Political scientists have shown that the state of the national economy has a substantial impact on voting behavior and election outcomes. Pioneers in this line of scholarship viewed their work as a bracing antidote to pessimistic concerns that the unsophisticated masses would be swayed by slick campaign ads or demagogic appeals. For example, Gerald Kramer characterized his findings as demonstrating "that election outcomes are in substantial party responsive to objective changes occurring under the incumbent party; they are not 'irrational,' or random, or solely the product of past loyalties and habits, or of campaign rhetoric and merchandising." In the same vein, V. O. Key Jr. interpreted retrospective voting as evidence for his "perverse and unorthodox argument" that "voters are not fools."

This optimistic interpretation of economic voting seems to me to be misconceived. While it is certainly true that American voters respond to "objective changes occurring under the incumbent party," they do so in ways that produce an immensely consequential partisan bias in economic accountability. While they may not be fools, they are short-sighted, and their myopia often leads them to cast votes that are damaging to their economic interests.

The notion that citizens can maintain an accurate "running tally" (as Morris Fiorina nicely put it) of the successes and failures of the incumbent administration sounds much more plausible than the alternative notion that they formulate detailed views about major issues of public policy and use those views to weigh the strengths and weaknesses of the competing candidates' platforms. However, even this less ambitious understanding of how voters decide ignores a variety of complexities. How do voters know whether and how the incumbent's policies have actually contributed to good or bad conditions? (My Princeton colleague Christopher Achen and I have found that voters routinely punish incumbents for events over which they clearly have no control, including droughts, floods, and shark attacks.)

In chapter 4 of Unequal Democracy I show that voters are remarkably myopic. Economic conditions have a powerful impact on presidential election outcomes - but only economic conditions in the year of the election. Economic conditions in the third year of each four-year term have much less effect, and economic conditions in the second year of each four-year term (the first year in which a president's policies are likely to have any significant effect) have no discernible effect on election outcomes. As a perceptive foreign observer, Moiseide Ostrogorski, surmised more than a century ago, the American voter "sees only present advantages."

If economic conditions in presidential election years were representative of the incumbent party's performance more generally, this quirk in voters' time horizons would only be interesting to psychologists. What makes it politically significant is the fact that conditions in election years are not representative of the parties' economic performances more generally. As I noted the other day, Democratic presidents have produced substantially more real income growth for middle-class and working poor families over the past 60 years. However, Republican presidents have actually produced significantly more real income growth for middle-class and working poor families in presidential election years. As a result, voters' myopia produces a powerful boost to Republican candidates - enough of a boost to have been decisive in three of the nine Republican victories of the post-war era.

Why have Republican presidents produced so much more income growth in election years than Democratic presidents have? Certainly some of their success is due to intentional political manipulation, as when Richard Nixon issued checks reflecting higher Social Security benefits in October 1972 but delayed the corresponding increase in the payroll tax until after the election. The stimulus checks reaching millions of voters this month will likewise go some way toward insulating the incumbent party from economic discontent in November.

However, both parties can and do engage in election-year manipulation of the economy. The Republicans' marked advantage in election-year performance has much more to do with partisan differences in patterns of income growth earlier in each administration. Democratic presidents typically use their post-election "honeymoon" periods to implement expansionary policies producing robust economic growth in the second years of their administrations; the resulting economic booms cannot be sustained indefinitely, and real income growth usually slows considerably in the third and fourth years of Democratic administrations. In contrast, Republican presidents typically preside over significant economic contractions early in their terms, with income growth recovering as the next election approaches.

Overall, middle-class and poor voters do substantially better under Democrats. But their short-term focus leads them to forget or ignore much of the good fortune they experience under Democrats and much of the economic distress they experience under Republicans. By forgetting history, they often vote to repeat it, reelecting Republicans who preside over even less income growth in their second terms than in their first terms (as Eisenhower, Nixon, and Reagan all did and Bush probably will).