top of page
  • Marcos Cintra

Livro: Bank transactions: pathway to the single tax ideal - A modern tax technology; the Brazilian experience with a bank transactions tax (1993-2007) (I/III)



The urgent need for a new tax system has been one of the most intensely debated topics on Brazil’s list of pressing domestic issues. Clearly, tax reform can no longer be postponed, since it addresses a major element of the country’s economy and is a decisive factor in determining choices of all economic agents, both public and private.

Over the past fifteen years, debate on the Brazilian tax system has been greatly intensified. During this period several propositions for tax reform have been introduced, creating a clear division of opinions between two schools of thought on the subject. On one side, stands orthodoxy, based on traditional concepts of public finance and on conventional canons of tax law. Some of these concepts and canons have been superseded by the effects of recent technological advances, most importantly, electronic information and new means of asset transfers.

On the other side of the divide is the innovative and anti-dogmatic school of thought which proposes the elimination of conventional tax models and which is epitomized by the resurgence of the age-old concept of the Single Tax, which, in its modern version makes extensive use of non-declaratory taxes and of electronic technology.

The first school of thought – associated with the use of conventional declaratory taxes – believes that “old taxes are good taxes”. This school mistakenly sustains (so claim such conservative reformers) the continuation of paradigms which, inadvertently to them, have become outdated, and which have been superseded by the peculiar impacts of modern economies, characterized by globalization and by the overwhelming effects of the digital information age. The great Brazilian economist, diplomat, and public figure Roberto Campos, an active participant in the tax reform debate, once stated that that to defend this school of thought is to engage in a melancholic and poorly disguised exercise of trying “to perfect the obsolete.”

The second school of thought calls for the elimination of declaratory taxes and for their substitution by electronic taxes operating through the bits and bytes of the data-processing centers and clearinghouses of the banking system, such as a bank transactions tax. Traditionalists call this an audacious proposition, bordering on illusion. Despite the proven capacity of such taxes to generate impressive amounts of revenue and to show an almost universal pattern of incidence and coverage, researchers and defenders of this school of thought usually draw the wrath of traditionalists who oppose it. The guardians of orthodoxy, the bureaucratic establishment, and the recurrent tax evaders refuse to relinquish their decades-old professional and intellectual investments – despite the fact that all evidence proves them increasingly obsolete.

In this book, the terms “declaratory” and “non-declaratory” tax are used here after to express the distinguishing features of conventional versus non-conventional taxes. Conventional taxes, in current use throughout the world, make extensive use, by the taxpayer, of self-prepared filing of paper tax-returns based on tax accounting procedures set up by tax authorities. The non-conventional, non-declaratory taxes make no use of paper tax filings, and are usually collected automatically, administratively, by electronic means, such as the bank transaction tax in use in Brazil from 1993 to 2007.

The clash between these two tax paradigms, the declaratory versus the nondeclaratory tax system (which might be rephrased as the “with” versus the “without” paperwork tax system) draws to the surface questions concerning not only the deep changes that are occurring within the modern world economic environment, but also the academic posture of taxation (and even of public finance) as a science.

Thomas Kuhn says that a field of study becomes a science when a community of experts consensually accepts a paradigm – that is, a set of problems and uniform standards of approach – with a foundational theory and a common set of explanatory and interpretative traditions. “The authority of a scientific proposition is founded on its capacity to generate consensus within a given community. This consensus, for its part, does not depend on whether the scientific propositions provide an indisputable vision of the intimate configuration of reality. It does, however, depend on whether its development has been guided by demarcation criteria that are authoritatively prevalent in the environs of that community.” Kuhn goes on to state that, “It is for this reason that paradigms distinguish themselves by their incommensurability. If each paradigm sets forth the conditions of the scientific nature of the knowledge produced in its environs, the proofs invoked in favor of other paradigms tend to be disqualified a priori.” In other words, a proposal that contradicts “conventional wisdom”, paraphrasing John Kenneth Galbraith, is summarily considered “unscientific”, not because of lack of objective analysis of its scientism, but simply because it does not apply methods and models considered “correct”, “truthful” or “evident”.

This perspective of conceptual advances within the domain of science (which can be applied, mutatis mutandis, to life in society) allows us, furthermore, to understand the defensive arguments of “traditionalists,” who tend to reject the singletax proposal on the pretext that, “if it were good, it would have already been adopted by more advanced economies.” This sad argument acknowledges the inertial weight of entrenched concepts of tax systems or, inversely, it ignores the revolution that electronic technology has inspired in some countries, but not in all of them. For example, Brazil has a banking system that is significantly more modern than that of most of the advanced economies, including the United States, and this is the foundation that supports the paradigm shift towards the single-tax concept. Furthermore, such an attitude ignores that there are cultural, social, political, and economic differences among countries that make some urgently need a new tax system, while others do not, as least not with the same intensity.

Regarding tax systems, it becomes increasingly evident that the conventional paradigm is gradually becoming exhausted. In tax matters, the conventional paradigm is following the steps described by Thomas Kuhn to justify a “scientific revolution”: old beliefs become less capable of providing answers to concrete problems, and for each solved problem others appear of even greater complexity.

An illustrative example is found in the changing perception of tax administrators regarding the Income Tax. After the Second World War the global income tax became almost universally used. “This tax was an ideal instrument for the time and came to be seen by many policymakers and tax experts as a “dream tax”. In the United States, 90 per cent of taxpayers had considered the income tax as a fair tax during World War Two, according to survey data published by the American Enterprise Institute (2005)”. Nevertheless, perceptions about this form of taxation are gradually changing because of new circumstances present in the world, but also because of some characteristics of the income tax which were persistently ignored by policy makers. “It was considered an efficient tax because most economists dismissed its potential negative effects on work effort and incentives. Few academic articles, if any, dealt with these potential disincentives. Furthermore, though it now seems strange, books on income taxation did not even mention ´tax evasion´ or ´the underground economy´ as potential problems associated with income taxes”.

The conventional tax paradigm faces a serious crisis due to its incapacity to provide explanations, diagnostics, justifications, and solutions to new facts and circumstances that are rising on the contemporary economic scenario. Indeed, what we see is the erosion of traditional mechanisms of tax collection. Such mechanisms are based on the notion that the taxpayer is a potential defrauder, until proven otherwise, and this has led to the creation of a significant number of control, inspection, auditing and surveillance systems that turn out to be expensive, complex, and highly bureaucratic, but nonetheless, incapable of preventing tax evasion.

In truth, the outcome of this debate tends to become more predictable, insofar as two fundamental phenomena of modern history will impose their inevitable consequences in favor of the non-declaratory system. These two phenomena are: first, the technological revolution of the information age; and secondly (but no less important), the current globalization of world economic relations.

The information age has profoundly altered the aggregate production function of modern economies. Decision-making has been greatly streamlined by the increasing number of methods for processing massive amounts of information. Data collection and analysis have improved through increased sophistication in electronic processing. The supply and control of massive amounts of information have become key decision-making inputs for modern businesses. Furthermore, the use of paper currency is being steadily replaced by electronic money; the concept of wealth and money is being constantly redefined. These changes bring into stark relief the precariousness of tax reporting and the handicraft mechanisms used in conventional tax systems, which, historically, were developed in response to the technological and organizational environment that existed immediately following the industrial revolution.

Furthermore, growth of the service sector’s share of GDP has significantly reduced the effectiveness of the tax collection, auditing, and control mechanisms currently in use. The productive sector has become ever more intangible and dematerialized, and this has only stressed the dwindling effectiveness of conventional mechanisms for tax assessment and enforcement. In fact, intangible services traded over the Internet (as for example, new accounting software, with high initial production cost but currently reduced to bits and bytes for delivery and utilization) are actually beyond the reach of tax authorities, kept outside the realm of such type of exchange. It becomes increasingly more difficult to levy specific taxes on trade of either products or services if the resulting payment transaction takes place in a tax haven, where no specified origin or destination of any good or service can be readily identified. At that moment a non-declaratory tax, such as a bank transaction tax, begins to make sense because it is levied on that agent’s banking activity and not on its reported accounting statements.

Traditional tax models assume that production, and its resulting taxable income, is carried out through manual production processes (or later, through mechanical production processes) concentrated within finite geographical spaces, centered in organizational structures that are autonomous, independent, and subject to domestic rules established by a sovereign State. This is the world of the industrial revolution, later modified by mass production, where production and exchange are strategically concentrated on a relatively small number of large national corporations. Tax assessment and enforcement have, therefore, to be directed and adjusted to that reality.

But that kind of a world is swiftly dying, a fact readily visible to anyone versed in the realities of world globalization.

Historically, the entire universe of individuals and businesses, of all sizes and in all sectors, soon became subjected to the obligation to pay taxes. Levying taxes across the board greatly expanded the pool of taxpayers. Whereas the taxpayers’ universe had previously consisted of those few large units of production and exchange that typified the early stages of the industrial economy, soon it began to encompass all businesses and individuals in modern societies. Tax collection, assessment, and control functions now demand operations on a scale wholly incompatible with the declaratory, bureaucratic, paper filing systems typical of the traditional tax method of “self-assessment, self-levying, self-collection, and public audits,” which typify conventional declaratory tax systems.

The electronic revolution provided an indispensable new instrument for collecting and analyzing the enormous mass of data and information needed for tax control, monitoring, and collection processes. But such technological change is not restricted to a mere increase in speed of data processing within the old tax paradigm, although it has been successfully serving this purpose. Now it becomes possible to underscore the creation of a new tax paradigm, of a new tax species, such as payment taxes, which were never possible before.

Brazil’s current banking and payment systems are among the most modern known in the world, and this enables them to bring about such paradigm shift. The importance of the information age is not limited to being an auxiliary method for controlling, auditing, and analyzing tax data. Its importance extends beyond this, as it became a determining factor in the conceptualization of new taxation models, primarily in configuring new tax bases, such as bank transactions, electronic flows, telephone pulses, electronic wavelengths, and other intangible bases, which are impossible to be reached by conventional taxes.


A second factor to demand deep changes in conventional tax models is globalization. This is a multi-faceted, complex element, which is having a strong impact on economic and social life of humankind. According to José Eduardo Faria, globalization has been responsible for the “relativity of several important concepts, principles, and categories – such as sovereignty, legality, the hierarchy of laws, subjective rights, formal equality, citizenship, balance of powers, security, and certainty – that have been heavily affected by economic, social, political, and cultural changes that have taken place largely apart from legal structures, judicial mechanisms, institutional structures, democratic procedures, and the capacity for regulation, control, management, direction, planning, and concession-granting of nation-states.”

Indeed, globalization has weakened the power of national public administrations, by decentralizing and fragmenting the decision-making capacity of traditional governments. Even more visibly, it has “debilitated the taxation and regulation capacity of governments.” José Eduardo Faria argues that, “within this highly unstable scenario, positive law... came to face a cruel dilemma: if it remains concerned with its logical integrity and with its formal rationality, in view of all these profound and intense changes, it runs the risk of not accompanying the dynamics of facts, of becoming functionally ineffective and, ultimately, socially discredited, ignored, and (in the worst case) even disposable. If it allows itself to be seduced into attempting directly to control and discipline all sectors of social, economic, and political life that are increasingly tense, unstable, unpredictable, heterogeneous, and complex... it runs the risk of becoming disfigured as a normative reference.”

The divorce between the conceptual foundations of government that emerged from the post-war period, and the realities of modern world globalization brings out what José Eduardo Faria called the “systemic ungovernability” of the traditional State. This begs the question: to what extent are traditional taxation models assimilating this new reality, marked so deeply by the information revolution and by intensive globalization?

The traditional tax system presupposes that the taxpayer is a nuclear firm (the same principle applies to individuals) that produces tangible goods with one or not more than a few physical facilities concentrated within a single national State (or tax territory), surrounded by suppliers and buyers that have the same basic characteristics. ILLUSTRATION 1 above describes this situation. In this system, it is easy to assess the taxpaying capacity of the nuclear company. It is also simple to enforce tax regulations by cross-referencing data with peripheral supplier and buyer companies, or individuals.

The situation is radically different, however, if the operational strategies of these businesses are executed through decentralized networks that spread across several nation-states, producing both tangible goods and, ever more frequently, services, as shown in ILLUSTRATION 2. By their very nature, services are intangible, highly mobile and easily transported through electronic media. The illustration below shows the operational complexity of these businesses operating globally, involving federal and external variables, international trade and logistics, cross-ownership of investment capital, fast technological change and market-share strategies. These factors imply the need for increased sophistication of concepts and of operational methods that are not adequately addressed by conventional tax models.


Examples of such challenges to the conventional tax paradigm are the growing incapacity of nation-states to deal with problems created by “tax havens”, by increasingly complex means for laundering money, and by the uncontrolled flows of foreign funds between companies belonging to a single global conglomerate (transfer-prices).

In testimony before the Special Commission on Cumulative Taxation in theChamber of Deputies, on April 2, 2002, the Secretary of the Federal Revenue, Everardo Maciel, stated that: “the extraordinary changes occurring throughout the world also explain the large differences occurring between nations. One of these changes is globalization, which has brought very intense transformations. It is important to remember that one-third of foreign trade takes place between companies; another third is comprised of transactions within multinational corporations(…) These factors demonstrate the growing importance of these multinational firms, which raise a modern and worrisome question about the future of the corporate income tax or about the so called “ transfer price” (…) Today, some countries assert the existence of transfer-prices point to the unlikely survival of taxes such as the corporate income tax in the future. Comments frequently made in the press, in international seminars, and in international tax conferences point to this fact as something new, which calls for a review of traditional tax models, most of which are of Anglo-Saxon extraction.

Globalization has, therefore, significantly changed the social, political, and economic environment in which tax systems must operate. The main changes have been the extraordinary growth in international trade of goods and services, increased mobility of labor and capital, and growth of multinational, transnational and international companies. Tax administrators nowadays speak of taxation on world bases. Tax competition between countries has mushroomed. Unfortunately, such changes have gone in the direction of increasing complexity, interdependence and fiscal competition between countries. “Tax termites”15, such as electronic and internet commerce, plastic and electronic money, transfer pricing, tax havens, foreign shopping, and complex financial instruments have contributed to decrease the revenue raising efficiency of national governments. “The work of ´fiscal termites´ (is) busily gnawing at the foundations of the tax systems” 16  Firms and people do not hesitate to abandon countries where they are located to seek any point on the planet that offers less progressive and lower taxes.

Tanzi believes that the effect of the fiscal termites in national economies is to decrease tax revenues. In fact, this has not been occurring. The tax burden has been increasing worldwide, but at the cost of tax shifting and increasing burden on less mobile taxpayers, such as wage earners and producers of non-tradables, worsening the domestic patterns of incidence and equity.

The perplexity facing tax administrators when confronted with such difficulties in preserving their national taxing capacity has led to a twofold solution: one, is objectively trying to typify each possible problem or situation (which is obviously impossible to enumerate and extremely costly to operate). The problem is becoming so acute to the point of motivating governments, especially in the European Union, to discuss the creation of a super national layer of global government, capable of coordinating, or more appropriately, of attenuating, through unconditional or supervised delegation, the tensions and stresses that are evolved in international tax relations among national states.

The second alternative is to endow tax authorities with subjective power to analyze each situation on a case by case basis, as they arise, and thereby decide what should be considered legitimate tax planning and what should be considered an illegitimate “legal” form of evasion.

If the first line of conduct implies high compliance and administrative costs due to the mushrooming bureaucracy that would probably result from it, the second alternative would imply juridical insecurity and potentially mistaken or arbitrary judgments.

Needless to say that such “solutions” may greatly increase the compliance and administrative costs of tax systems throughout the world, which, in turn, could induce the growth of evasion and of the informal economy. Thus, tax evasion and the flight toward the underground economy would further reduce the taxing capacity of national governments.

Edgar Feige, a pioneer in the study and measurement of the underground economy, coined the term, tax revolt, stating that: “the irregular economy appears to have little respect for conventional geopolitical boundaries. Indeed, it is being increasingly noticed in almost all developed societies.” Feige says further: “I wish to note that I began this investigation suspecting that the irregular economy was smaller than previous estimates had suggested. I am now convinced that the irregular economy is indeed of staggering proportions and growing rapidly.”

Tax reformers in a country like Brazil should not become prisoners of conventional wisdom, nor be restricted to old tax models which have surpassed their useful lives. A country’s tax system must be able to adapt to the dynamics of the modern economic world. Taxation falls on ever changing economic bases, and not on consolidated juridical facts . Tax reform, therefore, should allow for enough flexibility and realism to be able to adjust itself to a society’s environment, and to its social, economic, political, and cultural characteristics.

“Since around 1980, the annoyance of taxpayers worldwide has been directed with increasing intensity not only at the high levels of taxation, but also at the complexity and instability of the tax systems. This annoyance has become a major factor in the changing attitudes of citizens towards taxation recorded in many countries during that period. In addition therefore to the level of taxation, such issues as complexity, instability and fairness of the tax systems have become important in many countries…instability, inefficiency and absence of fiscal coherence have characterized the tax systems.”

Such dissatisfaction with conventional tax systems, which are still being used extensively around the word, cause even more amazement as they still find economists who strongly uphold them, despite all evidence to the contrary. The amazement at this state of affairs is precisely described by the following statement:

“no one would design such a system on purpose and nobody did. Only an historical explanation of how it came about can be offered as justification. That is not a justification, but a demonstration of how seemingly individually rational decisions can have absurd effects in aggregate”.

In other words, citizens and policy makers are in search of a new “tax technology”, paraphrasing Vito Tanzi. Maybe he is foreseeing the future, although with a certain bias in emphasis, when he mentions that “the discovery of value-added taxes in the 1950´s and its widespread use in later years must be considered the most important technological development in taxation in the past 50 years. [But also] … gross assets taxes and taxes on financial transactions have been less important technological developments in Latin America.”

The first part of his statement is gradually becoming less true, although in the past it has certainly helped to improve tax systems in the world. The second part, however, is becoming an increasingly crucial technological development in taxation, as will be demonstrated in this text.

Unfortunately, “the first law of finance is inertia”, as we are painfully reminded by Prof. Richard Bird. “It is surprising that the many governments in the world, most of which are trying to raise more revenues, have not come up with more ingenious ways of doing so. The lure of the familiar and the apparent desire of most governments- like most people- not to be the first to do anything new doubtlessly account for the relative lack of fiscal innovation in the last 50 years.… For the most part, however, a first lesson suggested by history is that the fiscal problems of the next 50 years will probably have to be dealt with using taxes very much like those on hand today. As with most social and political institutions, there seems to be little or no chance of a quick technological fix.”

Prof. Joseph Stiglitz seems equally skeptical about this issue when he states that “I do not see that any likely changes in technology in the near future will have a revolutionary effect on the design of our tax system”.

In spite of the impressive weight of such opinions, we hope this text will confirm Vito Tanzi´s remarks on the technological significance of both the electronic age and of its offspring, the financial transaction taxes, in constructing future tax systems in the world.



The publication of the article entitled “For a tax revolution”, in the Folha de São Paulo, in January 1990, was a turning point in the debate on tax reform in Brazil. The article introduced the Single Tax on Bank Transactions. I proposed that current paper-ridden (declaratory) taxes be replaced by a single paperless (non-declaratory) bank transaction tax.


The Single Tax is a centuries-old idea. If first appeared in the 18th century when the physiocrats argued for taxation of land as the sole source of government revenue. In the 19th century, Canada and the United States also discussed similar ideas. France discussed a single tax in the post-war period, and in the 1990’s, in Brazil this same proposal reappeared in a new format, as a bank transaction tax.

Historically, the difficulty (and the recurrent failures) involved in applying single tax proposals has been to find a tax base that, by itself, is broad enough to generate sufficient revenue for the government without requiring that tax rates be so high as to stimulate evasion

The modern concept of bank transactions meets this requirement.

Several countries have implemented a tax on bank transactions, such as Argentina, Colombia, Venezuela and Australia, among others, although not as a single levy, as has been proposed in Brazil. As a regulatory tax on the financial markets, with the purpose of slowing down the flow of speculative money, both internally and on an international scale, transaction taxes have either been used, or are being considered for use, in a large number of countries around the world, such as India, Australia, Austria, France, Germany, Italy, Japan, Switzerland, UK, USA, Malaysia, New Zealand, Singapore and many others.

The financial, or bank transaction was first noticed as a potential tax base after the advent of digital currency, as it began to replace paper currency, and as banking transactions began to be processed electronically. The success in the search for such a single tax base and the solidification of an economic system based on digital money made possible, and maybe inevitable, the birth of the Single Tax proposal in Brazil.

Though the Single Tax has a long and respectable tradition in the evolution of economic thought, it has never come to fruition because, before the growth of digital transactions, no society in history had ever satisfied two basic conditions necessary for its effective implementation, which modern bank transactions make possible. The first condition is a highly digitized and technologically sophisticated banking sector with a nationwide system for clearing checks and other documents. The second condition is a cultural predisposition within a society not to use paper currency, replacing it with digital currency. In other words, digital bank transactions made possible the birth of a broad tax base, large enough to generate enough revenue to support the public sector of modern societies.

Brazil is the only country that fully meets these two requirements. It has one of the most advanced and digitized banking systems in the world, with technological standards that surpasses those found in developed countries such as the United States and the European Union. In truth, “Brazil is the banking benchmark of the world”. Further, it is one of world’s economies that is least dependent on paper currency. And culturally, it has already absorbed the unavoidable replacement of paper currency by fiduciary money, such as checks, electronic debit and credit cards, internet and email transactions, and other forms of digital currency. The Single Tax proposal stresses two fundamental points.

First, it reduces all taxes to one single tax. All others would be extinguished, except the extra-fiscal taxes such as customs fees and other non-revenue taxes used as instruments of economic policy by the government. No longer would there be an individual or corporate income tax, or a sales tax such as the ICMS (a state valueadded tax), or the ISS (a municipal turnover tax on services). Wages would not be subjected to any withholdings whatsoever, either as an advance toward income tax or to finance social security. Corporations and individuals would no longer have to file tax returns of any kind.

Secondly, the fundamental premise of this proposal lies in transferring the tax base exclusively to banking transactions, ending the multiplicity of tax bases that exist today. Every time an economic agent makes a payment through the banking system there will be a tax incidence assessed to the value of the transaction. The tax will be divided into equal parts and charged both to the issuer and the beneficiary of the payment. And, importantly, it will not be charged to transactions in the financial and capital markets.


The impact of the Single Tax on Bank Transactions model has triggered a nationwide movement to reform Brazil’s tax structure. Those who favor a paper-free tax system embrace the Single Tax proposal, whereas defenders of paper-driven taxes discredit it, stressing its undesirable cumulativeness.

The single tax has countless advantages as a taxation system. Auditing becomes simpler; taxation criteria are more transparent; bureaucratic and compliance costs both to the public and to the private sectors are lessened. The simplification of the fiscal process becomes evident when all revenue is concentrated in a single tax, levied on a single tax base. Public administration costs decrease.

Only recently have economists and public officials begun to estimate auditing and other administrative costs related to tax collection in Brazil. The results of such studies are leading to important conclusions about the advantages and disadvantages of alternative tax models.

In the United States, federal tax collection costs equal 0.5% of revenue. For personal income tax, the compliance costs for individual taxpayers represent from 5 to 7% of the revenue raised by the federal and state income tax systems combined. Administrative tax costs in the United States are estimated at 1.13% of revenue.

Compliance costs related to sales tax are estimated to be 3.93% of revenue. In 1986, the cost of fiscal administration in France was 3 to 4% of revenue, or 1.5% of GDP, not including private compliance costs. Data from research conducted in other countries and reported at the International Fiscal Association Conference in Rio de Janeiro in 1988 are reproduced in TABLE 1.


In Brazil, tax administration costs to the government are probably much higher, not only because of the inefficiency of the tax collection apparatus, but also because of the multiplicity of fiscal obligations to which individuals and corporations are subjected. Add to these the costs of tax reporting to which private agents are subjected in Brazil, and it is no exaggeration to state that total costs can be as high as 20% of tax revenue. This is unproductive effort, which translates solely into expenditures, without in any way contributing to increases in production or social well-being.

It is worthwhile noting the statements made by former Secretary of Federal Revenue, Everardo Maciel, while testifying before the Comissão Parlamentar de Inquérito [Parliamentary Inquiry Committee] (CPI) on May 8, 2002. The Secretary’s sympathy for the CPMF (a bank transactions tax used in Brazil since 1996 and which would be the hegemonic tax in the single tax model) is noteworthy. He said, “my presence here is solely to quickly state for the record that the bank debit transaction tax (CPMF) has been an extremely valuable instrument from a revenue collection standpoint, precisely because it manages to produce public revenue at low cost, with extreme efficiency, and, additionally, serves primarily as an auxiliary instrument for tax auditing.”

Nevertheless, when asked about the Single Tax model, Secretary Maciel stated, “Even if I were totally favorable that the CPMF be converted into a permanent tax, I recognize nonetheless that if we go to the trough too eagerly, that is, if its tax rates increases, we could begin to induce ever more sophisticated, ever more elaborate tax evasion procedures. My experience tells me that anytime the rate increases, anytime fiscal pressure turns heavy, taxpayers will seek ways to free themselves of it, and usually through tax evasion. The second point, and Deputy Marcos Cintra knows this, I do not believe that the Single Tax is the best solution for the tax system. We have a large cast of alternatives and options. Rest assured, your Excellencies that every time we build a tax system around a single point, taxpayers will try to run away from that point; they will try to find a way to dodge it. So, we must always have somewhere else to go; if we do not reach it through this avenue, we will reach it through another. And that has been the history of taxes throughout the world; this is how tax theories developed. But I think that, today, the bank debit transaction tax (CPMF) occupies a place of capital importance, a place of distinction in tax theory, especially for taxation in countries that have weak tax collection traditions, as is the case in Brazil.”

Though Secretary Everardo Maciel argues for the permanence of a bank transactions tax, the fear of possible evasion blocks him from fully supporting the Single Tax. Ironically, the bank transaction tax has been showing strong evidence to be capable of minimizing, if not eliminating, tax avoidance in Brazil.

It is also worth noting an opinion that claims that criticisms of the Single Tax are born of “...small and easily correctible details which are enumerated in order to bombard the most brilliant idea, I dare say, that has ever arisen on tax matters in modern times, in the era of financial capitalism, not by the fact that it is single, but primarily because of the characteristics of the tax, which is practically impervious to evasion. Therein resides the fear of its creation.”


Less than three years after the Single Tax proposal was published, it was quickly misused by the government into becoming one more tax laid atop the many others already existing in Brazil.

Indeed, ignoring the single tax philosophy, in 1993 the government proposed the creation of a bank transaction tax (the IPMF - Provisional Bank Debit Tax - instituted that year), initially for the purpose of helping eliminate the public deficit, and later with the express purpose of financing health expenditures (called CPMF, created in 1996). The Brazilian stabilization plan (known as the Real Plan) put an end to the inflationary tax, and the federal government chose to increase tax revenue in order to balance the budget. President Fernando Henrique Cardoso’s fiscal policy was responsible for raising tax revenue from 27.9% of GDP in 1994 to almost 32% of GDP in 2001 (in 2008 it reached 36% of GDP). The CPMF greatly contributed to this result, raising revenues at an extremely low collection cost to the Government.

Criticism soon began to be heard, claiming that such a tax would be harmful to domestic production, especially due to its cumulativeness. Even though historical facts proved false many catastrophic prophecies attributed to a tax on bank transactions, such belief persists up to the present moment, to the point of making public opinion believe that the main distortion that must be eradicated from Brazil’s tax system is the cumulativeness of taxes such as the CPMF.

To rebut such argument I published an article in Folha de São Paulo [the most important newspaper in the country] in which I defended the CPMF as an efficient tax mechanism, despite its misuse by the Government.

The CPMF achieved several intents: to guarantee fiscal equilibrium, to fund public health expenditures, to capitalize an anti-poverty fund, to detect tax evaders, and to finance the impact of an increase in the minimum wage. Nevertheless, its opponents persistently refuse to acknowledge its qualities, such as bringing relief to taxpayers from necessary increases in other taxes. After all, if the CPMF did not exist, conventional taxes, which are almost always inefficient and inequitable, would necessarily have higher rates, adding to the heavy burden already borne by taxpayers.

The CPMF, which is the backbone of the Single Tax proposal, is a revenueeffective tax. Despite the fact that some economic distortions can be attributed to it, its advantages far outweigh its disadvantages.

The usual objection to the CPMF has to do with the harmful impacts of its cumulativeness on the financial markets and on foreign trade. This is an undeserved criticism. There is nothing to prevent the government from exempting exports, from taxing imports in identical conditions as those applied to domestic products, and from exempting the turnover in financial and capital markets from taxation. It is worth mentioning that government authorities have been making considerable efforts to improve tax exemption mechanisms for exports and to levy an Economic Equalization Contribution on imports in order to remove any tax discrimination against domestic production.

The CPMF has one undeniable advantage, conveniently ignored by several of its critics: it eliminates the greatest anomaly present in the current Brazilian tax system, namely, the artificial differences in production costs caused by widespread tax evasion. Because the Brazilian tax system provides generous possibilities for evasion, the pattern of tax incidence becomes extremely uneven, leading to even more serious economic distortion than the alleged changes in relative prices caused by turnover taxes such as the CPMF. In fact, by making evasion practically impossible, the CPMF attenuates this serious distortion, as will be shown in the next chapter.

Nevertheless, it irritates and enrages powerful interest groups because of this evasion-proof form of operation. For the CPMF, the cost of tax avoidance usually exceeds tax savings. This is the greatest advantage of this type of non-declaratory tax. Because of its evasion-free characteristic, it allows for low rates.

This tax also displeases tax collectors, tax accountants, and attorneys who both defend and prosecute tax evaders, as it makes their intervention in the fiscal process unnecessary. This type of tax reduces costs, eliminates corruption, and results in a pattern of incidence exactly proportional to the volume of financial transactions performed by taxpayers. By doing so, those who presently are disproportionately overtaxed, such as payroll wage earners will be taxed less, heavily while tax evaders will pay their share. This is the essence of the Single Tax proposal.

The issue of cumulativeness is an easily refutable criticism raised against this type of electronic tax, as will be shown later. What must be stressed is that the main objective of tax reform in Brazil is to eliminate the main source of strong economic distortions in Brazil, which is tax avoidance and tax evasion, which are encouraged by the conventional taxes that make up its current tax system. Furthermore, the increasing complexity and irrationality the Government has introduced into the Brazilian tax system in the last decades encourages informality and other forms of tax avoidance, producing corruption and an unfair pattern of tax incidence.


Productivity and competitiveness

A BNDES paper states that cumulative taxes such as a bank transaction tax “are easier to collect and pay.”, whereas valued added taxes are “more complex to calculate and even to comprehend.”

Arguing their opposition to cumulative taxes, the authors list two of their undesirable characteristics, supposedly inexistent in VATs. They say cumulative taxes “are most damaging to the competitiveness of domestic production because of the difficulty in exempting their incidence on exported goods and because of the advantage they grant to imports, which usually are not subjected to the same treatment in the country of origin."

Concerning this observation, it is interesting to note the reaction of Professor J. A. Scheinkman when invited to lecture on trade competitiveness and tax harmonization in Brazil. He said, “Competitiveness is a notion that does not make sense for a country as a whole. All countries have greater or lesser competitiveness in different products.” He adds, “The idea that the tax system… affects competitiveness, as I see it, does not make sense.”

Professor Scheinkman demonstrates that tax evasion and the informal economy are factors that depress an economy’s productivity. If a tax system induces high rates of tax evasion and avoidance, productivity loses its correlation with investments in technology, or with administrative and managerial efficiency. A company that has low production costs may be less “competitive” when compared to a company that evades taxes, even if the tax evader has significantly higher cost of production. This causes inefficient companies to survive and depresses a country’s economic productivity. Because Brazil’s tax system encourages tax evasion and the flight to the informal economy, it “depresses productivity in a very significant way”. We see, therefore, that “national competitiveness” is not hurt by cumulativeness, but rather by a tax system that induces tax evasion, as usually occurs when conventional declaratory taxes are employed.

He adds, “The need for tax reform has nothing to do with matters of the country’s integration into a trade bloc,” and, “we need a tax reform that is taken seriously, that lowers the high rates prevalent in Brazil which make people simply avoid and evade taxes.”

In other words, removing cumulative taxes will not increase the economy’s productivity and competitiveness. Their elimination will result in the need for higher rates of conventional taxes in order to keep revenues constant and, therefore will lead to increased tax evasion. The great villain of the current tax system is not cumulativeness per se, but rather tax evasion that results from the complexity and high rates inherent in current declaratory tax models.

It should also be noted that adequate tax policy can fully remove the “disadvantages” of cumulative taxes pointed out in the BNDES paper mentioned above. In fact, the tax reform the Government announced in July 2001 moved in exactly these two directions; that is, zero-rating for exports and the creation of a bank transaction tax on imported goods and services. The objective of these measures was to guarantee absolute isonomy between domestic and foreign producers, which redresses the two criticisms of the CPMF presented by the authors of the BNDES paper.

Allocative Efficiency

Cumulative taxes are often criticized on the basis of comparisons with valueadded taxes. In general tax analysts follow the usual text-book conclusions that make extensive use of optimal tax theory in reaching normative conclusions about their respective impacts on allocative efficiency. Such conclusions, however, are fragile to the extent that such theoretical work depends heavily on strong assumptions, which are seldom, or never, found in the real world.

Good economic analysis requires that each type of tax be evaluated not only for its intrinsic characteristics, but must also take account of the empirical circumstances surrounding its application. Failure to consider these circumstances, coupled with a naïve, automatic and uncontested acceptance of the simplifying hypothesis found in the theoretical compendia of public finance, implies running the risk of making gross mistakes. Such is the case when the VAT is discussed in Brazil.

One advantage claimed on behalf of VATs is that they cause fewer distortions in relative prices than would be caused by cumulative taxes. However, for this statement to be true, one must accept the premise that perfectly competitive markets exist, such as assumed in conventional optimal tax theories, based on excess-burden analytics.

We know, however, that such a hypothesis is essentially heuristic and that, in practice, markets do not meet the criteria needed to be considered perfect. As stated by the Federal Revenue Service “the superiority of value-added taxes in terms of distortionary impacts is readily recognized in the case of easily administered tax systems that are nationally harmonized, with low evasion rates…and with one or two tax rates. However, given the actual restrictions, and considering that the ideal situation cannot be easily reached in the short run, it is wise to adjust the debate and avoid making decisions which can hurt the system”.

In an interesting paper that seeks to establish normative conclusions on the allocative impact of different taxes, Cláudia and Ibrahim Eris use Leontief’s linear model in seeking tax policy guidelines. They write, “the task of ranking taxes as ‘better’ or ‘worse’ is very complex, even in simple models such as the one adopted above; and the literature on this matter has been reduced to a few scarce studies that are sometimes even erroneous.” The authors conclude by saying, “truthfully, the world of Public Finance is a second-best world, and as such the traditional graphs of utility frontiers are often irrelevant, because the system’s distortions place the economy at a point below such frontiers. The upward movement of the utility frontier says nothing about the point (below the frontier) in which the economy finds itself.”

On the validity of policy prescriptions of optimal tax theory it is worth quoting Frank Hahn, who says “…while these studies have increased our understanding of what is involved, the tax formulas which they contain cannot be taken very seriously…Welfare economics is the grammar of arguments about policy, not the policy.”

On this same line of thought Sandmo states that “The theory obviously has its limitations. It is at its best in yielding rules for the optimal structuring of a given tax system and has less to contribute to the discussion of major problems of tax reform, which typically involves the choice between alternative tax systems. A difficulty with the extension of the theory to cover these global problems is that the costs of administration have not been incorporated into the theory; this is one aspect of the neglect of transactions costs in the theory of general equilibrium….This raises the question of whether optimum tax formulae can have any claim to be taken seriously, given that they abstract from such central concerns as administrative costs and incomplete information….it may well be that we shall find the models of optimal taxation to be useful, even though we may have to supplement them with considerations which are exogenous to the models themselves.

Indeed, welfare theory demonstrates that society will not choose a point which is allocatively efficient if, compared to another situation (even if inefficient), it is capable of reaching a superior point in its social welfare function. In other words, even though ideally the VATs may theoretically introduce fewer distortions in relative prices, it is possible that cumulative taxes would be preferable if, for example, it can be proven that it decreases tax evasion, or that it requires a lower nominal tax rate to collect a given amount of revenue, and as a result the pattern of tax incidence is considered more acceptable to society, as demonstrated below.

It is generally assumed that efficiency is the only criterion for choosing a particular allocation of resources. However, even assuming a perfectly competitive economy, one may not state that a Pareto-efficient allocative situation resulting from such market configuration will necessarily maximize social welfare.

The implication of this statement is that one cannot guarantee that the use of a neutral tax, such as the VAT is assumed to be (even though in fact it may not always be so), will necessarily maximize the social utility function. Distributive considerations can make possible the attainment of a higher point in the social welfare function of a society through the use of a cumulative tax. In other words, from the standpoint of maximizing the social welfare function of an economy, the use of a tax that is non-neutral and “inefficient” from an allocative standpoint, and thus configuring a Pareto-inefficient situation, may be preferable to a Pareto-efficient position resulting from the use of a neutral tax. This possibility demonstrates the error contained is statements to the effect that VATs are necessarily better and always preferable to cumulative taxation.

A resource allocation is Pareto-efficient if, in order to improve one person’s position it is necessary to worsen the situation of at least one other individual. A situation is deemed Pareto-superior if, it is possible to improve one person’s initial welfare state without diminishing another’s.

The Fundamental Theorem of Social Welfare Economics says that if producers and consumers act competitively, strictly as price takers, a market in perfect competition will produce a Pareto-efficient situation.

To maximize utility, consumers will equate prices of goods and services to their respective marginal utilities; that is, with two products (X and Y) and two consumers (A and B), the marginal rate of substitution (MRS) between the products will be equal to their relative prices (P(X)/P(Y) If consumers are price takers,

MRS (A) = MRS (B) = P(X)/P(Y). (1)

This equality is a necessary condition for a Pareto-efficient situation in an exchange market economy.

In an Edgeworth Box (GRAPH 1), the points where the condition expressed by equation (1) is satisfied are found on the Contract Curve that results from the tangential points between the indifference curves for consumers A and B.


Assuming perfect competition and that the production of X and Y requires the use of scarce resources, firms will maximize their profits equating product prices with their marginal costs (MgC). Firms are also price takers. Thus, P(X) = MgC(X), and P(Y) = MgC(Y).

From the Transformation Curve (GRAPH 2), we know that a Pareto-efficient situation requires the impossibility of increasing production of one good without reducing production of another, from which results that the marginal rate of transformation (MRT) in production of both goods be equal to their relative prices.

MRT= MgC(X)/MgC(Y) = P(X)/P(Y). (2)

Considering equations (1) and (2), it follows that MRT= MRS (A) = MRS (B). Thus, producers and consumers, as price takers in competitive markets and acting so as to maximize profits and utilities, will produce a set of Pareto-efficient positions in production and in exchange.

GRAPH 3 depicts the points of competitive equilibrium in a production and exchange economy. Considering an initial factor endowment distribution between individuals A and B, competitive markets will result in such prices and costs of X and Y as given by the marginal rate of transformation T. It is important to remember that, starting at an initial point O of factor distribution, it will be possible to achieve a Pareto-efficient point located in the Contract Curve. Through exchange and production adjustments, the economy will move from the initial point O and attain a Pareto-efficient point E where competitive equilibrium is reached.


The question now is to find out whether the Pareto-efficient situation resulting from a given initial factor endowment and from the functioning of a competitive market will always be preferable to any other possible situation. The response is clearly negative.

A Pareto-inefficient point can be socially preferable to the point of competitive equilibrium. GRAPH 4 demonstrates such a situation. Given an initial endowment, competitive equilibrium is found at E1. Can one, however, state that this point is preferable to point M? E1 is a competitive equilibrium, and therefore Paretoefficient, solution, while M is a point not on the Contract Curve and, therefore, is inefficient.

Point M could be preferred if the Social Welfare Function attributes value to the pattern of wealth and income distribution among individuals A and B. Point E1 determines a strong concentration of wealth in favor of individual B. If society prefers a more equitable pattern of distribution, point M could be preferable, even though it is inefficient from an allocative standpoint. Pareto-efficiency alone is insufficient to assess social preferences of a society. I addition to a mere evaluation of allocative efficiency, it may become necessary to use other criteria for choosing a social optimum.


GRAPH 5 shows the Utility Possibility Frontier (UPC) between individuals A and B resulting from the Contract Curve in GRAPH 4. Given the utility functions of A and B, U(A) = U(X,Y), and U(B) = V(X,Y), each point on the Contract Curve determines a point on the UPC. Assuming a Social Welfare Function W = W(U(A), U(B)) that reflects the preferences of society, it is possible to construct the Social Indifference Curves (IS). The maximization of social welfare occurs at the Paretoefficiency point W. Pareto-efficient point E1, however, is inferior to Paretoinefficient point M, from the standpoint of social values of such a society. In this example, it prefers a more equitable income distribution, even though it implies an inefficient solution from the standpoint of competitive equilibrium.


The Fundamental Theorem of Social Welfare Economics proves that under perfect competition the market searches for an efficient competitive equilibrium at some point on the Utility Possibility Curve. Nothing guarantees, however, that it will be the point of maximum social utility.

It is possible to draw parallels between this situation and choices involving taxation.

Tax systems based on value-added taxes suffer from higher evasion rates than do cumulative tax systems, such as a bank transactions tax. Though we admit that theoretically VATs are neutral, and therefore more efficient from an allocative point of view (although this may not be true from an empirical standpoint), we cannot conclude that they necessarily result in a resource allocation pattern capable of maximizing the Social Utility Function. This stems from the perverse distributive consequences caused by VATs’ patterns of tax incidence, such as higher rates of evasion and higher administrative and operational costs, as compared to cumulative taxes on bank transactions.

In other words, there is a trade-off between tax efficiency and social evaluation of alternative patterns of tax incidence. A non-neutral tax, such as the bank transactions tax, may be preferable from the social standpoint because it drains less real resources from society due to its lower operational and compliance costs, and also because it does not encourage evasion, and therefore has a better pattern of incidence than is the case with VATs.

GRAPHs 6 and 7 demonstrate a situation in which a tax introduces distortions, but at the same time is preferred by society.

Initially, the economy is at competitive equilibrium E. It is a Pareto-optimal situation, with income distribution favoring individual B. However, the government seeks a fiscal policy that aims to redistribute income in favor of individual A, and therefore wants the economy to be located at a point in the shaded area, where social preferences for greater equity in income distribution are satisfied, preferably at some point along the contract curve within the area of the government’s preference.


One option is to adopt a conventional tax model, such as a VAT, which is considered to be neutral and efficient from the standpoint of relative-price determination, such as point E1. In order for this new equilibrium to be attained, relative prices would have to be P(X)1/P(Y)1, compatible with the competitive equilibrium E1.

However, if the conventional VAT tax model stimulates evasion, the change to point E1 will be frustrated, re-concentrating income and dislocating the new equilibrium, to E2, outside the government’s area of preference. In this sense, the option of a Pareto-inefficient solution as the point E* can be preferable, even if it is not a competitive equilibrium solution. In this situation the economy will stand at point E*, with relative prices that are incompatible with a Pareto-optimal situation given by the tangency point between the Indifference Curves I(A) and I(B) on the Contract Curve. However, this point (E*), as demonstrated in GRAPH 7, is preferable to points E and E2, despite the fact that it is Pareto-inferior relative to the solution originally desired by the government, point E1.


What these examples demonstrate is that one cannot state a priori that the best tax policy must necessary be made up of taxes that are considered allocatively efficient. In principle, neither cumulative nor any other type of tax system should be rejected strictly due to judgments about their allocative efficiency. There may be room for them in configuring a tax system capable of improving the social welfare of an economy. It is an empirical question.

A second reason why one cannot state a priori that a VAT is preferable to a turnover tax is that the assumptions required for a perfectly competitive market to exist are not usually met in the real world. What happens when the conditions for obtaining a Pareto optimum are not met? Richard Lipsey and Kelvin Lancaster, in pioneer writing on “second-best economics”, discussed this issue in 1956.

The second-best theory has shown that it is impossible to reliably rank different market situations without empirical analysis of each specific scenario. “The general theorem for the second best optimum states that if there is introduced into a general equilibrium system a constrain which prevents the attainment of one of the Paretian conditions, the other Paretian conditions, although still attainable, are, in general, no longer desirable. In other words, given that one of the Paretian optimum conditions cannot be fulfilled, then an optimum situation can be achieved only by departing from all other Paretian conditions. The optimum situation finally attained may be termed a second best optimum because it is achieved subject to a constraint which, by definition, prevents the attainment of a Paretian optimum. From this theorem there follows the important negative corollary that there is no a priori way to judge as between various situations in which some Paretian optimum conditions are fulfilled, while others are not. Specifically, it is not true that a situation in which more, but not all, of he optimum conditions are fulfilled is necessarily, or even likely, to be superior to a situation with in which fewer are fulfilled….It follows from the above that there is no a priori way to judge as between various situations in which none of the Paretian optimum conditions are fulfilled. In particular, it is not true than a situation in which all departures from the optimum conditions are of the same direction and magnitude is necessarily superior to one in which the deviations vary in direction and magnitude.”

The “second best” theorem proves that if any one of the conditions necessary for obtaining a Pareto Optimum situation is not satisfied, then the best possible situation ( the second-best) in general can only be attained if all other Pareto conditions are relaxed. In other words, as stated by Paul Samuelson “a given divergence in a subset of the optimum conditions necessitates alterations in the remaining ones”.

What is interesting in this theorem is the counterintuitive result that sometimes when one variable does not achieve its desired value, the best policy choice will involve moving other variables away from their first-best position.

Obviously, the postulates of the second best theorem are violated when critics of cumulative taxes, such as the bank transactions tax, declare, a priori, that valueadded taxes are more efficient than turnover taxes. Thus, the ranking of alternative tax regimes become problematic in the presence of market failures, tax avoidance, transaction costs, and other important departures from the postulates of perfect and complete markets.

According to second-best theory, as interpreted by J.A. Kay, “tax reform proposals must not be evaluated by counting the number of distortions, and arguments based on ‘double-taxation’ disregard the fact that it is the relative level of taxation, not the number of times the tax is levied, which is relevant in economic decision-making.”

One must recall that economic models impose conditions, hypothesis, and assumptions from which logical conclusions are drawn. The conclusions of a model depend on satisfying a set of given conditions for equilibrium. In economics, conditions of equilibrium depend on the behavior of consumers and producers who seek to simultaneously optimize their objective functions. Economic models usually assume a large number of behavioral hypotheses and conditions. One can assert that a second-best situation occurs when all conditions of equilibrium are not simultaneously met.

Lancaster and Lipsey demonstrated that usually when a single condition of equilibrium is not met, all other conditions must be altered, that is, satisfying all other conditions of equilibrium falls short of being the optimal behavior of economic agents. Therefore, it is possible to state that under such conditions, in order to achieve optimal equilibrium, the introduction of other distortions that correct the initial distortion may become necessary.

Therefore, given that the real world does not satisfy the rigorous requirements for achieving a Pareto optimum, it is easy to understand the futility of systematically avoiding the introduction of economic variables deemed to be allocatively inefficient on the arguments that society wishes to reach a perfectly competitive equilibrium. Usually, in order to achieve allocative efficiency targets that are almost always unachievable, such arguments ignore considerations about equity and income distribution, about fairness and about public and private costs. In so doing, issues of equity and ethics are sacrificed, albeit their importance when it comes to maximizing the social welfare function.

This is an important argument in the debate on tax reform in Brazil, where fierce battle is waged between those who support taxes that are deemed more efficient, such as the VAT, and those who call for bank transaction taxes, which are fairer because they are less prone to evasion.

In the debate, which involves matters of efficiency versus equity, the conclusions of the second-best theory become highly relevant, in that they demystify the need to attempt to satisfy competitive equilibrium conditions at any cost, even if this means reducing levels of social welfare, and compromising the social goals of tax justice and equity.

But perhaps, the most deadly criticism of the neo-classical analytics used in appraising the efficiency of taxation through the measurement of the excess burden caused by each type of tax (inspired by the path-breaking formulations of optimal taxation by Edgeworth and Ramsey) comes from the School of Public Choice led by James Buchanan and Richard Wagner.

“Excess burden is used widely throughout public finance for both normative and positive analyses…Both the normative and the positive uses of excess burden in contemporary fiscal theory start from the same analytical point of departure, where a tax is said to distort some margin of choice. The analytical task is then to appraise the extent of the loss brought by the distortion. Although (excess burden) is generally treated as a universal quality of all but lump-sum taxation…it is not a universal quality of taxation, but it is at most a contingent feature of a subset of the possible institutional frameworks within which fiscal outcomes emerge. The conventional excess-burden analytics transpose results from individual experiments where they do apply, onto market experiments where they do not apply… In an exchange each of the participants may well be modeled as maximizing utility, but there is nothing that is maximizing over the set of those participants…In democracies, taxation is something that citizens do to themselves…taxation emerges from inside the body politic of taxpayers…taxation in a democratic regime cannot lower utility for everyone …(excess-burden approaches to taxation) are inept because they show everyone losing utility…To the extent governance is mutually beneficial…there is no excess burden from taxation because taxation is the price that allows gains from trade to be exploited…” Wagner concludes by stating that “Within the context of democratic ideology, even if perhaps not democratic practice, taxation would seem to be judged good or bad according to its ability to facilitate or impede mutually profitable fiscal exchange and not according to some excess burden metric that is assessed independently of the fiscal process and the institutions that frame it."

Excess burden would seem truly to be a grin without a cat.

Such considerations give weight to tax appraisals which are based on each tax’s own advantages and disadvantages, independently of the alleged allocative distortions derived from conventional neoclassical analyses. As will be shown by the simulations presented in the next chapter, even according to conventional excessburden metrics, VATs can be more distortionary than cumulative taxes if we assume the existence of evasion and of differential tax rates among products.

Furthermore, it will be shown that they also introduce other serious distortions as they require comparatively higher tax rates for any given revenue target and impose extremely high operational and compliance costs. Cumulative taxes are levied on the total value of production, whereas VATs are levied solely on wages, profits, interests, and rents at each stage of the productive process. Consequently, for a given revenue target, the VATs need rates that are higher than those of cumulative taxes. This means taxing factors, including labor, at rates that are higher than those of cumulative taxes. Thus, it is clear that VATs discriminate against labor, especially in the highly labor intensive sectors, such as the service sector.

According to the well-known Meade Report on tax reform in the United Kingdom “the economist distinguishes between the ‘income effect’ and the ‘substitution effect’ of a tax burden.... ‘income’ effects are not a symptom of economic efficiency… but ‘substitution’ effects are an indication of economic inefficiencies and wastes… Avoidance of economic inefficiencies would involve avoidance of high marginal rate of tax where these substitution sensitivities were great. One corollary of this need to keep marginal tax rates down is a general presumption in favor of tax systems which provide a broad basis for revenue-raising purposes. To raise a given revenue by means of low rates of tax spread over a large tax base may be assumed to cause less marked ‘substitution’ distortions than to raise the same revenue by concentrating high rates of tax on a few activities, unless special circumstances suggest that those particular activities show exceptionally low substitution sensitivities”.

Regarding the allocative effects of a tax system, it should be observed that, ideally, in order to minimize distortions, taxes should not cause any changes in economic decisions that would have been made in the absence of taxes. The ideal tax system would minimize dead-weight tax loss. However, we know that only a tax on life, that is, a fixed value per capita tax would achieve this desideratum. But being an unacceptable option in modern societies, we are left with trying to minimize, not avoid, losses in efficiency. Thus, a basic rule would be the use of taxes with high average rates, but low marginal rates, as recommended by Martin Feldstein when he says that it is the marginal tax rates that determine their dead-weight loss caused by excess-burden. Because economic decisions are always made at the margin, the use of taxes with this characteristic would be more desirable than taxes that show equal average and marginal rates (such as VATs), or even marginal rates that are higher than average rates (such as a progressive income tax).

Thus we see that, by requiring lower marginal rates for raising a given revenue target, the bank transaction tax can be less distortionary than value-added taxes, which require significantly higher marginal rates. This assertion clears up the mistake that surrounds assertions about the supposedly natural evils of cumulativeness and the supposedly natural merits of value-added taxes. The Meade Report actually concludes that, “it is an impossible task to trace through the complete efficiency and distributional effects of a tax change in a complex economy in which there is a complicated network of market and productive inter-relationships between a large number of products and activities and in which there many kinds of market imperfection and of environmental and similar side effects.”

All in all it is important to remember when designing tax systems or when proposing tax reform, that theoretical models used by economists require strong assumptions for optimality, always highly unrealistic, which turns them into inadequate guides for policy purposes.

“The theory of optimal taxation has, for the past two decades, been the reigning normative approach to taxation. This paper argues that, in its current state, optimal tax theory is incomplete as a guide to action concerning many critical issues in tax policy. It is incomplete because it has not yet come to terms with taxation as a system of coercively collecting revenues from individuals who will tend to resist. The coercive nature of collecting taxes implies that the resource cost of implementing a tax system has been and will continue to be a critical determinant of appropriate tax policy…production efficiency is in general not desirable when there are constraints on how commodities and profits can be taxed…that the apparent triumph of production efficiency as a goal is somewhat surprising in view of the strong assumptions needed to demonstrate its desirability.

Joel Slemrod, the author of this quote presents us with a demolishing view of optimal tax theory as a guide to policy. Indeed, many of the critical issues in tax policy nowadays lie “outside the domain of optimal taxation theory” and of its efficiency implications. He goes on to say, “I believe that its critical problem is the failure to consider the technology of collecting taxes…the leap from the blackboard to the real world is a large one when it comes to taxation…integrating the issue of administrative ease into optimal tax theory will require a shift of emphasis away from the structure of preferences, which has been the principal focus of optimal tax theory, toward the technology of tax collection”.

Also, as quoted elsewhere in this text, Frank Hahn has stated that due to its inherent limitations optimal tax theory “cannot be taken very seriously” as policy prescriptions.

In sum, the entire optimal taxation theory, of which the excess burden analysis represents its main analytical construct, has been forcefully challenged by Public Choice economists such as James Buchanan and Richard Wagner. Their main argument states that conventional excess burden analyses are limited in their usefulness at most to a contingent subset of possible institutional frameworks, such as a “monopolistic or exploitative state to explain fiscal outcomes”, which represent and “secure advantages to the dominant classes or groups at the expense of the remainder of society. Such exploitive models assimilate the state more strongly to the practice of brigandage and rent extraction than to the supply of public goods”. According to them, excess burden analytics are not adequate for modern societies that act through a “cooperative state to explain fiscal outcomes… (through which) fiscal programs advance the common interests of everyone, evoking images of public goods and the benefit principle in the process”. As such, taxation does not allow for measurement of losses and of inefficiencies such as implied by excess burden analysis. Instead of being an autocratic regime that rulers impose on their subjects, in democracies it becomes a mutual and voluntary relationship among citizens, and “taxation is something that citizens do to themselves”. Thus, “taxation in a democratic regime cannot lower utility for everyone. At least some people must gain utility, and in the limit everyone could gain”. In conclusion, “excess burden analytics make no relevant analytical contribution and would seem to represent a grin without a cat…there is no excess burden from taxation because taxation is the price that allows gains from trade to be exploited.”

Enforcement and compliance

Another situation could result from comparing the high operational costs of the VATs to the cost of tax collecting through cumulative taxes such as bank transactions tax or the Single Tax. By being non-declaratory and electronically collected, they show extremely low operational costs both to the public and to the private sectors.

The administrative costs of VATs are absurdly high, especially in federative countries such as Brazil. From an operational perspective, they are complex and inefficient if imposed by sub-national governments. VATs are appropriate for centralized countries. Few federative nations use them, and the ones that do, incur high costs and enormous bureaucratic complications, as happens in Brazil and Canada. The United States have kept away from them for good reason. And in Brazil the attempt to unify the ICMS (a VAT administered by each of the States) is warranted by the unfortunate experience of having in place an absurdly complex system.

A paper written in the United Kingdom showed that even in countries with centralized administration the operational costs of VATs are high. After the personal income tax, value-added taxes are the most costly to collect, absorbing 4.72% of revenue, as shown TABLE 2. Excise taxes carry the lowest operational costs. Because they have essentially non-declaratory characteristics, their operational cost is only 0.45% of revenue.


Following this line of analysis, the conclusions of the International Symposium on Tax Reform, held in São Paulo and attended by some of the world’s major tax professionals (including Arnold Harberger, Charles McLure, Richard Bird, James Buchanan, Vito Tanzi, Anwar Shah, and John Edwards) were unanimous in condemning the Brazilian tax system for empowering sub-national government units to collect VATs.

Furthermore, in federative countries this type of tax leads to federal tax competition, and to the raising of customs barriers between states and even between cities. It is the paradise on earth for tax attorneys and corrupt tax collectors.

Everardo Maciel, former Secretary of the Federal Revenue, said, “we committed a grave tax policy mistake in this country during the 1960s. Brazil was extremely reckless and bold when it introduced … the value-added tax, the ICM… and imposed it down the line until the retail sector. Certainly the country was not prepared for this type of tax mainly because of its pervasive culture of tax evasion. Another error was committed…making the ICM replace the IVC (a turnover sales tax collected by states). Basically, the mistake lay in conferring on the States the competence for collecting a value-added tax on consumption, which resulted in two almost insoluble problems. The first was related to tax exemptions for exports… the second appeared in interstate transactions. Value-added taxes on consumption do not lend themselves to co-existing comfortably with federal customs barriers.”

Ives Gandra teaches us that the Gordian knot in tax reform in Brazil rests in the ICMS. It is a VAT-type tax. The vast majority of countries throughout the world have adopted this type of tax, but almost always under the responsibility of the central government. Rarely are they used in federations such as Brazil. The ICMS should be a federal tax, not a state tax, and stubbornness in keeping it in the hands of the states has caused serious distortions in its operation.

In Brazil, organized groups use the tax system as a weapon in their political battles for conquering greater economic space. The criticisms raised by certain groups opposed to more innovative tax reform proposals, such as the Single Tax, are markedly pseudo-scientific. As such, they must be interpreted as being merely an instrument in the struggle for economic power.

Intuitive perceptions indicate that tax evasion, fraud and avoidance vary in direct proportion to the level of nominal tax rates and to the complexity of the tax system in use. Although theoretical models that simulate tax compliance do not come to clear conclusions or to correct indications of taxpayers behavior when faced with varying tax rates or with changing levels of difficulty in tax assessment procedures, it is safe to say that under usual assumptions about risk preferences a rise in the marginal tax rate or in revenue-neutral progressivity increases the flight to the underground economy and increases the tax gap. That is, the higher the rates, the greater the stimulus and the reward for tax evasion and avoidance, such as suggested by the Laffer Curve construct . It is easy therefore to conclude that VATs stimulate tax evasion and avoidance more intensely than do taxes that require lower rates, such as certain non-declaratory cumulative taxes. VAT models of taxation are becoming increasingly fraud-ridden, even in countries with high ethical standards, as most of the European countries.

As tax avoidance and evasion increase, new rounds of rate hikes and of bureaucratic controls become necessary. Therefore, it is not surprising that in countries where the VAT has been imposed, compliance and administrative costs have risen continually, leading to rate increases. Paraphrasing Henry Simons, the experience with VATs, especially in developing countries, is like “dipping deeply with a sieve”.

In Brazil, VATs were initially set at 12% and today are at 17% for most products; in some cases they reach 25 or even 30%. Therefore, because they require higher rates and because they stimulate tax avoidance, VATs show undesirable patterns of incidence. Some taxpayers are taxed in excess, whereas many others pay too little, or less than they should.

In Brazil, the tax system stimulates collusion against the public sector, sales without invoices, fake transactions, and “tax planning”. This system creates largescale allocative distortions, in which production costs and competitive capacity no longer define production efficiency. On the contrary, the competitive capacity of businesses depends to a great extent on the administrator’s skill in practicing crafty tax avoidance schemes more effectively than his competitors. This stimulates survival of the most cunning and domination by the corrupt. Market selectivity no longer rewards the most efficient. It is not perchance that in economies with high tax avoidance rates, the tax burden falls disproportionately on those who are least equipped to avoid taxes, such as wage earners. In Brazil, payroll income accounts for 52% of the tax burden, but labor income represents only 27% of national income.

Enforcement and compliance costs are becoming crucial elements in appraising the efficiency of tax systems. “The significance of compliance costs and enforcement difficulties warrants rethinking basic questions of tax design; perhaps tax rules or even tax systems that are desirable in principle should be redesigned in practice, sacrificing the original equity and efficiency goals to some extent for the sake of improving tax administration”.

Differential incidence

Cumulative taxes may certainly cause typical distortions. They introduce changes in relative prices, although such negative effects are greatly eased by their low marginal rates. They are less transparent because they are embedded in production and in input prices, and they become invisible to the consumer. In the case of exports, cumulative taxes require more complex mechanisms for zero-rating, although this is a manageable technical problem if detailed inter-industry relations matrices are available.

Another alleged distortion in evaluating bank transactions taxes stems from the assumption that being collected cumulatively at each stage of the production process they imply high tax burdens on products with “long” production chains. This is a mistake. Production chains should never be described as “short” or “long” – they are always infinite in their number of transaction links. Production of any good or service involves contributions from all other sectors of the economy. It is a circular process that necessarily uses inputs from several other sectors which, for their part, also need inputs from other sectors, and so on. Therefore, production chains are always infinite, never ending.

What truly determines the tax burden of a cumulative tax is the relationship between the value of inputs and the value-added at each stage of the production process. For example, if a given production sector buys inputs at a certain price and adds an equal amount of value, the cumulativeness carried by the prior production phases is totally imbedded in the price of the acquired inputs. The value-added at this step of production does not carry any cumulative effect, and only in the following stage of production, when the final product of this stage becomes an input, it begins to carry the cumulativeness of previous stages of taxation.

TABLE 3 reflects this fact, assuming a value-added of 100% of the value of acquired inputs. In the example, we assume that the product’s final value is R$ 100, and the bank transaction tax has a rate of 1% on each bank credit and debit transaction. The data shows that the turnover effects dissipate rapidly when we analyze the tax carried over from previous production stages, following a decreasing geometric progression. In this example, the total value of tax accumulated into the product’s final price is R$ 3.8646; that is, the tax burden on final price is 3.8646%.


One will notice that under the conditions specified in the example, cumulativeness carried along the production chain quickly dissipates, reaching a value of merely five cents of a real (R$ 0.05), in phase T-5, and quickly moving to near-zero values as we move backward in the chain. It can thus be seen that the accumulation of taxes occurs with a much less alarming intensity than critics of bank transaction taxes would have one believe. In phase T-3 the value of the accumulated tax is little more than 5% of the total tax burden on final price, as can be seen in ILLUSTRATION 3.


Taking an extreme example, in which the value-added in each step is only 10% of the value of acquired inputs, the tax burden contained in the final price reaches 18.1066%. Notice that even in this case the tax carried over from the previous phase of the production chain also drops quickly to near-zero values as we move backwards in time. In phase T-6 the value of the tax is only 5% of the total tax burden (ILLUSTRATION 4).


ILLUSTRATION 5 shows the impact of cumulativeness on the production chain for several levels of value-added.

Critics claim that the CPMF goes against all modern principles of tax theory and contradicts everything other countries are doing. The first part of this comment is false, and the second, irrelevant, and deserves no response.


The myth about the inevitable evils of cumulativeness must be placed in proper perspective. No tax is neutral, whether cumulative or on valued added. All taxes have advantages and disadvantages, as is widely acknowledged both by critics and supporters of bank transaction taxes. And comparing advantages to disadvantages, cumulative taxes such as the CPMF show a clearly positive balance. There is no evidence, either in Brazil or in other countries, that bank transaction taxes are more distortionary than other available alternatives. They do not discriminate against wages, they have rates that are structurally lower than those of the VATs and, therefore, discourage tax evasion and corruption. Furthermore, they have extremely low operational costs, almost zero in the case of electronic bank debit taxes such as the CPMF.


Since the Single Tax proposal was first presented in the early nineties, it has been known that transaction taxes impact interest rates in direct proportion to the rate of turnover of market financial transactions. This could imply losses to investors and would require a corresponding increase in interest rates to offset this effect. Furthermore, extending the investment term in order to neutralize the turnover effect would imply an unjustified discrimination against short-term investments, although proponents of the Tobin Tax see this as a necessary instrument to slow down the flow of speculative money around the world.

I have insistently demonstrated that, because of its cumulative nature, borrowers would have to bear increased costs. For this reason I proposed a specific methodology for applying the Single Tax on financial transactions, on the capital markets, and on the stock exchange.

Financial transactions should be considered as rental payments on capital. Thus, just as the value of a good being rented in the real market is not taxed by a transaction tax (only the value of the flow of services is taxed), the principal in a financial transaction should also not be reached by such a tax (only the interest payments should). In a residential rental, the transaction tax is imposed solely on the value of the rental flow, and not on the stock value of the real estate being rented out. Likewise, in financial transactions only the interest earnings should be taxed, and not the capital that produced the earnings. If this were not the case, shorter transaction periods would bear higher impact of this tax. This would make short-term transactions in the stock market, such as day-trade, too expensive, even impracticable.

According to our proposal, financial transactions would take place only through special bank accounts, similar to savings accounts. These special accounts, unlike regular checking accounts, would only be allowed to have credits or debits from other special accounts, or from the accountholder’s own checking account. Credits in the special accounts would be exempt from the tax. Whenever funds are debited from a special account and credited to a checking account of the same account holder, that portion of the transferred value that represents real returns earned during the period in which the funds were held in the special accounts would be automatically taxed. Once those funds are released for other uses from the checking account they will be taxed like any other transaction. It is worth noticing, that in order to improve the application of the CPMF in Brazil, the Central Bank adopted similar procedures in 2005.

Some people fear that applying a transaction tax without those special precautions could drastically reduce the already modest returns of short-term funds (which are rolled over daily), causing a massive exodus to cash deposits, where they would be less taxed. The monetary base could suffer a contraction as the result of the increase in the compulsory deposits of the banking system in the Central Bank since the rate of compulsory deposits for cash accounts are approximately 50% higher than those required from short term investment funds. This would also cause a reduction in the financial system’s supply of loan able funds, causing economic contraction.

As mentioned before, the initial implementation of the CPMF did not avoid these inevitable distortions. The correct operational system contained in the Single Tax proposal would exempt financial transactions insofar as they occur strictly within the scope of the financial or capital markets. Taxation would only occur upon transfer of real returns to the investor’s own checking accounts, when they become available for other uses. Thus, short-term transactions would not be discriminated, borrowers from banks would not be punished, the fear of possible migration to foreign stock market funds would be reduced and trade would not flee abroad in order to escape the CPMF.

Unfortunately the CPMF was introduced in Brazil nos as a Single Tax, but as an additional tax in a confusing and inefficient system. And despite the growing evidence that it had negative impact on financial markets, the government took too long to correct such distortions.


The Single Tax has countless advantages, such as enormous simplification and reduction of tax collection costs. This latter advantage is not restricted to a reduction in the size of the governmental apparatus, but also includes lower compliance costs to businesses, that currently use about 10% of their administrative staff to meet the requirements of tax reporting. Altogether, such costs now amount to as much as 20% of the country’s gross tax revenue, which presently amounts to 37% of GDP. This means that the impact of the Single Tax proposal in terms of releasing real resources to be put to other uses would be on the order of 7.4% of GDP. This amount is more than double the net capital earnings transfers to foreign income recipients – payments of interest, profits, and dividends. These are resources that could be channeled to productive investments and would be capable of leveraging economic growth, instead of being absorbed in government consumption activities and private administrative costs.

The Single Tax, if we take the experience of the bank debit transaction tax (CPMF) as an example, could lead to the virtual elimination of tax evasion and fiscal corruption, and to a reduction of the underground economy, at practically no cost to the public sector. Tax collection would take place automatically upon each debit or credit transaction within the banking system. Each credit and each debit account would be charged a fixed percentage of the value of the transaction. Thus, every time a check or other payment method is cleared, the system would automatically transfer the proceeds of the tax to the federal, state, and municipal treasuries, pursuant to predefined criteria. To make tax evasion practically disappear it would suffice to monitor the bank clearing systems.

Practical administrative and operational aspects of taxation are usually ignored, or simply assumed away as non-existent, by economists. Only recently have economists turned their attention to such issues. Public policy towards evasion reflects complex issues involving efficiency, equity and growing compliance and administrative costs, and generally need much more attention from economists than has happened in the past. “Tax evasion clearly complicates measures of the distortionary effect of taxation; given a fixed revenue requirement, evasion means that higher and more distortionary taxes on reported income may be needed, while unreported income largely escapes taxation and its distortionary effects”

Skinner and Slemrod say that “the cost of tax evasion include violations of horizontal equity, vertical equity, and efficiency…Increased enforcement will generate more revenue, but often at a substantial resource cost….The advantage of tax simplification is that it will generally reduce the loopholes that are breeding grounds for tax evasion schemes. Finally, reduced marginal tax rates have been associated with a significant decline in tax evasion.” Thus, the Single tax proposal stands on a superior level relative to conventional declaratory taxes such as the VATs, which are notoriously prone to corruption, high operational costs and growing complexity.

The most significant element of the Single Tax proposal is that, being evasion proof, the tax rate could be low. Using the statutory regulations that apply to the bank debit transaction tax (CPMF), the Single Tax revenue can be estimated by simple linear projection. Despite all existing constitutional immunities and even in the absence of adequate government auditing of bank payment systems, revenue from the CPMF was significant and indicates that a 3% rate on each side of each transaction is capable of generating around 40% of GNP in revenue. This amount is larger than the combined tax revenue from all federal, state, municipal, and social security taxes and contributions.

However, the initial Single Tax revenue forecasts assumed a universal tax base, which is significantly broader than the present tax base of the CPMF. Furthermore, precarious auditing of banking institutions is responsible for a significant part of the difference between the revenue estimates done in early studies of the Single Tax and actual revenue collected through the present system. The data available at the time of the first revenue estimates were done in the early nineties implied that, with a 1% rate, annual revenue from the Single Tax would be between US$ 69 billion and US$ 89 billion. At the then current exchange rate, these revenue levels accounted for between 19.6 and 25.3% of GDP. TABLE 4 reproduces the main results obtained by different research groups in their projections.

During the period immediately following the announcement of the Single Tax proposal, and following the publication of a study which contained the initial revenue estimates, a large banking institution began to provide the author with monthly volumes of debit and credit transactions related to its clients’ checking accounts. That banking institution’s accounting was the basis for transactions estimates, and made possible to obtain data equivalent to a significant sample of Brazil’s total banking system. These data were reported to the author monthly between June 1990 and May 1996. The reported values at December 2000 prices can be found in ANNEX I-B.


TABLE 5 shows the annual banking transaction data for that institution and, based on the estimates of its share of the national banking system, it was possible to estimate projected revenue of the Single Tax, at rates of 1% for debit and 1% for credit bank transactions.

The conclusion to be drawn from these projections is that, considering the low marginal rate for the Single Tax needed to replace current tax revenue, the incentive for tax evasion would virtually disappear. Tax evasion would become impossible except in cash or barter transactions. In these two cases the cost of evading taxes (transaction costs) would most likely exceed the benefit from tax evasion, thus removing the stimulus for attempting to circumvent the tax system. Furthermore, the regulations for the Single Tax should provide that transactions above a certain value must be processed through the banking system, under penalty of legal nullification if processed otherwise.


It is important to remember that under the Single Tax model any withdrawal or deposit of cash into or from the banking system would be charged an overtax at a rate that is equivalent to the revenue raised considering the number and value of transactions carried out with that cash amount until its return to the banking system. This operational procedure would deter the use of cash transactions as a means of tax evasion.

Such a taxation system would eliminate (or compensate for) tax evasion – the value of which is presently estimated to be between 30 and 40% of public revenue. There would be a noticeable reduction of production costs and prices, a fall in the costs of the public apparatus, and, potentially, a significant reduction in the tax burden. Certainly some of these gains would be once and for all, but would be sufficient to allow a significant fiscal adjustment and a noticeable recovery of the country’s investment capacity.


Romantic visions see taxation as an expression of the civic spirit of citizens, conscious of their rights and duties. Humanitarians have come to believe that the only way to redistribute wealth and income is through compensatory (or punitive) taxation of the more efficient and wealthier. Economists and political leaders seek through taxes, or through exemption from taxes, the pathway to stimulate economic growth. Ecologists and conservationists use the tax system as a form of environmental protection and of punishment of those who break preservationist rules. Urban and regional planners use taxes as inducement mechanisms to reach desirable social objectives. Farmers want to achieve land reform through taxation of large landowners. In a nutshell, everyone seeks in the tax system the solutions to their problems. As Everardo Maciel said, “this merely serves to demonstrate that the debate over taxation can take unpredictable turns, dictated by fortuitous reasons or impenetrable motives.”

Given these multiple objectives and the inevitable indetermination that stems from the existence of more objectives than instrumental variables to achieve them, the tax system has lost effectiveness in performing its essential function, that of raising public revenue.

We know that some taxes, to a greater or lesser degree, may perform regulatory functions. Some taxes were created with essentially non-revenue objectives, as is the case with import taxes, which exist fundamentally as instruments of industrial policy and for protection of domestic production. Revenue from these taxes is strictly a secondary objective. Others, such as the IPI (a federal value-added tax on industrial production) on tobacco and alcoholic beverages, combine revenue goals with social objectives of public health and safety.

Unfortunately, this non-fiscal perspective has influenced so intensely the Brazilian fiscal policy that its tax system has become unintelligible and has performed poorly in its primary revenue-raising function. The multiplicity of objectives to be met by the tax system has turned it into a highly complex, bureaucratic, expensive, inefficient, and highly corrupt system, and has become a strong inducement to a wide variety of non-compliance and evasion tactics.

“This problem has been recently highlighted by a Report to the President of the United States on tax reform, ´Simple, Fair and Pro-Growth: Proposals to fix America´s Tax System´ prepared by the President’s Advisory Panel on Federal Tax Reform (November 2005). The Report suggests that legislators have lost sight of the fact that the fundamental purpose of the tax system is to finance public spending. Other goals have distracted the system from its fundamental purpose”.

From a fiscal standpoint, it is essential to collect revenue as efficiently, economically, and simply as possible. For this very same reason the Single Tax proposal gains significance as a basis for Brazil’s tax reform.

Theoretical formalism, which is much appreciated by staff economists who seek to identify and measure the allocative and distributive impacts of taxes with meticulous precision, is proving itself increasingly misleading as a script for tax reform, given that economic reality does not always adjust itself to the ideal economic models designed in the realm of high abstraction. In the words of Mangabeira Unger, the academic perspective unfolds in the midst of “edifying and tranquilizing illusions”. But “the world is wild and obscure”. The world of perfect competition does not exist.

Along this same line of reasoning, Delfim Netto states that economic science creates the impression of being “…a body of progressive knowledge, a ‘hard science’.” He further says, “What all this sophistication has forgotten is that its conclusions depend upon two implicit postulates: 1) that tax evasion does not exist; in other words, that each citizen is a prisoner of rigid social rules that cast the tax evader into opprobrium, and 2) that collection of these taxes has no costs; that is, they flow naturally and smoothly to the coffers of the treasury… When one considers the falseness of these two postulates, one begins to doubt the quality of suggested recommendations and to have greater intellectual respect for ‘non-declaratory tax’ proposals…”

The rescue of the concept of revenue as the fundamental and primary goal of any tax system is supported by two articles published in Folha de São Paulo, by Prof. Roberto Mangabeira Unger.

In “Taxes and Paradoxes” the author confirms the need to rescue the revenue function of taxes, when he states that indirect taxes, even cumulative ones, can “generate a lot of money with little economic disruption”, whereas direct and progressive taxes, so dear to staff economists, “such as individual income tax, do not produce the expected revenue. Neither can it do so, for the time being, without engendering disincentives, capital flight, and devastating tax avoidance.” Unger goes further and says that the essential objective of a good tax is to generate “money for the State to invest in social issues.”

In another article by Roberto Mangabeira Unger, titled “Tax Reform (1)”, the author confirms the theory espoused by supporters of the Single Tax, that redistribution of income “is done more efficiently from the expenditures side of the budget than from the revenues side supported by a progressive tax model”, thus demystifying the academic theory that progressive taxes are a necessary condition for good tax reform.

Mangabeira Unger’s conclusions and proposals on tax reform, different from the Single Tax, feed into a system made up of consumption, inheritance, and financial gains taxes. But the fundamental point is that the premises that restore the focus of taxes on revenue, displacing its non-fiscal effects, are the same that support the defense both of the Single Tax and of Mangabeira Unger’s proposals.

“Governments will rediscover that the objective of taxation is to provide revenue for the state to meet its obligation and not to engage in social engineering through the tax system”.


A bank transaction tax is a good tax if it is used as a single tax; but is a bad tax if added onto many others. The CPMF was created as one more tax to be added to Brazil’s fiscal paraphernalia.

The Government disfigured the proposal of the Single Tax on bank transactions. Conceived to be the basic tax for the entire fiscal system, it was ultimately reduced to a dishonorable role as one appendix in the nation’s confusing tax structure. The Government acted as a rapist who, in its brutality, sees nothing but the immediate object of its appetite. It completely ignored the virtues of the Single Tax – such as reducing bureaucracy, instilling morality, and promoting development. Instead, it adopted the tax solely for its high capacity to raise revenue.

However, even if spurious, the CPMF produced a worthy result, in that it allows an accurate evaluation of the efficiency of the bank transactions tax, which is the foundation for the Single Tax construct.

Described by adversaries as a hated cascading tax, the CPMF is called by every offensive name that can be given to an innocent tax: dumb, unfair, anti-production, anti-savings, and harmful to exports. Several political analysts, financial reporters, and especially collectors and other tax professionals criticize cumulative taxes, while heaping praise on value-added taxes, such as the ICMS. They support VATs as if they were the eighth wonder of the world. They consider them to be fair, neutral, and efficient.

Reading the literature on the tax reform debate, a rookie economist who believes everything said by adversaries of turnover taxes, cannot avoid the impression that the simple elimination of turnover taxes would redress all that is wrong with it, and that with a simple flick of the wand Brazil’s tax system would become rational, fair, modern, and efficient.

There is nothing more mediocre than to accept, without rigorous critical evaluation, the prejudices and clichés contained in these opinions. The CPMF and its experience in Brazil brought to light many positive aspects of this tax.

It is not necessary to overemphasize its virtues. All one needs to do is consider that with rates of merely 0.38% (0.30% from June 2000 to March 2001) and with practically no costs to the government or to the taxpayer, the CPMF raised R$ 14.4 billion in 2000, approximately R$ 17.1 billion in 2001, and R$ 36.2 billion in 2007. Taxes that are highly complex and carry high administrative costs, such as the IPI (a federal value-added tax on industrial products) and the corporate income tax, generated revenue of only R$ 18.8 billion and R$ 17.6 billion in 2000, respectively; R$ 19.4 billion and R$ 17 billion in 2001 and R$ 32.9 billion and R$ 70 billion in 2007.

The CPMF is universal, evasion-proof, and it grasps all economic agents, eliminating the inequity of self-assessed, paper-driven taxes that impose high tax burdens on some taxpayers while favoring tax evaders and expert tax planners who enjoy markedly lower individual burdens. The CPMF is able to eliminate the greatest anomaly present in the current tax system, which is artificial production cost differences and consequently unfair competition caused by widespread tax evasion. Tax avoidance and evasion distort desirable patterns of distribution of the tax burden, and this distortion is more serious and disrupting than the alleged distortions in relative prices that a turnover tax, such as the CPMF, might be causing to Brazil’s economy.

Concerning this issue, Everardo Maciel says, “The literature of public finance is full of examples of economic distortions caused by taxes. What is not stressed, however, is that the comparisons presume a context in which tax evasion is nonexistent or of little relevance. This, however, is not the reality in emerging countries. In these countries, to parody a well-known aphorism, one might say, create a tax, create evasion.” He continues, “ultimately, what we intend to assert is that tax evasion in emerging countries is the greatest economic distortion caused by taxes, far superior to any other.”

Notwithstanding, IPEA (a Brazilian government-sponsored think tank) states the following: “The injurious elements of cumulative taxation can be classified into two groups: it harms the allocation of resources and the competitiveness of domestic goods. These distortions are due to the fact that this type of tax unintentionally and uncontrollably alters the economy’s relative prices.”

To respond to this assertion, we should invert the argument and ask its authors if the changes in relative prices introduced by value-added taxes are intentional and controllable in an environment with widespread tax evasion and avoidance, as happens in Brazil An inevitable conclusion is that taxes that are easily evaded, such as VAT-style self-assessed taxes, certainly create even more unintentional and uncontrollable changes in relative prices, because nothing is as unpredictable or uncontrollable as tax evasion.

The Brazilian economic environment has greatly changed in the present computerized and globalized world. Thus, one should not imagine that conventional taxes, created during the time when information technology was based on paper, on accounting ledgers, on physical transportation, on economic isolationism and on political fragmentation, such as prevailed during most of the 20th century, will be able to avoid widespread tax evasion and its dramatic consequences. Conservative tax policies will only deepen such inadequacies in the future.

In a country like Brazil, that suffers from all sorts of administrative deficiencies, from a slow and inefficient judicial system, from a weak and discredited tax auditing apparatus, and from a deeply rooted culture of tax evasion, it is easy to understand the reason for so much criticism aimed at the CPMF. It corrects such anomalies. After all, for rent seekers, it is preferable to “pay” taxes on profits and on valueadded, since they can be easily manipulated by delinquent taxpayers, than to have a tax system that eliminates privileges, prevents avoidance, and turns universal the set of taxpayers in the country.

The CPMF has been used successfully to achieve extremely important objectives, such as fiscal balance and currency stability. Still, there is generalized rejection against it although, if it did not exist, conventional taxes, which are always more inefficient and inequitable, would require even higher rates than they do at the moment. Therefore, the criticisms about cumulativeness need to be more deeply analyzed, and cannot be uncritically accepted.


One of the most frequently raised questions about a bank transaction tax has to do with its progressiveness. Critics claim it is regressive.

Actually, because it is a cumulative (turnover) tax, products that involve a greater number of transactions along the productive chain – with more roundabout production methods – and those that add less value at each stage, will be more heavily taxed. Thus, the Single Tax system should have a natural degree of progressiveness given that wage goods – staple products that make up the demand bundle for lower income families– would tend to have a lower tax burden than that of relatively more sophisticated products. Wage goods usually have less roundabout production chains, with less processing and a high rate of added value relative to the value of inputs at each production stage.

Another interesting feature of the Single Tax proposal is that income and production become no longer the main components of the tax base, as happens in conventional tax systems. The tax base would shift to financial transactions. Thus, productive activities become less taxed, and those that involve mere asset transfers, that currently are notoriously under-taxed, such as estate and personal property transactions, would be more heavily taxed.

The Single Tax proposal has, therefore, some essential characteristics that must be stressed: it ensures tax collection; it eliminates tax evasion and fiscal corruption; it increases efficiency of tax collection; it frees up significant resources in the private and public sectors; it is a comprehensive system; and it exhibits natural progressiveness.

Maria da Conceição Tavares evaluated the alleged regressiveness of bank transaction taxes, considering their incidence by income brackets. In her article “Imposto sobre circulação financeira” (a Tax on Financial Circulation) the author says, “The argument that the tax would basically penalize the middle class is unjustified. This is a tax that primarily penalizes individuals who turn the financial circulation of their savings into an extra and often considerable source of income.” She further states, “because they are one of the dynamic vectors in the economic restructuring and globalizing process, bank transactions constitute one of the few potential bases for future taxation in which it is possible to anchor public revenue increases without punishing the productive sectors and the needier social segments”. Because of the difficulties in simulating transactions for businesses and financial institutions, the results presented in her article refer to a partial revenue base, restricted to individuals, on whom a tax similar to the CPMF would be levied at a rate of 0.25%. The conclusions of the exercise rebut arguments that a bank transactions tax is unfair because it is regressive. Her conclusions (although they may have lost some validity given the increasing use of bank transactions by all income groups) are reproduced here, in full:

“1 - The lower-income groups, with average monthly income equal to 1 to 3 minimum wages and half of the group with average monthly income equal to 4 minimum wages – which together account for 70.6% of the reference population (income receiving individuals older than 10 years of age that are economically active) – are presumed not to use the banking system and therefore would not be directly affected by the tax.

2 - Of the remaining 29.4% that operate through the banking system, the tax burden falls predominantly on the higher-income groups (those with average monthly incomes that fall between 20 and 38.7 minimum wages).

This latter segment, which accounts for a scant 3.4% of the reference population and less than 12% of individuals with bank accounts, holds 29.2% of total income and would account for 63.5% of the IPMF revenue paid by individuals.

The group that has lower incomes (between 7.2 and 14.2 minimum wages per month), which represents 62% of the taxed universe and 18.2% of the reference population, accounts for 31.1% of revenue, whereas this group’s share of income is 38.6%.

Even the upper-middle segment, with average income of 14.2 minimum wages accounts for a smaller share of tax revenue than it should, given its total income. In other words, the argument that the tax would unfairly penalize the middle class is not supported. This is a tax that burdens those individuals who make bank transactions an extra and considerable source of income.

3 - The average effective rates on members of each group are also progressive, varying from 0.25% (affecting only that lowest-income portion of the group, which is taxed only once at the time of salary withdrawal), up to 0.70% levied on the group whose average monthly income is equal to 38.7 minimum wages.

Rate progressiveness is determined by values attributed to coefficients of financial circulation. The underlying theory is that greater savings coefficients correspond to higher income levels, and that the lion’s share of those savings goes into financial investments.

The portion of income that is earmarked for this purpose is expressed by the financial investments coefficient. This coefficient, in turn, is associated with a greater number and volume of transactions; in other words, a higher turnover rate of financial credits and debits. The relationship between the volume of transactions performed during the year and income determines the magnitude of the financial circulation coefficient.

4 - Finally, the index of progressiveness, presented in the simulation (which expresses the relationship among differentials of taxation and of average income among various taxable groups), shows absolute values that are increasing and greater than unit.

This indicates that higher-income sectors not only pay relatively more taxes, but they also pay at proportions that are much higher than the differences between their average income and that of other groups.”

Conceição Tavares’s simulation truly demonstrates that an electronic transactions tax is proportional, or slightly progressive. It burdens more intensely those who have greater resources.

From the standpoint of corporations (which were not included in the exercise), the author says that, the greater the volume of bank cash withdrawals, that is, the greater the circulation coefficient and the turnover rate of financial capital, the greater will be the participation of the tax on the volume of income invested in the financial system.

Concerning the impact of the tax on prices, Tavares concludes that it should not be significant and it would not trigger (as it has not triggered) financial disintermediation.

Summarizing, Conceição Tavares says that an electronic tax is desirable, given that it does not create distortions in the productive structure and is levied proportionally on taxpayers. Furthermore, it reaches the informal sector and minimizes tax evasion. In other words, bank transaction taxes are shown to be reasonably progressive taxes in their patterns of incidence, directly contradicting those who accuse them of being regressive. The tax falls more heavily on rentiers, whether “formal” or “informal”. Maria da Conceição Tavares concludes by stating,

“Financial circulation is the tax base of the future, given that, in addition to its continual expansion, it allows for electronic controls and, therefore, should allow for less tax evasion than is allowed by current taxes.”

As a direct tax, the bank transaction tax – in its formal expression – is neither progressive nor regressive; it is proportional, as long as it has a single rate. This means that for each individual transaction, the single rate would guarantee incidence that is proportional to the value of the transaction. Indirectly, as it becomes an item in production costs, it is alleged to be regressive.

Zockun M. H. estimated that the CPMF accounted for 2.2% of family income for the lowest income bracket (two monthly minimum salaries), and only 1% for families in the highest bracket (thirty or more monthly minimum salaries).

Such results were not confirmed by other estimates, such as those of Paes and Bugarin showing a virtual proportionality in the CPMF incidence by income classes, varying from 1.31% and 1.33% of family income. Using family budget research by IBGE I estimated that the tax burden of the CPMF on total expenses were the following: 1.64% for families with 1.2 minimum salary of monthly earnings, 1.58% for the 3.20 minimum salaries bracket, 1.51% for the 6.5 minimum salaries bracket, and 1.41% for the 23 minimum salaries bracket. According to estimates by the Ministry of the Economy made public in Congressional hearings in 2007 the CPMF is a “redistributive” tax, both by income classes and by regions of the country. According to the presentation 72% of the CPMF is collected by companies, and only 28% by individuals; of the revenue collected from individuals 17% are collected from the richest 10% of the population, and only 2% of revenue comes from the poorest 50% of the population. Actually it is very closely a proportional tax if we take account of possible measurement and estimation errors.

However, what really interests economists is the evaluation of tax incidence from the perspective of the complete set of transactions performed by individuals in the market. In this sense, the bank transaction tax can have a natural progressiveness, inherent to the different patterns of expenditures of the various income brackets of Brazil’s population, as shown by Tavares.

Furthermore, a more equitable distribution of national income must not be sought solely through progressiveness in taxation, but rather through the final impact of the fiscal process, which is comprised not only of the pattern of taxation, but more importantly by the composition of public expenditures, which can be progressive or regressive. The Ministry of the Economy showed that the poor northern and northeastern regions of the country collect only 24% of the CPMF revenue, but receive 42% of the CPMF collected by the federal government. Thus, the CPMF is not as regressive as claimed.

The concept of tax progressiveness has been strongly attacked by several scholars. Indeed, “progressive taxation appears to have lost much of its political appeal… people became increasingly convinced that the economic costs of progressiveness were too high to make it worthwhile”. Furthermore, “what can be done through the tax system to redistribute income … no matter how extreme such redistribution might be, is unlikely to have much effect on the overall distribution of income”.

Ives Gandra points to “the noticeable trend of European economies to begin, gradually replacing direct taxation, which has always been considered socially equitable, for indirect taxation, believed by economists to be regressive and antisocial. And the most curious consequence of this trend is that countries that have begun to reduce direct taxation have shown an increase in investments; and increasing investment is socially fairer because it generates growth, jobs, and better social conditions, facilitating the exercise of labor rights. On the other hand, progressive direct taxation (...) ultimately causes recession and inflation, with unemployment, lower wages, and less possibility for a proper dialogue on the claims of the labor class. Europe, well into the 1980s, decided openly to head toward abandonment of ideological social justice theories, which are inhibitors of development, and to begin to thread the pathways of the practical theories of international competitiveness, the only [theories] that are truly equitable from a communitarian perspective. This is the reason for which the European Union is turning to two orders of taxes; that is, indirect taxes and social contributions, gradually reducing direct taxes, including income tax.”

On this same issue, Roberto Campos addresses the question of equity in Brazil’s tax system. He says, “Our fiscal ethics have been practically destroyed. The great American judge, Oliver Wendell Holmes, said that to pay taxes is to purchase civilization. In Brazil, it means acquiring annoyance. The taxpayer has three perceptions: a) the Government does not return reasonable services, even minimally; b) the fiscal system is extremely complex, with high bureaucratic costs and three levels of corruption; c) the federal revenue agency is wholly inequitable because government-owned companies, which are notoriously noncompliant, and the entire informal economy escape the tax burden. Only one-third of the economy, represented by organized businesses in the private sector and by formally registered payroll wage earners actually pays taxes. The other two-thirds are delinquents. Thus, the estimated fiscal burden of 24% of GDP (in 1991), which would seem reasonable in worldwide terms, is abusive when it is levied solely against the formal private sector.

Not to mention, of course, the inflation tax.”

Roberto Campos continues criticizing the excessive progressiveness of taxes: “It is a socialist superstition. Everyone must pay proportionally to their income. To impose on successful persons burdens that are more than proportional to their income is simply a confiscation, only understandable if: a) the wealth is undeserved and to be punished and is not, as often occurs, the result of greater diligence and creativity; b) if the Government were inspired by Puritan ideals, with unquestionable priorities, and were not a squandering spender. The best fiscal system is that which does not punish the rich, but which preserves for each person the maximum of incentives for its productive capacity. ‘Fiscal justice’ is much better served from the expenditures side than from the revenue side of the budget.” Indeed, “any serious fiscal attempts at poverty alleviation must be undertaken primarily on the expenditure side the budget.”

Mario Henrique Simonsen reaffirms the same concern about excessive progressiveness, saying, “Today, the merits of progressiveness are strongly contested. A good portion of developed countries has considerably reduced the number of progressive rates, as well as the maximum rate. And the trend seems to return to proportional taxation, with one single exception: the exemption limit below which the taxpayer is released from any tax. The fall of the myth of progressiveness is due to several factors. First, the distribution of wealth promoted by the Government is not a function merely of a single tax, but depends on a set of taxes, and most importantly on the composition of public spending. What good is there in having an income tax that is strongly progressive if other taxes live along side it that are strongly regressive? The best thing would be to merge them into a single proportional or averagely proportional tax. On the other hand, what good is a progressive tax system if public spending benefits the rich much more than the poor? It would be better, in that case, for the budget to shrink and for the market to handle the conflicts of interest of the rich. Truthfully, the great distributive task of the Government should be handled through the operations of public spending, offering education, health, and assistance to the most needy. By doing this, the enchantment of progressiveness would, at least to a great extent, be undone.

“Secondly, excessive progressiveness simply makes the taxpayer more disinterested in working and in assuming risk, which explains the stagnation produced by the welfare state of England’s Labor Party, which was rightly dismantled by Prime Minister Margaret Thatcher. Why work harder and run more risk if the Government takes 80% of the earnings, when these are positive? In the 1970s we discovered the obvious: highly progressive taxes engender laziness. “Thirdly, progressiveness creates the incentive for transferring fictitious income from one taxpayer who has a higher marginal rate to another who has a lower marginal rate. Suppose one individual, X, whose marginal rate is 50%, is a customer of physician Y, who has a 30% marginal rate, and let us allow for medical deductions, as is usual, to be deducted from taxable income. One additional $1 in medical receipts is worth 50 cents to the customer and costs only 30 cents to the physician. The natural incentive would be a fake receipt from the physician to the customer. The customer would give the physician a check, and the physician would return $1 in cash. This example of fictitious transfer of income is merely one among thousands in a progressive tax system – and there is no screen so finely meshed that it is capable of preventing it.”

On this same issue, Prof. Roberto Mangabeira Unger goes further, asserting that,

“in the short run, and under the conditions of most contemporary societies, the progressive structure of taxation is irrelevant, when not harmful.”

Mangabeira Unger goes on, stating: “a comparative study of taxation and public spending reveals a shocking fact. There is an almost inverse relationship between the theoretical fairness of tax systems and the success each of them may actually have in funding social expenditures for investment and for income equalization. In places where there is, in fact, more redistribution, such as France, indirect and ‘unfair’ taxation of consumption serves as a major source of public revenue. In countries where inequalities are stronger and social spending is restricted, such as in the United States, homage to progressiveness in taxation prevails. (…) In the name of “fairness”, the first step must be to abolish the income tax for individuals and corporations, along with all other taxes that burden production and wages and torture the middle class. The paradox that delights the thinker bores the practical man. This is one of the reasons why the reformative actions of practical men regularly produce paradoxical effects. Both political life and the academic milieu are being exercised in the midst of edifying and tranquilizing illusions. The world is wild and obscure. To confront it one must be possessed by a passion that takes us outside of ourselves and places in our hands the trumpet Joshua blew before the walls of Jericho. Do you think, reader, that basic information such as this, about the inverse relationship between tax progressiveness in revenue and in spending, would occupy the center of the attention of scholars of public finance and tax law? You are wrong. Sheltered in their analytical apparatus, few allow themselves to be surprised by reality.”

Ultimately, the question to be answered is whether bank transactions taxes are fair. The evidence presented in this text points to the unequivocal advantage of bank transactions taxes compared to conventional tax bases, which have been fast losing efficiency as revenue collectors and as income redistributors. Further, they tend to stimulate tax avoidance, in addition to carrying high operational and compliance costs.


In this section it will be shown that usual criticism of the Single Tax concept is mostly wrong, and often biased and unfair. Below, we summarize some of the most commonly heard allegations.


The Single Tax’s structure is accused of not meeting the requirements for vertical equity. It was shown, though, that it is flexible enough to accommodate reasonable progressiveness, if so desired, by exempting operations that add to values below a given floor, during a given time period, or even by using progressive rate structures for different transaction value brackets. Though this possibility does not deserve our support, it could be easily implemented.

The distribution of Brazil’s tax burden by income groups reveals the extremely regressive nature of our present tax system. The supposedly progressive income tax only reaches the income of middle-class wage-earners in the formal sector, and fails to reach other income recipients, showing therefore a limited potential for income redistribution.

The Single Tax, by using the filter of bank transactions, inexorably reaches all types of income. It is ultimately more equitable and more progressive than, for instance, Brazil’s tortuous income tax. As for indirect taxes embedded in prices, the Single Tax is not comparatively more regressive, and induces less allocative distortions than do conventional systems for taxing consumption, as shown in this text.


The Single Tax is indisputably a cumulative tax, incident on bank transactions at each successive stage of the economic process. But this in no way discredits it as a good tax. The non-cumulativeness requirement for good tax system is a mindless fetish. No tax is perfectly non-cumulative, except those found in theoretical constructs divorced from reality.

Value-added taxes would be impracticable if they did not contemplate, as they actually do wherever they are practiced, the most diverse types of exceptions, exemptions and special regimes, which give them appreciable degrees of cumulativeness.

Brazil has a plethora of cumulative taxes, of which, curiously, some are execrated, others tolerated, while others still are fully appreciated, such as the special tax regime for small firms. The Single Tax is not different from these on this aspect, but it does exhibit the notable advantages described above – it’s simple, cheap, gentle…

Furthermore, the well-known conclusions of the second-best theory, as well as those of modern optimal tax theory, demonstrate that one cannot state a priori that a cumulative tax is less efficient than non-cumulative taxes. It is most likely that, for a given amount of revenue, a cumulative tax, which requires a low rate, is preferable to a value-added tax with a high rate, as shown in this text.

Incentive to excessive vertical integration

The weight of the Single Tax in the composition of final prices is inversely proportional to the ratio of value-added relative to purchased value of inputs, at each stage of the production process. However, incentives to vertical integration resulting from this peculiarity may be less present in the Single Tax world than in the tax structure that exists today. Coeteris paribus, the incentive to vertically integrate is obviously present in a Single Tax system, but it is a marginal thrust compared to the heavy cumulative burden that exists nowadays. One need only see that the PIS/Cofins (a former turnover tax on gross sales) alone had a statutory rate of 3.65%, and an effective rate of 3.79%.

The incentive for vertical integration in production is intensified as the turnover rate increases. If we consider the Single Tax system’s low marginal rate, it is improbable that this integration process would go beyond what can be predicted by reasons strictly related to economies of scale and other types of externalities. Verticalization beyond what would be justified in a neutral economic environment implies costs, against which the tax savings would have to be compared.

Furthermore, our simulations show that distortions in relative prices caused by the Single Tax are lower than those present in the current system, as was shown before. Actually, the decision to vertically integrate hinges preponderantly on technological reasons, such as gains of specialization and scale, compared to which the weight of the Single Tax should not be very significant.

Discrimination against domestic production

Imported products reach the final consumer after a few, one or two, intermediate domestic transactions. Thus, critics say, the Single Tax on imported products is significantly lighter than the tax load carried by locally produced goods. It could be argued that either for imports or for domestic production the tax load is relatively light compared to present levels of taxation. On the other hand, if necessary, such differences in tax regime can be compensated by customs duties, and by application of countervailing taxes on imports, as provided under international laws that govern foreign trade.

Tax exporting

It is true that it is easier to exempt taxes on exports under the value-added tax system than under the Single tax. But exempting exports it is also feasible under the Single Tax system, though it requires more elaborate procedures. Tax exonerations would have to be calculated by empirically monitoring production chains, or by using Leontief input-product matrices, and thereby offering exporters tax credits, rebates, returns, or equivalent subsidies – not too different than what is nowadays practiced elsewhere, as demonstrated in various empirical studies.

It is false that the Single Tax has an inherent anti-export bias. What hurts exports is not the existence of the tax, but rather carelessness in exempting it for exports.

Cluttering tax harmony

If most countries, including our trading partners (except, most noticeably, the United States) adopt the VAT model, and tend to avoid taxing exports or practice other types of discriminatory policy instruments, such as outright subsidies, it makes sense to search for tax formulas generally accepted in the context of foreign trade regulation.

On the other hand, the supposition that the Single Tax would be so dissonant with tax practices around the world, when compared to our trading partners’ systems, as to hamper trade and political rapprochement within regional commercial blocks is unfounded. As we have seen, the Single Tax is similar to gross sales taxes found in many countries.

The obsession with tax harmonization, seen as homogenization of tax systems, is something of a myth. In truth, the tax systems of different countries are profoundly heterogeneous for traditional, cultural, political, economic, and geographical reasons, and this has not hindered the accelerated growth of international trade.

Stimulus for banking disintermediation

At a reasonable level of taxation, the tax savings obtained by avoiding the banking system, and thereby having to carry out business transactions in cash or barter, does not compensate for the resulting additional transaction costs such as cash storage and transportation, the lack of safety, the risks of counterfeits, the illegality of foreign currency transactions, etc. To this, we add measures such as surcharges for cash withdrawals and deposits and other dissuasive measures that have been mentioned throughout this text.

Social injustice

At least 30 million individuals in Brazil do not have bank accounts, despite being participants in the active labor force. They do not qualify for bank accounts – i.e., they are illiterate or lack stable income sources, permanent addresses, property, etc. In the Single Tax world, critics claim that these individuals would be harmed by surcharges or any other penalizing measure applied to cash transactions.

We rebut this objection along two lines. First, it is possible to universalize access to the use of electronic money through the use of debit cards, even if the use of checking accounts is subject to restrictions. Second, such people already suffer from a regressive tax system, with heavy fiscal overcharge, that would decrease under the Single Tax system. Because they have a strong propensity to consume, they are victims, under the current system, of an extremely high indirect tax burden imbedded in prices. This burden would certainly be reduced as the Single Tax replaces current consumption taxes.

Excessive indirect tax burden on consumer prices

Under the current consumption tax system, which is subject to exuberant evasion, substantial portions of the price paid by consumers are taxes that, in fact, are not collected by the government. To a great extent, they are appropriated by the tax evaders themselves. By replacing current taxes with the Single Tax, tax evasion would be eliminated and the tax burden, built into prices, would be more lightly distributed. The anticipated effect would be a fall in prices, which would benefit, first and foremost, the neediest segments of the population, who spend all their income in consumption goods.

Abandoning fiscal policy

The proposed Single Tax system is accused of giving up the use of tax instruments for economic policy making. In general, however, the use of tax instruments to achieve redistributive or greater social equity purposes does not achieve the goals desired by society and by policy makers.

The Single tax actually does away with the use of tax incentives, or disincentives, as instruments of economic policy. However, policies related to prices, income, as well as growth and anti-cyclical policies can still be carried out through the utilization of non-fiscal instruments such as monetary, credit, consumer relations regulations, and even direct subsidy policies and cash transfers. All of these instruments are more transparent and more subject to social controls than obscure tax benefits, which usually do not achieve the intended purposes of tax policy makers.

The loss of tax instruments does not immobilize economic policy; it only makes it less tortuous. Policy makers must adapt to a new paradigm of economic policy, one that does not count with tax instruments for intervening in production, consumption, and investment, but which can be both more powerful and more efficient than tax incentives.

It is not our intention to abolish tax instrument in customs policy and in regulating financial markets. In many countries taxes on foreign trade are administered as customs duties, and customs administration is usually separate from federal tax administration. Foreign trade duties would not be abolished.

Single tax base benefits property and wealth

The Single Tax unifies all income and consumption bases, but allows the property base to escape taxation. Critics say that property owners would escape taxation tax more easily than would those whose savings are concentrated in the financial market. They would also escape taxation by avoiding financial transactions through the use of barter trade.

We respond that, first, barter, or “in kind” type of trade is hardly a realistic alternative in the modern world. Furthermore, this type of economic relationship can easily be avoided by proper regulation. Also, property and wealth taxation have been showing a declining trend relative to other taxes in the world. Brazil has a certain aversion for this type of taxation due to it being costly to administer and unproductive in terms of revenue collection.

But surely the Single Tax model does not preclude its concomitant use, if desired.

Weakening the federative principle

Replacing municipal and state taxes with the Single Tax inevitably raises questions about weakening the Federation. The Single Tax could not work on sub national geographical bases because it would privilege areas that enjoy high concentration of banks. This excludes the possibility of transferring the Single Tax to state or municipal control. It can only be a federal tax, and the sharing of revenue must occur according to predominantly political criteria.

Under our proposal’s approach of gradual adoption, the Single Tax would initially be implemented only at the federal level. Later, states and municipalities could also be included in the Single Tax model, thus postponing to a later time the difficult question of revenue sharing among them.

There are many models of federalism, with greater and lesser degrees of autonomy for decentralized political entities. From a strictly financial point of view, it seems that a constitutional guarantee of revenue sharing among the various federated entities, adhering to a negotiated proportionality system, would be sufficient to ensure federative autonomy. However, we cannot disregard that, according to a respectable current of jurists, the Brazilian federative model would be inseparable from the relative taxing autonomy of the federated entities, which means they must have their own taxing jurisdiction and authority, with the power to determine the variables that make up their own taxes and their respective administration.

It is not convenient to ignore the legal/political institutionalism anchored in our historical tradition. This is the reason we, when in doubt, may prefer the gradual path, with the introduction of the Single Tax in phases, beginning exclusively in the federal sphere and postponing facing the problems associated with the federative issue to when states and municipalities join in this tax model at a later point in time. This alternative abstains from suppressing state and municipal tax jurisdictions. The constitutional provisions will remain intact. Governors and legislators, both regional and local, responding to the demands of their respective populations, will decide whether to make full use of their tax jurisdictions and authority, or whether they prefer to abstain from using them, adhering to the federal Single Tax model.




bottom of page