Let me begin the initial post-election analysis by thanking Rachel Pagano, Ben Wessel, the Republican and Democratic student groups, Owen Witek and Sarah Pfander who manned the computer all night, all the media staff, everyone who bought me beer, Wenbo for the food, and of course my colleague Bert Johnson for his usual astute insights and commentary. It was another great night at the Grille – greater for some than for others- and we also broke a record for participation on the Presidential Power website – thanks to all the commentators, particularly Conor, Tarsi and Chris, who kept the updates coming. Kudos to all.
Now on to the Results. We wake up to a Republican-controlled House, as Republicans gained – so far – 60 seats, the biggest pickup for any party since 1946, which puts them at least at 239 seats, with the likelihood as of this writing that they will reach 243, far more than the 218 needed for a majority. The Senate remains under Democratic control, despite at least a 6-seat Republican gain there. As I expected, Colorado is simply too close to call, and results are pending in Washington and Alaska, although the latter should stay Republican (it appears Murkowski might win there as write-in candidate!). So we are probably looking at a 4-to-5 seat Democrat margin in the Senate, 53—47 or 52-48, pending the results in the remaining three states. The split outcome in the Senate and House is not a surprise, but the absolute size of the Republican seat gain in the House exceeded the Labor Day mean projections of all political scientists, although it fell within the confidence interval of most of their projections. As I noted in an earlier blog, political scientists are an inherently conservative lot when it comes to forecasting unusual electoral outcomes and, at least in the House, this was an unusually large gain by historical standards. My quick look at the data indicates that there have only been three greater swings in the House since it achieved its current size of 435 in 1913 – those occurred in 1938, 1942 and 1946. The Senate Republican increase of 6-7 seats is more modest, historically speaking, having been eclipsed most recently in the 1980 Reagan election, but still a significant gain.
What explains the massive size of the Republican wave? In retrospect, I think the source of the wave dates to the end of the Bush administration, with the initial bank bailout. The decision by the Bush administration – one ratified by Democrats – to stabilize the banks that were viewed by many taxpayers as a victim of their own bad judgment did not sit well with voters and – as the economic crisis worsened – it triggered a populist backlash that we’ve seen frequently in American history. The most visible manifestation of this movement was the Tea Party. Many on the Left initially dismissed the Tea Party movement because they were partially misled by the media coverage that disproportionately focused on its fringe elements. But, at its core, the Tea Party was driven by a deep concern and even anger over the growing size of government, as reflected in TARP, the auto bailout, the stimulus bill and finally, health care, and the idea that government was more interested in colluding with big business – banks and financial centers, the auto industry, insurance and pharmaceutical companies, etc. – than in protecting voters. The Tea Party, however, was only the most visible manifestation of this dissatisfaction – a dissatisfaction that grew larger when the economy showed little job growth well into the Obama administration. At its heart, this election turned on the economy.
At least that’s my first read. As evidence, let’s take a look at the House exit polls, keeping in mind John Sides’ warning that the reasons people give for why they voted as they did are not necessarily accurate. We should note at the outset that turnout was down among those groups – youths, Latinos, African-Americans – that formed a core part of the Obama coalition in 2010. At the same time, it was up from 2008 among older voters and they skewed heavily Republican. Indeed, Republican turnout at 36% of the voters was more than the proportion of all adults who call themselves Republicans, and equal to Democrat turnout on Tuesday, despite the Democrats’ registration advantage. However, independents (28%) were the key here, as they broke decisively for Republican candidates 55%-39%.
Note as well that self-professed conservatives, at 41%, were the biggest voting bloc, with moderates coming in second at 39%. Liberals, by contrast, constituted only 20% of the voters. Obama voters constituted only 46% of House voters, not much more than McCain voters (45%) – another change from 2008. Republicans regained their support lost in 2008 among suburban voters, and broke even with Democrats among voters in the West, while retaining their dominance in the South and, to a lesser extent, the Midwest.
So, the composition of the midterm electorate, as I expected, favored the Republicans. This is consistent with the surge and decline thesis that explains the president’s seat loss in the first midterm more generally and which I’ve talked about extensively in previous posts.
A second factor affecting midterms is that, for some voters, the election is viewed as a referendum on the president. There is evidence that this dynamic was at play yesterday. Among House voters, fully 39% strongly disapproved of the job Obama is doing, while only 23% strongly approved. Those disapproving went almost entirely for Republican candidates. Thirty-seven percent of House voters said their vote was meant to express opposition to Obama – only 24% said they voted in support of Obama. By a margin of 52-44%, House voters thought Obama’s leadership had hurt the country.
Both these factors – the composition of the electorate, and attitudes toward Obama – were shaped by voters’ dominant issue concern: that government was growing too large against the backdrop of a stagnant economy. By a 56-38% margin, voters thought government was doing too much, and the “doing too much” crowd voted overwhelmingly Republican, 76-21%. By an almost equally overwhelming 74-24%, voters expressed anger/dissatisfaction with the federal government, and the angry/dissatisfied crowd went Republican 64-33%.
What about the Tea Party movement? Forty percent of voters supported it – higher than the average support I was detecting among all registered voters across numerous polls – while 31% opposed and 25% were neutral. Neutral voters went slightly (49-47%) for Republicans, but Tea Party supporters broke strongly for Republican candidates 86-11%. However, more than half of the voters (56%) said their vote was not meant to send a message, for or against, the Tea Party. In short, the Tea Party movement was not the issue for most voters in this election – the economy was. Fully 62% of voters said the economy was the most important issue facing the country – far more than any other issue, and they went Republican 53%-44%. A plurality of voters want the next Congress to cut the deficit (39%) and another 37% want it focus on creating jobs – core Tea Party issues. Interestingly, voters were divided roughly into thirds on whether the stimulus package had helped, hurt or made no difference to the economy. Most voters blamed either Wall St. 35%, or George Bush (30%), for the state of the economy more than Barack Obama (23%). But the Democrats still felt the repercussions of a bad economy. Efforts by pundits like Frank Rich and E.J. Dionne to make the Tea Party an issue in this election went nowhere.
Interestingly, 48% of Americans want to repeal the health care law, 16% leave as is, and 31% expand it. Given the divided control of government, I fully expect that no changes will be made to health care, if at all, until after 2012.
Finally, by a slim margin, a plurality of voters (39%) wants to extend the Bush tax cuts for all Americans.
Exit polls offer a first, albeit potentially misleading look at what voters were thinking on Tuesday, so we need to be careful in extrapolating from the results. Nonetheless, the picture they paint is one consistent with political science forecast models more generally: the midterm electorate was less favorable to Democrats in general (surge and decline), and to Obama specifically (referendum). To this we might add a third factor, although I have less data on this: the tendency for a more moderate electorate to split government control between what they view as two relatively extreme parties. By my count, the next two years of divided government means that voters have chosen to split control 38 of the last 68 years, or more than half the time.
The biggest factor driving the vote on Tuesday, however, was the one that incited the Tea Party movement back in late 2008: concern over the size of government against the backdrop of a stagnant economy. Simply put, with unemployment hovering above 9%, voters were in the mood for change, and they took their frustrations out on the party in power. The Tea Party was simply a manifestation of this deeper, more widespread voter dissatisfaction – not a cause of it.
If I get a chance, I’ll do a separate analysis of the Senate exit polling data. For now, however, we face divided government again. The difference is that this time we have a Democratic President and Senate, but a Republican House – a combination we’ve not seen since before the Civil War, if I’m remembering my history. Because the House in its modern incarnation is designed to empower the majority party, the political dynamics for the next two years will be interesting, to put it mildly. In a future post I’ll try to preview what I think will happen.
In the meantime, let the 2012 presidential campaign begin!
ADDENDUM: Jeff Garafano emails to say Bennet is declaring victory in Colorado. If true, this pushes the Democrats to at least 52 in the Senate, but I am waiting to see whether this goes to a recount.
UPDATE 2:05 pm. I am told that Colorado is unlikely to go to a recount – Bennet’s lead is big enough that no recount will be triggered and Buck is unlikely to pursue/pay for one on his own.