The moral, it appears, is "be careful for what you wish for." How did the wish come true? How did partisanship reverse course? I here discuss some of the multiple causes of ideological polarization in the House of Representatives. I pay particular attention to causes stemming from the faithful representation of constituencies by elected members, particularly the realignment of the parties in the South and changes in district constituencies wrought by redistricting. Polarization in the House will likely continue throughout the next decade, as there will be little change in these underlying causes.

The current level of partisanship in the House of Representatives is comparable to the levels experienced prior to World War II (Poole and Rosenthal 1997). ¹ Much this polarization can be traced to a key cleavage of the Democratic New Deal coalition between the southern and non-southern factions of the Democratic Party. During the decline and resurgence of partisanship, the mean ideological voting scores of Democrats elected from districts outside of the South remained unchanged. Southern Democrats, on the other hand, increasingly resisted extensions of the New Deal and grew steadily more conservative (Sinclair 1982), until the late 1960s when they reversed course towards a liberal direction.

The resolution of the ideological split in the Democratic Party is a familiar story to students of American politics (Rohde 1991; Sinclair 1995). By virtue of Democratic dominance in the South and a system of seniority and property rights on House committees that rewarded the electoral security, Southern Democrats dominated the ranking leadership positions on committees. Conservative committee chairs used their gate-keeping authority to stifle liberal legislation and southern members formed ad-hoc coalitions with Republicans when voting on the floor. As the number of non-southern, liberal Democrats grew this situation became untenable. Reforms of the early 1970s stripped power from the position of committee chair, in some cases chairmen were removed, and power was relocated under the collective control of the, now majority non-southern, party caucus. With control wrested from the conservative wing, liberals could steer policy in a direction more to their liking.

The liberalization of the voting behavior of Southern Democratic representatives was not simply the adoption of a new outlook on life. When Democratic politics took a liberal turn, southern voters and some politicians left the Democrats in favor of the more ideologically appealing Republican Party. But the Democratic Party did not die in the South. Some southerners would still vote for a yellow dog over any alternative, and comparatively liberal voters, blacks, poor whites, and migrant Yankees, entered the southern electorate and countered the exodus from the Democratic Party (Hood, Kidd, and Morris 1999). 

Today, two-party competition is alive in the South. Whereas just two decades earlier Southern Democrats were considerably more conservative than their colleagues, Southern Democrats are now only slightly more conservative. If the nationalization of American politics plays out, Southern Democrats will become ideologically indistinguishable from their peers, slightly reinforcing the ideological polarization of the House of Representatives.

The polarization of the House has also been affected by transformations in the aggregation of constituencies into districts inside and outside of the South. The most controversial change, sparking a number of Supreme Court rulings in the 1990s, is the rise in the number of majority-minority districts. Majority-minority districts are special districts mandated by the Voting Rights Act to provide discriminated minorities a chance to elect candidates of their choice. Such districts usually contain a majority of minority voters, hence the name. Majority-minority districts tend to be heavily Democratic since minorities and their neighbors tend to be members of the Democratic Party. Whether majority-minority districts are in the aggregate beneficial or harmful to the policy preferences of minorities is a matter of debate (e.g., Thernstrom 1987; Cameron, Epstein and O'Halloran 1996); however, one thing is certain: representatives elected from majority-minority districts are among the most liberal members of the House of Representatives. The emergence of majority-minority districts has thereby drawn the average ideological voting score of members of the Democratic Party to the left and contributed to the overall polarization of the House of Representatives. 

Since the adoption of the Voting Rights Act of 1965, the number of majority-minority districts steadily increased until recent Supreme Court rulings struck down redistricting plans that maximized minority representation (e.g., Shaw v Reno; Miller v Johnson). Although many have interpreted the Supreme Court rulings as an end to racial gerrymandering, the Court has also embraced a non-retrogression principle that sets a floor on the number of majority-minority districts equal to their 1980s levels. The contribution of majority-minority districts to the polarization of the House of Representatives will continue following the 2001 redistricting until the Court is willing to relax the non-retrogression principle; even then incumbents of majority-minority districts will likely fight to keep their constituencies intact.

The bizarre shapes of majority-minority districts make them highly visible gerrymanders. However, there have been more subtle gerrymanders on behalf of incumbents and political parties affecting the polarization of the House of Representatives. Using partisan registration, and an imputation where such statistics are not available, we are able to measure the impact of redistricting on the partisan composition of districts over the past three rounds of redistricting at the start of each decade (McDonald and Grofman 1999). The partisan composition of Democratic and Republican districts became more disparate, even when removing majority-minority districts from the analysis. The average difference in the percentage of Democratic registration of districts electing Republicans and districts electing Democrats increased from 11.7% in 1972, to 12.3% in 1982, to 14.0% in 1992. This national trend is mirrored in nearly every state with a divided congressional delegation; the few exceptions are states with a low number of districts, with few opportunities to gerrymander.

There is a direct link between the ideological voting score of a member and the partisan composition of the district, as filtered through the party of the member. Confirming the two-constituency hypothesis (Huntington 1950; Fiorina 1974), Democrats elected from comparable districts are more liberal than Republicans are. Within party, members elected from districts with more Democrats are more liberal. The relationship is quite strong, consistent with a modified median voter theorem (Downs 1957; Aranson, Hinich and Ordeshook 1974). Thus, through faithful representation of constituencies, the increased disparity in the partisan composition of districts has contributed to the overall polarization of the House.

Another result of the increased disparity of district constituencies is fewer competitive congressional districts. According to Congressional Quarterly Weekly Report, the number of competitive districts remained steady at 64 in both 1972 and 1992. However, the growing number of competitive seats in the South confounds the national numbers. In 1972 there were only 10 competitive southern races; by 1992 the number had risen to 34. Creating safer, more politically homogeneous districts is akin to proportional representation. Since there is a balance of partisan support in the electorate, creating safer districts produces numerical parity in the House, which heightens the need for the party leadership to enforce party discipline, especially among members elected from competitive districts, who tend to be the most centrist. The numerical balance in the House therefore further contributes to polarization through the enforcement of party discipline on moderate members.

A simple explanation as to why district constituencies increased disparity is that technological advancements provide more opportunities to gerrymander. But the means is not complete without a motive. Incumbents and political parties use redistricting as a tool to achieve their electoral goals of reelection for incumbents and, in the case of drawing congressional districts, seat maximization for parties. The way in which these goals are obtained is sensitive to the voting behavior of the electorate. Over the past 30 years the electorate dealigned, self-described partisans are less likely to vote for candidates that are offered by their parties. In order to offset the increased volatility of the electorate, incumbents and political parties are motivated to stack more partisans into districts during redistricting (McDonald 1998).

The increased polarization of political parties may have affected partisan loyalty in the electorate; recent polls show an upturn in partisan loyalty in the 1990s. Looking ahead to the 2001 redistricting, the new-found partisanship of voters will pose a dilemma for incumbents and parties. To take advantage of the increased loyalty of voters, incumbents and parties should draw more heterogeneous districts than currently exist. But incumbents want to keep their reelection constituencies intact during redistricting (Fenno 1978). If incumbents truly run the show (Cain 1984), then the levels of disparity should persist, although the need for further increases is unlikely.

Barring unforeseen disruptive political events, the ideological polarization of the House of Representatives is likely to persist throughout the next decade. The South is likely to remain competitive, the number of majority-minority districts is unlikely to decrease, and incumbents are likely to preserve their constituencies as best they can in the next round of redistricting. The balance of partisanship in the House is likely to persist as well. Legislative battles in the next decade will occur in a partisan charged atmosphere with lines drawn along representatives elected from competitive districts. Since these pivotal members will be decisive on floor votes, the institutional caucus rules reinforcing polarization of the House are likely to remain in place, if not strengthened, to enforce party discipline. Faced with cross-pressures from their constituents and party leadership, the position of these centrist members will be unenviable.

• Ideology is measured using the first dimension of Poole and Rosenthal's (1997) NOMINATE scores. The polarization of the House is measured as the difference between the mean voting scores of Democrats and Republicans.

Aranson, Peter H., Melvin J. Hinich, Peter C. Ordeshook. 1974. "Election Goals and Strategies: Equivalent and Nonequivalent Candidate Objectives" American Political Science Review 68(1): 135-152.

Congressional Quarterly, Inc. 1972. Congressional Quarterly Weekly Report 30(34): 2057-2060.

Congressional Quarterly, Inc. 1992. Congressional Quarterly Weekly Report 50(36): 2836-2841.

Cain, Bruce E. 1984. The Reapportionment Puzzle. Berkeley: University of California Press. 

Cameron, Charles, David Epstein and Sharyn O'Halloran. 1996. "Do Majority-Minority Districts Maximize Substantive Black Representation in Congress?" American Political Science Review 90(4): 794-813.

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper.

Fenno, Richard F. 1978. Home Style: House Members in Their Districts. Boston: Little, Brown. 

Fiorina, Morris P. 1974. Representatives, Roll Calls and Constituencies. Lexington: Lexington Books.

Gephart, Richard. 1998. Statement in The Arkansas Democrat-Gazette, 19 Dec. 1998. 

Hood, M.V. III, Quentin Kidd, and Irwin L. Morris. 1999. "Of Byrd[s] and Bumpers, Using Democratic Senators to Analyze Political Change in the South, 1960-1995." American Journal of Political Science 43(2): 465-487.

Huntington, Samuel. 1950. "A Revised Theory of American Party Politics." American Political Science Review 44:669-77.

McDonald, Michael P. 1998. "Optimal Gerrymandering and the Dealignment of the American Electorate." Paper presented at the annual meeting of the American Political Science Association. Boston.

McDonald, Michael P. and Bernard Grofman. 1999. "Redistricting and the Polarization of the House of Representatives." Paper presented at the annual meeting of the Midwest Political Science Association. Chicago. 

Poole, Keith T. and Howard Rosenthal. 1997. Congress: A Political-Economic History of Roll Call Voting. New York: Oxford University Press.

Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press. 

Sinclair, Barbara. 1982. Congressional Realignment. Austin: University of Texas Press.

Sinclair, Barbara. 1995. Legislators, Leaders, and Lawmaking: The U.S. House of Representatives in the Postreform Era. Baltimore: Johns Hopkins University Press.

Thernstrom, Abigail. 1987. Whose Votes Count? Affirmative Action and Minority Voting Rights. Cambridge: Harvard University Press.