Advertising Higher Earnings during Recruitment Processes for Community Service Providers in Uganda

Researchers:
Fieldwork by:
Sample:
11,708 women
Timeline:
2012 - 2014
Target group:
  • Job seekers
  • Women and girls
  • Adults
Outcome of interest:
  • Employment
  • Provider Performance
  • Worker satisfaction
Intervention type:
  • Recruitment and hiring
  • Monetary incentives
AEA RCT registration number:
AEARCTR-0001852
Data:
ICPSR
Partners:

Advertising higher wages during recruitment processes may not always attract high-performing candidates. The earning numbers advertised could change candidates’ perception of the job and alter their decision to apply. Researchers conducted a randomized evaluation to test the impact of expected earnings on candidates’ perception of a community health worker position in Uganda, and on the resulting size and composition of the applicant pool. With limited information about the job, higher expected earnings attracted a larger pool of applicants, but also discouraged those with pro-social motivations from applying.

Policy issue

In many low- and middle-income countries, access to public services, such as health care, is limited in rural areas. A popular strategy among organizations to reach remote communities is the community-based approach, which recruits workers locally and trains them to serve in their own communities. The job responsibilities of workers hired in the community-based approach often combine pro-social tasks, which are intended to benefit others, with profit-generating tasks that support the organization’s financial sustainability. These tasks typically involve selling products for a small profit, a portion of which is used to compensate the worker. However, organizations concerned with worker effort and retention may be interested in recruiting workers that are more aligned with the pro-social nature of the job. Previous research suggests that higher wages may give the impression that a job is more difficult or demanding, and that job listings advertising higher wages may receive more applications. Does increasing the advertised pay of jobs during recruitment attract workers with more private ambitions, such as career advancement or earning money, at the cost of attracting those with more pro-social motivations? How does recruiting less socially motivated workers affect their job performance, and the quality of the organization’s services?

Context of the evaluation

Rural communities in Uganda often do not have access to public services. Accordingly, performing community services work as a part-time secondary job is common in Uganda. In 2014, a survey found that 84 percent of non-governmental organization workers in rural western Uganda reported working in agriculture or a self-employed business as their primary occupation, while 98 percent of community health workers hired by the Ugandan Ministry of Health reported their community role as a secondary occupation.

BRAC, one of the world’s largest non-governmental organizations, operates a Community Health Promoter (CHP) program in Uganda. At the beginning of the evaluation, focused specifically on rural areas, BRAC’s operations in Uganda included 128 branches and 2,500 CHPs throughout the country. Each branch includes multiple CHPs under one supervisor, who is in charge of visiting each CHP in the field each week and organizing a one-day refresher training course each month. The role of the CHP combines social-oriented tasks: visiting households, educating household members on health issues, providing pre- and post-natal checks, with business-oriented tasks: selling health products like pain relievers, soap, and oral rehydration salts, as well as household goods like salt, oil, and napkins.

BRAC in Uganda recruits CHPs solely from members of its in-country microfinance groups, which are composed of all women. Being a CHP is a part-time job with flexible hours, often performed on top of a different primary occupation. Monthly earnings vary widely between CHPs. A 2012 BRAC survey found that incomes range from as low as 7 UGX (US$0.002) to as much as 200,000 UGX (US$58.82) per month. The survey also found that the average CHP dedicates fifteen hours per week to the job.

A Community Health Worker teaches nutrition to members of local households
© Stephan Gladieu / World Bank / CC BY-NC-ND 2.0

Details of the intervention

Researchers collaborated with BRAC to evaluate the impact of variation in expected wages on candidates’ perception of job characteristics, the likelihood of a candidate applying for the position, and the traits and performance of accepted applicants. Across different villages, participants were randomly assigned to the following pay groups, with compensation offers remaining constant within a given village:

Information Intervention

  1. High-Pay Group (2,270 participants): Participants were shown leaflets with the CHP job description, indicating that they would earn up to 200,000 UGX ($US58.82) per month.
  2. Medium-Pay Group (2,279 participants): Participants were shown leaflets with the CHP job description, indicating that they would earn up to 30,000 UGX ($US8.82) per month.
  3. Low-Pay Group (2,296 participants): Participants were shown leaflets with the CHP job description, indicating that they would earn at least 7,000 UGX ($US2.06) per month.

After reading the leaflets, half the participants answered questions about their expected earnings for the CHP position and the other half answered questions about their perceptions of the job. Participants’ responses helped the research team to study the impact of changing expected earnings on various aspects of the role, such as how CHPs allocate time across tasks, how many hours they work, and how difficult they perceived the job to be. In a separate intervention, participants learned about the job through the leaflet, but were offered a chance to apply for the job. Similar to the previous intervention, applicants were randomly assigned into one of the following three pay groups:

Recruitment Intervention

  1. High-Pay Group (1,614 participants): Participants were shown leaflets with the CHP job description, indicating that they would earn up to 200,000 UGX ($US58.82) per month.
  2. Medium-Pay Group (1,632 participants): Participants were shown leaflets with the CHP job description, indicating that they would earn up to 30,000 UGX ($US8.82) per month.
  3. Low-Pay Group (1,617 participants): Participants were shown leaflets with the CHP job description, indicating that they would earn at least 7,000 UGX ($US2.06) per month.

Microfinance groups were randomly assigned within their branches to different pay groups.

In each branch, a BRAC officer in charge of recruitment made an unexpected visit to newly formed microfinance groups and advertised the CHP position to group members. All members were asked to answer questions about their pro-social preferences, education, main occupation, etc. and received a copy of the same leaflet that was given to participants in the information intervention. Then, each member spoke with the BRAC officer privately in a separate room and was asked whether she wanted to apply. After interviewing all the interested candidates, the BRAC officer appointed one candidate to the CHP position. Researchers studied the impact of changing expected earnings on the likelihood of a candidate applying for the position, and the degree of pro-social motivation of applicants and appointed CHPs.

Results and policy lessons

Fewer socially motivated participants applied when advertised wages were particularly high. Although higher wages increased the total number of applicants and workers hired, workers attracted by the potential for higher earnings also had lower performance and retention.

Application pool size: In the recruitment segment, an average of two people applied for each vacancy in the low-pay group. By comparison, an average of three people applied for each vacancy in the high-pay group. The medium-pay group also saw more applicants than the low-pay group, suggesting that higher advertised pay increased the likelihood of an individual applying for a job.

Motivation of applicants: In the information segment, candidates in the high-pay group were 6.9 percentage points more likely than the comparison group to perceive the CHP position as one in which private goals and money-earning take priority over social goals like improving community health. This represented a 17 percent increase from 40.3 percent of people in the comparison group who shared the same perception. Candidates who had never volunteered in the health sector were more likely to apply when a higher wage was advertised to them. Similarly, those who had volunteered in the health sector before were less likely to apply when the advertised salary was high. These results suggest that in cases of incomplete information about a job, large increases in pay signaled to pro-social candidates that the job was more oriented towards private benefits, and may have reduced their interest in applying.

Staff retention: Low-pay CHPs stayed in their jobs longer than medium- and high-pay CHPs did. After two years, high-pay CHPs were about 14 percentage points more likely to quit than low-pay CHPs, corresponding to a 72 percent higher dropout rate.

Employee performance: Higher performance of CHPs was associated with more pro-social attitudes. Accordingly, low-pay CHPs performed better than medium- and high-pay ones. Low-pay CHPs performed more pre- and postnatal checks, one of the key responsibilities of the role, than their peers who received higher earnings. Similarly, low-pay CHPs received higher performance scores.

Utilizing these results, BRAC made changes to its recruitment guidelines and practices. The organization now emphasizes more that the CHP role is a volunteering position, while downplaying the incentives.