“There is a widespread belief that disadvantages in cognitive ability and family background are difficult to overcome.”

In The Rebirth of Education: Schooling Ain’t Learning,1 development economist Lant Pritchett describes the challenges of education in underdeveloped countries. Because so many of the problems that he identifies are endemic to centralized, state-run education systems, I view the book as an instant classic in a genre that might be termed “applied libertarianism.”

Pritchett does not advocate that government get out of education. On Milton Friedman‘s idea of vouchers, Pritchett offers no more than lukewarm support. Pritchett views the market-oriented reforms implemented in Chile as a failure. At the same time, he is unflinching in his criticism of the education policies that are popular with the educational establishment.

Pritchett offers a first-person anecdote describing a meeting he witnessed in Uttar Pradesh, a poor province in India, between a school principal and dissatisfied villagers. The principal responds to their complaints by saying,

“It is not our fault. We do what we can with your children. But you [are] right, you are brutes and donkeys. The children of donkeys are also donkeys. We cannot be expected to teach your children. They come from your homes to school stupid and you cannot expect that they will come home from school anything other than stupid.”

While few would be so blunt, I think that many educators of low-achieving students are inclined to say, “It is not our fault.” There is a widespread belief that disadvantages in cognitive ability and family background are difficult to overcome.

Pritchett is a strong proponent of the use of randomized controlled trials to assess educational interventions, and his book is filled with the results of such trials, mostly conducted in developing nations. Many such experiments fail to yield quantitatively significant results. Although Pritchett does not say so, these results can be viewed as confirming the insensitive school principal’s basic outlook.

Perhaps the most striking example of an intervention with disappointing results is school enrollment itself. He writes,

The astounding fact is that the average developing-country adult in 2010 had more years of schooling, 7.2, than the average adult in an advanced country in 1960, 6.8… Haiti [now has a more-schooled population] than France and Italy had in 1960.

For more on these topics, see the EconTalk podcast episodes Pritchett on Education in Poor Countries and Hanushek on Educational Quality and Economic Growth. See also Education Vouchers, by Pedro Schwartz. Library of Economics and Liberty, June 3, 2013 and Education by Linda Gorman in the Concise Encyclopedia of Economics.

Notwithstanding this phenomenal increase in the number of school years completed in underdeveloped countries, there has been little corresponding increase in the ability to read or do arithmetic. Pritchett describes a “flat learning profile,” meaning that additional years of schooling result in only small increments of additional skills. One illustrative example of this profile is Tanzania.

By the end of grade seven in Tanzania, only 41 percent of students are proficient in the [grade two learning standards]… At the current shallow grade learning profile, achieving universal secondary schooling would not even produce universal grade two education.

In India, “The percentage of grade five students who cannot read a story at the grade two level has been rising over the past seven years.”

Specific examples of interventions that have failed to show quantitatively significant results in controlled experiments include:

  • higher pay for teachers
  • all teachers having diplomas
  • all teachers having recent undergraduate training
  • more textbooks for students
  • desks for all students
  • class sizes of 25

On school spending in general, Pritchett writes,

… the basic data tell the big-picture story: countries have exactly the same measured learning results with very different levels of spending. And countries have very different learning outcomes with exactly the same levels of spending… the United States spends PPP $105,000 per child and Poland spends PPP $39,000, and they get nearly the same results. Conversely, Spain spends PPP $74,000 and Finland spends slightly less, PPP $71,000, yet Finland outperforms Spain by 50 points (half a student standard deviation).

He notes that in the United States, spending on education has soared, and yet NAEP (National Assessment of Educational Progress) scores did not move between 1971 and 2004. Other OECD countries tend to show similarly minimal improvement, sometimes with very large increases in spending.

However, there are some results that do appear to suggest robust improvements in education. Pritchett points to studies that suggest that tutoring and selective remedial instruction can make significant differences. Also, making school performance data available to parents can help, particularly among private schools. He writes,

Researchers in Pakistan ran an experiment of choosing some villages in which every child was tested in grade three, and “report cards” for both public and private schools were provided to parents… many of the private schools that were “bad” on the initial report card improved substantially—by, on average 0.34 effect sizes… many private schools that were revealed to be bad just closed… good private schools that stayed open did not improve much, though they did lower fees… there was some modest improvement in government schools, but no closures. The learning gains in bad government schools were only at 0.078 effect size—4.5 times smaller than the learning gains in bad private schools.

Indeed, Pritchett makes several points that go against the conventional view that looks favorably on government provision of education. Analyzing another experiment conducted in Pakistan, he writes,

The striking thing is that not only are government schools rent extracting, or paying more for inputs and teacher wages than needed, they appear to be actually value subtracting: a teacher in the government system does absolutely worse at producing child learning than a teacher completely on his or her own.

Pritchett criticizes economists who account for government’s role in schooling based on theories of externalities.

The basic structures of the schooling systems were therefore laid down not by technical consideration of what would lead to the efficient production of skills or by any of the ideas of public economics about externalities and market failures… The centralization of France, the federalization of Germany, the localism of the United States, and “choice” in Holland were not the result of debates over the relative technical efficacy of these different systems but of the differences between the state and the population in ideas about legitimate socialization.

Pritchett asks why governments provide public schools, rather than simply provide funding to enable parents to enroll children in school. He argues that the answer is that schooling arises in the context of a nation state, where governments have a strong interest in the values that are inculcated through school. He notes, “The historical rise of the modern school was therefore everywhere and always a contest for the control of socialization.”

Pritchett denies that there are scale economies in schooling that would justify centralized school systems. He notes that most professional service providers, such as dentists, lawyers, and architects operate in small firms.

Pritchett believes that schools can learn to improve. However, he points out that experience-based learning takes place not so much within schools as it does within a system that allows for entry and exit. The arrival of new schools that are good and the closing of old schools that are bad is the fastest road to improvement.

In conclusion, Pritchett believes that the ideal school system would have the characteristics of a vibrant emergent order.

  • Open: How is the entry and exit of providers of schooling structured?
  • Locally operated: Do those who manage schools and teach in schools… have autonomy over how their school is operated?
  • Performance pressured: Are there clear, measured, achievable outcome metrics?
  • Professionally networked: Do teachers feel a common professional ethos and linkages among themselves?
  • Technically supported: Are the schools, principals, and teachers given access to the technical support they need?
  • Flexibly financed: Can resources flow naturally (without top-down decision-making) into those schools and activities within schools that have proved to be effective?

He says that these characteristics can be achieved with a variety of approaches, including community-controlled schools, private providers, small governmental jurisdictions, and charter schools. Which approach would work best in a given country depends on the overall cultural and institutional context.

I strongly recommend The Rebirth of Education because of the practical, experience-based analysis that supports emergent, decentralized solutions. Pritchett explains clearly and with convincing evidence that the government-engineered approach is flawed, and that local knowledge along with the “exit” option offers the best hope. Although he is focused only on education in developing countries, I think it should be clear that his analysis applies equally to developed countries and to all issues of public policy.


Lant Pritchett, The Rebirth of Education: Schooling Ain’t Learning. Center for Global Development (Columbia): 2013.


*Arnold Kling has a Ph.D. in economics from the Massachusetts Institute of Technology. He is the author of five books, including Crisis of Abundance: Rethinking How We Pay for Health Care; Invisible Wealth: The Hidden Story of How Markets Work; and Unchecked and Unbalanced: How the Discrepancy Between Knowledge and Power Caused the Financial Crisis and Threatens Democracy. He contributed to EconLog from January 2003 through August 2012.

For more articles by Arnold Kling, see the Archive.