Robert Putnam, and the complexities of diversity

I’ve been enjoying reactions to my piece in the Boston Globe this week, some through bloglinks and comments, others through email. While these ideas aren’t especially radical to my Afrophile blogger friends, they’ve offered some of the best reactions to the idea. Hash of White African identifies incremental infrastructure as a “radical idea” for change in Africa, suggesting a set of other radical ideas:

* Mobile phones as platform for business by Tradenet.biz
* South African news is being transformed by blogging (Amatomu) and new ways of reporting (The Sunday Times)
* News is starting to be reported on video using no computers, only mobile phones

Ory wonders whether I’m building a bridge too far in trying to extrapolate from mobiles to really, really big infrastructure:

I’m especially thinking about big ticket items like roads and power generation/transmission that don’t have the same option of being subsidized by the consumer. Also the incremental and even moreso the “pico” approach does not address needs on a macro-level that most countries in Africa need to address, having power in the village is great but if factories / business can’t operate at an optimal level because of power shortages or costs there’s still a big conundrum that needs be address (and unfortunately the role of government can’t be wished away).

Too true, and questions I need to work through as I develop any of these ideas further.

In a funny way, one of the best things about publishing a piece in the Globe is that it guaranteed that I would encounter Michael Jonas’s article, “The Downside of Diversity”, published in the same section. Jonas takes a close look at some very provocative research published by Harvard political scientist Robert Putnam. Putnam is well known for his concern for the importance of civic institutions in the US – not just political parties, but bowling leagues – and the fact that these civic institutions are important not only for the health of our democracy but for individual health. In a new set of research analyzing the results of a massive study, the Social Capital Community Benchmark Survey, Putnam turns up an intriguing and disturbing result: Americans living in diverse communities tend to “hunker down” and participate less in civic life than Americans living in more homogenous communities.

It’s worth looking closely at precisely what Putnam found before getting into the arguments about whether his research crosses lines from academe into advocacy. In a lecture and paper titled “E Pluribus Unum: Diversity and Community in the Twenty-first Century”, Putnam outlines the two popular models for understanding what happens in diverse communities. The “contact model”, based largely on studies of the integration of the US Army, suggests that ethnic biases can be countered by contact. Soldiers who served in non-integrated units were far more hostile to the idea of integrated units than soldiers serving in those units – encountering people from another ethnic background and working together went a long way to easing ethic fears.

Most research on diverse communities doesn’t find this bridging result, however. Putnam outlines the “conflict” hypothesis and cites cases around the world where ethnic heterogeneity is associated with lower trust and cooperation between groups. Many of the conflict hypothesis papers have an implicit assumption – that mistrust of the “outgroup” is related to increased trust of the “in-group”. In other words, if I’m part of a small group of white Americans living in Accra, we’re likely to mistrust the larger “outgroup” and have increased solidarity between our ethnically homogenous ingroup.

What Putnam and colleages found in their study is even more disconcerting – people living in highly diverse communities seem to distrust both their ingroup and outgroup more than people living in low diversity communities. Ask people in rural North Dakota (where “celebrating ‘diversity’ means inviting a few Norwegians to the annual Swedish picnic”) whether they trust their neighbors “a lot”, and more than 80% of people say they do. The same question in North Minneapolis, and the answer drops to 20%. There’s a similar effect when asking people about whether they trust people of other races “a lot” – 55% of rural North Dakotans do, while less than 25% of North Minneapoleans do.

This lack of trust correlates to less civic and political involvement – less confidence in local government, less confidence in a personal ability to effect the political process, less registering to vote, less charity and volunteering.

There’s lots of possible objections to these findings. Perhaps ethnically diverse areas are more crime-ridden than homogenous ones and this leads to suspicion and mistrust? Maybe it’s the effect of economic diversity, not ethnic diversity? Maybe there are fewer resources designed to bring a community together in these areas? Putnam disposes of nearly all these concerns using multivariate analysis, controlling for economic diversity, crime rates, education, citizenship, residential mobility and other possible factors – the pattern of a relationship between diversity and decreased participation persists.

So far, Putnam’s paper looks like a great win for the anti-immigration activists. Now Bill O’Reilly can argue that not only will immigrants give us all leprosy, but that we’ll all be so hunkered down that we’ll stop voting. It’s pretty clear from the structure of the paper that Putnam was deeply worried about this outcome. Before presenting the findings, he makes an argument that, in the medium to long term, diversity through immigration helps lead to creativity, more rapid economic growth and has enormous benefits to developing nations, perhaps obviating the need for traditional development aid.

This framing is what Jonas’s piece focuses on. He points out that Putnam had basic results from his surveys as early as 2000 and took six years to search for possible explanations for the phenomenon. Finding that the results stand up to very thorough statistical tests, Putnam published a paper that argues that diversity seems to make us uneasy, but is inevitable and good for us in the long run. Needless to say, this hasn’t made everyone happy… and it’s made some people a little too happy for Putnam’s taste, including noted bigot David Duke. Conservative commentators have been questioning the way that Putnam has “spun” the research, suggesting that the conclusions were simply uncomfortable for Putnam and that he would have preferred not to release them.

In truth, there’s a really interesting intellectual paradox identified here. If Putnam’s findings are true, it’s very hard to explain why diverse, multiethnic cities are such fertile ground for creativity, arts, and innovation. Is there a subset of people who thrive on contact rather than conflict? Are communities challenged and then strengthened by ethnic diversity?

I’m looking forward to reading Putnam’s findings in more depth, looking closely at the survey results for different cities. I want to offer one possible factor that I suspect Putnam may not have looked at, which is the possible role of “virtual social capital”. Putnam is studying engagement with the local, physical world. I wonder what the correlation between ethnic diversity and non-local engagement is. As Lant Pritchett points out in “Let Their People Come”, technology has made it far more possible for migrants to be involved with their home communities through low-cost telephony, videconferencing, financial remittance, etc. Is it possible that we’re seeing a slowing in assimilation because it’s possible for Cambodians, Somalis and Mexicans to remain engaged with their home communities longer? Might it also be possible that we see some people who are “hunkered down” in ethnically diverse areas becoming more engaged online? It would be interesting to try to control these results for digital divide issues and for online usage as well as for the factors Putnam has already wrung out of the list of possible explanations.