Iowa gave Barack Obama a resounding 9.5-point victory over John McCain in 2008. Four years later, Obama’s margin shrunk to 5.8 points against Mitt Romney. But in 2016, something odd happened.

Donald Trump carried Iowa by 9.4 points — a dramatic change in the state’s recent voting behavior and close to the same winning margin as Obama’s eight years earlier.

Iowa had gone Republican before, of course. It voted narrowly for George W. Bush in 2004, and it backed Ronald Reagan twice, in 1980 and 1984.

But Michael Dukakis carried it comfortably in 1988, and Al Gore’s narrow victory in 2000 seemed to suggest Iowa was shifting ever so slightly toward the Democrats. That is what made 2016 so noteworthy.

Did Trump’s solid victory say something about the state’s fundamental partisan bent?

Yes and no. As reporter Paige Godden and the folks at Iowa Starting Line note, a stunning 31 Iowa counties voted twice for Obama but flipped to Trump in 2016. Their statewide map shows the bulk of the shifts occurring in eastern Iowa (and particularly northeast Iowa), areas of traditional Democratic strength.

While much of western Iowa has more in common with Nebraska and South Dakota (including its Republican bent), northeast Iowa has more in common with Minnesota and Wisconsin, and eastern Iowa looks toward Illinois.

The shift in eastern and northeast Iowa in 2016 probably reflects the shift seen elsewhere among working-class voters who were once reliably Democratic but responded enthusiastically to Trump’s message.

Blue-collar shift

According to a 2017 analysis by the Center for Economic and Policy Research, Iowa ranked third in the nation with blue-collar jobs as a share of total nonfarm employment in each state. (The group defined blue-collar jobs as those in goods-producing sectors — manufacturing, construction, mining and logging.)

Given that, maybe Iowa’s dramatic shift was not so surprising. But there is more to the story. And it suggests that the state’s swing was not quite as dramatic as it first looked.

Trump drew almost 801,000 votes in Iowa, more than 20,000 fewer votes than Obama did in 2008 (823,000 votes) or in 2012 (829,000 votes).

Trump’s showing, in percentage terms, was right in line with those presidential nominees who carried Iowa in the previous five contests.

He drew 51.2 percent, while Obama drew 52 percent in 2012 (against Romney) and 53.9 percent in 2008 (against McCain). George W. Bush drew 49.9 percent in 2004 (against John Kerry) and Gore drew 48.5 percent in 2000 against Bush.

In 1996, Bill Clinton drew 50.3 percent in Iowa against Bob Dole.

So while Trump’s margin of victory was large (9.4 points), especially for a once-hypercompetitive state, the percentage of the vote that he received (51.2 percent) was not particularly noteworthy.

How could that be? As in a handful of other competitive Midwest and Great Lakes states, Trump’s margin of victory was more a reflection of Hillary Clinton’s weak performance than his own strength.

Clinton drew just 41.7 percent of the vote in Iowa, a showing far worse than those by other recent losing candidates, including Romney (46.2 percent), McCain (44.4 percent), Kerry (49.2 percent) and George W. Bush (48.2 percent).

The only recent losing nominees to draw a lower vote percentage in Iowa than Clinton were Dole (39.9 percent in 1996) and George H.W. Bush (37.3 percent in 1992), both of whom lost a substantial number of votes to independent Ross Perot.

Looking ahead

Is this good news for Democrats, putting Iowa back in play in 2020?

Hillary Clinton’s showing in Iowa reflected a unique weakness, and Democrats have a chance to nominate someone who doesn’t have her baggage, which would improve their prospects in the state.

But Trump’s margin was substantial, and Democrats would need to bring out voters who stayed home in 2016 and also win back some former Democrats who backed Trump three years ago. That’s quite a challenge.

So far, there is little reason to believe that more than a trickle of 2016 Trump voters are ready to defect (or, more correctly, return to the Democratic Party), and that, plus the state’s rural Republican voters throughout the state, make it difficult to see Iowa supporting the Democratic nominee in November.

Of course, 2020 certainly should see a much closer race in Iowa if the Democrats nominate someone who can turn out new voters, rally the party base and appeal to swing voters upset with Trump’s language and governing style.

Trump’s margin in Iowa is likely to shrink in November to the mid- or even low-single digits. That means he is still likely to carry the state (barring new developments), but it also means that Iowa is worth keeping an eye on, if only as a test case as to whether Democrats are rallying behind their party’s nominee.