Knowledge that Transforms

Old age, like beauty, is in the eye of the beholder. If by old age we mean the period of life characterized by a significant reduction in market work, then a major phenomenon of the last three decades has been that the old among us have gotten considerably younger. Here we briefly review this phenomenon, discuss a parallel growth in broadly defined income support, both public and private, and then provide a mechanism for relating the two.

In a recent article in this Review, Alex Cukierman and Paul Wachtel (henceforth C-W) develop a theoretical relationship between the cross-sectional variance of expectations about inflation and the variance in aggregate demand and in the inflation rate over time. They also present some empirical evidence in support of these linkages. The purpose of this comment is to indicate several weaknesses in their analysis. First, an error in C-W's theoretical analysis leads to an error in their Proposition 3(b). A correction renders the set of conditions for the variance in expected inflation as they define it to move together with the variance in actual inflation even more complex than their Proposition 3 suggests. Second, C-W use the Livingston data in some tests of their Propositions 2 and 3, which link the variance of expected inflation as they define it to aggregate demand variance and to the variance of inflation. In doing so, they misinterpret the Livingston Survey in the context of their model. Actually, the model has much more straightforward implications for the variance in the Livingston inflation forecasts than those derived by C-W. Third, upon closer examination, their empirical evidence from the Livingston Survey data is not consistent with these implications of their model. This is because a minor C-W data mistake makes their preliminary evidence look more striking than it actually is, and because their regressions are marred by extreme serial correlation correcting for which alters their results.