In a recent Health Care Cost Monitor post, Kristin Voigt and Harald Schmidt contend that offering employees who use wellness programs in the workplace reduced insurance premiums and deductibles threatens the fairness and affordability of health insurance for those most in need. Depending on how the program is structured, this is not necessarily the case, but Voigt and Schmidt provide compelling support for their assertion.

More fundamentally, however, we ought to ask ourselves whether financial incentives to improve health-related behaviors work. If they do not work, then they will clearly fail to reduce health care costs. The evidence is decidedly mixed.

There is some evidence that financial incentives can affect simple behavior changes, such as medical compliance. For example, in a systematic review of 11 randomized controlled trials that had used financial incentives, 10 demonstrated a positive effect. In these studies, the incentives ranged from $5 to $1,000 for interventions including adherence to a tuberculosis medical regimen, dental care for children, immunization, postpartum appointments among indigent adolescents, the completion of a treatment program for cocaine dependency, and antihypertensive treatment.

Other studies have found a positive effect from small financial incentives for mammography, required hepatitis B vaccines among drug users, and keeping clinic appointments for depression. But other studies have failed to find financial incentives effective for medical compliance.

For complex behavior change of the type often included in wellness programs, such as smoking cessation and weight loss, there is almost no good evidence for a sustained, positive effect. For example, in a systematic review of 17 studies on financial incentives for smoking cessation, none of the studies found significantly higher quit rates after six months among people who had financial rewards compared with those who did not.

However, a large trial published last year randomized 878 people to either a control group or an incentives group that received $100 for completing a smoking cessation program, a further $250 for being abstinent six months into the trial, and an additional $400 if they remained abstinent six months after that. At 18 months from the initiation of the trial, the quit rate in the incentive group was significantly higher than that for the controls – 9.4% versus 3.6%.

That trial is but a single study. A sustained effect has also rarely been observed for weight loss. For instance, a systematic review of nine randomized trials on the use of incentives to reduce obesity rates found that the larger incentives produced better results, but that the difference was not statistically significant. People offered financial incentives of less than 1.2% of their disposable incomes had a mean weight change of precisely zero after 12 months. People who received financial incentives of 1.2% or more of their disposable income had a mean weight loss of 2.4 pounds at 12 months, and only 1.5 pounds at 18 months.

The same researchers who conducted the smoking cessation trial also conducted a study of financial incentives for weight loss, but their long-term results were not overly auspicious. For instance, seven months after the initiation of the trial, the mean weight loss in the control group was 4.4 pounds, compared with between 6.2 and 9.2 pounds in the incentive groups.

Overall, therefore, financial incentives for specific areas of medical compliance appear to show some promise, but there is insufficient good evidence that they work in any sustained way to alter broader lifestyle behaviors. It may be that the literature to date uses incentives over too short a time frame to affect these complex behaviors. It is also possible that the size of the incentives offered in the studies was too small. Moreover, the cost, let alone the cost-effectiveness of the incentives, is rarely explicitly recorded.

But what is clear is that the lack of good evidence that financial incentives generate a meaningful positive effect on complex behavior suggests that we cannot conclude that they will save health care costs. Therefore, governments should be cautious in endorsing them.

Adam Oliver is RCUK Senior Academic Fellow in Health Economics and Policy at the London School of Economics and Political Science. Formerly, he was a Japanese Ministry of Educa­tion research scholar at Keio University in Tokyo and a Commonwealth Fund Harkness Fellow at Columbia University. He was also founding coeditor of the journal Health Economics, Policy, and Law. a.​j.​oliver@​lse.​ac.​uk; 44 (0)20 7955 6471.