With rigorous economic research and practical policy solutions, we focus on the issues and institutions that are critical to global development. Explore our core themes and topics to learn more about our work.
In timely and incisive analysis, our experts parse the latest development news and devise practical solutions to new and emerging challenges. Our events convene the top thinkers and doers in global development.
Empirical development economics, economic reform, international migration, poverty, social policy and development.
Lant Pritchett is a senior fellow at the Center for Global Development and professor of the practice of international development at Harvard's Kennedy School of Government, where he taught from 2000 to 2004 and from 2007 onward. Before rejoining the Kennedy School in 2007, he was lead socio-economist in the social development group of the South Asia region of the World Bank. He occupied various other positions at the World Bank during his tenure there, beginning in 1988. Pritchett was a team member on a number of prominent World Bank publications including Economic Growth in the 1990s: Learning from a Decade of Reforms (2005); Making Services Work for Poor People (World Development Report 2004); Assessing Aid: What Works, What Doesn't and Why (with David Dollar, 1998); and Infrastructure for Development (World Development Report 1994). He has published two books with Center for Global Development, Let Their People Come (2006) and The Rebirth of Education (2013). Pritchett has published over a hundred articles and papers (with more than 25 co-authors) on a wide range of topics, including state capability, labor mobility, and education, among many others. Originally from Idaho, Pritchett is the father of three children and now lives in an empty nest with his wife of 31 years.
The last decade has seen considerable progress enrolling children in schools worldwide: today most people live in countries on track to meet the Millennium Development Goal of 100% primary completion by 2015. Sadly, enrollment doesn’t necessarily equal learning. A new report by the CGD Study Group on Measuring Learning Outcomes shows a shockingly wide gap between education inputs and learning outcomes. The report, Schooling is Not Education: Using Assessment to Change the Politics of Non-Learning, finds the learning crisis reflects systemic issues in education sectors worldwide.
President Nancy Birsdall, Senior Fellow Lant Pritchett, and Visiting Fellow Scott Morris are quoted in a New York Times piece on the feasibility of eradicating poverty.
From the article:
At a news conference during the spring meetings of the International Monetary Fund and the World Bank in late April, Jim Yong Kim held up a piece of paper with the year “2030” scribbled on it in pen. “This is it,” said Kim, the genial American physician who took over as president of the World Bank last summer. “This is the global target to end poverty.”
It sounds like the sort of airy, ambitious goal that is greeted by standing ovations but is ultimately unlikely to ever materialize. Development experts don’t see it that way, though. The end of extreme poverty might very well be within reach. “It’s not by any means pie-in-the-sky,” says Scott Morris, who formerly managed the Obama administration’s relations with development institutions. When I asked Jeffrey Sachs, the development economist, if the target seemed feasible, he said, “I absolutely believe so.” And Nancy Birdsall, president of the Center for Global Development, the powerful Washington policy group, told me, “In many ways, it’s a very modest goal.”
In part, this is because the bar is set very low. The World Bank aims to raise just about everyone on Earth above the $1.25-a-day income threshold. In Zambia, an average person living in such dire poverty might be able to afford, on a given day, two or three plates of cornmeal porridge, a tomato, a mango, a spoonful each of oil and sugar, a bit of chicken or fish, maybe a handful of nuts. But he would have just pocket change to spend on transportation, housing, education and everything else. The 1.2 billion people living in such extreme poverty, according to researchers at the Massachusetts Institute of Technology, might own land, but they are not very likely to own durable goods or productive assets — things like bicycles — that might help them raise themselves out of poverty. In such families, about half or three-quarters of income goes toward food.
Of course, making it above the $1.25-a-day mark doesn’t guarantee a white picket fence and a Caddy in the driveway — indeed it doesn’t even guarantee a proper meal. For that reason, some economists have criticized the bank for setting its targets too low. “It’s small,” Pritchett says. “It’s penurious. It’s charity-like. It’s not development.” He says that the billions who live on a bit more than $1.25 a day are still deeply impoverished by any reasonable standard. “Why are we focused on a line, above which nothing happens, set by some technocrats in Washington?” Another 1.2 billion live on between $1.25 and $2 a day, an only slightly less dire form of deprivation.
For the poor living in poor countries, particularly the profoundly unstable ones, gains have been harder-fought and slower, a trend that the World Bank’s own economists describe as worrisome. But that is not to play down the successes so far. In 2008, for the first time since the bank started measuring the statistics, the number of people living in dire poverty and the dire-poverty rate fell in every region around the world. Extreme poverty in sub-Saharan Africa has at last dipped below the 50 percent mark. Still, many within the development world doubt the ability of NGOs to cure the world’s most troubled nations of their woes. “I don’t think we have a recipe for fixing the Congo or South Sudan or Afghanistan,” says Birdsall, of the Center for Global Development.
In an interview, Kim sounded energetic and optimistic about the prospect that the great brute force of growth would keep on lifting hundreds of millions out of poverty — and about the bank’s role in nursing the process along. Given how big the world is, how big the goal is and how diverse economies are, it would take a multipronged approach, he said. For parts of sub-Saharan Africa, it would mean huge electrification projects. For China, it would mean smarter urbanization and clean energy. For India, it would mean enormous infrastructure investments that the World Bank could help finance. It also might mean replicating what has worked for those big, quick-growing emerging economies in poorer, poverty-stricken developing ones.
Read it here.
Most of the world’s children now live in countries on track to meet the Millennium Development Goal of universal primary completion by 2015. Countries have indeed made great progress getting kids in school, but behind that progress is a problem: many children are hardly learning anything in school. Some measures of learning are just dismal. In India, for example, only about one-third of children in grade 5 can perform long division. Nearly one-half cannot read a grade 2 text, and one in five cannot follow a grade 1 text.
What is to be done? Broadly speaking, schools, governments, and donors need to focus more on actual learning goals, not just filling seats. This report of the CGD Study Group on Measuring Learning Outcomes shows how to make some headway in that direction. Governments need to develop comparable, public learning assessments. Civil society should engage at the grassroots to demand accountability. Donors can play a secondary role by pegging funding to results or experimenting with different strategies. And the UN and other multilaterals should set global standards against which national efforts can be measured. One option is to establish a global learning goal as part of the post-2015 development agenda.
The last decade has seen considerable progress enrolling children in schools worldwide: today most people live in countries on track to meet the Millennium Development Goal of 100% primary completion by 2015.
Sadly, enrollment doesn’t necessarily equal learning. A new report by the CGD Study Group on Measuring Learning Outcomes shows a shockingly wide gap between education inputs and learning outcomes – many children finish primary school unable to read, write or do simple addition. The report, Schooling is Not Education: Using Assessment to Change the Politics of Non-Learning, finds the learning crisis reflects systemic issues in education sectors worldwide. It recommends strong assessment regimes as part of the solution.
On May 9, CGD president Nancy Birdsall will chair a conversation on the report with Alice Albright, chair of the Global Partnership for Education; CGD Study Group co-chair Lant Pritchett; Project director Charles Kenny; and other members of the Study Group. They will discuss the findings and implications for education in the post-2015 development agenda.
Here we extend the basic idea of rigorous impact evaluation—the use of a valid counterfactual to make judgments about causality—to emphasize that the techniques of impact evaluation can be directly useful to implementing organizations (as opposed to impact evaluation being seen by implementing organizations as only an external threat to their funding).
CGD is pleased to welcome the David M. Kennedy Center for International Studies for an event honoring Lant Pritchett, recipient of the 2013 Distinguished Service Award.
Since 1984, the David M. Kennedy Center for International Studies at Brigham Young University has presented the Distinguished Service award to individuals who are among the most influential in their field and have made a significant contribution to our global understanding.
Following the presentation of the award, Pritchett will deliver a lecture titled "The Battle for 2015." As we approach the 2015 expiration date of the Millennium Development Goals (MDGs), rich and poor countries alike are reassessing development priorities. Pritchett will discuss what constitutes "development" and the instruments the international community should rely on to implement a new agenda. His remarks will focus on the dilemmas facing developed country governments -- and the US government in particular -- in the run up to the post-MDG era.
Now that a bipartisan group of U.S. senators and the president have made specific proposals to reform immigration law, you’ll see months of controversy. As is frequent in politics, the most controversial policy step will be among the least important.
I’m talking about the provisions, in both proposals, for an “earned path to citizenship” for the 11 million people residing in the U.S. without legal permission. (People who like this plan call it “regularization”, people who dislike it call it “amnesty”.) Showcase policy steps like this don’t affect unauthorized immigration nearly as much as the pathways for U.S. employers to smoothly, flexibly hire foreign labor when they need to.
Read this twice: I am not saying that the status of unauthorized immigrants isn’t important. I’m saying that the policy step of a path-to-citizenship is unlikely to be an important determinant of the unauthorized immigrant population, even in the short term.
This is clear from U.S. history. In 1986, President Ronald Reagan created a path to citizenship for most of the 3 million unauthorized immigrants then living in America when he signed the Simpson-Mazzoli Act. Watch what happened to the unauthorized immigrant population. By 1990, just four years later, you couldn’t even tell that anything had happened:
“These experiences show the importance of enacting a legalization program only in the context of comprehensive immigration reform designed to reduce future unauthorized inflows as much as possible. Flows increased following IRCA because the amnesty did little to reduce the demand for unauthorized workers […]”
The extreme controversy over the 1986 path to citizenship was like worrying about how to pay today’s $100 debt while ignoring the mounting expenses that will give you a $1,000 debt next year. Leaders truly concerned about America and the people who live here have a responsibility to offer long-term solutions.
We don’t know yet exactly what the Senators and the president propose as a long-term solution. The Senators’ proposal claims that it “will provide businesses with the ability to hire lower-skilled workers in a timely manner when Americans are unavailable or unwilling to fill those jobs.” How they do this, and whether or not it truly serves U.S. employers’ needs, will do by far the most to determine whether or not tomorrow’s America is a land of unauthorized immigrants. Orrenius and Zavodny propose a flexible system of auctioned work permits that would be a boon to U.S. enterprise and a big stimulus for the U.S. economy, particularly agriculture (details in their book). CGD’s Lant Pritchett discusses some of the global economic benefits of guest-work arrangements here.
But the exact way of handling future flows is less important than an overall focus on working with American employers. Any immigration reform proposal serious about reducing unauthorized immigration must focus on creating tools for U.S. business, not chains for U.S. business. Past attempts at addressing future flows have been chains instead of tools. Today’s H-2A visa for seasonal agricultural work is a good example. It is so cumbersome and expensive to use that the vast majority of American farmers—who depend critically on migrant labor—refuse to use it. The program doesn’t help farmers get the labor they need, doesn’t help connect them to workers by facilitating recruitment; it exclusively puts up costly barriers for farmers to fight through in order to access legal migrant labor.
If you’re concerned about the phenomenon of unauthorized immigration or the plight of unauthorized immigrants, pay less attention to the path-to-citizenship. Keep one question foremost in your mind as you read the forthcoming details of the Senators’ and president’s proposals: What are they offering U.S. farmers to get the labor they need without going under? What are they offering U.S. parents to get the childcare they need without breaking the bank? These are the provisions that will shape unauthorized immigration for tomorrow’s America.
The Social Progress Index is an effort of the Social Progress Imperative to create a new and better way to compare the human and social development performance of countries. High on their agenda is to not use GDP per capita or other measures of national development, but rather focus on direct measures of human well-being. And it turns out to be a useful measure of the importance of national development. Let me explain.
The Social Progress Index (SPI) has three components:
1. basic human needs,
2. foundations of wellbeing, and
Each of these three components are built of four sub-indicators which are each themselves built up from raw data. For instance, 1) basic human needs has four sub-components:
1.1 nutrition and basic medical care,
1.2 water and sanitation,
1.4 personal safety.
The sub-component 1.2 water and sanitation is based on data on:
1.2.a access to piped water,
1.2.b rural access to improved water source, and
1.2.c access to improved sanitation.
In apparent contrast to the SPI approach, in earlier papers with Michael Woolcock and Matt Andrews, I have argued one definition of national development is a four-fold transformation of nation-states towards (1) a more productive economy, (2) a more responsive polity, (3) a more capable administration, and (4) more equality of treatment of all citizens. With Charles Kenny I argued that high levels of national development were highly correlated with nearly every proposed normative measure of human development and hence drive towards greater national development was typically a powerful source of gains in human development.This new SPI is a fresh chance to assess how strong the connection is between national development and direct measures of human well-being.
Turns out, this new Social Progress Index (SPI) is almost perfectly correlated with national development. If one predicts the Social Progress Index using just three measures: (ln) GDP per capita (productive economy), the POLITY index of autocracy/democracy (responsive polity) and World Governance Indicator of Government Effectiveness (capable administration) the correlation across 128 countries of the actual values of SPI with the predicted values is .961 (an R-Squared of .924). A correlation of .96 in cross-section data is amazingly, astoundingly high. Most people would feel that two measures correlated at .96 are measuring the same thing, not even highly correlated measures of different things. (In fact, often two measures of the same thing—like “years of schooling of the adult population”—don’t have cross-national correlations as high as.96 due to pure measurement error.)
Figure 1 shows the scatter-plot of the actual values of the Social Progress Index and the predicted values (the predicted values are the National Development Index which best predicts SPI).
The boxes are drawn in the graph to emphasize the white space. White space in this graph means something—it is the combination of country experiences in social progress-national development that just don’t happen. The northwest section is white space because no country in the bottom third of national development (the west on NDI) is anywhere near the top third of social progress (north on SPI). Similarly the southeast section is white space because there is no country in the top third of national development (east on NDI) anywhere near the bottom third of social progress (south on SPI). And pretty much all the countries in the middle third by national development are at or near the middle third on social progress.
One can imagine two “pure” development strategies. ‘Due North’ means doing as well as possible on Social Progress while holding national development fixed. ‘Due East’ means going after national development while acquiring only the Social Progress gains expected from the north-east sloping relationship.
It is obvious that for countries with low national development (the bottom third) the Due North strategy of focusing on increasing Social Progress while holding national development fixed has sharply limited prospects. A typical country in the bottom third, like Mozambique, has an actual and predicted SPI of about 46. The regression residuals have a standard error of about 4 so that if Madagascar had SPI performance one standard error better than predicted for its NDI its SPI would be 50. That would move it to ahead of Pakistan and Zimbabwe (at 49) to about the level of Tanzania or Uganda. Alternatively, the 90th percentile of the residuals is 5.75. So if Madagascar were among the 10 percent best SPI performers for its level of NDI it would be at 51.75, reaching Rwanda, but would not achieve the Social Progress of Laos or Bangladesh. Even if Madagascar were the best performer of SPI for its level of national development it would still be in the world’s bottom third on Social Progress.
In contrast, the predicted gains from national development progress are both powerful and unlimited on the upside. So, for instance, if Madagascar improved its GDP per capita, government effectiveness and polity scores by one standard deviation each, its predicted Social Progress would rise to 61.6—ahead of Egypt (60.74) and approaching Morocco (61.9) or China (62.1). And if Madagascar had experienced the growth performance of Korea and had Korea’s GDP per capita that alone would raise the predicted SPI to 68.4—ahead of Thailand or Turkey.
Of course, there is no reason why a country could not have both national development progress and improve its SPI relative to NDI as the same time. For instance, moving ahead in NDI by one standard deviation and having SPI higher for NDI, Madagascar reaches 65.6.
Figure 2: Illustrating various scenarios for Madagascar (a lower third SPI country)
Technical Points: One Big and Two Small
The big point is that correlation is not causation. I am not using statistical procedures (either OLS or other) to resolve whether NDI causes SPI or vice versa. Causality is resolved by theory. The basic theory of the consumer tells us an expansion in consumption possibilities (the budget set) causes higher consumption of all normal goods (budget expansion paths are upward sloping). Higher consumption does not, per se, cause higher income or expand the budget set. Many of the SPI items are consumption items (e.g. piped water) and no one has a causal theory that suggests these are causes (much less the predominant causes) of higher output per worker, GDP or better government. Hence there is a well worked out and empirically validated causal theory of how NDI causes SPI but not (for many/most items in SPI) vice versa. (That said, some SPI items could be argued and we’ll argue that out another day).
Now to the small points. First, the fit is amazingly good given how much measurement error people think there is in measuring concepts like “social progress” or “national development.” Suppose the true relationship was that national development and social progress were perfectly correlated and hence national development predicted social progress exactly, but that each was a concept measured relative to the true concept with a small amount of measurement error—a noise to noise plus signal of 4 percent. Then the observed R-Square would be .924 just from measurement error alone. Second, I did all this empirical work in a few hours and hence have had no time to data mine or functional form mine or goodness of fit mine in any way (for instance, the observation for El Salvador (SLV) on GDPPC just looks way too low to me and hence it looks like a good SPI performer for is level of national development, and if I am right then fixing this would raise the R2; similarly India appears to underperform in part because the WGI measure thinks India has a more effective government than Indonesia, Vietnam or Argentina (which I have my doubts about), and a lower number for India would improve the R2. I just put natural log of PWT8.0 GDPPC (for which the latest data is 2011), the WDI number for Government Effectiveness (and maybe some other “state capability” indicator would give better fit), and the raw POLITY score. With any effort I could crank the R-Squared (a bit) higher but that this is just the “slap in the standard data in the standard way and see what happens” goodness of fit which makes .96 even more amazing.
One of the big debates about efforts like the MDGs or SDGs that drive donor agendas is how much progress can be achieved “programmatically”—by the application of specific projects, programs or policies at existing levels of national development—and how much of progress on human development will require national development (in its multiple dimensions of the four-fold transformation). The very tight association of national development and measures of human development or social progress is an important fact that should inform this debate.
FOR IMMEDIATE RELEASE
Schooling Ain’t Learning – And What to Do About It:
New Book Exposes Education Failures around the World
More information about the book
Book Launch-September 30
Buy the book
Washington, DC – A global push to get all kids enrolled in school has been largely successful—most countries will meet or nearly meet the Millennium Development Goal that each child “complete a full course of primary schooling” by 2015. But a new book by Lant Pritchett from the Center for Global Development documents a deeply disturbing reality: for millions of children in the developing world schooling is not producing “education” in any real sense.
Consider these examples from the book, The Rebirth of Education: Schooling Ain’t Learning:
· In India less than half of children surveyed in grade 5 could read a grade 2 level story, one in four could not read a simple sentence, and only slightly more than half could do subtraction.
· In Tanzania six-out-of-ten students who took the 2012 examination for secondary school completers failed.
· In Pakistan a child who enters fifth grade not knowing how to do simple division has only a one-in-six chance of learning in an entire year of schooling.
· Pritchett not only sounds the alarm. He goes on to diagnose these failures and propose a potentially transformative new approach to education.
The result of years of research, including time Pritchett spent studying schools in India, the book offers shocking new analysis and data about the current state of education in developing countries and a trenchant critique of the global focus on enrollment rather than learning.
“We often see education as one of the most powerful tools for escaping poverty,” says Nancy Birdsall, president of CGD. “In this important book, Lant Pritchett reveals the great divide between schooling and learning and reminds readers that our goal is not to get students into classrooms but rather to prepare young people to become productive members of the community.”
Pritchett begins his book with a story from India. In 2006, MIT researchers conducted a rigorous evaluation of schools in Uttar Pradesh, one of India’s poorest states, only to discover that many fifth-graders could not read a simple story, do basic division, or even recognize letters of the alphabet. At a village meeting Pritchett listened as the father of a boy who was unable to read rose to address the school principal:
“You have betrayed us. I have worked like a brute my whole life because, without school, I had no skills other than those of a donkey. But you told us that if I sent my son to school, his life would be different than mine. For five years I have kept him from the fields and sent him to your school. Only now I find out that he is thirteen years old and doesn’t know anything. His life won’t be different. He will labor like a brute, just like me.”
The principal responds: “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 stupid and you cannot expect that they will be home from school anything other than stupid.”
India is not alone. Even in middle-income countries with high average years of schooling, between one-third and two-thirds of 15-year-old students do not meet even the most basic math, reading, and science learning goals. In Brazil, Turkey, Indonesia, and Mexico, over 50 percent of 15-year-old students with over five years of schooling do not meet math goals. When compared to their counterparts in rich countries, the educational divide is even clearer: 15-year-old students from Thailand, Mexico, Mauritius, and Chile fall below the 20th percentile of students in Denmark. Students from Qatar, Ghana, Saudi Arabia, and El Salvador fall below the 5th percentile when compared to their counterparts in Australia.
“This problem couldn’t be more important,” says Pritchett. “A child who finishes school at age 15 this year and plans to work until age 65 will be in the labor force until the year 2063. These children are emerging from primary schooling or even secondary schooling with so few skills that they are unprepared for today’s economy, much less for the economy of 2030 or 2063. Their lack of basic education is a burden they will bear for decades.”
Part of the problem is the emphasis on inputs instead of outputs. Desks and chairs, pencils and textbooks, students and teachers may look like a school but they don’t always add up to learning. Pritchett borrows a term from biology, “isomorphic mimicry,” to describe this phenomenon of looking like something else without acquiring the core functions, as when a non-poisonous butterfly evolves to look similar to a poisonous species to avoid being eaten by birds.
In describing education systems, Pritchett borrows from The Starfish and the Spider: The Unstoppable Power of Leaderless Organizations, by Ori Brafman and Rod Beckstrom. In many countries, Pritchett says, the schooling system is controlled by a large government-owned, top-down bureaucracy – a “spider.” This spider dictates everything: which schools get built, which programs get funded, which teacher gets assigned to which school. Despite the bureaucracy’s extensive reach – the spider’s web – all decisions are made in one centralized location – the spider’s brain.
The Uttar Pradesh school principal’s rebuttal reflects one of the problems of the spider school system, Pritchett observes. In a top-down system, all the power rests with administrators. But educating children requires a system far more complex and flexible than a top-down bureaucracy, what Pritchett calls a “starfish system.”
A starfish, unlike a spider, is a radically decentralized organism – some species of starfish have no brain at all and a starfish’s parts are loosely connected and controlled by local actions. Pritchett praises starfish systems for being locally operated, performance-driven, and open.
The problem is that form follows function. Instead of focusing on what education should look like Pritchett urges a renewed focus on what schools are meant to do. This can only be achieved by encouraging school systems to measure learning outcomes – and allowing local schools the freedom to create schools that best meet learning goals.
He makes clear that there is no single solution that will solve the problem for all schools. Rather the “pivot to learning” he advocates will require school systems that are more like starfish and less like spiders: open, locally operated, performance-pressured, professional networked, and technically and financially supported.
The Rebirth of Education has received wide praise:
“With abundant data, experience, and clear thinking, Pritchett makes a compelling case for why more of the same won’t cut it anymore, how we need to think deeply about how change happens and who can drive it, and why we need to be suspicious of experts and blueprints,” says Rakesh Rajani, founder and head of Twaweza, a Tanzanian NGO.
"Lant Pritchett's path-breaking and courageous work exposes the scandal of education policy in development, which contents itself with achieving quantitative targets on student enrollment even when no real education is happening,” said William Easterly, Professor of Economics at New York University. “Nobody reading this book will ever think about education the same way again."
“Lant Pritchett’s recommendations will disappoint both orthodox economists and orthodox educators since they do not reinforce any of the standard recipes. But those willing to be convinced by Pritchett’s logic and the particular blend of caring and impishness that characterizes his writing will be justifiably alarmed, then enlightened, and finally filled with hope,” says Luis Crouch, chief technical officer, International Development Group. “I urge all my colleagues to read it immediately.”
The Rebirth of Education: Schooling Ain’t Learning will be released on Monday, September 30 at an event hosted by the Center for Global Development. More information about the book and the launch event can be found on the CGD website.
Notes for Editors:
Members of the media interested in attending the launch event should contact media relations associate Catherine An by calling 202-416-4040 or emailing email@example.com. The book launch is slated for Monday, September 30 at 4 pm at the Center for Global Development.
About the Book:
The Rebirth of Education: Schooling Ain’t Learning may be ordered through Brookings Institution Press. ISBN: 978-1-933286-77-8.
Some economists, with their recent fad for "evaluation", have managed to get themselves deeply confused about what the "conditional" in "conditional cash transfer" (CCT) is really about. They often interpret the "effectiveness" of CCTs relative to the action/behavior/outcome that was conditioned on—for example, the impact of schooling-conditioned transfers on enrollment rates.
But the key question is whether the value of the conditionality in CCT lies in the political symbolism provided by the condition that supports the transfer, or whether the condition has some additional benefits (or what mix of the two).
A policy entrepreneur who wants well-targeted cash transfers will encounter two obstacles. One is meddling politicians who want to manipulate cash transfers, with motives based in clientelism, patrimonialism, or partisan electoral politics. A second is "right wing" political forces who believe that transfers to "the poor" are morally objectionable as they create perverse incentives by reducing the need for work, financing bad habits (drink, tobacco) or otherwise go to the “undeserving” poor.
The "conditionality" of cash transfers is a brilliant symbolic political ploy to neutralize both meddling politicians and right-wing objections. The recipients of CCTs are "deserving" poor because they are (a) chosen by "scientific" criteria like a proxy means test (which therefore cannot be meddled with politically) and (b) the cash transfer isn't just a "handout" but rather the poor have to do something socially valuable to be deserving poor, they have to "invest in human capital."
The role of conditionality in making a well-targeted cash transfer politically feasible might be completely independent of whether the conditionality itself improves recipient household welfare. In fact, conditionality could be valuable from a social welfare function view even if the conditional transfer proved to be less welfare enhancing than a non-conditional transfer. This would be the case whenever making transfers conditional raises the amount and improves the targeting of the politically feasible transfer sufficiently above the largest politically feasible unconditional transfer. One conditional dollar that is politically feasible to deliver might do more social good than an infinite number of unconditional dollars that are politically infeasible to deliver.
In the logic of politically motivated conditions, a policy entrepreneur who wants to raise the well-being of the poor would seek to maximize the symbolic value of the conditioned behavior while minimizing the welfare loss to households of actually engaging in that behavior. The idea would be to pick a behavior with potent symbolic value that nearly all households in the targeted category are doing anyway. This means the welfare loss from the condition is low, since it is already the unconditioned maximizing choice of most households. Notice that in this political logic if an activity with potent symbolism could be found, the optimal "conditioning" would be infra-marginal for nearly all transfer recipients—that is, by design the conditionality would not change most recipients’ behavior.
Of course, another way to think about CCTs is that imposing conditionality has big welfare effects over and above the transfer itself. This is an odd way for economists to think, as the value of the conditionality then depends on changing people’s behavior with a binding condition. This is odd position for economists who are usually an intellectual bastion in favor of choice and against the symbolic appeal of paternalism to those who condescend to the poor. A binding condition only increases a non-paternalistic social welfare function if households are (for some reason) not welfare maximizing already or if there are external effects. Of course many sophisticated arguments could be made about spillovers or information failures or externalities but there has never been any evidence for the magnitude of these effects. Moreover, the condition is often to attend schooling that is already massively subsidized and provided (near) free of charge so that the argument would have to be the externalities/market failures are not sufficiently offset by the existing subsidy—an empirical standard that has never been documented anywhere.
Notice that these two logics create almost directly opposing views on the ex ante design and the ex post evaluation of impact. The political logic suggests creating conditions that are easy to meet but symbolically important so that the “impact” should be small and the “impact” of the conditionality could be interpreted as the welfare loss needed to politically sustain the program. The “economic” logic suggests creating conditions that require behavior change and are not infra-marginal and the “impact” of the conditionality is a measure of the welfare gains from forcing poor people to change their behavior.
The genius of the CCTs in Mexico and in Brazil was not about how to get kids in school but rather about how to use the fact that almost all kids already were in school to generate welfare loss minimizing but nevertheless politically powerful symbolism to expand cash transfers. Many of these transfers were conditioned on the enrollment of children in age groups with near-universal enrollment. (In Brazil in 2001, for example, enrollment at ages 9–12 exceeded 95%.) As a strategy for getting kids into school, as a recent J-PAL note attests, this makes CCTs highly cost-ineffective. Does this make the transfers a silly giveaway or a political master stroke—maximizing the political symbolic value of a cash transfer while minimizing the burden on the recipients of conditions? Ask yourself: What did the designers think?
Miguel Székely, one of the designers of PROGRESA, writes, “The real underlying objective of external donors/investors in some circumstances might not be generating impact, but rather making the statement that they are supporting a particular cause. In such cases, the objective might well be a noble and legitimate one, and the measure of success will be the flow or resources itself, rather than its final impact, but neither the donor/investor nor the executor might have incentives to invest in evaluation.”
In other words, one common narrative—that the scaling up of CCTs is a good example of evidence based policy making because the use of randomization in the design of PROGRESA provided solid evidence that it was an effective program and hence other countries adopted a CCT because of this solid evidence—has it almost exactly backwards. The impact evaluation proved that PROGRESA was cost ineffective if it was considered as a mechanism to increase schooling. Everyone involved in the design knew this. They were not imposing the conditionality to get the behavior conditioned upon, but to get the transfer itself.
What PROGRESA proved that was convincing was about the political effectiveness of conditions to the implementation and supportof cash transfers. Adding conditions to cash transfers allowed a well-designed cash transfer to have political traction against opponents. But this was learned from the experience itself—which is why the Brazilian experience added to the persuasive power of the evidence that adding the conditionality to cash transfers is good politics even though it had no experiment attached.
Ironically, what was really learned from the experience of PROGRESA is that having a rigorous experiment attached to your program can be great politics, as it makes the program seem technocratic and scientific and cool and adds to the symbolic value by demonstrating compliance in attendance (even where that was not an increase versus a counter-factual). This increases your ability to resist partisan political meddling in design and implementation—even if you don’t learn anything particularly special from the experiment.
The next few posts on education are a bit unusual, in a good way I hope, but unusual entrants into the blogosphere. As part of the CGD initiative on education in the developing world and the pivot from schooling to learning, we are going to post links to and discussions of some of the new empirical evidence that is emerging. However, the new evidence on learning trajectories--the gains in skills/capabilities/knowledge as students progress through grades--both requires some common background and, to my view, challenges some of the fundamental assumptions about the schooling experience. So this first post describes a measure--the learning gain per year of schooling compared to a standard deviation of student mastery across students in a given year of schooling--that is intimately related to both the achievement of learning goals and to the fundamental design of schools. This puts the post somewhere between a paper reporting new results and describing results appearing in other papers.
As Mark Twain said: “It ain’t what you don’t know that gets you in trouble, its what you do know that just ain’t so.”
The entire structure of elementary schooling—from the physical layout of classrooms to the curriculum to the way teachers are assigned to a child’s school daily routine—is based on what a child having completed “Third Grade” (or “standard 3” or “level 3” ) means. What a third grade child does in school is learn the third grade curriculum and what a third grade teacher does is teach the third grade curriculum. The curriculum of schools is designed so that a child learns mastery of subject area concepts and skills and their applications in sequence so that third grade builds on second grade and prepares for fourth grade. This is premised on two perfectly reasonable ideas.
First, that concepts and skills are cumulative and are best mastered in a certain order. An excessively simple example: a child needs to recognize letters before they can read words and need to recognize and read words in order to understand sequences of words as sentences and be able to read sentences fluently (that is, at a minimum speed) in order to understand meaning, and so on.
Second, that group based instruction is effective (and efficient) when instructional groups have reasonable homogenous skill sets.
Curricular graded classroom instruction—the foundation of most schools—is therefore premised on knowing as an empirical fact that a child being in “third grade” or “fifth grade” conveys substantial information about their actual knowledge and skills. Tragically, increasing bodies of research suggest that in many developing countries what we know about “grade” just ain’t so.
The ratio of the gain in skill across grades to the dispersion of skill within a grade is central to what we mean by grade—and yet is often unmeasured. I show that when this ratio of gain to dispersion is low that “grade” means little or nothing in two empirically precise ways: (1) “grade completed” has little predictive power for measured skills/knowledge and (2) the overlap of skills/knowledge across grades is near complete so that “grade” does not represent an homogenous group for instruction.
Let me start with hypothetical examples using simulations to illustrate the ratio and its empirical consequences and then show that data from South Asia and Africa are consistent with very small ratios of grade gain to student dispersion.
So, suppose we have an agreed upon valid and reliable measure of student mastery of any curricular domain (e.g. reading, mathematics, science, history). We will assume that this measure of domain mastery has a Gaussian normal distribution and normalize the measure so that the standard deviation in that measure of curricular mastery across students in a given grade is 100. These assumptions imply that if the average score of third graders is 300 then the bulk of these students (68 percent) will have scores between 200 and 400 (and 16 percent below 200 and 16 percent above 400). Let us also assume that this measure has an average of 200 when a child is in first grade.
Now, suppose we measured on this same scale the mastery of children in grades 1 to 6. We could then plot the distribution of these scores grade by grade. What that would look like depends on the magnitude of the gain in mastery—what could be called learning—from year to year on this scale—hence relative to the (constant) student standard deviation. Figures 1a and Figure 1b show two possibilities. In Figure 1a the learning gain is 80 points per year and hence the ratio of gain in skills to dispersion of skills is .8 of a student standard deviation per year. In this scenario “being in third grade” means something—most third grades have greater mastery than the typical first grader but less than the typical fifth grader.
Figure 1a: Illustrative distributions of student capability across grades—learning gains large (=.8) relative to within grade student differences
Figure 1b is exactly like 1a (on the same scale) with the only difference being that the learning gain per grade is only 20 points per year—or .2 of a standard deviation. Now “being in third grade” is not very informative at all. A child in third grade that is at the lower end of typical performance (a student standard deviation lower, or the 16th percentile) has substantially less mastery that the first grade average (140 vs 200). Conversely, a child in third grade at the upper end of typical performance (a student standard deviation higher, or the 84th percentile) would have mastery much higher than the fifth grade average (340 vs 280).
Figure 1b: Illustrative distributions of student capability across grades—learning gains small (=.2) relative to within grade student differences
This can set up two empirical measures of what “grade” might mean and precisely how grade might mean little or nothing.
The first is the size of the predictive power of grade for skill mastery. Suppose I want you to guess whether or not a child has some skill—like reading a story of a given difficulty or doing three digit subtraction. Further, before you guess I will answer one question about the child—boy or girl, age, height, color of eyes, name, or, perhaps, grade completed. What is the best question to ask and how much does it improve your guess? Most people’s gut instinct, especially if they know the grade at which a given skill is taught in school, the grade completed is the best piece of information and knowing a child’s grade completed improves one’s guess a lot. That intuition is correct is the ratio of learning gain per grade to student dispersion within grade is large. But, if the learning gain is small compared to the students’ skill dispersion in a grade, then knowing a child’s grade actually helps little in predicting what they know.
The usual statistic to understand the predictive content of a piece of information it is (incremental) R-Squared: how much lower is my predictive error when I based my guess on X? R-squared varies from zero, where X means nothing, to one, where X reveals all. In our simple simulation above we can compute for each ratio of grade learning gain to student dispersion the R-Squared of knowing a child’s grade in predicting their concept/skill mastery. Figure 2 shows that is the gain per grade is small then grade means little or nothing for predicting learning. If the gain per year is only .2, then knowing a child’s grade improves one’s guess by less than 10 percent. If one “knows” that “grade matters” and based on that knowledge designs a school on the basis that “third grade” is informative for what a child in third grade can do and is ready to learn, then when learning progress is weaker than assumed the whole structure can fail.
Figure 2: If learning gain per grade relative to dispersion is low then knowing a child’s grade has little predictive ability for what they skills/concepts/capabilities they have (and hence are ready to learn)
The second measure of what grade means is the extent to which “grades” represent homogenous groups for instruction. Let me define by the “core” of a grade the level where the bulk of the students are. In this simulation let us for now define the “core” as students within a standard deviation of the average. We can ask: “how much does the core of one grade overlap with the core of another?” If progress from grade to grade is small, then the overlap is large, and if progress from grade to grade is large, then most students in a higher grade will have moved beyond those of a lower grade.
Figure 3 illustrates the notion of “core overlap” using grades 3 and 4 of the simulation with a learning gain per grade to dispersion of .2 (gain of 20, student standard deviation 100). In this case the “core” of grade 3 is 140 to 340 and the “core” of grade 4 is 160 to 260. Hence the overlap of the “core” in these adjacent grades is 160 (those at the lower end of grade 4 core) to 240 (the upper end of the grade 3 core) which is a range of 180 out of a core of 200. So 92 percent of the students in grade 4’s core were also in grade 3’s core.
Figure 3: The “instructional core” of grades has extensive overlap with previous grades if the progress per year is low
The essence of a sequenced grade curriculum is that each grade builds on what was learned in the previous grade. So by design some amount of instruction is review and deepening of mastery of previously acquired skills and some amount of instruction is introducing new material. But when learning progress is slow “third grade” ceases to mean anything as nearly all of what is done in “third grade” is known by both those in second grade and those in fourth (and fifth) grade.
Figure 4: Connection between overlap of adjacent grades “instructional core” and the ratio of learning gain to dispersion
In this first post, all I have shown are the arithmetically inevitable consequences of the relationship between the pace of progress in learning and the within grade dispersion across students. Nearly every aspect of the modern school is built on the presumption that the learning gain per year is large enough to make group grade based instruction on a sequenced curriculum an effective (and perhaps efficient) approach to teaching and learning. But if the learning gain is too low then everything that is “known” about the “right” way to do schooling may well be wrong. In my next blog, I’ll show that many strands of empirical evidence emerging from South Asia and Africa suggest that what we know just ain’t so.
The “just right” approach for the mobility of low-skill labor looks to avoid either “too hard”—expecting countries to make legally binding commitments to a global protocol—or “too soft”—no global mechanisms for reducing restrictions on labor mobility. We propose a “bundled” organization that works with existing bilateral labor agreements and partners as part of an organization capable of analysis and advocacy.
The Obama Administration, whether by design or by accident, has opened the door for the first time in the World Bank’s history to the possibility of a real contest over the merits of its nominee to take the helm there compared to a nominee from the developing world. All three candidates have experience working on development (and that is a refreshing change from the tradition of financiers and political heavyweights at the helm). But their strengths are different. In the case of Kim, the U.S. nominee, and Ngozi Okonjo-Iweala, their training, their experience, their instincts, and their worldviews are completely different.
Jim Kim is an MD -anthropologist and most recently a university president who has devoted most of his career to finding better ways to deliver better health care to the world’s poor. He is steeped in the tradition of medical science: deeply committed to measuring and evaluating what works on the basis of evidence before intervening—an approach to learning that has recently been taken up enthusiastically among micro-economists inside and outside the World Bank. He believes in the power of popular movements to overcome financial and behavioral obstacles to better health regardless of poverty or hopeless politics in poor countries – as illustrated in the case of the work he led at WHO to extend treatment of HIV/AIDs to millions of people.
Ngozi Okonjo-Iweala is a Harvard and MIT-trained economist, former finance and foreign affairs minister and current economic czarina of Nigeria, and longtime staff member of the World Bank, where she rose to become for all practical purposes the deputy to Robert Zoellick before returning to her own country. She has devoted her career to changing economy-wide (“macro”) policies and practices around the world -- in countries like her own and in the rich world as well – in support of the kind of broad-based growth in poor and middle-income countries that empowers the poor to escape poverty. She has practiced good economics in the hard soil of tough politics in Nigeria, fighting high-level corruption at personal risk, recently working to eliminate gasoline subsidies that benefit the car-owning rich while sapping the public budget of resources to serve the poor. (Full disclosure: Ngozi is also a member of CGD’s board of directors.)
For the first time, these two and the third candidate (José Antonio Ocampo of Colombia now at Columbia University, nominated by Brazil* -- more on that in another post!), can be asked to publicly set out his or her vision for its next decade. Also for the first time, members of the bank’s board will interview the candidates. Government officials and development advocates will have a similar opportunity, assuming each of the candidates travels to the world’s major capitals to make their respective cases, as Agustin Carstens and Christine Legarde did recently in their competition to head the IMF.
Here are three questions worth asking each of the candidates to be the next president of the world’s premier development institution.
1.Should the World Bank bring its institutional and financial clout and expertise to the problems of the global commons (and a broader set of “global public goods” (GPGs) in economics lingo) that are especially costly to the livelihoods and welfare of the world’s poor: climate change; drug resistance; the missing Green Revolution in Africa; Would you as president seek a new mandate and new capital to deal explicitly with the development challenge a deteriorating global commons poses, as some have advised, asking the Bank to become a catalyst for global public goods. Would you as president seek capital from China, Brazil and other developing countries with ample currency reserves to support a new GPG arm of the Bank based in Mumbai, Sao Paulo or Shanghai?
The Bank is engaged in cross-border work – including some infrastructure and watershed programs that involve more than one country, and it manages several global programs including Climate Investment Funds and a small carbon-trading platform. But its global engagement is ad hoc, at the behest of one or another member country that sets up a special fund to support “global” programs. It lacks a clear mandate and instrument to do more. History, habit and politics have tethered its operational machinery to its bread and butter country-specific loans and grants for more than 50 years after all. Without leadership from the top, it may well miss the boat on a pressing set of 21st century challenges to the elimination of global poverty.
Kim has thought about challenges that transcend country borders in his work on drug-resistant tuberculosis. Okonjo-Iweala has thought about the issue in the context of "green venture funds" with public backing for “deal-packaging” and first-loss guarantees that would catalyze institutional and sovereign wealth fund investments in clean energy for the poor. She inspired and led the Bank’s work to recover stolen assets that crooked heads of state – often from oil and mineral-rich nations like her own -- have with impunity deposited in western banks..
2.What is your vision for the future of IDA, the bank’s fund for grants and concessional loans to the poorest countries? Should it be bigger in 10 years or smaller?
Under current rules low-income countries such as Pakistan, India, Nigeria and Vietnam are likely to have graduated from IDA by 2025. Remaining countries eligible for IDA will be mostly flailing and failed states suffering a toxic combination of internal conflict and corrupt or ineffectual leadership. Should the freed-up IDA resources be committed to countries like Somalia and Afghanistan, in the interests of global security, or go mostly to countries like Honduras, Liberia, and Rwanda, where they are more likely to improve lives in the short run and more likely to help end dependence on aid? Should some IDA resources be deployed to deal with global commons problems, where the traditional Bank country loan is not useful? Or should IDA funds be used to subsidize loans to countries like Brazil, Egypt and India which though “middle-income” have huge numbers of poor people?
Kim believes services can be delivered to people effectively in tough environments. He is not naïve about the dependence of service delivery on appropriate policy and politics but his starting point is that there is a way to get it done if the right management and execution lessons can be found, adapted and brought to bear. Okonjo-Iweala is also in her way a radical optimist on improving the lives of the poor. But she is probably equally focused on getting the macroeconomics and politics right to raise incomes of the poor. She led the difficult campaign to replenish IDA funds just over a year ago at the World Bank (the replenishment amounted to $30 billion or $45 billion depending on accounting definitions). In the short run at least she would be probably more aware of the tradeoffs over time and place and challenges that are implicit in the question posed above.
3. What is your vision for the role of the Bank in the dynamic emerging economies, and for their role in the governance and management of the Bank? Will you encourage the United States to support changes in governancesuch as a reduction in its veto power, and encourage the Europeans to give up their unreasonably large number of board chairs – in the interest of engaging China and other developing countries more actively in the development challenge for the next decade? Will you support the initiative of the BRICs to make the IFC, the private sector arm of the World Bank Group an independent entity with own president? Do you think the Bank needs more capital even if it comes from China and other countries, reducing the influence of the U.S and Europe, steady and benign supporters for over 50 years? Would you consider a change in the selection process for your successor, such as the introduction of double majority voting (under which your successor would need both a majority of weighted votes and a majority of country votes)?
The underlying shift in global influence and power toward the dynamic emerging economies (on which global growth and prosperity in rich countries now heavily depend) puts a premium on the World Bank president’s negotiating and consensus-building skills. Consider one source of tension: At the Inter-American Development Bank, Brazil and other developing country members have had the votes to avoid transferring net income from its regular income-earning loan facilities to its concessional facility. But with control of the World Bank in the hands of the traditional trans-Atlantic powers since its founding, that is a practice that has been routine at the World Bank. The most recent IDA replenishment included billions of dollars to come from the IFC’s and the IDBRD’s net income. Brazil and other developing countries reasonably argue that the transfers have been a non-transparent way for the rich economies to minimize their direct contributions to IDA, while indirectly raising the costs of borrowing from the regular facility. The option in the future is for China and Brazil to contribute directly to IDA. . . .and then of course to take their place at the table where IDA policies and programs are set.
This last set of questions goes to the heart of the issue of the role and standing of the World Bank in a shifting geopolitical landscape. Kim, as a Korean-American with experience working in Peru and Haiti and at WHO, has presumably thought about geopoliltics. Did he demonstrate at Dartmouth convening and consensus-building skills a World Bank president needs? Okonjo-Iweala grappled with geopolitics when, as Nigerian Finance Minister, she negotiated that country’s $30 billion debt relief package in Paris, Washington and New York. And she did so at the World Bank, when she held the high-level management position of Secretary of the Board. Does she have those consensus-building skills?
I know all three candidates enough to be confident they have qualities that could make them effective as president of the world's preeminent development institution. They all have shown in one realm or another that they are effective leaders of men and women. For me the key question has to do with their vision for the World Bank of the future. Will the next president focus on business as usual, but better? Or will she or he lead the bank in new directions for a new era?
*Correction: The Brazil chair nominated José Antonio Ocampo on behalf of the Dominican Republic; the DR is one of the countries Brazil represents at the Board of the World Bank
The last decade has seen considerable progress enrolling children in schools worldwide: today most people live in countries on track to meet the Millennium Development Goal of 100% primary completion by 2015. Sadly, enrollment doesn’t necessarily equal learning. A new report by the CGD Study Group on Measuring Learning Outcomes shows a shockingly wide gap between education inputs and learning outcomes. The report, Schooling is Not Education: Using Assessment to Change the Politics of Non-Learning, finds the learning crisis reflects systemic issues in education sectors worldwide.