Henry Farrell sets up a new blog to publicize political science research.
Melinda Henneberger says that Democrats would win more women voters if they were more moderate on the abortion issue. I’m not sure if she’s right or not, but this article has been much discussed around here.
Robert Putnam is a big favorite of mine, though his latest study is very disturbing. His five-year study shows that immigration and ethnic diversity have a
devastating short- and medium-term influence on the social capital, fabric
of associations, trust, and neighborliness that create and sustain communities.
He fears that his work on the surprisingly negative effects of diversity
will become part of the immigration debate, even though he finds that in
the long run, people do forge new communities and new ties. (Thanks, Amy P.)
Apt. 11d 
I respect Putnam for coming forward with this data, even though it’s not the result that he wanted. One of the more interesting aspects of his findings is that this breakdown in community happens within groups, as well as between them. So in a diverse neighborhood, not only do you distrust and avoid people from outside your ethnic community, but you also begin to distrust and avoid members of your own community. The good news is that it’s not permanent, and that eventually a new community identity is formed. (However, there are many areas of the world where this final step has not occurred, and is not likely to happen any time soon!)
Needless to say, this study has implications for the current immigration debate. There’s been very lengthy and heated discussion at janegalt.net. Megan has said (while arguing against restrictionists) that we contemporary Americans underestimate how difficult and lengthy assimilation was for Italians, Irish, etc. I agree and suggest that this is actually an argument for caution with regard to new immigration, since the entry of each new immigrant marks the beginning of a decades-long process.
LikeLike
It’s a long time since I followed the debate, but didn’t Will Kymlicka and Keith Banting come to the conclusion that neighbourhood diversity didn’t have an impact on people’s willingness to support a redistributive welfare state? If they’re right, then combined with Putnam’s analysis (assuming he’s correct) something very complicated is happening in those neighbourhoods.
LikeLike
Sorry I forgot to add a link. It’s here.
LikeLike
I favor abortion on demand, but I’m sure that Henneberger is right. The Dems take their absolutist posture on this at huge cost to their ability to win elections.
Clinton, on this as so many other issues, threaded the needle on electability with ‘safe, legal, and rare’. He did damn-all on ‘rare’ while in office, which is why he is remembered so fondly by the abortion-rights crowd, but he did understand how many voters feel about it.
I am myself an abortion-on-demand absolutist, as I said – it seems to me that human personality, or ensoulment, or whatever you want to call the boundary between human life and not-human-life, happens after birth in the interplay of experience with the potentiality of the infant. But I think my position is a minority one, and not one which any party could profit from taking in an election.
Several women who have had abortions have talked about it with me, and so I’m going to generalize from them – they think about it, and what could have been, and there’s some sadness. It doesn’t FEEL like just a minor procedure, even if they were abortion-on-demand absolutists before.
I think Hillary Clinton has been doing the most, of the Dems, to make a formulation which can respect the sort of feeling which the women I have talked to have.
LikeLike
re: Putnam. I have a lot of questions about Putnam’s study after reading the brief, one-sided article in City Journal.
Did he control for socio-economics? I imagine trust is lower in all poor neighborhoods, regardless of diversity.
Did trust vary depending on the ethnicity of the groups in the neighborhood? Did Latinos generate more distrust than Indians? Did English speakers generate less distrust?
Did he control for location? I imagine that areas like New York City are so used to diversity that an influx of a new immigrant group wouldn’t affect them.
Do trust levels go down initially, but then go back after time? If so, who cares?
I think only fools would say that immigration is painless. Of course, it’s a rough road with competition for limited resources and with fear of change and difference. However, some things are more important than neighborliness. I think that Putnam’s study shows the shallowness of neighborliness.
LikeLike
Seriously?
Laura McKenna links to the Robert Putnam study thusly: Robert Putnam is a big favorite of mine, though his latest study is very disturbing. His five-year study shows that immigration and ethnic diversity have a devastating short- and medium-term influe…
LikeLike
Laura, you can find Putnam’s article here. (Hope that html works.)
LikeLike
I grew up in Pennsylvania, a rare state where you can be an electable Democrat and pro-life.
But former PA Governor Bob Casey wasn’t allowed to speak at the 1992 Democratic National Convention because he’s pro-life. (Contrast this with the Republicans, who allowed pro-aborts Guliani and Pataki to speak during prime-time in 2004.) Clinton sold out most lefty special interest groups (as well as his own wife) at some point during his presidency, but he never wavered on killing unborn children at any stage. The Left holds abortion-on-demand sacred, even when it costs them elections.
LikeLike
Mrs. Ewer, actually, several pro-life Democrats did speak during the convention. Casey wasn’t allowed to speak (well, the claim was because he didn’t support the ticket, but I’ll take the scenario that makes the Dems look the worst) because he was planning to talk about abortion in his speech.
Have any pro-choice Republicans been able to talk about abortion in their convention speeches? Did Giuliani talk about being pro-choice? Did Pataki talk about abortion? No.
Personally, I do believe abortion on demand is sacred. That is because I believe most women choose not to have unplanned pregnancies and abortions on a regular basis. I think that when they seek abortions, it’s because they feel it is the best choice for them given the circumstances.
I also believe in open heart surgery on demand, for the record.
LikeLike
“However, some things are more important than neighborliness. I think that Putnam’s study shows the shallowness of neighborliness.”
I will grant that some things are more important than “neighborliness,” Laura, but I think that only because the very phrase “neighborliness” captures only a very small portion of what the whole concern for “civil society” and “civic trust” is attempting to get at. I haven’t read the book by Kymlicka and Banting which Ciaran linked to above, but I have read articles which preceded the book’s publication, and my understanding of them is that all they could show was that multiculturalism did not significantly affect support for the universal (and therefore abstract) principles of welfare and redistribution. In other words, ethnic diversity and stress did not change existing levels of support for economic justice. Now, economic justice is nothing to sneeze at–indeed, if you are (as Kymlicka certainly is) a philosophical liberal, then all that really matters in the end is economic (and, presumply, political) support for individual liberty, since that is what enables people to choose or not to choose to be part of something. Hence the great importance of making certain that the welfare state can make peace with multiculturalism. But economic justice, expressed through welfare payments, is only one part of social justice. What about people’s willingness to make use of (and support through tax dollars) public spaces and parks and events where everyone can interact equally? What about the functionability of common institutions like public schools, which so often–as you know–depend upon “human capital”; is such equalizing capital just as forthcoming in communities stressed by diversity? And so on, and so forth.
I really ought to blog on this. Long comment made short: Putnam may well have failed to control for many things; I don’t know. But setting economic justice against mere “neighborliness” fails to capture the full set of worrisome social concerns which his examination of diversity is presumably truly after.
LikeLike
It looks like Robert Putnam has shown that ethnic diversity makes people less gullible.
LikeLike
Thanks for that link to the Putnam study, Wendy and Ciaran for the Will Kymlicka and Keith Banting book.
I’m going to have spend some time looking over that Putnam study, but you all should take a look at it. It’s a really fine example of how political science should work – it’s well written, interesting, with thought spend on policy implications.
Putnam’s study seems to be very different from Kymlicka’s and Banting’s work. I suppose you could still retain liberal values, while losing trust in your neighbors.
LikeLike
Although I hate to admit it, this study rings true to me.
When you attend meetings at church, for example, you can’t help but notice how people show up. It’s almost comical, watching the various different ethnic groups arrive at different times based on their idea of acceptable punctuality. (As an old German I’m always the first to arrive.) We all have to figure out for ourselves what it means to be punctual in our new group. And sometimes the resulting mild conflict drives people off, or causes them to ‘turtle’.
What this study perhaps does not reflect is that, in my experience, the great social fabric you find in, say, South Dakota, is not all its cracked up to be. There’s a reason young men from South Dakota end up in Chicago when they come out of the closet!
LikeLike
Dear Laura,
I agree heartily with RAF on the importance of community. In fact, I would suggest that community without economic equity would be far preferable to economic equity without community. Putnam mentions, by the way, that neighborhood diversity is accompanied by a rise in local activism, along with a decrease in trust and community involvement. Arguably, what is happening here is that residents are suddenly more wary of their neighbors, and more interested in elbowing their way forward to get their slice of the pie ahead of their neighbors. Meanwhile, in less diverse neighborhoods, residents are less assertive, trusting that they are going to be treated fairly. (Maybe that explains “What’s Wrong With Kansas?”)
While it is common in some circles to idolize community, I think that “community” and “conformity” are actually two sides of the same coin, and that community comes with a price. For there to be a friendly, neighborly, low-friction public life, individuality has to be suppressed. You can’t both enjoy untrammeled individualism (of whatever kind) and community, too. If you disagree with me and think your warm, supportive community is different, try painting your house orange with purple polka dots and putting a car up on blocks in your front yard, or enthusiastically laying out your original new argument against gay marriage to all your Bobo friends.
LikeLike
Having only skimmed the Putnam paper, I find myself with one gigantic question:
Has he controlled for urban vs. rural environments? Or, more specifically, for the overall size of the community? In a small town, you have a lot more trust of your neighbors, regardless of race, because there is by necessity a higher density of ties. In most urban and suburban areas, high population densities mean lower density of ties between individuals. I don’t know most of the people who live on the next block (even, for that matter, on my own block), so I’m not sure how much I would trust them, regardless of race.
However, in a small town, everyone knows everyone–one time, when I was a kid, a book arrived at the post office with its address label torn off. It was an anatomy textbook. The postman knew that I was “smart” (his son was one year behind me in school, though we were not friends) and interested in science, so he brought it by our house to see if we had ordered it. No, but thanks for asking.
Similarly, my parents are in the last stages of building their dream house in the same rural area where I grew up. The total cost of this house has probably crossed the $500,000 mark. Although they had a general contractor who did most of the work, they also dealt with a mason, a water/septic specialist, and a landscaper, among others (listening to them talk about some of these folks is like a page from A Year in Provence, which I’m pretty sure they haven’t read). NOT A SINGLE CONTRACTOR wanted to sign a contract. The entire house was built on handshake deals. That level of trust has little to do with ethnic homogeneity and everything to do with the size of a community.
The communities with the lowest levels of trust are not only highly ethnically diverse, but they are also urban communities with high populations. So, did I miss the data showing that ethnically homogeneous high population areas are more trusting than ethnically diverse high population areas (please do control for income) or that ethnically homogeneous high population areas are more trusting than ethnically diverse low population areas? If not, then I’m going to cry “hidden variable” and take the study’s conclusions with a large cube of salt.
LikeLike
Yeah, my quick skim came up with the same questions, Mrs. Coulter. A good part of the paper compares North Dakota with San Francisco, which raised the same red flags for me. I didn’t have the time to check out his figures to see if there was more info in there.
LikeLike
I’ve enjoyed this debate, but may I respectfully ask you to read my article, not just journalistic summaries, before reaching conclusions. For example, my analysis absolutely does control for city size, poverty, and dozens of other potential “causes”.
Bob Putnam
LikeLike
Professor Putnam,
Give that most of the mechanisms discussed in the study are perception-driven, are you considering sampling for subjective estimates of diversity and some of the other relevant independent variables, such as crime? Data on subjective estimates of diversity might also provide some way of cutting into the categorical boundary changes you hypothesize occur over time (although all of the findings here would be strengthened, as you suggest, by diachronic analysis).
LikeLike
Absolutely, I need time to read through the research properly.
I’m still mulling over whether or not your conclusions square with my anecdotal experiences. I was formally in a highly diverse neighborhood in Manhattan, where the Dominicans, Hipsters, Russians, German Jews were in very close contact. The community organized a Harvest festival, afternoon playdates, a parent listserv, Kareoke nights. And now I’m living in a small town where everyone is Irish/Italian Catholics (like myself), and I still don’t know the names of the weirdos who live across the street from us. Anecdotes, I know, so that’s why I’ll read your study tonight.
Do your findings affect your views of social capital? If social capital is so fragile that it can’t withstand the knocks of a new face a the door, then is it something that we should care about?
LikeLike
“If social capital is so fragile that it can’t withstand the knocks of a new face at the door, then is it something that we should care about?”
Again, Laura, I think you’re (intentionally?) minimizing the scope of the issue here. “Social capital” is not solely, nor even primarily, something reflected in and/or manifest by one’s friendliness with the weirdoes across the street, or one’s responsiveness to knocks from strangers at the door. It has to do with basic levels of mutual comprehension and trust. That’s not to say that your response to the conclusions of his study aren’t necessarily accurate, just that it seems to me that you’re presuming that since social capital must mean X, and you anecdotally can have good functioning (even egalitarian) communities without X, that must mean it isn’t necessary to worry about social capital. I think you’re skipping some steps.
LikeLike
Putnam’s latest work is reinforcing in my mind two points:
(i) the very serious limitations of survey research.
What’s with the faith we put in these “trust” questions, and self-reporting of engagement in various activities?
I don’t care how (apparently) consistent your findings are across various surveys, nor how nuanced your question wording is, nor how big the battery of clever questions is, nor how stunning the loadings are in your impeccable factor analysis. This just screams out for finding clever instruments (although statisticians are lurking in the background to warn us about too-easy acceptance of instrumental variable approaches).
There just has got to be a better way actually to measure trust and personal engagement than self-reporting and association membership rolls. Some of the experimental data Putnam reports in passing points to a better approach, I think, but of course that has the distinctive limits of artificially crafted experimental settings.
We should be looking for natural experiments to test models of trust, and we actually need far more precise micro-level models of what goes on in peoples’ heads, which leads me to my second gripe …
(ii) the profound limits of what I’ll call (with apologies to March and Olson) the ‘garbage can’ approach to social science research.
This modeling approach plagues both the ‘social capital’ and ‘political culture’ literatures.
You know what I mean.
A not-ridiculously unfair general characterization might go as follows:
“Hey, we’ve got this somewhat murky but pleasingly intuitive causal model in our heads about what’s going on in the very messy world of citizen beliefs and motivations. So, we’ll just “operationalize” our concepts, then plug some of these survey-derived variables into a logit or probit or heckman selection or whatever sexy new estimation strategy is in vogue, stick in a bunch of variables that seem intuitively relevant as “controls” and then run the damn thing through the software.”
Chris Achen has said smart things about how we ought to think about causality in constructing models, and what that would demand by way of empirical testing. What he says seems mostly to fall on deaf ears when it comes to studying social capital and political culture. Similarly so for Jim Johnson’s concerns about the conceptual problems lurking in political culture research – any survey researchers listening?
… But, having indelicately and imprudently uttered complaints (i) and (ii) …
I was recently messing around with my own ‘garbage can’ model of some of the trust measures in the World Values Survey. They have the usual “can’t be too careful”-type trust questions, but there are some interesting questions in there that aren’t directly about trust per se, yet ask whether the respondent would be concerned about new neighbors with X trait/belief/etc.
So, I ran a straightforward logit estimation on an embarrassingly crude model (likelihood of mentioning race as a concern about neighbors, regressed on town size, controlling for the usual SES suspects typically used in social capital and political behaviour research).
The data has all the problems with survey research I’ve complained about, and the model has all the annoying features of the ‘garbage can’ approach I’ve bemoaned.
All the same, what I find is that, when plausible control variables are held at their means, I get the following estimates of the likelihood of mentioning neighbor’s race as a concern, for a range of countries in the latest full wave of the survey (mostly these are North American and European countries):
Predicted Likelihood of Concern about Neighbour’s Race, by Town Size (other variables held at their means)
Town Size p(y) UB LB
under 2,000 0.29 0.28 0.31
2000 – 5000 0.26 0.25 0.27
5000-10000 0.23 0.23 0.24
10 – 20,000 0.21 0.20 0.21
20 – 50,000 0.18 0.18 0.19
50-100,000 0.16 0.16 0.17
100-500,000 0.14 0.14 0.15
over 500,000 0.13 0.12 0.13
Now to be sure, this model sucks (a sharp and soon-to-be-on-the-job-market sociologist from Toronto pointed out to me that, among this model’s many flaws, “race” doesn’t really mean the same thing, or have anything like the same salience, even across European respondents, let alone different towns and cities across Europe and the Americas). And the data have all sorts of problems.
However, lacking any humility or good sense (along with an utter lack of modeling acumen), I’ll conjecture that my off-the-cuff model’s suckitude is in the same ballpark as that plaguing some of the studies that Putnam cites approvingly.
Yet my results suggest that — self-reported levels of generalized or localized trust be damned — race is less likely to be on respondents’ cognitive radar as the size of their settlement increases.
But again, I don’t think this is the way to get at either the pscyhology of trust and its consequences for economic performance and democratic citizenship.
LikeLike
Russell – People are going to have different gut reactions when they first glance at this report. For some, they are going to make judgments about immigration. Others are going to minimize the importance of social capital, trust, neighborliness. It may be throwing the baby out with the bath water, but it’s a gut reaction that some people are going to get. They are going to look at these findings and say that neighborliness and trust are ephemeral things and perhaps we shouldn’t prize this value that only turns on when people are around like minded people. Yes, this is a gut reaction and I just wrote a pro-human capital post a couple of weeks ago, so don’t mind me.
I’m curious if some diverse areas have more social capital than others. Are there any examples of diversity w/some level of social capital? Are there any lessons that can be learned? For example, do schools make a difference.
LikeLike
Another way of thinking about this might be to say that we, as Americans, are surprisingly crappy at figuring out how to get along with new people. You’d think we’d have learned this by now. (Or perhaps I’m just too impatient?)
LikeLike
OK, so now that I’ve had the time to actually read the paper more thoroughly, I see that a serious attempt was made to deal with the questions I raised. (Note to self: don’t be such a hothead.)
A calmer question: this is not time series data, right? Because it strikes me that it would be really interesting to look at how shifts in diversity correlate with levels of trust. Another thing I was wondering about: diverse neighborhoods with lots of immigrants tend to be highly transient, which makes it difficult to build both strong and weak ties in the community. Is it possible to separate out that effect?
LikeLike
There are two transience variables: Census Tract Percent Living Same Town and Respondent’s decades in the community.
LikeLike
Jen: “What this study perhaps does not reflect is that, in my experience, the great social fabric you find in, say, South Dakota, is not all its cracked up to be. There’s a reason young men from South Dakota end up in Chicago when they come out of the closet!”
and
Laura: “However, some things are more important than neighborliness. I think that Putnam’s study shows the shallowness of neighborliness.”
and
Laura again: “If social capital is so fragile that it can’t withstand the knocks of a new face at the door, then is it something that we should care about?”
Depends on who “we” is. “Neighborliness” is not an absolute good – but then again, neither is diversity. People have different desires and priorities. I fail to see why a young gay guy’s being happier in Chicago is some argument against the particular social fabric of a South Dakota town – any more than anybody else’s preference for the high-trust, “neighborly” SD town is an indictment of Chicago. I prefer the crackling Babel of a great diverse city myself. Does this mean every friggin’ place in America therefore needs to be made over to meet my tastes? (And they are tastes, mind you – not morals.)
One more point. Amy P. states”…that community comes with a price. For there to be a friendly, neighborly, low-friction public life, individuality has to be suppressed.“. This is true, to varying degrees. But why can’t people order their priorities and make their own decisions about what they want to pay for what? Big country, after all. I really don’t have a problem with homogeneous South Dakota towns and great diverse metropoleis existing in the same space-time continuum. I will also point out that engineered, politically enforced diversity exacts its own price in freedom.
LikeLike