Share

In 2014, Mark Lowcock, then head of the UK’s Department for International Development, pulled off an unexpected coup: securing an agreement between donor governments on new rules for counting official loans as aid. Some neat diplomatic footwork is needed again now, because negotiations over extending this agreement to donors’ investments in the private sector are threatening to fall apart. Among the consequences could be that the UK walks away from using internationally agreed standards for measuring aid and starts to create its own statistics. Other countries may follow.    

A better measurement system for unbiased aid allocation

Consistent aid data has intrinsic value, and the details of disagreements between countries are interesting, but the bigger picture is this: using aid to mobilize private finance is one half of the globally-agreed strategy for financing the Sustainable Development Goals (the other half is domestic revenue mobilization in developing countries). If donor governments do not get enough recognition for putting money into the private sector, they may allocate too little, and if aid statistics are inflated, they may allocate too much. The world needs an accurate system of measurement so that allocation decisions are undistorted and driven by expected development impact alone. And besides quantities, quality matters too: governments are also disagreeing over safeguards intended to protect the developmental nature of aid and, crucially, ensure that it does not become a means of providing state aid to domestic firms by the back door. But there is also a risk of imposing constraints on development finance institutions (DFI) that will limit their positive impact.

This is a complex topic with much for governments to disagree about, but negotiations are stuck on square one: how to count loans as aid. This new CGD note lays out all the gory details—here we present the nub of it. It should be stressed that negotiations are ongoing and these proposals are subject to change.

Where we are now

Governments are stuck between a rock and a hard place. The rock is the 2014 deal for loans to the official sector (sovereigns, municipalities, state-owned enterprises) which has set a benchmark that some countries refuse to undermine. The hard place is the regulated system of official export credits and lending rates observed in private markets, which define loans that should not be counted as aid. The problem is that, at least in the current low interest rate environment, the 2014 benchmark implies some loans made by export credit agencies, and sometimes private financiers, could count as aid.  

A further twist is that there are actually two proposed methods for counting private sector aid: the “institutional” method, which counts the full face-value of money governments give to their DFIs; and the “instrument” method, which calculates a number for aid—known as the ‘grant element’ of a loan—in every transaction between a DFI and a private enterprise. The deal currently on the table deviates from the 2014 official-sector deal, using a “private sector surcharge” to push down the ODA content of some shorter-term loans in the instrument method, in attempt to minimise inconsistency with the market, and export credits. Those countries which do a lot of lending and intend to use the instrument method feel unfairly treated by others who intend to use the institutional method which they, in turn, regard as an egregious scheme for inflating aid numbers.    

The truth is that there is no perfect answer. Even if donors carefully calculated aid using parameters tailored to each transaction, there would still be disagreement over what those parameters should be and no objective means of resolution. But an administrative system for creating aid data needs to be simpler, and compute aid by sorting investments into several boxes. There is a danger of getting hung up on a few transactions that might look wrong and forgetting the bulk. Export credits are not always the right comparator because DFIs often lend to riskier enterprises. Plus, the current low interest rate environment will not last forever. A deal should be struck that meets holdouts halfway on the quantitative side of instrument method, to be followed by qualitative safeguards to isolate transactions stray close to state aid, and ensure that only those with a clear development rationale, and which private investors would not finance themselves, get counted as aid.

The path forward

I have suggested two options for reaching a reasonable compromise:

  1. a system that deviates from the 2014 deal by identifying sectors where lending to a private enterprise is notably more or less risky than lending to the state; or,
  2. a more finely differentiated system of private sector risk premia that looks less like the 2014 official sector deal but which creates scope for giving the holdouts some of what they want, but not everything.

Although not currently the sticking point in negotiations, the proposals for equity need revising too. The introduction of a cap on negative ODA in the instrument method threatens to transform equity investing into an ODA pump. If I had my way, the institutional method would also be revised so that governments report an ex-ante grant element estimate when they inject capital into their DFIs, based on the institution’s targeted rates of financial return and its geographic coverage. But that may be too much to ask. 

A last word on transparency

A compromise is worthwhile on the assumption that most reported PSI will be genuine aid, even if somewhat inflated. This assumption needs to be verified. As things stand some DAC member governments want to retain the right to redact some transaction level data for PSIs on the grounds of confidentiality. The DAC must find a way to permit public scrutiny of how much aid is being claimed from what investments on what terms in what places, even if based on averages at some level of aggregation. DFIs should also publish the development rationale for each investment they make. Secrecy will only strengthen suspicions that ODA is being misused.