³Ô¹ÏºÚÁÏ

Problem

Formal education expanded rapidly across many low- and middle-income countries in the late twentieth century. As a result, a substantial share of the young people now entering urban labour markets have far more schooling than their parents did. But qualifications alone don't determine who gets ahead, and little is known about which non-cognitive attributes employers demand from educated young professionals. Part of the difficulty is measurement: the way that someone behaves as a manager is normally observed only among people who already hold the job. If managerial behaviour is shaped by background, and firms consistently prefer one style over others, then behaviour could be quietly locking some people out of professional careers — and passing exclusion from one generation to the next.

Science

We built a new way to measure how young professionals behave as managers before they're ever hired as one. We invited 982 young professionals in Addis Ababa to a video studio, where paid actors played workplace characters: an absent employee, a supplier failing to deliver, a worker asking for a pay rise. We asked each participant to respond on camera, as if they were the manager in the room. Two coders, working independently, classified every response by the action taken, the source of authority invoked, the justification given and the tone used.  We then showed those recordings to hiring managers at 576 established Ethiopian firms, who ranked the candidates they saw. Using a statistical model that groups people by patterns in their behaviour, we describe four distinct management styles; we term these 'rule-based', 'affiliative', 'power-based' and 'discretion-based'. Rule-based managers follow formal company policy and hold a clear, assertive line. Affiliative managers seek common ground and often give way. Power-based managers also refuse requests, but on their own personal authority. Discretion-based managers refuse without offering a clear rule or reason. The statistical model describes each young professional as being a mixture of these styles.

Outcome

Firms strongly preferred the rule-based style. We estimate that about 70% of firms would rank it first for entry-level managers and about 20% the power-based style; almost none would put the affiliative or discretion-based styles first. The same preference appears whether candidates are judged as future employees or as future entrepreneurs. Firms do differ somewhat from one another, but we find no meaningful differences by sector, by public or private ownership, or by the assessing manager's own management-practice score — a contrast with existing evidence that firms differ in the senior leaders they want. Workers agreed: in focus groups with 66 employees across 17 firms, each shown five candidates spanning the four core styles plus a mixed profile, 57 of 132 rankings placed the rule-based candidate first, well ahead of any other. Employees valued it because they expected to be treated fairly and consistently. They ranked the affiliative style last, and follow-up discussion suggested they saw managers who seek common ground as impressionable and prone to favouritism.

The rule-based style is also, descriptively, the style of people who have already got on. Rule-based managers reported higher reservation wages and profits, had spent less time unemployed, had moved between jobs less often, and were more likely to hold a management position already and to have a degree. And they were more likely to come from more educated families: among participants who had not received a workplace placement, the rule-based style was 8 percentage points less common for those whose parents had not finished primary school.

That last comparison is not causal on its own. But all our participants had taken part, five years earlier, in an experiment that randomly assigned some of them to shadow a manager in an established firm — so we can test whether exposure causes the style. It does. Those randomly assigned to a placement were more rule-based and less discretion-based, and the effect came entirely from participants whose parents had not finished primary school, among whom the rule-based share rose by 5.6 percentage points. The placement shrank the gap in style between those from less- and more-educated families from 8 percentage points to 2.3. Hiring managers responded accordingly: among candidates with no placement, one from a less-educated family had only a 45% chance of being ranked above an otherwise comparable candidate from a more-educated one — even though managers had no way of observing anyone's family background. Among candidates who had received a placement, that gap is no longer statistically distinguishable from zero, and a placement raised a candidate's chance of being ranked above an otherwise comparable candidate without one to 55%.

Change

Policy to help disadvantaged jobseekers has traditionally focused on qualifications or on helping people signal the skills they already have. Our research points to a third path: helping young professionals develop the professional behaviours that firms want. Time inside an established firm is one route the evidence supports, and its benefits fall precisely on those least likely to have absorbed these norms at home. That makes a case for designing and targeting early workplace exposure — internships, placements, shadowing — at the young people least likely to get it otherwise.

Girum Abebe, International Finance Corporation; Marcel Fafchamps, Stanford University; Michael Koelle, OECD; Simon Quinn, ³Ô¹ÏºÚÁÏ; Tom Schwantje, Bocconi University

'Management style under the spotlight: evidence from studio recordings', working paper, February 2026