Copenhagen Consensus Center
Home Menu

Post-2015 Consensus: Education Perspective, Krafft Glewwe


Overall, we generally agree with Psacharopoulos on the value of improving school quality, and we share his skepticism of vocational and technical education. Yet we would also argue that the evidence for expanding pre-primary and primary education in Sub-Saharan Africa is not as strong as he claims.

The task set is a very difficult one and we question the very idea of this undertaking, largely because the field of education suffers particularly from output-driven goals rather than outcomes. Setting goals that focus on outputs or proxies, such as enrolment rates, creates an incentive to attain that specific goal, which will not necessarily lead to the essential goals or benefits of education, such as human capital, knowledge, and skills. 

It is this ‘how to proceed’ that is crucially absent from Psacharopoulos’s paper. For instance, improving school quality by increasing student test scores is stated to have a benefit-cost ratio between 3.0 and 5.0, but how to improve test scores and school quality is effectively not discussed. Even if ways to achieve more years of schooling can be found and costed, calculating the benefit-cost ratio can be misleading given the lack of data on externalities, measurement errors (particularly in developing countries), the absence of the self-employed from the estimate and the inclusion of government workers, whose salary reflects government policy rather than productivity.

We are therefore concerned about the likely inaccuracy of the quoted figures. This is compounded by the low quality of the evidence presented (and perhaps the quality of evidence available). A further factor is the uncertainty around estimates and potential outliers. In particular, the benefit-cost ratio of preschool programs in sub-Saharan Africa is based on a study in Kenya, with a ratio an order of magnitude higher than others, with no adequate justification. 

A weakness of the challenge paper is that it provides no economic justification for public subsidy of education, since social rates of return quoted are lower than private rates of return. Although Psacharopolous alludes to wider social benefits, he is not able to provide sufficient hard evidence to support his case. 

Research from developed countries shows that the best investment a country can make is likely to be in pre-primary education, particularly for disadvantaged children. This is the most cost-effective stage to invest in and the opportunity cost of children’s time is low at this age. Good quality pre-primary education consistently improves child development and enhances later attainment. On the other hand, investing in universal primary education in sub-Saharan Africa is problematic, given that half the 57 million children out of school are in conflict zones. For those outside conflict zones, primary education is likely to be a worthwhile investment, but schools must be of sufficiently high quality to allow significant learning. In the case of secondary education, there are unfortunately too few jobs requiring formal education to make this a good investment in some less developed countries.

The issue of education quality is important but priorities for cost-effective interventions are difficult to set. However, in all cases studied it is important to assure that teachers are both present and knowledgeable. This in itself may also help to reduce dropout rates.

While we have pointed a number of difficulties with the underlying challenge and with Psacharopolous’s approach, ultimately we come to similar priority targets. Pre-primary education, primary education, and school quality investments are the best education goals. However, there is also a real need for high-quality longitudinal studies to assess the benefits of interventions which may only pay off much later in life.