Public economics and development action an introduction to a special issue in International Tax and Public Finance

Tax, and public-sector matters more generally, is high on the agenda of international development. This is clearly reflected in the Sustainable Development Goals (SDGs) approved by the United Nations General Assembly in September of 2015. SDG17 addresses the need for improving domestic resource mobilization (DRM) directly, and most of the other SDGs cannot be achieved without adequate tax and spending policies. To give just a few examples, SDG10 (reduced inequalities) will depend on government capacity to redistribute income, whereas SDG8 (decent work) requires that tax systems do not create an unnecessarily large burden on economic efficiency. Together these goals reflect the classic efficiency–equity trade-off, which is at the heart of public economics research and policy analysis. Finally, unless all households have sufficient market income (a highly unlikely scenario), the very first SDG (eradicating poverty) requires the presence of social safety nets, which must be financed by public monies. This is the backdrop, against which UNU-WIDER organized a WIDER Development Conference on the theme of “Public Economics for Development” in Maputo, Mozambique, 5–6 July 2017. The conference was wide-ranging, including papers and keynote lectures on all areas of public economics, as applied to developing country contexts. This special issue includes five studies from the conference. In what follows, we first reflect briefly on a set of key issues when researching the public sector in developing countries before summarizing the selected articles.

Tax, and public-sector matters more generally, is high on the agenda of international development. This is clearly reflected in the Sustainable Development Goals (SDGs) approved by the United Nations General Assembly in September of 2015. SDG17 addresses the need for improving domestic resource mobilization (DRM) directly, and most of the other SDGs cannot be achieved without adequate tax and spending policies. To give just a few examples, SDG10 (reduced inequalities) will depend on government capacity to redistribute income, whereas SDG8 (decent work) requires that tax systems do not create an unnecessarily large burden on economic efficiency. Together these goals reflect the classic efficiency-equity trade-off, which is at the heart of public economics research and policy analysis. Finally, unless all households have sufficient market income (a highly unlikely scenario), the very first SDG (eradicating poverty) requires the presence of social safety nets, which must be financed by public monies. This is the backdrop, against which UNU-WIDER organized a WIDER Development Conference on the theme of "Public Economics for Development" in Maputo, Mozambique, 5-6 July 2017. The conference was wide-ranging, including papers and keynote lectures on all areas of public economics, as applied to developing country contexts. This special issue includes five studies from the conference. In what follows, we first reflect briefly on a set of key issues when researching the public sector in developing countries before summarizing the selected articles.

Key issues in public-sector economics for developing countries
First, low-income countries tend to have lower quality institutions, and the constraints the governments face in developing countries are in many ways different from those in their high-income counterparts. This means that public finance solutions that work well in developed countries are not necessarily suitable for developing economies. An example is the progressive income tax system. It is unlikely to work well in contexts where many work outside the modern sector. The implication is that redistribution must rely in large measure on expenditure-side policies. The corporate income tax is also likely to be too complicated a tool vis-à-vis the taxing of small-and medium-sized businesses (SMEs), suggesting that simpler presumptive taxes should be considered instead.
Second, shortcomings in administrative capacity also relate to low fiscal capacity. Although the tax-to-GDP ratio has, on average, increased among low-income countries (See Fig. 1), one can safely state that the current level (around 15 per cent of GDP) remains too low to finance necessary development spending. While increasing the tax take is, on this background, a key goal in the tax and development agenda, such efforts must not create unnecessary distortions on private-sector development.
Third, and closely related to what we said above, tax systems in developing countries need to be designed with the small formal sector in mind. Taxes, almost by definition, fall on the formal sector, and although the tax-to-GDP rate may be low, tax rates on formal employers and employees may in many cases well be quite substantial. Increasing the tax base to cover the present informal sector is, thus, an indispensable step to raise tax revenues in developing countries.
Fourth, a particular feature of developing economies is the phenomenon of the "missing middle" (see e.g. Dharmapala et al. 2011). The notion of the missing middle refers to the empirical regularity that the distribution of firms is dominated by a large share of largely informal small firms and some large (even multinational) companies, whereas the medium-sized firm category is largely missing. This creates two pressures to taxing businesses: One needs to worry about building tax systems that create incentives for small businesses to grow, and at the same time, it is vital to make sure that large, multinational, companies pay a fair share of taxes. The latter goal is particularly challenging. It is widely expected that these companies often engage in tax planning via international profit shifting. Finally, because of the extraordinarily high levels of inequality in many developing countries, distributional concerns are even more pressing than in developed countries. Yet, in addressing these concerns, analysts often face a severe lack of statistical evidence. The UN SDGs called for a data revolution, inter alia because it is common that a clear and detailed picture of the severity of inequality is missing. This is so for many reasons, including that survey-based measures of inequality are highly likely to underestimate the true state of inequality. 1

Studies in this issue
The articles in this issue provide a broad picture of some of the main challenges in tax and development research. The authors rely on different methodologies, ranging from theory via macro-level empirical work to studies using administrative micro-data sets. They cover institutions, inequality, international taxation, and the taxation of SMEs.
The first article "What Determines Administrative Capacity in Developing Countries?", by Ricciuti, Savoia and Sen, 2 sets the scene by examining how to enhance the quality of public-sector management in developing countries. This, if anything, is a top priority for development. Even ample public revenues will not result in effective development action and outcomes if not spent wisely. The authors use data from a recent project on Public Expenditure Financial Accountability (PEFA) and the Polity IV database to explore the long-run determinants of public-sector management capacity. They first establish that the constraints to the executive, as measured by the Polity project, are positively associated with good practices in public-sector management. Using a host of econometric techniques, including instrumental variables regressions, they argue that there is also a causal link from political institutions, such as a stronger system of checks and balances on the executive, on improved practices of public spending. This is an important finding. While progress in changing political institutions is by no means straightforward, it is not impossible either, and policymakers should take note of the finding.
The second study "Inequality, Good Governance and Endemic Corruption", by Epstein and Gang, expands on the theme of institutional quality. They set up a game-theoretic model between two constituencies (the rich and the poor) and the tax administration, in circumstances of high inequality and less-than-perfect admin-istrative capacity. A key characteristic in their framework is the endogeneity of the tax enforcement level, which the two competing groups try to influence. One of the results is the possibility of a poverty trap, an endogenously selected low enforcement state when other, better, equilibria would have been attainable. The paper highlights some of the forces that lead to chosen tax capacity and in this way provides pointers as to how to tackle weak enforcement.
The third article "The Effect of Top Incomes on Inequality in South Africa", by Hundenborn, Woolard, and Jellema, relates as well to inequality, this time to its measurement. As alluded to above, a common worry when measuring top incomes in developing countries is that the surveys used for this purpose may not capture very well the actual incomes of the rich. The novelty this paper offers is the combination of survey data and administrative information, and especially matching the two using recent statistical techniques. The paper demonstrates how to do this in the context of South Africa, and the results reveal that while, quite surprisingly, the survey over-reports incomes of upper middle classes, top incomes are indeed under-reported. Using information from 2 years, the authors then demonstrate that income inequality has fallen slightly in South Africa in recent years.
Turning to international tax issues, the fourth article, entitled "Estimating the Scale of Profit Shifting and Tax Revenue Losses Related to Foreign Direct Investment", by Janský and Palanský, uses macro-level data from the UNU-WIDER Government Revenue Dataset 3 and re-estimates some of the recent, influential, regression estimations used to gauge the extent of international profit shifting. Their results confirm earlier findings that substantial international income shifting takes place across countries. In addition, the extent of the profit shifting appears to harm poorer countries relatively more, when measured as a share of GDP, noting that there could also be other country-specific differences in the relative significance of lost revenues. Finally, the fifth paper "How do Small Firms Respond to Tax Schedule Discontinuities? Evidence from South African Tax Registers", by Boonzaaier, Harju, Matikka, and Pirttilä, examines again taxation of business income, this time focusing on SMEs using administrative taxpayer level data from South Africa. The country operates a special tax regime for small business corporations (SBCs), where the corporate income tax is progressive and has lower rates than the standard 28 per cent CIT rate. Using the bunching estimator, the authors demonstrate sharp reactions of SBCs to the tax incentives, translating into substantial elasticities. The key question is whether these reactions reflect true economic choices (such as those related to investment and employment) or reporting behaviour. The authors argue that the sharp bunching and immediate responses to the change in the tax incentives likely reflect the fact that reporting behaviour is responsible for much of the observed bunching.
Arguably, the articles in this issue speak to some of the key policy matters in public-sector economics for developing countries. We hope that the readers of this issue will find the articles informative and that they will stimulate further research in the challenging area of public economics and development action in developing countries.