The author attempting to analyze the interplay between politics and economics is faced with a large number of problems, ranging from an always too small and idiosyncratic data set to the lack of any generally accepted theory of the process which has testable implications. My aims have been (1) to show that past studies have typically treated inflation in an extremely naive way, disregarding the fact that it may have political pluses if unanticipated, and (2) to test the sensitivity of established formulations to what I believe is the more useful view of voter behavior given in Model II. Since Model I and Model II represent polar extremes of amnesiac and eidetic behavior, tests of intermediate lag structures are clearly in order. Because of administration peculiarities, however, lost data points will pose a serious problem.
Next on the research agenda is a look at the process by which the electorate forms its expectations and evaluations. Do events of the recent past indicate that administrations have lost the ability to
fool
people with fiscal and monetary policies because the public's expectations have been rationally formulated? (McCallum, 1977) If so, then there will be a lower likelihood that a president can successfully control his popularity by using monetary and fiscal policies. Or is the recent past an odd slice of history in which external events have temporarily eclipsed the administration's power to control the economy? If so, then the next few decades of data will be well worth examining.