In past 25 years, Mass. has become more politically polarized

Massachusetts may be a liberal state, but it has some pretty conservative regions. Broad swaths of red covered Central and Southeast Massachusetts in last month’s gubernatorial election, with clusters of blue in the West, the far Cape, and greater Boston.

A generation ago the map looked very different; flecks of red and blue could be found just about everywhere in Massachusetts. But over the last 25 years the state has become more polarized. Political differences that were widely distributed are now more geographically concentrated.

Continue reading below

After the November election, I asked readers for their thoughts on the underlying cause of Massachusetts’ recent political polarization. I then took those theories to the data and tested them. Here’s what I found:

1990 to 2014

One way to see how much the political grography of Massachusetts has changed is by comparing Republican Charlie Baker’s victory last month with Republican Bill Weld’s win in 1990. Those two elections share a number of basic characteristics. In 1990, incumbent Governor Michael Dukakis was stepping down, just as Governor Patrick did this cycle. And in both cases, that vacancy in the corner office left an opening for a moderate Republican.

Yet, Weld’s path to victory looked nothing like Baker’s. Whereas the 2014 map is broken into pretty distinct bands of red and blue, what stands out about the 1990 map is the absence of any obvious geographical pattern. Central Massachusetts was split, as were the East and West.

Data: Two-party vote share. View interactive version

Data: Two-party vote share. View interactive version

If you look at a map showing how much each town has shifted between 1990 and 2014, the pattern is even clearer. Town colors aren’t changing. Whole regions are.

Data: Normalized percentage point change in voting. View interactive version

Do these shifts matter?

It’s important to keep in mind that these are maps of towns, not people. Boston may look about the same size as Sheffield, but on Election Day, Bostonians cast over 100 times as many votes.

Also, whatever forces are reshaping Massachusetts, they’re likely to be different from those that have reshaped US politics. Massachusetts towns can’t be gerrymandered, the way US congressional districts regularly are. And the nationwide red-blue divide reflects the history of race relations in a way that’s not really relevant to what we’re seeing in the Bay State.

Still, there is at least one common concern. The concentration of views in particular regions could mean fewer daily encounters with people who have different views. When MassINC pollster Steve Koczela looked at how Senate results have changed over time, he found that there are more one-party towns in Massachusetts today than at any time since the 1970s.

What caused this shift?

Last month, when I asked readers to share their thoughts, they responded in earnest. Some pointed to income as an explanation, others to a split between urban and rural areas, and still others to commuting patterns and cultural affinities.

Looking at the numbers, though, some of the leading theories just didn’t hold up.

It’s not about changes in income. Over the last 25 years, Massachusetts has shed manufacturing jobs and developed an economy more reliant on technology and health care. It’s reasonable to think that this transition would leave some regions behind, and that this would affect their residents’ politics. But data show almost no correlation between changes in the median income of particular towns and changes in their voting habits. Palmer and Dalton, for instance, are two towns where incomes have declined significantly since 1990. One of them has gotten more conservative, the other more liberal (you can guess which is which from the geography).

It’s not about a growing urban/rural divide. Population density is actually less important today than it was in 1990. Perhaps the cleanest way to think of this is just to keep in mind that those towns in the Berkshires which are trending left are often just as rural as the right-moving towns of Central Massachusetts.

What are some other possible causes?

One possibility, raised by reader “salemgal,” is that people are increasingly choosing to live near like-minded folks. As salemgal put it: “Assuming I had the money, as someone who leans a bit left of center, I would not seek to live in Weston and would much prefer Newton or Lincoln where I would have more in common with my neighbors.”

Another possibility is that education, more than income, is driving political preferences. Or, alternatively, that declining union membership is playing a key role. I’m eager to look into these more deeply, but first wanted to hear again from readers — to see if the findings thus far spark any new thoughts. Is it about age? Economics? Social preferences? Something else? Email me or share your thoughts in the comments. And if my further research proves you right, you’ll get full credit.

Evan Horowitz digs through data to find information that illuminates the policy issues facing Massachusetts and the United States. He can be reached at Follow him on Twitter @GlobeHorowitz