IE 11 is not supported. For an optimal experience visit our site on another browser.

Great Lakes states voted against Hillary Clinton, not for Trump

An NBC News/Marist poll shows Michigan, Pennsylvania and Wisconsin might be harder for President Donald Trump this cycle.

WASHINGTON — A lot of things had to fall right for Donald Trump to win the presidency in 2016, but the biggest key was his ability to win important Northern states that had voted Democratic for several cycles. In particular, Trump flipped Michigan, Pennsylvania and Wisconsin, which had voted Democratic in every presidential election since 1992.

The latest NBC News/Marist polls from Michigan and Wisconsin suggest that it may be difficult to repeat that feat in November. The surveys show that Democratic nominee Joe Biden has solid leads over Trump among likely voters in both states — 52 percent to 44 percent in Michigan and 54 percent to 44 percent in Wisconsin.

Those polls are the latest in a stream of surveys in recent weeks that suggest that 2016's results may have been something of an outlier and more about Trump's opponent, Hillary Clinton, than about Trump.

There is reason to see 2016 as Trump's shaking up the electorate in those three critical states. The sheer size of the swing of the vote in them was truly impressive. Trump's winning margins were small, but the changes from four years earlier were large.

In Michigan, the margin went from Democrat Barack Obama's winning by 9 points in 2012 to Trump's winning by 0.2 points in 2016. Pennsylvania went from Obama's winning by 5 points to Trump's winning by 0.7 points. And in Wisconsin, the flip was from Obama by 7 points to Trump by 0.8 points.

Those numbers show a Republican erasing the historic edge of his Democratic opponent in the Great Lakes region. And they suggest that Trump transformed the vote in those key states, if only for one cycle.

But look at the numbers a different way and you see a very different story. Consider the percentage of the vote Trump got in those three states compared to the percentages Republican presidential nominee Mitt Romney got four years earlier.

In each case Trump did better, but not markedly so. In Michigan, Trump did 2 points better than Romney. In Pennsylvania and Wisconsin, Trump did about 1 point better than Romney. In each state, Trump's improvement over Romney accounted for less than a quarter of the total swing.

That is to say, when you look closely at the swings in those states, they were less about voters moving toward Trump than they were about voters moving away from Clinton.

Of course, this fall Clinton is not on the ballot for the Democrats; Biden is. And the current polling averages in those states look a lot more like 2012 than they do 2016.

In Michigan, the FiveThirtyEight.com polling average shows Biden leading Trump by 50 percent to 43 percent, a 7-point edge. In 2012, Obama won Michigan by 9 points.

In Pennsylvania, the polling average shows Biden ahead of Trump by 50 percent to 45 percent, a 5-point edge. In 2012, Obama won Pennsylvania by 5 points.

And in Wisconsin, the polling average shows Biden with 50 percent of the vote and Trump with 44 percent, a 6-point lead. In 2012, Obama won Wisconsin by 7 points.

The similarities are striking, and it's also worth noting that Biden is at 50 percent in each of those states in the FiveThirtyEight averages. In 2016, Clinton didn't get above 47 percent of the vote in any of them on Election Day. In the state polls, Biden looks a lot more like Obama than Clinton.

And beyond those big three states, there are signs that others in Great Lakes/Big Ten country look different this year from four years ago. Iowa and Ohio are good examples. In 2016, Trump won both by large margins — 9 points in Iowa and 8 points in Ohio. But now they both look like toss-ups.

In Iowa, the president is up by 47 percent to 46 percent, a 1-point edge, in the FiveThirtyEight polling average. In Ohio, it's reversed, and Biden holds a narrow 1-point lead, 48 percent to 47 percent in the average. In essence, the races are tied in those states.

It should be noted that Trump's current percentage in those states is almost exactly what Romney got out of them in 2012 — it's 1 point better in Iowa and the same in Ohio.

That's a lot of data and election comparisons, but, taken together, the numbers point to one central theme. While Trump has rightly gotten a lot of credit for flipping key states in 2016, not all of those flips were because of him alone. His opponent seems to have been a major factor, perhaps a bigger factor, and one he won't have helping him in November.

Right now, 2020's poll numbers look more like Obama's 2012 win than Clinton's 2016 loss. And the differences may be about Trump's opponent as much as they are about Trump himself.