Accueil Revue Numéro Article

Revue française d’études américaines

2001/4 (no90)

  • Pages : 128
  • ISBN : 9782701129181
  • DOI : 10.3917/rfea.090.0010
  • Éditeur : Belin

ALERTES EMAIL - REVUE Revue française d’études américaines

Votre alerte a bien été prise en compte.

Vous recevrez un email à chaque nouvelle parution d'un numéro de cette revue.


Article précédent Pages 10 - 18 Article suivant

Having failed repeatedly to get rid of the Electoral College, an eighteenth century institution that was originally intended to insulate Presidential selection from popular opinion, it was just a matter of time before a US Presidential election would eventually produce different popular and Electoral College outcomes. All it would take was a close election in which one party won more of its state Electoral College votes by narrower margins than the other party. Always a theoretical possibility, it had not actually happened in the twentieth century until the 2000 election. [1][1] This point is disputed by some who claim that Kennedy... But why was the 2000 election so close that it caused the outcome that the US had hoped to avoid?


The answers lie in the gradual evolution of the US system over the last two decades towards a position of relative parity between the two parties, the failure of the Democrats to capitalize on the incumbency advantage that normally accrues to a party that presides over economically prosperous times, and the increasing skill of political campaigns in targeting resources efficiently where they matter the most.


There was no dramatic shift in 2000 in the underlying sociology of American elections. Rather, many of the characteristic trends seen in the 1992 and 1996 Presidential contests reappeared in 2000 despite attempts by both candidates to re-position themselves strategically. The consequence of these evolving patterns is that Democrats and Republicans have now reached a position of relative parity, giving independents and moderates the tie-breaking power to determine the final electoral outcome.


An important stable trend in recent years is the modest income divide in current US partisanship. In the last twenty-five years, Republicans have only once received as much as two-thirds support from the highest income voters (i.e. Reagan in 1988), and since the early nineties, barely more than half. [2][2] All polling: the figures come from exit poll data compiled... The same can be said about the Democrats and the lowest income groups. The reason for this is the cross-cutting effects of two dimensions of US politics. One is the traditional left-right split over taxes and government spending, and the other is the social division over such issues as abortion, gun control, women’s rights and family values. A significant proportion of highly paid, secular professionals who are moderate to conservative on economic issues and moderate to liberal on social issues have gravitated into the Democratic party, and vice versa, of low income, rural and small town working class into the Republican party because of its religious right appeal.


This pattern seems to be relatively immune to the strategic nuances of the various Presidential candidates. Where Bill Clinton decidedly catered to the middle of the ideological spectrum, Al Gore attempted to position himself as more of a populist—by opposing Bush’s across the board tax cut, proposing costly new spending programs such as prescription drug benefits for the elderly, and tying himself more closely in terms of both operations and programs to the unions. On the other side, Bush’s strong conservative credentials, his visit to Bob Jones University and a tax cut plan that was clearly aimed at giving substantial relief to wealthier groups should have increased his upper income vote in 2000, but it did not. Bush won only 54% of those making over 100,000 dollars versus 43% for Gore as compared with the support that Ford (63%), Reagan (64% and 69%) and the elder Bush (62%) received from the highest income voters from 1976 to 1988. Similarly, Gore received 57% of the lowest income vote as compared to 59% for Clinton in 1996, 55% for Mondale in 1984 and 58% for Carter in 1976.


There were other continuities between 2000 and the two previous elections as well. The gender gap was prominent once again. In 1976, Ford and Carter got identical shares of the male and female vote. But beginning with Reagan in the eighties, men began to prefer the Republican candidate by larger margins than women (Kaufman and Petrocik 1999; Abramowitz 1995). By 1996, men and women favored different candidates with men choosing the Republican Bob Dole 44% to 43% and women supporting the Democrat Bill Clinton 54% to 38%. In 2000, despite the fact that Al Gore and George Bush both deliberately downplayed abortion to avoid offending swing voters in mid-western states, women’s and pro-choice groups rallied to Gore’s side in the last month of the campaign, giving him a 54% to 43% margin while men favored Bush 53% to 42%.


Why did the gender gap persist when the candidates themselves in 2000 sought to downplay the abortion issue? The answer is that woman’s groups paid little attention to the campaign rhetoric, and based on Bush’ record and prior statements about his judicial philosophy, they concluded that he fully intended to appoint conservatives to the Supreme Court who would vote with Justices Clarence Thomas and Antonin Scalia to restrict abortion rights. Given the importance of the Supreme Court and the likelihood that the next administration might have the opportunity to appoint as many as three new justices in the next four years, the abortion issue clearly polarized voters by gender in 2000 just as it had in 1996.


The regional variations of the past two elections were also evident. George Bush’s strongest support came from the South and the Western Plains States. Al Gore, like Clinton before him, got his core support in the Northeast, the North Central and Far West states. It has not always been thus. In 1976, Carter failed to carry California, but won the South and the Northeast. By contrast, Clinton carried California in both 1992 and 1996, as did Al Gore in 2000, but only managed an even split in the South. Things were so bad in the South in 2000 that the Democrats did not even carry Clinton’s home state of Arkansas or Gore’s home state of Tennessee.


What does the sharp regional variation in the Presidential vote mean? Most likely, it reflects the great cultural divisions in the US, which are particularly salient in times of relative economic prosperity. Republicans derive their core support in areas where the Protestant Religious Right flourishes whereas the Democrats do better with Jews and Catholics, and in the more secular areas. White Protestants favored the Southwestern candidate George Bush 63% to 34% over the southeastern candidate Al Gore. No Democratic Presidential candidate since Jimmy Carter in 1976 has ever received more than the 40% of the White Protestant vote. Carter, a White Protestant who taught Sunday School and infused his Presidency with religious values, only got 31% of the White Protestant vote in 1980. Since White Protestants constitute close to half of the US electorate, the Republican advantage with them is a major obstacle for the Democrats.


A corollary to the regional divide is the tendency for the Republicans to run strongly in the small rural communities. Rural voters number about a quarter of the electorate. They favored Ronald Reagan and the senior Bush by comfortable margins in the eighties. Clinton was able to win many of them back in 1992 and 1996, but in 2000, Gore lost the rural vote 59% to 37% and the small town vote 59% to 38%. Gore’s strengths were in the large and medium size cities (equaling about 29% of the voters), and he narrowly edged Bush for the suburban vote (43% of the total electorate). Many of these rural areas are in the West and the South. We can speculate whether Bush’s advantage in the small towns and rural areas may have stemmed in part from his pledge to restore family values and national honor in the wake of eight years of Clinton’s ethical misadventures.


To some degree, the regional divisions also reflect the racial divide in America. Blacks in the South and West favored Gore by the same wide margins that they did in the East and Midwest, but Southern Blacks only compose 5% of the total voters whereas Southern Whites equaled 23% of the electorate. Outside the south, the largest concentrations of Blacks and Latinos are in the large urban areas. Blacks in 2000 supported Gore 90% to 8%, Latinos 67% to 31% and Asians 54% to 41%. The inability of the Republican Party to win large numbers of nonwhite voters is a major concern for the party, and one that Bush, to his credit, has attempted to address in the diversity of his cabinet and White House appointments. Since the nonwhite share of the electorate is growing through immigration and natural birth rates, the Republican Party cannot afford in the long run to be perceived as hostile to nonwhites.


The dilemma for the Republican Party is how to avoid alienating its white Protestant base in its effort to include the growing nonwhite population in its ranks. The party owes its success in the South at least partly to its opposition to liberal civil rights policies such as affirmative action, forced busing to integrate schools and re-distributive policies (e.g. welfare programs) that benefited poor urban nonwhite populations. Sticking to its base might be the safest course in the short run, but the sorry state of the California Republican Party is a visible warning to the national party that alienating nonwhite voters could be a perilous tactic in the long run. A series of anti-immigrant Republican sponsored initiative measures has polarized nonwhite voters against Republican candidates at the same time that California is become majority minority. As a consequence, Republican statewide candidates have failed miserably in the last four years and the Democrats have built up large margins in the state legislature.


Another state that poses interesting strategic dilemmas for the parties due to mixed demographic trends is Florida. The Republicans have a solid base vote among white Protestants (e.g. the Panhandle region in the northwest that came in after the networks had declared Florida for Gore). Cubans, unlike the Mexican-Americans, also strongly favor the Republican party, especially in the wake of Elian Gonzales incident.


But growing numbers of retirees from more moderate and Democratic states in the Northeast and nonwhites in the urban areas made the state a toss-up in 2000 despite Jeb Bush being Florida’s Governor.


In sum, the 2000 election, for all its notoriety and drama, did little to change the underlying sociology of US electoral politics. As things stand presently, neither of the two major parties has a dominant position. Bush won the Electoral College vote, but lost the popular vote by a half a million votes. The 2000 Election also left the House and the Senate with very thin margins. Slightly more voters call themselves Democrats (39%) than Republicans (35%), and a growing block (26%) regard themselves as Independents. Moreover, half the electorate labels itself moderate as opposed to liberal or conservative.


In effect, the two parties, and liberals and conservatives, cancel each other out, and the critical choices are made by the independents and moderates. Of the 26% who label themselves independent, 15% are moderate independents. They split their vote 48% to 45% for Gore in 2000. Under such conditions, opposing parties with nearly equal strength and an evenly divided swing voter block, a close election was a likely outcome.

The Failure to Capitalize on Economic Conditions


Even granting the point that the parties are equally matched, the Democrats had one big advantage going into the 2000 election, namely, the thriving economy. Historically, incumbent administrations in the US have been rewarded by strong economic conditions and punished for bad ones (McKuen, Erikson and Stimson 1992; Clark and Stewart 1994; Norporth 1996; Nadeau, Niemi, Fan and Amoto 1999; Shah, Watts, Domke, Fon and Fibison 1999). Hence, there was every reason to think that Gore’s candidacy would be the beneficiary of the country’s strong economic performance under President Clinton. Based on a growth rate in real GDP of 5.3% (the fourth best economic record in the post-war period), all of the reputable political science forecasting models predicted a comfortable Gore victory. Although Gore did ultimately win the popular vote, the margin of victory was as much as six points below the predicted margin of some of the models (Wlezian 2001).


Leaving aside for the moment such important underlying epistemological questions as whether elections can be predicted at all, or whether models that rely only on structural factors like the state of the economy and omit the short-term effects of campaigns can ever succeed, it is a mystery that this election did not follow the normal historical pattern. How can this be explained? There are several possibilities.


This first is the “tactical error” theory (Holbrook 2001; Campbell 2001). The essence of this explanation is that the economy did not help Al Gore as much as it should have because he did not let it help him. Seeking to distance himself from the harmful aspects of Clinton’s legacy (e.g. Monica Lewinsky), Gore went too far and failed to remind voters of the Administration’s considerable achievements in managing the economy. He compounded this error by not allowing Clinton to campaign for him even though Clinton had demonstrated in his masterful farewell address at the Democratic Convention that he could make the case better than Gore could himself. The political science models that relate economic performance to presidential vote shares assume that candidates will run on and not away from their achievements. The failure of that seemingly uncontroversial assumption, the argument goes, accounts for the political underperformance of the economy in 2000.


The other mistake that Gore might have made was allowing Bush to get closer to the median voter with his compassionate conservatism appeal while Gore himself veered in a more populist direction. The plausibility that these kinds of tactical mistakes might have made a difference in the very close 2000 outcome highlights the weaknesses of political science forecasting models: they do not even try to account for the irrationality or incompetence of campaign decisions and hence are vulnerable to mis-prediction.


A second type of explanation can be called the “bad measures” thesis. It holds that while the theory of economic determinism is basically correct, the measures of economic performance are fatally flawed (Lewis-Beck and Tien 2001). Some recent research by Larry Bartels and John Zaller argues that instead of the percent change in GDP per capita, the proper measure of consumer anxiety is the annual growth rate of real disposable income (RDI), (Bartels and Zeller 2001). They demonstrate that while the US economy was very strong in 2000 by the GDP measure, it ranked only 8 out of the 14 post-war elections in terms of RDI. Moreover, they maintain, a slowdown in RDI in the last two quarters before the election was probably a major factor in Gore’s defeat while Gore’s weaknesses as a candidate could only account for a small amount of the underperformance.


Finally, there are various structural factors that worked against the Democrats in 2000. It was certainly problematic that Green Party candidate Ralph Nader was able to sustain a campaign from the left while Reform Party candidate Pat Buchanan’s campaign from the right failed miserably. Clearly, it would be a mistake to assume that all of Nader’s vote would have gone to Gore, but it is likely that some of it would have, and the Nader vote might have made the difference, particularly in close states like Florida. In addition, incumbent administrations lose support over time the longer they are in office (Brace and Hinckley 1991). Since the Democrats had been in office for eight years, Gore may have lost a half point or more as a result.


The final verdict on what happened to the incumbency advantage that the Democrats should have had by normal historical standards must await further careful scholarship. It is quite possible that scholars and pollsters underestimated the negative effects of the impeachment scandal on the election, because it had had so little impact on Clinton’s job ratings (Norpath 2001). For the purposes of the present discussion, the critical point is that Gore’s failure to reach the standard benchmark performance made the election much closer than it should have been.

Strategic Targeting and Soft Money


The other notable feature of the 2000 election was the way that soft money in particular has been used to target the most competitive areas of the country. This strategy heightened the overall competitiveness of the election. Since the seventies, federal elections have been regulated by a law (i.e. the 1974 amendments to the Federal Election Campaign Act) that limited contributions from parties, interest groups and individuals to candidates, and provided for public financing of Presidential elections. This system began to fray in the 1996 election, but fell apart completely in 2000. As a result, so-called “soft” money became extremely important to both parties.


In the language of modern American politics, campaign donations regulated under the Federal Election Campaign Act are called “hard” money. The limitations were fixed then and have not been adjusted since for inflation or the rising real costs of elections. As a consequence, both parties have become more dependent upon “soft” money–i.e. money that can be given in unlimited amounts to the parties for so-called party-building activities under a provision in the law. Essentially, as the limitations upon hard money have become more stringent over time because they have not been adjusted in twenty-five years, the parties and their candidates have become more dependent upon the soft money. In the 2000 election, the Republicans spent a total of $691,804,099, and 35% of that was soft money. The Democrats spent $513,059,203, of which 45% was soft money.


Soft money was important to the outcome in 2000 for several reasons. First, the fact that the money can be given to the parties in unlimited amounts undercut the effect of the hard money limitations. In other words, it pumped a lot more money into the election than there would have been otherwise. Secondly, the definition of party building activities includes so-called issue advocacy tv spots in which one side attacks the other but avoids using the magic words “elect” or “vote for”. This meant that much of it was spent on tv commercials intended to win over swing voters. Lastly, this soft money was transferred into the most competitive states rather than evenly across the board. Thus, states where Al Gore had overwhelming leads (e.g. New York or Massachusetts) received less money from the Republican party and little attention or visits from George Bush during the campaign. Similarly, states where Bush had large leads received little money or attention from the Democrats. Instead, the money and campaign efforts were focused in the battleground states like Florida, Ohio, Pennsylvania, Missouri, Washington, and the like.


In a situation of evenly balanced parties, the campaigning apparatus has optimized itself around the “persuadable voter” and the swing states. The soft money phenomenon reinforces the strategic efficiency of modern American elections. Used to bolster campaign efforts in states and areas where it matters the most, it has been shaped by the incentives of an Electoral College strategy (Shaw 1999a; 1999b). Without the Electoral College incentive, Bush’s and Gore’s campaign organizations would have focused their vote-maximizing efforts on the most populous states rather than the most competitive states per se. But since under the current system the votes that matter most are the state victories, the candidates targeted their attention on the states where the outcomes were most in doubt. By playing the Electoral College strategy as efficiently as they did in an election that was as close as the 2000 contest was, the candidates contributed to the prospects of a divergent popular and Electoral College winner.



In the end, the 2000 election will likely be remembered for its institutional legacy—ballot and campaign finance reform—rather than for any sign of realignment. Clinton was unable to translate his achievements into a more favorable electoral alignment for the Democrats, and Bush was able to capitalize on tactical errors by Al Gore and a perceived weakening of the economy to win back what his father had lost. The prospects of reforming the Electoral College are no greater than they were before 2000 since it would take a constitutional amendment and the legislatures in the smaller states have already indicated that they would refuse to go along with a change. While the polls indicate that most Americans would prefer to change to direct election by popular vote, the Electoral College victory has not seriously impeded the Bush Presidency. Given the parity of the parties and the stability of the underlying electoral sociology, one cannot rule out the possibility that the events of 2000 might repeat themselves in the near future.


  • Abramowitz, Alan. “It’s Abortion, Stupid: Policy Voting in the 1992 Presidential Election,” The Journal of Politics 57 1 (February 1995).
  • Bartels, Larry M. & John Zaller. “Presidential Vote Models: A Recount,” PS: Political Science and Politics 34 1 (March 2001).
  • Brace, Paul & Barbara Hinckley. Approval: Constraints Within and Across Presidencies,” Journal of Politics 53 4 (November 1991).
  • Campbell, James E. “The Referen-dum That Didn’t Happen: The Forecasts of the 2000 Presidential Election,” PS: Political Science and Politics 34 1 (March 2001).
  • Clarke, Harold D. & Marianne C. Stewart. “Prospections, Retrospec-tions and Rationality: the Bankers Model of Presidential Approval Reconsidered,” American Journal of Political Science 38 4 (November 1994).
  • Holbrook, Thomas M. “Forecasting the Mixed Economic Signals: A Cautionary Tale,” PS: Political Science and Politics 34 1 (March 2001).
  • Kaufmann, Karen M. & John R. Petrocik. “The Changing Politics of American Men: Understanding the Sources of the Gender Gap,” American Journal of Political Science 43 3 (July 1999).
  • Lewis-Beck, Michael S. & Charles Tien. “Modeling the Future: Lessons from the Gore Forecast,” PS: Political Science and Politics 34 1 (March 2001).
  • Mckuen, Michael B., Robert Erikson & James A. Stimson. “Peasants or Bankers? The American Electorate and the US Economy,” American Political Science Review 86 3 (September 1992).
  • Nadeau, Richard, Richard G. Niemi, David P. Fan & Timothy Amato. “Elite Economic Forecasts, Economic News, Mass Economic Judgments and Presidential Approval,” Journal of Politic 61 1 (February 1999).
  • Norpoth, Helmut. “Presidents and the Prospective Voter,” Journal of Politics 58 3 (August 1996) ; “Primary Colors: A Mixed Blessing for Al Gore,” PS: Political Science and Politics 34 1 (March 2001).
  • Shah, Dhavan V., Mark D. Watts, David Domke, David P. Fan & Michael Fibison. “News Coverage, Economic Cues, and the Public’s Presidential Preferences, 1984-1996,” Journal of Politics 61 4 (November 1999).
  • Shaw, Daron R. (a) “The Effect of TV Ads and Candidate Appearances on Statewide Presidential Votes, 1988-96,” American Political Science Review 93 2 (June 1999) ; (b) “The Methods Behind the Madness: Presidential Electoral College Strategies, 1988-96,” Journal of Politics 61 4 (November 1999).
  • Wlezien, Christopher. “On Forecasting the Presidential Vote,” PS: Political Science and Politic 34 1 (March 2001).



This point is disputed by some who claim that Kennedy actually received fewer popular votes than Nixon in 1960. See Lawrence D. Langley and Neal R. Pierce, The Electoral College Prime (New Haven: Yale UP, 1996) and Brian T. Gaines, “Popular Myths about the Popular Vote-Electoral College Splits,” PS: Political Science and Politics 34 1 (March 2001).


All polling: the figures come from exit poll data compiled by The New York Times. See Marjorie Connelly, “Who Voted: A Portrait of American Politics, 1976-2000,” The New York Times, November 12, 2000.



Analyse de sociologie électorale, cet article souligne l'importance des continuités dans les résultats des élections 2000 par rapport à celles du passé (notamment celles de 1992 et 1996). Aucun changement profond ne peut véritablement être décelé. Mais les élections de 2000 révèlent cependant quelques surprises et évolutions. Les Démocrates n'ont pas été en mesure de traduire les succès en matière économique du gouvernement Clinton en une victoire dans les urnes. Défiant les modèles traditionnels de prédiction électorale, George W. Bush a instrumentalisé le thème de l'économie à son profit. Par ailleurs, les perspectives de réforme du collège électoral et du système du « winner takes all » ne semblent pas être en bonne voie. Compte tenu de l'état actuel de quasi-parité entre les deux grands parties et de la stabilité du paysage socio-électoral, les candidats indépendants et membres des partis périphériques sont appelés à exercer une influence non négligeable lors des prochaines échéances.


  • Sociologie électorale
  • Minorités
  • Partis politiques
  • Fédéralisme
  • Géographie électorale


  • Elections
  • Political parties
  • Minorities
  • Women
  • Electoral behaviors

Plan de l'article

  1. The Failure to Capitalize on Economic Conditions
  2. Strategic Targeting and Soft Money
  3. Conclusion

Article précédent Pages 10 - 18 Article suivant
© 2010-2017