Donald Trump’s compulsion to exaggerate the magnitude of his election victory––talking about a landslide that didn’t happen and miscounting electoral votes––has distracted from the true strategic importance of his triumph.

From one perspective, the size of Trump’s win wasn’t so impressive. He received 45.9 percent of the popular vote––two points less than his opponent and four points shy of a majority––and he captured fewer electoral votes than the winners of seven of the last nine elections.

From another angle, Trump’s win was very impressive. He won the battleground states of Ohio, Florida, North Carolina and Iowa––states that swing back and forth by small margins. But, more importantly, he cracked the Democratic “firewall” states of Michigan, Pennsylvania and Wisconsin. He put these 46 electoral votes into the GOP basket for the first time since Ronald Reagan.

Trump was the first Republican presidential nominee since 1988 to carry Michigan and Pennsylvania, something George W. Bush, Mitt Romney, John McCain and Bob Dole failed to do. He was also the first Republican since 1984 to snag Wisconsin.

Michigan––the hardest of these states for the GOP to crack––is a case in point. Barack Obama won it in 2008 by 16 points, and in 2012 by 10 points against Mitt Romney, who had deep familial roots in the state. Bill Clinton easily won Michigan in his two successful presidential bids, by 14 points in 1992 and by 13 points in 1996. Even Al Gore and John Kerry carried it.

Going into 2016, Michigan looked like a lost cause for the GOP. Trump turned the tables by reaching beyond the Republican base and upping his party’s vote among not only independents and moderates, but also among Democrats, liberals, and union members.

According to exit polling, Romney won Michigan’s independents by only a point. Four years later, Trump ran up an eye-popping 16-point margin. Trump also did seven points better among liberals and five points better among Democrats than did Romney.

Digging deeper, Romney lost union households in Michigan by 33 points. Trump slashed that Democratic advantage to only 13 points. Moreover, Trump did 11 points better than Romney with voters who want government to “do more.”

At the same time, Trump received a lesser share of Republicans in Michigan than did Romney (90 percent vs. 96 percent), and a smaller slice of conservatives (79 percent vs. 83 percent).

Trump’s strength with white men was greater than Romney’s (64 vs. 58 percent), but his campaign did not spark the massive white male turnout that observers assumed. In fact, white men made up a smaller sector of Michigan’s electorate when Trump led the GOP ticket than when Romney did (36 percent vs. 38 percent.)

It was Trump’s change message that gave him victory: 83 percent of voters in Michigan and Pennsylvania and 84 percent in Wisconsin said Trump, and not Hillary Clinton, would bring change. And here’s the kicker: More voters said they wanted change from their next president than either the right experience or good judgment.

The big question for the future of American politics is whether Trump’s capacity to win blue states is a one-trick pony, based on the unique circumstances of 2016, or the start of a major shift?

Political scientists tell us that great partisan realignments endure over multiple elections, so it’s premature to determine whether such a shuffling of the deck is occurring. The viability of Trump’s electoral coalition depends upon his success as president, just as the Republican realignment of the 1860s depended upon Lincoln’s leadership and the Democratic realignment of the 1930s rested upon FDR’s popularity.

Despite the frenetic early weeks of the new administration, Trump and GOP congressional leaders have yet to hack their way through lots of thorny issues. How they handle tax reform, health care, budget cuts, immigration policy, infrastructure funding and a myriad of potentially explosive investigations will make or break them.

Another unsettled question is whether Democrats will learn how to be an effective opposition party in these unpredictable times or whether they’ll move so far left that they slide off the end of the ideological see-saw?

If the 2016 election teaches us anything, it is that each party needs to attract voters beyond its core base. That’s how Democrats can rebuild their electoral-vote firewall. It’s also how Republicans can dismantle it.

Ron Faucheux is a political analyst and author. He publishes LunchtimePolitics.com, a daily newsletter on polls, and runs Clarus Research Group, a nonpartisan survey research firm.