University of Wisconsin (working paper here) also examines the recent literature on central bank coordination. He points out that the identifying the source of shocks is necessary to assess the benefits of cooperation to address them, and suggests that financial sector shocks may be most relevant for modeling open-economy coordination. But widespread cooperation could undercut the ability of a central bank to credibly commit to a single target, such as an inflation target.

Policymakers in emerging markets who must deal with the consequences of policies in advanced economies have been particularly mindful of their spillover effects. Raghuram Rajan, for example, who is back at the University of Chicago after serving as head of India’s central bank, has urged the Federal Reserve and other central banks to take into account the impact that their policies have on other nations, particularly when unwinding their Quantitative Easing asset purchases. He pointed out: “Recipient countries are not being irrational when they protest both the initiation of unconventional policy as well as an exit whose pace is driven solely by conditions in the source country.”

If international cooperation is viewed as a bargaining game, what incentives do the advanced economies have for cooperative behavior in light of the asymmetries among nations? Engel points out that in such circumstances, “…the emerging markets may believe that they have too little say in this implicit agreement, which is to say that they may perceives themselves as having too little weight in the bargaining game.” Conversely, central banks in the upper-income countries may in ordinary circumstances see little need to extend the scope of their decision-making outside their borders.

This attitude changes, however, when a crisis occurs, as Frederic Mishkin of Columbia shows in his examination of the response of central bankers to the global financial crisis. The Federal Reserve established swap lines to provide dollars to foreign central banks in countries where domestic banks faced a withdrawal of the funding they had used to acquire dollar-denominated assets. In addition, six central banks—the Federal Reserve, the Bank of Canada, the Bank of England, the European Central Bank, the Sveriges Riksbank and the Swiss National Bank—announced a coordinated reduction of their policy rates. Coordination becomes quite relevant in a world of sudden stops and capital flight.

The need for such activities could increase if there is a global financial cycle, as Hélène Rey of the London Business School has stated. She presents evidence of the impact of global volatility, as measured by VIX, on international asset prices and capital flows. An important determinant of such volatility is monetary policy in the center countries. Rey agrees with Rajan that: “Central bankers of systemically important countries should pay more attention to their collective policy stance and its implications for the rest of the world.”

Perhaps a better motivation for the need for joint action comes from Charles Kindleberger’s list of the responsibilities that a hegemonic power such as Great Britain played in the period before World War I. These included acting as a lender of last resort during a financial crisis; indeed, it was the lack of such an international lender in the 1930s that Kindleberger believed was an important contributory factor to the Great Depression. Since the end of World War II the U.S. has vacillated in this role while the international monetary system has moved from crisis to crisis. Meanwhile, offshore credits denominated in dollars have grown in size, and could conceivably constrain the Federal Reserve’s ability to undertake purely domestic measures.

A policy of “America First” that means “America Only First and Last” ignores the fragility of the international financial system. Just as there are no atheists in foxholes, no one doubts the merits of coordination when there is a disruption of global markets. But suffering another crisis would be an expensive reminder that the best time to minimize systemic risk is before a crisis erupts.