#RSAcommunity - Zachary Neal and the problem of diverse communities

Liveblogged notes of Zachary Neal‘s talk on community integration and cohesion at the RSA.

In this talk he’s going to focus on micro networks. Are diverse communities possible? Tha answer’s grim: no. But there is a bright side…

He’s been thinking about community policy in the US; it’s fragmented and piecemeal. It’s more clearly articulated in the UK. In 2001 the Home Office came out with a report on community cohesion, which lead to the Commission on Integration & Cohesion. In 2010, the Cabinet Office made it clear it was important as part of the Big Society rubric.

This is the right direction – but there’s a hidden problem, a policy paradox. It’s not clear how integration and cohesion interlock. Are more integrated communities more cohesive? Or are more integrated communities less cohesive?

Neighbourhoods

In segregated communities, similar people live near one another.  In integrated communities, different sorts of people are more evenly mixed through the neighbourhood.

Social networks

In fragmented communities, people have disconnected social networks. In cohesive communities, people have dense special networks.

Making Friends

How do people develop social networks; how do they come together?

Homophily – birds of a feather flock together. This is a nearly universal characteristic – it applies to animals, cities and protein interactions. It can be stinger or weaker. But it’s not about aversion. It’s more about opportunities to meet.

Proximity – near things are more related than far things. Works for all sorts of things, but people especially.

They create hypothetical communities, and think about what the social networks might look like, assuming moderate homophile and proximity. Moderately segregated communities are moderately cohesive. Highly segregated communities are more cohesive. They see this time after time. And on the other end of the spectrum, highly integrated communities are much less cohesive.

Conclusion: homophily and proximity means that making communities more integrated makes them less cohesive.

The Policy Problem

Are we stuck with this? Or can we shift to a world of integrated, cohesive communities? At any strength, homophily and proximity push against this. So, can we get rid of homophily? Can you imagine a world where you only became friends with people unlike you? Unlikely.

The other possibility is getting rid of proximity – making people more likely to become friends with people a long way away. Again, seems unlikely.

To create a integrated, cohesive world people need to avoid their neighbours, or avoid “birds of a feather”. But is that a world we want to live in? It seems to him that it’s not a world he wants to live in, or is it clear it’s even possible.

Is our policy initiative aiming for an unobtainable goal? Should we be striving for a balance instead? Could some communities benefit from more integration, some from more cohesion?

Q&A

Is there a Goldilocks point where you have sufficient cohesion, without becoming a monoculture?

It’s hard to identify that. Maximising cohesion is not necessarily our goal. Cohesivie communities tend to be very stagnant. Ideas stay within them, they don’t innovate. More fragmented networks mean you receive lots of different information, opening the way to innovation.

How possible is it to change the tradeoff through skilled network interventions?

The easiest work – under the name the contact hypothesis – worked poorly. The way way to break down boundaries is through friends of friends, not forcing unlike people to live next to each other. It’s difficult to create an intervention to create this friends of friends, though. We understand what’s need, but not how to do it.

Is a better understanding of social networks relevant to policy?

For centuries governments have been collecting census data and using it to set policy. The problem is that census data treats each individual separately – we need to look at how people relate to one another. That move sue beyond the simplistic individual analysis. Social networks are providing us with those tools. Pretty much everything we do is driven by the people we know.

In the states, we see naturally occurring retirement communities. They’re not moving, just finding each other and supporting each other.

The internet and faster transport are eroding the proximity effect. Now it’s possible to carry on long-distance friendships without meeting, or to form retirement community snot based on spacial proximity.

We’re seeing two types of relationships emerge online. There are those relationships that become offline relationships, and then we’re seeing the low level “Facebook” relationship, formed with just a click. Use of the internet to form real world relationships is one way of reversing these trends.

Who funds you?

This is unfunded work.

*Is computer analysis of networks is incredibly naive – possibly even wrong? *

This is an early version of a much larger model that will include many other characteristics. This models will never capture what’s going on in people’s heads. It’s a purely structural models – that gives us some idea of the boundaries within which policy can be set. There’s nothing random in networks – just things that are hard to predict and things are very hard to predict.

Is the term “proximity” a problem? Facilities can bring people together, but not at the same scale you’re talking about. Is the very idea of neighbourhood a problem in this?

In this model proximity just means the things immediately around your house. Your point is that proximity can mean proximity to facilities. Public schools can allow parents to form relationship and networks around that school. Charter schools create more fragmented networks. The way we design these public facities can effect the social networks in the area.

What about Universities? Or social media?

Universities are one of those nuclei that networks form around. But there’s still an element of homophily, around university education, around subject matter. Online social networks don’t seem to be translating into offline relationships. They could be used to reduce the effect of proximity, though, through maintaining relationships established face to face over greater distance.

My thoughts

*I really want to read his book. The model he’s presenting sounds like a good, but simplistic start on understanding the variables underlying community – that can’t quite stand up the claims being made, because there are more factors in play that the model accounts for. His approach to the effects of online networking on relationships seemed simplistic and on the borderline of wrong – but it feels like he’s doing good work challenging some of the assumptions around community policy. *