A lot of people working in global development—including many at CGD like me—would like to see the World Bank’s lending arm for richer developing countries play a larger role financing global public goods like climate change mitigation and pandemic preparedness. That would take a larger IBRD, likely requiring a capital increase. But that only deals with one half—the supply side—of the financing challenge. The World Bank also has a demand problem: a lot of countries don’t want to borrow from it for the kind of long-term sustainable development projects that deliver global benefits. That has implications when it comes to how much World Bank shareholders can ask in return for providing additional capital.
Net IBRD lending over 2010-2020 averaged about $8.8 billion dollars, with disbursements running at an average of about $20 billion a year. That’s a lot of money, but we should put it in perspective: think of the $100 billion in annual new climate finance rich countries have promised, the trillion-plus in annual investment needed to meet the Sustainable Development Goals or, simply, the size of current IBRD borrower economies—about $29 trillion dollars.
Figure 1 displays total IBRD outstanding loans as a percentage of current borrower aggregate GNI and the country median of IBRD loans as a percentage of borrower GNI over time. As a proportion of aggregate GNI, the portfolio has fallen from 4.0 percent in 1987 to 0.7 percent in 2020 while the median borrower’s IBRD loans were worth 5.2 percent of GNI in 1987and dropped to 1.8 percent in 2020. This small-and-getting-relatively-smaller organization we want to both foster economic development in middle-income countries and solve all of the world’s collective action problems. We’d better hope both tasks are easy.
Figure 1. The Incredible Shrinking IBRD
A capital increase (along with a more aggressive use of existing capital) is surely an important part of the solution: the IBRD couldn’t return to anywhere near its relative size of the 1990s without it. But there is also a demand constraint to be overcome. Look at the largest middle-income countries, the ones with an outsized role in meeting many global challenges: India, China, Brazil, Indonesia, Nigeria, and Pakistan. Only in Brazil and Indonesia are outstanding loans worth more than 1 percent of GNI. Overall, there are only 9 countries where IBRD loans are worth more than 5 percent of GNI (the median level in 1987), and 30 countries where they are worth more than 2 percent of GNI (Figure 2). Some of the countries with low borrowing volumes are pariahs, some are new borrowers that have had little time to build up their IBRD loan portfolios, some (like China) are soon to graduate from borrowing. But those factors do not add up to an adequate explanation.
Figure 2. Where the World Bank Mattered in 2020
And this limited uptake isn’t primarily driven by IBRD lending controls to reduce risk. The countries where current loans account for more than 5 percent of the IBRD’s total outstanding portfolio are Egypt, Turkey, Colombia, China, Brazil, Mexico, India, and Indonesia. None is at above 10 percent of the portfolio and only India and Indonesia are above 8 percent. The Bank has a single borrower limit set at $23.5 billion for borrowers that are below an income of $7,065 at which the graduation process begin, but only China, Brazil, India, Mexico, and Indonesia have outstanding loans worth more than half of that limit.
Instead, while the World Bank likes to see itself as a long-term development partner, its middle-income borrowers clearly see it at least as much as a lender of penultimate resort (before the IMF). It appears that the average IBRD borrower increasingly prefers to obtain finance from its regional multilateral development bank (Figure 3, in nominal USD). In 2020, African countries received $4.7 billion gross disbursements in non-concessional finance from the AfDB and $3.5 billion from the IBRD, countries in the Americas received $12.3 billion non-concessional finance from the IDB and $8.7 billion from the IBRD, Asian countries $18.4 billion from the ADB and $11.0 billion from the IBRD. Furthermore, countries largely only resort to the World Bank for significant net borrowing in times of crisis—a proper job for the IMF, not the IBRD. Twenty years ago, the World Bank’s own Dilip Ratha wrote a paper on The Demand for World Bank Lending and found that IBRD commitments were positively related to an increase in debt service payments and inversely related to the level of reserves of the borrowing country, and little seems to have changed. Figure 4 displays the percentage growth of the IBRD portfolio in absolute dollar amounts and as a percentage of current (nominal USD) borrower GNI. Lending spiked on these measures during the Asian, Russian, and Latin American financial crises in 1997-98, the global financial crisis of 2008-10, and the pandemic crisis of the last few years.
Figure 3. We’d Rather Go Regional ($bn OOF Commitments, RDBs and IBRD).
Figure 4. Trusted Development Partner or Lender of Penultimate Resort?
That a lot of eligible countries simply don’t borrow very much from the IBRD might suggest the need for a rethink of the current country engagement model. When total outstanding loans are a percentage point of GNI, implying new operations worth a fraction of that amount, the whole World Bank tome-burdened bureaucrat-heavy institutional superstructure of country economic memoranda, country partnership frameworks, country procurement reviews, independent evaluation group country evaluations and perhaps even in some cases country offices, directors, managers, and program officers might be a bit excessive. Certainly, the clear assumption of the average country partnership framework that the World Bank is going to have a major cross-sectoral role in shaping development prospects might need tempering.
And that a lot of countries don’t see the attraction in borrowing very much from the Bank is a problem if we want the Bank to do more. Perhaps it is because finance ministries don’t want the hassle of dealing with the World Bank and/or can borrow from the market at similar rates, perhaps it is because line ministries don’t want the conditions or timing uncertainties that come with IBRD lending (even in an emergency), perhaps it is because Bank country teams don’t want to fund what client countries want to invest in. Former World Banker Johannes Linn suggested a list of reasons in 2004, most of which still seem relevant. But if you look at recent World Bank client surveys of some big IBRD clients, two things stand out. First, “World Bank Group’s processes too slow and complex” ranks near the top of a question about the institution’s greatest weakness, along with “too influenced by developed countries” (to be fair, similar results about processes appear to apply across multilateral development banks according to a recent ODI report: policy conditionality safeguards, complex procurement, and financial management systems are all considerable disincentives to lending from a client perspective). Second, when asked “where should the World Bank focus its resources?” respondents rank global issues like pandemics and climate considerably below issues like job creation, education, governance, and growth (although the ODI report suggests some evidence of stronger demand for climate finance from borrowers).
CGD blog posts reflect the views of the authors, drawing on prior research and experience in their areas of expertise. CGD is a nonpartisan, independent organization and does not take institutional positions.