"Structures, Processes and Governance in Tax Policy-Making" By Carlo Cottarelli

March 8, 2012

By Carlo Cottarelli, Director, Fiscal Affairs Department, International Monetary Fund
Saïd Business School
Oxford, March 8, 2012

Let me first thank Mike Devereux and his colleagues at the Center for Business Taxation for giving me the opportunity to speak to you this evening.

The IMF is best known for its surveillance and program work with countries—very much to the fore at the moment, of course. What is less well known is that we also advise governments by providing technical assistance at their request in many areas, including tax policy and administration. In the last two years, the Fiscal Affairs Department delivered tax policy and revenue administration missions to 99 member countries, from headquarters, and we oversaw our eight regional assistance centers, organized conferences and seminars, and published various analyses of current tax issues.

Traditionally the bulk of our technical assistance has been to low and middle income countries, both in response to crises and to support longer term revenue mobilization. More recently, we find the demand for our advice increasing from advanced countries, in particular in Europe. This advice spans a wide range of tax issues, the general thrust being to move towards taxation of consumption, rents and polluters in order to raise revenues and support growth, while furthering the authorities’ equity objectives and promoting transparency and good governance. Tonight I do not want to talk about this advice itself, but rather to share with you some of the lessons we have learned on what works in reforming tax policies. I will highlight six lessons.

The first lesson is that public understanding and support are crucial to success. Tax policy reform will usually create winners and losers, and the losers can be vocal and well organized in resisting reform. The chances of successful reform are increased if parliamentarians, the private sector and civil society understand its objectives. While our technical assistance advice is confidential, we attach increasing importance to building public understanding through outreach activities. In Japan, for instance, we have long advocated an increase in the 5 percent VAT rate to help resolve serious fiscal challenges. In addition to highlighting this in our annual surveillance reports on Japan, and associated press conferences by senior Fund management, IMF staff have presented their arguments to influential tax commissions and in published papers and public events. We like to think that this helped smooth the way for the government's recent announcement of its intention to double the rate by 2014.

We have come to appreciate that champions are vital to tax reform in the face of vested interests. Often, one or two knowledgeable and energetic individuals can make the difference between success and failure. In Iceland, the Minister of Finance at the height of the crisis successfully championed reforms to raise revenues from all segments of the tax system. Liberia succeeded in overhauling tax incentives and re-designing the fiscal regime for natural resources, thanks in no small part to a dynamic Minister and Deputy Minister of Finance.

Transparency also plays an important role in stimulating public debate and scrutiny. As an example, we encourage and assist countries to quantify revenue losses from tax expenditures, for publication in annual budget statements. The aim is to subject tax expenditures to the same scrutiny as spending policies. The need to do better in this area, I should stress, is not restricted to lower income economies.

The second lesson is that good tax administration is good tax policy. Successful tax policy design must recognize administrative capacity. Effective tax administrations that make proper use of withholding and third-party information are often a pre-requisite for expanding the tax base.

The administration constraint bites particularly in fragile and post-conflict countries, where our advice focuses initially on collecting revenues at the border. Weak or corrupt tax administrations undermine trust between governments and citizens, eroding taxpayer morale, with negative consequences for voluntary compliance. Unfortunately there are no “quick fixes,” and strengthening tax administrations requires multi-year efforts to get the basics right.

But policy makers should recognize the reverse causation. Simple, transparent tax systems with broad bases simplify the job of tax administrations and reduce opportunities for non-compliance.

My third lesson: tax policy making is often about the law of second best—and often the tenth best. Successful tax policy reform must recognize not only administration capacity, but should also be tailored more broadly to country circumstances, implying that we must sometimes compromise certain cherished principles.

First principles would argue that base broadening should take precedence over increasing headline tax rates to boost revenues. But in many countries, base broadening may not be feasible in the short term due to lobbying by powerful interest groups. The U.S. tax system provides plenty of examples. The tax deductibility of mortgage interest payments is estimated to cost $100 billion per year in forgone revenues, is regressive and may have contributed to excessive leverage in the U.S. housing market. Yet despite numerous commissions recommending its abolition, no progress has been made.

Instead, when revenue needs are large, countries tend to increase tax rates, despite concerns about the effects on economic activity. The U.K. recently chose to increase the main VAT rate to 20 percent rather than streamline expensive and poorly targeted VAT exemptions and zero rates.

The lesson here is that even when politicians and civil servants understand what constitutes first best reforms, these can be hard to implement. As tax policy advisers and practitioners, we must recognize these constraints and at least adopt some degree of modesty.

Indeed, occasionally the roles are reversed, and we find ourselves advising that the first best may not be the most appropriate reform. China has excellent tax policy making capacity and the authorities are always eager to be on the cutting edge. We recently provided advice on measuring tax gaps, but found ourselves advocating a simple approach initially, despite the authorities’ enthusiasm for immediately adopting sophisticated measurement techniques.

The fourth lesson is that sequencing and packaging can determine the success of reforms. We advocate that reforms to raise revenues and improve efficiency be accompanied by measures to protect the poor. It is often possible to abolish tax exemptions while compensating low income households through targeted income transfers, but an announcement to introduce compensation at some future date will lack credibility. So it is important that compensatory measures are introduced simultaneously.

A mistake that governments sometimes make is to give away the sweeteners—for example, reducing rates—before implementing revenue-raising reforms such as streamlining exemptions. Inevitably, the difficult reforms are not implemented.

A related question is whether ambitious tax reforms are more likely to succeed if proposed as a “Big Bang” or if the measures are staggered over time. The answer will depend on a country’s implementation capacity. The abolition of mortgage interest tax relief in the U.K. provides a successful example of gradualism. Beginning in 1974, ceilings on the size of loans and restrictions on the tax rate at which relief could be claimed chipped away at the value of the relief, paving the way for its abolition in 2000.

But often, the “Big Bang” has more chance of success on political economy grounds.

The political window for tax reform after an election is usually short. There is evidence from Latin America that reforms are often an all-or-nothing process.1 The argument is that by introducing wide ranging reforms in a single step, there is a greater likelihood that groups which lose from some measures will simultaneously gain from others. Contrast this with a gradual approach, which risks being stopped at each stage by the group being hurt. In these circumstances, political constraints may outweigh economic arguments for a gradual approach.

The Slovak Republic’s tax reform in 2004 provides a good example. The reform introduced a 19 percent flat tax on income, to improve the efficiency of the tax system, accompanied by base broadening and an increase in the tax free allowance, to avoid increasing the tax burden on the very poor. The VAT was simultaneously reformed. The package was revenue neutral, but shifted the tax burden from direct to indirect taxes without hurting the poor.

I will call my fifth lesson “keeping up with the neighbors.” The design of tax systems in neighboring countries has an increasingly powerful influence on your own tax design as the forces of globalization intensify tax competition pressures. We provide tax policy technical assistance to Caribbean countries, for instance, where countries use low effective tax rates in the tourism sector to compete in attracting hotel chains.

Our technical assistance recognizes that effective tax reforms often require action at the regional level. This does not necessarily imply tax harmonization, but it may imply interim steps such as codes of conduct to prevent harmful tax practices.

Regional cooperation can be a slow process, but with patience it can yield results. Since 2005, we have supported a regional working group in Central America. The group has recently had a notable success with ministers of finance agreeing to a declaration of good practice in granting tax incentives. We hope that this will form the basis for a more binding code of conduct in future.

My sixth and final lesson is that strong tax policy making requires strong institutions. This is a topic that is discussed at length in the conference paper. Unfortunately we find that tax policy-making capacity is very low in many countries—sometimes shockingly so in relatively large countries. We are increasingly focusing on developing tax policy training as a core element of our technical assistance.

Successful tax reform usually requires a strong and specialized tax policy function within government, which in our experience is most logically located in the ministry of finance, although close cooperation with the tax authority is vital. In low-income countries, a scarcity of well trained staff is a constraint, and we advise establishing small units that can be grown over time.

Clearly a strong tax policy unit cannot be a panacea for weak political leadership or inefficient tax administration. However, tax policy units can help governments to resist opportunistic behavior by providing robust analysis and a well articulated reform strategy.

To conclude, our experience is that tax policy advice centers on bridging the gap between what is optimal and what is feasible. Our tax policy missions are no place for purists!

It is also important to acknowledge that there are rarely any “quick wins.” As I said, strong policy requires strong institutions, as well as transparency and debate to build public support, and regional coordination to counter harmful tax competition. All of these elements take time to establish.

I could cite examples of countries from every region in the world where, despite large and sustained levels of technical assistance, tax collections seem to remain fixed as a law of nature. But I could also cite enough examples of countries that succeeded in reforming their tax system (Canada, China, El Salvador, Iceland, Latvia, Tanzania and Vietnam) to remain optimistic that good tax policy making is alive and well. We will certainly need plenty of optimism in the years to come.

Thank you for your time, and I hope that you enjoy the rest of the conference.

1 Martinelli and Tommasi (1997), “Sequencing of Economic Reforms in the Presence of Political Constraints,” Economics and Politics, Volume 9, No. 2.


Public Affairs    Media Relations
E-mail: publicaffairs@imf.org E-mail: media@imf.org
Fax: 202-623-6220 Phone: 202-623-7100