« do as I say, not as I do | Main | April What? An Open Thread »

April 02, 2011

Comments

so what about the hypothesis that Catholics make their neighbours more liberal?

Or maybe culturally diverse areas are more liberal.

I'm not that certain that a vote for Obama was a vote for liberalism -- however defined. Among a lot of us it was a vote against John McCain and Sarah Palin, because Obama is not liberal and never has been. He cast himself as middle of the road, but his policy history was to the right of that, and has continued so. The lesser of two evils is not necessarily a positive good.

The other half of this is the idea that Catholics = liberalism. Since most liberal attitudes and views within the Church have been stomped out by the current and immediately past pope, and the openness of Vatican II has been largely closed down again, and many of the people who were happy with Vatican II have left, there's a far higher proportion of conservative Catholics now than there were, say, twenty or thirty years ago. They're the ones who are still there.

Either Pew's demographic data has serious differences, methodology-wise, with that of a Gallup survey a mere 3 years prior, or there was a HUGE influx of Catholics into the state of California.

I'd want to know why there's a disconnect. Gallup, back in 2004, said 27% of the population of California were Catholic. Pew's results say that 1.4 million Catholics moved into the state, or there was a mass conversion.

Poll methodology could account. There's a large-ish difference between people who claim to be Catholics, and people who attend mass other than on Easter and Christmas.

That aside, it's completely unsurprising to me that Catholics might have different preferences in different regions of the country.

Slarti:

Gallup does not mention conducting its interviews in languages other than English. The Pew survey does interviews in both English and Spanish. My bet is that Gallup is substantially undercounting Catholics (and others) whose native language is Spanish -- a group that is extremely large in California.

To a large degree, I have to agree with fiddler. Obama is a liberal only if you assume that McCain and Palin are moderates. On anything that I would consider a reasonable scale (say the average of political philosophies over the last 30-40 years) Obama counts as definitely middle of the road. He's somewhat liberal on some issues, somewhat conservativce on others.

I can see only two ways to define Obama as liberal:
1) Define moderate/centerist as where McCain is. (Which might have been true in 2000, but definitely was not in 2008.)
2) Define all blacks as liberal by definition. (With, perhaps, a few token exceptions for blacks on the really extreme right.)
Other than that, Obama's statements, and especially his actions since taking office, look really, really centerist. A few exceptions, sure, but exceptions both right and left of center.

Oh, yes. And note that on the subject of Same Sex Marriage, Obama is actually slightly to the right of what polls as the average American view. He can see some kind of official recognition of gay relationships, but is not ready to go with gay marriage.

Among a lot of us it was a vote against John McCain and Sarah Palin [....]
Sure.
[...] because Obama is not liberal and never has been. He cast himself as middle of the road, but his policy history was to the right of that, and has continued so.
This is one of those depends-where-you're-coming-from, where-are-you-on-axes/spectrum, questions.

It's helpful if you define "middle of the road," "right," and "liberal" in something clearer than "this is how I think" terms. What are your personal definitions of these terms, if you have time to give a short version?

Because otherwise, golly this leaves quite an open-ended debate. Which may be your intentional comment hook. :-)

If not, it sure can work. :-)

It's very arguable! :-)

Remember, Bill Ayers still controls Obama!

More sanely, I think it's perfectly defensible to describe Obama's history as pragmatically liberal.

For benchmarking, have you read Dreams From My Father?

Crap, I left the italic tag open, and I don't have the password to fix it. Doctor Science, I apologize, but you or Slart will have to go into my comment and close my open tag so all browsers will see it as fixed.

What I just did by fixing the following in this comment will fix the general problem, which affects the sidebar as well, for some browsers, but not others. Only going into my previous comment will be a complete fix.

Thanks, and sorry about that!

Done! bwa-ha-ha, the power.

As I said, votes for Obama rather than McCain are a *proxy* for liberalism -- they're county-level data that will, to a considerable degree, correlate with support for LGBT rights. Not completely, but there is no county-level data set that is better.

fiddler:

The other half of this is the idea that Catholics = liberalism. Since most liberal attitudes and views within the Church have been stomped out by the current and immediately past pope, and the openness of Vatican II has been largely closed down again, and many of the people who were happy with Vatican II have left, there's a far higher proportion of conservative Catholics now than there were, say, twenty or thirty years ago. They're the ones who are still there.
What idea that "Catholics=liberalism"? The map data show that Catholicism and liberalism tend to co-occur at the population level. The report says that Catholics are not more conservative about LGBT rights than other people in their states, and that in the country as a whole Catholics tend to be more liberal on this issue. What are you disagreeing with?

Back in my home town in Southern Mississippi, the number of Catholics has increased dramatically, if noting the number of people sporting ash on their forehead at the Local Walmart during Lent was any indication, so for at least that area, the 2000 map seems to miss that out. That may be a Katrina phenomenon, but it makes me wonder about the 10 year gap in other places, which, along with the underreporting phenomena noted above, could hide some interesting changes.

I'm also wondering about the composition of US Catholics and their beliefs. It has been said that Hispanics tend to be conservative about social issues, which I assume includes those who have arrived more recently, However, Latin America has been the center of liberation theology and I wonder if some immigrants take up some of those issues, which would make them more 'liberal', if you could total up liberal beliefs and put them on a scale. I don't know if liberation theology is sympathetic to LGTB issues but I wonder if it would have an effect on those issues. It may be the case that folks who might be inclined to view things in terms of liberation theology might not be among the likely people to come to the US, or might their status prevent them from influencing things in their parishes. I'm thinking of the outsize influence that European immigrants had in anarchist/socialist movements at the turn of the century, but it may be that the perception of recent immigrant minorities as illegal might prevent the kind of leadership that you saw with Emma Goldman and Johan Most.

This (the breakdown of Catholics into various beliefs in Church doctrine, which then may be viewed as being liberal or conservative) may be what fiddler is getting at when suggesting Catholicism=liberalism. I don't know how you would go about presenting some sort of data on this in a way that wouldn't invite criticism, but I would note that when I see some opinion piece by a person that lists their Catholic credentials in major news outlets, it is often pretty much centered in a US conservative outlook. (there is one guy in particular, often appears as a talking head, a real blowhard, but google searches fail me. Still, is he just chosen because he is controversial or because he represents a substantial opinion among people in the Church?)

In trying to find the name, I came across this blog post that was a retrospective by a person who was 'the Catholic writer' for the LSU student newspaper. I'm not claiming he or the reaction to him is representative, Louisiana has a unique population in terms of religion and in terms of internal state demographics, so it is a unique case, but seeing the mix of opinions he has, it makes me wonder about the mix of opinions that we might define (or reject) as prototypically liberal or prototypical Catholic

Gallup does not mention conducting its interviews in languages other than English. The Pew survey does interviews in both English and Spanish. My bet is that Gallup is substantially undercounting Catholics (and others) whose native language is Spanish -- a group that is extremely large in California.

That's a reasonable explanation; thanks. But the demographics of the rest (non-Catholic, IOW) of the state of California would still have to compare with that of Michigan, which I'd guess it probably doesn't.

Quibble:

Your "General" columns should take the Catholic parts of the population out if you want a compaarison between Catholics and others. Else you count Catholics twice. You really want "other," I think.

This doesn't matter much here. Being a sport, I recalculated the tables that way and came up with 35%, 45%, and 50% of "General" under SSM, and 60%, 60%, 73% under CU. The "not a sin" numbers are 42% and 39%.

The biggest change is in the national "General" numbers. Not a big deal as to your overall point, maybe, but it would be nice to know the sample sizes. Also, I have a pet peeve. I really would like to see the first decimal place on stuff like this, as well as most political polls. There's a difference between 51.4-48.6 and 50.6-49.4.

I was Catholic until my Deacon informed me he would tolerate no "Liberal" nonsense in "His" church. Now I am just Liberal and Progressive and, regardless of what the Conservatives have to say, I am Christian! (To the Conservatives I am a Left Wing Liberal Nut Case who is both Godless and Mentally Deranged.) Our agenda at our Conservative Diocese was anti-gay, anti-abortion and anti-Liberal everything. I wondered what ever happened to Jesus at church. Someone kidnapped him from the cross and strapped Him in the voting booth. That's why I avoid churches like the plague today. No "Studies, No Statistics...Just the hard facts of experience, folks."

http://thelimingview.blogspot.com/

Done! bwa-ha-ha, the power.
Thanks!

Drink deeply!

I'd have to work a while to come up with a definition of liberal more rigorous than I know it when I see it, but I have been watching politics for a few years. Long enough to remember Eisenhower dimly. Since on his policy positions Obama falls somewhere between Nixon and Regan, I don't see how he fits any definition liberal. Except perhaps the one used by most of the mass media, which seems to be "anyone not insanely reactionary".

there is one guy in particular, often appears as a talking head, a real blowhard, but google searches fail me.

Bill Donohue?

How liberal Obama actually is has little bearing on the usability of votes for Obama as a regional proxy for liberalism. In 2008, liberal voters generally voted for Obama, except for the small fraction who voted for Green or other small-party or independent candidates. Conservative voters were much more likely to vote for McCain. That correlation is what the original post is using the map for.

Eyeballing the map, it looks as if Obama votes and Catholic-heavy regions match up well, with the following exceptions:

1. Majority-African-American areas in the South, which were Obama-voting but not Catholic (which might actually be a region where Obama votes don't track cultural liberalism);

2. some white liberal areas on the West Coast, also Obama-voting but not Catholic;

3. Louisiana, more Catholic than average but McCain-voting.

All this seems to make logical sense.

Comparing US and European political conditions regarding LGBT rights implemented can show effects of catholics on liberalism. Western Europe is in general majority catholic population except GB.
The countries that are more conservative are those where higher percentages of non-catholics is causing reactionary defensive conservativeness, in order to stop the change into protestants. Like Germany and Switzerland. Countries like Spain, Ireland where catholics themselves are more conservative are due to historical nation preservation impulses. Ireland from English protestants, and Spain from Arab attacks and also from 1936 Civil War.
Socialism, LGBT rights, no death sentences and drug use freedoms shows much stronger European then US liberalism. It can be construed as effects of catholic liberalism.

I forgot to mention abortion rights as European stronger liberalism. Of course i am talking about population, not about church hierarchy, which is highly conservative due to desperately trying to preserve itself against what they preach. Jesus's teachings are very strongly against hierarchy.

What idea that "Catholics=liberalism"?
Since your point, as I took it -- "the most you can say is that Catholics don't seem different from the non-Catholics they live among" was that Catholicism and political belief are fairly orthogonal, I won't disagree with you, but simply note that when I see the words "Catholic" and "liberal" juxtaposed -- although "Catholic" and "leftist" might be better, and even better would be "Catholic" and "social justice" -- I immediately think of Dorothy Day.

Bill Donahue?

pin pon! Matt goes on to Final Jeopardy!

NYTimes on him here

the perennially indignant president of the Catholic League for Religious and Civil Rights

Nice tun of phrase.

Here is the man speaking for himself.

Sexual libertines, from the Marquis de Sade to radical gay activists, have sought to pervert society by acting out on their own perversions. What motivates them most of all is a pathological hatred of Christianity. They know, deep down, that what they are doing is wrong, and they shudder at the dreaded words, "Thou Shalt Not." But they continue with their death-style anyway.

Love the title on this one Criticism of the Catholic Church is unfair.

And one more Catholic Church's issue is homosexuality, not pedophilia which explains that the reason the Church has so many problems is because of homosexuality.

I realize he is essentially self appointed, so I can't draw conclusions about Catholics from him, but still, I wonder how much of it would be supported by the Catholic Church.

Of course, to prove that I am not a serious person, when I think of Catholic anything, I always first think of Sally Field

I have no scientific data but the impression I got over the years is that in the US the members of the catholic base ('laypeople') are more liberal than Juan Average world catholic while the clerus is more conservative. Over here in Germany it is a bit more complicated. We have both large groups of true liberals and extreme reactionaries in both the laypeople and the clerus. The Bischofskonferenz (Conference of Catholic Bishops)is a notorious battlefield. The liberals have a slight majority but the reactionaries call on Rome all the time to counter that). I see no clear geographical pattern except for some proverbial 'black' spots where it would be very unwise to be on certain days of the year, if you are not a catholic fanatic.
---
The US may also be a special case because of the strong conservative protestant presence. In Europe protestants are on average more liberal than catholics in the same area. Maybe US catholics are more or less the same as over here and the difference is that US protestants are much more conservative than their European counterparts.

Dr science, did you come to your conclusions by applying the chi squared statistic (to test for significance) to the data in your table? It would be the appropriate scientific thing to do. If not, I'm not sure that there is real difference in some of those %s.

Doctor Science, you asked: What idea that "Catholics=liberalism"?

Quoting the post:
If votes for Obama are a proxy for liberalism -- and I don't see why they shouldn't be -- then Catholicism and liberalism tend to co-occur: counties with a lot of liberals frequently also have a lot of Catholics, and vice versa....

Where am I finding that Catholicism in your post = liberalism? Through logic. If votes for Obama = liberalism, and counties with many Catholics vote liberal, then the Catholics in those counties (if they are the majority, or a plurality, which isn't given) must be voting liberally -- unless there is some way to sort out both religious affiliation and the strength of piety within the population of voters, which I think may not be possible.

Speaking as a former Catholic: yes, Catholics in different parts of the country probably vote differently -- because Catholicism has largely been a religion of immigrants from a wide variety of countries. Even though up to 50 years ago all of the Masses were in Latin, there were distinct cultural differences among different immigrant groups, enough that in many cities there are a lot more Catholic churches than one might otherwise expect: one built by the Italian immigrants, and one built by the Irish immigrants, and one built by the Polish immigrants, and so on for each nationality. There is of course a difference between the number of church members in a census and the ones in the pews, in every generation. The working-class Catholics in the Northeast may have more similar political views regardless of culture because many of them worked in factories and joined unions. I would expect that Catholics in the Southwest, with a different cultural history and social background, might vote more conservatively than Catholics in the East.
Also, for what it's worth, Obama did receive endorsement from major unions, which may have influenced working-class Catholic voters despite the outsourcing of much formerly American factory work in the past few years.

the chi squared statistic (to test for significance) to the data in your table

avedis, I've always heard it referred to as the chi-square test, though I am only familiar with it from linguistics, so I'm wondering what field that you do that used the term in that way?

I go either way, so to speak.

For the purposes of conversation, you should consider my field to be parameter estimation.

fiddler:

OK, I see what you mean -- but you're not being careful enough for my taste, you're using "=" loosely in a conversation where we're trying to figure out what "=" means.

If votes for Obama = liberalism

This is not what I said. I said they are a *proxy* for liberalism, they can be expected to co-occur -- specifically, the set of liberals is almost entirely contained within the set of Obama voters. We see that the set of Catholics strongly overlaps with Obama voters, and thus can be expected to strongly overlap the set of liberals.

What I'm seeing in this discussion is a very great reluctance for people to accept that there is a null hypothesis, much less get to the point of deciding whether it's been disproved or not. The null hypothesis is *always* "nothing is happening", and in this case that translates to "religion doesn't make a difference".

I am contending that that's what we're seeing: religion doesn't make a difference. Or at least, the data do not show religion making a difference. We may *feel* that religion is a driver on these issues, but that's not supported by the actual evidence.

Bernard:

Since I'm taking my figures from the Public Religion Research report, I'm using their terminology. You may want to dig in a bit deeper to figure out exactly how they derived their numbers.

Also, I have a pet peeve. I really would like to see the first decimal place on stuff like this, as well as most political polls.

Alas, that usually would be a lie. The margin of error in these polls is much too large for decimal values to be meaninful. On the polls the PPRI did themselves:

The margin of error for the national sample is +/-3.0 percentage points at the 95% level of confidence. The margins of error for subgroups are larger than the margin of error for the national sample. For the Catholic subsample (n=210), the margin of error is +/-­7.0 points at the 95% level of confidence.
Ouch. The Pew Report margin of error is much better -- 1.5% for Catholics, for instance -- but it only looks at the distribution of Catholics, not their opinions.

Okay, I see where you're coming from -- and we were using different meanings for 'proxy', which is where I was supposing you meant equality.

Dr. S.,

Since I'm taking my figures from the Public Religion Research report, I'm using their terminology. You may want to dig in a bit deeper to figure out exactly how they derived their numbers.

I can't quite tell from the report. But since they talk about a "Catholic subsample" I suspect they included Catholics in "general."

Alas, that usually would be a lie. The margin of error in these polls is much too large for decimal values to be meaninful.

I don't think so. The mean of the sample, which is an estimate of the population mean, is a definite number. There is no margin of error associated with it. If you poll 1000 voters, and 546 are for Claghorn and 454 are for Foghorn then the poll results are 54.6 to 45.4. There is no uncertainty about that. And to illustrate my peeve, notice that a swing of two voters, or maybe one, from Claghorn to Foghorn changes the reported results from 55-45 to 54-46. Two is not 1% of 1000.

What is uncertain is how close that is likely to be to the population mean, and the margin of error measures that.

Very interesting and compelling points. Has anyone done any of those sophisticated statistical tests using regression or whatnot to determine the residual effect on LGBT-friendliness of an individual being catholic after normalizing or considering location. I would also like to see it broken down by ethnicity, national origin, age, sex, church attendance. Maybe the salient fact about catholics isn't that they belong to the catholic church, but that they don't belong to fundamentalist evangelical churches. Would be nice to have a longitudinal study that addressed that.

The mean of the sample, which is an estimate of the population mean, is a definite number. There is no margin of error associated with it.

This sounds a little pedantic, Bernard; hopefully I'm not misunderstanding you.

No statistical parameter computed from sample data is uncertain; it is what it is. What's uncertain is how well that parameter represents the larger population from which the sample was taken. I believe the "margin of error" Doctor Science was referring to were the 95% error bounds on the estimate of the mean from the sample in question.

If you deny that there's any margin of error in the estimate of the mean, then we're about to have an argument over principles of sample statistics that were explored nearly a century ago. Hopefully this is not the point you were attempting to make, though; this is more a statement of how I took your comment, along with an invitation for you to revise & extend if needed.

Slarti,

I agree with everything you said. Perhaps I was unclear.

You write,

No statistical parameter computed from sample data is uncertain; it is what it is.

Yes, though I'd call it a "sample statistic" rather than a "statistical parameter." So I disagree that it would be inaccurate, or unreasonably precise, to report that statistic exactly. It is what it is.

I believe the "margin of error" Doctor Science was referring to were the 95% error bounds on the estimate of the mean from the sample in question.

Yes.

If you deny that there's any margin of error in the estimate of the mean, then we're about to have an argument over principles of sample statistics that were explored nearly a century ago.

I would not dream of denying such a thing (or of provoking an argument with a martial arts expert even if I did have such a dream). What I am saying (or trying to say) is exactly that the margin of error is associated with the quality of the estimate. But it is not associated with the estimate itself.

If I estimate that someone weighs 186.8 lbs, that's my estimate. You can report my estimate as 187 lbs if you like, but it is more, not less, accurate to report it as 186.8. The fact that there is measurement error, so my estimate is likely to be off, is true, but irrelevant to the accuracy of your report. It is relevant to the accuracy of my estimate, but that's a different thing. That my scale is imperfect affects the latter, not the former.

So it is here. The hypothetical poll I described had a result of 54.6 to 45.4. Again, you can round it in reporting, but that loses, rather than gains, accuracy. The problem I have with rounding poll results is that it can occasionally be misleading, especially misrepresenting shifts when the numbers from the next poll are reported.

As to being pedantic, well, sometimes I am, and this might be such an occasion. I don't think so, but that's the nature of a pedant after all. So maybe we have no disagreement about that either.

Thanks for elaborating, Bernard. I did not mean "pedantic" as a euphemism for "annoying", even if it appeared that way. Probably a better word would have sufficed; the sense I wanted to convey was that you were making an observation that was true, but was more nit-picky than useful.

It turned out that you weren't making that point at all, though, so it's just as well that I didn't express myself more exactly.

Anyway, thanks for clarifying.

Yes, though I'd call it a "sample statistic" rather than a "statistical parameter."

I was using "parameter" in a loose sense; we could be talking about any statistical moment, here.

This horse is dead now, agreed?

James:
Has anyone done any of those sophisticated statistical tests using regression or whatnot to determine the residual effect on LGBT-friendliness of an individual being catholic after normalizing or considering location.

To a first approximation, the effect is: none. Perhaps Bernard could do something more detailed, but the basic point of my post is that the data don't look like they'll supoort *any* effect of Catholicism. I don't think we have access to the kind of raw data needed to look at income, etc., though I agree that would be interesting.

Maybe the salient fact about catholics isn't that they belong to the catholic church, but that they don't belong to fundamentalist evangelical churches.

That's possible, but once again, you are dodging the null hypothesis, that religion is *not* the driving force. You're looking for a way for religion to be crucial, when the data do not point in that direction.

Maybe, as with this Presbyterian minister, the factor of overwhelming importance is whether you talk to people you know are gay.

I can't stress enough how rough polling is. I've done a lot of it, and people fudge like mad. They respond casually, they'd change their minds and give a different answer an hour later, they don't think things through, sometimes they're deliberately bullsh*ting for any number of reasons, the wording of the question matters, their opinion will change the next day with new info, let alone next week, and so on.

Polling, no matter how carefully done, is only the roughest of snapshots.

This isn't to invalidate any large points made, because all of life is somewhat ad hoc, but I'd stress that trying to get down to fine numbers indeed doesn't work, whatever the numbers themselves say, because the numbers reflect rather changeable opinions and very rough answers.

"The term, then, is obviously a relative one; my pedantry is your scholarship, his reasonable accuracy, her irreducible minimum of education, & someone else's ignorance." -- H. W. Fowler

Slarti,

The horse is well and truly dead.

I live in Boston, perhaps the most Catholic city in the nation. The Catholics I know- actual practicing ones who go to mass more than twice a year and send their kids to CCD or Catholic School- they just do not believe in certain big chunks of official doctrine. And they do not for the most part have a big problem with gays, or any more of such than their generation and daily contacts would predict. The survey corresponds exactly to my experience. Catholics I know do not stand out from the background population in re gay issues, and that is because Catholics I know do not believe in what the Church teaches.

Dr Science, thanks very much for this. I'm not certain that you've proved the null hypothesis is true (the underlying data is a bit shaky), but I can't find anything to argue against the simpler point: the null hypothesis has not been disproved.

The null hypothesis also makes a ton of sense -- much more than the claim that there is a significant correlation between being Catholic and being pro-gay rights. "Less anti-gay than you probably expected" I'll believe; "significantly more pro-gay than non-Catholics" is a bridge too far.

kent:

Well, the thing about a null hypothesis is that you *never* prove it's true. Its function is to just sit there and be disproven -- or not.

mwing:

I admit, my gut reaction is to agree with you, and point out that since Humanae Vitae in particular Catholics have gotten *really, really good* at ignoring Church doctrine about sexuality.

But the data (what there is of it) suggests that Catholics may not be distinctive this way, but that Americans of *all* denominations are similar in our ability to ignore whatever religious doctrine doesn't mesh with our surrounding, local culture.

Dr. Science: Touche.... The positive argument that I was trying (and failing) to suggest needs more data is basically the one that you mention, um, needs more data, in your later comment: "Americans of *all* denominations are similar in our ability to ignore whatever religious doctrine doesn't mesh with our surrounding, local culture."

I'd love to get better data to back you up on this point ... although I'm pretty sure it's true even without any data. In fact, I'm pretty sure it can be expanded well beyond "Americans" and may be a characteristic of "humans." :)

The Catholic Church that exists today is not the same as the Catholic Church I once knew and loved. It was gutted, modernized and politicized by vatican II. My only "Agenda" is the saving of souls. I think the Church should not be an extension of any political party and I certainly don't like being asked to sign political petitions in the Church vestibule following Holy Mass. It is disgusting.

http://thelimingview.blogspot.com/

The comments to this entry are closed.

Blog powered by Typepad