Feed on
  1. American church membership has risen throughout most of the past two centuries, from 17 percent of the population at the time of the American Revolution to 34 percent in the mid-19th century to over 60 percent today.
  2. The fraction of the U.S. population employed as clergy has remained stable at 1.2 per thousand for the past 150years. It seems to have been on the increase since 1970 (see figure below).
  3. Since the 1930s, the percentage of Americans attending church every week has remained stable (only Catholic attendance dropped after the changes in the Vatican in the 1960s).
  4. Religious contributions have remained stable at one percent of GDP since the 1950s, which accounts for half of all charitable giving in America.
  5. Religion seems not to be the province of the poor or uninformed. Rates of religious belief and religious activity (in cross-sectional analyses) do not decline with income, and rise with education. But styles of religion certainly do vary.
  6. Most members of “extremist” sects do not show signs of mental disorders.
  7. College professors are less religious than the population writ large.
  8. Worldwide, and in America, the fastest growing sects are strict, sectarian and theologically conservative.

That’s from a 1998 review by Larry Iannaccone. I’d love to see an update. By the way, an area of study ripe for scholarship would be a public choice analysis of religious institutions – particularly of the large, dominant institutions in a country.

3 Responses to “Empirical Regularities in Religion”

  1. Brad Samples says:

    I cannot access the article, how does he define ‘church’? Does it include only ‘mainstream’ religions (I’m thinking Abrahamic ones, Hinduism, Buddhism, maybe some smaller Asian ones) or does it cover paganism and the other fringe religions?

  2. Michael says:

    I had kinda the same question about what constitutes a church. Looking at the article (it’s JSTOR, so I can access it from my school, but you need to have some sort of registration), it appears that it is survey data, so a church is defined by the surveyee instead of the surveyer.
    Another aspect I wonder is how churches have changed over time. It could be that most churches back in the 1700 & 1800’s were very “fire and brimstone” in the condemnation of sin. Today there are several churches that are more of a feel-good place where sin isn’t really discussed. Naturally, more people would be willing to go to these churches. I have had discussions with people on this, and there are many who are all to willing to sacrifice doctrine in order to get a body in the pew. The survey does include non-Christian religions, too. And being highly biased about this area, I believe there is a significant difference between types of churches. None-the-less, it is an interesting article.

  3. wintercow20 says:

    We would need to look at the Gallup polls he is citing. I believe it is a liberal interpretation of church. While I agree that there are serious definitions of what constitutes a church that “condemns sin”, in terms of measuring the “religiosity” of Americans I am not sure a measure excluding more liberal definitions would be accurate too. The reason being that choices 100 and 200 years ago were restricted for various reasons, and societal factors played a role in what church you attended. So I am not sure that strong adherence to “traditional” religions back then are representative of “true” religiosity … whatever that means.

Leave a Reply