Economic fatalism and elections

It's an old myth that economics determine elections.

Yet this myth has recently been recycled from New York Times' columnist Paul Krugman to The Washington Post's Ezra Klein. By implication, their argument absolves Democrats of responsibility for their formidable headwinds. But it also evokes deeper questions that transcend the coming election cycle and concern how we understand every cycle.

This is, actually, a billion-dollar question. That's the amount spent on national campaigns in the U.S. each cycle – to say nothing of political journalism.

The economic fatalists see elections in simple terms. "Midterm elections, where turnout is crucial, aren't quite like presidential elections, where the economy is all," Krugman wrote Monday. Keep the word "all" in mind. One week earlier, Klein argued this same case. Headline: "It's Always the Economy, Stupid."

But it's not always the economy.

In fact, the economy is not always the dominant environmental factor. Think 1952, 1968 and 2000 in presidential years. Think 1966, 1974 and 2006 in midterm cycles. Now hold the thought. We'll return to it.

Political scientists have forecast elections for generations. These professorial seers rely on myriad environmental elements: presidential approval ratings, incumbency, seismic social and national security issues, as well as the economy.

"The economy is a junior partner compared to preference polling and incumbency," said the University of Buffalo-SUNY's James Campbell of forecasting models. Campbell has spent years analyzing different models.

Emory political scientist Alan Abramowitz agreed. In his words, "The point is not that the economy isn't important, of course, but that it's one factor and not always the most important one." Abramowitz is also a leading expert on election forecasting.

Why is it only "one factor?" Consider Klein's case.

Klein heavily relied on a graph published along with his story. It illustrated the correlation between income growth and the vote. But one should always look at the outliers. Two jumped out on this graph: 1952 and 1968. Both are classic examples of good economies not saving the party that controlled the White House.

The big factors of the 1952 race: Korean War, a deeply unpopular incumbent and Dwight Eisenhower's towering personality. Had Ike run as a Democrat, once thought possible, Democrats might have had their sixth consecutive term.

The good economy proved of little weight in the late sixties as well. In the 1966 midterm, Democrats were sunk by the unpopularity of Lyndon Johnson, his Great Society and social issues. Two years later, those social issues (race, urban upheaval) and the Vietnam War particularly took to the fore of the American mind.

The mistake arises from the absolutist view of economic influence. George Washington University's John Sides created the graph in Klein's story. Sides tellingly drew a more conventional conclusion from his research. "Elections are strongly related to the economy," he wrote in an email exchange. And in Sides' terms, few experts would disagree.

But Klein looked at the upper right of Sides' graph, saw a correlation and argued grand causation. There were three landslide elections in that corner: 1964, 1972 and 1984. "We understand elections in terms of candidates, but it seems awfully convenient that the three worst candidates happened to end up in the three most impossible election years," Klein wrote.

It's also "awfully convenient" to ignore that all three of these campaigns had popular incumbents. There were those other factors as well, like the assassination of John Kennedy.

Richard Nixon had a good economy in 1972. But other presidents had better. Yet George McGovern won one state, and only about three-in-ten whites that year. Replace McGovern with Henry Jackson and, to borrow from a young Pat Buchanan, Nixon's strategy of "square America" versus "radical America" would have been significantly undercut. Replace Barry Goldwater with Nelson Rockefeller; 1964 is a closer race.

The economy was relevant in these years. For these long shots, strong economies made their long odds longer.

Yet the more one looks at the data, the more economic fatalists' larger case falls apart. The Post graph relied on real disposable income. That's Americans' after tax income, when adjusted for inflation. It includes benefits packages, like health insurance. Studies suggest it's the best of imperfect political indicators (including unemployment).

Income growth correlates to election outcomes. But the correlation is weaker than Klein contends. It's also far from determinative.

Real income growth, per capita, was actually larger in the 2000 election year than in 1972. If income growth is destiny, 2000 should have been the landslide. Income growth was significantly larger in 1988 than the greater blowout of 1980. This correlation is even weaker in midterm years.

Yet no serious forecaster disregards economic omens. This is also why the president's approval rating is so helpful. It incorporates economic views. Voters assess the status quo based on their priorities. The economy tops the list in these hard times. But in other times, say Watergate's impact in 1974, other issues move up the list.

That returns us to the present debate. How much does economic fate determine Democrats' near-term fate?

The president has kept his campaign promises. Yet those promises were made before the stock market crash. Imagine an investor approaching the market today as if there was no crash.

After all, 2008 was that rare exception. Barack Obama owed his mandate to the market crash; he sustained a majority in the Gallup tracking poll only after Sept 15. But by spring 2009, the president invested that mandate in health care. The president's core priority differed from Americans' priority.

To be fair, Krugman has not argued Obama is powerless before this economic storm. Krugman advocated a more economically engaged president–when Obama was popular–and larger stimulus.

But Obama did not focus year one on the economy. He did not push that new New Deal to ease the blue-collar crisis, as others and myself advocated. Instead, China implemented a New Deal like stimulus. It's one reason they recovered far more quickly.

By contrast, Franklin Roosevelt was consumed by the economy in year one. The New Deal largely chopped away at the Great Depression after 1934. But Americans believed FDR was doing what he could. It's a key reason Democrats bucked trends and won the 1934 midterm.

Political fortunetellers deserve healthy skepticism partly because of this human factor. Models were notoriously wrong in years like 1982 and 2000; Al Gore was expected to defeat Bush by a clear, some said large, margin. Why were these models wrong? A critical reason, they relied too heavily on the economy.

Klein wrote that academic forecasters get, "within 2 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."

The most widely known economic prognosticator is Yale economist Ray Fair. Even Fair does not only rely on the economy. But he does rely on it more than most. Fair correctly predicted George Bush's victory in 1988 by, yes, about 2 points. But by 1992, Fair wrongly forecast Bush's vote margin by nearly double digits. He predicted Al Gore would win with 51%. By 2004, Fair once again greatly overshot the GOP vote.

The best models in one cycle are not necessarily the most accurate in the following cycles. And several models are significantly more accurate than Fair's. One reason, they rely less heavily on the economy.

The economic fatalists are demonstrably wrong. But forecasters are right to note the large influence of environmental factors.

Campaigns are however consequential. Political operatives and reporters (often greatly) exaggerate that consequence. But as James Campbell calculated: the frontrunner's lead in June of the election year has declined or evaporated by Election Day in 13 of 16 elections from 1948 to 2008. Campbell estimates that "campaign effects" decide about one of every five elections.

This is why politics is also a game of inches. Campaigns matter on the margins. But many a president has been made on those margins.

© RealClearPolitics.com