The Impact of China's AIIB on the World Bank

Founded alongside the rise of the United States at the close of World War II, the World Bank has remained throughout its history the leading international development organization. Over the years, however, it has been criticized for several problems, including unsustainable development policies,Footnote1 long project approval times,Footnote2 and inadequate financing capacity for infrastructure projects.Footnote3 It has also been associated with intrusive policy conditionality.Footnote4

In 2016, rising China established the Asian Infrastructure Investment Bank (AIIB), with fifty-seven founding members. This new institution joins two dozen multilateral development banks (MDBs) that have been formed since 1944.Footnote5 Its ostensible goal is to fill a shortfall in infrastructure investment in Asia and beyond.Footnote6 Yet the AIIB stands out for several reasons: its size and ambition; the absence of the United States and Japan; and, perhaps most importantly, Chinese leadership of the institution. The AIIB might represent real competition for the World Bank and for Western political influence in the developing world.

Of course, the prospect of alternative funding does not necessarily force governments to shift away from the West's foremost development organization. The AIIB has largely aligned its lending standards with those of other MDBs,Footnote7 and it might simply fit well within the existing MDB complex with little disruption to the World Bank.

Moreover, the World Bank might adapt. Scholars have found that Western donors have scaled back their demands of recipient countries in response to competition.Footnote8 Perhaps accordingly, Humphrey and Michaelowa find that Chinese foreign aid has not been a “game changer” for World Bank lending.Footnote9 If anything, Zeitz finds that the World Bank provides a larger share of infrastructure-intensive projects to countries receiving more Chinese aid.Footnote10

Still, governments dissatisfied with the World Bank may be looking for new options.Footnote11 To assess this possibility we consider a specific group of developing countries that joined the AIIB: the founding members. In considering the founders, we follow Broz, Zhang, and Wang who emphasize the importance of key countries undertaking costly action to signal their interest in new leadership of the global economy.Footnote12 They focus on countries that sent high-level delegations to a key summit on China's Belt and Road Initiative. In studying the founding members of the AIIB, we consider a costlier signal. The action constituted no mere summit meeting, but rather the founding of an institution—perhaps a sign that they are looking for alternatives to the World Bank.

For these developing countries that helped found the AIIB, we hypothesize fewer World Bank projects in the area of infrastructure—the specialty of the AIIB. In testing our hypothesis, we pursue an innovative identification strategy, using a recently developed method, the Dynamic Multilevel Latent Factor Model (DM-LFM).Footnote13 This model serves as an alternative to the difference-in-differences approach when the parallel-trends assumption is violated. Unlike the synthetic control method for comparative case studies, this approach accommodates multiple treated units and also corrects for biases induced by unit-specific time trends. We further analyze the data using difference-in-differences, negative binomial, Poisson, and two-way fixed effects models, as well as the generalized synthetic control (Gsynth) method.Footnote14 Within the literature on the rise of China, this paper is the first (to our knowledge) to employ this full set of methods, and specifically to use the approach of Pang, Liu, and Xu.Footnote15

Analyzing data on World Bank projects for 155 recipient countries from 1992 to 2019, we find at least short-run evidence supporting our hypothesis: a drop in participation in World Bank infrastructure programs by the developing founding members of the AIIB. Specifically, we estimate an average reduction of 22 percent in the annual number of new World Bank infrastructure projects in these countries.

Note that while we also estimate a drop in World Bank loan commitments for the founders, the estimated effect is imprecise and lags behind the effect on the number of projects. We can more confidently report a drop in projects. As our interviews suggest, the number of projects better proxies for government interaction with World Bank staff, its cumbersome approval processes, and its intrusive policy conditions.

The estimated effect is attenuated in 2019, suggesting that it may be temporary. Unfortunately, analysis of 2020 data is confounded by the pandemic. Kilby and McWhirter present evidence that World Bank lending follows a distinct pattern in this period.Footnote16 We leave the investigation of the impact of COVID-19 to future research.

Still, our methodology shows at least a temporary effect through 2019 for the developing AIIB founders. The DM-LFM accounts for both time- and unit-specific treatment and covariate effects, and addresses potential bias due to unobserved time-varying confounders by estimating latent factors. The DM-LFM relies on the key assumption of latent ignorability, which assumes that treatment assignment is ignorable conditional on observed covariates and unobserved latent variables.Footnote17 Estimates could be biased through a feedback effect, however, if the decision to become AIIB founders and the 2016 establishment of the AIIB were determined by countries’ previous borrowing from the World Bank. To address this possible selection bias, we conduct placebo diagnostics. The estimated negative effect emerges only with the founding of the AIIB in 2016, not before, which furthers confidence in our results.

Our approach does not enable us to distinguish whether the effect is driven by demand (the decisions of the AIIB founders) or supply (the decisions of the World Bank). Our limited interview evidence suggests that the mechanism runs through the demand channel, which is consistent with the bold move of the founders to establish the AIIB.Footnote18 As for a supply channel, we analyze key dependent variables from the literature—such as US voting behavior on the World Bank executive board, and levels of World Bank conditionality. We find no convincing evidence that the World Bank is punishing the founders. Furthermore, we find an effect of AIIB founding membership only on infrastructure projects, not on non-infrastructure projects—where AIIB founders still rely on World Bank projects because the AIIB has not yet focused on them.Footnote19

We conclude that the effect we estimate is driven by demand. When it comes to infrastructure, the AIIB founders have begun to turn their backs on the World Bank.

We recognize that this finding could be a result of “crowding out,” if AIIB projects simply replace World Bank projects. But not all the developing AIIB founders actually participated in AIIB projects during our sample period, and, interestingly, our results are not driven by AIIB project participation—the effect holds for both sets of founders. While we expect all the developing founders to draw on AIIB funding eventually, these countries appear to have become emboldened by the founding of the AIIB to distance themselves from the US-led World Bank—some of them even before entering into AIIB projects.

The AIIB founders may have turned away from these World Bank projects as a costly signal to encourage World Bank reforms. On the other hand, they might genuinely prefer China's institution and be willing to forgo the benefits of working with the World Bank to avoid the costs (slow approval and intrusive policy advice) as they look forward to working with the AIIB.Footnote20

The World Bank and the Founding of the AIIB

The World Bank's largest shareholder, the United States, publicly opposed countries’ joining the AIIB in the run-up to its founding in 2016.Footnote21 This opposition was perhaps not without reason. Some scholars contend that the very purpose of the AIIB is to supplant the old development institutions—along with the influence of the United States and Japan.Footnote22 No small operation, the AIIB is often portrayed as “China's World Bank.”Footnote23 Its president, Jin Liqun, has touted it as a “sophisticated, clean and efficient bank.”Footnote24

The World Bank is better established, but the AIIB has some advantages. Developing countries have long complained about the World Bank's project approval procedures and extensive social and environmental standards.Footnote25 The World Bank confesses that “improving the timeliness of lending operations remains a challenge”—the time from concept to first disbursement averages more than two years.Footnote26 By contrast, the AIIB's average time from concept to approval is a mere seven months,Footnote27 providing a “lean” and “clean” source of infrastructure financing.Footnote28

Though still in the early phases, the AIIB has already approved 172 projects, with 34 additional proposals under consideration (at this writing). Total financing for these projects runs into the tens of billions of US dollars. These projects go to more than thirty countries around the world; most are in Asia and the South Pacific, but many are in Eastern Europe, and some in Africa.Footnote29