A test for pure or apparent ideology in congressional voting

A politician's preferences do matter. He is not simply a captured agent speaking only for his constituent's interests. Some latitude exists for him to shirk his responsibility to serve his constituents' interests and to cast votes consistent with his own personal ideology. More important to our understanding and modeling of voting behavior, shirking for ideological consumption is seen to increase with age once a threshold of about 53 years old is reached. Shirking is not a phenomena of the politicians' final term in office, rather it likely exists in every term and begins to increase significantly two terms prior to the expected horizon. Sixty percent of the sample of senators were seen to exhibit increasing ideological consumption with age. Other studies have attempted to distinguish the investment and consumption components of ideological measures by more appropriately characterizing the constituent interest variables (Peltzman, 1984) or purifying the ideological variable (Kalt and Zupan, 1988; Kau and Rubin, 1979). Of course, one might always ask, are the variables now properly characterized and completely pure? This paper takes a different approach. It shows that the ideological measure, as an explanatory tool, increases in importance with age. Our explanation is that there is a true ideological component imbedded in the measure and politicians do, in fact, shirk their constituents' interests to engage in ideological consumption. Because an approaching horizon lowers its price, ideological consumption increases with age. Copyright Kluwer Academic Publishers 1989