A new dataset is throwing fresh light on old assumptions about government revenue in developing countries
We know that public finance statistics are open to interpretation. And we know that creative accounting can turn a fiscal deficit into a surplus, or obscure the extent of under-investment in public infrastructure. But most of us expect that the fiscal fundamentals are reported accurately for most countries in the world. That is, we – as citizens, as economic advisers, as investors, as aid donors, as bankers, and as managers of the world economy – should be able lay our hands on accurate information about, for example, how much revenue Peru raises relative to Gross Domestic Production; how that figure has changed over the past 20 years, and how it compares with Mexico over the same period. And we imagine that we can access all this information online from some large well-resourced international organisation like the OECD or the IMF.
Provided we are interested in the richer countries of the world, we are not often disappointed. We can access reliable basic revenue figures for Austria, Belgium, Canada or the USA. But once we begin to look at the middle and lower income countries, the picture changes. Sometimes the problem is that governments, especially those awash with revenues from extractives (oil, gas or mining), deliberately hide or obscure the truth. But, equally often, the problem is that there are various official-looking sets of data that give very different numbers for what should be the same figure. The most telling evidence for this lies in the use of the IMF Government Finance Statistics. Unwary researchers often use this, because it is considered the standard source. But the IMF’s own staff know better. They know that many of these figures are unreliable, especially in respect of the revenues governments obtain from extractives. As with many researchers, they often construct their own data sets on public revenues from a variety of resources.
Given this, it remains a mystery why the IMF has not invested more into producing and sharing a more accurate dataset. That mission was undertaken by the International Centre for Tax and Development (ICTD), a global research network headquartered at the Institute of Development Studies in the UK. ICTD has just released a new Government Revenue Dataset, covering the period 1980-2010, that is more accurate and comprehensive than any of the alternatives. It’s the fruit of three years of work: comparing, sifting and sorting through a range of existing sources, and has been published alongside a detailed account of how the dataset was compiled, so that its strengths and limitations are visible.
At the launch today, at the Centre for Global Development in Washington DC, ICTD researchers will be discussing how they have used this data to re-examine some very live questions about government revenue in low-income countries. For example, the new data tells us that, despite the fears expressed by many leading scholars and international organisations, there is no evidence that providing aid to the governments of poor countries reduces the effort they put into raising their own tax revenue.
We hope our efforts will be of real interest well beyond the ranks of researchers. We want this new dataset to contribute to improving governance, especially in but not limited to, developing countries, generating more and better informed debate about how revenues are raised and spent. We are also looking for an organisation to step up and take the responsibility for maintaining and further improving this valuable global resource.
Mick Moore is chief executive of the International Centre for Tax and Development and professorial fellow at the Institute of Development Studies
