Daily Maverick Politics

Do you know what’s good for you?

Originally published in Daily Maverick

The sorts of people who complain about a nanny state are often the same sorts of people who know what they want, and have at least a rough idea of how to get it. By contrast, being denied a choice is less notable if it occurs in a context in which you don’t make many choices in any case.

Put another way – politically liberal folk who complain about state intrusion on their choices can be accused of an undue focus on “middle class problems”. When you have choices, it’s annoying to have them restricted. Unfortunately, this can manifest in both positive and negative ways, because for every liberal who wants to minimise state intrusion on private choice, there’s a hippie who doesn’t think they should vaccinate their kids.

The overlap here is with regard to our belief that we are being best placed to make decisions for ourselves and our families, and also sometimes our conviction that our model is the appropriate one for states to adopt.

Because I know what’s best for myself (or so I claim), I should be allowed to do it. And, if there are others out there who don’t know what’s best for themselves, they will over time – even perhaps generations – discover what they want and how to get it. The state’s role is to not get in the way of that self-actualisation.

Some take these arguments further than others. Some libertarians might argue that even prescriptions for medication are an undue restriction on my free choices. If I have consulted Doctor Google, and take responsibility for my choices, why may I not purchase medication without paying a 3rd-party R350 for a permission slip to do so?

I’ll leave the libertarian arguments to Ivo Vegter. For my part, I’m happy to identify as a liberal, but even that more moderate position is becoming increasingly difficult to justify in light of its idealistic underpinnings. I can recall having these debates in tutorial rooms in the early 90’s, where we wondered whether John Stuart Mill’s harm principle could be justified with reference to typical humans, instead of the very atypical sort of human represented by Mill.

Today, behavioural economics motivates for a far more pessimistic attitude towards self-awareness and rational choice for even those middle classes – never mind those for whom simply having choices is a luxury.

For those of you who don’t know it, the harm principle is summarised in this passage from On Liberty:

the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or mental, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinion of others, to do so would be wise, or even right.

For Mill, this made sense because we know our own desires and needs better than anyone else does. If others – like the state – were to estimate what those needs might be, they would have to do so by considering the average person’s interests. And of course, none of us think of ourselves as average (even though, as a matter of logic, most of us would have to be). So, to cut a very long and very interesting story short, we should be left alone to make our own mistakes, except in cases where we might cause harm to others.

This is good and well if some of us occasionally smoke ourselves to death or have motorbike accidents without wearing helmets. If reasonable precautions against harms to others are taken while smoking, your only interest in someone else’s smoking is the potential increased costs of your own medical treatment. Similarly with the wearing of a helmet, in that your only interest should be whether accidents without helmet cost more – and how those costs are covered – than accidents with helmets.

But what if we aren’t rational choosers? Or rather – seeing as we already know that we’re not – what if our irrationality is so profound that we typically make sub-optimal choices, or at least make sub-optimal choices reliably enough that something could be done about it?

The reason that this isn’t a traitorous question for a liberal to ask is because when we think of our liberty, there is perhaps a danger of thinking about being impeded in the pursuit of a particular choice, rather than thinking about how we maximise our liberty on aggregate, throughout the course of our lives.

As I mentioned in a previous column on the Western Cape’s “Get Tested” lottery incentive for HIV/AIDS testing, we are all prone to hyperbolic discounting (in short, underestimating the value of later rewards in favour of sooner ones), and interventions which involve telling people – albeit subtly – what’s good for them can have very positive results, as for example in the J-PAL immunisation intervention in rural India.

Imagine that the liberty of your current self is impeded through some government agency making it difficult for you to do something. Or rather don’t imagine, but remember the last time you needed to get a medical prescription. One way of perceiving these events is as violations of your current liberty. Another interpretation is however also possible, in which your future self might be rather grateful that your choices were restricted, seeing as she now gets to enjoy the liberties made possible (in an extreme version of the thought experiment) through still being alive.

There’s no doubt in my mind that a fully competent person should be free to make self-harming choices. The question, though, is whether we are as competent as we think, for the reasons I’ve hinted at above (more fully explored in this book review by Cass Sunstein). Or more important, perhaps, the question of whether we are competent enough, regardless of how competent we think we are.

If we are not competent enough, the focus moves to what we should do about it. One option is to allow for social engineering through natural selection, whereby we make our mistakes and live with the consequences of those mistakes. But even though liberals and libertarians haven’t historically been too concerned with political correctness, embracing this view might be a challenge in that it’s likely to be the poor and the uneducated that suffer most, simply through not having the luxury of the choices many of us take for granted.

And if we don’t go that route, consistency problems soon arise, in that there’s a small step between nudges, or “choice architecture”, and banning certain choices entirely. The Conly book, reviewed in the link above, argues for a strictly utilitarian calculation of which choices should be permitted and which not, with a strong bias towards freedom.

The mechanics of and legislation underpinning those calculations is clearly a source for concern, in that we might justifiably be afraid of a state encroaching ever further on our freedom. At the same time, though, as Sunstein points out: “when people are imposing serious risks on themselves, it is not enough to celebrate freedom of choice and ignore the consequences.”

Daily Maverick Morality Politics

Premier Zille is nudging people to Get Tested

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.