FE43DF93-13CD-493E-BEF2-A77A9E7D6926 Copenhagen Consensus Center Logo
Copenhagen Consensus Center

Post-2015 Consensus: Education Assessment, Psacharopoulos

Assessment Paper

Unsurprisingly, the MDG targets that are not likely to be achieved by 2015 are repeated with 2030 as a new target date. Today, there are currently 60 million children out of school, most of them in Sub-Saharan Africa.  Using an estimate of $300 per primary school student in Africa, an additional $17 billion per year would be needed to reach the zero targets by 2030. However, total international aid for basic education in low income countries in 2011 was just $5.8bn.  What is clear is that the aid allocated to education needs to be spent where it can do the most good and not funneled toward unrealistic targets. 

Investments in expanding any level and type of education pass in general a benefit-cost test. But some education investments are far more profitable than others. For example, studies have found an astonishing benefit-cost ratio of 77 in Kenya (at 3% discount rate) for preschool programs. Therefore, a more modest and pragmatic approach appears warranted. Achieving realistic targets would be better than idealistic ones, which in the end fall short.

Summary of the most high-yielding targets from the paper

Education Targets Benefit for Every
Dollar Spent
Increase the preschool enrollment ratio in Sub-Saharan Africa from the present 18% to 59% $33
Increase the primary enrollment ratio in Sub-Saharan Africa from the present 75% to 100% $7
Increase student test scores by one standard deviation $4
Ensure secondary school completion $4


The Post-2015 MDG discourse has generated a number of education goals to be fulfilled by 2030.  The aim of this paper is to have a closer look at such targets and identify the most concrete and prominent ones that are amenable to cost-benefit analysis.  Based on existing research, the targets are evaluated according to findings in the economics of education literature. 

An overall conclusion is that the vast majority of the Post-2015 education targets could not be achieved by 2030, hence confirming the need to prioritize on the basis of cost-benefit analysis. Preschool and primary education in Africa have the highest benefit-cost ratios.

Historical perspective

There is a long history of international organizations setting quantitative targets for education.  As of today, none of these targets has been achieved.

For example, in 1961 a high-level UNESCO conference set a goal of primary enrolment in Africa being raised from 40% to 100% by 1980. In reality, enrolment stood at just 56% in that year. In 1990, UNESCO and other international organisations launched the ‘Education for All’ campaign (EFA), with a goal of universal primary education by 2000. By 1999, net enrolment in Africa was only 57%. The target was reaffirmed and the date shifted to 2015 at the UN Millennium Summit. However, today there are still about 60 million children out of school and over 100 million young people aged 15 to 24 who lack basic literacy skills.

Setting unrealistic targets is not only a problem for developing countries. The European Council meeting in Lisbon in 2000 set a goal that the proportion of early school leavers should be no more than 10% by 2010. By 2008, only six of the 27 Member States had met the target and Eurostat data from 2013 shows that 14% of 18-24 year olds are still early leavers from education and training.

Unfortunately, the lessons of grandiose but unrealistic education target setting have not influenced the current post-2015 MDG proposals. Instead, lack of finance is the most cited reason for failing to meet targets, accompanied by calls for increased foreign aid. However, there are many other reasons for targets being missed. In some cultures, parents to not allow girls to attend school after puberty, while in the poorest communities child labor is needed to supplement family income. Other factors include a lack of understanding of the lifetime value of education, the political climate, a high number of orphans or children with only one parent, or schooling which is simply too poor to achieve much.

Education in the Post-2015 agenda

Proposals put forward are of a very general nature, for example:

  • Establish Satisfactory Education System Accessible to All at All Levels
  • Lifelong Learning
  • Continued Pursuit of Lifelong Learning
  • Equal Right to Education
  • Socio-economic Equality

MDG targets that are not likely to be achieved by 2015 are repeated with 2030 as a new target date. These include reducing adult illiteracy by 50%, having 100% of young people completing primary and lower secondary education and improving school readiness by reducing the number of children not having preschool provision by 50%.

Many targets are expressed in very general terms which defy rigorous economic analysis. However, the range of goals proposed can be grouped into the following major clusters:

  • Preschool
  • Primary
  • Secondary
  • Tertiary
  • Education quality
  • Vocational education
  • Education finance

Cost-benefit analysis of the proposals

There are a number of alternative approaches to making a comparison of the costs and benefits of different levels of educational attainment. For example, comparing university and secondary education, the benefits are the additional lifetime earnings of graduates, while the costs are measured by the expense of keeping the student in education plus the earnings foregone during this time. There are three alternative metrics which express the difference: the internal rate of return, the net present value (NPV) and the benefit-cost ratio.

As an alternative to a full discounting approach, the so-called ‘Mincerian’ or earnings function method can be used. In this, a function of log-wages is expressed in terms of the number of years of schooling and years of labor market experience. It is a slightly inferior method, since it assumes a flat age-earnings profile, but this simplification can also be used in a short-cut version of the full rate of return calculation. The Mincerian approach has become the preferred one in recent years.

In the empirical literature the vast majority of education cost-benefit studies measure rates of return, because it is easy to make comparisons across countries, exchanges and discount rates.  A distinction must be made between the private rate of return, which looks purely at costs and benefits for the student and the social rate of return which also takes account of the state’s cost to provide education and expresses benefits as gross earnings.

In the mainstream economics of education, the social rate of return is traditionally expressed as the private rate adjusted for the full cost of schooling, jointly borne by the individual and the state. These returns, however, are called ‘narrow-social’, because they fail to take account of benefits beyond those accruing to the individual. These externalities which are part of ‘wide-social’ returns can include, for example, lower fertility or lives saved because of improved sanitation conditions followed by a more educated woman who may never participate in the formal labor market.

Unfortunately, the scant empirical evidence on the external effects of education means that social rates of return are usually of the narrow type. Nevertheless, the distinction is an important one.

Cost-benefit evidence on the proposals

Early childhood education:  There are many studies on the effect of preschooling on eventual educational attainment and adult earnings, but very few contain cost-benefit analysis.  However, due to longitudinal data availability on preschoolers in the United States, this country’s early years’ education has been the subject of extensive cost-benefit analysis which also allows estimates of wide social returns.  Three studies gave rates of return of between 7 and 18%, with benefit-cost ratios of 2.7, 6.9 and 7.2 (3% discount rate).

The American studies show that kindergarten has an important effect on adult earnings. A study of graduates in Jamaica also found that a preschool intervention increased average earnings by 42%. In Indonesia, the achievement gap between children from richer and poorer backgrounds was reduced, allowing the poorer children to make better progress later in the education system. Other studies gave benefit-cost ratios of 3.0 and 3.7 for preschool programs in the Philippines and Bolivia, and an astonishing 77 in Kenya (all at 3% discount rate).

Main school system: There are three compilations available of the social rates of return (narrow base) for investment in education in over 100 countries. The World Bank published compilations in 2004 and 2012, and there is an OECD one from 2013.

The first World Bank compilation, based on the full discounting method, shows private returns to be higher than narrow social returns, reflecting the regressive nature of public subsidy, which increases with the level of education.

Returns are inversely related to a country’s prosperity, being lowest for OECD countries and highest for Latin America and the Caribbean and sub-Saharan Africa. Based on the social calculation, primary education shows the highest returns, followed by secondary and then higher education. For Africa, for example, the social returns are 25.4, 18.4 and 11.3 respectively for these three levels of attainment.

In 2012, the World Bank used the Mincerian method and so only private returns are given. In this case, returns are highest for tertiary education: for example 21.9, 10.8 and 13.4% for tertiary, secondary and primary education respectively in sub-Saharan Africa, 1.8, 6.9 and 10.3% for the world as a whole. However, this method grossly underestimates true returns because of the tacit inclusion of foregone earnings to the cost of keeping children in school.

The OECD figures are only for upper secondary and tertiary education, given the universality of primary education among its members. The (narrow) social return is 11.2% for tertiary and 8.4% for secondary education.

We should also note a common finding arising from other studies that educating girls has a higher rate of return than that for boys. A 1992 report, for example, showed the wide social return on girls’ education in Pakistan to exceed 20%.

Education quality: Economic analysis of school quality is sparser than for quantity, since studies typically do not consider the cost of delivering improvements. In fact, many econometric studies show that increased resources have not improved test scores. In a survey of 376 education production functions relating school resources to student achievement, most studies report negative or insignificant effects of expenditure per student, teacher salaries or class size. Institutional changes such as the introduction of monitoring and evaluation systems, central examinations, teacher incentives and accountability are more likely to improve school quality, although difficult to cost.

From the relatively meagre evidence available, we can roughly estimate a 15% average social return to investments in school quality improvements, giving benefit-cost ratios of 5 and 3 for school quality at 3% and 5% discount rates, respectively.

Vocational education: In many countries, the returns to academic education are higher than for vocational training. In a large World Bank follow-up study of students in the technical-vocational curriculum stream of secondary education in Colombia and Tanzania, it was found that the graduates did not seek or find employment in the sector they studied.   Perhaps counterintuitively, general secondary education is more profitable than vocational education.  The reason is that whereas general and vocational secondary school graduates have more or less equal earnings after graduation, the vocational track of secondary schools costs about twice that of the general track. Similar results have been found in other countries.

Education financing: One of the broad goals in the Post-2015 MDG proposals is for ‘sufficient financing’ to be devoted to education. However, financing is a means of achieving goals, not a goal in itself and cannot therefore be the subject of cost-benefit analysis.

Lessons from the literature: A solid conclusion we can draw is that investments in expanding any level and type of education in general passes a cost-benefit test at either a 3% or 5% discount rate. However, some are more profitable than others. Economics Nobel Laureate James Heckman concluded on the basis of a large body of rigorous evidence that skills are acquired most efficiently in the early stages and levels of education.

Present conditions

On the eve of the MDG-2015 deadline for achieving universal primary education, there are about 60 million children out of school, more than half of them in sub-Saharan Africa.  23% of primary school age children in the region are out of school, and even advanced industrial countries fall short of the 100% net enrolment ratio. Preprimary rates are lower and vary widely between regions; in sub-Saharan Africa, enrolment stands at just 18%, compared to 81% in industrialized countries.

There are about 70 million children without access to secondary education, mostly in South Asian and sub-Saharan Africa. For the world as a whole, the latest available enrolment ratio is 62.7%, with a range from 24.7% in sub-Saharan Africa to 92.2% in the EU. There are even bigger disparities when it comes to tertiary education, although there are no specific MDG targets.

Regarding the quality of schooling, 123 million 15-24 year olds lack basic reading and writing skills, and over 60% of these are women. In Pakistan less than half of grade 5 children in Balochistan could solve a two-digit subtraction, compared to 73% in the wealthier province of Punjab. Standards vary widely across the world.

Assessing the Post-2015 MDG proposals

The most suitable basis for benefit-cost assessment is the World Bank compilation published in 2004, giving social rates of return. However, figures derived from this are at the bottom end of the range of wider social return estimates since they take no account of externalities. The benefit-cost ratios for marginal expansion of education under current conditions are above 1 for all levels and regions, but highest for primary schooling.

Universal primary education is the most prominent target in the Post-2015 goals, with sub-Saharan Africa representing the greatest challenge. Average costs will rise and benefit-cost ratios fall as enrolment increases: building more schools in rural areas and hiring qualified teachers puts a strain on resources, hence raising the marginal cost. Assuming the elasticity of the marginal cost curve to be 1 and noting that the cost increases only affect about 20% of the total cost of schooling (the rest being foregone earnings) we find that benefit-cost ratios for Africa between 2015 and 2030 vary from about 5 to 9, depending on the discount rate. Applying the same approach across the world, the benefit-cost ratios range from 3.6 to 6.3.

A further target is a reduction of 50% in the proportion of children not attending early childhood care and education programs. Looking again first at sub-Saharan Africa and using half the reported Kenyan benefit-cost ratio of 77 (to exclude the nutrition element), adjusted benefit-cost ratios are in the range 28 to 39.

The figures in this paper exclude externalities, but we know that educating one member of society is associated with a series of benefits that accrue not only to the educated person but also to others. We have solid evidence that parents’ education has a positive effect on health and child survival.  Children of better-educated parents have a higher chance of survival and are more likely to go to school and receive regular health checks. More educated women have lower maternal and infant mortality rates and improved reproductive health. The case is particularly strong for mothers’ education. Beyond health, it has been found that each additional year of education on average reduces a country’s chances of falling into civil war by 3.6%.

What will it cost?

A number of bodies have estimated the cost of providing education for all to lower secondary level. In 2010 UNESCO estimated that it would take another $16 billion per year in external financing to achieve basic education in low income countries by 2015.  Their latest estimate is for an additional $29 billion per year. Since the UNESCO estimate was for a shortfall of $38bn annually to achieve universal EFA goals by 2015, we can take it that this represents a conservative figure to meet the new goals by 2030, as these are likely to be more ambitious.

There are currently 57 million children out of school, most of them in sub-Saharan Africa.  Using an estimate of $300 per primary school student in Africa, an additional $17 billion per year would be needed to reach the zero target by 2030.  However, total international aid for basic education in low income countries in 2011 was just $5.8bn.  On the other hand, UNESCO) reports that the cost of 250 million children not learning the basics is equivalent to an economic loss of $129bn. 


These estimates should, of course, be considered as approximations. In addition, for the benefits to flow from education, a number of other conditions must prevail; for example, a country should be free from conflict and have established protection of property rights.

The generality, ambiguity and optimism of the Post-2015 MDG targets in the present discourse do not augur well for their implementation by 2030, if not well beyond and a more modest and pragmatic approach appears warranted. Achieving realistic targets would be better than missing idealistic ones. Perhaps ‘Education for All’ should be replaced by ‘Education for Some’, that is, the most needy. But such a mundane term would never fly in international parlance.