Later On

A blog written for those whose interests more or less match mine.

Interesting article on cash incentives

with 5 comments

My conservative readers will find this article right down their alley: people willing to make big changes for cash incentives, just as a conservative would predict. And liberals will endorse the idea for the social benefits that accrue. At last: something on which we can all agree. Still, as the article notes, some people object strenuously. I think that this in an indication that sometimes even a mildly complex idea idea—in fact, “complex” is an overstatement—is too much for some people to grasp.

“Anger over NHS plan to give addicts iPods,” ran the headline. The UK’s National Health Service is notoriously hard up, so news that government advisers were suggesting doctors offer drug addicts prizes as an incentive to stay clean was certain to raise some hackles. Why should “these people” with “self-inflicted” problems get priority, a patients’ advocate was quoted as asking in the article, published this July in The Sunday Times. “This country really is on its head,” concluded a reader in the newspaper’s online comments section.

Such reactions are typical when anyone raises the idea that addicts should be rewarded for changing their ways. Yet the fury over the proposal fails to account for one critical fact: incentive schemes work. And not just with drug abusers. Rewards have been used to help smokers quit, persuade parents to keep their children in school and boost uptake of healthcare. Far from being a waste of money, all the indications are that they can save money by reducing crime and other problems associated with poverty and drug use.

With over a decade’s worth of positive evidence, incentive schemes are finally getting the recognition they deserve.

This summer, New York became the first city in a developed nation to try to alleviate poverty by offering incentives to improve people’s engagement in areas such as education, health and employment. Other experiments are assessing whether rewards could help overweight people slim down and persuade the chronically ill to stick to medical regimes. Advocates say that incentive schemes are ready to change many areas of social policy – but only if the deep-seated objections of headline writers and others can be overcome.

“In the United States, you’re supposed to buckle down and work hard,” says Warren Bickel, a psychiatrist at the University of Arkansas for Medical Sciences in Little Rock, who has studied incentive schemes. “Paying people to be good? It goes against common-sense views about how society should work.”

One of the first assaults on these common-sense views began a decade ago in Mexico. Like many developing nations, the country had tried to tackle poverty using top-down projects, such as subsidies for staple foods and healthcare. Most of them were unsuccessful, in part because they failed to reach those most in need of help. During the 1990s, the Mexican government decided to experiment with a radical alternative: low-income families would receive cash payments, to be spent on whatever they liked, if they met certain goals. By enrolling a child in primary school and ensuring they attended regularly, for example, a family could earn up to 165 pesos a month (just over $20). Since it is harder to keep older kids in school, the reward rose by up to 140 pesos for high-school students. Another 125 pesos were on offer for families if their children had regular health check-ups. In all, a family could supplement its income by a maximum of 750 pesos, an amount chosen to equal 20 per cent of the average monthly expenditure of a typical family enrolling on the scheme.

The success of government initiatives like this one can be tough to evaluate, since politicians are rarely frank about failed policies. Progresa – the name is an abbreviation of the Spanish for education, health and nutrition programme – was an exception. Right from the start, the Mexican government was committed to having independent academics involved in designing and rating the scheme. Its findings were remarkable: within a few years the number of girls enrolled in high school jumped around 14 per cent to as much as 76 per cent; clinics that monitored babies’ development received up to 60 per cent more visits; and illness rates among children dropped 12 per cent.

In the following years, other Latin American nations launched similar schemes, dubbed conditional cash-transfer programmes, and recorded similar successes. The World Bank has become an enthusiastic supporter of these schemes, and is trialling the cash-transfer idea in several African countries including Ethiopia, Swaziland and Mozambique. “We’re seeing dramatic improvements from these schemes,” says Ruth Levine of the Center for Global Development, a think-tank based in Washington DC.

Now the cash-transfer concept is coming to New York City. With around 40 per cent of families on an annual income of less than $20,000, the city needs the project, says James Riccio, one of the researchers involved. More than 5100 families are being recruited by Riccio and colleagues at MDRC, a New York-based organisation dedicated to evaluating public policies. They have drawn up a long list of incentives, including $200 for each family member who attends an annual health check-up, $50 a month for each high-school child who attends regularly, an extra $150 monthly for adults in full-time work, and up to $3000 if they undertake training while in employment.

The $50-million project is an ambitious and risky undertaking, especially for mayor Michael Bloomberg, who has backed it politically and with his own money. What works in Latin America may not transfer to the Bronx, where life is tough but poverty is not life-threatening in the same way as it is in the slums of Mexico City. Yet Bloomberg can draw confidence from incentive experiments that have targeted very different groups to those in Mexico, and achieved similarly impressive successes.

It was the crack-cocaine epidemic of the late 1980s that prompted drugs researchers to experiment with incentives, says Stephen Higgins, an expert in addictive behaviour at the University of Vermont in Burlington. Two decades later, there is still no guaranteed way to get users off crack, but cash rewards do help. One typical experiment found that two-thirds of users quit during a 16-week programme during which they could earn up to $1300 by regularly providing drug-free urine samples. Only 40 per cent of those having cognitive therapy and 27 per cent on methadone managed to do the same (Archives Of General Psychiatry, vol 59, p 817). “The empirical evidence is now overwhelming,” says Higgins.

As incentive schemes built up a reputation, experts in other kinds of addictive behaviour became interested. Smoking was next. Tobacco consumption has dropped in recent decades in many developed nations, but is still a huge public health problem. More than 400,000 Americans die of tobacco-related illnesses every year, yet only 3 per cent quit in any one year. Studies by Kevin Volpp from the University of Pennsylvania in Philadelphia indicate that incentives could improve things.

In 2003, Volpp and colleagues recruited 170 smokers, each lighting up at least 10 cigarettes a day, and offered them a programme of five classes to help them give up. Half were also given $20 after each class and $100 if they then quit smoking. Seventy-five days after the course finished, 16 per cent of people in the incentives group were not smoking. Among those that didn’t get the money, the figure was just 5 per cent (Cancer Epidemiology Biomarkers & Prevention, vol 15, p 12).

So why do such rewards work? On one level, the answer is straightforward. People adjust their behaviour to achieve pleasurable goals, such as food or sex. Pursuing hard drugs is an extreme goal, but the underlying theory applies just the same. Each hit of heroin brings intense pleasure, which encourages users to change their behaviour so that they can get more and more of it, until eventually they become conditioned to seek the drug – what psychologists call operant conditioning. In this context, a financial reward is simply an alternative hit, a pleasurable pay-off that competes with the heroin for attention.

It surely can’t be that simple, though. In almost all incentive schemes, the money on offer is so minimal that it is hard to see why addicts care. In the cocaine experiment, users initially received just $2.50 for staying clean. Similarly, smokers know that the money they will save by quitting is far greater than what’s on offer in an incentives programme. More telling still is new research by Higgins showing that incentives helped persuade 40 per cent of pregnant smokers to forego cigarettes until they had given birth, compared with 9 per cent of controls. “That’s crazy,” he says. “Why would a pregnant woman smoke, knowing it will affect the foetus, but stop because I offer her a couple of bucks?”

The key to understanding this disconnect is in the way we process rewards. If offered $100 now, or $105 in two months’ time, most people opt for the immediate payment. This might not make sense in purely financial terms, but it is human nature to place a lower value on money that we will receive in the future – economists call it “discounting the future”. The lower the value someone places on future income, the greater the discount rate they are said to be using.

One thing that unites the groups targeted by incentives is that they all seem to have higher than normal discount rates. In the case of drug users, one study found that heroin addicts preferred a hypothetical cash reward of $400 now, above one of $1000 delivered one year hence. For non-addicts, it took a delay of five years to make the choice worthwhile (Experimental and Clinical Psychopharmacology, vol 5, p 256). The instant high that drugs offer appears to lock users into a world where the immediate is paramount and the future darkens rapidly. Because addicts discount future benefits so heavily, it is much easier to change their behaviour with offers of immediate financial reward than with promises of a healthier life a year from now.

The evidence for high discount rates among low-income groups is more patchy. Nevertheless, it is enough to convince many economists. One study, for example, found that families in the bottom fifth of earners used a discount rate 50 per cent higher than those in the top 5 per cent (The Journal of Political Economy, vol 99, p 54). That makes sense, says Higgins, as poor families are often dominated by immediate needs, such as where the next meal is coming from.

All this suggests that the New York scheme has a good chance of making a difference to the behaviour and lives of at least some of the city’s poorest citizens. Better still, the realisation that incentives tap into our natural tendency to give a lower value to future rewards relative to immediate ones, indicates that they could have wide applications. Volpp is exploring some of these. He describes incentives as a “very promising” way to help obese people lose weight – though he is reluctant to discuss details while the research is under way. He also suspects incentives could help people stick to a regime of prescription drugs, and is doing a study with stroke patients on blood-thinning medication, a notoriously problematic group.

This is all well and good, but one question looms large over all these schemes: what happens when the rewards stop? For the social programmes, the long-term goal is not just to get poor children into school, but to ensure that future generations are less poor. This aim could be scuppered by many factors. If the labour market in developing nations cannot provide jobs for the newly educated children of low-income families, for example, little may change. If circumstances in poor areas do not improve, a new round of incentives may be required to get the next generation of children into school. Only decades of monitoring will show whether cash-transfer programmes actually provide long-term benefits.

Studies of drug users and smokers can be evaluated much more quickly. Here, the results do raise questions about permanent change, since some addicts from both groups relapse when the rewards are removed. Worse still, when researchers at the University of Oxford reviewed 15 smoking studies in 2005, they concluded that none showed statistically significant evidence of improved quitting rates beyond the initial six months (The Cochrane Database of Systematic Reviews, DOI: 10.1002/14651858.CD004307).

If addicts need permanent cash incentives to stay off drugs, are the programmes worth starting in the first place? It probably depends on how you perceive addiction. Many experts view drug abuse as a chronic illness requiring long-term treatment. They point out that society does not baulk at providing a life-long supply of drugs to patients with other serious conditions. Even some established treatments for addictions, such as group therapy for alcoholics, are potentially indefinite – so why not monetary rewards?

Once monetary rewards have been used to wean a user off drugs, the same techniques should simply be adjusted and used to prevent them relapsing, argues Kenneth Silverman, a psychiatrist at Johns Hopkins University in Baltimore, Maryland. This is exactly the approach he uses at the Therapeutic Workplace, a centre where Baltimore addicts can receive employment training, work opportunities and cash in return for staying clean. And his experience suggests that it does seem to work. In one case, he was able to help keep six out of 20 young mothers completely free of heroin and cocaine for three years (Experimental and Clinical Psychopharmacology, vol 10, p 228). To put this into perspective, only one woman in the control group stayed clean that long.

Investment in long-term incentive programmes may also make good economic sense, given the costs of treating the health problems that addictions cause, plus the money ploughed into tackling crime associated with illegal drug use. Little work has been done on this issue, but what is clear is that incentives schemes cost around the same as other therapies, such as methadone treatment, that are considered good investments because they reduce crime.

Yet none of that will matter if researchers cannot convince the public and politicians that incentive schemes are worth trying. In the US, where almost all the research on incentives and addictions has been done, the technique is used in only a few isolated drug-treatment programmes. That is partly because the idea is still seen as novel, says Higgins. The New York City experiment could therefore be something of a breakthrough. So too could a decision by the NHS to introduce UK-wide financial incentives to help recovering drug addicts. The latter is still up for discussion, but the need for a decision has been brought home to the British public following revelations last month that many drug-treatment clinics are so desperate to improve clean-up rates that they are offering extra doses of methadone and antidepressants as incentives – a policy that has been roundly criticised in the media.

If either the US or UK schemes prove successful and cost-effective, financial incentives could take off. “It behoves us to consider these approaches,” says Volpp, although even he accepts that incentives are likely to remain controversial no matter how successful they prove to be. “There is a lot of resistance to paying people to do something they should be doing anyway.”

Written by Leisureguy

24 November 2007 at 2:50 pm

5 Responses

Subscribe to comments with RSS.

  1. How would a liberal necessarily see the accrual of social benefits (this is the USA we’re talking about, not Mexico)? In a country where across the board, people are developing a deeper and deeper sense of entitlement, this is a terrible idea. Simply passing out incentives will not guarantee the intended social benefit — what happens if families seek out loopholes, start abusing crystal meth to lose weight, and force their already uninterested and uninspired kids to school every day just to pad their handout check? “Welfare babies,” anyone?

    I’m sorry, but life is tough. This is akin to paying your child a allowance for just keeping his room clean. That’s expected behavior. It almost makes more sense to revoke entitlements if someone neglects the most basic of social responsibilities. Society doesn’t need to be in the business of providing “instant gratification replacement therapy” for people who’d be better served if they’re left to take their own initiative.



    24 November 2007 at 4:38 pm

  2. Well, I was struck by the fact that it works. Expectations that it will not work surely run aground on evidence that it does. And some people have not developed sufficient social responsibility and can’t see the (long-term) benefit of cleaning up their act. What this approach offers is a way to mitigate significantly a problem society faces, and a way to help people learn social responsibility and reap some long-term benefits. In particular, by ensuring that the children are schooled, fed, and get medical care, the new generation may avoid the problems their parents had.

    Evidence is always useful. Where is the evidence that people are developing a deeper and deeper sense of entitlement? How do you measure such a thing, anyway? The bad things you suggest might happen might also not happen. The program in New York will be studied closely and then we will have facts rather than fears to use in evaluating the program.

    I’m sorry, but life is based on facts, not indignation. There is an underclass in the US, and this is an approach to dealing with that problem. I think it deserves a chance, based on the evidence to date.



    24 November 2007 at 4:49 pm

  3. Mike…all I’m saying is that this hasn’t actually been applied to the US yet. It works in Latin American countries where the circumstances of poverty are much different, and families respond in different ways depending on the availability (or lack thereof) of public and private programs, handouts, etc. The underclass in the US can already subsist on these things to a greater extent than the underclass in Mexico. I might be horribly confused as to what things are actually like, but in my head I see a poor Mexican family that doesn’t have to send their child to work if they can replace the income in some way. So that kid gets to go to school and it all works out. In the US… I’m just not so sure the circumstances are like that.

    And I’m not trying to be indignant… I just don’t see the evidence that this works in the USA. I hope I didn’t piss you off in this or an earlier reply.



    24 November 2007 at 4:59 pm

  4. The last time I checked, the UK wasn’t a third world country.

    In fact, at two quid per dollar (a 33% increase in value over the last four years) they are doing quite well economically.



    24 November 2007 at 5:21 pm

  5. Sorry—my tone was grumpy, I see in re-reading my post. Here’s a relevant passage from the article:

    The $50-million project is an ambitious and risky undertaking, especially for mayor Michael Bloomberg, who has backed it politically and with his own money. What works in Latin America may not transfer to the Bronx, where life is tough but poverty is not life-threatening in the same way as it is in the slums of Mexico City. Yet Bloomberg can draw confidence from incentive experiments that have targeted very different groups to those in Mexico, and achieved similarly impressive successes.

    In any event, within a year or two, evidence should be available about what in the program works and what does not. That will be very interesting indeed.

    Bloomberg has had some excellent successes. The “311” idea was a stroke of genius and it’s working like a charm. So maybe he’ll be right on this one, too. We can hope.



    24 November 2007 at 6:40 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: