What is meaningful transparency for NGOs?

There has been an interesting debate online about what kind of transparency should be expected of NGOs for projects in which they receive government funding.

It began with a guest post on Bill Easterly’s Aid Watch blog by Till Bruckner. Till asked USAID to release budget details of NGO projects they fund. After a long delay, Till got the information he asked for, but a lot of the spending details were blacked out. Even though NGOs have publicly committed themselves to transparency and accountability, there was a lot of variation between them about what information was withheld.

This led Scott Gilmore at the Peace Dividend Trust to ask what information it is reasonable to expect NGOs to publish. In his view, we shouldn’t care what NGOs spend, provided they achieve results; NGOs have to retain some private information about their costs to enable them to compete for contracts.

Then Till replied to Scott, saying his purpose was to analyse how much aid in Georgia was actually spent there, and how much ended up back in the donor country. For this he needed quite specific budget details. He also challenges Scott’s view that aid agencies should retain proprietary information to enable to compete, on the grounds that sharing this information will encourage learning across organisations, and discourage back-room dealing for contracts.

Scott came back, saying that Till was partially right. Given his research question, Till was right to ask for this budget information (though Scott still does not think it would be very useful for combating corruption). But Scott continues to believe that NGOs are entitled to preserve some secrecy about the details of their work so that they can compete for projects.

Till’s post contains a splendid observation about the meaning of corruption which is worth reproducing in full:

“The aid industry has created a system that conveniently defines corruption so that expats can live a good life within the rules, whereas locals on far smaller salaries and with larger family commitments frequently get branded as corrupt for breaking these rules. In my experience, Afghan villagers do not share this narrow legalistic definition of corruption. When a project fails to deliver benefits to the poor, and the expat project manager at the same time lives a life of (locally) unimaginable luxury on designated poverty alleviation funds, villagers logically conclude that the project is failing due to corruption: instead of helping them as originally promised, the NGO is only helping itself. NGOs’ arrogant attitude – “we’re accountable by our own standards so we don’t need to tell you where the money goes” – does little to change this perception.”

It is also worth reading the discussion from several commenters.

So, what should we conclude from this debate? Our research on aid transparency from the past two years, published on this site, may shed some light. Here are five things we have learned:

First, people want access to this information for many reasons.

Till needed information to find out how much aid is spent in the recipient country, a perfectly respectable ask. We have conducted interviews in recipient countries with people who seek aid information for specific reasons: finance ministries want to improve resource allocation; central Banks want to manage flows of foreign exchange and the impact on domestic liquidity; line Ministries want to plan their own activities to complement those of aid agencies; MPs want to improve accountability for services; civil society groups want to hold government and NGOs to account; anti-corruption campaigners want to see how contracts were awarded, and how money was spent; private firms want to spot gaps in the market; communities want to shape choices that affect their communities; academics want to research the impact of aid, or how it has been used. And in industrialised countries, academics want information for research so that we can understand what works and what does not; taxpayers want information about how their money has been used; and policy-makers want to know whether what they are doing is effective.

We’ve learned it is important to not make assumptions about why information might be important, and in our work we have tried not to privilege one or other of these needs. All are real and legitimate, and the information sought is proportionate to the different benefits.

Second, people increasingly do not want only the information packaged for them by organisations.

It is both desirable and proper that NGOs and aid agencies should use their information to tell their story about their performance, about their overhead costs, or about how they used money. But people want to be able to recreate calculations themselves, or do a different calculation, from the underlying data; if an organisation wants us to accept its conclusions, then it has to publish the underlying information from which they are drawn.

This point of view became mainstream in the UK with the publication of details of MPs’ expenses. Parliament had wanted to publish summary totals (e.g. amount spent on accommodation, amount spent on entertaining), even though they had the specific details available. When the details were eventually leaked, it turned out that the most important information was not in the aggregates but in the individual transactions. It was the details that contained information that some MPs had redefined which home was their primary residence, and that others had claimed for the cost of a duck-house for their garden. If the information had been published only packaged as parliament wanted, some of the most useful information would not have been available.

Third, information is more useful when it is comparable.

In this case, Till wanted to find out information about the amount of aid that arrived in Georgia, and the amount that stayed in the donor country. That task is made harder if organisations publish information which differs in scope, definition and level of detail. When trying to establish whether aid agencies are making good value for money choices, you need certain basic information on a comparable basis.

Fourth, transparency is cheaper than secrecy

Look at all the effort USAID had to go through to redact (that is, black out) various parts of these contracts. Demands for information are only growing, by far the most cost effective approach for aid agencies is to collect and record information on the assumption that it will all be published proactively and immediately. Aid agencies increasingly have this information in computerised form, which is technically straightforward and cheap for them to publish online.

Fifth, the burden of proof is shifting to those who would keep information secret.

In the past, researchers needed to explain why they sought information. Organisations would contemplate these reasons and, if convinced, would release the data. As the costs of publication are falling, this process is changing (removing the “disproportionate cost” argument) and the assumption grows that the public is entitled to know how its money is being spent.

One area of disagreement between Till and Scott has been whether it is acceptable for NGOs to maintain their precise cost structure as a kind of trade secret, so that they can compete for contracts in the future. We have no research to offer on that issue. Our instincts are that, as Till says, the public is entitled to learn how NGOs funded with public money pursue socially important objectives. Transparency is likely to make competition between suppliers more effective and more efficient for the taxpayer, rather than less, and to reduce the likelihood of contracts being awarded to inefficient firms or on the basis of corrupt deals. There may be some details which should be withheld on privacy grounds (such as the salaries of individuals) or on security grounds (especially for people working in stabilisation and post-conflict environments). But in general, we think that when public money is being spent, the details should be published. Firms who would rather keep their business practices secret can choose not to bid to spend public money.