What really happened in Maryland

For Democrats, election night 2014 offered a surfeit of unpleasant surprises. Perhaps most surprising of all was the outcome of Maryland’s gubernatorial race (disclosure: our firm was not involved).

After all, Maryland is one of the bluest states in the country — the second most Democratic in Gallup’s party identification data, with Democrats holding 74 percent of state Senate seats, 69 percent of seats in the House of Delegates, 88 percent of U.S. House seats and both seats in the U.S. Senate.

The three public polls in October found Democrat Anthony Brown leading Republican Larry Hogan in the governor’s race by 7, 9 and 13 points. So no one expected Hogan to win by nearly 5 points.

What happened?

One may reasonably wonder why I’m reopening that wound now, six months after the election, when the shock has worn off and we are already well on to 2016. My answer is three-fold.

First, because individual-level data is only now becoming available, we can learn much more about these races now than we could before. While we may be bored with them and on to the next campaign, real learning can only occur now.

Second, the answer bears directly on the debate between persuasion and turnout to which I devoted four columns last year.

Finally, two illuminating pieces of work, one from Catalist chief scientist Yair Ghitza and another by Jared McDonald of the University of Maryland, have recently become available.

Many initial analyses followed the meme I attacked before the election — “it’s all about turnout,” they argued. A New York Times headline summed up the consensus: “Drop in Democratic turnout was the difference in the Maryland governor’s race.” The Baltimore Business Journal (“Low Democratic turnout propelled Larry Hogan to victory in Maryland governor’s race”) and other outlets echoed that lesson.

We now know it’s not true.

McDonald notes that, while turnout in Maryland was lower among Democrats than among Republicans, with 54.5 percent of the actual electorate consisting of registered Democrats, Brown could have won without a single vote from a Republican or an independent.

Moreover, using a post-election Washington Post/University of Maryland poll, he shows that Hogan would still have won if the turnout had matched recent midterms — from 2010, or from 2006, or from 2002.

If turnout was not the culprit, then what was?

Brown simply lost lots of Democrats and was overwhelmed among independents. Some 23 percent of Democrats defected to Hogan, while only 6 percent of Republicans crossed party lines for Brown.

Hogan Democrats were different from other Democrats in several ways, including the fact that they put a higher priority on taxes.

In other races, Maryland independents have leaned toward Democrats, but in this contest they went for Hogan by a massive 61 to 33 percent.

Post-election polls have flaws, of course; fortunately, Ghitza takes a very different approach to the data — using multilevel regression with variables included on Catalist’s data file — and comes to similar conclusions.

Ghitza’s data also indicates turnout was not the malefactor.

Yes, Democratic turnout dropped 8 points from four years ago, when Martin O’Malley was reelected overwhelmingly as governor, but Democrats made up just 2 percent less of the total electorate in 2014 than they had in 2010, with Republicans rising by 2 points. Registered Democrats still made up 54 percent of actual voters, while just 32 percent were Republicans.

Even with the turnout in a presidential year with Barack Obama at the top of the ticket, Brown would have lost if he bled Democrats and Republicans the way he did in 2014.

By Ghitza’s calculation, 21 percent of Democrats defected to Hogan in this race, compared with just 16 percent when O’Malley was reelected. And Hogan did a massive 18 points less well among independents than O’Malley did four years before.

While better turnout would certainly have helped Brown, both these analyses make clear that persuasion made the ultimate difference. And they remind us that it’s worth waiting for the data before jumping to conclusions.

Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the minority leader of the Senate and the Democratic whip in the House.

Whether winning for you means getting more votes than your opponent, selling more product, changing public policy, raising more money or generating more activism, The Mellman Group transforms data into winning strategies.