if you actually read the paper, Stiglitz's conclusions are fine. the research design has obvious limitations, such as the utility of historical data, but stiglitz is always honest about the implications of these limitations, and makes an excellent point about the government being indirectly...