In my last two articles (here and here) I’ve touched on the topic of spatial segregation in cities. Many commentators point with alarm to the increasingly sharp divide between the inner city and the suburbs of Australian cities in terms of income, education and values.
They’re right to be concerned about rising inequality and how it manifests spatially. But we also need to be careful about assuming all geographical clustering or clumping of like-minded people is necessarily a bad thing and should therefore be actively discouraged by policy.
The fact is people like to associate with others who they have a lot in common with; usually people who’re like them. As I’ve discussed before, big cities tend to foster networks that are less diverse, not more (Are bigger cities less diverse?).
That might seem counter-intuitive because cities are rightly renowned as crucibles of heterogeneity. Part of the “diversity dividend” of big cities is there’s greater exposure to new and different ideas with scope for synergistic connections.
But another part is cities make it much easier for people to find and associate with other people who’re like them, no matter how “specialised” their interests. They don’t need to mix with people who’re an imperfect match like they’d have to if they lived in a small town.
Even with modern transport and communications technologies, people in cities still find value in living close to other like-minded people. These geographical clusters are rarely comprised entirely of one group, but there’s usually a significant over-representation that gives an area a distinctive character.
The classic example is migrants, who throughout history have tended to live in distinct ethnic and cultural pockets.
There are considerable advantages for migrants in settling among people of their own ethnic background. It gives access to rich information networks about jobs, potential marriage partners, access to housing finance and the myriad other essentials for adapting to life in a new country.
The grown-up children of migrants tend to be more geographically mobile than their parents, but for the first wave of settlers the benefit of having family and a community close by who speak the same language, practise the same religion and live by familiar cultural norms must be immense. Areas of ethnic concentration are often rich in social capital and positive externalities.
In fact lots of groups benefit from spatial proximity. Single people like living near other single people. That means they’re more likely to meet others with similar interests, but it also attracts entrepreneurs who provide the lifestyle services and facilities singles want.
Families also benefit from living near other families. Concentration supports a greater variety, depth and specialisation of infrastructure like child care, schools, clubs and sporting facilities. It facilitates shared child-minding and play-dates. Parents of young children know how much easier it is to entertain other parents with same-age children.
Sydney’s Inner West, Brisbane’s Inner South and Melbourne’s Inner North are all areas that support a particular set of values, as exemplified by the high vote they gave The Greens at Saturday’s election.
If you live in these areas you’re more likely to meet people whose values are similar. That might matter in how the local school is run, what sort of shops are in the high St, and whether council is active in supporting cycling, public art and child care.
In short, there are big benefits in living close to people who’re like you. As I noted here, researchers at Wellesley College and the University of Kansas call it the “similarity-attraction effect”:
it influences everything from whom we date and hire to where we choose to live. The bigger the pond, the more likely we are—consciously or not—to swim around until we find a group of like and like-minded people.
There can also be advantages in keeping out of the way of people who are very different from you. If there’s enough of them they can frustrate your values on public goods policy e.g. schools. In some cases they might be downright hostile or offensive to members of your group.
In fact there can be disadvantages in “too much” diversity. In his book on social capital, Disconnected (which I’ve discussed before), Andrew Leigh points out that there is a negative correlation between trust and ethnic diversity:
Residents of multi-racial neighbourhoods are more likely to agree that ‘you can’t be too careful in dealing with most Australians’. In particular, neighbourhoods where many languages are spoken tend to have lower levels of trust…
Not everyone necessarily benefits from spatial clustering, though. The poor might get no advantage from congregating or it might even make them worse off. Nor do they always choose to cluster – the better-off might move out of certain suburbs leaving those with fewer resources behind.
I don’t know if disadvantaged citizens are better off living in mixed areas or not; there seems to be mixed evidence on this point. I suspect though that living in intense concentrations of disadvantage is worse.
The optimal level of clustering at the local level is very likely a complex matter. It probably depends on factors like the group who’s clustering, the variable or dimension along which it occurs, the degree of clustering, the spatial area, and the extent to which other groups are crowded out.
Last time I noted that segregation in US metros, as measured by the dissimilarity index, had fallen from 76 in 1970 to 56 in 2010. A score of 0 means the mix of blacks and whites is the same in every Census tract. Forced segregation is damaging to individuals and society, but I suspect few blacks or whites would choose a dissimilarity score close to zero.
Diversity and specialisation are the key advantage of cities. But it isn’t necessary or desirable that cities are uniformly diverse at all geographical scales from the street to the metropolitan area. The way cities have evolved suggests the advantages of diversity can be realised at the regional or metropolitan scale and the advantages of propinquity at the local scale.