Modern American Catholics reject many Catholic teachings:

According to a recent Gallup Poll, 78 percent of American Catholics
support allowing Catholics to use birth control, 63 percent think
priests should be able to marry, and 55 percent think women should be
ordained as priests. Last week Gallup reported that more Catholics than
non-Catholics believe that homosexual behavior, divorce, and stem-cell
and human-embryo research are morally acceptable.

Question: Why don’t these Catholics switch to a church that shares their doctrinal views?  For example, why not just go Episcopalian? 

Perhaps this seems like a stupid question.  Many Catholics will object, “We can’t ‘just switch,’ because Catholic is what we are.”  But what exactly does this indignant response mean?  It can’t literally mean that Catholic identification is written in stone; the Church has lost adherents over doctrine before.  Remember the Reformation?

What it means, rather, is that in matters of religion people have both doctrinal preferences and denominational preference.  Even if a new religion came along that exactly agreed with their current religion, most people would still strictly prefer their current religions.  From a slightly different perspective, this means that people will – within some range – stay loyal to one religion even though a competing religion is, in purely doctrinal terms, a better fit. 

As a result, religious leaders have some slack to deviate from the rank-and-file.  And when you think about it, using this slack to push unpopular doctrines may be relatively benign.  A religion with a devoted flock could also use its slack to cover up child abuse, knowing that few members will switch service providers if the ugly truth comes out. 

I bring this all up because this story about religion is exactly analogous to the recently-discussed Lindbeck-Weibull model of voter choice.  The key assumption in the Lindbeck-Weibull model is that voters have both policy preferences and party preferences.  Even if Democrats and Republicans had identical platforms, many Republicans would still strictly prefer Republican candidates.  Why?  Because many Republicans are Republican in the same sense that Catholics are Catholic.  Belonging to this group is “part of who they are.”  And if enough people have these party preferences, the result is that political leaders have some slack to deviate from the rank-and-file on matters of both policy and common decency.  Indeed, if party preferences are sufficiently lop-sided, the result is one-party democracy.

The Lindbeck-Weibull model is pretty obscure even in economics.  But the more I reflect on it, the more this simple model seems to explain a lot of facts that other models of religious and political competition can’t.