As "globalization" took off in the public imagination in the 1990s, many economists (myself included) viewed this not only as a great manifestation of the gains from trade, but also as a disciplinary check on politicians and policy. Policies which are counter-productive to wealth creation will lead to capital moving away from the counter-productive and to the more productivity friendly policy regimes. Countries will lose both creative people and capital when they adopt stupid policies, and attract them when they follow more "friendly" policies for creativity, entrepreneurship, and investment. Globalization will discipline governments through a competitive process to attract a tax base, etc. In short, a process analogous to the Tiebout process associated with local public economies is applicable to the global economy. The best book on this, in my opinion, was Quicksilver Capital by Richard McKenzie and Dwight Lee.
Unfortunately, Tiebout competition can be derailed when there are effective mechanisms of collusion among the various levels of government. The collusion among governments subvert the competitive pressures that Tiebout identified as my colleague Bryan Caplan demonstrated in one of his first academic papers. As market-preserving federalism breaks down due to government collusion with respect to key policy variables, the discipline of competing for a tax-base is muted. Capital doesn't exist as quickly.
In the NYT today Thomas Friedman argues the current global financial system is more analogous to bumper to bumper traffic where one accident can result in a long traffic jam, rather than continuous competitive pressure caused by the ability of individuals to move and investment in more favorable locations.
What do you think? Are the forces of quicksilver capital defeated by the collusive behavior of leading governments to check the disciplinary competitive pressures for good economic policies? Or do you think this is the wrong way to think about international political economy and globalization?