The Pew Research Center is out with a major new study of religion in America, along with a nifty interactive tool to explore the data. With a staggering 35,000 respondents, there’s no shortage of ways to slice the data. Here in the Boston area, the story mirrors the national trend: The percentage of Christians are declining, and the share of Americans who are not affiliated with any religion is rising.

As the Boston Globe’s Crux notes, Catholicism is bearing more of the brunt of this decline than other denominations. Pew found that for every new member joining the church, 6.5 are leaving. This trend is playing out even in the traditionally Catholic Boston Metro Area, where unaffiliated now slightly outnumber Catholics. (Nonetheless, Boston and Massachusetts are still among the most Catholic areas in the nation.) And, as Crux points out, the demographics of the Catholics who remain are shifting: They are getting older but at the same time less white. Hispanics now comprise 41 percent of American Catholics.

These demographic changes have bearing on the political landscape as well. The identity and ideology of the “Catholic” voter is shifting. We looked at this issue back in 2012, during the U.S. Senate race between Scott Brown and Elizabeth Warren. Overall, Brown led Warren by 7 points among Catholic voters, while Warren led among the nonaffiliated voters. But much of that lead came from white, non-Hispanic Catholics — the traditional profile of a Massachusetts Catholic voter. As the demographics continue to shift, it will be harder and harder to pigeonhole and politick towards a typical “Catholic” voter, even in a state like Massachusetts.


Casino Votes in Massachusetts Breaking Along Income Lines

Yesterday, voters in Brockton approved a casino in the city by the narrowest of margins; yes beat out no by 143 voters, or 1 percent of the 14,000 votes cast. That 50 percent support is on the low end of the 50-55 percent range that MPG president Steve Koczela predicted based on how previous casino votes have correlated with household income, adjusted by the cost of living in the communities where they have been held.


We’ll update our prediction model and this chart with the Brockton results and see what it predicts for upcoming votes in New Bedford and Somerset.


MPG is hiring!

Like the Topline? Interested in polling, politics and policy? Apply for our (paid) internship. Details on our website.




Stiff Upper Lip

The polls missed badly in the UK general elections last week, and the postmortems are flying fast and furious. HuffPost has a great summary with lots of navel gazing links.

In the aftermath, some in Britain are even calling for a ban on pre-election polls altogether.


2016, the Saga Continues

Two national polls tell different stories for Hillary Clinton. The CBS/NYTimes poll saw Clinton’s favorable rating actually increase in the wake of stories about her family foundation. But the NBC/Wall Street Journal found Clinton’s unfavorables ticking up. Still, Clinton remains the only candidate tested that has a net favorable rating, and she wins in hypothetical match-ups against Republican hopefuls.

Up in New Hampshire, UNH and WMUR find no real leader in a crowded Republican field. But Clinton has lost ground to the GOP pack.

And PPP’s latest national Republican primary poll, just out today, finds Scott Walker ahead and Marco Rubio vaulted into second.



The Republican-controlled Congress has a dismal 19 percent approval rating, according to Gallup.

Harvard/Social Sphere pollster John Della Volpe dissects the millennial vote in a Boston Globe editorial.

Over at HuffPost, Kathleen Weldon looks at the history of public opinion on nuclear energy.


——————————-Nerd Alert Tearline———————————


MPG president Steve Koczela is off to sunny Florida for the annual conference of the American Association of Public Opinion Research, or AAPOR for short. Follow the #AAPOR hashtag on Twitter and let the statistical wonkery flow over you.

Steve will be presenting a paper drawn from our WBUR Tracking Poll during last year’s governor’s race. It’s about different ways of asking voters whether they are Democrats, Republicans, or independents, and how those responses differ from how voters are actually registered on the rolls.

Another example of an AAPOR paper: Pew looked at results from one of its major surveys and found differences between telephone and online responses to some questions. This “mode effect” has important implications as more surveys shift from telephone to the web.