Bright Simons: Is the World Bank saving or harming Ghana?

There is always much talk and jargon about “reform” of the “global development finance architecture” but it really all boils down to convincing richer folks in the “Global North” to release more cash to relatively poorer folks in the “Global South”.

That convincing has had many ups and downs. Whilst the World Bank has celebrated the replenishment of its grants and soft loans pot, the IDA, in 2021 as the biggest ever in history, and is looking forward to an even bigger inflow for the next replenishment cycle (which will be known as IDA21), development activists at the global level point to the massive gap between the $5 billion more per year the Bank committed to spend recently and the trillions of dollars experts say are needed to align needs in the developing world with the climate and resilience agenda ($4 trillion, says the UN).

For those of us whose activism mostly focuses on the country level, though, we sense a major gap in the global discourse: when the likes of the World Bank get more money, that doesn’t necessarily translate to bigger and better investments in developing countries. Various factors much closer to home than in Washington appear to conspire and create a “constriction” in the flow of allocated money. Attempts to clear up the pipe and speed up “disbursements” of the money, on the other hand, can lead to a deterioration in the quality of development projects and compromise the impact on people’s lives.

To examine this idea carefully in Ghana, the prime focus of my own policy activism, I have been painstakingly probing the outcomes of the World Bank’s investments there over the last two decades for a paper that has just been published by Paris-based Finance for Development Lab (FDL).

I conclude that the results presented by the World Bank’s reports on these investments substantially deviate from the reality on the ground about 70% of the time.

This analysis excludes certain investments made through the IFC, the World Bank’s private sector arm, or the guarantees issued by the World Bank’s MIGA, even though there are other reasons to be concerned about the IFC’s growing penchant for prioritizing malls and luxurious apartment complexes over social enterprises.

The “Disbursement” Bogey

In the early 2010s, the World Bank was struggling to disburse the funds it had allocated/committed to Ghana. The disbursement rate was languishing around 10%. In the ensuing years, efforts were made to considerably boost disbursements. Today, the global Bank’s ~$4 billion Ghana portfolio has an average disbursement rate higher than 48%.

I have selected a few case studies discussed in the FDL paper mentioned above based on their stellar ratings by World Bank staff. The idea being that if these are among the best projects from a disbursement and outcomes perspective, then there is even greater concern about the rest of the portfolio.

Ghana Energy Development & Access Program/Energy Sector Reform Program

In 2007, the World Bank launched a $220 million project called GEDAP. One of its major aims was to transform the Electricity Company of Ghana (ECG). From 2010 onwards, the World Bank has expanded the scope and ticket of GEDAP to double down on the intervention. Having started with a $70 million investment, an extra $60 million soon followed. In GEDAP assessments, the World Bank touts progress in lowering ECG’s commercial losses and assures evaluators that progress on improving bill collections, especially for power consumed by government entities, is imminent.

An enhanced effort, the Energy Sector Recovery Program, launched in 2019, with even higher focus on persistent ECG arrears accumulation and value chain debt. Boasting 88% disbursement rates and a “moderately satisfactory” rating, one might easily assume that these interventions have made a significant difference to ECG’s operational situation.

According to the World Bank’s official assessment of the results of these investments, a $55 million injection to upgrade billing and revenue protection, a “Commercial Management System”, and “Advanced Metering Infrastructure”, have been critical to salvaging ECG’s finances. For example, a provision was made in the program for 156,000 smart retail-level meters and 25,000 bulk meters to support revenue management. A timeline of January 2020 for all these powerful measures to bear fruit was indicated in the official project update.

Unfortunately, the facts on the ground, amply attested to by Ghana’s Auditor General and the ongoing rolling blackouts in Ghana today, do not bear out these assessments. Official audit data shows that ECG procured smart meters costing more than $145 million without competitive tender, contrary to its own procurement policies. Consequently, Ghana’s Auditor General failed the entire exercise on Value for Money grounds. Some contractors failed to perform on contracts, and the frustratingly long waiting times for securing a meter have not improved. ECG’s losses have actually increased since GEDAP commenced, to an average of about 30% over the project horizon. Taking full account of collection failures, the loss ratio actually exceeds 50% in some quarters. Very recently, ECG’s regulators fined the company’s board members millions of Ghana Cedis out of frustration with the company’s persistent flouting of regulatory directives.

e-Ghana/e-Transform

The e-Ghana project began in 2006, was reviewed when the new government assumed office in 2010, and wrapped up in 2014, whereupon a follow-up digitalisation program, e-Transform, was put together to carry forward the broad vision of ICT-driven development in Ghana.  Besides the World Bank, the EU chipped in, as also did the British and Danish governments (about 40% of the ~$60 million ticket).

When the government changed again in 2017, the usual review occurred. The digital ID component was dropped because a politically favoured vendor was not keen to play to World Bank dictates. Newer focuses like cybersecurity, digitalisation of the postal system, and the digital transformation of the national hydromet infrastructure were introduced to complement the core e-government initiatives. By 2020, disbursements had jacked up to 74%. There are projections of the disbursement rate hitting 96% in 2024. Here is how a few of the key modules are doing.

• GIFMIS

One of the most important reforms carried over from e-Ghana to e-Transform was the automation of the payments and payments-monitoring aspects of public financial management, through a platform called GIFMIS.

The World Bank’s assessment documents describe GIFMIS as “one of the major accomplishments of the project”. GIFMIS was said to have “added several major functions: treasury, budget formulation and execution, financial reporting and more transparent use of funds”.

GIFMIS was also hailed as having connected all key government Ministries and regional treasuries and spending units, and thus for having “improved the efficiency of government functions, in this case planning the national budget.”

The EU and the European bilateral donors disputed the World Bank’s findings. They insisted that their reading of the joint audits showed that at best GIFMIS was at 60% readiness. They pointed to persistent public financial management (PFM) challenges – like arrears, overspending, extra-budget allocations, and slippages – that a well-functioning automated payments platform should be able to address but GIFMIS seemingly couldn’t. Unable to make progress with the World Bank, the EU and the bilateral donors withdrew from the joint effort and transitioned to observers.

Recent reports by Ghana’s Auditor General also paint a very different picture of GIFMIS. It turns out that more than 86% of government payments meant to be covered by the system have been bypassing it, defeating its very rationale for existence.

• e-Procurement

The e-Procurement module of e-Transform was launched on 30th April 2019 with a unique selling point of cutting procurement costs for the government by $10 million a year over the ensuing decade. Analysts in Ghana estimate losses to the state due to poor procurement of $3 billion per year, the same as the country’s IMF bailout package. Over 600 public and state-controlled institutions were expected to make use of the new platform.

We examined 4000 of the 4875 tender results posted on the GHANEP e-procurement website. Based on our sample, we estimate that 95% of all tenders posted on the site are for amounts less than $100,000. We found hundreds of entries of less than a $100, and some below a dollar. Many records do not even feature award amounts, rendering them useless for analysis and defeating the very purpose of a transparent e-procurement platform.

For example, virtually every entry related to Bekwai Municipal Hospital and Nsawam Government Hospital was for a token amount of the cedi equivalent of a few dollars or even less than a dollar. Virtually all entries related to Asamankese Government Hospital, Oda Government Hospital, Sefwi Asafo College of Health, Asankragwa Nursing & Midwifery College, and Kade Government Hospital have no dollar amount mentioned at all. Public health institutions are overrepresented most probably because the reimbursement procedures of the national health insurance mechanism have made use of similar systems familiar to personnel. More than 60% of all entries relate to about 5 government owned health facilities.

Unsurprisingly, procurement activity related to the vast majority of World Bank funded projects do not feature on the GHANEP site. In fact, our extensive search in March 2024 revealed only one World Bank funded procurement activity featured, a Ghana Health Service office equipment purchase in December 2023, with reference number GHSHQ/2023/ODG/20. It goes without saying that all the multimillion-dollar government projects that have given Ghanaian activists serious grief over the last half a decade such as SML, the National Cathedral, and the Bank of Ghana motels do not feature.

• e-Health

The Integrated e-Health system promised in the e-Transform initiative was to deliver on telemedicine, mHealth, and training at multiple levels of the health ecosystem to enable service interoperability. Instead, the strategy never really took off in any concrete form. The clear calls for interoperability among digital solutions has been totally abandoned for a monopoly system owned by a US-based Ghanaian entrepreneur and developed in closed-source fashion by a contractor in India.

The platform, known as Lightwave, was declared by ministerial fiat as the only acceptable electronic health solution across the entire health network, even though many of the modules needed for fully-fledged health management operations were still under development. Even more perplexing, existing ehealth solutions provided by other companies were forced out, again by Ministerial fiat. The result has been extreme vendor lock-in and a decimation of the e-health innovation marketplace in Ghana.

• e-Justice

The e-Justice module went live in March 2019. In their assessments justifying the need to extend and expand e-Transform, World Bank staff hailed tremendous early successes: 43000 legal cases had been filed electronically within 2 years of launch.  Our survey of legal practitioners however indicate that the vast majority of processes related to court functions are still operated manually.

An evaluation by independent analysts show that the website has fallen into functional disrepair, the “effective launch” date has been pushed back to 2026, and the court authorities are now looking to engage consultants to implement a change management system.

• e-Immigration

Of the different e-Transform modules, the e-Immigration system presents the most bewildering account. It was envisaged to deliver automated immigration clearance at the airport (using biometrically-enabled e-gates), digital visa processing, and the phasing out of paper-based procedures across all borders (including land and sea). In short, a big deal. The e-gates submodule alone was budgeted at nearly $20 million. Central to all this was a Secure Border Management System (SBMS) meant to replace a US-donated platform on the grounds of enhanced data security.

6 years after SBMS was expected to launch, the web version of the US-donated system continues to be the primary immigration clearance solution in use at Ghana’s sole international airport. Despite claims to the contrary in the official World Bank records about the project, the fact on the ground is that no SBMS was rolled out. The $16.3 million e-gates that the official records claim were already functional and just needed to be transferred from terminal 2 to terminal 3 of the international airport, at the cost of an extra $2.9 million, have not been deployed to automate immigration clearance five years on. The multi-million-dollar electronic visa management system launched, according to World Bank records, in February 2019, failed to deploy to most of Ghana’s diplomatic missions abroad. The individual missions have had to engage service providers to build and manage separate systems at their own cost.

Despite good intentions, reality has not been too kind

It should now be apparent to the reader that official World Bank reports diverge substantially from the lived reality of domestic activists. Even though it was already apparent in 2020 that most of the touted deliverables on e-Transform were mostly on paper, the World Bank, in 2020, agreed to extend it for a further 4 years and even doubled the commitment from $97 million to $115 million.

It is also evident that World Bank staff operate based on assumptions of the intent of political actors that are not grounded in reality. For example, in one of their assessments of the status of the institutional reforms considered necessary for the success of e-Transform, the World Bank team gave the impression that the “broadcasting law” deliverable in the program would be passed by Parliament sometime in 2014. Ten years on, lacking any serious domestic political constituency, the law has still not been passed.

Clearly, World Bank staff regularly misdiagnose the political economy of the country when designing and monitoring projects.

For example, software platforms (called “TRIPS” and “e-Register”) developed under e-Ghana/e-Transform for the tax authorities and the Ghanaian companies’ registry by a company called GCNet and launched with great fanfare in 2013 and 2015 respectively. They quickly fell into disuse and never went mainstream because GCNet progressively lost political capital. Eventually, GCNet lost its legacy port automation business to a new politically favoured contractor, and was forced to concentrate on the Company Registry automation effort only to be hobbled by employee strikes, leading to a shutdown of the e-Register system. TRIPS, on the other hand, was eventually scrapped altogether and a new system built by a new vendor, called ITAS, was introduced in 2023. The tender leading to the selection of the successful vendor, a consortium constituted by TATA and local ICT firm, IPMC, was impugned by Ghana’s Central Tender Review Committee, an advisory body, leading to protracted contracting delays.

Sometimes, projects do get off the ground nicely but fail the sustainability test. One of the main hopes in rolling out e-Transform was to turn Ghana into a Business Process Outsourcing (BPO) hub and innovation powerhouse through effective incubation of startups. “Regional Innovation Centers” were to be set up in the capital of each of the ten administrative provinces in Ghana at that time. In e-Transform project assessment documents, the evaluators highlighted considerable progress by listing startups that have benefited from the initiative and are poised to make major contributions to the ecosystem. No resilience metrics were provided. Yet, 90%+ of the 25 startups listed as evidence of the success of the innovation program had ceased to operate within 3 years.

The same World Bank assessments used the growing presence of companies like QAI, Teletech, Comviva, Tech Mahindra, and ACS, to back claims that the BPO strategy had been wildly successful.

A few years later, the idea of decentralising the program to all provincial capitals was shelved and with that the concept of using them as e-government enablement platforms. The effort was consolidated into two facilities in Ghana’s two largest cities, which today mainly function as office space for well-connected startups looking for subsidised rent. The BPO momentum has almost fizzled out.

The lesson here is that disbursement rates for World Bank and other global development finance flows cannot be improved without fundamental improvements to the quality of governance. Trying to do otherwise is not merely wasteful, it can even be counterproductive.

By contributing to Ghana’s high input but low outcome resource use strategy, the World Bank may well be weakening social solidarity. Ghana’s current growth model is one that has seen the growth elasticity of poverty (the corresponding decline in poverty as a result of economic growth) decline by a staggering 25 times from -1.18 to -0.07. Consequently, both Ghana’s poverty gap and inequality measure (GINI coefficient) have worsened considerably.

Hence, whilst Ghana’s current World Bank average disbursement ratio of 48.6% (having accelerated from 16.6% in 2018) is markedly higher than Kenya’s 17.7%, the latest available measure of the share of “problem projects” in Ghana, at a five-year average of 30%, is significantly higher than the comparable Africa-wide average of 26%. Comparatively, Kenya has seen its GINI coefficient fall from 45 in 2005 to roughly 35 today. The latest measure of Ghana’s World Bank portfolio commitments at risk at 38% is likewise considerably higher than Kenya’s 22.2%, or even the Africa-wide average of 26%. To repeat: improving disbursement of development finance at all cost does not always generate positive social outcomes.

One could make the argument that waste and inefficiency are inherent in the nature of all government bureaucracy, included that of the so-called Global North. The difference is that Global North governments account primarily to domestic stakeholders for their development finance decisions. In the Global South, certainly in Ghana and most African countries, the likes of the World Bank and the IMF are critical stakeholders that are nonetheless fully captured in domestic accountability regimes. To the extent that countries like Ghana depend heavily on external public finance actors, waste and inefficiency are compounded by the gaps in domestic oversight.

Thus, alongside the push for more money to be pumped through global development finance channels, like the World Bank, we all ought to pay equal attention to the role of domestic activists in ensuring that the money will actually go towards real social impact. Such a move would require intentional resourcing of such “citizen and community verification and rating service providers”.

It is common in the commercial world for investors to pay ratings agencies to produce rigorous information on the jurisdictions they want to invest in and the governments they want to lend to. The World Bank and other development agencies ought to take a leaf from this playbook and invest in grassroots “community rating agencies” that can provide not just information, as classical ratings agencies do, but also the necessary political pressure to right wrongs in project execution proactively, or even preemptively.

And for added bonus, such a model may even escape the neocolonialist baggage often associated with World Bank – IMF “conditionality”.