From the Globe, beginning with this article on the socio-economic distribution of the closed parishes

The rate of closings is higher in nonwhite neighborhoods. The typical parish being closed or merged is in a neighborhood that is 16 percent nonwhite, while the typical parish being spared is 9 percent nonwhite. And parishes with a non-English Mass had a 27 percent chance of being affected, compared to 19 percent for the rest.

Churches that serve immigrant populations and offer non-English Masses set to be closed include the only parish with German and Latin Masses (Holy Trinity in Boston). Spared are all the churches with a Mass specifically for the Cambodian, Cantonese, Cape Verdean, Creole, Korean, Mandarin, Nigerian, or Vietnamese communities.

The neighborhoods losing churches are no poorer than those keeping theirs. The typical household income in the neighborhoods around the parishes being closed or merged ($52,900) is nearly identical to the income in the neighborhoods of parishes remaining open ($53,300). The Globe’s measurement is imprecise, using the 2000 Census income for residents of the census tract containing the parish address, not the actual parish boundaries.

The economic fairness was achieved through O’Malley’s cluster system, in which groups of neighboring parishes were forced to offer up at least one parish for possible closure. The most affluent parish in the cluster was targeted just as often as the poorest, the Globe found. O’Malley generally followed those recommendations.

What most distinguishes the closing churches is poor attendance, as judged by a census taken at Mass on several Sundays last October. The closing and merged churches had an average attendance of 559 that month, compared with 1,068 at the churches remaining open. In most clusters, the least-attended church was closed.

And continuing with more articles linked on the sidebar.

Boston Herald coverage here.

More from Beliefnet and our partners