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.
Technology, infrastructure, governance and anticorruption, human development, subjective wellbeing/happiness
Charles Kenny is a senior fellow and the director of technology and development at the Center for Global Development. His current work focuses on gender and development, the role of technology in development, governance and anticorruption and the post-2015 development agenda. He has published articles, chapters and books on issues including what we know about the causes of economic growth, the link between economic growth and broader development, the causes of improvements in global health, the link between economic growth and happiness, the end of the Malthusian trap, the role of communications technologies in development, the ‘digital divide,’ corruption, and progress towards the Millennium Development Goals. He is the author of the book "Getting Better: Why Global Development is Succeeding, and How We Can Improve the World Even More" and “The Upside of Down: Why the Rise of the Rest is Great for the West.” He has been a contributing editor at Foreign Policy magazine and a regular contributor to Business Week magazine. Kenny was previously at the World Bank, where his assignments included working with the VP for the Middle East and North Africa Region, coordinating work on governance and anticorruption in infrastructure and natural resources, and managing a number of investment and technical assistance projects covering telecommunications and the Internet.
In Navigation by Judgment, Dan Honig argues that high-quality implementation of foreign aid programs often requires contextual information that cannot be seen by those in distant headquarters. Tight controls and a focus on reaching pre-set measurable targets often prevent front-line workers from using skill, local knowledge, and creativity to solve problems in ways that maximize the impact of foreign aid.
Over 1 billion women lack access to financial services due to economic and social barriers, time and mobility constraints, and discrimination in service provision. Financial services delivered digitally can address these barriers by providing women with safe and accessible channels. This event will look at the recent evidence and emerging technologies that work to empower women economically.
Pascale Hélène Dubois will discuss the global impact of World Bank investigation and prevention activities and then join a panel with Kathrin Frauscher, Deputy and Program Director, Open Contracting Partnership and Hasan Tuluy, Partnership for Transparency Board Director, former World Bank Vice President, to dive deeper into what more can be done at the World Bank and other international institutions to combat corruption.
Technological advances in fields such as artificial intelligence and automation have the potential to fundamentally alter prevailing economic trends. While the effects of these changes are the subject of great debate in the developed world, less discussed has been how they will impact the developing world. Speakers will explore what emerging technologies mean for both the traditional models of development and the future of job creation in developing countries.
Christal Morehouse, Alla Volkova, and Silvia Fierăscu of The Open Society Foundation (OSF) have just issued a fascinating report on the gender breakdown of speakers at 23 high-level conferences across Europe—including nine forums, six conferences, four meetings, three summits, and a games and an ideas lab (which adds up to 24 because the World Economic Forum Annual Meeting is a Meeting at a Forum—doubtless why it costs $71,000 to attend).
The OSF team tracks 12,600 speaking roles between 2012 and mid-2017 and finds that across conferences with data for 2012 and 2016 there’s a welcome trend: 14 of the 15 conferences had a larger proportion of women speaking in the later year. But across the sample they still find that overall men occupied 74 percent of speaking roles and women only 26 percent. Women were better represented amongst keynotes and moderators than panelists—where only 24 percent were women.
Some gender equality leaders stand out: Chatham House had 44 percent women speakers at its 2016 London Conference, for example. Chatham House’s Laura Dunkley reports that the institution:
recently launched an internal gender awareness action plan. Covering three strands of Chatham House’s activities, the work will aim to build a toolkit for think tanks operating within international affairs…Specifically on convening and debate, Chatham House is committed to no all-male speaker panels at events and encouraging staff to uphold this commitment when speaking externally. Chatham House places equal importance on creating a greater mix of gender representation in participants, as well as ensuring events and workshops are designed to facilitate and encourage inputs and contributions from all participants.
We look forward to seeing—and learning from—this toolkit. And the OSF report is well worth reading in full both for the data and ideas on how to improve. But OSF doesn’t report on the proportion of European conference panels that are all-male—Euromanels, if you will. So we used the OSF’s dataset to calculate that.
Gender Breakdown of European Conference Panels
In short: the Euromanel is alive and well. Of the 2,282 panels in the database we looked at, 903 (or 40 percent) were all male. At the other end of the scale, 56 (2.5 percent) were all women. In the middle 280 (12.3 percent) were 50/50 gender-balanced (only possible on panels with even numbers of participants, of course) and 154 (6.8 percent) had more women than men but did have a man. Some comforting data: over time (between 2012 and 2017) the proportion of Euromanels has gone from 43 percent to 33 percent (though note that is over a changing set of conferences).
Thanks to OSF for quantifying and highlighting the conference speaker gender gap in Europe. Doubtless the situation is little different in the United States, and we’d love to see data on that.
Data notes: We used the OSF data on panel membership alone (excluding keynotes, moderators, and others), dropping panels that appeared to have only one panelist and a few observations that were potentially duplicate or where the panel name was too generic to ensure it was a separate event from other listed panels. You can download our code here.
Last year on International Women’s Day, we talked about labor and financial equality as a prerequisite for women’s empowerment. This year, we’re being a little more introspective. Many nonprofits working in global development advocate for women’s empowerment and gender equality (this one included). But do they follow through when it comes to their own staff and management? Tax reporting requirements in the US mean that we can at least partially answer that question as it relates to remuneration, and the preliminary answer is “not yet.”
To compile a database of organizations and employees in international development we started with three websites: the Inside Philanthropy list of global development funders, the 2016 Global Go To Think Tank Index Report list of “Top International Development Think Tanks” (p. 90), and the list of members at Interaction, an alliance organization of international development NGOs. We then randomized the order of the names in each of the three lists and searched in list order for the organization’s IRS Form 990.
We extracted data on the organization’s total revenue, the fiscal year of the 990, and the gender and pay of the listed employees who worked at least 30 hours a week (which excluded most board and trustee members). Not every organization was eligible for analysis: we excluded organizations with no available tax form and/or fewer than five salaried and full time (at least 30 hours) employees. For the think tanks list, we excluded non-US organizations which would not have 990s available. We continued down the list for each group until we had 10 organizations of each type to make up our sample to analyze. All but two of the Form 990s we examined were from 2015, the exceptions were from 2016.
The resulting database covers 110 NGO employees, 151 think tank employees, and 64 foundation employees (an average of 10.8 per institution). We report below what proportion of those key/high paid employees were women in each group, the average pay of key/high paid women expressed as a percentage of the average pay of key/high paid men, the percentage of organizations for which a woman is listed as the highest paid employee, the percentage of organizations where more or equal numbers of key/high paid women were listed as men and the percentage of organizations where the average pay of key/high paid women was the same or more as the average for men in the organization.
Some caveats regarding data and interpretation of results:
We are limited by the data that nonprofits self-report on their forms: in that regard, our database includes more than just the top five highest paid employees in think tanks and NGOs, it also includes other “key” employees that worked 30 hours a week or more. For the foundation lists, it’s important to note that they are only required to list the top five highest paid employees which creates a smaller sample pool.
Our sample size is small: out of 202 foundations, 130 think tanks, and 182 NGOs we only look at 10 for each group. This means the margin of error of our estimates will be large, particularly on questions at the institutional level (on the question “what percentage of best paid employees are women,” for example, we can only be 90 percent confident that the mean of the full group of foundations listed by Inside Philanthropy is within about 25 percentage points of the mean we report from our sample).
The pool of organizations includes nonprofits that only work on development issues as one part of a much broader portfolio of activities. That’s particularly true of the think tanks and foundations.
The purpose of the exercise is not to shame specific nonprofits, but rather to (imperfectly) paint a sector-wide picture. To this end, we’ll break from standard CGD practice and not release the underlying dataset (however, please note: this information is all in the public domain). But for the sake of transparency, we also report CGD’s 2015 numbers on a separate line (it was not part of the random sample of think tanks).
In short, the data we’ve collected suggests that nonprofits involved in international development still have some way to go in terms of gender equality. Here are our key findings:
Less than a third of people in our (small) sample of key/high paid employees in US think tanks are women, although in our sample of foundations there was parity between the number of men and women.
Key/high paid women appear to be paid less than key/high-paid men in all three of our sample groups, and very few of the highest-paid employees in organizations connected to global development are women—a total of 5 out of 30 across the three groups.
Looking across organizations, it is impressive that 6 out of 10 of our foundations list more than or equal numbers of high-paid or key women as they do men and 4 out of 10 foundations pay more or the same on average to listed women as to men. But only 2 out of 10 think tanks and the same proportion of NGOs pay more or the same on average to listed women as to men.
Percentage of Women amongst High Paid EmployeesPercentage of Women amongst Orgs' Single Highest Paid EmployeesAvg Women's Pay (as a % of Avg Men's Pay amongst High Paid)Percentage of Orgs with >49% Women in High Paid GroupPercentage of Orgs with Avg Women's Pay >= Men's (amongst High Paid)
Back to caveats: these numbers can only be taken as suggestive given the small size of our sample. We didn’t do more data collection for this blog post because attempts to automate the process failed—though many thanks to CGD’s Mike Brown for trying—and the manual exercise is reasonably labor intensive. We’d happily be a part of a (funded) exercise that scaled up the analysis, looked at trends over time, and examined factors that lie behind variation in women’s representation at top pay levels on the nonprofit space. There is more information to analyze including expanding our gender analysis to leadership positions and board membership. It’s also crucial to look at this through the lens of racial diversity: while women make up 48 percent of board members in nonprofits, only 20 percent of board members are people of color (compared to a 38 percent population share in the US as a whole).
And it would be great to be able to look at trends over time. That can add important details to the story: a recent study of the World Bank’s staff found that the average woman earns 77 percent as much as the average man, with the big reason for that gap being that women tended to have entered at lower employment grade levels in the institution (women and men who started at the same time and grade in the Bank have fairly similar wages years later).
But if these numbers broadly hold up, they suggest US institutions involved in international development aren’t sufficiently practicing what they preach when it comes to the importance of diversity and equality to outcomes—and they are less effective as a result. More (and more fairly remunerated) women in nonprofit leadership will change what the sector does for the better.
We welcome comments about methodology or concerns with our data source—or anything else!
CORRECTIONS AND UPDATE: In checking our methodology (prompted by Laurence Chandy, to whom thanks), we realize that we underestimated the impact of the new PPP lines on poverty due to two mistakes in the computer code underlying the original version of this post: we used 2010 CPI figures where we meant to use 2011, and conversely, we used 2011 population figures where we meant to use 2010. An explanation of the error and the changed numbers are presented below, along with access to the original estimates and all code. We’ve used strikethrough and square brackets in the text to show where language and numbers have changed, the pictures refer to the corrected data. We’ve also taken the opportunity to add some final thoughts on the discussion in an update below.
On Tuesday night, the International Comparison Project released the latest purchasing power parity numbers for the world’s economies. The vast majority of the planet slept right on as if nothing had happened.* And they were right. But the new numbers still suggest the size and distribution of world income looks considerably different than we previously thought. The World Bank will produce new official estimates in the coming days, but our preliminary estimates suggest the share of people in the developing world living below the absolute poverty line of $1.25 per day in 2010 “fell” by nearly half, from about 19.7 percent to 11.2 8.9, thanks to the revisions.
Poor countries are somewhat richer than we thought…
Global poverty numbers involve two sets of data: national income and consumption surveys (collated in the World Bank’s PovcalNet) and international data about prices around the world. The ICP is in charge of this second set of data. It compares what people buy and at what local currency price they buy those things to come up with a ‘purchasing power parity’ exchange rate, a ratio that is designed to equalize the power of a rupee to buy what Indians buy with the power of a dollar to buy what an American buys. Tuesday, the ICP released their estimates for what those purchasing power exchange rates looked like in 2011.
In short, the new PPP numbers suggest a lot of poor countries are richer than we thought. For example, The Economist now estimates that China’s economy will be bigger than the US economy by year’s end. (This much to the delight of Eclipse author and colleague Arvind Subramanian, whose forecasts now look better than anyone else’s.) But it isn’t just China. India’s 2011 current GDP PPP per capita from the World Bank World Development Indicators is $3,677. The new ICP number: $4,735. Bangladesh’s 2011 GDP PPP per capita according to the WDI is $1,733; the ICP suggests that number should be $2,800. Nigeria goes from $2,485 to $3,146.
…And a lot fewer people are living on less than $1.25 a day…
The new purchasing power estimates also have an impact on the number of people under the international poverty line. If poor people in India can buy more with their rupee than we thought, that means they’re richer than we thought. We can use the poverty dataset Sarah and Justin recently scraped from PovcalNet to work out an estimate of how many fewer absolute poor people there are than we previously thought. We take the most recent PovcalNet survey data and looks at the number of people under $1.25 a day based on ICP2005 compared to ICP2011 numbers (in each case adjusted to the survey year using national CPI data).
To come up with a global 2010 estimate for the impact of the new ICP numbers on poverty, we use the approach that PovcalNet takes to ‘line up’ data based on national surveys done in different years — adjusting the income numbers by the change in private consumption per capita from national accounts data (more on the practice and pitfalls of this approach from Shaohua Chen and Martin Ravallion here). Note that a 2011 international dollar isn’t worth what it was in 2005. Allowing for US inflation, a 2005 $1.25 poverty line would be closer to $1.44 in 2011, so we set a ‘new’ international poverty line for 2011 US dollars of $1.44. The figure below (with a log scale) shows the ‘lined up’ data for countries — how much their estimated 2010 poverty (measured as $1.25 in 2005 dollars) increased or fell thanks to the ICP revisions.
The estimated number of $1.25 poor in India in 2010 falls from 396 393 million to 148 102 million, thanks to the revisions. The number of poor people in Nigeria goes from 88 87 to 60 51 million. And pretty much every country is sloping downward — using the same national survey data the new ICP numbers suggest much lower rates of absolute poverty.
The new numbers also suggest that Africa is home to considerably more poor people than any other region — the pie chart below shows the old and new regional breakdowns. South Asia’s share of the world’s absolute poor was estimated at about 49 percent on Tuesday. By Wednesday morning it had declined to 34 29 percent.
…But Estimating Poverty Numbers is Hard.
It should be noted that the estimates of global poverty based on the 2005 ICP and PovcalNet data we produce are not the same as the 2010 numbers reported on the World Bank website, but they are close: when we match latest survey data country numbers for poverty using the 2005 ICP our estimate is 19.7 percent and the World Bank estimate is 20.6 percent. (Two reasons for the discrepancy: (i) we cannot follow the World Bank in estimating poverty for the DR Congo and Central African Republic because the WDI doesn’t have the national accounts data for us to do it and (ii) we use a more recent income survey for Nigeria.) The [corrected] Stata code and data we used for the calculation are available here. The originals are available here. We hope that people [re-]check — and extend on — our estimates.**
What lies behind the dramatic changes in calculated GDP and poverty rates? A big factor may be that the national inflation rates used to convert incomes into 2005 PPP dollars in the last few years appear to be higher than the rate of inflation reflected in the baskets of goods and services measured by the two rounds of ICP surveys: Pakistan’s PPP conversion rate for GDP was 19.1 Rupees to the dollar in 2005 and 24.4 in 2011 — a gentle increase of 28 percent. The Consumer Price Index in Pakistan has gone up 102 percent over that same period. That might reflect changing or inadequate ICP commodity baskets or consumption data in one or both years, or mismeasurement of prices by Pakistan’s statistical agencies. But whatever the reason, it appears to apply to a lot of countries. Very few places saw PPP conversion rates climb close to or more than CPIs between 2005 and 2011, which is why poverty rates based on the 2011 PPP numbers tend to be lower.
And that leads to a caution — any exercise in calculating global income and poverty numbers has to come larded with caveats. Angus Deaton’s big concern with purchasing power parity is that it isn’t very meaningful to say we are equalizing the power of a rupee to buy what (especially poor) Indians buy with the power of a dollar to buy what an American buys because (especially poor) Indians and Americans don’t buy very much at all of the same stuff. Poor people and rich people consume different things. There are also worries that countries (and China in particular) may be economical with the truth when reporting their survey data. And lining up data to measure poverty in a common year using private consumption per capita data from the national accounts is a highly imperfect approach.
Perhaps most importantly, it is worth repeating what didn’t change between Tuesday and Wednesday. The people who have just been classified as ‘not absolutely poor’ don’t actually have any more money than they did yesterday, and will still struggle in terms of getting a decent job, and many still face grim daily tradeoffs between buying school supplies or ensuring their kids are well nourished. In fact, if the new PPP numbers suggest anything it is that the quality of health or education or access to services associated with a given income has just gone down.
The new numbers are good news for people who care about poverty, but they matter much less for people who are poor.
*They were all sleeping if you adjust for geographic solar-facing parity.
**Don Sillers at USAID is producing a new Visual Basic tool to allow everyone to make their own poverty calculations — including the impact of PPP changes — more easily.
DETAILS OF CORRECTION: In the original version of this blog we presented numbers that missed out a year of CPI data (2011) in the calculation of the adjustments caused by the update. This error led to numbers that underestimated the potential impact of the revisions on the 2005 $1.25 PPP poverty. We also used 2011 population numbers rather than 2010 numbers. These errors were pointed out by Laurence Chandy of Brookings (who has his own post on the new PPPs and poverty), and we are very grateful to him. It is a sign of how many poor people are bunched near the 2005 $1.25 PPP line that missing one year of CPI data can have such a dramatic impact on the numbers.
We have egg on our faces. We're very sorry to those who quoted our original estimates and are contacting a number of them to notify them of the mistake and updated post. The experience certainly confirmed the benefit of CGD's policy to publish data and methods including Stata files wherever possible –for those who want to compare old with new, the original files for this post are here, and a list of changes by country here.
UPDATE: The post, comments here and the subsequent Brookings post have illustrated the potential range of adjustments to extreme poverty numbers ‘caused’ by the PPP revisions last week. If you use the approach we adopted above (one option that Angus Deaton proposed at the end of his AEA presidential address), poverty fell by about half. If you use 2005(ish) national poverty lines updated for the new PPPs (another option that Angus Deaton proposed in his AEA presidential address, and the ‘minimal requirement’ proposed by Martin in the comments below) poverty fell by around a fifth (this is close to the Table 1C result from the Brookings blog). If you use Michael Lipton's FEM approach (see comments for details) or any other approach that doesn't involve the PPP conversion at all, by construction estimates of poverty wouldn’t have moved at all.
When it comes to the new official World Bank numbers, we’d certainly prefer an approach that (at the least) adjusts for the actual purchasing power of poor people (see Kaushik Basu on this point here). But going forward, if the World Bank extreme poverty numbers are going to be compatible with the World Bank’s mission of ending extreme poverty, we suspect the Bank’s approach to extreme poverty calculations is going to have to change. Previously, that approach included updating to the latest poverty lines of the world’s poorest countries. But it is implausible to imagine the poorest countries in the world in 2030, however fast they grow between now and then, would declare they had no poor people. That would be required if the world were to meet a global zero extreme poverty goal based on the average of the poorest countries’ 2030 poverty lines.
So it is a good time to have a discussion of how much change we want in how the poverty numbers are calculated, and if we can stick to that approach going forward to 2030. The World Bank has set a zero poverty goal, now it needs to fix the goalposts. (And it should be transparent with the data and methodology so everyone can know how close to the goal the world is.) We hope, despite the calculation errors and the heat and light, this post and the comments below have contributed to that debate.
All data used in this post can be found here (insert link to dta file attached). Raw data available here.
Corruption in aid has been a major topic of discussion in Washington in recent weeks. One of the questions reportedly from the Presidential transition team to the State Department was: “With so much corruption in Africa, how much of our funding is stolen?” During the nomination hearings for Rex Tillerson to be Secretary of State, Senator Rand Paul provided one answer: seventy percent of aid is “stolen off the top.”
The question is a fair one to ask. The bad news is that the short answer is “we don’t know.” The better news is that the slightly longer answer is “nowhere near 70 percent.” And the best news is that if we spent more time tracking the results of aid projects, we’d have a much better idea of where corruption was a problem and if our efforts to reduce it were working.
Statistics about corruption are hard to verify and open to considerable dispute. I've written before about a recent Supreme Court case where the justices strongly disagreed about what counts as 'corrupt,' for example. But also, for obvious reasons, people don’t tend to advertise the fact that they are involved in corruption. That all makes measurement hard. Nonetheless, three indicators of corruption might help answer how widespread the problem really is:
investigative cases of particular aid projects;
survey evidence about bribe payments; and
'expert perceptions' about the general state of corruption in a country.
What does each indicator suggest about the extent of corruption?
1. Investigative cases
The World Bank’s Sanctions Evaluation and Suspension Office keeps track of cases where World Bank investigations have uncovered evidence of fraud and corruption. An analysis of cases between 2007 and 2012 found sanctionable fraud or corruption in 157 contracts worth $245 million, of which less than a third of contracts showed evidence of sanctionable corruption. The World Bank’s lending volume is about $40 billion a year, so this suggests less than a third of contracts collectively worth about 0.1 percent of volumes over the period involved discovered and sanctionable corruption.
2. Survey evidence
Of course, that investigative cases only capture ‘discovered’ corruption is a huge issue. It is likely that the great majority of corruption isn’t uncovered by investigators. So what about asking firms that work on aid contracts?
Sadly, we don’t have survey evidence of corruption specifically involving businesses involved in aid contracting, but we do have surveys of corrupt payments by firms to government officials for government contracts in general from the World Bank. Enterprise surveys ask firm managers how much are the “gifts” expected in return for winning a government contract in their sector. In 22 countries, the average is more than 5 percent of the value of the contract. In 38 countries, the average is between 2 and 5 percent; in 52 countries the average is below 2 percent. You’d hope aid would be less affected by corruption than local spending given all of the oversight apparatus from investigators general to procurement experts involved, but in truth there is limited evidence they have a big impact, so maybe the numbers are about the same for aid-financed contracts. Given the $161 billion global aid business, an average of (say) five percent being lost to corruption adds up to around $8 billion—a real loss. But not the $113 billion that would be suggested by Senator Paul’s figure.
3. Perceptions measures
The trouble with measures of bribe payments, however, is that they only capture one form of corruption, and one of its impacts. They don’t capture officials simply stealing funds on their own account. And if the bribe goes to cover up substandard work which means the aid-financed road falls apart or the aid-financed drugs don’t work, the impact of corruption is far bigger than the bribe payment made. Again, we don’t have a measure of this broader corruption specific to aid flows, but you can ask a bunch of people how corrupt they think a country is in general as one potential indicator. Using one such ‘expert perceptions’ measure, Bill Easterly at one point calculated that 76 percent of US foreign aid went to countries judged to be corrupt.
Sadly, as Easterly himself noted, "It is inherently impossible to calculate the number 'x percent of aid is stolen' from the numbers that I use on how much aid goes to corrupt countries." And that’s not to mention the considerable problems with regard to the accuracy—and even the meaning—of a single measure of how a small group of people feel about overall corruption levels in a country.
So, using available corruption measures we’re left with three answers to “how much of our funding is lost to corruption”: one almost-certain underestimate of “less than 0.1 percent,” one possible underestimate of one part of the problem of “a few percentage points,” and one arguable interpretation that doesn’t really help that “a lot of aid goes to corrupt countries.”
Measuring Lost Aid through Outcomes
Luckily there's another, better way to estimate how much aid is 'lost': look at outcomes. If the aid program manages to buy all of the things it is meant to buy and deliver them where they are meant to go at a reasonable price, the aid funds can’t have been lost to corruption. Of course, measuring missed outcomes would be an over-estimate of aid lost to corruption specifically—a lot of other things can mean aid fails to deliver from bad luck to incompetence.
But at least it might provide an upper-end figure.
One partial measure of outcomes is project ratings. Take the World Bank, where project outcomes are rated on a six-point scale from highly unsatisfactory to highly satisfactory. Assume all projects rated in the bottom three categories are ‘lost’ and that adds up to 461 out of 1,617 projects completed 2010-2015, or about 29 percent. Be a little more generous and assume only the bottom two categories are ‘lost’ and that drops to around nine percent.
Again, Sarah Dykstra, Justin Sandefur, Amanda Glassman, and I looked at evidence for waste in GAVI—an aid institution dedicated to expanding vaccination coverage in developing countries. While we found plenty of evidence that developing countries were spending less on vaccines themselves when GAVI gave them vaccines at no or low cost, we estimated actual waste (vaccines not leading to vaccination) at between 0.5 percent and 15 percent of vaccines delivered, depending on the vaccine.
Or following on through the causal chain from aid financing to outcomes, countries that were the focus of attention from the President’s Emergency Plan for AIDS Relief (PEPFAR) saw the annual change in the number of HIV-related deaths between 2004 and 2007 that was 10.5 percent lower than other African countries. If the money had been lost to corruption, we simply wouldn’t have seen these results.
All of the evidence we have—on corruption cases, bribes, and outcomes alike—suggests corruption is a problem for aid, but nowhere near as big a problem as suggested by Senator Paul. That said, we still don’t have enough evidence on results to come up with any conclusive overall numbers on ‘the percentage of aid that delivers the impact it was designed to.’ And we should. If the administration’s concerns over aid and corruption ended up improving the results focus of aid programs, that would be great news for many reasons—but not least because it would help reduce the real impact that corruption can have on development.
Under managing director Christine Lagarde, the International Monetary Fund (IMF) has become a champion for gender equality. This note examines how much the IMF’s dialogue with its member countries has changed as a result of the labeling of gender as a "macrocritical" issue. In short, there has been increased attention to the issue as reflected in word counts and discussion of women’s labor force participation, but there is still a long way to go.
Development is easy, right? All poor countries have to do is mimic the things that work in rich countries and they’ll evolve into fully functional states. If only it were that simple. My guest this week is Lant Pritchett, a non-resident fellow at the Center for Global Development and chair of the Harvard Kennedy School’s Master’s program in international development. His latest work looks at how the basic functions of government fail to improve in some developing countries (a dynamic he defines as a “state capability trap”). Part of the problem, says Lant, is that donors often insist on transplanting institutions that work in developed countries into environments where those institutions don’t fit at all.
Despite decades of development assistance, on a wide variety of indicators of how well governments provide certain services—policing, delivering the mail, building roads, etc.—some countries are simply stuck in the mud. Lant’s work meticulously illustrates the depths of the problem. “We thought we would be able to replicate the development process very fast. We thought, these [countries] are going to develop in about 10 – 20 years,” explains Lant. “At the current rate of progress, it will take literally thousands of years for many developing countries to reach Singapore’s level of capability. That’s the capabilities trap.”
In our conversation, Lant unpacks the problems inherent in what he calls “isomorphic mimicry”: building institutions and processes in weak states that look like those found in functional states. “They pretend to do the reforms that look like the kind of reforms that successful [countries] do, but without their core underlying functionalities,” says Lant. “Instead, countries wind up with all the trappings of a capable system—institutions, agencies, and ministries—without its functionalities.”
How to break this bad habit? To start with, Lant says, the development community needs to understand that a lot of what it’s been doing hasn’t worked: “It’s just surreal, the disjunction between any grounding in what the empirical realities of the acquisition of state capabilities have been, and the way in which development plans of official organizations often assume that capabilities can be grown.” To paint a more realistic image, Lant suggests that it’s important to develop indicators that showcase real progress (and are resistant to the mere appearance of progress). In the education sector for example, instead of measuring enrollment rates, he suggests more effort to measure learning (see Charles Kenny).
Listen to the Wonkcast to hear our full conversation, including a discussion of Andy Sumner’s concept of the “New Bottom Billion” late in the interview. Have something to add? Ideas for future interviews? Post a comment below, or send me an email. If you use iTunes, you can subscribe to get new episodes delivered straight to your computer every week.
My thanks to Will McKitterick for his production assistance on the Wonkcast recording and for drafting this blog post.
Angus Deaton’s new book The Great Escape is a must-read for those interested in development simply because it is written by Professor Deaton, a world-leading expert in trends in global quality of life. I’m not all of the way through it but have found it fascinating so far, including his argument that aid doesn’t work (mostly). Add to the anti-aid pile another must-read by Bill Easterly, The Tyranny of Experts, which will come out in March 2014 and appears to involve a tripling down on his own argument that aid has failed.
When two people as luminary as Easterly and Deaton agree on something and you don’t, that’s a reason to check you aren’t the one in the dark — or at the least not wearing rose-tinted Ray-Bans. But I do think the evidence shows that aid can and has worked, even if much of it hasn’t. Indeed, it may even be that most aid doesn’t work, or at least is far from as efficacious as it might be. But that’s a reason to focus on quality, not a reason to give up.
Much of the last 50 years of development economics has focused not only on ‘how much’ but, critically, on ‘how well.’ So, for example, you can have very high investment rates and absolutely dismal economic growth if most of that investment goes to white elephants — or hardly functioning steel plants in the case of Nigeria. And the economic returns to sticking kids in school can be dire if none of those kids learn anything, not least because the educational returns to hiring more teachers or buying more books will make no difference if the teachers are unable or unwilling to teach. It would be surprising if the same general lesson didn’t apply to aid.
And we know the quality issue is a big one for the aid sector. Just on the donor side, there’s ‘aid’ that’s debt relief on loans that aren’t being repaid, ‘aid’ that’s spent hiring third-rate consultants at home to provide unwanted advice to countries they’ve never previously visited, ‘aid’ that pays for domestic agricultural produce and shipping that sums to multiples of the cost of that produce in recipient markets, ‘aid’ that pays for overstuffed bureaucracies mandating safeguards and processes that tie up recipient governments and slows down delivery. Then there’s ‘aid’ delivered to prop up friendly kleptocrats or buy seats on the Security Council or guarantee arms sales. And, of course, recipients use the outputs of aid projects no more — and probably less — effectively than they do investments generated by local funding.
These quality concerns matter ever more to aid impact because in terms of pure volume of resources aid is becoming ever less important. Government spending in developing countries now equals $5.9 trillion a year. That compares to DAC ODA of $0.15 trillion. Add in the issue of fungibility, and that means much aid effectively finances a small percentage of marginal investment projects on a government’s wish list. You’d hardly expect that to have a huge impact.
So how do you increase ‘quality’ — by which I mean delivering aid that might make a meaningful impact? The results of the academic aid effectiveness literature to date have been only so much use. They suggest, for example, that aid appears to more effective in promoting economic growth when given to richer countries currently receiving little aid, with stronger institutions and a healthy macroeconomic position. In other words, aid works best where it is least needed.
On the donor side, there is little hard evidence to confirm that the Paris agenda has been the key to highly effective aid, although surely giving aid only where it is wanted, working transparently and in partnership, and focusing on results are all good things. Examination of World Bank project performance, confirming the importance of ‘good countries’ to project outcomes as measured by the Bank’s IEG, also suggests that big, complex projects run by inexperienced task managers fail more frequently. But all of these factors together explain only 12 percent (approximately) of the variation in project outcomes.
So perhaps it is better to take a step back and ask, ‘how do we make aid do more than finance a small percentage of marginal investment projects on the government wish list?’ I think the answer involves safety, leverage, and learning:
As the number of low-income countries with extremely limited domestic revenues falls, the plausible role for aid as the ‘safety net provider of last resort’ in countries left behind actually grows. It’s a role that aid already plays with vaccines, which helps explain why GAVI-eligible low-income countries have higher vaccination rates than do lower-middle-income countries. And it is a role that in the future might extend to $1.25/day poverty.
Outside those situations, aid will only have an impact if it leverages other spending, either in developing countries or at the global level. In developing countries, this likely involves leveraging private financing through guarantees and partial investments — as long as those investments wouldn’t happen anyway — or leveraging more public investment through cash-on-delivery programs. Cash on delivery could perhaps be particularly aimed at global public good provision like lower pollution, reduced deforestation and unsustainable fishing, or reduced communicable disease spread. Another way for aid to have an outsized global impact is in the form of investment in and testing of new technologies for development: vaccines for neglected tropical diseases, robust forms of off-grid energy, drought resistant crops, and so on.
Finally, aid might have a role in promoting learning. If we’ve found one thing from 60 years of ODA, it is that there is no one-size-fits-all solution to development at the macro or micro level. Knowledge is itself a global public good, and there’s clearly a need for an immense amount more of it. Aid that tries new approaches, and properly evaluates them, is hugely valuable.
The traditional investment project–driven model of aid looks increasingly irrelevant. Frankly, the evidence suggests it has never worked particularly well. That’s why Bill Easterly and Angus Deaton can conclude aid is a waste of resources. But I’d take away a different lesson. Aid can work, sometimesspectacularly. And if we refocused aid, we’d see even more impact. So when it comes to evaluating aid, I think the key question is ‘does it focus on safety, leverage, and learning?’ If not, it is likely grist for Deaton and Easterly’s mill.
The Financing for Development conference in Addis Ababa in July represents one of President Obama’s last major opportunities to secure his development legacy. This memo offers 14 proposals from the Center's experts for commitments the United States Government should consider advancing for the Conference on Financing for Development. Each of the proposals has the opportunity to yield tremendous development impact, some as standalone USG commitments, and others as commitments ripe for broader cooperation.
There are more schools worldwide than ever before, but are children really learning? Charles Kenny investigates the broken link between schools and learning and suggests some proven methods for improving outcomes in education.
This paper attempts a first-cut listing of global public goods and international spillover activities, as well as providing some data on their global distribution alongside basic correlational analysis. Few if any goods are “pure” global public goods and there is a spectrum of the extent of spillovers. Some global public goods are not well measured. The listing is far from exhaustive, nor is it based on rigorous selection criteria. But it does suggest considerable diversity in trends, levels and sources of public good and spillover activities.
The Sustainable Development Goals are an ambitious set of targets for global development progress by 2030 that were agreed by the United Nations in 2015. A review of the literature on meeting "zero targets" suggests very high costs compared to available resources, but also that in many cases there remains a considerable gap between financing known technical solutions and achieving the outcomes called for in the SDGs. In some cases, we (even) lack the technical solutions required to achieve the zero targets, suggesting the need for research and development of new approaches.
How much innovation will be needed to meet the United Nations’ Sustainable Development Goals? Our results suggest that (i) best performers are considerably outperforming the average performance at a given income level, suggesting considerable progress could be achieved through policy change but that (ii) the targets set in the SDGs are unlikely to be met by 2030 without very rapid, ubiquitous technological progress alongside economic growth.
This analysis examines the relationship between legal reform and social norms surrounding homosexuality. First, about a fifth of the variation in individual preferences can be explained at a country level. Second, using a difference-in-differences strategy, legalizing homosexuality improves how individuals view the tone of their communities. Third, we provide further evidence supporting a legal origins argument by examining former colonies. We conclude that adopting legal reform can improve societal attitudes.
This paper analyzes six waves of responses from the World Values Survey to understand the determinants of beliefs about women’s roles in society and their relationship with the legal system and outcomes.
Governments buy about $9 trillion worth of goods and services a year, and their procurement policies are increasingly subject to international standards and institutional regulation. Using a database of World Bank financed contracts, we explore the impact of a relatively minor procurement rule governing advertising on competition using regression discontinuity design and matching methods. Our findings suggest the potential for more significant and strongly enforced transparency initiatives to have a sizeable effect on procurement outcomes.
In 1996, Burkina Faso enacted legislation banning the practice of female genital mutilation/cutting (FGM/C). Much of the qualitative literature surrounding FGM/C discounts the impact of legal change on what is considered a social/cultural issue. We use data from the Demographic and Health Surveys DHS(VI) in Burkina Faso to test for a discontinuous change in the likelihood of being cut in the year the law was passed. We ﬁnd robust evidence for a substantial drop in hazard rates in 1996 and investigate the heterogeneous impact of the law by region, religion, and ethnicity.