Climate Finance and the World Bank

World Bank President Ajay Banga has declared ambitious targets for climate finance in the Bank, having announced at COP 28 that the share of climate finance in WBG commitments would reach 45% by 2025.  There are problems with these targets that cannot be solved by vapid declarations at the COP gymkhanas or at the Hoover Institution: (i) the definition of climate finance; (ii) the path to greater climate impact, whether it is mitigation or adaptation; (iii) client demand for climate finance; (iv) staff incentives in project preparation and management; (v) the craven hypocrisy of the Bank’s largest shareholder; and (vi) pressure from the public to act. I start with point (vi: pressure from the public) because there happens to be a recent example which might mislead the unwary.

Pressure from the public

Oxfam has recently attacked^[1] the World Bank’s climate portfolio, alleging that large amounts have gone missing. Bank senior management is too craven to defend its own work, with the institutional weakness being, as ever, that many of those managers have only vague acquaintance with what the institution does, so it is left to me stand up for the Bank, at least in most of the instances cited by Oxfam.^[2]

The charges

A press release from Kate Donald, head of Oxfam International’s Washington office, stated:

“An Oxfam audit of the World Bank’s 2017-2023 climate finance portfolio found that between $24 billion and $41 billion in climate finance went unaccounted for between the time projects were approved and when they closed.

There is no clear public record showing where this money went or how it was used, which makes any assessment of its impact impossible. It also remains unclear whether these funds were even spent on climate-related initiatives intended to help low- and middle-income countries protect people from the impacts of the climate crisis and invest in clean energy.”

The evidence

I extracted the data from Oxfam’s published database, which its staff constructed from World Bank information^[3] and, apparently, advice from a few Bank experts.^[4] The Oxfam project sample (n = 732) showed US$28.660 billion as “budget” — the sum of projected WB financing in all components for a given operation — and US$27.209 billion as actual.^[5] The “actual” item is the sum of spending by component in a given operation before it closed.

Oxfam created a variable to express the relative deviation between budget and actual as abs(100 * (actual - budget)/budget). The headline allegation-—“between $24 billion and $41 billion in climate finance went unaccounted for”-- is the product of the relative absolute deviation in each project multiplied by its budget and then summed across the portfolio.

I verified Oxfam’s analysis by reconstructing part of their database. Taking the larger operations, I created a sub-sample of 72 operations with US$19.448 billion in budgeted WB financing which is 68% of the budget value of the Oxfam sample (data file attached, with the original Oxfam calculations and some extensions I have added).

First, deviations between budgeted and actual funding occur for practical reasons. Borrowers can cancel components, all or in part. Clients request reallocations. There are cost overruns or underruns, the latter often owing to depreciations in the local currency against the loan currency. Oxfam shows such deviations but wrongly interprets them to mean that the money “went unaccounted for”. The complete project documentation shows what happened—reallocations, cancellations, unit cost changes, exchange rate movements, and other reasons. An example of a cancellation and reallocation is the India PMGSY Rural Roads Project (project code P124639). IBRD and IDA gave program financing of US$1.860 billion of which US$0.225 million was canceled^[6] giving an actual amount spent of US$1.635 billion.

Second, the great majority of WB financed operations in the Oxfam review are obviously not ‘climate finance’ and no one claimed that they were.^[7] In my sub-sample (n = 72 projects, budget value of US$19.448 billion), the Bank assigned climate mitigation weights to 25 projects (budget of US$4.152 billion); and climate adaptation weights to 29 projects (budget of US$4.768). Oxfam’s procedure—to assign climate weights of 100 to every project—is wrong.

Third, many of the questioned projects have a small climate footprint: health programs (7 programs, unrated or zero climate rating), education (3, unrated or zero climate ratings), urban and housing (7, unrated or low ratings). Many are in emergencies, notably COVID response (Bolivia), war zones (Yemen, South Sudan, Afghanistan) or natural disasters (Nepal earthquake) and would in general have zero mitigation ratings and low adaptation ratings. Projects with obvious adaptation effects — Nigeria: Soil Erosion and Watershed Management (adaptation rating of 98 ; the two Ethiopia Safety Net programs (10, 54); Cambodia: Southeast Asia Disaster Risk Management (100); Myanmar: Flood and Landslide Emergency Recovery (46) — do have climate ratings.

Fourth, Oxfam claims that there is no “clear public record” about how the funds were used. This is really nonsense. By using the internal codes (e.g., P173984 for Bolivia: COVID-19 Crisis Emergency Social Safety Nets Project^[8])  anyone can find the project appraisal document, annual audits, the procurement plans, the supervision reports, and many of the contracts with details of what was spent on each component.

What the allegations mean

Oxfam pretends that all operations in its sample were 100 % climate finance, which is false, and then massages the funding numbers to arrive at deliberately misleading headline conclusions. It relies on the facts that most people have no idea how the Bank operates^[9], have no notion of ‘climate finance’, and would be unable to read and understand the reports from which the allegations derive. The Oxfam claims, taken at face value, mean practically nothing, but they do raise serious problems about the Bank’s roles in climate finance and in decarbonization.

1. The Bank’s portfolio has little on climate mitigation; in my sub-sample of US$18.348 billion in actual commitments, about 6 percent was spent on projects with mitigation ratings.^[10] 2. The adaptation portfolio is better in the sense that it is larger, constituting some 14% of the actual commitments in my sub-sample. The problem is that adaptation work suffers from high monitoring costs — how much do we really know about the standards of adaptation work and how much do we know about its effects ?[11]

3. The Bank’s President is bragging again (at the Hoover Institution, of all places) about the Bank’s plans to expand climate lending. Yet we see in the Bank’s recently issued targets that its annual mitigation goals are only 0.31 Gt CO2 eq, which is perhaps 0.5% of annual global GHG emissions.^[12]

4. From the Bank’s limited climate finance commitments to date, it is hard to see how the institution can achieve its targets, especially in adaptation, without damaging the credibility of project evaluation.

5. Ajay Banga recently accepted credit on behalf of the Bank for mitigation research on methane released from Vietnamese rice paddies. The purpose of this claim is to show that the Bank is being responsive to the global heating emergency. This is doubtful in several senses: (1) it has not been confirmed scientifically that the new paddy production methods (known as AWD: ‘alternate-wetting-and-drying’) save much methane; (2) the Bank’s role in the work was small;^[13] (3) the Bank should not be taking credit for what has been achieved by scientists working on AWD for years; and (4) again, even if we accept the claims about Vietnam, extrapolating them to other rice-growing situations will be costly and might not work.

I will say more in following posts on “Climate Finance and the World Bank”

* * *

[1]. Press release of October 17 https://www.oxfam.org/en/press-releases/41-billion-world-bank-climate-finance-unaccounted-oxfam-finds. The Oxfam report can be found at https://oxfamilibrary.openrepository.com/bitstream/handle/10546/621658/bp-climate-finance-unchecked-241017-en.pdf?sequence=6.

[2]. Lest anyone think that I have somehow been co-opted by the iniquitous Washington Consensus, a companion piece will review serious operational risks in at least one of the operations questioned by Oxfam.

[3]. The database constructed by Oxfam is at https://oxfamilibrary.openrepository.com/bitstream/handle/10546/621658/bp-climate-finance-unchecked-accompanying-data-241017.xlsx;jsessionid=2DD7E9904DECB27F902E53DBC12750FB?sequence=2

[4]. My honest thanks to one of the Oxfam people who explained their methods … but we are going to have to agree to disagree.

[5]. The Oxfam database in the sheet ‘budget_actual’ shows values of US$35 trillion for each of the 10 components of the NIGERIA: Kaduna State Economic Transformation for Results (P161998) when the true ‘budget’ value should be US$35 million for each component, giving a project total of US$350 million which is what the ICR shows (link at https://documents1.worldbank.org/curated/en/099205008182215477/pdf/BOSIB0e19d6c830de0a8f805e8120d9c273.pdf)

[6]. Lest some of our more excitable NGOs run shrieking to the Inspection Panel about “$225 million unaccounted for”, the project completion report (https://documents1.worldbank.org/curated/en/797521647873402937/pdf/India-PMGSY-Rural-Roads-Project.pdf) shows what happened. There were two cancellations; first, to reallocate IDA funding to help with emergency response work in the state of Uttarakhand (page 10, para 13); and second because of  contracting delays caused by the COVID pandemic (pages 15-16, para 33).

[7]. Now, getting ahead of the story a bit, it is undeniable that many of our distinguished global leaders and their hired academics lie all the time about what they are doing in climate finance. In a future episode, I will show how the WB tells such fairy tales but it cannot be said, as Oxfam does in this screed, that the Bank was so guilty in the projects under review.

[8]. The code P173948 can be used to find at least 26 documents at the link https://documents.worldbank.org/en/publication/documents-reports/documentlist?keyword_select=allwords&srt=score&order=desc&qterm=P173984&lang_exact=)

[9]. See for example the deranged reactions on LinkedIn to a re-post of Oxfam’s press release at https://www.linkedin.com/posts/activity-7256634933891207168-_vdz?utm_source=share&utm_medium=member_desktop. There is least one Oxfam Board member who was a senior Bank manager and should know better.

[10]. Notable were Argentina — Guarantees for Renewable Energy (P159901; guarantees made of US$0.395 billion, rated “Satisfactory” on Outcome and Bank performance); and Zambia — Solar Energy (IDA guarantee of US$5.7 million, rated “Satisfactory” on Outcome and on Bank performance, with US$105.5 million in co-financing).

[11] The ICR (March 25, 2019) for the Rwanda Agriculture Sector Program 3 (P148927) is alarming; the report says, “While reporting during implementation as part of the Aide Memoires and ISRs has been sparse, it appears that core environmental and social principles have been adhered to.”

[12]. The source for the annual mitigation target of 0.3 Gt CO2 eq is at the link https://scorecard.worldbank.org/en/data/result-details/CSC_RES_GHG_EMS_YEAR?orgCode=ALL&refareatype=REGION&refareacode=ACW&age=_T&disability=_T&sex=_T

[13]. For example, farmers have used complex water management systems in the Tonkin Delta for more than a century as Rene Dumont (“La culture du riz dans le delta de Tonkin”) showed in 1935.