Dani Rodrik argues that global institutions are not always the best answer to global problems. He favors domestic policies that limit the spillovers from other countries arguing that "This option may seem protectionist, but it could ultimately ensure a more durable globalization":
Don’t Count on Global Governance, by Dani Rodrik, Commentary, Project Syndicate: ...The absence of global institutions – acting as lender of last resort or serving up coordinated fiscal stimulus – aggravated the crisis and delayed the recovery. And now, go-it alone fiscal, monetary, and exchange-rate policies are spilling across national borders, threatening currency wars and protectionism.
How we deal with these challenges is the greatest economic question of our time. One approach, favored by technocrats and most policymakers – at least until domestic politics intrudes – is to seek solace in ever-greater global governance. Global problems, after all, require global solutions, which means strengthening international organizations like the International Monetary Fund,... the G-20, and negotiating stricter international codes and standards (as has occurred with capital-adequacy requirements, for example).
Another approach is to recognize that global governance is bound to remain incomplete, and to moderate the side-effects through a more cautious form of economic globalization. This strategy entails throwing some sand in the wheels of the global economy in order to expand room for domestic policy and limit the impact of adverse spillovers from other countries’ actions. This option may seem protectionist, but it could ultimately ensure a more durable globalization.
Many of the world economy’s troubles today originate from our unwillingness to recognize that domestic policy objectives will ultimately trump global responsibilities...
Economists teach the virtues of open trade because it benefits us... Exposing the domestic economy to global markets ... brings its own rewards. A world economy made up of countries that pursue their own national interests will perhaps not be hyper-globalized, but it will, by and large, be an open one.
Certainly, the global economy needs some traffic rules where there are clear cross-border spillovers. But the balance between national prerogatives and international rules must make a virtue of political reality. If we veer too far toward global governance, we will end up with meaningless rules that beg to be flouted.
Just a quick, on the run comment: Financial regulation is one area where I think global institutions are needed, and there is also a need for a global lender of last resort.