ماموریت های طرفدار اجتماعی و انگیزش کارگر: مطالعه تجربی

Do employees work harder if their job has the right mission? In a laboratory labor market experiment, we test whether subjects provide higher effort if they can choose the mission of their job. We observe that subjects do not provide higher effort than in a control treatment. Surprised by this finding, we run a second experiment in which subjects can choose whether they want to work on a job with their preferred mission or not. A subgroup of agents (roughly one third) is willing to do so even if this option is more costly than choosing the alternative job. Moreover, we find that these subjects provide substantially higher effort. These results suggest that some workers can be motivated by missions and that selection into mission-oriented organizations is an important factor to explain empirical findings of lower wages and high motivation in these organizations.

Plenty of evidence suggests that there are workers who care about the mission of their job. Recent survey studies, for example, show that workers in the public sector care about the usefulness of their job for society Frank and Lewis (2004), and that altruistic motivation is an important motive for volunteering (see e.g., Burns et al. (2006) or Carpenter and Knowles Myers (2010)).1 Analyzing British Household Panel data, Gregg et al. (2011) find that workers in the non-profit sector are more likely to do unpaid overtime work than workers in the for-profit sector, and Fehrler (2010) shows that teachers in Swiss Waldorf schools, private schools with a special pedagogic profile, strongly identify themselves with their schools’ missions and accept to work for far lower wages than public school teachers. Evidence from a non-OECD country is presented by Serra et al. (2011) who find that pro-social motivations predict the choice of Ethiopian health professionals to work in the non-profit sector and that workers in this sector earn less than their colleagues in the for-profit sector. Nyborg and Zhang (2013) present evidence which suggests that some workers in the private sector care about the mission of their employer, too. Analyzing data on Norwegian firms, they find that the firm's reputation for social responsibility is associated with lower wages, controlling for many other factors.2 How mission induced motivation affects principal agent problems has been studied in a number of theoretical papers.3 In several of these studies, public sector workers are modeled as agents who care about the mission of their jobs (e.g., Francois (2000), Dixit (2002), and Prendergast (2007)). Dixit (2002) concludes that agencies could save on monetary incentives to get the same level of effort as private sector firms. Extending this idea to other sectors, Besley and Ghatak (2005) develop a model in which workers provide more effort if they are matched with an employer with their preferred mission, which in turn makes it optimal for the employers to lower monetary performance incentives and offer different contracts. In their model, employer missions and performance pay are perfect substitutes. They also discuss policy implications and conjecture that the decentralization of a school market, for example, might lead to a substantial gain in efficiency through better matches of teachers and school profiles. In this study, we first ask the question whether any workforce could potentially be motivated to provide more effort with the right mission. If so, employers might be able to save on monetary incentives. We test this hypothesis in a laboratory labor market experiment, in which the subjects can choose their job mission. We test whether they provide higher effort and whether this has an effect on contract choice. Our implementation of job missions follows Besley and Ghatak (2005) model, in which some workers care about the output of their job. In our mission choice treatment, agents generate a donation to an NGO of their choice. In our control treatment, subjects generate an extra pay-off to a randomly drawn student from the students register of the University of Zurich. In the first group we, thus, have a simple matching mechanism of missions and motivations, while the total output generated under the same contract and effort choices is the same in both groups.4 In each treatment, half of the subjects play the role of the employer and the other half the role of the worker. Having also the role of the employer played by subjects allows us to test the prediction of different contract choices, in addition to the prediction of different effort provision under the same contract, across treatments. Employers offer contracts consisting of a fix wage and a piece rate.5 Then, workers choose their effort level which determines pay-offs and donations. Our main results are the following. Workers do not provide higher effort in the mission choice than in the control treatment. In neither treatment effort provision is higher than the optimal effort provision of a purely self-interested worker. Consequently, employers cannot save on monetary incentives in the mission choice treatment and the contracts they offer are not different from those in the control treatment. A related study, focusing on the motivations behind pro-social effort provision, finds similar evidence. Tonin and Vlassopoulos (2010) measure effort provision in a real effort (data entry) task in a field experiment, in which subjects generate a donation for an NGO of their choice in addition to their own pay-off. They find a slightly higher effort provision than in a control treatment. However, the effect is very small and only significant for female participants. Moreover, they find that this effect is motivated by warm glow, i.e., utility from the act of giving itself, rather than by pure altruism. These experimental findings and the empirical evidence of motivated workers in mission-oriented organizations, discussed above, seem contradictory. However, it might be the case that selection of mission motivated workers into corresponding jobs explains the observations. If there is only a subgroup of workers who can be motivated, employers with strong missions have an incentive to screen workers. Paying lower wages than the market wage is a possible screening mechanism for worker motivation (Delfgaauw and Dur, 2007 and Brekke and Nyborg, 2008).6 To study selection, we run a second experiment in which all subjects are workers. They are offered two contracts each period by the experimenter. The contracts in the first periods all pay the same piece rate but differ in their fix wage. Under one contract in each period the workers can generate a donation to an NGO of their choice in addition to their own income. Under the second contract they generate a donation to a randomly chosen student. By varying the difference in the fix wages between the two contracts over 20 periods we can measure how much a subject is willing to pay to work for her preferred NGO. In some periods we also vary the piece rate to see whether potential differences also occur at different piece rate levels. Moreover, we can compare effort choices of the subjects that choose the NGO contract with effort provision of the subjects who do not. Doing so we account for the fact that workers self-select into different sectors and jobs and that this process might lead to workforces that differ in their responsiveness to pro-social missions. We thus address a potentially important aspect that has not been addressed in previous experimental studies.7 In addition to the between-subject comparison of subjects choosing or not choosing the NGO contract we can compare effort choices within subjects for all subjects that choose either contract type at least once (that is 68 out of 70 subjects). This within-subject comparison, in which we control for selection, can be compared to the between-subject differences from the first experiment, in which selection was ruled out by design. We observe that about one third of the subjects choose the NGO contract even when it is costly to do so (some of them, however, only if the cost is very small). This group provides higher effort than the other subjects and their effort provision is higher than the optimal effort level of a purely self-interested worker.8 Our within-subject comparison shows an insignificantly higher effort provision under the NGO contract than under the alternative contract. This finding is consistent with the zero effect finding of our first experiment. Overall, these results suggest that the scope for an increase in effort of a given workforce through the provision of missions is small. However, a motivated subgroup of workers exists and self-selects into the mission-oriented jobs. Offering low salaries appears to work as a screening device for motivated workers. The paper proceeds with the theoretical background in Section 2, followed by the experimental design in Section 3, the presentation of the results in Section 4, and the conclusion in Section 5.

Empirical evidence suggests that a part of the workforce is motivated by the missions of their jobs. This idea is also supported by theoretical arguments. Besley and Ghatak (2005) conclude that substantial efficiency gains might be feasible through improved matching of workers and employers with the right missions. In this paper, we first ask the question whether a given group of workers can be motivated to provide higher effort by letting them choose the mission of their job. We test this in a laboratory experiment in which subjects can generate a donation to an NGO of their choice in addition to their monetary pay-off. In this setting, which should guarantee good matches of workers and missions, we do not observe increased effort provision. Tonin and Vlassopoulos (2010) find similar evidence. This suggests that improved matching of an existing workforce with their preferred missions, most likely, does not lead to a substantial increase in effort provision. However, these results are at odds with empirical findings of high worker motivation in various organizations with strong missions (e.g., Carpenter and Knowles Myers, 2010, Fehrler, 2010, Gregg et al., 2011 and Serra et al., 2011). Sorting might be the reason for these seemingly contradictory findings. Our second experiment shows that a subgroup of subjects (about a third) is willing to give up monetary rewards for the option to generate a donation to their preferred NGO. Moreover, we find that these subjects also provide substantially higher effort. This suggests that some individuals can indeed be motivated by the mission of their employer. Therefore, sorting of motivated workers into mission-oriented jobs is important, and offering a lower wage appears to be an effective screening device for employers with strong missions. The results from our second experiment are consistent with the empirical findings of both high motivation and low wages in mission-oriented organizations and add a potentially important element to the empirical literature on social incentives. However, it should be kept in mind that we use a student sample and study sorting in a laboratory labor experiment. It is likely that other incentives influence sorting in labor markets outside the laboratory, in addition to the incentives studied here. It might also be the case that the individuals who decide whether to work for a NGO respond differently to the studied incentives than the students in the laboratory. The student subjects obviously face a decision that is much less consequential for their future lives than the choice of an employer outside the laboratory. They also self-select into participation in a laboratory experiment for different reasons than the reasons that make workers enter the labor market. This might lead to a sample of individuals in the laboratory that is different to the workforce outside the laboratory with respect to their mission motivation. Future studies – especially field studies with different target populations – will therefore be useful to investigate the external validity of our results.