Democracy and the National Vote

Similar to Circadian, the moon, and sunspots, U.S. elections occur on a regular cycle.  For those in the U.S. this is a natural part of a democracy.  However, in other nations, national elections are not necessarily regularly scheduled.  In Britain, for example, the prime minister requests the Queen to call elections at a time the P.M. believes is most political advantageous.  It cannot be longer than five years, but it may be much shorter.  A no confidence vote in the Parliament mandates immediate elections.

Another interesting aspect about our process is that the U.S. considers itself the premium democracy that other counties are encouraged to emulate.  However, our presidential election does not meet the test of a majority driven, democratic process.  For example, the nomination process now involves a series of state primaries and caucuses, each with different rules, which replaced the national caucus system of the early 1800s. In recent elections, small states with early primaries and caucuses, like Iowa and New Hampshire, have a disproportionate influence over the process.  Howard Dean’s loss in Iowa doomed his chances.  Knowing the importance of the early state selection, states, including Wisconsin, have moved their primaries earlier in the year—a process that is known as “front loading.”  Attempts to create a more systematic national primary or regional primary day have failed.

Once the candidates are selected, our electoral process comprises 50 state elections plus D.C. rather than one national election.   In each state but two, the candidate that wins the state wins all the electoral votes of that state.  The consequence is that strategically, candidates spend more time in competitive states than in states that are less competitive even if these states have many more voters.   The consequence is that the voters in the “battleground” states are more important than the significantly larger number of voters in the large, but less competitive states.  The ultimate result could be that the candidate wins the Electoral College vote, where each state has the votes equivalent to the number in their congressional delegation, but loses the national popular vote total, which of course is what happened in 2000.  Even without the Florida debacle, the legitimacy of a president selected who did not win a majority is shaken. 

Some historians have written that the founders thought that the Electoral College system would be eventually replaced and expected that after George Washington no candidate would receive a majority.  The consequence would be that the House of Representatives, where each state would have one vote, would be the forum for selecting the president (Roche, 1961).  Despite the views of the founders, we still have the Electoral College and the House has rarely been involved. 

Beyond the process, questionable vote counting has challenged democracy in the U.S.  On the presidential level, the impact of Mayor Daley on the vote in Chicago in 1960 is legendary.  More recently, counting votes in Florida with hanging chads does not represent a mature democracy.  After the 2000 election, Congress enacted a law granting funds to states to improve their voting process. (Much of the money has been delayed in being allocated.) However, the touch screen process, adopted by many communities, has been challenged for accuracy, its inability to create a backup paper record, and its vulnerability to hacking.  For the first time in history, the U.S has invited international observers to monitor the election. 

Public Opinion Polls and Forecasting the Election 

The science of public opinion polling over the last half century has developed with polls becoming more accurate.   It is a long time since a poll by the Literary Digest inaccurately predicted Landon’s win over Roosevelt in 1936.  It is intuitively difficult to accept how the opinions of a small group of people can accurately reflect the larger group, a relationship which has been mathematically verified. But results have shown that polls are accurate.  Elections are more difficult for polling than general opinions because we have to add an additional unknown—who is going to vote.  Thus polls differ a bit on the extraction from the data of “likely voters.”  Further, in presidential elections the polls have problems because we really do not have a national election but 51 separate elections.  Therefore the proper polls would be ones taken in each state rather than a national poll.   States vary in the quality of the reported state-level polls. 

Regardless of increasing accuracy of public opinion polling, the presidential polls are having a harder time.  For one, the response rate has dropped.  The fundamental principle of polls is that everyone has an equal chance of being in the sample.  The more people who refuse to be questioned, the less accurate the survey.  The response rate fall off has occurred because of the proliferation of telemarketers, who are hated on the level of HMOs; the increased number of polls; the increasing concern with privacy; and the use of answering machines and caller ID to screen calls, although the latter two have had much less affect according to research.  In addition, the development of more political campaigns using push polling, whose aim is to promote candidates has reduced the receptivity of people to legitimate polling efforts.   The Pew Research Center found that interviews were completed in 38% of sampled households this year versus 58% in 1997 (Pew Research Center 2004). 

Technological developments have impacted poll accuracy that rely upon land lines.  An increasing number of people (although still representing a small percent) have only a cell phone.  At this time, polls do not reach cell phones.  Pollsters are conferring about what to do as this trend escalates.  Today, it may be significant in swing states. 

Lastly, close elections, which this is predicted to be, presents a number of problems.  For one, polls only predict within a confidence range.  The actual results may fall within this range, but be reversed.  Further, close elections typically increase turnout.  It is possible that someone who has not voted in recent elections will now vote.  Poll results using likely voters as their base may have problems because typically past voting turnout is among the principal criteria in extracting likely voters from the results.  

Despite the problems for polls for this year’s election, the variability of poll results in this election is small, leading to more confidence in the results.   Keeping in mind the caveats expressed above, the discussion in this paper relies upon various polls that have been considered the most creditable.  

Kerry versus Bush:  Issues and Strength 

There are many potential issues that will impact this election.  Based upon a number of polls, it is clear that Iraq, terrorism/homeland security, the economy (especially jobs and unemployment) and health care (including Medicare and Prescription Drugs) are the leading issues that voters say will influence their vote.   The NBC/ Wall Street Journal poll’s results below are similar to others. 

 As for trends for the three top issues, President Bush is losing support.  This is especially evident in confidence in him over Kerry to provide homeland security, where the President had been consistently ahead.  With the jobs picture showing signs of weakening in August, Kerry has been able to expand his lead on this issue. 

Polarization of the Electorate 

Many commentators have noted that there are fewer undecided voters at this time of the election year than typical.   In the swing states, the task for the candidates is to appeal to the small segment of the decided voters, which is crucial given how close the election appears to be, and most significantly mobilize their base of support.  It is not only how many people support the candidate, but most importantly how many supporters turn out to vote.  Traditionally, the Republicans have had an advantage here.  Differential turnout is related to the socio-economic status of the supporters with the Republicans, having a larger number of upper income and more educated voters.  Kerry supporters will have to work hard to get groups who traditionally vote Democratic, such as African-Americans, out to vote. 

Is the electorate more polarized than in the past?  One way of examining this is through the strength of party affiliation.  Since the 1950s, attachment to the parties has weakened as the graph below shows.  This has allowed candidates to gain votes from those who identify with the opposite party. However, beginning in the 1980s, we see a trend where party identification has again strengthened, although still not to the levels of the 1950s, but certainly stronger than in the 70s, illustrating polarization. Along with this change have been demographic changes in party attachment.  The most notable change is the gender gap between the parties, especially evident on racial lines.  White women are more likely to be Democratic while white males Republican.   With this change, there has been some disconnect between economic status and party affiliation.  Some analysts see a significant dose of social issues as differentiating the parties.

Another polarization view is how party members evaluate the candidates on various issues.  Issue differences by party are not atypical, but in this election, surveys have shown that it is even stronger than usual.  For example, disapproval of the President’s Iraq policy is much greater among Democrats than Republicans and events, which have not been too favorable to the President’s policy, have moved evaluations very slightly.  On questions related to the voters perceptions of the candidates (asked by the Annenberg Public Policy Center) such as which candidate is more optimistic, cares about people like me, inspiring, a stronger leader, trustworthy, and more likely to share my values, among others, the results show polar opposite evaluations by Democrats and Republicans.  Most of the responses show 80% favoring their own party’s candidate.

Ideological division could also lead to voter polarization.  Is the electorate divided more in conservative and liberal camps than in the moderate category?  The data from the University of Michigan National Election Survey shows a drop in moderates (moderates or slight liberal or slightly conservative) by 10% in 2002 in contrast to 1972.  Rather than half the electorate considering themselves in the middle, now 40% do, a significant percent but less than in the earlier period.  In addition, 22% in 2002 said they had not thought about their ideological position.  If we add these to the moderate category, a substantial portion of the electorate is in neither the conservative or liberal camps., which counters the polarization view.  The significant ideological change has been with those who consider themselves conservatives (more than slightly conservative), which went from 11% to 25%.  However, a presidential candidate cannot rely solely on their ideological group to get elected and therefore must make a play for the center.  Clearly, neither Goldwater nor McGovern was successful at doing this.  In 1976, it was argued that Dole was also not able to get back to the center after the primaries.   George W. Bush, who has taken a number of conservative stands, may similarly have difficulty appealing to this moderate group.  Social policy positions, such as pressing for a Constitutional Amendment opposing gay marriage, have alienated the Log Cabin Republicans, a gay Republican organization, who chose not to endorse the President for reelection.  In 2000 they claim, one million gay individuals voted for Bush most of who will not in 2004 because of his position on the amendment.  Prior Republican candidates had been careful not to alienate this group.  A similar conservative position on stem cell research may reduce his appeal to moderates. 

The Pew Research Center for the People and the Press noted that the polarization also can be seen in the change in news viewing habits.  Fox News has seen a significant increase in audience from 17% to 25% of those who regularly watch the news.  Their audience is more conservative (52% describe themselves as conservative) and more Republican than the other networks with an increase in this imbalance.  CNN, the first all news TV network, has had a shift in its viewers, according to Pew research, with its audience being composed of a larger percent Democrats than in earlier periods. 

One aspect of polarization is clearly evident.  With reapportionment, parties have drawn safe districts that are very homogeneous for one party or the other.   The consequence is that Congressional districts and state legislative districts, especially in Wisconsin, have little competition.  Individuals elected from such districts are far less likely to compromise with the consequence that lawmaking is paralyzed by party strife.  In Wisconsin, one-third of the Assembly seats are not contested, another seven have only primary competition, and in another three, there is only third party competition.  Four of the Senate races have no main party competition.  Nationally, competitive Congressional districts are rare. 

Demographic Division Between the Parties 

As the following chart shows, based upon polls by the Pew Research Center, the Democrats now lead the Republicans 33% to 29% in party affiliation, the same as in 2000. After 9/11, Democratic identifiers had decline, but in recent years they have recovered.  If you add independents who say they lean to one party or the other, Democrats expanded their lead to 41% to 47%. 

The Democratic advantage among women is evident with 37% of women declaring a Democratic affiliation with 27% being Republicans.  For men, Republicans slightly outnumber Democrats.   In terms of religious affiliation, the greatest difference between parties is that White Evangelical Protestants are more likely to be Republican while Jews, Democratic.  Other religious groups break fairly closely.  According to the Pew data, whites are 5% more likely to be Republican and Blacks, significantly Democratic.  This is not surprising.  What is interesting and significant for the election is Hispanics because their party affiliation has fluctuated and many are located in key states, especially Florida.  The Pew Poll shows 40% of Hispanics considering themselves Democrats with 20% being Republican.  In ideological terms, liberals are overwhelming Democratic with conservatives being Republican, reinforcing a polarized view of the electorate.  Those considering themselves moderates are more likely to be Democrats than Republicans, a division that could significantly impact the election if Bush is seen as too conservative. 

Of the religious groups, there has been interest in Catholic support because John Kerry is a Catholic, the first to run in the general election for President since John Kennedy.  Moral issues raised related to abortion, stem cell research, and gay marriage also may influence Catholic voters.   The Gallop Poll survey, taken at the end of July, shows Kerry leading Bush among Catholics 54% to 41%.  Most interesting is that Catholics who attend church more regularly are more likely to vote for Bush.   Catholics attending every week support Bush by 52% to 42%.  However, this lead reverses for Catholic attending fairly regularly (nearly every week or monthly) who support Kerry by 50% to 42%.  Infrequent church attendees overwhelmingly support Kerry by a majority of 57% to 37%.  This percent is also seen among Hispanics, who are mainly Catholic, and support Kerry over Bush by 58% to 38%.  Interestingly for Hispanics, extent of church attendance does not impact their intended vote. 

Jewish voters have been significantly Democratic since the New Deal.  However, there had been an erosion of Democratic support during the Reagan years.  Clinton regained the Jewish vote for the party.  Bush had low ratings in 2000.  In an August poll of Jewish voters, released by the National Jewish Democratic Council, Senator Kerry is favored by 75%to 25% over Bush.  This poll surveyed voters using the web (although it did use a random sample ) so there may be some skew because it was received only by those who have computers.   However the results were compatible with other polls taken.

Source: The Pew Research Center for the People and the Press

Regional Support of the Candidates 

It is expected that Kerry will do well in the northeastern and West Coast states.  The Midwest states have been considered the swing states, where the greatest focus of the campaigns has been directed with the bulk of the political advertising aired in these states.  Wisconsin, a swing state, has seen extensive advertising while voters in less competitive states have viewed comparatively few (Wisconsin Advertising Project 2004).  Kerry may have some advantage in the swing states in comparison to Gore in 2000 because many of these states are also the ones that have lost jobs in recent years and the job issue is more of a concern than in 2000.  Although many economic indicators show a recovery, job creation has been a significant weak spot.   

The South, which at one time was considered a solid Democratic region because of the politics of race, has moved in recent decades into the Republican fold.  It is here that President Bush, focusing on issues such as faith based government service provision, gay marriage, and gun control, hopes to mobilize his supporters to keep the states in his column.  In recent times, only Democrats from the South such as Johnson, Carter, and Clinton have done well.  Gore was not able to take advantage of this, losing even his home state of Tennessee.  However, Edwards is more of a populist candidate.  Whether a Southern vice president will be able to overcome the view of Kerry as a liberal from the North as Johnson did for Kennedy in 1960 remains to be seen.  The Hispanic presence in the South could aid Kerry because recent Bush Administration travel restrictions to Cuba have angered Hispanics as did the return to Cuba of a child who had left with his mother who died escaping Cuba brought unfavorable views of the Clinton Administration, harming Gore’s election chances. Virginia, last supporting a Democratic presidential candidate in 1964, has changed with large population increases in the more competitive northern D.C. suburbs, the election of a Democratic governor, and the large number of veterans to whom Kerry has appealed (Benenson 2004: 1446). 

The Southwest, typically Republican, may be more favorable to Democrats.  Gore won only New Mexico by about 300 votes.  Demographic changes with an increased Hispanic population, such as in Colorado, and liberals moving from the West Coast may make this region competitive.  Booming resorts have attracted a number of low income workers.  With population growth recorded in the 2000 Census, these states have gained importance as they have increased their number of votes in the Electoral College (Schlesinger and Jordan 2004: A1). 

The Nader Factor 

Given the closeness of the 2000 election and Ralph Nader’s role in some close states, his impact on the 2004 election could be significant.  In Florida, Nader received only 1.6% of the vote, but Bush’s recorded margin was only .01%.  In Wisconsin Nader received 3.6% of the vote.   Gore did win the state but with the slim margin of .2%.  In New Hampshire if most of Nader’s vote went to Gore, Gore would have won the state.   

Nader could still play an important role in swing states.   His overall impact will be less because he did not receive the Green Party’s nomination as he did in 2000.  As an independent, he will be on fewer ballots.  Republicans have worked in several states to secure the needed signatures to get Nader on the ballot, which they see as a way to divide the liberal vote.  Given what some considered his “spoiler” role in 2000 and his statement that he would not seek the Green Party’s nomination (although he had hoped for their endorsement), some of his former supporters have urged him to withdraw this time for fear that his candidacy may aid the President’s reelection (Wilfong 2004: 1452). 

Nader’s 2.7% of the vote in 2000 trailed many third party candidates in the past.  Former President Theodore Roosevelt, running as a third party candidate in 1912, received 27% of the votes and carried seven states.  Wisconsin’s La Follette in 1912 won 13 electoral votes with nearly 17% of the vote.   More recently, George Wallace in his best race in 1968 won five states with 10 million votes, possibly tipping a close election to Richard Nixon.  Ross Perot similarly was a factor in Clinton’s win in 1992, having gained 19 % of the vote, much of it may have gone to George H.W. Bush.  Therefore, given the projected close margin, it is possible that Nader could be a factor.  However, if he fails to make the ballot in may swing states, his role will be diminished. 

In conclusion, the election is projected to be one of the closest in history. Kerry has benefited from opposition to Bush’s policies, but has had difficulty in defining himself to the voters.  Attempting to let voters know more about Kerry, personally and politically, was the theme of the Democratic National Convention.  A pro defense posture was emphasized but one that would be better executed than the current administration.  Socially, the emphasis was on policies that would not tilt to the wealthy as they claim is the result of Bush policies and would help the middle and working classes within the framework of a balanced budget (Cochran 2004).  With perceived significant differences in policy in many areas, the race’s outcome will be significant to the direction the U.S. takes in social, economic, and international policies in the next four years.

Sources Cited

Benenson, Bob. 2004. “Playing the Long-Shot States.” Congressional Quarterly Weekly.       June 19.  

Cochran, John. “Democrats Polish Message, Renew Push for Swing Voters.” 2004

Congressional Quarterly Weekly.  July 31. 

National Election Studies. 2004. University of Michigan. http://www.umich.edu/~nes/nesguide/nesguide.htm 

Pew Center for the People and the Press. 2004. http://people-press.org/ 

Roche, John P. 1961.  “The Founding Fathers: A Reform Caucus in Action.” American Political Science Review.  LV:799-816. 

Schlesinger, Jacob and Miriam Jordan. 2004. “Population Shifts in West Shape Kerry’s Strategy.”  The Wall Street Journal.  July 21.  

Wilfong, Catherine.  2004.  “Nader’s Great Green Divide.”  Congressional Quarterly Weekly.  June 19. 

Wisconsin Advertising Project. 2004. http://polisci.wisc.edu/tvadvertising/