Ruy Teixeira on the elections

Michael Pollak mpollak at
Wed Nov 13 10:52:43 PST 2002

[This guy is one of the very few psephologists I really respect. Even in weeks like this, when I couldn't be more exasperated with the stupidities of polls, he manages to convince me that there's something in there iff you do it very carefully.]

Public Opinion Watch October 28-November 8, 2002

In this edition of Public Opinion Watch:

- Election 2002: What, Why, and Where (and How the Democrats Can Do


Gallup poll of 1,221 adults for CNN/USA Today, released November 4, 2002 (conducted October 31 to November 3, 2002), 2002 election results, including vote by county

Committee for the Study of the American Electorate, "Turnout Modestly Higher," November 8, 2002

Greenberg Quinlan Rosner poll of 1,992 registered voters, including 1,763 voters 2002 voters for Campaign for America's Future, released November 8, 2002 (conducted November 5-6, 2002)

"Romp and Roll," National Journal 2002 Election Issue, November 9, 2002

Public Opinion Watch says: it's time to organize our thinking about the election of 2002 (those who wish to panic may leave the room). To that end, we will go over four topics, in the following order.

- What happened? - Why did it happen? - Where did it happen? - What does it all mean, especially for future Democratic politics?

What Happened?

Consider the following numbers: Senate: -2 House: -4 Governors: +3

These are, of course, the Democratic net losses (and gains) in the 2002 election. Repeat them to yourself several times; they do not seem, on the face of it, to indicate a Republican tsunami that swept all away before it. In fact, these numbers suggest the partisan balance of the country, at least in terms of voting and public support, has changed only slightly. This is different from, say, 1994, when huge Republican gains (52 House seats, 9 Senate seats, 10 governors) really did dramatically change the partisan balance.

But, when the country is divided as evenly as it currently is, small changes can have big effects. That's just the mathematics of the situation: going from 49-51 to 51-49 in the Senate makes a big difference-much more than going from 47-49 to 49-51 or from 51-49 to 53-47. So the very evenness of partisan division in the country lends itself to sudden lurches in political power driven by only small switches in public sentiment. And that's what we had in this election.

So why are Democrats so upset-if not suicidal-about this election? One factor, of course, is that the effect of changing the Senate from razor-thin Democratic to razor-thin Republican control is greatly magnified by pre-existing Republican control of the House and the Presidency. Unified control of Congress and the Executive gives the Republicans an undeniable and very large lift in terms of implementing their agenda and Democrats are (reasonably) very worried about this.

Another factor is how the Democrats lost. In every Senate race (except South Dakota and Arkansas) that was close or deemed to have the potential to be close, the Democrats lost. They lost Missouri 50 percent to 49 percent, Minnesota 50 percent to 47, New Hampshire 51 percent to 47 percent, Colorado 51 percent to 46 percent, Georgia 53 percent to 46 percent, North Carolina 54 percent to 45 percent, and Texas 55 percent to 43 percent. That hurts, because the Democrats had expectations-or at any rate, reasonable hopes-in each and every one of these elections. No one thought, of course, that they would win them all. But they sure didn't expect to lose them all, either.

And why did they have high expectations? That gets to the third factor: the economy has been performing poorly and that should have benefitted the "out party," the Democrats. According to Greenberg Quinlan Rosner's (GQR's) post-election poll, 70 percent of voters on election day thought that economic conditions were only fair or poor. And no wonder voters thought that: the stock market has collapsed, decimating many voters' 401(k) retirement accounts, unemployment is up, the corporate sector is racked with scandals, the budget surplus is gone, economic growth is slowing, and consumer confidence is at its lowest level since 1993. Add to this the traditional advantage of the out party in midterm elections and it's no wonder the Democrats had some optimism about Tuesday's election results.

Why Did It Happen?

So, why did it happen? How did the Democrats manage to sustain these losses when as little as a week before the election it looked like they might actually gain a seat or two in the Senate, not lose control? What happened in the last week-really in the last five days-was a pro-Republican surge among likely voters, which was neatly captured in a Gallup poll released on November 4. In that poll, the Republicans had a six-point advantage on the generic Congressional question ("Who would you vote for if the election were held tomorrow?"). That reversed a three-point deficit for the Republicans that Gallup had measured ten days earlier.

What happened in the last five days, of course, was President Bush's barnstorming tour of battleground states, where he rallied the troops and excoriated the Democrats for standing in the way of his efforts to make America secure in a perilous world. In fact, if you look at Bush's standard stump speech during this tour, only the first seven paragraphs were devoted to taxes and the economy; the other twenty paragraphs and the greatest applause lines were reserved for homeland security and the struggle against terrorism.

That focus worked. It did help rally the Republican faithful and reminded swing voters about the war on terror-something they strongly supported and something they associated much more with President Bush and the Republicans than with Democrats. Indeed, the Democrats basically had ceded the issue to the Republicans with their hasty approval of the Iraq resolution in Congress, which appeared motivated by nothing more than a desire to get the issue out of the way so they could talk about prescription drugs and Social Security. And the Democrats' national security bona fides were not, shall we say, enhanced by the prolonged fight over the Homeland Security Bill, which stalled over worker protection and unionization issues.

The way the Republican identification with the war on terror boosted their vote totals was vividly illustrated in the GQR poll. In that poll, the top reason for voting, or considering voting, Republican in the election was "to support the war on terrorism and a strong military." And the second most common reason was "to support President Bush." Being preferred by thirty-nine points as the party best able to keep America strong clearly pays dividends. How successful was the Democratic effort to counter this advantage by focusing on domestic issues? Not very. In the GQR poll, the Republicans were preferred over the Democrats by three points on the economy and by one point on the budget and deficits. Since the economy was the number one issue in the election, according to most pre-election polls, this was obviously quite a problem for the Democrats.

Of course, the Democrats didn't offer much, or really any, plan on the economy. Indeed, in the GQR poll, two-thirds of voters agreed that "candidates did not set out clear positions on how to deal with the economy." Democrats assumed that clear positions on the economy weren't necessary-negative perceptions about the economy would help them, regardless of whether they actually said anything specific about it. They were wrong.

And what about prescription drugs, the issue on which so many Democratic hopes rested? It's true that voters tended to favor the Democrats as the party better able to handle the issue; the problem was that voters couldn't figure out how the two parties' current approaches to prescription drugs actually differed. The GQR poll found a plurality (43 percent) agreeing that "This year, the Republican and Democratic candidates supported a prescription drug benefit for seniors." Just 34 percent thought "the Republican and Democratic candidates disagreed about prescription drugs." This isn't exactly the product differentiation you want on your signature issue.

Where Did It Happen?

So the Democrats' domestic program -- such as it was -- proved an inadequate counterweight to a Republican surge based on national security in the final days of the campaign. Where did this hurt them the most-cause them to lose when they might perhaps have won?

As one might expect, their closest Senate losses were in closely divided "blue" states (Minnesota) or in red states where the Democrats are developing a competitive position (Missouri, New Hampshire, Colorado). Thus, at the edges of the blue coalition the Democrats were seeking to develop, the Republicans were able to hold the line and even make gains. And, of course, they picked up one state (Georgia) that is solid red and where the national security issue clearly tipped the race.

On the governors' level, the story was different. Here national security concerns were less likely to be relevant and therefore less likely to hurt the Democrats' efforts to put together a majority coalition. This is shown by the Democrats' ability to pick up ten seats, including four medium-to-large midwestern states that are critical to maintaining their coalition (Illinois, Pennsylvania, Michigan, Wisconsin) and a couple of sizable states (Tennessee and, especially, Arizona) where the Democrats can reasonably hope to be competitive in the future. In contrast, the Republicans did pick up seven seats, but only one of them can be considered a critical part of their coalition (Georgia) and only one was in a swing state (tiny New Hampshire).

Looking inside some of the Senate states the Democrats lost, preliminary analysis suggests that a couple of things were going on. First, while turnout across the United States was up from 1998 (from 37.6 percent to 39.3 percent of the voting-age population, according to the Committee for the Study of the American Electorate), it appears to be the case that in strongly Democratic large cities the increase in turnout was less than in strongly Republican areas in cities' outer suburbs or in rural areas. For example, in Missouri, the increase in votes cast over 1998 is less in the strongly Democratic city of St. Louis than in the strongly Republican suburb of St. Charles county or, especially, in rural and fervently Republican Cape Girardeau county. The same pattern can be seen in Minnesota, where many of the more rural counties cast almost as many votes in 2002 as in 2000, while the more urban counties lagged behind.

But the other part of the story, and perhaps more important, was the reduction of Democratic support in Democratic-leaning suburban or mixed suburban-urban counties, where Republican success in picking up swing voters was likely to manifest itself. For example, in St. Louis county in Missouri, Jean Carnahan's margin over Jim Talent was only three points, down from the eight points her late husband carried the county by in 2000. And in Hennepin and Ramsay counties in Minnesota, Walter Mondale's margins over Jim Coleman were substantially less than Mark Dayton's over Rod Grams in 2000 (eleven points and ten points less, respectively). But in completely urban St. Louis city in Missouri, the Democratic margin was much the same as in 2000.

So, it seems likely that a failure of core Democratic areas to match turnout increases in heavily Republican areas plus a shift away from the Democrats in Democratic-leaning suburbs were the factors responsible for many of the Democrats' key losses. Together they meant that the Democrats could not prevent highly mobilized Republican areas, which both turned out at high rates and supported Republicans at high levels, from dominating these electoral contests.

What Does It All Mean?

The logic of the analysis presented here suggests several thing the Democrats need to pay attention to:

- Democrats need a national security policy that is a plausible alternative to the Republicans. Voters want to know what the Democrats are going to do to make them feel safer; so far the Democrats have really had no answer other than to support the president and occasionally mutter criticisms of how he's going about things. Moving forward to 2004, they will need to develop a more imaginative approach or Republicans will continue to dominate the issue, with predictable negative political effects.

- They need a domestic policy agenda that goes beyond prescription drugs and defending Social Security. That starts with a clear approach to improving the economy and should probably also include new programs in education, where the Democrats opened up a wide lead on the issue in the months prior to the election, but then had no advantage at all among voters on election day. Call it E2-and then throw in the environment and energy, where they continue to enjoy large advantages, to make it E4. E4 has got to be better than prescription drugs, in particular, which seems to do more to confuse voters than rally them to the Democrats.

- Finally, avoid an unproductive debate on mobilization versus winning over swing voters. Both are clearly important and a plausible national security program combined with an imaginative domestic agenda would go far toward accomplishing both goals.


Ruy Teixeira is a Senior Fellow at The Century Foundation.

Public Opinion Watch covers newly released polls, as well as key newspaper and magazine articles that make use of polling data. If you've ever wondered what to make of the blizzard of survey data covered in the newspapers--and whether the newspapers themselves know what they're talking about--you'll want to check out this feature on a regular basis. Each edition will combine noteworthy findings and trends from the latest polling data with analysis of the misinterpretations and misrepresentations to which polling data are so often subject.


If at anytime you wish to unsubscribe from a list please email list at with Unjoin and the NAME OF THE LIST[Public Opinion] from which you wish to be removed.

Thank you for being a member of The Century Foundation's Public Opinion Watch listserv.

** Please visit The Century Foundation on the web at**

More information about the lbo-talk mailing list