Unintended Consequences of Merit-Based Teacher Reform in Colombia

While school enrollment rates have increased in nearly all developing countries in recent decades, student performance has remained largely stagnant. For instance, by 2010, elementary education in Colombia achieved near-universal enrollment, yet Colombian students’ scores at the 75th percentile fall below the 25th percentile scores for students in OECD countries. These learning gaps have prompted policymakers to reform the hiring mechanisms of public teachers in an effort to improve student outcomes. In The Unintended Consequences of Merit-Based Teacher Selection: Evidence from Large-Scale Reform in Colombia (NBER Working Paper 33008), researchers Matias Busso, Sebastián Montaño, Juan S. Muñoz-Morales, and Nolan G. Pope use administrative records from Colombia’s Ministry of National Education to examine how a nationwide merit-based teacher hiring reform affected the educational outcomes of public school students.

A merit-based teacher hiring reform succeeded in recruiting teachers with strong cognitive skills, but the replacement of more experienced teachers led to a decline in student outcomes.

The reform, implemented in 2005, replaced a decentralized teacher hiring system with a centralized merit-based approach. The new system selected new teachers based on a standardized entrance test of subject knowledge and teaching aptitude. While private schools maintained hiring autonomy, public schools — which educate 70 percent of high school students — were required to follow these new regulations for their teacher hiring. The reform additionally increased starting salaries for teachers to make public school teaching positions more attractive.

Teachers hired under the new system seemed to be more qualified as they scored, on average, 17 percentile points higher on pre-college examinations than teachers hired under the previous system. However, the reform unintendedly increased teacher turnover and decreased the overall stock of teacher experience, as experienced teachers were replaced by novice teachers. The proportion of teachers with less than five years’ experience rose from 10 percent in 2002 to 30 percent by 2008.

Prior to the reform, from 2000 to 2004, the gap in test scores between public and private school students was stable, evolving in parallel. Post-reform test scores of public school students decreased by 8.2 percent of a standard deviation compared to private school students’ scores. This decline is equivalent to that associated with having a teacher who performs 1 standard deviation below average. Mathematics and English scores showed the largest decreases at 14 and 16 percent of a standard deviation, respectively. The likelihood of enrolling in college within six months of graduation fell by 3.3 percentage points, a 21 percent reduction, and college graduation rates within six years of high school graduation decreased by 0.9 percentage points (a 10 percent reduction).

The researchers’ findings suggest that more exposure to novice teachers after the reform was associated with larger declines in public school student outcomes. In schools where more novice teachers were hired, the decline in student outcomes was more than twice as large as in schools where no novice teachers arrived.

— Leonardo Vasquez