Hostname: page-component-7c8c6479df-p566r Total loading time: 0 Render date: 2024-03-28T16:32:58.630Z Has data issue: false hasContentIssue false

Constituency Service and Incumbency Advantage

Published online by Cambridge University Press:  27 January 2009

Extract

Numerous scholars have documented a dramatic increase in incumbency advantage in US congressional elections and also state legislative elections over the past four decades. For example, Gelman and King show that incumbents in the House of Representatives now receive about twelve extra percentage points solely as a result of holding congressional office during the campaign; the comparable figure for most of the first half of this century was only 2 per cent. This advantage of incumbency has made members of the US House and many state legislators nearly invulnerable to electoral defeat.

Type
Notes and Comments
Copyright
Copyright © Cambridge University Press 1991

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Erikson, Robert S., ‘The Advantage of Incumbency in Congressional Elections’, Polity, 3 (1971), 395405CrossRefGoogle Scholar; Erikson, Robert S., ‘Malapportionment, Gerrymandering, and Party Fortunes in Congressional Elections’. American Political Science Review, 66 (1972), 1234–55CrossRefGoogle Scholar; Cover, Albert D. and Mayhew, David R., ‘Congressional Dynamics and the Decline of Competitive Congressional Elections’, in Dodd, Lawrence C. and Oppenheimer, Bruce I., eds, Congress Reconsidered (New York: Praeger, 1977), pp. 5472Google Scholar: Born, Richard, ‘Generational Replacement and the Growth of Incumbent Reelection in the US House’, American Political Science Review, 73 (1979), 811–17CrossRefGoogle Scholar; Payne, James L., ‘The Personal Electoral Advantage of House Incumbents’, American Politics Quarterly, 8 (1980), 375–98CrossRefGoogle Scholar: Alford, John R. and Brady, David W., ‘Partisan and Incumbent Advantage in US House Elections 1846–1986’, (Houston, Tex.: Center for the Study of Institution and Values, Rice University, 1988)Google Scholar; Alford, John R. and Hibbing, John R., ‘Increased Incumbency Advantage in the House’. Journal of Politics, 43 (1981), 1042–61CrossRefGoogle Scholar; Garand, James C. and Gross, Donald A., ‘Change in the Vote Margins for Congressional Candidates: A Specification of the Historical Trends’, American Political Science Review, 78 (1984), 1730CrossRefGoogle Scholar; Ferejohn, John A., ‘On the Decline of Competition in Congressional Elections’, American Political Science Review, 28 (1977), 127–46CrossRefGoogle Scholar; Nelson, Candice, ‘The Effects of Incumbency on Voting in Congressional Elections’, Political Science Quarterly, 93 (19781979), 665–78CrossRefGoogle Scholar: Mann, Thomas E. and Wolfinger, Raymond E., ‘Candidates and Parties in Congressional Elections’, American Political Science Review, 74 (1980), 617–32CrossRefGoogle Scholar; Krehbiel, Keith and Wright, John R., ‘The Incumbency Effect in Congressional Elections: A Test of Two Explanations’, American Journal of Political Science, 27 (1983), 140–57CrossRefGoogle Scholar; Cain, Bruce, Ferejohn, John and Fiorina, Morris, The Personal Vote: Constituency Service and Electoral Independence (Cambridge, Mass.: Harvard University Press, 1987).CrossRefGoogle Scholar

2 Gelman, Andrew and King, Gary, ‘Estimating Incumbency Advantage Without Bias’, American Journal of Political Science, in press.Google Scholar

3 Ornstein, Norman J., Mann, Thomas E. and Malbin, Michael J., Vital Statistics on Congress, 1989–1990 (Washington, DC: American Enterprise Institute, 1990), p. 144.Google Scholar

4 Cain, , Ferejohn, and Fiorina, , The Personal Vote, p. 123.Google Scholar

5 Cain, , Ferejohn, and Fiorina, , The Personal Vote.Google Scholar

6 See Blydenburgh, John, ‘A Controlled Experiment to Measure the Effects of Personal Contact Campaigning’, Midwest Journal of Political Science, 15 (1971), 365–81CrossRefGoogle Scholar: Fermo, Richard F., Home Style: House Members and Their Districts (Boston, Mass.: Little, Brown 1978)Google Scholar; Cover, Albert D. and Blumberg, Bruce S., ‘Baby Books and Ballots: The Impact of Congressional Mail on Constituent Opinion’, American Political Science Review, 76 (1982), 347–59CrossRefGoogle Scholar; Cook, Timothy E., ‘Review Essay’, American Political Science Review, 77 (1983), 1017–18CrossRefGoogle Scholar; Ragsdale, Lyn and Cook, Timothy, ‘Representatives’ Actions and Challengers' Reactions: Limits to Candidate Connections in the House' (paper delivered at the annual meeting of the Midwest Political Science Association, Chicago, 1984)Google Scholar: Feldman, Paul and Jondrow, James, ‘Congressional Elections and Local Federal Spending’, American Journal of Political Science, 28 (1984), 147–64.CrossRefGoogle Scholar

7 The strongest opponents of this near consensus are Johannes, John R. and McAdams, John C., ‘The Congressional Incumbency Effect: Is It Casework, Policy Compatibility, or Something Else?American Journal of Political Science, 25 (1981), 543–67.CrossRefGoogle Scholar

8 Cain, , Ferejohn, and Fiorina, , The Personal Vote.Google Scholar

9 Jewell, Malcolm E. and Breaux, David, ‘The Effect of Incumbency on State Legislative Elections’, Legislative Studies Quarterly 13 (1989). 495514.CrossRefGoogle Scholar

10 Alford, and Brady, , Partisan and Incumbent Advantage.Google Scholar

11 See Gelman, and King, , ‘Estimating Incumbency Advantage without Bias’.Google Scholar

12 Rivers, Douglas and Fiorina, Morris P., ‘Constituency Service, Reputation, and the Incumbency Advantage’, in Fiorina, Morris P. and Rohde, David, eds, Home Style and Washington Work (Ann Arbor: University of Michigan Press, 1989).Google Scholar

13 Note that using a lagged value of P as a control, as some scholars have done, is not a solution to this endogeneity problem. Indeed, in some cases, introducing an additional control variable can actually cause more bias and inefficiency than less, especially when testing causal hypotheses.

14 This measure was studied in the US Congress by Polsby, Nelson W., ‘The Institutionalization of the House of Representatives’, American Political Science Review, 62 (1968), 144–68, Table 6.CrossRefGoogle Scholar

15 In addition to the endogeneity problems caused by using constituency service at the district level, no measure of district-level incumbency advantage has ever been proposed. I therefore move to the legislature as the level of analysis. Because this produces only a single number for each election year, we would need a very long time-series of congressional election years to study the effect of constituency service. I therefore move to state legislative data.

16 Gelman, and King, , ‘Estimating Incumbency Advantage’Google Scholar, also show that every other measure of incumbency advantage proposed in the literature is biased or inconsistent.

17 These data are available from the ICPSR and were originally collected under the direction of Professor Malcolm Jewell.

18 See Silverman, B. W., Density Estimation for Statistics and Data Analysis (London: Chapman and Hall. 1986)CrossRefGoogle Scholar. When one draws a histogram, a choice must be made as to the size and number of bars, which effectively provides a compromise between averaging to find patterns (few bars) and accurately representing the original data (many bars). For kernel density estimates, this same compromise is made by the value of the ‘smoothing parameter’. For the example in the text, I use half the value of the standard error for each observation as its smoothing parameter.

19 The budget and salary data are taken from Brace, Paul, ‘The American Statehouse Transformed: Longitudinal and Cross-Sectional Analysis of Changing Legislative Resources’ (New York: New York University, mimeo, 1987).Google Scholar

20 Strictly speaking, the disturbance term in this model comes only from the uncertainly in estimating I. That is, Î = χβ + ε. One could add another disturbance term, allowing I itself, rather than just Î, to be a random variable: E(Î) = 1 = χβ + ε. In theory, this would produce slightly more efficient estimates (if this alternative model is true), but using the simpler model, even if this model is right, causes no statistical bias. Furthermore, a recent paper demonstrates that estimators of the two models produce virtually identical inferences in practice, both for coefficients and standard errors. See Jackson, John E., ‘Estimation of Variable Coefficient Models’ (paper presented at the annual meeting of the Midwest Political Science Association, Chicago, 1990).Google Scholar

21 I also did tests for autocorrelation and nonstationarity in incumbency advantage and in the explanatory variables. These tests included additional lag terms, linear and quadratic trend terms, using difference of incumbency advantage and other techniques. None of these materially altered my substantive conclusions.

22 The boot-strapped standard error is 0.006. so we can have confidence that even if certain model assumptions are invalid, this coefficient is still significantly greater than zero. I also tested for a nonlinear relationship by using the log of the budget and salary figures. These nonlinear results were somewhat stronger than those presented in Table 2, indicating diminishing returns for successively larger increments in the legislative operating budget. The results in the text are the best linear approximation to this nonlinear result and also are simpler to explain and interpret.