What was the problem?
In Zambia, as in most of sub-Saharan Africa, the healthcare system is undermined by shortages of qualified staff, particularly in rural areas.
In 2010 the Zambia Ministry of Health created a new post of 'Community Health Assistant' to address this shortage. Successful appointees would become civil servants, receive a year's training and be appointed to serve in rural areas, where they would be eligible for promotion to higher cadres within the ministry.
Promotion prospects were expected to attract talented and motivated workers. However, there was a risk that these would attract the “wrong type” of workers, who have a loose connection with their local communities, perform poorly on the job and abandon their posts for better-paid jobs in urban areas, as had been the norm in most of sub-Saharan Africa.
What did we do?
LSE Professor of Economics Oriana Bandiera spent over four years analysing employee motivation and her work was awarded the 2007 IZA Young Labor Economist Award. On the basis of this work, Bandiera was approached by Nava Ashraf, Associate Professor in the Negotations, Organizations and Markets Unit at Harvard Business School, to collaborate on research comparing schemes for motivating urban health workers in Zambia.
The researchers tested four different strategies to motivate 1,200 agents hired to distribute condoms. They found that rewards, especially non-financial rewards, leveraged the agents’ motivation for the cause, making highly motivated agents sell more condoms.
Based on these research findings, the Zambia Ministry of Health contacted the researchers to collaborate in the design of recruitment and motivation strategies for its first 314 community health assistants in July 2010. The collaboration then continued as three randomised controlled trials, designed by Bandiera and her colleagues, were implemented by the Ministry with technical support from the researchers.
The first trial was designed to measure the effect of the job's 'mission statement' on the selection, retention and performance of the community health assistants. In half of the districts the recruitment poster stated that the work would 'help the community'. In the other half the poster suggested that the new post would lead to rapid career advancement. Once the trainees began working in the field in July 2012, the skill levels, social motivations and ambitions of the two sets of recruits were compared against performance measures.
The second trial was designed to measure the effects of peer comparisons during training, while the third trial assessed the effects of autonomy at work (see descriptions below).
'The research has…fundamentally altered staff recruitment and retention strategies within the Zambia Ministry of Health. These changes are significant, affecting the recruitment and motivation of 5,000 new community health assistants and informing similar changes in the recruitment and motivation of all new nurses, doctors and civil servants.'
After a year, the results of trial one demonstrated a noticeable difference in performance between the two groups. Applicants recruited on the basis of 'career opportunity' had more advanced skills, stronger scientific backgrounds and higher performance ratings. Both groups, however, attained similar scores in terms of commitment to the local community and retention rates were similar in the two groups. Career incentives, far from selecting the "wrong" types, attract talented workers who deliver health services effectively.
The second trial formed part of the year-long training schedule. Four kinds of non-financial reward were offered to incentivise trainees to study and the motivational effects of each one were measured using feedback and ranking. Peer comparison feedback was found to reduce performance whereas professional and community recognition enhanced it. In the light of these mixed results, the Ministry of Health decided not to implement social recognition schemes.
Trial three began in December 2012 and aimed to quantify the effect of work autonomy on performance. The community health assistants were split into two groups. In the first group performance goals were set for them by their supervisors, whereas in the second group they were allowed to set their own goals. The hypothesis is that goal-setting can be an effective tool to foster performance and retention. This experiment is on-going.
The results of these trials have influenced the Zambia Ministry of Health in its approach to the recruitment and motivation of community health assistants. As a result of trial one, the Zambia Ministry of Health is using the career advancement poster in its ongoing recruitment efforts, and as a result of trial two, it has decided to abandon social recognition and reward schemes once workers are deployed in the field.
The research has therefore fundamentally altered staff recruitment and retention strategies within the Zambia Ministry of Health. These changes are significant, affecting the recruitment and motivation of 5,000 new community health assistants and informing similar changes in the recruitment and motivation of all new nurses, doctors and civil servants.
The Ministry has also embraced the idea that field trials are a useful way to assess and develop employment policy and has welcomed continued collaboration with the research team on evidence-based policymaking.