Not only do studies show conflicting conclusions or different interpretations of existing information, but the sheer weight of the evidence makes it difficult to move. Any dramatic change in a large and important program is fraught with difficulties, and the more research that emerges the more the difficulties become apparent. Research also eliminates some of the simplistic arguments that would make privatization appealing. For example, work by Geanakoplos, Mitchell, and Zeldes (1998) carefully distinguishes between privatization, prefunding, and diversification and debunks the notion that privatization alone can lead to higher returns. My sense is that at least one if not more of the authors of that study may favor individual accounts, but their work demonstrates clearly that this preference has to be based on something other than an improvement in returns.
[Added to the 2000 Social Security Chronicle.]