As submitted to the Daily Maverick
As T.O. Molefe pointed out in a column last week, we can legitimately question whether a campaign like Get Tested will have any long-term effect on the willingness of South Africans to get tested for HIV. Launched by Premier Helen Zille to coincide with 16 Days of Activism for No Violence Against Women and Children, Get Tested offers a financial incentive for knowing your HIV status. Anyone who volunteers for HIV testing between November 28 and December 9 stands a chance of winning a first prize of R50 000 or one of five R10 000 runner-up prizes.
You might recall that Zille attracted a significant amount of criticism for her stated intentions of criminalising HIV transmission. As was the case then, an apparent problem with Get Tested is that according to Molefe, it tries “to achieve individual thought and responsibility by taking away thought and usurping responsibility”. Furthermore, Molefe draws on the work of Daniel Pink in arguing that the effects of the incentive will only be short-term, falling away once the prizes are no longer available.
Both of these criticisms are to my mind at best not compelling, and at worst unfounded. While it’s true that behavioural economics is a contested science, the contestation more typically relates to whether findings in experimental conditions cross over into informing policy in the real world, and to whether the rise of “choice architects” – in other words, those who design the “nudges” – present an illiberal incursion into our freedom to choose.
Pink’s claim that carrots and sticks are less effective, and potentially harmful, in addressing complex 21st century problems raises questions relating to the design of these nudges. These questions don’t however rule out the potential usefulness of nudges. If the desired behavioural change depends on tweaking complex motivational forces, we might struggle to achieve it – but there seems little difference in kind from more simple behaviours. The question is one of degree only.
Of course, there might be problems too subtle to address via these sorts of nudges, but given that we’re talking about a science that arguably only came into existence in 1979, with Kahneman and Tversky’s “Prospect theory: An Analysis of Decision Under Risk”, it seems rather hasty to claim that the experimental work conducted by George Ainslie, Gary Becker and a host of others has been premised on problems already superseded by newer and more complex ones.
Despite this question, it remains true that our models of intertemporal choice behaviour and hyperbolic discounting (in this context, the choice habits that could result in people being tested now for the chance of winning money, instead of for their own long-term benefit) are imperfect. They could perhaps even be critically flawed, and as I mentioned earlier, applicable more to experimental conditions than to real-world policy.
But the successes of these sorts of interventions – at least on pragmatic criteria – seem to be far outweighing their failures. Long-term trials such as the 3-year J-PAL immunisation intervention in rural India have been tremendously successful. Offering parents 1kg of lentils and a set of metal meal plates upon completion of a course of immunisation for their child resulted in a more-than-doubled immunisation rate at half the cost, thanks to economies of scale resulting from increased uptake.
There are similar tales of success with HIV testing in South Africa. The 2008 intervention by the Men By the Side of the Road charity, which offered unemployed men R75 to test for HIV had a 100% uptake in the group it targeted, along with a waiting list in the thousands. The Discovery Health/Sunday Times Right to Know campaign (July 2008 to June 2009) attracted 55 000 volunteers for testing, thanks to the promise of one person per month becoming R100 000 richer.
But the question is of course sustainability, and whether these incentives have a long-term impact on behaviour. And here is where I’d urge Zille’s critics to a little more temperance, in that if relatively low incentives of one R50 000 prize and five R10 000 prizes end up resulting in significantly increased rates of testing, the R100 000 total spend could end up being a very worthwhile (and sustainable) investment, given that the Province currently spends a reported R661 million on its HIV/AIDS programmes per year.
What of the criticism that choice-architecture of this sort is somehow grossly illiberal, “taking away thought and usurping responsibility”? First, it’s perhaps worth noting that the liberal goal of securing freedoms is surely maximised by both good health, and by decreasing expenditure on preventable diseases so that the funds can be used elsewhere. The question is how we secure those goods, and whether we impinge on liberties excessively in doing so (as we would do by criminalising HIV).
While it’s true, as Molefe says, that HIV testing should be its own incentive, I’d suggest that those of us who know this already get tested. Offering people the chance of a cash prize for HIV testing can only increase the likelihood that any first-time testers can be exposed to that message, resulting in a probable net gain in terms of awareness. It’s not the case that the Get Tested campaign is replacing ordinary awareness campaigns – it’s a supplement to them, and one that seems worth trialling given its low cost.
Furthermore, it’s a supplemental sort of intervention that is widely accepted, rarely attracting the sort of criticisms Zille is confronting here. What, for example, do critics of choice architecture think of Discovery Health’s Vitality programme, which by Molefe’s reasoning “sends the unintentional message that [good health], something which is an incentive in itself, is so abnormal and exceptional a behaviour that those who get tested are entitled to a reward”?
While nudges like the Get Tested campaign are sometimes accused of violating human autonomy, it’s difficult to be sympathetic to these charges. South Africans retain their choice to be tested or not – and if some who would otherwise not be tested now come forward, they do so because they autonomously desire money. Their exercising that autonomy is potentially to all of our benefits, given the level of public expenditure on AIDS.
In short, the harms of Get Tested are unclear while the possible benefits are not. Given that it will cost us little in terms of money – and nothing, as far as I can tell, in terms of liberty – we should give it a chance. If it ends up working, some of those entitlement horses might eventually end up being able to find their own way to the watering-hole.