VoxEU Column Development

No margin, no mission? Motivating agents in the social sector

What motivates agents in the social sector? Recent theoretical advances indicate that financial incentives might crowd out intrinsic motivation and reduce performance. This column describes a field experiment where hairdressers in Zambia were motivated in different ways to sell condoms. It finds that social recognition can be a more powerful performance motivator than financial gains.

Many organisations hire agents to perform pro-social tasks, namely tasks that entail benefits for others in society. This is mostly true for non-profit and other mission-driven organisations, but is becoming increasingly relevant for for-profit organisations that often engage in pro-social activities alongside their main line of business.

In addition, many mission-driven organisations in the developing world implement social programmes by engaging community members to perform pro-social tasks in addition to their main income-generating activities.

Despite its practical relevance, there is limited evidence on determinants of performance on pro-social tasks and in the social sector in general, and, in particular, little is known on how to best motivate agents in this sector.

Why might pro-social tasks be different?

Theoretical literature suggests that evidence from the private sector (Lazear 2000, Bandiera et al 2005, 2007) may not suffice when designing effective incentive mechanisms for pro-social tasks. First, mission-driven organisations, unlike private-sector businesses, likely engage workers whose interests are aligned with the organisation’s mission, and so these workers may be inherently less responsive to financial incentives (Besley and Ghatak 2005). Further, if workers are intrinsically motivated towards pro-social tasks, mission-driven organisations, unlike private-sector companies, have to worry that financial incentives can crowd out motivation, and thus reduce overall performance (Bénabou and Tirole 2003, 2006), especially when financial incentives are low-powered (Gneezy and Rustichini 2000). These considerations raise the need for an alternative reward scheme in the work of mission-driven organisations, which can give workers a non-monetary stake in success and leverage, not crowd out, their intrinsic motivation for the pro-social task.

Study setting: Selling female condoms in Lusaka, Zambia

Our field experiment in Zambia (Ashraf et al 2012) tests the effect of both financial and non-financial rewards on workers’ promotion and sale of female condoms, which we class as a pro-social task because correct and consistent condom use is a major driver of HIV prevention in sub-Saharan Africa.1 We collaborated with a public health organisation in Lusaka that hires and trains hairdressers and barbers to sell condoms in their shops. Hairstylists were chosen for the task because they have a personal relationship with clients that allows for successful targeting of condoms to ‘at risk’ customers. Prior to the study, 42% of all hairstylists in Lusaka reported discussing HIV/AIDS with clients, and 45% reported discussing romantic relationships with clients. Hair salons are also numerous and present in every neighbourhood in Lusaka.

The experiment

The field experiment randomly assigned 1,222 agents located in 200 distinct neighbourhoods to one of four groups. Agents in the control group received a standard volunteer contract offered by NGOs, and agents in the three treatment groups received small financial rewards, large financial rewards, and non-financial rewards, respectively, in exchange for selling condoms to clients. High and low financial rewards were awarded as a margin on each condom sold, and in the non-financial scheme, stylists received social recognition for each set of condoms sold, in the form of stars posted on a thermometer displaying condom sales (‘star’ treatment, Figure 1). We tracked condom sales for one year, at monthly frequency.

Figure 1. Poster on which ‘stars’ treatment is displayed


Stylists rewarded with stars sold twice as many condoms as those in the control group. That they did so every month of the year rules out that differences are driven by a novelty effect. Further analysis shows that the stars treatment leveraged intrinsic motivation for the cause. Indeed, it was more than twice as effective for stylists who showed high motivation for the HIV/AIDS cause prior to the programme. Also, the stars treatment appeared to facilitate social comparison among hairstylists located near one another, as the treatment was more effective for stylists surrounded by other stylists in the same treatment. Lastly, in contrast to existing laboratory evidence, we found no evidence that financial incentives crowded out stylists’ intrinsic motivation (Figure 2). On the contrary, high financial rewards were more effective for stylists who showed high motivation for the HIV/AIDS cause prior to the programme.

Figure 2. Female condom sales by incentive treatment


Our findings open new avenues for mission-driven organisations worldwide involved in the distribution of health and other ‘social goods’ to design innovative, cheaper, and more effective methods to motivate their workers. Further investigation can offer insights into the features of non-financial rewards like that stars treatment that contribute most to their success.

While the results described here build on a particular programme and a specific non-financial motivation scheme, the general design principles are easily replicable and adaptable to other settings. Rewards varied linearly with condom sales, which minimised the discouragement that could have come with a non-linear rewards scheme. Moreover, rewards were made clearly visible to third parties thus allowing social comparisons between different stylists engaged in the same task, which proved effective at eliciting greater sales efforts. Finally, stylists were awarded by a reputable and well-known organisation, which might have contributed to their value.

An obvious limitation to using non-financial rewards is that they cannot replace money as the main medium of compensation, and are thus of limited use in situations where performance pay is a large share of total pay. However, our findings suggest that a non-financial incentive can be a cost-effective means to motivate workers in settings where the fraction of variable pay over total pay is small. Ultimately, to assess whether non-financial rewards can be effective in other settings, future research must provide further evidence on how the nature of a reward – financial or non-financial – interacts with the nature of the task – pro-social or not – to attract, motivate, and retain high performing workers in mission-driven organisations.


Ashraf, N, O Bandiera and K Jack (2012), "No Margin, no Mission? A Field Experiment on Incentives for Pro-Social Tasks", CEPR Discussion Paper 8834

Bandiera, O, I Rasul, and I Barankay (2005). “Social Preferences and the Response to Incentives: Evidence from Personnel Data”, Quarterly Journal of Economics 120(3): 917–62.

Bandiera, O, I Rasul,and I Barankay (2007), “Incentives for Managers and Inequality Among Workers: Evidence from a Firm Level Experiment”, Quarterly Journal of Economics 122: 729–75.

Bénabou, R and J Tirole (2003), "Intrinsic and Extrinsic Motivation", Review of Economic Studies 70(3): 489–520.

Bénabou, R and J Tirole (2006), "Incentives and Prosocial Behavior", American Eco- nomic Review 96(5): 1652–78.

Besley, T and M Ghatak (2005), "Competition and Incentives with Motivated Agents", American Economic Review 95(3): 616–36.

Gneezy, U and A Rustichini (2000), “Pay Enough or Don’t Pay At All”, Quarterly Journal of Economics 115(3): 791–810.

Government Republic of Zambia (2010), Zambia Country Report: Monitoring the Declaration of Commitment on HIV and AIDS and the Universal Access Biennial Report.

Lazear, E (2000), “Performance Pay and Productivity”, American Economic Review 90(5): 1346–61.

UNAIDS (2010), “Report on the Global AIDS Epidemic.” Geneva: Joint United Nations Programme on HIV/AIDS.


1 Zambia has one of the world’s highest adult HIV prevalence rates at 14.3% (Ministry of Health, Zambia 2010). It is estimated that in 2009, one million Zambians were living with HIV and 45,000 died of HIV-related causes (UNAIDS 2010).

2,106 Reads