How to Establish a Tax Policy Unit
Author: Martin Grote1
  • 1, International Monetary Fund

This paper explores that in developing economies, sufficient tax revenue is necessary to finance spending on health care, education, and infrastructure—all of which are prerequisites for economic growth and development. However, it is not simply the revenue ratio that matters; the quality of the revenue system is also essential for delivering fair and efficient outcomes. To design a revenue system that fosters sustainable economic and social development and enjoys broad public support, it is essential for tax reform proposals to be carefully assessed, quantitatively analyzed, and openly debated. This requires that decision makers and all stakeholders in the debate have access to the best available facts, data, and independent evidence-based analysis, including about the impact of tax reforms on revenue, the income distribution, and economic performance. The central institutional actor in the decision making process—the executive—is best supported in this process by what is generally called a tax policy unit (TPU). TPUs are tasked to guide and inform the tax policy debate, based on facts, independent data analysis, and multidisciplinary efforts.


This paper explores that in developing economies, sufficient tax revenue is necessary to finance spending on health care, education, and infrastructure—all of which are prerequisites for economic growth and development. However, it is not simply the revenue ratio that matters; the quality of the revenue system is also essential for delivering fair and efficient outcomes. To design a revenue system that fosters sustainable economic and social development and enjoys broad public support, it is essential for tax reform proposals to be carefully assessed, quantitatively analyzed, and openly debated. This requires that decision makers and all stakeholders in the debate have access to the best available facts, data, and independent evidence-based analysis, including about the impact of tax reforms on revenue, the income distribution, and economic performance. The central institutional actor in the decision making process—the executive—is best supported in this process by what is generally called a tax policy unit (TPU). TPUs are tasked to guide and inform the tax policy debate, based on facts, independent data analysis, and multidisciplinary efforts.

The design of a country’s tax system plays a critical role in its economic and social development, as recognized by the United Nations: its sustainable development agenda identifies strengthened domestic revenue mobilization as a key priority. Clearly, in developing economies, sufficient tax revenue is necessary to finance spending on health care, education, and infrastructure—all of which are prerequisites for economic growth and development. However, it is not simply the revenue ratio that matters; the quality of the revenue system is also essential for delivering fair and efficient outcomes.

To design a revenue system that fosters sustainable economic and social development and enjoys broad public support, it is essential for tax reform proposals to be carefully assessed, quantitatively analyzed, and openly debated. This requires that decision makers and all stakeholders in the debate have access to the best available facts, data, and independent evidence-based analysis, including about the impact of tax reforms on revenue, the income distribution, and economic performance. The central institutional actor in the decision making process—the executive—is best supported in this process by what is generally called a tax policy unit (TPU). TPUs are tasked to guide and inform the tax policy debate, based on facts, independent data analysis, and multidisciplinary efforts. TPUs also generally maintain oversight to ensure integrity of the tax system, and play a critical role in informing stakeholders along the path towards a coherent, manageable, fair, and efficient design of the tax system.

Many developing economies lack a well-functioning TPU. Some emerging market economies (Brazil, South Africa) and low-income countries (Burundi) have established TPUs or some specialization in tax policy analysis. Yet, in most developing economies, the government lacks such a unit; it also suffers from severe data constraints and/or has limited capacity to perform data analysis.1 This lack of information complicates decision making, hampers the debate about tax reform, and can lead to ill-designed tax policies or a tax system that fails the test of legitimacy in the eyes of the public—which can be detrimental to tax morale and compliance. Under these circumstances, even a small TPU (which could gradually be expanded) can be critical for making progress, not only in supporting good tax policy, but also in identifying the needs for capacity development, and coordinating efforts of donor countries and technical assistance providers.2

Based on several decades of experience in technical assistance on tax policy by IMF staff, this note provides advice and guidance for establishing a TPU. It relies largely on collective experience of tax policy experts, mainly from advanced and some emerging market economies, as well as a sparse literature on the issue, to draw lessons for countries that seek to make progress. The note covers such aspects as the role and functions of a TPU, its organizational structure, its location within government, as well as its relationships with other organs of state and external stakeholders. The analysis here is also informed by a review and comparison of various jurisdictions’ tax policy design function—sometimes organized in a TPU—in 25 high-, middle-, and low-income countries (see Annex 1).

Several lessons about what has worked well, where, and when, emanate from the analysis. Given, however, that institutional structures and processes of tax policy formulation vary significantly across countries, the study concludes that the structure of a TPU should be tailored to the needs, capacity, circumstances, and opportunities of individual countries. The note ends with ten general lessons for countries seeking to establish or further develop a well-functioning TPU.


A TPU generally encompasses four key functions: (1) guide general tax design and associated public consultations; (2) perform revenue and economic impact analyses; (3) initiate, participate, and oversee manifestation of policy content during legal drafting processes; and (4) contribute to international tax obligations by, among other things, supporting the country team that is negotiating tax treaties with economic impact analysis. Each of these functions is important, both during a period of major structural tax reform and when implementing smaller year-to-year tax changes.3

Guiding Tax Reform

TPUs play a critical role in guiding tax policy reform and producing objective analysis of the available tax options.4 They generally deal with analyzing and evaluating ideas for policy changes or new policies arising from many sources: parliament and government officials, business, labor organizations, and the general public. Within the government, key players in this process include the ministry of finance, which identifies and estimates public expenditure and revenue; other ministries, which attempt to achieve policy objectives through the tax system; and the revenue administration, whose role is to enforce the law and improve tax compliance. Often, these institutions’ objectives with respect to tax policy proposals compete, and complex policy and administrative trade-offs arise. The TPU plays a pivotal role in clarifying these trade-offs and guiding the decision making process under the government’s tax policy agenda. It should do so by providing impartial analysis and recommendations on how to best serve the public interest. In addition, a TPU can facilitate—through its tax analysis—a broader economic policy coordination function across ministries and agencies in terms of trade policy (tariffs), regional and local government own-source revenue development, sectoral strategies (extractive industries), unemployment and labor market developments, and entrepreneurship.

Regarding the implementation of tax changes, a TPU’s work program should be realistic, and its priorities should be based on measurable criteria (for example, preparing achievable revenue-raising strategies and targets in support of the medium-term budget framework). A pragmatic approach would be to rank priorities by distinguishing different categories of tax changes, for example: (1) items deemed essential and requiring legislation during the next parliamentary session to correct major drafting deficiencies in the law that are being aggressively exploited as tax loopholes; (2) items that are still important, but not critical, which should be legislated as soon as practicable, and certainly within the next 12 to 18 months; (3) items that, while still important, are less urgent and which can be deferred.

The TPU plays a leading role in explaining in a non-politicized way the economic rationale and intent behind tax policy changes and tax legislation. However, it is mainly the revenue administration that drafts interpretation notes that provide taxpayers and revenue administration officials with guidance and direction in applying the existing tax codes and regulations. In this regard, a TPU should advise the revenue administration and the ministry of finance on the revenue implications of draft interpretations before their publication. Policy interpretations or guidance notes are usually published to inform and guide taxpayers and their advisors, and this secondary or subsidiary legislation is an essential element of creating certainty vis-à-vis the tax law and its application. However, great care is required that discretionary powers of the tax administration are not inappropriately expanded without legislative oversight—for example, a TPU could monitor and alert the executive if such discretions exceed a certain predetermined materiality level.5

Revenue and Economic Impact Analysis

TPUs generally perform quantitative and qualitative analyses, often as part of the ministry of finance’s broader budget development process. These analyses should be rigorous, systematic, and based on transparent and up-to-date methodologies. The underlying methodological tools must be made publicly available so that they can be replicated and scrutinized by external stakeholders (for example, academics, civil society organizations and businesses).6 The following analyses are often conducted by a TPU (Bird 2004, UNDP 2008, Martinez-Vazquez and Heredia-Ortiz 2009); among them, the first two are most essential and of critical importance, also for developing economies:

  • Revenue forecasting is essential for every TPU. It has two aspects:

    • Projecting a baseline of tax receipts over the budget cycle for a given set of policy parameters. These are usually based on a combination of micro and macro data and play a vital role in the budget preparation process. They are also often used for other purposes, such as to establish collection targets for the revenue administration. Baseline projections are usually made twice a year: first for the annual budget cycle and, second, during the middle of the fiscal year to take stock and assess the variance from initial budget estimates for medium-term expenditure budget purposes.7

    • Predicting revenues under proposed tax law amendments relative to current law (revenue impact analysis), is often based on microsimulation models, which can infer the revenue effects of detailed parametric changes in the tax code on specific groups of taxpayers. Aggregating those effects provides insight into the revenue implications.

  • Tax expenditure analysis refers to the calculation of the revenue forgone due to specific provisions in the tax code in deviation of a benchmark system—such as exclusions, deductions, tax credits, deferrals, and preferential tax rates. Given that tax expenditures may constitute a sizable share of total public expenditure, tax expenditure analysis is vital for informing decision makers and other stakeholders about the costs and benefits of specific tax preferences, ideally annually or every second year. Most advanced economies perform an annual tax expenditure review, often published as part of the budget. This type of analysis is less common in developing economies (Figure 1).

  • Distributional analysis refers to the implications for income distribution among households, individuals, and businesses. This type of analysis is generally based on microsimulation models, set at the level of individual households or corporations. Sometimes, these models incorporate behavioral effects to provide ’dynamic scoring’—that is, forecasts including the impact of behavioral responses to tax policies. If market price responses are also captured by the model (which is the case in general equilibrium models), the analysis can take account of the incidence effects of taxes—that is, the ultimate effects on individuals after taking account of tax shifting effects that occur through market responses to the tax. The economic incidence of a tax change might be quite different from the statutory incidence and can be very relevant for the distributional implications of a tax change.8

  • Economic impact analysis often refers to the macroeconomic implications of tax policy changes on the labor market, investment, aggregate inflation, economic growth, and so on. This is generally carried out by employing partial and/or general equilibrium models that capture various behavioral responses to taxes and include possible interactions between markets. Different models will generally need to be employed for different taxes, for example, an elaborate household model for personal tax reform and an elaborate business sector model for corporate tax reform. Parameters are often estimated based on econometric data analysis or calibrated on the basis of existing empirical literature. Some models also capture international spillover effects of tax policies, such as through trade, the international allocation of labor and capital, and international profit shifting by multinational companies.

  • Cost-benefit analysis is the most comprehensive assessment for determining the desirability of certain tax provisions (such as tax incentives) based on a broad welfare concept. Other types of analysis are usually used as input for cost-benefit analysis, for example, a tax expenditure study is a necessary input to determine the direct fiscal costs of a certain tax provision (see IMF and others 2015).

  • Tax indicator analysis refers to the use of simulated indicators capturing certain distortions of the tax system (which can generally be compared across countries). One such indicator is the effective tax rate on business investment, which can be usefully employed to reveal distortions of income taxes, such as corporate taxes, withholding taxes, and personal income taxes (IMF and others 2015). Another indicator is the labor tax wedge, which summarizes the overall impact of taxes and social security contributions on the wedge between the cost of labor and the net wage for the employee—as an indication of the distortion to the labor market (OECD 2016).

Figure 1.
Figure 1.

Percentage of Countries where Tax Expenditures Are Periodically Calculated

Citation: IMF How To Notes 2017, 002; 10.5089/9781484323816.061.A001

Source: World Bank (2015).

TPUs are not always sufficiently resourced to employ the full range of economic models and economic analyses mentioned here. However, the most essential ones, such as revenue forecasting, tax expenditure reviews, and some form of costing tax policy changes (that is, revenue impact) are generally undertaken. In developing economies, it is advisable to focus initially on capabilities that improve the accuracy of revenue forecasting, quantitative assessments of tax changes, and calculating tax incentives’ revenue losses.

As countries advance on these fronts, models analyzing economic and distributional impacts could be developed, and cost-benefit analyses for major adjustments of tax instruments be conducted. It is quite common that governments outsource more refined analysis to a group of academics, external consultants, or a government agency (for example, autonomous fiscal institutes) that specialize in data analysis and economic modeling. As indicated in Annex 1, countries that have specialized fiscal analysis institutions include Italy, the Netherlands, Norway, Spain, and the United Kingdom. Moreover, some analyses are more appropriately conducted by the revenue administration, such as tax gap analysis (mostly conducted for value added tax or corporate income tax, as elaborated in detail in the Fiscal Affairs Department’s gap analysis program which constitutes a model for value added tax gap estimation)—or a methodology for revealing the untaxed share of the tax base due to non-compliance. The results inform discussions on revenue administrations’ compliance risk management strategies and complement tax expenditure analyses, which seek to identify revenue effects associated with policy design (the tax policy gap).

Legal Drafting

Legal drafting is a complex stage in the tax design process. Based on the policy formulation of intended tax changes, legal drafters must translate these into concrete and transparent law. On one hand, the applied drafting language must be tight to reflect the policy’s intent fully, without creating loopholes for tax avoidance and evasion. On the other hand, language must be informed by the interpretations of a jurisdiction’s courts. Due to the complex nature of tax legislation, it is important for revenue administration officials to be closely involved in the drafting of tax legislative amendments and to stay engaged throughout the legislative process.

Drafting can be managed by the office of the attorney general, yet be undertaken through a drafting committee with different participants, including TPU staff. Sometimes, a separate parliamentary office oversees legislative drafting standards or undertakes the drafting role directly. In some jurisdictions, all legislation is drafted by parliament’s counsel (for example, the U.K. Office of Parliamentary Counsel), but in close cooperation with tax lawyers of the TPU and revenue administration. The TPU will also manage the legislative program and participate in reading debates on draft tax legislation in a parliament’s standing or select committee responsible for the deliberation on tax law amendments.

Treaty Negotiations and Participation in International Tax Fora

International tax law expertise is needed for dealing with international taxation, such as negotiating bilateral tax treaties, considering appropriate source rules, anti-avoidance rules (thin capitalization and controlled foreign company rules), and transfer pricing. Ideally, this process requires knowledge about the costs and benefits of certain provisions in a tax treaty, calling for specific data analysis by the TPU on such issues. For example, the TPU should over time begin to get a better understanding of the effects of tax treaty provisions on inbound and outbound investments, as well as on tax revenue obtained from such flows. It should also assess risks associated with treaty shopping. Moreover, the TPU should contribute to discussions about internationally agreed standards, such as binding directives (for example, used in the European Union or the West-Africa Economic and Monetary Union) and be familiar with standards and guidelines developed by the Organization for Economic Co-operation and Development (OECD), Group of Twenty (G20) and the United Nations (UN). TPU senior staff should also be involved in international initiatives on tax coordination in respect of the G20/OECD inclusive framework on base erosion and profit shifting (BEPS), the global forum on automatic exchange of information, and the UN Committee of Experts on International Cooperation in Tax Matters. Countries’ tax treaty negotiation teams, constituted on an inter-agency basis, are commonly led by high-ranking revenue administration officials or a senior tax counsel in the Treasury, who should rely on the TPU’s quantitative analysis when negotiating treaty benefits for the country.

Governance Placement

TPUs are commonly placed within the executive branch of government, but there are exceptions. In Russia, for example, the TPU was initially situated in an academic environment and resourced through an independent foundation.9 This independence can have the advantage of providing protection against overly restrictive political influences of the analyses and forecasts by the TPU. However, an external TPU may suffer from limited access to confidential taxpayer data or a lack of information about controversial reform proposals. Instead of full delegation of all TPU functions to an independent institution, certain tasks can be outsourced (such as the development of economic models or certain types of analysis, for example, a tax gap study or an economic impact analysis). In some countries, tax policies are analyzed by an independent ministry of planning. However, this is not generally seen as best practice, as the staff of such a ministry might be insufficiently informed about broader fiscal policies. Thus, it can lead to advice that is inconsistent with the tax system’s overall integrity or with the wider fiscal policy design (Bird 2003).

In most countries, the TPU is placed in the ministry of finance. Out of the 25 countries listed in Annex 1, only two (Croatia and New Zealand) have assigned the tax policy design function to the revenue administration. In Argentina, Brazil, Chile, Guatemala, Peru, and Tanzania, well-organized tax research units are embedded in the revenue administration. These units perform similar or complementary analysis of tax reform proposals, for example, by reviewing revenue collections against targets, assessing and reporting on tax evasion and compliance trends, and producing tax expenditure reports. In these countries, there is generally a close working relationship between the revenue administration and the Treasury (responsible for tax policy design) in formulating tax policy and amending tax laws.

Placing the responsibility for the tax policy function within the ministry of finance—as opposed to the revenue administration—has two main advantages. First, it is consistent with the ministry of finance’s general responsibilities for fiscal policy and macroeconomic management. Indeed, TPUs located in the ministry of finance are often part of a wider fiscal policy unit that formulates both tax and expenditure policy. The fiscal policy unit’s functions and outputs are varied but are usually clustered around six core areas: macroeconomic analysis and forecasting; tax policy analysis; revenue forecasting and taxpayer statistics; expenditure planning and analysis; public debt management; and the analysis of intergovernmental fiscal relations (Martinez-Vazquez and Heredia-Ortiz 2009). Importantly, while tax policies preferred and advocated by the revenue administration can be expedient from an administrative perspective, they might give too little weight to economic, social, or distributional impacts or the wider fiscal stance. By placing it in the ministry of finance, the TPU is better able to serve the public interest in all its dimensions.

A second advantage of locating a TPU in the ministry of finance instead of the revenue administration is that the latter may be naturally more risk-adverse to policy changes (especially if they are frequent) that necessitate substantial adjustments to administrative and information technology systems. It is for those reasons that the revenue administration might be more inclined than the ministry of finance to refrain from proactively advocating tax policy changes that could improve the overall architecture of the tax system or that respond quickly to external changes. Still, a TPU needs to engage and consult continuously with the revenue administration to reap the benefits of “insider knowledge” with respect to taxpayer behavior and compliance, and to access taxpayer data generated by the revenue administration in line with an explicit taxpayer data exchange protocol. Institutional rivalry between the revenue administration and TPU should be minimized; if left unaddressed, it could undermine joint efforts in revenue forecasting and impact analysis and ultimately lead to poorly informed or highly biased tax policy debates and policy choices.

Staffing and Organization

The size of TPUs tend to vary considerably among countries, depending on the size of the country, its stage of economic development, and the governments’ capacities (that is, resources) regarding the process of policymaking. Figure 2 gives an example of a mid-sized TPU of approximately 35 staff. Alternative structures are also observed: for instance, direct and indirect tax experts are also sometimes responsible for the quantitative analysis of their respective areas; or the revenue forecasting group may cooperate closely with the macroeconomic forecasters in the budget office to assess the overall revenue envelope. In some advanced economies, the number of staff could be substantially higher. For example, Canada’s TPU employs more than 150 staff. In developing economies, however, the TPU or division dealing with tax policy design is much smaller—for example, Kenya’s National Treasury currently employs about 10 staff in its tax division.10

Figure 2.
Figure 2.

Example of an Integrated Tax Policy Unit

Citation: IMF How To Notes 2017, 002; 10.5089/9781484323816.061.A001

Note: The head of the tax policy department is synonymous to the head of a TPU.

A TPU should be staffed with interdisciplinary expertise—economists, statisticians, lawyers, accountants, and tax administrators (Bird 2004). At least some analysts should have practical experience in revenue administration and legal drafting. The interdisciplinary approach is important to guard against a narrow policy orientation.

TPU staff should be employed on a long-term basis; their positions require ample investment in human capital development through in-house training, which calls for firm commitment from both the staff and the organization. Experts should impart their knowledge on junior staff, evidencing the organization’s commitment to the lifelong training and development of staff. The TPU should offer market-based salaries to attract qualified staff and retain scarce skills. This is often problematic in developing economies. For example, some African countries face a limited pool of individuals with skills in accountancy, economics, statistics, and tax law. In addition, even when trained within the TPU, staff are often drawn away to the private sector or to a semi-autonomous revenue administration that offers more competitive wages. A TPU with a dynamic internship program, offering bright finishing PhD students practical work experience for a few years, could counter some of these destabilizing effects of the labor market.

The African Tax Institute

The African Tax Institute was born out of the fortuitous union of two initiatives: first, the United States Agency for International Development (USAID) and the U.S. Treasury supported, through financial and technical resource transfers, the creation of the South African tax policy unit in the National Treasury. Second, the Harvard International Tax Program promoted the establishment of an academic tax institution serving Africa. In 2002, the South African National Treasury, a private U.S. donor, and the USAID funded and cooperated with the University of Pretoria to establish the Southern African Tax Institute. Due to its wide regional acceptance and recognition for its successful tax training modules, the Southern African Tax Institute evolved into the African Tax Institute. The Institute is devoted to training, research, and technical assistance in areas of tax policy and tax administration for Africa. The Institute employs international tax experts, including from the International Monetary Fund, on a short-term basis.

Where foreign donors initially set up a TPU, it is critical to pair local staff with foreign experts and international advisors who initially run and manage it. These cooperative efforts encourage domestic staff to publish working papers jointly with international experts, which can raise the quality and usefulness of reports (Martinez-Vazquez and Heredia-Ortiz 2009). Analytical capacity can be provided by local academic institutions to preserve capacity for the future and encourage fresh inflow of academically trained staff, especially in cases where donor-funded skills transfers are expected to dry up after some time. An example is the African Tax Institute (Box 1).

Interactions with the Revenue Administration and other External Relations

The relationship between the TPU and the legal/ technical and operational departments within the revenue administration is critical. Tax policy analysts should consult with the revenue administration’s operational managers on the feasibility of tax policy proposals under consideration. Conversely, managers in the revenue administration should be responsible for alerting the TPU of possible weaknesses in the tax legislation that are identified while processing and auditing tax declarations, and for ensuring that anonymized taxpayer data is made available regularly in the format and within timeframes required by the TPU. Figure 3 illustrates the nature of these interdependencies. It shows the greater relative influence of the ministry of finance during the early stages of the policy development process when broad policy issues are being decided upon. Subsequently, the revenue administration has greater influence, as matters of detailed operational design of the legislation are being developed.11

Figure 3.
Figure 3.

The Tax Policy Development Life Cycle

Citation: IMF How To Notes 2017, 002; 10.5089/9781484323816.061.A001

Countries have developed different ways to organize collaboration between the TPU and the revenue administration. For example, Australia has broken down barriers by organizing preparatory work through project committees (Box 2). Such committees can foster a coordinated and effective approach to revenue forecasting, policy analysis, and guidance of the tax debate.

In the interest of well-considered and practical taxation policies, the relationship of the TPU and the operational side of the revenue administration should be formalized under a firm institutional arrangement (for example, a protocol). This is especially important if the revenue administration is organized as a semi-autonomous revenue authority. A mutually binding protocol could delineate how their respective inputs would be coordinated and organized. The protocol also needs to provide for conflict resolution when the Treasury and revenue administration cannot agree on a common tax policy stance and/or legislative matter. In such situations, the Treasury, as the leading policy institution, must ensure that the revenue administration’s opposing view is provided to the executive (the minister) in the format approved by the revenue administration. The minister should provide direction, and utilize input from the legislative counsel or a tax advisory body (for example, Australian Board of Taxation, Turkish Tax Council, and so on).

Australian Integration Model in Tax Policy Design

The Australian integration model in tax policy design and the drive towards transparent public consultations was initially applied to business tax reforms. The approach emphasized the need to move from a “sequential” taxation design process (different units organized in terms of labor specialization) to a holistic approach by combining units of different institutions in project committees or cross-functional teams. In formulating new tax proposals, the design committees draw their membership from the Treasury’s tax policy unit, the revenue administration and its various specialist units, and parliament’s counsel. These units maintain their separate identities, accountabilities, and core functions, but cooperate in committees or teams that are tasked to review separate aspects of the tax system or a specific tax policy issue. This tax design approach delegates, first, to the Treasury’s tax policy unit the primary accountability for tax policy development and for ensuring that the final products of the integrated design process achieve their stated policy objectives. It is also responsible for the oversight of tax laws and regulations. Second, the Australian Taxation Office is primarily responsible for the design and development of the legislative, administrative, and compliance systems to give practical effect to stated tax policy in addition to the implementation and ongoing administration of these systems. Finally, the Office of Parliamentary Counsel has primary responsibility for detailed tax design and drafting of tax legislation. These tax design working committees need to be resourced properly (especially if external expertise is drawn into such a process); the research brief needs to be drawn up and agreed upon by responsible government institutions, team leaders need to be appointed, and effective project management arrangements need to be established.

The TPU should have reasonable access to taxpayer data to ensure its successful operation. Given that revenue administrations’ collection activities generate data on the financial affairs of taxpayers, access to such data is legally protected and often subject to strict secrecy rules. Consequently, access rights must be carefully balanced within such cooperation protocols. For that purpose, the identity of taxpayers must be concealed or, in the case of large taxpayers, information needs to be merged with aggregate taxpayer data if it would allow for their identification (for example, if only one big oil company or brewery is operating in the jurisdiction). Access to taxpayer data is best achieved through the signing of binding protocols, as informal understandings or traditional practices based on an individual’s conduct and willingness to cooperate in a taxpayer data exchange may not be sufficient. The alternative of allowing TPU staff to take so-called oaths of secrecy to access confidential and individualized taxpayer data from the revenue administration is neither common nor favored by revenue administrations.12

The TPU should also coordinate with other stake-holders to ensure a smooth and sound tax reform process that receives broad public support. External consultation with the private sector, the statistical office, and parliament’s legislative counsel are important—and the evidence in Annex 1 suggests that this is almost universal. TPUs should be judicious as to which other institutions contribute to tax policy formulation so that hints of state capture by private sector interests do not arise (McIntyre and Oldman 1975). Also, communication is important to build support across party lines. Stakeholders should be informed about the need for certain tax policy reforms through active information campaigns. It is desirable that the TPU’s research and policy discussion papers be widely available on the government’s website (D’Ascenzo 2002, Du Toit 2004).

TPU and Advisory Boards

If major structural changes to the tax system are contemplated, the appointment of a special “one-off” commission may assist the government in deflecting political difficulties during the tax reform process. This has sometimes proven an effective tool for tax reform. Australia, Canada, Columbia, Germany, India, Italy, Japan, the Netherlands, New Zealand, Norway, Poland, South Africa, Sweden and the United Kingdom (mostly through the Institute for Fiscal Studies) have complemented the Treasury and revenue administration tax policy machinery with appointed commissions of inquiry into the tax system. Commissions often consist of panels of tax practitioners, business leaders, and academics. Some make use of tax specialists from the TPU, as they depend on their knowledge and secretarial inputs. Yet, tax reform commissions are not always successful. For instance, experience suggests that governments often appoint tax commissions to deflect popular resentment of the tax system, thereby postponing hard decisions. Also, tax-reform-by-commission may allow the government to cherry-pick recommendations acceptable to the executive while ignoring those elements of the proposed tax reform that underpin the tax system’s integrity.

Countries may also establish permanent review bodies or tax advisory boards, which report to the ministry of finance and serve as screening devices in assessing public support (mainly from tax practitioners and businesses) before draft tax legislation is tabled for wider public comment. In Australia, for example, a non-statutory Board of Taxation is tasked with advising the treasurer on improving the tax system. The Board comprises ten members, seven of whom, including the chairman, have been appointed from the nongovernmental sector. The Board is supported by a secretariat provided by the Treasury. Permanent advisory committees may, however, sometimes run the risk of lobbying for special interests (or perceived as doing so), which should of course be avoided; the latter can be achieved through governance rules that excuse members from deliberating on matters in which they have a direct interest. South Africa and Turkey have put in place similar advisory boards, which provide an institutional platform for engagement with taxpayers on identifying inconsistencies in tax laws and seeking ways to alleviate compliance burdens.

How to Establish a TPU?

Most advanced economies—in contrast to many developing economies—have well-functioning TPUs that provide revenue forecasts and economic impact analyses, support legal drafting and tax treaty negotiations, and guide the tax policy debate. For developing economies, establishing a TPU is a vital step toward building an effective revenue-raising system. To that end, experience yields a number of lessons that may be of assistance to countries in the process of establishing or advancing a TPU:

  • 1. Establishing a TPU requires broad support in government. There should be wide political consensus regarding the importance of effective, efficient, and fair domestic revenue mobilization. The ministry of finance should convey to other organs of state the TPU’s central role in designing tax policies.

  • 2. A TPU may start on a small scale and grow gradually. Low-income countries with limited resources and weak institutionalized capacity could begin with a small TPU comprising perhaps five staff—for example, two analysts focusing on direct and indirect taxes, one tax lawyer, and two economists in the revenue analysis section (focusing on tax expenditure budgeting and revenue projections). This could gradually be expanded with more specialized functions.

  • 3. The revenue analysis section of the TPU may initially operate as a technical working group with delegated staff, for example, from the statistical bureau or similar institution, thereby leveraging scarce skills and knowledge in government until the TPU’s own capacity is expanded.

  • 4. The drafting of tax laws may initially be assigned to the attorney generals’ office, with growing involvement of the TPU over time. The legal drafting process could be executed by an interdepartmental committee with treasury and revenue administration representation. Ultimately, as tax lawyers join the TPU, the unit could draft tax laws itself, or provide expert advice during the tax law drafting process where that function is in another government agency or parliamentary body.

  • 5. The TPU should ultimately employ an interdisciplinary group of experts, with qualifications in economics, econometrics, statistics, tax law, and accountancy. Staff should have advanced academic qualifications, together with experience in government, budgeting, and taxation.

  • 6. TPU staff positions need clear job descriptions and performance agreements, aligned with the short- and medium-term strategic plans of the unit.

  • 7. TPU leadership must have solid experience in tax policy theory and contemporary tax debates. The unit will be responsible for overseeing a country’s tax policy development and implementation. TPU leadership will engage with legislative counsel, parliament’s standing or select committees on public finance, taxpayer associations, and practitioners. To provide objective advice to the minister, TPU leadership should be apolitical.

  • 8. The TPU should be supported by a binding agreement with the revenue administration about accessing sanitized taxpayer data for modeling purposes—while honoring important taxpayer confidentiality protocols enforced by the revenue administration.

  • 9. A newly created TPU should disseminate information publicly, for example, through releasing tax analysis working papers and aggregated tax statistics (for example, revenue structure, sectors’ revenue contributions, and distributional analyses). This will improve the TPU’s public profile and perceptions about its objectivity.

  • 10. Complex analyses may be contracted out to local academic institutions as part of long- term academic backstopping. This should include an internship program managed by the TPU to identify, attract, and ultimately retain competent analysts. TPU management should play a strong mentoring role and encourage life-long training of staff.

The translation of these lessons into operational advice on establishing a TPU requires an assessment of country-specific circumstances, capabilities, and preferences. The IMF stands ready to provide such tailored technical assistance where desired.

How to Establish a Tax Policy Unit
Author: Martin Grote