The World Bank has just released its 2017 Private Participation in Infrastructure Report.

Here are some key findings pertaining to Sub-Saharan Africa (SSA).

Only the SSA region saw “declining investment” by private sector actors into public infrastructure. 2017 recorded the second lowest level of private participation in SSA over the last decade.

Indonesia alone received nearly EIGHT TIMES as much private investment into large scale infrastructure projects as the whole of SS Africa combined!

Whilst Ghana and Rwanda are frontrunners in 2017, it is important to note that the pledges of investment in the case of Ghana are somewhat shaky. Analysts would recall that in 2016 pledges of $2.05 billion for the Tema Port expansion and the Amandi Power Plant couldn’t materialise in 2017 due to ongoing structuring difficulties. The $550 million pledged for the Tema LNG Terminal, which is the basis of the country’s strong performance in the 2017 World Bank PPI report, is likewise mired in administrative delays. Legal and commercial structuring continue to be a major challenge for many Public-Private Partnership (PPP) projects in Africa, generating expensive and distracting legal disputes by the hundreds.

Another insight that is also relevant for the points I intend to make in this brief note is the continued decline of government contribution to PPP ICT projects. In the last five years only one-third of large-scale ICT infrastructure projects received government support.

On the whole, private participation in infrastructure (PPI) projects, especially in SS Africa, continue to be critical. PPI drives foreign investment, which ensures the availability of resources to invest in the first place in Africa’s many cash-crunched economies. PPIs can also bring top-notch expertise and project management capacity simply not available locally.

But that is far from saying that PPI or PPP activity by themselves necessarily produce these benefits. It depends entirely on the principles and strategic logic governing the specific PPI/PPP activity. Let me illustrate this point with a case study from Ghana, in the ICT infrastructure area.

When the Bank of Ghana called for restricted bids in 2016 for a “retail payment system infrastructure”, it was clear from the context that the winner of the tender was going to be working on mobile money interoperability. The Central Bank’s commercial subsidiary, GHIPSS, had already been working on the other components of the system, such as RTGS and ACH. The missing piece therefore was connecting the mobile money platforms of the telcos to the GHIPSS infrastructure. Mobile Money interoperability (the ability of a Mobile Money – MoMo – subscriber on one mobile/telecom network to send funds directly to the wallet of a subscriber on another network) is believed by experts to be essential to the financial inclusion agenda. Interoperability however requires that all telecom networks in the territory accede to a joint protocol for managing settlement, disputes, and sharing of fees (if any) etc. The process requires technical capacity, the procurement of which can imply some costs.

The Bank of Ghana (BOG) decided that the project shouldn’t be paid for from public funds. A private investor should foot the bill and then recover the funds from the mobile money “industry”, in essence from the telecom networks that provide MoMo services in Ghana.

All well and good. However, alarm bells should have rung immediately the BOG saw the bids of the three companies they invited to the restricted tender. They claim that they emailed two other companies to submit tenders but the two didn’t respond. As for why the BOG simply didn’t publish an open tender, no one has so far provided a cogent reason.

One company said the job could be done for $3.2 million. Another said they could do it for $1.2 million. The third, an obscure integrator called Sibton, said it would cost $1.1 BILLION, but that they would raise the needed funds themselves and recover through levies charged to the industry.

The Bank of Ghana, making reference to its Request for Proposals, immediately disqualified the first two companies on the ground that those companies had asked the Central Bank to invest in various aspects of the project, and the Bank had made it clear in its RFP that it did not want any financial exposure.

But if indeed this was the BOG’s attitude, why was it getting involved in choosing a monopoly provider to foist a system on the mobile networks in the first place? If all it wanted was a “facilitation” role, then it could have simply followed precedent around the continent by convening the mobile networks and strongly encouraged them to undertake the interconnections and integrations themselves, as indeed they do for voice calls and other similar services. A number of SS African countries such as Madagascar, Kenya, and Tanzania, had indeed already gone this route.

Looking at the ridiculous gap between the costs quoted by the first two companies and Sibton, did it not occur to the BOG that the cost of interoperability was not that high and therefore that there was something completely off about the Sibton deal?

Anyway, insisting that the contract sum is ZERO, it awarded an exclusive contract to Sibton, authorising the company to spend $1.1 billion (per Sibton’s own preferred exchange rate) to build the interoperability switch and recoup the investment through a revenue sharearrangement with the mobile networks.

The BOG did not bother to establish the capacity of the company to raise this massive amount of money.

Even though the company had stated that it will spend $400 million on electricity and water and $92 million on insurance, no alarm bells rang. Even though no financial institution had committed to the project, the BOG saw nothing problematic. In fact, the BOG consented to an agreement whereby the financial proposals made by Sibton in its tender submission were incorporated by reference, verbatim, as commercial terms. This meant that Sibton was guaranteed to recoup the $1.1 billion plus a reasonable return on investment. Sibton had not been coy about the mean of this recoupment: a levy on mobile money operators.

How much was Sibton hoping to squeeze out of mobile money? An average of $40 million a year for the first 5 years. It was then to escalate this amount such that over the life of the contract (15 years in the first instance, and a further 10-year extension at its discretion) it would make back $1.1 billion plus profits. In fact, the contract was drafted in a manner that allows Sibton to adjust the revenue share amount (called “tariff” in the agreement) to ensure what its mouthy lawyers call “financial equilibrium”.

That point is critical to understanding the logic of this PPP: the contract and the financial proposal (which, as you might recall, had been incorporated into the contract by reference) were emphatic that Sibton shall spend $1.1 billion on the project (though no mechanism was provided to verify how much it actually spends). It must then make enough money from charging the mobile money operators until it has recovered the $1.1 billion and made a reasonable return on its investment, beginning with revenues of $200 million in the first 5 years.

The weirdest problem with this bizarre arrangement is obviously that it takes away all the risks away from the private operator since they are guaranteed that fees shall continuously adjust to ensure that come what may they will recover their investment plus profit. Meanwhile, the said investment had been pegged at a level that is at best several hundred times the cost of what the networks themselves would spend in skilled labour if they were to handle the interconnections themselves.

To understand the sheer monstrosity of the scheme, consider this.

The most profitable telecom company in Ghana is MTN. In its most recent financial disclosure, it revealed a profit figure of about $53 million. If we consider that the profit margins on its mobile money product is similar to its other products, then its profits from mobile money was a measly $7.2 million. But let’s even follow the crowd and assume that mobile money is wildly profitable in Africa. Let’s assume that rather than the 7% margin it makes on its general business, it makes 30% on mobile money. This will still imply a profit of only $31 million per annum on mobile money.

Now here is the meat. MTN in Ghana, according to the latest figures this author has seen, has nearly 90% of mobile money accounts and 92% of deposits! So practically, it makes most of the profits in this nascent industry. It is safe to say that the entire profit outturn in the “mobile money industry” in Ghana (on a 30% profit margin assumption) is not more than $40 million per annum.

For the Sibton deal to have worked, therefore, virtually all the profits in the industry would have had to be whittled away to satisfy a contractor providing very little value in the ecosystem. Or, more likely, the mobile network operators would simply have passed on the costs to the mobile money subscribers.

A PPP/PPI approach to managing MoMo interoperability in Ghana, in the way that the authorities designed it, would thus have cost more and delivered far less to consumers and the industry. Thankfully, there has been a change of guard at the financial helm in Ghana. A decision was taken to continue with the regulator-led approach by implementing a mediating switch (rather than allow the telecom networks to implement multiple, bilateral, interconnections as is the case elsewhere). However the design and construction of the switch has been awarded to the same subsidiary of the Bank of Ghana already managing other aspects of financial sector interoperability, GHIPSS. The cost will be less than $4.5 million upon completion, and already phase one of the effort (allowing wallet to wallet transfers) has been completed.

It is clear, even from this brief case study, that PPP/PPI models of infrastructural transformation in Africa shall offer scant relief to a continent reeling from heavy underinvestment unless serious attention is paid to how risks and rewards are shared between the public and private sectors, with a view to ensuring sound governance and administrative propriety and integrity.