As Prepared for Delivery
It’s October. Here in DC, that means that the annual meetings of the World Bank and the International Monetary Fund (IMF) are here! For me and my team at the U.S. Department of the Treasury, this is like Christmas or, perhaps, the Super Bowl.
Finance ministers and central bankers from all over the world come to town. Secretary Yellen meets counterparts. Think tanks and investment banks hold workshops. Distinguished academics debate various hot topics. It requires an immense amount of preparation. But it’s great.
The annual meetings are also a time to review the international economic policymaking process. One of my core responsibilities is to help oversee our relationship with the IMF, an institution that lends out tens of billions of dollars each year, tied to economic policy advice, and often with the goal of helping countries avoid, or recover from, crises. The IMF is clearly one of the most impactful elements of the global financial system, so it’s important to reflect on how to make it work even better.
In that spirit and toward that end, I wanted to kick off the season with what I consider to be five underappreciated facts on how the IMF is run – or what I refer to in the title as governance – and what I think we should do about them.
Fact 1: The IMF staff, and the IMF board, are (too) polite.
Don’t get me wrong, I’m a Midwesterner. I love politeness. But Keynes is reported to have said that the IMF had to serve as a “ruthless truth teller”. And I agree that it should do so even when it is uncomfortable, whether for borrower countries with programs or for their official creditors.
For example, the Fund often lends based on promises made by official bilateral creditors to forgive debt or to offer new financing, promises referred to as financing assurances. But even though this financing can be central to hold the IMF program together, the details on financing assurances – who gave them, when, in what form, and any failures to deliver on these commitments – are not always publicly disclosed.
In recent years, programs in Argentina, Ecuador, and Suriname, for example, all moved ahead on the basis of financing assurances that ultimately were either not delivered, or delivered with significant delays, something that was only addressed in passing by the Fund.
Just this past year, I traveled to Quito and to Islamabad to discuss with their reform-minded Finance Ministers the possible design of new IMF programs for Ecuador and Pakistan. In both cases, real resolve and decisive action will be needed to sustainably improve their fiscal situations, and the path of their adjustment will depend on the delivery of official financing. And yet, as with those earlier programs, the Fund’s papers on these new programs were similarly “polite”.
In the case of Ecuador, China has not provided financing assurances beyond the required first 12 months of the program. Rather than plainly detailing the risks this implies, including mentioning China by name, the program documents obliquely state that authorities are in discussions with their “main bilateral creditor.” In the case of Pakistan, program documents describe the total pooled financing coming from official bilateral lenders, but do not offer a creditor-by-creditor breakdown. We strongly support both of these programs, but for insiders and outsiders alike, this politeness can obscure key factors determining a program’s success. It also reduces the incentive for creditors to deliver, and ultimately to honor, their assurances in a timely manner.
Helpfully, the IMF recently took some big steps to fix this. Financing assurances reviews will now be more intensive in programs involving debt restructurings. This is an important improvement, but why stop there? A clear articulation of the details around financing assurances makes sense in all programs. I hope the Fund rigorously applies this new policy, plainly characterizes the relevant promises made and fulfilled or unfulfilled by official creditors, and ultimately, considers expanding its application so it becomes standard practice.
Another area where brutal, if impolite, honesty will be increasingly important relates to the IMF’s precautionary lending programs, notably the Flexible Credit Line (FCL). Countries that the IMF staff and Board judge to have very strong macroeconomic fundamentals and institutional quality, with sustained track records of sound policies, can pre-qualify for access to financing. Those countries can then draw down that line if needed without undertaking any required reforms or meeting other conditions, akin to a credit line from a bank. To date, six countries have had these FCLs, and some are quite large – Mexico’s 2016 FCL, for example, was for nearly $90 billion.
Last year, IMF staff recommended that, up to a size limit, these be treated as permanent, subject to periodic Board renewal, with no expectation or encouragement of qualifying countries to drop their credit lines. After all, countries pay a fee for these lines, and their strong macroeconomic positions and institutions merit a conditionality-free facility. The United States was proud to support this recommendation. It is good economics for countries to be able to purchase insurance in this way, and it is appropriate for the IMF to be the institution to provide it.
But this new policy will only constitute an improvement if the IMF staff and Board do not politely rubber stamp requests to renew these credit lines and, instead, seriously assess whether the country, in fact, remains qualified. If policies, or politics, or even shocks out of a countries’ control impair its economic situation or governance so that the borrower should no longer qualify for the credit line, the Fund must be prepared to acknowledge this.
The Fund is also sometimes too polite in its surveillance responsibilities. For example, as described in its Articles of Agreement, the IMF was established in part to “facilitate the expansion and balanced growth of international trade” and to “maintain orderly exchange arrangements among members, and to avoid competitive exchange depreciation”. However, if you read the Fund’s Article IV Reports, you will find that these sorts of topics get less attention than they deserve, including in globally significant economies like China’s. For example, the IMF does not publicly comment on the role of state-owned banks in managing China’s exchange rate or on why changes in the People’s Bank of China’s (PBOC’s) balance sheet don’t line up with reserve transactions in China’s balance of payments data. And while the Fund has built admirable expertise in compiling alternative measures of China’s local government debt, far beyond what authorities publish, it has neglected to apply this same analytical rigor to quantifying China’s industrial policies.
Finally, we the shareholders are also often too polite when voting on programs. A norm at the IMF Board has been to abstain, rather than vote no, to express disagreement. Why are “no” votes so rare? Perhaps early engagement between Board members, staff, and IMF leadership means that most programs only reach a vote if they command broad support. But it is important for the governance of any organization to have the ability to clearly voice opposition, in a non-stigmatized way, that is also reflected in the final tally. While I still expect it to be a very rare occurrence, I am convinced there are times when IMF members should ignore the niceties of an abstention, and vote “no” rather than accept a weak or problematic program.
Fact 2: The IMF is (sometimes) too risk averse!
Given the IMF’s role as a global lender of last resort to troubled economies, it may not shock to learn that, in many ways, it, and its Board, have a very conservative approach to risk. And for the most part, that makes sense and is what we support. But there are a few ways in which, by slightly relaxing its approach, the Fund may be more effective in accomplishing its mission.
One example is the Fund’s perspective on the possibility of having to tolerate some period of arrears. If the alternative is making disbursements to an off-track program where a country has not completed needed reforms, the IMF must choose arrears – the threat of arrears should not outweigh the need to maintain high standards. Only by designing strong programs and standing firmly behind required conditions can the IMF credibly affirm that countries are on a sustainable path. If a program is clearly off-track and disbursements are made simply so the country can pay back the Fund, it may in fact worsen the poor performance and make ultimate repayment less likely. Such behavior could also reduce other countries’ motivations to enact reforms and keep their own programs on track.
Another example is the Fund’s heavy reliance on co-financing from official bilateral creditors and multilateral development banks rather than having larger exposures on its own. Co-financing certainly has benefits – it can help share risk and marshal pressure and advice from other creditors to help the borrowing country adjust. But there are downsides too. Countries on the brink of macroeconomic disaster are sometimes forced to go hat-in-hand to bilateral creditors, which are often hesitant to lend more. In the worst cases, creditors use their significant leverage over the debtor, which correctly views the creditor’s co-financing as key to advancing their Fund program.
Though a more thorough analysis should consider factors such as the scale of shocks hitting program countries and compositional differences in program recipients, by some metrics, it appears that the size of IMF programs might be declining. Financing from the Fund averaged about 1.5 to 2 percent of GDP in the 1980s and 1990s before jumping above 3 percent during the global financial crisis. Average program sizes have since fallen to around 1 percent over the past few years, a period when we have assessed several IMF programs to potentially be too small and relying too much on co-financing. The IMF should keep a careful eye on this dynamic and, in appropriate cases, increase its exposure with bigger programs and a commensurate increase in associated conditionality and required reforms. In this vein, we also welcome a review on the IMF’s Exceptional Access policy, which was last reviewed more than twenty years ago, to make sure the IMF is bearing risk appropriately in these critical cases.
The IMF and its shareholders should also think through other possibilities to get more out of their existing resources. It should continue to guard its rock-solid balance sheet, a bedrock protection that we and other major shareholders have long relied on to keep lending costs low and predictable. But the IMF’s precautionary balances — the buffer it keeps to protect against large-scale defaults on its lending — significantly exceed the Fund’s target. As such, the Fund might consider using some of these excess funds to shore up the sustainability of the Poverty Reduction and Growth Trust, the IMF’s tool to lend concessionally to low-income countries.
Lastly, earlier this year the IMF completed a helpful review of its Lending into Official Arrears policy. The updated policy will afford the IMF with new options in some cases to move ahead with programs even if a large creditor does not agree to restructure its debt along with other creditors in a timely manner. Though such cases inherently involve more risk to the Fund, we encourage the IMF to use this policy when needed to support its membership.
Fact 3: Not all financing flows supporting IMF programs are created equal.
In April, Under Secretary Shambaugh spoke about the U.S. Vision for Global Debt and Development Finance, which has since formed the basis for the Nairobi-Washington Vision. He called on the official creditor community to provide net positive financing flows on a coordinated basis in support of borrower countries with IMF programs. If resources flow out of countries trying to pursue appropriate reforms, they may have to forgo needed investments for their development, and will have a harder time keeping their programs on track. Our hope is that official bilateral creditors, who are themselves IMF members and part of a global community aiming to support development, collectively maintain their financing support to countries during a program. That would be the right thing to do.
The Vision specifically calls for new support to be transparent and readily accessible. The reason this is so important is that not all financing flows are the same. In particular, the purchase of strategic assets or opaque project lending with conditions attached – such as requirements on how inputs are sourced or which labor is used – are generally not in line with the Vision, even though they may technically result in net inflows.
Bilateral swap lines that carry onerous use restrictions or where the terms are not publicly disclosed are also not generally in line with the Vision. And IMF data on gross reserves often include the value of PBOC swaps, even though such swaps often come with opaque restrictions on their use and potentially do not satisfy the IMF’s rule that reserve assets must be “readily available and controlled by the monetary authorities.”
IMF country teams have approached this issue in different ways. In the case of Sri Lanka, IMF staff have noted in footnotes that the central bank’s PBOC swap assets are currently unusable. For Laos, Fund staff indicated in a footnote that it is uncertain whether swap assets are tied to specific purposes, but that “if [PBOC swap] disbursements are tied to specific uses or not under full control of [the central bank], then they should not be counted” as reserves. In Suriname, staff created a “usable gross reserves” category that excluded the PBOC swap, while still counting swap assets toward gross reserves. Meanwhile, in Argentina, staff have continued to accept Argentina’s classification of its full PBOC swap line in its gross reserves, but staff exclude the “unactivated share of the PBOC swap” from their estimate of Argentina’s “FX liquidity.”
This confusing and inconsistent treatment partly reflects a lack of reporting by the PBOC on the details of each of its swap arrangements. The IMF’s treatment of swaps should be consistent and compliant with its stated policy when it calculates reserves, analyzes debt sustainability, and secures financing assurances.
Collectively, we should continue to cultivate enthusiasm and secure commitments from the official sector to invest in support of Fund programs. But we must make sure the result is the expansion of transparent, credible, and on-budget financing flows or debt relief to countries undertaking reforms, not potentially damaging forms of lending.
Fact 4: Third-party providers can sometimes help IMF programs, but sometimes can’t.
Supporting low-income countries, including those covered under the Fund’s Fragile and Conflict-Affected States strategy or where a government has a history of misuse of funds, is one of the most important things the IMF does. But programs in these countries are often also the most difficult because the IMF must minimize any chance of unwittingly abetting the drivers of corruption, violence, or illegal activity.
In these cases, the IMF has sometimes appealed to third party implementation as a way to spend funds or administer reforms without overly relying on the government or other risky actors. This is generally a good idea. If the third party is assessed to be more transparent, more efficient, and more law abiding, it is, by construction, a less risky way to provide critical government services or to address urgent humanitarian needs than simply cutting a check to the local authorities.
But we must remember that money is fungible. And this means that third-party implementation is no panacea. Imagine the IMF wishes to support a country whose budget has deteriorated due to rising prices of social spending, but where it worries that the government might misdirect IMF resources to buy weapons. Giving money to a third party may not help, even if it faithfully and transparently implements as promised. After all, with that social spending now off the government’s budget, the government’s fiscal constraint will be relaxed. It can simply reallocate that amount of money however it wants.
Third party implementation will be most effective, and is more likely to successfully act as a safeguard, when it is designed to perform a service that the government would otherwise not be doing, or where it can be additive to the government’s efforts, rather than simply substituting funds so the government can shift its spending. Given the importance of the IMF’s work in Fragile and Conflict-Affected States, and given the subtleties of this issue, the IMF should consider developing a policy on when and how best to use, or to avoid, third-party implementation.
Fact 5: The IMF’s got a really tough job … and thankfully has great people doing it!
Perhaps I should have started with this one. I’ve just encouraged the Fund to communicate more directly even when uncomfortable, to take more risk in certain situations but to be extra careful while doing so, to be nuanced in discerning which financial flows to encourage and which to discourage, and to actively deploy safeguards while staying attentive to their limitations. Each issue is obviously subtle and tricky. Like I said, the IMF has a really tough job.
Luckily, the Fund’s staff are passionate, hard-working, well-trained, and dedicated caretakers of the international financial order. They are open to feedback, and more than capable of pushing back when it’s off the mark. At Treasury, I have a fantastic and experienced team, deep in the issues, and able to brief on macroeconomies all across the globe. Even still, while preparing to visit a country, I find the advice of the experts at the IMF and its Article IV surveillance reports to be an obvious first stop and an indispensable resource.
I can’t think of another institution as successful as the IMF in promoting stability and growth and spreading macroeconomic expertise, and it has now been doing so for more than 80 years. I am confident that, with continued support and commitment to the institution, and with occasional tweaks to its policies and governance, the IMF will continue to do so for a long time to come.
Thank you.