Fullerton, A. and Dixon, J. 2010. "Generational Conflict or Methodological Artifact? Reconsidering the Relationship between Age and Policy Attitudes in the U.S., 1984-2008", Public Opinion Quarterly 74(4): 643-673.
The authors in this piece limit their analysis to issue areas that they claim are particularly salient for age-based divisions: education, health care spending (particularly Medicare) and Social Security. They acknowledge that without some form of methodological wizardry, that it is impossible to distinguish between age (i.e., life-cycle effects), period (based on independent variables that vary over time), and cohort (driven by particular historical events common to all born at a particular time) effects. They spend some pretty serious time justifying their methodological choices, and do so in a manner that is clear even to those of us who are not deeply immersed in the cutting edge of public opinion analysis methods, so well done there. They find that when they change methods, that the conventional wisdom of Campbell goes poof. In short, the life-cycle hypothesis is an artifact of bad methods. When the data is properly disaggregated, then the effect changes. The remainder of their results match the expected outcomes.
My take: I get the notion that in order to analyze data in certain ways, we have to engage in manipulation. Sometimes this is justified. Sometimes it isn't. This is one case where, because of the clarity of the authors' explanations, both as to the shortcomings of existing methods and why they think their methods are better, the legerdemain is justified. However, I have to strongly disagree with their explanation of their results. They argue that the Greatest Generation experienced two different views of government during their formative years: the laissez-faire approach of the Roaring Twenties and the Hoover administration, and the New Deal, which gives them an appreciation for what government can do and a skepticism toward its overreach. But this cohort in fact experience the failure of laissez-faire. Any skepticism about government spending that this cohort expresses seems unlikely to come from their experiences in Hoovervilles, since it was government programs like the CCC and WPA that got them out of those shantytowns.