Career Choices in Rigid Labor Markets: Political Connection as an Insurance Device

Political connections shape access to economic opportunities in many developing countries, yet their influence extends beyond labor-market outcomes to earlier decisions about education and careers. Using household survey data from Benin, we document that individuals with political connections are more likely to aspire to top-tier public-sector jobs, less likely to pursue entrepreneurship, less likely to drop out of school, and more likely to aim for higher education. These patterns suggest that connections affect behavior well before labor-market entry. However, observational evidence alone cannot determine whether these differences arise from access to connections themselves or from beliefs about how jobs are allocated.

The theoretical framework posits that labor-market outcomes depend primarily on individual qualifications and ability, but that political connections can partially substitute for these determinants. By reducing job search costs and increasing the likelihood of securing desirable positions—especially in the public sector—connections make certain career paths more attainable. Anticipating these advantages, individuals adjust their educational and occupational choices toward jobs that connections can help secure. Because such jobs typically require minimum educational credentials, individuals—even those with relatively low academic performance—may aim to meet these thresholds rather than invest in higher levels of human capital. In this way, political connections shape incentives by altering both access to jobs and the perceived returns to education.

The intervention is an in-the-field survey experiment designed to isolate this mechanism by exogenously shifting perceptions about job access. The study targets parents and secondary-school students in Benin and presents them with structured scenarios describing the transition from school to the labor market.

The information treatments rely on carefully standardized vignettes presented to both students and parents, with parallel structures across groups. All students receive a baseline scenario describing an individual (or child) completing secondary school and choosing among similar post-secondary options, with job access formally tied to educational performance. In the control condition, job access is framed as competitive and merit-based, highlighting the role of exams, individual ability, and the costs associated with job search. The experimental variation lies in how job access is framed. In the treatment conditions, the wording shifts to portray job access as more predictable and implicitly facilitated through informal channels, particularly through political network for public sector jobs. For parents, the scenarios embed the household’s decision over educational investments, while for students they focus on educational aspirations and career choices. Importantly, political connections are not explicitly mentioned, especially for students, allowing participants to infer the role of networks indirectly.

By randomizing these informational environments, the intervention identifies how beliefs about job access—rather than observed connections themselves—shape expectations and forward-looking decisions prior to labor-market entry.

2026-04-28

2026-05-31

Parents: Investment in Children and Career Orientations

Students: Career choices and School Performance

Parents

Investment in Children Education: This outcome measures parental investment in the child’s education over the short term. Parents are asked to indicate which actions they plan to take in the next 6 to 12 months (selecting all that apply). The options capture different dimensions of educational investment, including financial investments (such as paying for tutoring, exam preparation materials, or school fees), time reallocation (such as reducing the child’s household responsibilities to free up study time), and encouragement of longer-term educational attainment (such as supporting continued education at the university or technical level). The question also includes an alternative reflecting lower educational investment—encouraging the child to enter the labor market earlier—as well as the option of taking no action. Together, these responses provide a multidimensional measure of parental commitment to human capital investment.

Career Orientations: This outcome measures parents’ career orientation for their child. Parents are asked to indicate which career path they would most encourage their child to pursue. The response options distinguish between employment in the public sector, employment in the private sector, entrepreneurship, and uncertainty. This measure captures parents’ preferences over broad occupational trajectories and reflects how they prioritize stability, returns to education, and perceived access to opportunities across sectors.

Students

Career choices: This outcome measures students’ career choices. Students are asked to indicate which path they would choose if they were in the situation described. The response options include employment in the public sector, employment in the private sector, entrepreneurship, vocational training, and uncertainty. This measure captures students’ preferences over alternative career trajectories and reflects how they evaluate different opportunities in terms of returns, accessibility, and risk prior to labor-market entry.

School Performance: This outcome measures students’ effort toward improving school performance. Students are asked to report the likelihood that they would increase their study effort to achieve the minimum qualification required, as well as how they would respond if they encountered academic difficulties. The response options capture several dimensions of effort, including attending additional classes or tutoring, retaking exams if necessary, and reducing leisure time to study. Each behavior is measured on a Likert scale ranging from “very unlikely” to “very likely.” Taken together, these items provide a multidimensional measure of students’ willingness to exert effort to improve academic performance and meet educational thresholds.

Students: Job Prospects and Search Costs across sectors, Perception about labor market fairness

Students

Job Prospects and Search Costs: This outcome measures students’ beliefs about job prospects and job search costs across sectors. Students are asked to report, on a scale from 0 to 100, the perceived likelihood of finding a job after graduation in the public and private sectors, as well as the expected costs associated with job search in each sector. These measures capture two key dimensions of labor-market perceptions: expected job accessibility (probability of employment) and the costs of securing employment. By comparing beliefs across sectors, this outcome provides insight into how students perceive differences in opportunities and barriers, which are central to understanding their educational and career decisions.

Perception about labor market fairness: This outcome measures students’ perceptions of fairness in the labor market, specifically in the public sector. Students are asked to evaluate how fair they perceive the hiring process to be, using a Likert scale ranging from “very unfair” to “very fair.” This measure captures beliefs about the extent to which hiring is based on merit versus other factors, such as informal networks or connections, and provides insight into how students perceive the integrity of public-sector recruitment.

We test whether perceived access to political connections shapes educational investment, student effort, and career choices prior to labor-market entry. We expect that households exposed to information suggesting that job access is facilitated through networks will be more likely to orient students toward public-sector careers. Because these jobs still require minimum educational credentials, we also expect higher incentives to invest in education and exert effort to meet these thresholds. Finally, we expect the treatment to shift beliefs about labor markets, making public-sector jobs appear more accessible and less costly to obtain.

To test these hypotheses, we implement a randomized control trial with households in Benin. The study targets households with at least one adult household head and one student aged 15 or older. Within each household, we interview one household head and one randomly selected eligible student, allowing us to jointly capture parental decisions and student responses.

Households are randomly assigned to one of two study conditions, with randomization conducted at the household level and blocked by village to ensure balance across local environments. Assigning treatment at the household level ensures that both the household head and the student are exposed to the same informational environment, thereby avoiding conflicting signals within the household and preserving the coherence of decision-making.

In the competition-based labor-market condition, which serves as the control group, respondents receive vignettes describing a labor market in which job access depends on academic performance, exam success, qualifications, and the costs of job search. In the political-connection condition, respondents receive parallel vignettes in which job access appears more predictable and implicitly facilitated through informal networks, particularly for public-sector employment. Importantly, the relationship between political connections and public sector employment is never mentioned; instead, the treatment varies how job access is described, allowing participants to infer the role of networks indirectly.

We then measure a set of primary and secondary outcomes immediately following the information treatment. For parents, primary outcomes capture educational investment and career orientation for their child, while for students they include career choices and intended academic effort. Secondary outcomes, measured for both groups, capture beliefs about labor-market conditions, including perceived job access, job-search costs, and fairness in hiring across sectors. Together, these outcomes allow us to evaluate how changes in perceived job access influence behavior and beliefs prior to labor-market entry.

Not available

We implement randomization directly in the field as part of a nationally representative survey of 1,862 households. We first randomly select 62 villages across regions and then randomly select 30 households within each village. After collecting background information on households and household heads, we assign households to study conditions using the household head’s month of birth. Households whose heads were born in even-numbered months—February, April, June, August, October, and December—are assigned to the political-connection vignette, while households whose heads were born in odd-numbered months are assigned to the competition-based labor-market vignette. This procedure assigns treatment at the household level, ensuring that both the household head and the selected student receive the same vignette. Birth months are unrelated to the experimental outcomes, and each household has a one-half probability of assignment to either the treatment or control condition.

Households

No

1865 households

1865 households

Given the direct in-the-field randomization nature of the experiment, we use simple randomization procedure and do not know yet how many households will fall in the treatment group and how many will be assigned to control group. We will have this information at the end of the field work, which is ongoing.