Principal investigator: Bruno Benevit
Original title: Losing Prosociality in the Quest for Talent? Sorting, Selection, and Productivity in the Delivery of Public Services
Authors: Nava Ashraf, Oriana Bandiera, Edward Davenport and Scott S. Lee
Location of the Intervention: Zambia
Sample Size: Adult 1.585
Sector: Health Insurance
Primary Variable of Interest: Pro-social motivation
Type of Intervention: Career benefits
Methodology: OLS, DID
Summary
O trade-offs The relationship between the skills and commitment of health workers can be a crucial aspect in explaining the quality of health services, indicating a duality between extrinsic and intrinsic rewards. To analyze how these incentives interrelate, the authors of this article conducted an experiment with a national recruitment drive in Zambia to assess whether offering career benefits leads to the hiring of less pro-social professionals. The results demonstrate that this phenomenon was only observed among low-skilled workers, and that workers who opted for the career path are more talented and equally pro-social. Additionally, the treatment led to improvements in several patient health indicators and worker productivity in the treated districts.
- Policy Problem
The quality of public services is a critical aspect for economic development. This topic is highly relevant when related to the provision of health services in rural areas, generating debates about whether the selection of health agents involves a... trade-offs between qualifications and skills compared to pro-social inclination (Ashraf et al.According to the authors, while economic incentives (extrinsic rewards) are more attractive to highly qualified and skilled healthcare workers, greater local affinity and pro-social inclination (intrinsic rewards) motivate the choice of professionals who are more committed and engaged with the role they perform.
In Zambia, much of primary healthcare in remote rural areas is provided by informal workers linked to religious and charitable organizations. To professionalize healthcare delivery, the Zambian government created a new public health worker position called Community Health Assistant (CHA). The first recruitment and training for this new role took place in 2010. CHAs are expected to conduct home visits and referrals to health units. Therefore, this professional needs both the technical skills for triage, history taking, first aid, and referral, and the social skills for counseling, support, and guidance of patients.
In this sense, offering career benefits in the selection process for these agents may attract more qualified health workers, but with less pro-social engagement. Although both attributes positively affect service delivery, the marginal gains of these characteristics in the quality of service delivery and health indicators are uncertain. Furthermore, due to the shortage of medical professionals, the application of this measure may imply large disparities in the quality of service provided, highlighting the importance of evaluating this public policy.
- Implementation and Evaluation Context
The experiment was conducted concurrently with a national recruitment campaign for a new healthcare position in Zambia. To determine whether career benefits attract talent over pro-social motivation, the experiment used two types of posters: one for the treated group and another for the control group. Recruitment was carried out randomly across 48 districts stratified by province and district education level. The poster for the treated districts recruited agents for the new career worker position, while the poster for the control districts targeted those interested in the informal employment model with non-governmental organizations, which was then prevalent. The two types of advertisements differed only in the way they presented career opportunities, requiring the same prerequisites for both models.
During the recruitment process, the selection panel observed general skills, cognitive (IQ) and non-cognitive skills (ambition, persistence, and work ethic), and pro-social skills, determined by the perceived usefulness generated by helping others. Candidates' cognitive skills were measured based on their final 12th-grade exam scores and the number of subjects taken in biology and other natural sciences. Non-cognitive skills considered professional ambition, measured according to the candidates' self-declared career aspirations. To measure pro-sociality, the authors considered the candidate's self-reported willingness to remain in the community long-term (5 to 10 years after the interview), in addition to the result of the alignment of interests on the "Inclusion of Others in the Self (IOS)" scale (Aron). et al..
- Policy/Program Details
In the program's first year, 2010, the government initiated the recruitment and training process for each of the 167 communities across 48 districts. Selection panels were responsible for choosing two candidates to serve as community health workers at each community's health post. These panels consisted of five members: the district health officer, a representative from the health center associated with the health post, and three members of the local neighborhood health committee. In total, the recruitment process received 2.457 applications, of which 1.585 participants were considered for the analyses conducted in this study. Ultimately, 334 candidates were recruited as primary positions and 413 as reserve positions for 47 districts (one district was excluded from the sample for not submitting candidates for community health worker positions). The selected health professionals become contractually obligated to the positions for one year, and community health workers must wait two years before applying for higher positions.
- Method
In order to control for possible selection bias regarding skills between the treatment and control groups, surveys were conducted before and after the health workers' training to verify differences in the ambition and pro-social preferences of the recruits. The authors also conducted a test to verify the existence of a difference in the importance given to general and pro-social skills between the panels of the treatment and control groups during the selection process.
Subsequently, the article presented a series of analyses related to the productivity of health services provided using the ordinary least squares method. Through the ordinary least squares (OLS) method, the first set of analyses verified the impact of treatment (the effect of career opportunities for the Community Health Agent) on the number of visits in the first 18 months after treatment. Concurrently, the authors analyzed whether there was a difference in services provided after the end of the 1-year contract, observing the presence of treatment compensation effects that affected the quality of visits, community meetings, and emergency calls.
The second set of analyses examined whether the treatment affected the utilization of health unit facilities in the treated districts. To this end, the difference-in-differences (DID) method was used to verify the effect of the treatment on health units after the intervention, that is, the integration of the community health agents (CHAs) into these units.
Finally, the OLS method was adopted for the third set of analyses to verify the impact of the treatment in terms of health practices, disease incidence, and anthropometric indicators of malnutrition and disease for mothers and children in the communities.
- Main results
Analysis of recruits' preferences indicates that the groups differed before training, but there was convergence in agents' preferences after training, and only candidates with low skill levels from the treated districts showed significant and minor differences in pro-sociality compared to the control districts. Overall, this behavior indicates comparability between candidates from both groups, allowing us to separate the effect of selection from the effect of incentives associated with the job modality. Additionally, the results indicated that the selection panels for both recruitments valued the candidates' characteristics in the same way.
When analyzing the impact of career opportunities for Community Health Agents (CHAs) on the number of visits in the first 18 months, the authors identified a large effect caused by the treatment, such that 13.818 visits would have been carried out if the agents in the control group had made as many visits as the treated group. Regarding the effect on the quality of care after the initial 18 months, the results showed that CHAs dedicate the same amount of time to care, are equally likely to visit the target patients (mothers and children), and carry out more home visits and more follow-up visits. Furthermore, no compensatory effects were identified in the performance of secondary activities; on the contrary, positive effects of CHAs were observed in the number of community meetings and patients seen at the health post.
The results of the analyses on the effect of the treatment on the use of health unit facilities showed significant positive results in several aspects. Districts that had Community Health Agents (CHAs) observed significant increases of 30% in the number of births performed in health centers, 24% in the number of visits to children, 22% in the number of children weighed, and 20% in the number of children vaccinated against polio.
The intervention also resulted in improvements regarding patients' health practices, disease incidence, and malnutrition rates. The presence of community health agents (CHAs) led to an increased likelihood of children being breastfed until age 2, adopting appropriate practices for children's waste management, and children being up-to-date with their immunization schedule. In terms of health outcomes, children showed a decrease in the likelihood of having a cough in the two weeks prior to the interview and a reduction in the incidence of malnutrition indicators.
- Lessons in Public Policy
This article presented evidence that offering public service positions with career opportunities for community work attracts community health workers. In general, this behavior does not imply a reduction in the level of pro-social behavior of community health workers, and the availability of these positions led to improvements in the quality and frequency of services provided in districts with community health workers. Access to these services resulted in improvements in the care and health indicators of children in the respective communities.
Although the results of this study demonstrate the effectiveness of this type of policy, the authors emphasize that selection mechanisms are not usually random. In this sense, policymakers must be careful when adopting strategies that properly select the most suitable candidates for the desired positions, requiring appropriate incentives for both recruiters and recruiters.
References
Aron, A., McLaughlin-Volpe, T., Mashek, D., Lewandowski, G., Wright, SC, and Aron, EN (2004) Including others in the self. European Review of Social Psychology, 15(1), 101-132.
Ashraf, N., Bandiera, O., Davenport, E., and Lee, S.S. (2020) Losing Prosociality in the Quest for Talent? Sorting, Selection, and Productivity in the Delivery of Public Services. American Economic Review, 110(5), 1355-1394.
