Ideas to Action:

Independent research for global prosperity


Views from the Center


Development Finance Institutions (DFIs) exist to promote development by investing in the poorest, least developed countries. They often route those investments via holding companies or private equity funds domiciled in tax havens. On the face of it, that seems absurd: tax havens are widely seen as a drain on development, depriving cash-strapped governments of billions of dollars in public revenue. Should the shareholders of DFIs demand a stop to this practice, as civil society urges?

In a new paper I argue that whilst widespread opposition to DFIs investing via tax havens is understandable, it is misguided. Banning the use of tax havens would do more harm than good. The argument rests on the answers to two questions: what do DFIs use tax havens for, and what are the alternatives to using them? The short answers are that tax havens are used to compensate for shortcomings in the legal systems of the countries that DFIs invest in, and that if DFIs stopped using tax havens, they would use onshore financial centres in rich countries instead. And that would not help developing countries one jot.

Why should DFIs continue to use tax havens?

To simplify matters, let’s say we care about two things: the amount of tax that developing countries raise from foreign investors and the quantity of investment in capital-scarce countries. We care because taxes pay for public services and investment creates better jobs and produces goods such as renewable energy.

Most of what we hear about tax havens has to do with multinational corporations shifting profits to low-tax jurisdictions. But in the context of where investors put their holding companies, what matters more are bilateral tax treaties. These are what determine whether investors pay a reduced rate on dividends or are exempt from capital gains tax, for example. Only a few tax havens have advantageous tax treaty networks, but established financial centres like London and Amsterdam have extensive networks. So as a rule, swapping a tax haven for an onshore financial centre would not help developing countries gather more tax.

A case in point is a recent accusation that a structure that CDC has invested in, via the fund manager Actis, has deprived the Ugandan government of £38m in capital gains tax by using holding companies based in Mauritius. As it happens, the UK also has a tax treaty with Uganda, so if Actis had put its holding companies in London, no capital gains taxes on the sale of equity would have been due in Uganda (and interest on the shareholder loans would have been taxed in London, not Uganda).

How about requiring DFIs to invest directly, without using holding companies or fund managers? Holding companies and funds should be “tax neutral” in the sense that their use does not result in investors paying more tax than they would when investing directly—this is why they will always be domiciled in specialist financial centres that allow for that (and meet legal and regulatory requirements). DFIs say they invest directly when they can, but that holding companies are sometimes useful for legal reasons and delegation to specialist fund managers works better when trying to invest in small businesses. The Ugandan power generator Umeme was once part of a multinational group, so some sort of cross-border holding company was inevitable. Less investment would reach developing countries if DFIs stopped using holding companies and delegating to funds.

Instead of avoiding tax havens, we need to reform tax treaty networks in favour of poor countries

At this point you may have two reactions: Why not regard London and Amsterdam as tax havens too? And how do we know the claims DFIs make, that the use of tax havens typically does not deprive developing countries of tax revenues but serves a useful purpose, are true?

The fact that problems with treaty networks and secrecy extend beyond stereotypical tax havens points towards the need for broader systemic reforms. Blacklisting a handful of islands is not the goal—raising standards however investments are structured, is. Multinational initiatives like the Global Forum on Transparency and Exchange of Information for Tax Purposes and the Common Reporting Standard are addressing problems around secrecy, but the points of leakage in bilateral tax treaty networks have not yet been fully addressed. To widespread acclaim, the OECD has introduced something called the “multilateral instrument” (MLI) to simultaneously update multiple tax treaties, and has used it to address certain weaknesses that enable “treaty shopping.” To really fix the problems that lie beneath routing investments via third-party jurisdictions, we need a “development MLI” to focus on the needs of the poorest countries, including minimum tax rates on dividends, interest, and capital gains.

The lack of evidence around the tax impact of using offshore vehicles is a thorny problem. The question of whether poor countries are deprived of tax revenues cannot be answered without specifying a realistic alternative, and that requires more than data—it requires informed but subjective judgements. But if it is wrong to ask DFIs to stop using tax havens, it is right to ask them to do more to enable public scrutiny of those investments. DFIs must invest effort into defining the tax implications of the offshore structures they use, and make that information public.


CGD blog posts reflect the views of the authors drawing on prior research and experience in their areas of expertise. CGD does not take institutional positions.