The Discreet (but dubious) Charm of Tax Treaties Jayati Ghosh

At first sight, treaties preventing double taxation appear to be self-evidently fair: why should any individual or company pay taxes twice over, on the same income?

Income taxes are typically collected by local or national tax authorities on incomes earned within their jurisdiction but complications arise when these stem from traded goods and services or arise in a context of cross-border mobility particularly foreign investment. Double-taxation treaties, typically bilateral, purportedly resolve competing claims to tax revenue from cross-border investments between home and destination countries.

And such treaties are widely accepted, without much question, by governments and publics alike. But, as with so many other international economic agreements, this hides important inequalities working against the interests of lower-income countries.

Shifting resources

Martin Hearson’s Imposing Standards lays bare this inequality with clarity and penetrating detail. The book exposes how the network of bilateral tax treaties has become yet another means of shifting much-needed resources from capital-importing lower-income countries to capital-exporting richer ones.

Currently, there are more than 3,000 such treaties, covering 82 per cent of the world’s foreign direct investment. They almost inevitably take precedence over domestic law, even though they are negotiated (often in secret) by administrative authorities with no legislative powers.

Tax treaties are widely seen as instruments by which low-income countries compete for inward investment, with the suggestion of a tradeoff between their attractiveness to investors and the tax revenues they can raise. But in fact there is little evidence that tax treaties have a positive impact on investment in lower-income countries: empirical findings are mixed and suggest that the impact is minimal or irrelevant, compared with other factors which drive inward investment.

Taxing rights restricted

Many treaties are effectively designed to restrict the host country’s taxing rights over foreign investors. This can occur through rules on ‘permanent establishment’ setting relatively high activity thresholds for foreign companies, below which countries cannot tax them. For example, Mongolia cannot tax Chinese construction companies unless they stay for 18 months, rendering many exempt.

In many cases, incomes such as royalties, pensions and varieties of capital gain are taxable only in the home country of the multinational. For example, Uganda cannot levy capital-gains tax on Dutch residents should they sell holding companies there.

Maximum tax rates on cross-border transactions are often applied, to withholding taxes on dividends, interest payments, royalties and service fees. For example, by default the Philippines taxes dividend payments abroad at 30 per cent and interest payments at 20 per cent, but some of its tax treaties reduce these rates to 5, even zero per cent.

There are also typically rules setting out how exactly business profits can be calculated for tax purposes, which reduce taxable amounts relative to domestic companies. Concerns about unequal tax-sharing powers across source and home countries are particularly important for many mineral exporters. In addition, signatories to these agreements have to eliminate any remaining double taxation of their own residents when calculating the tax liability.

Transferring cost

All this means, Hearson writes, that ‘the real impact’ of tax treaties is often not to alleviate double taxation but to transfer some of the cost from the capital-exporting country to the capital importer, or to reduce the effective tax rate for investors operating across the two countries. Research by the International Monetary Fund has found that each additional tax treaty involving an African country is associated with a 3 per cent reduction in corporate-tax revenues and the IMF has even argued that countries ‘would be well-advised to sign [tax] treaties only with considerable caution’.

So while one might have thought bilateral tax treaties would naturally be initiated by either party, in most cases capital-exporting countries drive the negotiations. Indeed, home countries of multinationals compete with each other to give their own companies a competitive edge through advantageous tax treaties.

Because the lower-income countries have tended to be rule-takers rather than rule-makers, in general such tax ‘co-operation’ transfers some or most of the costs of double-taxation relief from the capital-exporting state to the capital importer. Essentially, the political and technical aspects of tax treaties operate to limit taxation in capital-importing states, which thus bear most of the fiscal burden.

This comes wrapped in supposedly technocratic concepts and standards embedded in model treaties. Yet, as Hearson warns, ‘Expert knowledge about international tax is far from neutral.’

Persuasion and coercion

So why do countries sign such constraining treaties, which come with substantial fiscal costs and uncertain, relatively low and often questionable gains? The answer lies in a combination of ignorance, competition, persuasion and coercion.

Ignorance has led many governments in lower-income countries, especially those with less technical capacity and in the early phases of entering into such treaties, to adopt tax treaties without much examination, in the hope that this will send positive signals to global investors as to their ‘market-friendliness’. Competition is directly related to this: many lower-income countries sign on simply because their peers, in the same region or at similar levels of development, have done so.

Persuasion is often implicit. Since the 1920s, higher-income countries have dominated the development of model tax treaties, first at the League of Nations and then the Organisation for Economic Co-operation and Development, giving them a strong first-mover advantage. The OECD’s model tax treaty has become the norm being generally seen as the ‘acceptable’ way to tax multinationals, despite the bias against lower-income countries, and so prevailing over the slightly less skewed United Nations model.

Coercion in turn can take many forms. As part of treaty negotiations, some capital-exporting countries are reported to have threatened to withdraw tax-related technical assistance even aid funds. Withholding of essential information from tax authorities is another means of coercion: it appears, according to Hearson, that just to obtain information from less co-operative secrecy jurisdictions, some countries have found themselves obliged to sign a treaty restricting their taxing rights, opening them up to ‘treaty shopping’. This has been found to be so with the Netherlands, the United States (which has a number of ‘tax haven’ states within), Singapore and the Seychelles.

But things may be changing. Several countries Indonesia, Senegal, South Africa, Rwanda, Argentina, Mongolia, Zambia and Malawi have cancelled or renegotiated tax treaties, while Uganda is apparently undertaking a review. It is time for many more lower-income countries to realise that they have (again) been taken for a ride in this supposedly technocratic economic exercise.

(This article was originally published in the Social Europe on May 15, 2023)