Australian economist Andrew Leigh has entered into public discussion with Noel Pearson about Aboriginal inequality by proposing that randomised trials should be initiated for those educational innovations supposedly aimed at improving outcomes for disadvantaged groups. He takes his cue from Harvard economist Roland Fryer, who is well known for testing the effectiveness of cash rewards on academic achievement. Provocatively, Leigh has called proponents of the approach ‘the randomistas‘. (You can also hear about randomised education trials).
The rhetoric of the scientific method sounds very sensible. After all, if it works in medicine, why shouldn’t it work in education? Indeed we can go further: since we don’t accept medicine that hasn’t been tested in a randomised trial framework, why should we accept education without similar confirmation of its effectiveness? The point Fryer and now Leigh have been making is that much educational policy is ideologically driven, rather than evidence driven. We need proper evidence and for them, randomisation is the mark of proper evidence.
So is there any downside?
Well, it’s worth noting the reportage on the Harlem Children’s Zone, a set of charter schools that apparently produce ‘miraculous’ results.
Fryer and his colleague Will Dobbie have just finished a rigorous assessment of the charter schools operated by the Harlem Children’s Zone. They compared students in these schools to students in New York City as a whole and to comparable students who entered the lottery to get into the Harlem Children’s Zone schools, but weren’t selected.
This tells us two important things. First, there is a lottery to determine who gets in to one of these schools. Second, the children who aren’t lucky, don’t gain the benefit. They are stuck with the old system, presumably quite a lot less miraculous.
So the down side to the randomised trials argument is that to do the tests we need to influence children’s lives even more than we already do, by introducing a new element of contrived randomness into their educational careers. Children who benefit from the Harlem Children’s Zone aren’t merely lucky as in ‘lucky to be born in America’. They are lucky as in ‘the subjects of an overt social experiment based on an actual lottery, providing benefits to some by deliberately taking them away from others, whilst claiming that this dis-endowment was somehow random and therefore morally neutral’.
Despite the tone of scepticism I’ve introduced, this might not be a problem. It is clear that in many areas of life, people and institutions accept contrived randomness as the mundane background against which life is lived. Jury duty is a good example of contrived randomness which is fairly well accepted (I might try to wriggle out of it, but I’m not going to campaign for its abolition). This is indeed just the kind of situation that university research ethics committees might be expected to address.
On the other hand, it might be a very big problem. People who object might not be do so not because they hate the scientific method and want every child to fail. They could well be opposed because they see the deliberate division of children into lottery winners and lottery losers (when there was no such lottery previously) as an unacceptable form of Fatalism, or ‘chancism’ as public administration professor Christopher Hood puts it. He writes:
turning organizations into a gaming machine has the potential to foster fatalistic attitudes towards the underlying control system.(1998: 66)
Furthermore, disadvantaged groups already have a long history of contrived randomness as a tool of oppression. Aboriginal communities, for example have suffered terribly through the bureaucratic random reshuffling of families, otherwise known as the stolen generations. In this instance as in many others, the ‘random’ element was often a spuriously disinterested smokescreen for the pernicious intent of the policy.
This opens up the question of whether educational experiments can ever be disinterested. This might be the worthy aspiration of some, but it is emphatically not the aspiration of many, who already know, quite implacably, what the answers are. This indeed was the basis of the Northern Territory ‘Intervention‘ which sought to impose ‘ preconceived solutions on communities. We might now decry the Intervention, but it has given all interventions a bad name.
Before randomised trials in education are widely accepted, it will be necessary more broadly to legitimate the contrived randomisation of everyday life. As a result of past experiences this is likely to be an uphill struggle which risks further entrenching the vicious cycle of Fatalist policy.
Now read:
Four Cultures 