Skip to content

The Do-Nothing Discipline

How political science fell into the thrall of fundamentalist forecasting

In November 2008, most Americans acknowledged George W. Bush’s presidency as a failure. The Republican Party, unquestionably complicit in that failure, was down for the count; Barack Obama defeated Republican John McCain, a popular war hero, decisively. Conservatism itself seemed rocked on its heels, perhaps for good. When recently inaugurated President Obama delivered an address to a joint session of Congress to tout his $787 billion stimulus package, just 8 percent of TV viewers said they disapproved of Obama’s performance. By sharp contrast, “rising star” Republican Governor Bobby Jindal of Louisiana delivered a lackluster response, which featured an attack on a $140 million outlay for “something called ‘volcano monitoring’”—just a month before a volcanic eruption sent a sixty-thousand-foot cloud of ash over Alaska. Conservative New York Times columnist David Brooks called Jindal’s “stale, government-is-the-problem” ideology “a disaster for the Republican Party.”

For once, Brooks seemed to be attuned to the mood of the public. By May 2009, a Gallup analysis concluded that the “decline in Republican Party affiliation among Americans . . . has occurred among nearly every major demographic subgroup.” As Time magazine reporter Michael Grunwald observed at the time, “Republicans have the desperate aura of an endangered species . . . the electorate is getting less white, less rural, less Christian—in short, less demographically Republican.” The last time an American political party looked this weak was 1932, when Herbert Hoover’s Depression-era defeat, combined with the rise of the New Deal coalition, led to Democratic control of the House of Representatives for fifty-eight of the next sixty-four years. All signs suggested that Democrats in the Obama years could be on a similar path to long-term majority control of Congress and national politics at large.

So much, it seems, for the invincible new Democratic majority. In short order, nearly all the key indicators of political power began flipping the other way—toward the specter of long-term Republican domination. The Tea Party election of 2010 flipped the House of Representatives to Republican control. Almost eight years later, Republicans have majorities in both branches of Congress, Donald Trump is president, Neil Gorsuch is set to be a long-term fixture on the Supreme Court, and Republicans control sixty-seven of the ninety-eight partisan state legislatures. President Obama and his media fans needed excuses. Fortunately for them, a complacent yet tirelessly self-promoting group of political scientists was up for the challenge.

Fundamentalists for Hire

These diligent analysts have no official governing body, nor a formal designation within the political science profession. They don’t convene separate conferences or publish a journal teeming with rationales for Democratic failure. But their methods are strikingly uniform—as are their arguments. They share a stolid insistence that statistical analysis of a handful of quantifiable facts from past elections can predict with scientific confidence voter behavior in the next one. Within this somewhat enclosed system of study, the main analytical challenge becomes just how much weight to give each fact under review—except in the case of the scholars who reduce the Delphic variables on hand to a single one: the rate of economic growth during the previous presidential terms.

Since these thinkers are fond of calling their fetishized, quantified variables “fundamentals,” let’s call them “fundamentalists.” We might as well because, for all their statistical sophistication, their accounts of human motivation are so narrow as to bring to mind the dogmatic conviction that dinosaur fossils are less than six thousand years old. As they gained prominence during the Obama era, fundamentalists argued that the political factors obsessed over by journalists and activists, even the day-to-day analysis of aggregated poll data favored by the new breed of data journalists such as Nate Silver and Nate Cohn, are essentially a distracting waste of time. They also—and this is important—consider the daily tactical give-and-take of campaigns, the grand machinations of strategy, wildly overrated and largely irrelevant.

The best-known fundamentalist model is the brainchild of Douglas Hibbs, whose globe-trotting academic career includes stints at Harvard University, Sweden’s University of Gothenburg, Addis Ababa University, and the Sorbonne. His model considers two variables: “weighted-average growth of per capita real disposable personal income over the term” (i.e., personal income) and “cumulative US military fatalities (scaled to population) owing to unprovoked, hostile deployments of American armed forces in foreign wars” (i.e., non-necessary American military deaths). The model is called “Bread and Peace.” Nate Silver has called it the “best of its kind.”

But other contenders have made nearly as fetching an impression. Take Alan Abramowitz of Emory University. Abramowitz’s model has a snappy campaign-style sobriquet, “Time for Change,” and it claims the power to “forecast the incumbent party’s share of the major party vote with a high degree of accuracy around three months before Election Day.” It relies on three factors: “the incumbent president’s approval rating at midyear (late June or early July) in the Gallup Poll, the growth rate of real GDP in the second quarter of the election year, and whether the incumbent president’s party has held the White House for one term or more than one term.” (In the 2012 election year, he added a fourth factor: polarization.) In the New York Times, which has a near-insatiable appetite for academic prognostication, political scientist and pundit Brendan Nyhan wrote that “Mr. Abramowitz’s model dominates the field in its forecasting accuracy.”

Then there’s the model of Christopher Wlezien of the University of Texas and Robert Erikson of Columbia University. It unfortunately lacks a catchy name and focuses on a correspondently colorless body of data: the Conference Board’s index of leading economic indicators. “In previous articles,” they explained last summer, “we have shown that the weighted growth in these individual indicators through the spring of the election year—quarter thirteen of the election cycle—is a strong indicator of the vote.”

So much, it seems, for the invincible new Democratic majority.

And finally, Helmut Norpoth of Stony Brook University is the architect of the “Primary Model,” which has its own snazzy web site. Norpoth boasts that his model has “correctly predicted the winner of the popular vote in all five presidential elections since it was introduced in 1996.” He bases its forecast of the popular vote winner in November on the performance of the nominee in early primary contests.

Beginning around 2010, models like these came to exert an almost mystical fascination for a certain sort of more-sophisticated-than-thou journalist. Vox maestro Ezra Klein, the wunderkind blogger-turned-journalist-turned-media-entrepreneur was prototypical. “For decades now,” he enthused that summer when he was manning the policy desk at the Washington Post’s web operation, “political scientists have been building election models that attempt to predict who will win in November without making any reference to candidates or campaigns. They can get within two points of the final vote, and they don’t need to know anything about the ads and the gaffes and the ground games. All they really need to know about is the economy.” Paul Krugman delivered a similar hosanna chorus a week later: “What political scientists, as opposed to pundits, tell us is that it really is the economy, stupid.”

Alibis ‘R’ Us

One can squint a bit and figure out why fundamentalist pseudoscience might appeal to an economist like Krugman. But why, exactly, has it transfixed a new generation of political journalists—particularly as it professes to render politics, as a popular call to civic arms and a battle of ideas, all but irrelevant? That’s a mystery—and the mystery deepens when you factor in (as it were) the less-than-stellar recent performance of the omniscient modeling caste. The fundamentalists’ most influential booster is John Sides, the co-founder of the political science blog The Monkey Cage, and a political scientist at George Washington University. But even he was forced to concede after the 2010 midterms that the models didn’t perform as advertised: “GOP gains are greater than we would expect, based solely on the economy and presidential approval.”

The models’ track record seemed to be less momentous for the Obama Democrats than the fact of their existence: here was an academically sanctioned narrative of performance at the polls that also doubled as an all-purpose political alibi. Democratic failure, in this cloistered universe of social causation, was not a function of the party’s deficient platform, or its ongoing fealty to the financial sector, or even its fundamental incompetence at politics. No, the simple numbers showing a still sluggish economic recovery after the 2008 meltdown all but set the Democrats on course to a bitter midterm rendezvous with destiny. It was surprising, given the supposed power of this trend, that the party managed to go out and campaign at all.

The majesty of the fundamentalist narrative was so intoxicating that few paused to note that it bore a rather distressing family resemblance to the discredited economic determinism of the Marxist catechism. Historical certainties of the determinist creed rendered human agency—intervention in the machinery of government on behalf of an engaged or outraged citizenry—all but a dead letter. But then again, as fundamentalist narrative wafted out of the centers of liberal power in Washington, such political shortcomings appeared to be a feature, not a bug. In a November 2010 post-midterm election interview aired on 60 Minutes, no less an eminence than presidential polymath Barack Obama echoed the key articles of the fundamentalist faith. “I think first and foremost [the midterm elections were] a referendum on the economy,” the president announced—and added for good measure that “the party in power was held responsible for an economy that is still underperforming.”

President Obama may not have been summoning the mood of the country—recall, after all, that Monkey Cage impresario Sides had noted that the Tea Party takeover overshot the path plotted by key economic indicators. But he definitely spoke for the consensus taking shape within his administration. Virtually all of Obama’s White House team were said to be fundamentalists. A few months ahead of the midterm balloting, in fact, then New York Times political reporter Matt Bai offered a tidy summary of their mounting self-pity and sense of passive victimhood. Bai noted that the “president’s allies . . . suggest that voters are blaming the party in power for a stubbornly sluggish economy.”Longtime Brookings Institution election analyst William Galston took note of the same lugubrious trend immediately after the 2010 midterms, noting that the “mainstream liberals” who were “the most supportive of the president” focused their post-mortems on Obama’s toxic inheritance from Bush. “In short, proponents of this view contend, Obama and the Democrats are mostly the victims of forces beyond their control.”

It’s hard to picture Franklin Roosevelt endorsing a similar self-undercutting view of things for the crucial 1934 midterms—which were indeed a referendum on the still unproven performance of the New Deal. Roosevelt might have found some narrow empirical justification for adopting a fatalist, emo-style view of the electorate’s mood back then. Instead, he seized the rhetorical high ground. His 1934 State of the Union address bemoaned “the ruthless exploitation of all labor, the encouragement of speculation with other people’s money.” (He also, not incidentally, ratcheted up accountability for the financiers who wrecked the U.S. economy prior to his election—something the Obama Justice Department conspicuously failed to do.)

Ink-Stained Fatalists

Whatever the deeper explanation for Obama’s half-hearted approach to the 2010 midterms, there’s no doubt that the armature of fundamentalist prediction was mobilized to rationalize it after the fact. Despite his leaning on the state of the economy as his post-midterm alibi of first resort, on the 2010 hustings Obama did not push the (honest!) message that Bush and the Republicans were principally responsible for the economic calamity. A majority of voters continued to tell pollsters that Bush bore greater responsibility for the economy than Obama through the end of 2012—but somehow Obama’s Democrats did not run on a “blame Bush” message in the pivotal 2010 midterms. Instead, the Obama wing of the Democratic Party believed it had no opportunity to persuade voters who already believed Bush a failed president to vote against the candidates of Bush’s parties. Why? Because . . . fundamentalism!

Fundamentalists’ influence continued to grow with leading pro-Obama pundits. Ezra Klein’s compatriot Matt Yglesias wrote in 2011, “I’m pretty much an economic determinist about election outcomes.” John Sides was declared “Blogger of the Year” by The Week magazine for his work as co-founder of The Monkey Cage. The Week’s laurel-bestowing editors claimed that just as “Nate Silver became a superstar during the 2008 election by bringing serious statistical rigor to the horse-race coverage of presidential polling,” Sides “is doing the same thing for voter behavior, injecting the wisdom of the academy into everyday political discourse.” In March 2012 Klein scaled to the summits of The New Yorker to argue that that political scientists’ dispassionate empirical calculations had proven that politics—presidential speeches, in this particular case—had no influence, pretty much, on anything.

Then, in 2013, The Monkey Cage became part of the Washington Post. “We think that it will be a great place to continue the blog’s mission of publicizing political science research,” Sides enthused about his new journalistic home. Andrew Gelman, a Columbia University professor of political science and Monkey Cage contributor, calls the site “a good place to reach a wider audience” since “for better or worse, political science does not have a high profile in the news media or, I think, in the public policy world.” He then explained, in the very kind of nonsignifying jargon that undermines the public profile of his discipline, that “political scientists’ inputs into the public discourse are sporadic.”

Some Monkey Cagers were firmly aligned with fundamentalist election prediction models; some were not. But they all shared the bedrock conviction that “science” had the power to reveal truths about politics far beneath the notice of the “distracted” members of the hoi polloi. A de facto corollary of this faith was that the more counterintuitive these truths proved to be, the better; this was punditry flowing from the great fount of D.C. journalism, which has all but patented the term “contrarian.” Then came the 2016 presidential election. How did the fundamentalists do? Consider, well, the facts.

Failure Is Not an Option

On Election Day 2016, John Sides, leveraging his real estate on the Washington Post web page as the fundamentalists’ biggest booster, published a greatly reassuring survey of the election landscape. A “comprehensive average of election forecasts points to a decisive Clinton victory,” the original Monkey Cager pronounced. In his breakdown of the ostensible data, Sides jumbled together poll-based forecasts with fundamentalists’ “predictions” to argue, essentially, that if one model using one methodology is “science,” a meta-analysis of as many as possible must be science squared. Thus his confident conclusion: “The average of forecasts is better than the typical individual forecast—and the average suggests a five point victory for Hillary Clinton.” But he also told readers to pretty much ignore the “statistical models of elections—predicated on the economic fundamentals,” which “have forecast a tight race,” in favor of the poll-based predictions of non-professors (and non-fundamentalists) like Nate Silver and Nate Cohn. Then, the next day, remarkably, he all but apologized to his political scientist brethren, figuring out a way to torture the data to argue that, “for now, a striking fact is that the election ended up looking a lot like an average of the fundamentals-based models from several months ago.”

Wlezien and Erikson are the guys who claimed they could predict the outcome of presidential elections based on the Conference Board’s index of leading economic indicators and—as the election nears—the “incumbent party candidate’s current share of trial-heat poll averages.” A fortnight before the election Wlezien and Erikson’s model suggested Trump had a one in ten chance of victory. But in an interview around the same time, Wlezien hedged, admitting that the model only worked under very narrow, almost laboratory-like conditions—which is to say, in 2016, not at all. “All of this only applies when the candidates are evenly matched and don’t make too many mistakes,” he said. “Donald Trump, a non-traditional candidate is running for president—which, of course, is the secret behind his success—and this makes the outcome less predictable. . . . He won’t automatically get every Republican’s vote.”

He received 90 percent of them, identical to John McCain in 2008 and barely below Romney’s 93 percent in 2012.

As for Abramowitz, on August 11, with his “Time for Change” model telling him Donald Trump would win the popular vote by 2.8 percent, he advised his readers to ignore it. Instead, he argued that “despite the prediction of the Time for Change, Clinton should be considered a strong favorite to win the 2016 presidential election.”

Hibbs’s “Bread and Peace” model, meanwhile, picked Clinton over Trump by 7.8 percent—which, while inaccurate, was actually better than its performance in 2012, when it predicted Obama would lose to Romney by 6.8 percentage points in an election Obama went on to win by 3.9 percent.

Obama was like Charlie Brown to Mitch McConnell and John Boehner’s Lucy, constantly seeking to kick the football of bipartisan accord.

Norpoth’s model had been engineered to pick the winner of the popular vote. He had given Trump an 87 percent chance of winning it, yet Clinton came out ahead by about three million votes. Nonetheless, Norpoth declared victory and enjoyed write-ups as the man who predicted Trump’s victory. “Vindication is a good word for it,” was Norpoth’s Trumpian answer to a Newsday reporter asking how he felt about his performance.

The failure of these predictions was easy to predict. The Brookings Institution’s Thomas Mann, a political science PhD and the former executive director of the American Political Science Association, noted last fall that “models eschewing any public opinion measures of the current political climate or standing of the candidates have been most susceptible to large errors in predicting both the winner and the incumbent-party share of the two-party national vote for the presidency.” And it wasn’t just the bizarre circumstances of 2016 that inspired well-justified skepticism about the fundamentalist faith. Can political scientists, Nate Silver asked in 2012, quoting one book’s claim, “predict winners and losers with amazing accuracy long before the campaigns start?” He concluded, “emphatically not. Some of their forecasts have been better than others, but their track record as a whole is very poor.” On the whole, depending on what models you look at and from what time frame, and whether economists dabbling in fundamentalism count and other issues too dull to merit close examination here, the best that can be said with confidence is that these models are not worse than a naïve assumption that all elections will end up in a 50-50 split between the two major parties.

Damn Lies and Statistics

Indeed, there’s a distinctly Trumpian quality to the self-justifications proffered by the fundamentalists. Just as Trump bragged before a TV interviewer that his transparently phony claim that James Comey’s White House meetings with the president might have been taped had psyched out Comey and played to Trump’s deal-making strengths, so, too, do fundamentalists embrace hyperbolic claims about their forecasting prowess in the service of the greater teleological advancement of their discipline. In the heady fundamentalist heyday of 2010, for example, Sides offered this jumbled defense of the puffery he and his colleagues enjoyed in the mainstream press:

It’s actually valuable to state the economic case as starkly as Klein and Krugman did. In my blogger persona, I sometimes do that too. There is strategic value in doing so. The view of many, if not most, political journalists is miles from where political science stands. If it takes a statement like Klein’s (or some of mine) to pull that view towards one that acknowledges that economic performance is the key driver, then that’s fine with me.

Such complacency obviously works to forgive a multitude of analytic sins—in pretty much the same way that fundamentalists have produced a body of determinist dogma that has been deployed to explain away any and all Democratic failures. Still, there’s an unmistakable irony lurking beneath the results-indifferent self-
satisfaction of Sides and his colleagues. For the main victory of fundamentalism in the marketplace of ideas has nothing to do with the merits of the movement’s work. No, it stems from something these same political scientists say exerts no meaningful influence in elections: “communications strategy.” Democratic Party political operatives probably have more to learn from fundamentalists about marketing than about how elections work in the real world. Professors likes Sides know they can generate a lot of attention when they cast aside the credibility generated by more careful academic colleagues and deploy strategic exaggeration (what Trump himself has memorably termed “truthful hyperbole”) on behalf of their forecasting prowess.

Despite the movement’s many claims to the contrary, fundamentalism is not a discipline founded in objective truth. Models are altered retroactively to belatedly take account of the actual outcome. There is no consistent agreement among the leading figures in the field about what to do about the inconvenient fact that economic statistics published in advance of an election are subject to revision after the election.

The very assumptions underlying the claims of the fundamentalists wouldn’t pass muster in any Introduction to Statistics in the first place. As any freshman stats student should be able to tell you, a data analysis can’t be reliable with a robust number of apples-to-apples units of analysis—the “small N” problem. Even Wikipedia says as much: “For example: a study where countries are the unit of analysis is limited in that are only a limited number of countries in the world (less than 200), less than necessary for some (probabilistic) statistical techniques.”

Two hundred! A lot of these models seeking to predict future presidential election are based upon evidence from less than a dozen past ones—elections that radically differ from one another in the most fundamental ways. Take, for instance, the presidential election of 1976. Back then, an insurgent outsider Democratic nominee, Jimmy Carter, beat the accidental presidential incumbent Gerald Ford, the beneficiary of two major and unprecedented resignations from the Republican executive branch. In order to compare that presidential contest with the dramatically outlying 2016 balloting in an apples-to-apples way, you have to believe a lot of massive changes either didn’t matter or were somehow already baked into the model-makers’ formulae. Start with some demographic data: the 1976 electorate was believed to have been 89 percent white and 1 percent Hispanic, quite unlike 2016. In 1976, both parties contained liberals and conservatives and the party system in the United States was experiencing new convulsions. The once-exclusively Democratic South had lately trended Republican—but that process was now scrambled by the first major party nominee from the deep South in more than a century. In 1976, cable news was a non-factor, let alone the emergence of post-cable media demographics like “cord cutters.” Mobile phones, social media, and online organizing were decades off, as was “fake news” in its many head-spinning iterations. And so on.

Forty years later, for these models to make sense, you have to believe the same thing about a long and dizzying list of cycle-specific contingencies. Here’s but a small sample from the general election of 2016: Wikileaks’ release of hacked Podesta and DNC emails; Anthony Weiner’s sexting problem leading James Comey to send a public letter reopening the investigation into Hillary Clinton’s email server eleven days before the election; Comey’s reticence to unveil the FBI’s investigation of whether Trump’s campaign colluded with Russia; the Clinton high command’s decision to pour resources into Ohio and Iowa while taking Michigan and Wisconsin for granted; and so on. Oh, and Clinton was the first woman ever nominated by a major political party, aiming to succeed the first African American president.

But: 1976 and 2016 certainly both had “weighted-average growth of per capita real disposable personal income over the term.” So why not model them as one singular, rarely deviating study in foreordained outcomes?

One can squint a bit and figure out why fundamentalist pseudoscience might appeal to an economist like Krugman.

Fundamentalists believe the underlying patterns in American elections, or at least the important underlying patterns, to be unchanging. One might think that relying on, for comparison’s sake, election data from an America in which California and Vermont were generally Republican and West Virginia reliably Democratic would give at least some empirical pause to anyone rolling out the exuberant fundamentalist schema to analyze the Clinton-Trump contest. And yet the fundamentalists blindly compared away.

The great question here is just how the fundamentalist movement commanded so much earnest attention in the midst of a truly transformative, high-stakes election. The models adopted and touted by the fundamentalists were never particularly useful for their own stated purpose of predicting election outcomes. And yet their public prominence has continued to grow. Why? Now that’s an analytic problem worth chewing over.

What, We Worry?

President Obama was first a victim of fundamentalism before he was its beneficiary. Recall that Obama was like Charlie Brown to Mitch McConnell and John Boehner’s Lucy, constantly seeking to kick the football of bipartisan accord, be it on health care reform or the “Grand Bargain” over the budget, only to have Republican leaders snatch the ball away at the last moment, snickering to their anti-government base as the president landed again and again on his back. Here, too, it’s hard not to infer that he was influenced by some key first-order assumptions of the fundamentalists. After all, how could mere political decisions influence partisan outcomes? In this schema, the pursuit of bipartisan comity at the expense of partisan villification appears costless. And once the political party he led began to collapse in 2010—thanks in no small part to the systematic and fervid conservative bid to politicize bipartisan failures on multiple legislative fronts—fundamentalism offered Obama a free pass. The joint refrain went something like this: “If Bush’s economy and McConnell’s obstruction made the election inevitable, how can you blame me for losses up and down the ballot?”

Some fundamentalist pundits continued to hew to the one true faith for similar reasons—it provides both an all-purpose rationale for Democratic failure and a welcome release from the demands of exercising actual political agency. President Obama was as popular with the establishment pundits of his own party as any president since at least Reagan. Democratic Party pundits looked up to Obama’s charismatic wonkiness. Excusing Obama’s failures via fundamentalism was much more appealing than concluding that his adoption of their preferred strategies in 2009-2010 had empowered abhorrent leaders like Boehner, McConnell, and Wisconsinites Paul Ryan and Scott Walker.

There’s a distinctly Trumpian quality to the self-justifications proffered by the fundamentalists.

But pundits also relished fundamentalism for its intellectual dividends. The fundamentalist creed gave them an aura of raffish contrarianism—and elevated them above the common run of daily reporters hopelessly spun by “sources.” Fundamentalism gave pundits with a comparative advantage at reading academic papers an “intellectual” basis for claiming that elections were better covered from a computer applying academic papers to current economic statistics than by talking to either voters or political operatives.

And now it’s 2018. The economy is fine, certainly not declining, and consumer confidence is generally higher than it was under Obama. Fundamentalists take pride in not reacting to the events of the day, but how else can anyone explain Trump’s historic, sustained unpopularity in the face of a solid economy?

One pronounced theme of post-election coverage has been that the Trump insurgency represented a rebuke to discredited elites. In the case of elite political science fundamentalists, we can only hope the rebuke lands and stings. For the second time in the span of little more than a decade, traumatized liberals will have to figure out how to piece together a durable Democratic governing coalition atop the smoldering wreckage of a catastrophic Republican presidency. As they undertake this monumental challenge, one of the first orders of business will be to leave Obama’s embrace of fundamentalism where it belongs: in the dustbin of history.