Predicting the Midterm Elections: A Freakonomics Quorum

This year’s midterm elections promise to be a bit more eventful than usual, with predictions of seismic change in Congress and in many statehouses, most of it in a blue-to-red direction. But predictions aren’t elections; and even if the predictions hold true, what happens next?

So we checked in with some clever people who care a lot about electoral politics and asked them to risk their reputations by answering the following questions:

What’s your prediction for the outcome of the upcoming midterm elections? What will happen/change as a result of the outcome?

Here are their replies. Thanks to all of them for participating. I have to admit that I am particularly fond of this gem, from political scientist Seth Masket: “First, many D.C. pundits will claim that voters sent Washington a message that they want the two parties to overcome their bitter divisions and work together for the betterment of the country.??Second, this will not happen.” And Justin Wolfers‘s prediction is, simply put, a work of art. Enjoy.

Seth Masket is a professor of political science at the University of Denver.

“First, many D.C. pundits will claim that voters sent Washington a message that they want the two parties to overcome their bitter divisions and work together for the betterment of the country.??Second, this will not happen.”

Midterm elections tend to be less easily predictable than presidential elections, if for no other reason than midterm elections are not a single national event – they involve simultaneous elections in different states influenced by different rules, party systems, and political and economic environments. Nonetheless, we can draw some reasonably good predictions based on just a few key things that we know influence people’s votes: economic growth and the president’s popularity.

I prefer a simple model with just those two variables as predictors. I could probably be a bit more precise if I also included some polling results, but polls tend to be poor predictors of elections until a month or so before Election Day.??By contrast, my simple model uses pretty obvious predictor variables that are available by mid-summer of an election year.??In a sense, this model lets us see just how much information we can glean from the “fundamentals” of elections, which are apparent before most of the campaigning even begins.

I use the growth in real disposable income from the third quarter the year before the election to the second quarter in the year of the election, plus the president’s Gallup approval rating on Labor Day. Neither of those statistics is looking particularly great for the Democrats this year. Disposable income grew by 0.88 percent during this time period, and the average growth rate during midterm election seasons since 1950 is roughly double that. Also, the president’s approval rating of 45, while not as low as George W. Bush‘s in 2006 (39), is still well bellow the average of 53 during midterm election years.?Using this information, I forecast the Republicans picking up 40 seats in the House. This is, of course, just one more seat than the Republicans need to become the majority party.

I should note that this estimate, like most others, is rather “noisy.”??There’s a substantial error term either way, meaning the Republicans could win a more substantial majority, or the Democrats could hold onto the House by a handful of seats.??I estimate about a 53 percent chance that Republicans will take the majority.

The Senate is harder to forecast in the absence of polling, simply because we have so many fewer cases.??Only about 33 seats are up in a given election, and only about a third of those are considered competitive. Recent polling suggests that the Democrats will probably retain majority control of the Senate, but only by a margin of a few seats.

My forecast also predicts that Republicans will seize control of 16 state legislative chambers currently run by Democrats.

So, if, as I expect, Republicans manage to achieve narrow control of the House while the Democrats maintain a slim margin in the Senate, what will this mean? I feel far more confident in these predictions.??First, many D.C. pundits will claim that voters sent Washington a message that they want the two parties to overcome their bitter divisions and work together for the betterment of the country.??Second, this will not happen.??Indeed, narrow control by one party or the other will only make intense party discipline more important – the defection of just a few members can mean a major policy loss by the majority party.

Chris Lehane is a Democratic political consultant and crisis communications expert.

“The era’s ‘political roller coaster’ nature where the country careens from Republicans to Democrats as a consequence of neither party’s ability to address voter anxiety will lead to the chattering class questioning whether the two-party system is structurally capable of meeting the challenges of our times.”

With the qualifier that conventional wisdom in American politics is almost invariably wrong, here are five key trends that we can look for coming out of the mid-terms:

  1. Obama to Follow the Bush “Turn Out the Base” Re-Elect Playbook: Team Obama will approach the 2012 presidential election with the mindset that the composition of a presidential electorate is fundamentally different than that of a mid-term, where the fickle Independents are critical, and believe that their path to victory will be turning out the Democratic base and the so-called “Obama Voters.”? The challenge for this White House is that whereas President George W. Bush spent his first four years governing to, for and by the Republican base, President Obama is unfairly perceived by some in the Democratic base as not having delivered on the promise of his 2008 candidacy.
  2. Red State Mensheviks vs. Red State Bolsheviks: The Republican Party’s ongoing civil war will continue through the 2012 Republican presidential primary – and will immediately manifest itself on Nov. 3 in the form of a House Republican leadership fight.? Egged on by Elmer Gantry wannabes, ambitious House Republicans like Mike Pence, Eric Cantor and perhaps others are likely already sharpening their garrotes to challenge the current leader John Boehner.
  3. The Emergence of a “Vital Center” in the Senate: The mid-terms will result in a significant change in the ideological composition of the Senate – creating a Tea Party Caucus on the right and a de facto, tri-partisan, third party populated by the likes of Joe Lieberman, Ben Nelson, Olympia Snowe and Susan Collins that will be the center of gravity in the Senate.? As a subset of this dynamic, expect an all out full-court press post-Nov. 2 by Senate Republicans to try to flip some conservative Democrats over to the Republican side.
  4. Roller-Coaster Politics: The era’s “political roller coaster” nature where the country careens from Republicans to Democrats as a consequence of neither party’s ability to address voter anxiety will lead to the chattering class questioning whether the two-party system is structurally capable of meeting the challenges of our times.? Of course, such structural governing questions will bump into the structural electoral realities facing any real third-party candidacy, including dough, ballot access and an Electoral College system that is built for two parties.
  5. An Intervening Event: President Bill Clinton‘s handling of the Oklahoma City tragedy afforded him a second opportunity with the American public.? No president or party would ever hope or base their future on such a development, but the President will most certainly face a major, unexpected challenge of historic import in the next two years – and how he handles, responds and deals with that challenge could well dictate the remaining two years of the first term and beyond.

Robert S. Erikson is a professor of political science at Columbia University.

“The most certain thing one can say about this election is that the media agree on the narrative.”

When in 1994 the Republicans took control of Congress for the first time in 40 years, almost nobody saw it coming. When the Democrats regained control in 2006, this possibility did not even enter the discussion until the campaign’s final weeks. This time, the political world is prepared.? A Republican triumph will be no surprise in 2010.

The most certain thing one can say about this election is that the media agree on the narrative.? Election news is reported to fit the?story line that the Republicans will win the House and maybe the Senate. Of course, most events do fit the narrative and reinforce it.?? The economy stagnates, the polls favor the Republicans, tea-partiers are marching with enthusiasm and Democratic politicians have been peculiarly defensive.? Yet could it be that campaign news is filtered too much through the narrative of a Republican surge?? If events somehow begin to favor the Democrats, would they be noticed and interpreted objectively?

The narrative of a Republican success has been fueled by the continued struggles of the economy. But by itself the bad economy cannot explain the Republican surge.? Even in early Spring 2009, within a few months of President Obama’s inaugural, the “generic ballot” polls were suggesting that the Republicans would do well in 2010.? This was before any reasonable person could blame Obama for the economy, and at a time when attraction to the Republican brand was at its historic low.

Why, at the very start of this election cycle, were people already poised to vote for a party they did not much like against a party headed by a president who they generally liked? The best explanation is that this was an acceleration of the usual process that fuels midterm losses for the presidential party. Midterm voters try to brake the president’s ideological agenda by electing more of the opposition party to Congress.?? From the outset of this administration, policies like the stimulus and the takeover of auto companies galvanized the right and convinced more voters than usual of the need for an ideological correction.

If the Republicans control the new Congress, they should read the electorate’s message carefully and not misread it as a full endorsement of the Republican agenda.??? As the referee, the public prefers its policies to be somewhere in the ideological middle. An ideologically misplayed Republican victory could even work perversely to Obama’s advantage in 2012.? As with Clinton in 1996, Obama might benefit more by running against a runaway conservative Congress than by working with a slim and balky Democratic majority.

Before speculating in detail about what would happen with a Republican?Congress,?we should consider the small possibility that the dominant narrative is wrong.? The polling evidence is sufficiently ambiguous to offer a sliver of doubt?that the Republicans can gain the 40 seats needed to control the House. Nate Silver‘s prognostications and the election markets suggest that the odds of a Republican takeover are about 4-1.? That seems about right.

Karen Spencer is the president of Spencer-Roberts and Associates, a political consulting firm.

“And at this point, it looks like the voters want to even out the political parties in Congress.? They want new faces.”

When asked what my prediction is for the November election, I get nervous. Anything could happen to drive people to the polls.? And regardless of all the paid media, the cable pundits, the opinion articles and the Sunday talk shows, it’s the voters to who make the decisions.? Sometimes we forget that little point.? And at this point, it looks like the voters want to even out the political parties in Congress.? They want new faces.? Why? Because the old ones don’t seem to trust one another.? Congress (both parties) appear not to respect each other and don’t treat each other with respect and integrity.? As people expect of each other.?? And if these elected officials don’t trust one another, how can the public trust their elected officials?

If what I’m seeing now in the polling around the country is right, and those elected are more moderate Democrats or Republicans, I expect more deadlock in Washington. Moderates, centrists and Independents will continue to be unhappy because no one is listening to them.? Special interests are either pulling elected officials to the right or left.? And they go there because they get the most attention from the media.? I believe this “ping pong” election effect will continue until Congress?and the White House starting dealing with meaningful legislation for the middle class.

I’ve spent productive years working in D.C. both in Congress and the White House. This complete dysfunction of the last decade is confounding and disheartening. I simply don’t understand this dynamic.? Our Founding Fathers faced and expected this same dilemma in future generations but still produced amazing documents and legislation.? Our elected officials need to understand that they need to bring their own special interests to the table and settle them for the public interest. It is time to revamp our legislation for the 21st century.

Justin Wolfers is a professor of business and public policy at the University of Pennsylvania and a?regular contributor to this blog.

“The Democrats will retain control of the House and the Senate.? And I’m the only person in D.C. insightful enough to make this brave forecast.”

The Democrats will retain control of the House and the Senate.? And I’m the only person in D.C. insightful enough to make this brave forecast.

The talking heads on TV are nearly unanimous in saying instead that the Dems will lose the House.? Some hedge and say “it all depends,” and then list about a zillion factors.? (Of course it depends!)? But only I am brave enough to pound the table and say: “2010 is the year of the Democrat.”

Now the serious sources I trust suggest there’s a chance for the Dems; it’s not huge, but it is worth taking seriously.? InTrade says that the Dems are only a?one-in-eight to retain control, while uber-pundit?Nate Silver says there’s only a?one-in-five chance they retain control.? And Nate is?telling anyone who will listen that election forecasting is a difficult business, and that we should take uncertainty seriously.? The prediction markets agree.

But this lesson doesn’t sit well among the pundit classes.? It’s not the sort of story that makes for good TV: “Election hard to forecast; models come with errors.” Or perhaps “World complicated.? Highly-paid experts often wrong.”

But it’s true.? And one-in-eight events?do sometimes happen.? If this election is one of those times, then I’ll be the only person in the whole country to have called the 2010 election correctly.

If I’m right?? Well you can bet that I’ll beat the drums loudly and tell everyone in sight that I called it. I’ll blog it all week.? I’ll write an op-ed explaining my insights. I’ll go on to Jon Stewart‘s show to explain the fine art of psephology.? Hopefully you’ll be calling me the Nouriel Roubini of political punditry.? I’ll go on to a new life of lucrative speaking engagements and big book advances, while I beat back my coterie of devoted followers.

And if I’m wrong?? We both know there won’t be any real consequences.? I’ll be sure to sell some clever story.? You know, there was weather on election day (hot or cold, wet or dry – it all works!) and this messed with turnout.? Or perhaps, This Time Was Different, and my excellent forecast was knocked off course by our first black president, by rising cellphone penetration or a candidate who?may not be a witch.? I’ll remind you how I?nailed previous elections.? (Follow the links, you’ll see I’m doing it already!)? I’ll bluster and use long words like sociotropic, or perhaps heteroskedastic.? And I’ll remind you that my first name is Professor, and I went to a prestigious school.? More to the point, if I’m wrong, I’m sure we’ll all have forgotten by the time the 2012 election rolls around.? Shhhh… I won’t tell if you won’t.

So yes, my forecast is more about the marketplace for punditry than it is about this election.? I’m influenced strongly by my Penn colleague?Philip Tetlock, who has spent decades?pointing out just how bad expert political judgment is. Given these market failures, I would be a fool not to go for the gold.

So you heard it here first: The Dems will win.? But if they don’t, hopefully we can all just forget about it.

Alan Abramowitz is a professor of political science at Emory University.

“So what will all of this mean for the 112th Congress? Most likely, narrow majorities in both chambers, increased partisan polarization and gridlock.”

Republicans are poised to make major gains in the 2010 midterm elections. That’s because Democrats are facing a “triple whammy” this year. First, it’s a midterm election with a Democrat in the White House, and voters almost always turn against the president’s party in midterm elections. Second, Democrats have to defend a lot of seats in Republican-leaning districts that they picked up in the last two elections. Forty-seven House Democrats represent districts that were carried by both George W. Bush in 2004 and John McCain in 2008. And last but not least, there is a high level of discontent among the public over the direction of the country and especially the condition of the economy.

Put those three things together, and you’ve got a recipe for a wave election-one in which a lot of seats change hands. Recent polls show a generic Republican House candidate leading a generic Democratic House candidate among likely voters by an average of six to eight points. That’s very unusual-historically, a generic Democrat almost always beats a generic Republican-and it’s a dramatic reversal of the results in 2006 and 2008 when the generic Democratic candidate had a big lead during the final days of the campaign.

A simple forecasting model based on three predictors-the president’s party, the number of seats held by each party prior to the election and the generic ballot margin-produces very accurate predictions of the results of midterm House elections a couple of months before Election Day. In 2006, for example, this model predicted that Democrats would win a majority of House seats long before most of the pundits came to that conclusion. This year, the model predicts that Republicans will gain 45-50 seats in the House of Representatives, giving them a fairly narrow majority in the next Congress. The model is less accurate for the Senate where the overall outcome always hinges on a number of very close races, but Republicans should pick up at least 5 or 6 seats there and they could gain as many as 9 or 10.

So what will all of this mean for the 112th Congress? Most likely, narrow majorities in both chambers, increased partisan polarization and gridlock. Most of the Democratic losses will come from the party’s moderate wing because moderates tend to represent the most vulnerable districts and states. In almost every case, however, the Republicans taking those seats will not be moderates but strong conservatives, many elected with Tea Party support. So the likely result of the midterm election will be a smaller but more liberal Democratic caucus and a larger and more aggressively conservative Republican caucus.

It’s not a recipe for bipartisan cooperation in addressing the nation’s enormous challenges. Instead, what we’re likely to see in the next Congress is Republicans, pushed by the party’s Tea Party wing, seeking to extend the Bush tax cuts, drastically cut spending on domestic social program, and repeal almost every major piece of legislation passed by the current Congress. Expect Senate Democrats to spend most of their time filibustering to block GOP bills and, if that fails, expect President Obama to use his veto pen. There is little or no chance that any of the remaining items on the President’s agenda such as climate change legislation and immigration reform will be passed or that major presidential appointments will be confirmed or even come to a vote. And expect all of this to shape the 2012 presidential election, which will make the current campaign look like a garden party.

Leave A Comment

Comments are moderated and generally will be posted if they are on-topic and not abusive.



View All Comments »
  1. WTP says:

    Wolfers comment is clearly the most honest analysis by any pundit I have ever read. If only they would all admit that they are just trying to set themselves apart and make money, rather than drive a debate.

    Thumb up 0 Thumb down 0
  2. Richard Charnin says:

    Not one of the above mentioned the F-word: Election Fraud.

    ALL forecasters base their projections on the LV subsample which are generally quite accurate in predicting the recorded vote. But the recorded vote is bogus because of election fraud.

    Based on historical UNADJUSTED exit poll data and the True Vote Model, the Democrats do 3-4% better in the True Vote than the recorded vote.

    Only Election Fraud can cause the Democrats to lose the Senate.

    In the House, the pundits predict that the GOP will win approximately 235 seats. . Again, the LV polls have closely matched a fraudulent recorded vote.

    RV polls indicate that the GOP will win 222 seats. And that assumes an even split in theundecided vote. The Democrats may very well win the True Vote with 220-230 seats. But pollsters are paid to predict the recorded vote. And the Democrats will no doubt lose the recorded vote.

    Thumb up 0 Thumb down 0
  3. frankenduf says:

    i utterly agree with Richard Charnin- the real freakonomics story evolving is election fraud, which one may tease out looking at exit poll data, which was historically accurate until the 2000 cycle, when anomalies arose, coinciding with automated voting systems which leave no paper verification trail, and have not been systematically audited- freakonomics has exposed cheating in teaching grades and sports stats- it is curious to me why they do not expose the statistical facts to unmask election fraud

    Thumb up 0 Thumb down 0
  4. Drill-Baby-Drill Drill Team says:

    I am no Nostradamus–Beards are itchy–but I will have a go at Predictionomics.

    America is NOT a THREE PARTY system. A Third Party Parasitizes votes from its Ideological Brother. The much vaunted TEA Party will rob Republicans of valuable votes in close races.

    Tea Party candidate sucess will be less than 5%, but Sarah Palin will SPIN THIS and proclaim this as a victory and a mandate for her Presidential bid.

    The Democrats will gain seats in their Senate Majority.
    The Democrats will lose seats in the House, but still retain a Majority.

    It is like having a niggling Ralph Nader in the 2000 Presidential Election. Only it’s Sarah Palin and her Birthers. Call the Midwife.

    An Opposition needs to be consolidated and organized to overcome razor thin margins. Divided it Falls.

    Thumb up 0 Thumb down 0
  5. Jim Capl says:

    While I am not yet being called for interviews and projections, it is clearly only a matter of time once my views such as the following become more widely known: “Candidates with the highest vote totals will win close races by slimmest of margins” and you can quote me on that.

    Thumb up 0 Thumb down 0
  6. Eric M. Jones says:

    I always find it remarkable that the low-information voters are gathered up by the Republicans, essentially voting against their own best interest. I have come to an understanding on this:

    Low-information voters have exactly the same hope for upward economic and class mobility that an old fat drunk guy in the bar has for the cute waitress. They live on hope–and are more than willing to believe that their day will come–.

    Thumb up 0 Thumb down 0
  7. Steven Bearden says:

    I am curious about a statistical fact that could mean absolutely nothing. Isn’t one of the mantras of Freakonomics that correlation is not causation. You are talking about inconsistencies in two data sources where one is voluntary questioning. I would think that one would expect there to be a difference.
    Now, I’m not saying that there isn’t voter fraud, there very well could be. But I’ve never seen data that would convince me that it exists in a significant quantity to decisively sway an election.

    Thumb up 0 Thumb down 0
  8. Greg says:


    What you are misunderstanding is that many “low-information voters” live in largely rural areas, and have little to no desire to fund the Democrats typical legislation that largely benefits those living in urban/suburban areas.

    Thumb up 0 Thumb down 0