Taxation and development: what have we learned from fifty years of research?

Author:Bird, Richard M.

    From one perspective, the answer to the question posed in the title of this article is simple: we have of course learned a great deal about taxation and development over the last half-century. However it is equally obvious that we still have a lot more to learn. Indeed many policy suggestions emerging from research direct our attention to important matters that we do not as yet understand as well as we should. Moreover, even the best research answers to particular questions have usually turned out to be extremely difficult to apply in practice. As two prominent economists recently noted about tax research, "... a result from basic research is relevant for policy only if (a) it is based on economic mechanisms that are empirically relevant and first order to the problem, (b) it is reasonably robust to changes in the modeling assumptions, (c) the policy prescription is implementable (i.e. is socially acceptable and is not too complex) (Diamond and Saez 2011)." Few propositions even in the most recent research literature on taxation in developing countries--see, for example, Stiglitz (2010)--as yet come close to passing these tests.

    The middle ground between grand theory and case studies--the space within which policy-relevant approaches to the tax-development nexus lurk-remains largely undeveloped. The underbrush obscuring the critical political-economic linkages that dominate this middle ground needs to be further explored and cleared out so that we may begin to understand better how to improve tax policy and administration in the very heterogeneous settings found in different developing countries.

    This is not and cannot be a search for some new, simplistic "one size fits all" solution. Over the past fifty years both academic researchers and international institutions--sometimes following ideas suggested by research, sometimes responding to populist fads or pressures from the politically powerful--have issued many policy prescriptions with respect to how to improve economic growth and development in poor countries: increase capital investment; improve education; control population; liberalize trade and capital markets; reduce government controls on market activities; and so on, and on (Easterly 2002). Each of these policies has at times been marketed as a universal "silver bullet" that will result in improved economic performance wherever applied. Unfortunately, none has worked as advertised.

    What might be called the standard approach to tax and development has similarly undergone a number of major model changes over the years, with various stages in between and the extent of 'beta testing' or 'piloting' varying sharply from model to model. Despite the understandable desire of many to find a relatively simple model with which to understand and manipulate complex reality, it seems most unlikely that we shall ever be able to develop some magical fiscal medicine, the swallowing of which will always and everywhere lead to 'the improvement of mankind' (as an historian of thought (Robson 1968) once labeled the fundamental aim of John Stuart Mill's life and work). What this complex and changing world needs is not some non-existent 'universal fix' but rather a sort of fiscal medicine kit containing a variety of remedies and treatments that may help us cope with the wide variety of fiscal problems and needs that arise at different times and often in different ways in different developing countries.

    The aim of this brief article is thus both ambitious and modest. It is ambitious in that it provides a perspective on a half century of work by many people and institutions on many taxation issues in many developing countries and points out some questions that seem to call for more research. But its conclusion is modest in the sense that it emphasizes that even the best research is only one of many inputs in shaping public policy. Indeed, to some extent the problem is not so much to improve research on tax and development as to improve how we market what we learn to those who can, if they wish, put the knowledge to use.

    Consider, for example, the two most important changes in tax structures around the world in the last fifty years: the introduction of the VAT and the general lowering and flattening of statutory income tax rates. (2) Depending on one's political perspective and on how one interprets the economic analysis, each of these changes might be considered in any particular country--or perhaps more generally-to be either a fine example of how good tax research can influence tax policy or a horrible instance of how the rich and powerful can and do employ research selectively to support their personal interests.

    The worldwide downward shift of personal and corporate income tax rates has, for example, been supported both by developments in economic theory and by research results showing the large efficiency costs of high marginal tax rates. Similarly, to some extent at least the widespread adoption of VAT may reflect the research results suggesting that this way of taxing consumption is less economically distorting than most other forms of indirect taxation.

    On the other hand, economic research has seldom if ever been either the main or the only story with respect to either of the changes just mentioned (or, for that matter, any significant tax change). Economic research may provide valuable inputs into policy decisions, both because it is the only approach that focuses tightly on the important question of how efficiently we are using the resources we have to obtain the things we want and because it can-though often economists do not--say some useful things about the distributional outcomes that impact more immediately on policy decisions in most countries. However, as history shows, even the best economic research on taxation in itself never provides either a necessary or sufficient basis on which to develop and implement 'good' fiscal institutions.


    The three decades between the end of World War II and the first oil crisis of the mid-1970s were on the whole years of growth and prosperity in most developed countries. This experience first with successful government-led success in war and then with rapid and generally equalizing market-led growth inevitably shaped the ideas of the few people--mainly economists--who were beginning in the 1950s and 1960s to think about tax issues in the very heterogeneous group of developing countries, many of which were then in the process of emerging from colonial status.

    Most experts, perhaps inevitably, tended to suggest to developing countries looking for (or, at least, receiving) advice on tax matters not just what they thought had worked in their own countries but also what they considered to be even better- one might even say 'optimal'-systems suggested by new theoretical research. At the time, the accepted academic view of good tax policy was, more or less, that the ideal tax was a broad-based personal income tax with progressive rates that included capital gains in the tax base and was integrated with the corporate income tax in order to make both the decision to incorporate and the choice between debt and equity finance more neutral (Auerbach 2010).This concise summary tells the story: it was all about the income tax, at least in English-speaking countries. Indeed, at the time the report of Canada's Royal Commission on Taxation (1966) was called 'a landmark in the annals of taxation' by Harberger (1968) precisely because it was the most detailed attempt to turn these ideas into practical policy recommendations.

    Given this background, it is not surprising that what might be called Development Tax Model 1.0 basically set out a progressive comprehensive personal income tax as the ideal tax for developing, as for developed countries. Indirect consumption taxes were considered at best as a necessary evil, and both the international and sub-national aspects of taxation were generally neglected. Moreover, in line with the prevailing view of 'government as leader' most experts urged not just better (more progressive) but more taxes as necessary to development (Kaldor 1963). For twenty years or more such eminent foreign advisers as Kaldor, Shoup, Musgrave and their followers recommended packages based on this model to, more or less, all comers. (3)

    The outcome of all this advice was not impressive. Although the average tax ratio of developing countries rose from 14.8% of GDP in the 1970s to 17.0% in the 1980s, little of this increase came from the favored income tax and almost none of it from personal income taxes. (4) Indeed, income taxes became relatively less important in most countries as broad-based consumption taxes (usually in the form of VATs) became more important in the tax mix. Interestingly, over this same period, new developments in both theoretical and empirical public finance generally reinforced the lesser emphasis on income taxes and heavier reliance on consumption taxes that had emerged in practice. The result of the parallel developments in both practice and research was the gradual emergence of what might be called Development Tax Model 2.0 as, more or less, the fiscal component of the so-called Washington consensus that ruled the development policy roost after the early 1980s. (5)

    The main pillar of development tax policy in this model-which I have elsewhere called the BBLR (broad base low rate) model (Bird 2011a)-was no longer the personal income tax but the VAT (value-added tax), preferably (according to Model 2.0) imposed at a single rate and on a broad base (Ebrill et al. 2001). One reason for placing VAT at the fiscal center in many countries was the increasing emphasis on lower and more uniform import tariffs as part of trade liberalization. Income taxes, both personal and corporate, were still seen as important but with lower rates, broader bases and fewer incentives. More attention was paid to local...

To continue reading