Jay Cost: Is 1994 a Model for Democratic Success?

Roundup: Historians' Take

In recent weeks, the talk among pundits about the inevitability of Democratic triumph has simmered down. For some reason - one that I have not yet ascertained - Bush in the high-30s induces an entirely different storyline than Bush in the mid-30s. While the logic does not make sense, the result is nonetheless a move in the right direction. We seem to have returned to the much more sensible conversation that we were having in March: for the Democrats to win the House, they will actually have to do something.

What do they have to do? This is the $2.3 trillion question, one for which many answers abound. Most of the answers center on a date and a word. The date is 1994 - the most surprising of political years since Dewey defeated Truman. The word is "nationalize." The Democrats need to nationalize the 2006 election, just as the Republicans did in 1994.

Is that really an answer, though? Nationalizing an election is not like declaring shenanigans at a carnival. You do not simply walk up to a microphone in front of a camera, declare that the election is "nationalized," and away you go. Nationalizing an election is a tricky thing to do.

The reason for this is the pro-incumbent bias inherent to congressional elections. It has greatly increased in the last few decades. When people talk about nationalizing an election, what they usually mean is bringing the national mood to bear on local House races. To nationalize an election is to essentially reward or punish members of Congress based upon the President's job approval. This is a difficult task. Incumbents have become increasingly able to retain their seats despite whatever partisan winds that might be blowing. Part of this has to do with redistricting, but most of it has to do with the fact that our Congress has become professionalized. Being in Congress is now a profession - and so, therefore, is being a candidate. It is part of members' jobs to know their districts and to know what it takes to get to half-plus-one. Professional members can also hire consultants, pollsters and strategists who make their living based upon electoral victory.

But this bias existed before World War II, and it roots extend to the very foundation of the American experiment. Our system of government - in particular its doctrine of separated powers - makes it very difficult to identify who is responsible for any given policy or any given result. So many different actors are involved in any given situation that confidently identifying a causal chain is virtually impossible. Hurricane Katrina provides a perfect example. Who should we blame for Katrina: the state government, the city government, the county government, the President, the Congress, all of the above, none of the above? I think it is literally impossible to place blame with any precision. Most answers I have read ultimately hinge upon the partisanship of the writer - liberals are more likely to blame institutions operated by the Republicans; conservatives are more likely to blame institutions operated by the Democrats. The reality is that power is so divided that nobody is clearly to blame. This enables all parties involved to shirk any responsibility they might have.

So it goes with almost all policies and outcomes. Madison and the founders instituted this type of system to thwart tyranny. Their thinking was that a division of power would confound potential oppressors by pitting them against one another. It was successful; but, as a side-effect, it greatly diminishes responsibility. Identifying who is responsible is a difficult task, and it works to the advantage of incumbents twice over: they can believably take credit for the good stuff and believably deny responsibility for the bad. When you combine this structural feature with members who are very skilled at campaigning in districts full of fellow partisans, you can appreciate why the incumbency retention rate has recently hit the 99% mark.

There is another way to nationalize an election - one that is much easier. You do not have to pit yourself against James Madison when you nationalize open seat races. Recently, it is through open seats that most swings in the balance of power have occurred. When members of Congress decide not to run for reelection - and many times such a decision is predicated upon national political conditions - local races take on a much more "national" flavor. For starters, voters are more inclined to vote according to their partisanship. This is key, as the effect of the incumbent's advantage is his ability to get voters who are of the opposite party to support him. You also see a more explicit discussion of national issues. The freshmen class that entered Congress in 1992 was there largely because the election was nationalized, which in turn was due to the large number of retirements stemming from the House bank scandal.

Thus, to an extent, all midterms are "national:" in every year there are at least a few open seats. Unfortunately for the Democrats, in 2006 there are only about 5 open Republican seats from marginal districts - i.e. districts whose partisan composition does not heavily favor one side or another. What the Democrats therefore need is to nationalize the election in a way that Republican incumbents bear the brunt of the force. This is what the Republicans did in 1994. They grossed 56 Democratic seats that year. 22 of them came from open seat victories. A whopping 34 came from defeated Democratic incumbents, most of whom occupied marginal districts.

Just as the Republicans did in 1994, the Democrats have to find a way to counteract the advantages that incumbency confers. The question becomes: how did the Republicans manage this counterintuitive result?

They did this, in essence, by using Bill Clinton's knack for the art of persuasion against him. Clinton was able to persuade Democratic members of Congress to go on the record supporting left-center policies that their districts opposed. He thus made what is usually an opaque picture of responsibility crystal clear. This enabled the Republicans to do what is so rarely done: bring the national mood home to the district.

For how much the conventional wisdom asserts that Clinton only gained his political bearings in 1995, the fact remains that the 103rd Congress had an impressive number of achievements that Clinton spearheaded. Many of them were decidedly left-of-center: notable among these were the tax increase provisions in the Deficit Reduction Act, the Brady Handgun Violence Prevention Act, and the assault weapons ban provision in the Violent Crime Control and Law Enforcement Act. Clinton - in conjunction with the House leadership - was very adept at inducing Democrats from marginal House districts, where these measures were quite unpopular, to vote with the party. In roll call vote after roll call vote, Democrats - particularly from the South and the Mountain West - voted with the party and against their constituents.

These roll call votes were what landed Democrats in so much trouble on Election Day. They explicitly and clearly attached House members to that which had upset voters in these marginal districts. Republican challengers could say more than, "Our Democratic incumbent is part of the problem!" This is what every challenger in every race says every year. It is not effective because it does not offer a coherent or compelling causal argument. Republicans in 1994 offered something more. In race after race, Republican candidates declared, "Our Democratic incumbent is part of the problem; just look at his votes on H.R. 1025, H.R. 3355, H.R. 2264!" In other words, the roll call votes made what is normally a quiet and confused trail of responsibility loud and clear. Voters who were angry about the direction of the country came to believe, thanks to Republican campaigns based upon these votes, that their members were to blame for it. They thus developed bad impressions of their members, and, on Election Day, they voted Republican.

The evidence is pretty clear on this point. As Stanford political scientist David Brady has argued:

"For many...Democrats, especially those representing more conservative districts, the choice (to support or oppose Clinton's legislative agenda) was...difficult: vote for a decidedly liberal legislative agenda, and risk offending constituents, or vote against the party leadership. Those who decided to support the president's legislative agenda risked providing Republican challengers with an obvious line of attack...Where Clinton ran poorly in 1992, Democratic incumbents with pro-Clinton voting records in Congress were much more likely to be defeated than those with lower levels of presidential support."
Examining the roster of defeated Democratic incumbents, one will only see a few who had opposed Clinton on all three of the aforementioned measures. Almost all of the losers supported Clinton at least once, and they paid for it with their seats.
Would an identical strategy work this year? It might, but it might not; 2006 is different in many respects from 1994. As mentioned, the Republicans won 22 open Democratic seats that year. This year, the Democrats can only hope for 5 such seats. Further, there was a tension that year that is absent this year: namely, there were many districts that tended to vote Republican in the presidential election and Democratic in the House election. This gave Republicans many opportunities for pickups that are simply absent this year. More than 90% of Republican members are in districts that voted for Bush. Fortunately for Democrats, the number of seats they need to win is much lower. The Democrats had a 40-member majority going into the 1994 midterms. The Republicans have a 15-member majority today. ...

comments powered by Disqus