Whatever your views, there’s always a temptation to ruffle a few feathers among your peers. Among left-wingers in the UK, David Goodhart did just that in 2004: he argued that two cherished left-wing ideals were in conflict. In simple terms, ‘the Goodhardt hypothesis’ is that diversity undercuts support for the welfare state – partly because we become less willing to support taxes that go to people who are ‘not like us’, and partly because of the competition for scarce resources and opportunities among the most disadvantaged (social housing, welfare payments, low-skill jobs).
At heart this is about the relationship between diversity and trust. The idea that ‘multiculturalism’ undermines this trust between citizens has been a common theme in recent years – David Cameron was talking about the failure of multiculturalism earlier in 2011. (US readers may be more familiar with Bob Putnam’s ‘hunkering down’ version of the same argument). But two recent papers are in surprising agreement that the problem is not what we think, at least in the UK.
The two studies are Twigg et al 2010 and Sturgis et al 2010, both of which find a small negative effect of diversity on trust, which pales in comparison to the much larger effect of deprivation – which offers partial support to the Goodhart/Putnam thesis, but also room for critics to counter-attack. What caught my eye, though, were the ways that diversity, poverty and trust interacted – a pattern that can be seen in two charts.
Firstly, we have the figure from Twigg et al below. I’m not going to go into Theil scores etc; the necessary information is quite simple, although you have to remember which way round everything points. The horizontal axis is ‘deprivation score’ – more deprived areas are on the right of the figure. The vertical axis is the probability of perceiving low levels of social support. And as the lines go from top to bottom, they get more diverse.
What this shows is that diversity only harms trust in non-deprived areas. In deprived areas, there is no effect of diversity.
Interestingly, Sturgis et al find exactly the same thing – but their intriguing figure is one that takes this a stage further. The figure below looks at trust by deprivation and diversity, but also comparing people that know lots of people in their neighbourhood vs. people that know no people.
This shows that in deprived areas, diversity has no effect on trust among people that know lots of people in their neighbourhood. The largest effects are in non-deprived areas, for people that know no-one in their neighbourhood.
What’s interesting about all of this – other than the wonderful complex messiness of the real world – is that it contradicts the assumption among many British policymakers. Many people in the Labour Party have been arguing that they lost the last election because they were out of touch with the anti-immigrant feeling among working-class voters; these were people who were competing for housing, competing for jobs, and who didn’t reap the rewards of Britain’s multicultural society.
This may well be true – the New East End‘s partially flawed but compelling account of East London shows this competition in unflinching form. But the studies here show that it simply isn’t true to assume that trust is sytematically lower in poor, diverse areas than poor, non-diverse ones. The effects of ‘multiculturalism’ that are revealed by research are more complex than the one-sided portrayals that often abound in public debate.