A Wuffle, Thomas Brunell, and William Koetzle, University of California, Irvine
Maltzman, Sigelman and Binder (1996), in what we regard as monumental work, have reinvigorated the study of death in office. But their data analysis is incomplete. First, they fail to project realistically the trends in actual and expected congressional deaths illustrated in their paper. Second, like many congressional scholars, they do not pay sufficient attention to the Senate. We provide new evidence that the U.S. Senate may offer the solution to mankind's oldest quest, the fountain of youth (Gilgamesh n.d.; de Leon (1460-1521).2 Table 1 presents data on the proportion of U.S. Senators who died in office, by decade, from 1910 to 1990 (CQ 1995).
TABLE 1 Mortality of Sitting Senators By Decade ------------------------------------ Proportion of Senators Who Decade Die While in Office ------------------------------------ 1910 - 1919 .30 1920 - 1929 .25 1930 - 1939 .24 1940 - 1949 .28 1950 - 1959 .20 1960 - 1969 .15 1970 - 1979 .08 1980 - 1989 .03 ------------------------------------We plotted the senatorial mortality rate data reported in Table 1 versus year (for data grouped by decade) as shown in Figure 1.
The r2 for the regression of mortality rates on time is an astonishingly large .87.3 Rarely have we seen such a clear time trend. Senators are increasingly not dying in office.
However, as our colleague, Bernard Grofman, pointed out to us (personal communication, April 2, 1995), projecting this equation past 1990 would give us negative estimates of mortality rates as early as the year 2000. Given the strength of our empirical results and the scholarly commitment we share with Gelman and King (1990) to investigating incumbency effects without a priori bias, we do not reject the idea of revenants out of hand.4 Nonetheless, to obtain a mortality function with the more commonly accepted lower bound of zero, we have re-estimated the data with a logarithmic specification, by regressing the log odds of senatorial mortality versus time.5
Here we obtain
(1) log odds MORTALITY RATEThe r2 value for this logarithmic regression is .78 (with an adjusted r2 of .74.) While not quite as good as the previous linear fit, the fit of the logarithmic estimate is still impressive.
= 61.78 - .033 YEAR.
We may reexpress Equation (1) in more convenient form as
(2) MORTALITY RATEe61.78 - .033 YEAR = ------------------------------ 1 + e61.78 - .033 YEAR
Substituting in the value 2000 for YEAR, we find that senatorial mortality rate for that year is estimated to be about 2 percent. By 2050, estimated Senate mortality will be under one percent. By the year 3000, it will be virtually indistinguishable from zero.
The sophisticated logarithmic regression above provides clear and compelling evidence that, in the modern era, Senators who manage to retain office will be nearly immortal.6 Moreover, we may rule out a selection bias effect as the explanation for this phenomenon in that there is equally clear and compelling evidence that Senators who leave office do die.7
Our work provides a new and powerful substantive underpinning to Anthony Downs' oft-repeated yet sometimes disputed assertion that politicians have one motivation only --to retain office (Downs 1957; cf. Mayhew 1974). We learn from our findings that all previous work on incumbency advantage (e.g., Fiorina 1989; Gelman and King 1990; King and Gelman 1991) has completely missed the real nature of incumbency advantage, at least for the Senate. Thus, once our findings have been disseminated and their implications digested, we should see a drastic change in the willingness of Senators to seek the presidency. Our work also has clear implications for public policy in that it leads us to predict that the imposition of federal term limits will raise the national death toll.8
Although our empirical contribution is resolutely pre-mortem in spirit, it makes a valuable contribution to post-rationalist theory in the Wuffeauldian tradition.9 Of course, our findings need to be replicated for House members and for other bodies.10 Moreover, just exactly what it is about serving in the Senate (or being re-elected thereto) that keeps one from dying is a topic that must be left to future research.11
1. This paper was not prepared for delivery in Barbara Sinclair's class on the "U.S. Congress," UC Riverside, Winter Quarter, 1995. Nonetheless it was handed out there. The listing of authors is alphabetical by first name.
2. Our own work on this topic was written before we had the benefit of Maltzman, Sigelman and Binder's (1996) remarkable insights, but we have found no reason to change anything we had previously written.
3. Even the adjusted r is still a whopping .84.
4. Cf. Kubrick and Clarke (1968).
5. If we let M denote mortality rate, what we have done is to regress log odds MORTALITY RATE (= ln(M/1-M)) versus YEAR, for data grouped by decade.
6. See e.g., Strom Thurmond.
7. Witness Hubert Humphrey or Lyndon Johnson, for example. Thus, we may reject the selection-bias inspired hypothesis that senators do not die in office because only immortals (or near immortals) are elected to that office.
8. Even Cain's mind-boggling review of 666 (mutually contradictory) hypotheses about the effects of term limits (Cain, 1996) omits this one.
9. Stone (1995) compares and contrasts the Wuffeauldian paradigm (see e.g., Wuffle, 1979, 1982, 1984, 1988, 1989, 1992, 1993, 1997 forthcoming) to other work in the continental tradition. Stone's one sentence summary of the Wuffeuldian approach to the scientificistic method-"Truth is like a truffle; to find it we must dig around a lot-and then somehow get rid of all the dirt"-has not yet been improved upon.
10. Christopher Zorn, Ohio State University (personal communication, March 26, 1995) has suggested a similar pattern obtains for the U.S. Supreme Court.
11. In this context, consider the joke about the scorpion and the lawyer, the punchline to which is: "professional courtesy."
Cain, Bruce. 1996. "The Varying Impact of Legislative Term Limits." In Bernard Grofman (ed.) Term Limits: Public Choice Perspectives. Boston, MA: Kluwer Publishers.
Congressional Quarterly Guide to US. Elections. 1995. Washington, D.C.: Congressional Quarterly Press.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper and Row.
Fiorina, Morris. 1989. Congress: Keystone of the Federal Establishment, 2d Edition. New Haven, CT: Yale University Press.
Gelman, Andrew and Gary King. 1990. "Estimating Incumbency Advantage Without Bias." American Journal of Political Science 34:1142-1164.
King, Gary and Andrew Gelman. 1991. "Systematic Consequences of Incumbency Advantage in U.S. House Elections." American Journal of Political Science 35.
Kubrick, Stanley and Arthur C. Clarke. 1968. "2001: A Space Odyssey."
Maltzman, Forrest, Lee Sigelman and Sarah Binder. 1996. "Leaving Office Feet First: Death in Congress." PS 29(4):664-71.
Mayhew, David. 1974. Congress: The Electoral Connection. New Haven, Conn.: Yale University Press.
Stone, Alec. 1995. "Of Wuffle and Truffles: Gastronomic Investigations in the Neighborhood of the 1995 ECPR Meeting in Bordeaux." Unwritten manuscript, School of Social Sciences, UCI, April 30.
Wuffle, A. 1979. "Mo Fiorina's Advice to Children and Other Subordinates." Mathematics Magazine, Vol. 52, No. 5 (November), 292-297.
Wuffle, A. 1982. "The Pure Theory of Elevators." Mathematics Magazine, Vol, 55, No. 1 (January), 30-37.
Wuffle, A. 1984. "Should You Brush Your Teeth on November 6, 1984?: A Rational Choice Perspective." PS (Summer), 577-580.
Wuffle, A. 1988. "The 'Minimax Blame' Model for Voter Choice: A Guide to the Perplexed." PS (Summer), 85-88.
Wuffle, A. 1989. "Finagle's Law and the Finagle Point, a New Solution Concept for Two-Candidate Competition in Spatial Voting Games Without a Core." American Journal of Political Science, Vol. 33, No. 2, 427-454(with Scott Feld, Guillermo Owen, and Bernard Grofman).
Wuffle, A. 1992. "A Corollary to the Third Axiom of General Semantics." Journal of Theoretical Politics, Vol. 4, No. 2 (April), 238-240.
Wuffle, A. 1993. "The Political Economy of the Automobile: Four Approaches." Journal of Theoretical Politics, 5 No 3 (July), 409 -412.
Wuffle, A. 1997. 'Why Democrats Shouldn't Vote and Probably Shouldn't Register Either (with acknowledgments to R. Erikson)," Journal of Theoretical Politics, Forthcoming (with Christian Collet).
About the Authors
A Wuffle is a permanent assistant to professor, School of Social Sciences, University of California-Irvine, a position he has held since 1980. He has published more than anyone would wish, including such unjustly neglected classics of post-rationalist theory as "A Corollary to the Third Axiom of General Semantics," Journal of Theoretical Politics, 1993 and "Should You Brush Your Teeth on November 6, 1984," PS, 1984.
Thomas Brunell and William Koetzle are innocent (but very well-trained) political science graduate students at University of California-Irvine, whose empirical research the senior author of this paper is exploiting in accord with the best traditions of European scholarship. The two of them also wrote "Lipreading, Draft-Dodging and Perot-Noia: Party Campaigns in Editorial Cartoons." Harvard International Journal of Press/Politics. Vol 1 No. 4 (1996): 94-115, and already have a number of other singly authored and multiply authored articles currently under submission.