This is the short paper I wrote for UNDP based on my presentation last week.
Tax is already the biggest source of public finance in developing countries. African governments, for example, raise over ten times more revenue through taxation than through aid. Even if only the “low hanging fruit” are addressed, the result is likely to be significant increased public resources.
This is borne out by a brief study of recent trends in tax revenue in low and lower-middle income countries. A simple average across those countries for which sufficient data is available during the years 2002-9 shows an average annual increase in the tax/GDP ratio during this time of 0.23 percentage points. In a few countries the ratio fell; if we exclude these, the annual increase rises to 0.44 percentage points. If this latter figure is a reasonable estimate for the potential increase that could be achieved through a concerted effort across all low and lower-middle income countries, the result would be a year-on-year increase in public revenues of $22.5 billion.
The potential of taxation as part of the post-2015 settlement is therefore evident. This paper considers some of the challenges inherent in the implementation of policies to increase the tax/GDP ratio.
Challenge 1: matching tax potential to revenue need
There is surely potential to raise more tax revenue in every country, subject to the trade-offs discussed below. But that potential varies between countries. Potential corporation tax revenue from policy reforms depends on the size of a country’s private sector and its actual and potential levels of inward investment. These depend in part on factor endowments, which vary between countries.
A paper from the OECD Development Centre in 2011 compared the public resources needed to meet the Millennium Development Goals against a simple measure of each developing country’s ‘tax potential’. “Although at the global level the magnitude of potential additional resources available from improved tax collection is similar to that of the additional resources needed to achieve the goals,” it concluded, “on a country by country basis substantial external resources will still be needed.”
A targeted approach to development financing might therefore reallocate external finances away from those countries with the greatest potential for increased tax revenue, and towards those countries with the least potential for domestic resource mobilisation. This is an uncomfortable consideration, but one that merits examination.
Challenge 2: raising revenue in a progressive way
In an ideal world, certainly that advocated by NGOs, developing countries would increase their domestic revenue through progressive taxes that have little or no incidence on the poorest, and significant incidence on the richest. Special attention would be paid to the gendered impact of tax reforms: the fiscal system wouldn’t just be gender neutral, it would redistribute income from men to women to compensate for the opposite effect in the economy and society at large. But behind these simple aspirations is a host of conceptual and practical questions. For example:
What does progressivity mean? In its simplest formation, it means that the more a person earns, the more tax they should pay as a share of their income. This seems fair enough, but what, for example, about people with considerable wealth but little income?
The chart below shows the distributional impact of a budget in the UK. It is progressive for everyone except the poorest 10%. Is it still progressive? In developing countries, might it be enough simply to say that a tax is progressive if its incidence on the poor is lower than on the middle classes?
Where does progressivity matter? According to the IMF’s Mick Keen, “[w]hat ultimately matters…is the distributional impact of the full tax-spending system,” not that of individual taxes or even of the tax system as a whole. According to this logic, it’s OK to tax regressively if you spend progressively, but not everyone would agree.
Progressive relative to what? In the abstract? Relative to what went before it? Relative to the alternatives? One frequent argument in defence of the implementation of VAT in developing countries is that it is more progressive than the trade and excise taxes that it has replaced.
Tax revenue as a source of finance brings with it many benefits, but it is important to be aware of the trade-offs, too. It’s unlikely that an ambitious approach to increasing tax revenues can be realised solely through taxes levelled directly on wealthy individuals and companies. One core gap in most countries’ tax policy toolboxes is the capacity to conduct the distributional analyses of tax reforms, by income group or by gender, needed for stakeholders to fully understand this trade-off.
Challenge 3: re-examining the global distribution of taxing rights
Because their tax bases are less diverse than those of developed countries, developing countries are more dependent on taxes on companies. Tax avoidance, illicit capital flight and tax incentives are all factors that constrain developing countries’ current revenues, and because of their greater reliance on business taxes, they have a greater proportional impact than in developed countries. But there is another aspect that merits consideration. 
Across the globe, the majority of corporation tax revenues come from multinational companies. The division of these companies’ tax bases is determined by national policy and administration, and by tax treaties, all influenced by international tax rules, in particular the UN and OECD model tax treaties and the OECD transfer pricing guidelines
International rules are not politically neutral – they have implications for the distribution of taxing rights between different groups of countries. China and India, for example, have already implemented measures to increase their share of the tax base beyond what is allocated to them by international rules.
A global development finance settlement, especially one that considers aid and tax in the same breath, ought to include an explicit discussion of this point from a development perspective, before reaching a settlement on how these revenues are to be (re)distributed. A commitment to shift the distribution of taxing rights over multinationals towards developing countries would provide them with a sustainable increase in revenues, while increasing the benefits from inward investment. If combined with effective measures to address ‘base erosion and profit shifting’, this could even be a revenue-neutral intervention for developed countries.
‘What gets measured gets managed’: a conclusion in three types of target
Tax/GDP ratio. This is an attractive starting point to measure progress in domestic resource mobilisation, but it must recognise the different starting points and potentials in different countries. Asking each country increase its tax/GDP ratio by, say, three percentage points over 10 years might therefore be better than a notional target that may be too easy for some and too hard for others. Alternatively, an aggregate target tax/GDP ratio across all developing countries, with differential contributions by different countries to achieving it, might make sense.
Progressive taxation. Hand in hand with any target incentivising increased tax revenue must be one that measures and directs the share of taxation paid by the poor, and particularly by women. It must do so in a way that stimulates, rather than constrains, public debate on a topic that runs to the heart of political discourse. That most developing countries lack the analytical capability to measure this at present is precisely a reason to create such a target, which might conclude, for example, that women should be net beneficiaries of the fiscal system, or that the bottom third of the population by income should always pay a smaller share of their income in tax than the top third.
Corporation tax. There could be a global financing target focused on the corporate tax base. This could be, for example, an absolute increase of 20% in tax revenues from multinational companies in developing countries. At one end of the spectrum, this target could be achieved at a (relatively small) cost to developed countries through a shift in the allocation of the corporate tax base to developing countries; in that case, it may be revenue neutral for multinational companies themselves. At the other end, it could be obtained through reforms that increase companies’ overall tax payments by reducing ‘base erosion and profit shifting’ and eliminate tax incentives, increasing revenues in both developed and developing countries. An intermediate option would combine the two, keeping revenues constant in developed countries while increasing them in developing countries.
Martin Hearson 