Measuring Rents from Public Employment: Regression Discontinuity Evidence from Kenya

Abstract

Public employees in many developing economies earn much higher wages than similar privatesector workers. These wage premia may reflect an efficient return to effort or unobserved skills, or an inefficient rent causing labor misallocation. To distinguish these explanations, we exploit the Kenyan government's algorithm for hiring eighteen-thousand new teachers in 2010 in a regression discontinuity design. Fuzzy regression discontinuity estimates yield a civil-service wage premium of over 100 percent (not attributable to observed or unobserved skills), but no effect on motivation, suggesting rent-sharing as the most plausible explanation for the wage premium.

Enrollment without Learning: Teacher Effort, Knowledge, and Skill in Primary Schools in Africa

“…In practice, however, deviations from the official rule appear to be relatively common. For example, one-third of the 18,000 new teaching posts in Kenya in 2010 were misallocated, in the sense that district education officers deviated from the official algorithm to favor certain applicants (Barton, Bold, and Sandefur 2017).…”

“…There is also evidence suggesting that teachers are well-paid relative to other workers with similar educational background. Barton, Bold, and Sandefur (2017), for example, find, exploiting the Kenyan government's algorithm for hiring new teachers in 2010 in a regression discontinuity design, a civil service wage premium of over 100 percent.…”