Committee for the Study
of the American Electorate
Post-Election News Release

November 8, 2002

NOTES: The folloing post-election news release from the Committee for The Study of the American Electorate will be updated by the CSAE after there are more complete figures for states, notably in the west.

Beginning with this report and in all ensuing reports CSAE will be using a different denominator for the analysis of turnout. This denominator is more accurate because it eliminates non-citizens which were included in the old denominator. A full explanation of why this denominator is being used and why it is preferable to others which have been suggested is included in Note 2 in the notes section of this report. CSAE will for some time be providing charts based on the old Voting Age Population and the new one in its reports.

FOR RELEASE: November 8, 2002

IMMEDIATE

FOR FURTHER INFORMATION, CONTACT:
Curtis Gans
(202) 546-3221 (703) 478-1943

TURNOUT MODESTLY HIGHER
DEMOCRATS IN DEEP DOO-DOO
MANY QUESTIONS EMERGE

WASHINGTON, November 8 -- Fueled by grass-roots mobilization efforts in some hotly contested races, voter turnout increased modestly and likely temporarily in the 2002 mid-term election.

When final and official results are completed early next month, it is likely that an estimated 78.7 million citizens cast their ballots in the 2002 mid-term. Turnout, when the numbers finally are certified, will be 39.3 percent of eligible citizens, up modestly (1.7 percentage points) from the 37.6 percent who voted in 1998. About 121 million citizens did not cast ballots. Turnout went up in 31 states, but declined in 19 states and the District of Columbia, including some in some hotly contested statewide races. Only one state, Florida, had a record high turnout. Seven states set record lows.

Democratic turnout declined. Republican turnout increased modestly.

These are among the findings of a report on voter turnout and registration released today by the Committee for the Study of the American Electorate, a non-partisan, non-profit research organization based in Washington. The report is based on vote tabulation figures provided by the Associated Press at the close of its running of those figures Wednesday afternoon, November 6 and registration figures from 34 states and the District of Columbia which are final and official and provided by the chief election officer in each state. The vote tabulation figures are likely to be close to the final certified results, except in the three west coast states who were, at the time of this release, still counting substantial numbers of absentee and mail ballots and Texas whose tabulated results were only 90 percent complete.

The Turnout Story:

Among the other findings in this report:

The largest turnout increases were largely concentrated in states with high-profile close contests and where the candidates, parties and interest groups put unprecedented resources into the campaign, in general, and greater resources than in recent elections into grass-roots get-out-the-vote efforts. The greatest increases occurred in Tennessee (up 15.0 percentage points), which did not have a statewide election of consequence in 1998 but had high profile races for both Senator and governor this year, followed by New Hampshire (12.1), South Dakota (12.0) and Louisiana (8.0). Minnesota had the highest turnout (61.4 percent of eligible citizens), followed by South Dakota (61.3), Maine (50.6) and Vermont (50.0).

Turnout also declined in many states which had tight and expensive races. Turnout declined in California (an estimated 8.0 percentage points), Alaska (an estimated 6.5 percentage points), Colorado (5.6), Hawaii (4.5), Arizona (3.0), Wyoming (2.2), Wisconsin (1.2) and South Carolina (1.1) all of which had at least one race whose likely outcome was not clear until no earlier than the final week of the campaign.

Only one state, Florida at 42.9 percent of citizen-eligibles voting, set a new record high. Fully seven states California (at an estimated 31.5 percent of citizen-eligibles voting), Arizona (27.1), Colorado (40.2), Indiana (34.1), Montana (49.5), Nebraska (38.1) and New Mexico (39.4) set record lows.

The Republicans clearly outorganized the Democrats. Democratic turnout was down 1.3 percentage points or 7.7 percent from 16.4 percent in 1998 to 15.1 this year on the basis of aggregate House of Representatives vote. Republican turnout was up 0.5 percentage points in House races to 17.2 percent of citizen-eligibles in 2002 from 16.7 in 1998. (The Democratic deficit will diminish slightly and Republican gains will increase when final and official results are in).

Based on aggregate House turnout, the Democrats drew fewer voters than in 1998 in every region except the South. When final results are in, the Republicans will likely have gained in the South, New England, and the Middle Atlantic states; held their own in the farm and industrial midwest and lost only in the west.

The only regions in the nation in which the Democrats outpolled the Republicans were the far west and New England.

This election marks the third straight mid-term election (beginning in 1994) where the GOP has outpolled the Democrats in aggregate House of Representatives vote. Prior to 1994, the last time the GOP had done that was 1946.

While turnout was modestly up, overall turnout, based on the votes for the race or races receiving the highest number of votes is down more than 20 percent (10 percentage points since its apex of 49.3 in 1966). Democratic turnout is down 30.8 percent (8 percentage points) from its apex of 26.2 in 1962. Republican turnout is down 23.9 percent (6.3 percentage points) from is 26.4 percent apex in 1966.

(NOTE: Normally in its reports CSAE lists the various high and low levels for overall and partisan turnout and increases and decreases in these categories between the last election and this. But because there are substantial numbers of votes still to be counted, as of this writing, particularly in the west, these rankings are tentative and subject to change. In Section IV of this report, the charts on turnout based on highest state vote, there are rankings and differences which may be used with caution. Similarly in Section VI on House races, there are regional and partisan analyses that may be used. Normally also in its reports, CSAE’Äôs director, Curtis Gans, makes quotation marked commentaries throughout the report. In this report, these are saved for the analysis section below.)

The Registration Story:

Final registration figures from 34 states and the District of Columbia essentially corroborate the registration story CSAE released last week:

Registration is likely, when all numbers are in, to be down slightly from 64.6 percent of Voting Age Population to 64.1 percent this year. About 136.5 million Americans will likely have registered.

Democratic registration is down 1.8 percentage points to 30.4 percent of VAP in the 18 states and the District of Columbia which require partisan registration and have reported as of this writing, marking the ninth straight mid-term election in which Democratic registration has fallen. Democratic registration is down 34.5 percent (16 percentage points) from its 47.4 percent high in 1966 and has fallen in every region of the nation except New England where it has risen.

Republican registration held about constant between 1998 and 2002 (22.6 percent of VAP in 2002 in these states and 22.8 in 1998). Since the mid-1960's, however, GOP registration has increased nearly two-fold in the South while declining about 25 percent in the rest of the country.

Registration for third parties and independents has increased in these states by 1.6 percentage points from 15.8 in 1998 to 17.4 in 2002.

This also marks the ninth consecutive mid-term election in which registering for neither major party has increased. The increase has been more than four-fold since the 1960's.

(Note: As this went to press CSAE received registration figures for one other state, listed in Section VII of this report which changes the absolute numbers above, as they will change with every succeeding state report, but the trend lines are not likely to.)

Analysis:

There are two major points of analysis that fall within CSAE’Äôs bailiwick about turnout and about the political parties. A few words about each.

Turnout:

It would be nice to be able to say that with modest turnout increases in both 2000 and 2002 that the nation had turned the corner on turnout decline. It would also be wrong.

There is a glass half-full and half-empty story to be told about both elections. The half-full story is simple: there was a modest turnout increase in both 2002 and 2000. That increase was, in large part, caused by something that had almost been seen as extinct grassroots mobilizing and get-out-the vote activity in key states. And while the budgets for such activity in both parties in both elections did not come close to rivalling the moneys poured into political advertising, any commitment to personal contact activity is a welcome change.

But turnout is still more than 20 percent lower than it was in the 1960's and more than 25 percent lower, if one takes out the south which has experienced a rise in turnout because of the 1965 Voting Rights Act, the enfranchisement of African-Americans and real two-party competition. And, in this election, the turnout decline in 40 percent of the states, including many with highly competitive races, does not speak to a healthy return of civic and political vitality.

Turnout was slightly up in these two elections because of the extraordinary circumstances surrounding both the extraordinarily close (and advertised in poll after poll as so) 2000 Presidential election and the similarly close contest for control of both houses of Congress this year, both of which mandated extraordinary tactical grassroots measures for both parties in both elections.

But the underlying fact remains that the electorate is, by a more than 60 percent majority in mid-term elections and by nearly 50 percent in Presidential elections, largely disengaged from politics and that even in this election, as seen in the turnout figures for many states, that percentage is growing.

Many of the factors contributing to this are long-term the decline in trust in public officials fostered by their own high-profile conduct; the decreased coverage and increased cynicism in that coverage by major television outlets; the decline in the quality of education and the quantity of civic education within it; the fragmenting and atomizing effects of suburbanization, decline of community, single issue and identity politics and, most profoundly, by television, cable, satellite and the Internet; the erosion in strength of the major integrating institutions schools, churches, unions and political parties; and the move to self-seeking, anti-goverment, libertarian and consumerist values, to name but a few.

But at least one additional factor was evident in this campaign--the execrable conduct of our campaigns through the overwhelming use of attack and comparative advertising. One cannot explain why tight campaigns in South Carolina (two), Arizona, Colorado, Wyoming and Alaska would depress turnout without looking at the quality of those campaigns on the airwaves. One cannot explain why the close Senatorial race in Colorado not only depressed turnout in that state but ran behind the more one-sided and less negative campaign for governor unless one looks at the volume and virulence of the advertising campaign. Charles Krauthammer once wrote that American Airlines does not advertise on behalf of itself on the basis that Delta kills more people in crashes. Politics is the only industry that regularly and overwhelmingly denigrates its product. And by so doing, undermines real dialogue about issues, erodes respect for all leadership and drives citizens to the sidelines.

There will be no permanent reversal of the four-decade turnout decline unless and until we address the larger causal issues and also cease being one of the only democracies in the world which does not address the issue of political advertising as a matter of public policy.

Parties:

There are two stories from this year’Äôs figures one relating to the two-party system as a whole, the other, to the Democratic Party in particular.

For the better part of a decade CSAE has been highlighting in its reports the decaying state of the Democratic Party the continuing disinclination of citizens to identify themselves through registration with the party, the decline in the percentage of citizens who show up in the party’Äôs primaries, the equal erosion in their share of the Congressional vote, the loss of offices on every level in the nation. This is the year that the chickens came to roost, at least temporarily.

In the first election since the mid-1930s, a sitting President and his party, in the midst of an economic downturn, actually gained seats in the Senate, the House of Representatives, the state legislatures and avoided losing the majority of governorships the GOP has enjoyed since the anti-Clinton election of 1994.

This is a direct result of the nearly four-decade failure of the Democratic Party to put forward ideas, programs and messages commensurate with a major party seeking to govern the most powerful nation on earth. For two decades after the 1960's, their dominant advocacy was identity politics (advocating the rights of inclusion of specific demographic sub-groups in American society) and when that failed spectacularly in 1984, they have propelled themselves sideways and backwards by consultant and poll, by the ’Äúsmart’Äù politics of positioning themselves just to the left of the rightward drift of the Republican party but without a coherent, affirmative and durable outlook. They have essentially been ’Äúnot Rapoport.’Äù

But their lack of idealism has lost them the young, whose allegiance they had enjoyed from the 1930s through the early 1970s. Their lack of evident and continuing commitment to the concerns of those at the bottom of the income scale has led to the withdrawal of the poor from the political process. Minorities now feel taken for granted and the one thing that unites Democrats and differentiates them from their opposition the belief in the ability of popular government to act in the interests of the general welfare is almost totally absent from their advocacy.

Back in 1998, CSAE’Äôs director coined the term ’Äúa Seinfeld election, an election about nothing’Äù to describe that election. Several commentators have used that term with respect to this election. But they are wrong. The GOP, in many ways, defined this election. They hid their major weakness the uncertain and perhaps eroding economy and played to their strengths the battle against terrorism, potential war with Iraq, mainstream American values and the popularity of the President. They recruited, in many crucial elections, mainstream candidates and tempered their more extreme advocacy. Against which the Democrats offered little or nothing. They tried to make the economy an issue, but offered no program on it or, largely, anything else, except issues which appealed to seniors. As a result the GOP was able to mobilize its troops in large numbers and the Democrats were not. And, at least to a limited extent in a still divided country, the Republicans have a modest mandate.

This leaves the Democrats in a difficult position. For, if they proceed as they have been proceeding over the last few years offering modest programs and fighting only on the excesses of Republican policy and advocacy but letting things go through like the upper class parts of the President’Äôs tax cut they will likely again be perceived as the party without message. If they define a program and perspective and fight vigorously for it even from their minority position, they may be characterized as the obstructionists that produced a do-nothing Congress. But after a period in which their very lack of definition has cost them, the Democrats might choose the bolder course. The adage is you can’Äôt beat something with nothing and this year the Democrats proved it.

Which is not to say that the GOP is without its challenges. They will be seen as solely responsible for the next two years of governance and what transpires during that period. They face three specific dangers the ’ÄúGingrich factor,’Äù foreign policy hubris and the potential for a decaying economy. In 1994 when the Republicans were catapulted to power because of strong anti-Clinton feeling, particularly in the South, the new Speaker of the House, Newt Gingrich interpreted the GOP landslide of that year as a positive mandate for things erosion of environmental protection, action to curtail reproductive rights, among many others which were wildly unpopular and led to a substantial reversal in the GOP fortunes in 1996. That could happen again unless the Republicans are careful about what aspects of their agenda they choose to put forward. The newly minted muscular and interventionist approach of the Bush Administration to international affairs could both isolate the United States from the world community and strain domestic resources. And the economy’Äôs problems, left unaddressed, could bring down this Administration despite its present popularity.

Which is also to say that the key figure in this report, particularly with respect to registration, is the growing number of citizens who, through their registration, are saying ’Äúa pox on both your houses.’Äù This ever-growing group of citizens is not presently a coherent force, but should the two major parties continue to act as they have during large parts of the last decade and given appropriate leadership, the realignment that could occur might be active rather than the growing number of Americans sitting on the sidelines.

Two minor points:

1. Exit Polls:

It is tempting to say that the total failure of Voter News Service to provide exit poll information in this election couldn’Äôt have happened to a better bunch of people. Their past arrogance surpasseth understanding.

But, this election is the first in many years in which we have virtually no information on which demographic groups turned out and for whom and which party, because exit polls were the primary source of that information. And we don’Äôt have any information about why people voted the way they did.

The problem for many years has been the abuse of these very useful devices in order to declare winners as quickly as possible. In close elections like that of 2000, it leads to sufficient error that not only is the public getting bad information but it can actually affect, as it did in 2000, voting decisions during the election and political perceptions after. In one-sided elections, it leads to national declarations while citizens are still voting depressing the late vote and the for those who vote late, the will to vote; according to the large preponderance of evidence, possibly affecting the outcomes of elections below the level of the one projected and undermining the integrity of the process.

The networks did not do very well in their coverage of this year’Äôs election. They didn’Äôt issue any wrong proclamations about winners. But they also didn’Äôt report the real news of the night--the tabulated vote count--and thus their broadcasts lacked any coherence, drama or real reportage. But, if they went back to reporting that news and not their polling contrivances for declaring winners the actual vote count comes in at a sufficiently glacial pace so that they would never get it wrong and they would never be able to declare winners while citizens are still voting. And, those of us in the research community would be able to fully defend exit polls for the useful information they provide about the shape of the electorate and the citizens’Äô reasons for voting the way they did.

2. The Perils of Prognostication:

Each year at its end my friend, David Broder, writes a column about his errors the previous year. Last week, I produced a report which did not have any factual errors, but should suggest that immediate prognostication in this election was not my strong suit. I said correctly that turnout would be within two percentage points up or down of what it was in 1998. But then I went on to say how an increase in turnout would benefit the Democrats; Wrong! How they would likely be smiling more with respect to the Senate and governorships; Wrong! And finally how, while Minnesota would likely have a high turnout, it would not be higher than the Ventura race four years ago; Wrong! Not bad for one paragraph in one release.