Governing Global Tax Dodgers: The “Group of Four” and the Taxation of Multinational Corporations, 1970s–1980s

Abstract

During the 1970s, governments increasingly expressed concerns about the loss of revenue through the use of tax havens by both individuals and corporations. This article explores a covert international working group (the Group of Four) set up between France, Germany, the United Kingdom, and the United States in 1969 in response to such concerns. At regular meetings, officials exchanged information gathered by their respective tax authorities in auditing multinational companies. In the 1980s, under increasing pressure from governments in a now much more hostile climate to tax authorities, the Group’s work shifted away from multinationals and toward more general, technical questions. The history of the Group of Four illustrates the importance of the 1960s and 1970s as a period for regulating economic actors and the impact of broader circumstances on the success or failure of anti-tax avoidance measures.

In the fall of 2021, after decades of negotiations, more than 135 countries representing over 90 percent of the world economy agreed to set a 15 percent minimum tax on companies in a deal sponsored by the Organisation for Economic Co-operation and Development. The US secretary of the Treasury, Janet Yellen, heralded a “historic day” for citizens and governments across the world. ¹ The plan sought to create a more equal playing field for attracting multinational corporations and to end a half-century race to the bottom that had seen effective corporate tax rates decline dramatically, from a high of roughly 50 percent in the 1950s. ² The agreement was aimed to take the wind out of the sails of multinational companies pursuing tax avoidance strategies such as profit shifting to tax havens, the manipulation of prices paid and charged for transactions between different companies within a multinational company group (i.e., transfer prices), and abusing tax treaties. Critics quickly pointed out the shortcomings of the agreement: the rate of 15 percent was a far cry from what many advocates for reform had demanded and from the higher corporate tax rates of the midcentury. The benefits to be reaped from this redrawing of the corporate tax map would moreover disproportionately fall to high-income countries in the global north as opposed to poorer countries. Most importantly, the agreement’s fate was currently up in the air. Individual countries now have to implement the agreement and pass legislation accordingly. The process has stalled in various countries, including in the United States, where the current Congress is opposed to and unlikely to ratify the initiative by President Joe Biden and Treasury Secretary Yellen. ³

The 2021 agreement is the latest installment in a long history of efforts to join forces in combatting international tax avoidance and evasion. Most of these known efforts took place within the framework of different international and intergovernmental organizations. Looking back at a century of attempted fiscal regulation, the problem of taxing multinationals, and of international taxation more broadly, appears as a central but neglected feature of the entangled relationship between capitalism and global governance. Historians and other scholars of taxation have explored some of these efforts by looking at the work of the League of Nations in the interwar years, at the International Chamber of Commerce, the United Nations, and, from the 1950s and ’60s and going forward, at the OECD. ⁴ More recent multilateral initiatives at governing global tax matters have focused on the exchange of information to combat tax evasion, money laundering, and, after 9/11, terrorism. Based on archival materials in British and German archives, this essay reconstructs for the first time a different, less institutionalized, and hitherto unexplored example of governing global capitalism: a covert intergovernmental working group set up by the United States, United Kingdom, (West) Germany, and France (Group of Four) in 1970 to cooperate in matters regarding international taxation, primarily with the goal of sharing information and jointly investigating presumed tax evaders, especially multinational corporations. It thus reconstructs, for the first time, an important and initially successful attempt at governing international taxation and global capitalism. ⁵

Starting in the late 1960s, a strong public critique of multinational corporations and calls for intergovernmental regulation of these companies emerged. Such critiques and attempts by multinationals to fend off criticism by instead advancing proposals for voluntary self-regulation are well known. While the covert intergovernmental group explored here did not begin operating publicly for almost a decade, its work ought to be understood in the same context of rising discontent with multinational corporations. By the 1980s, in a now more hostile environment in which bodies such as the US Internal Revenue Service (IRS) faced scrutiny and criticism, the nature of the Group’s work moved away from more targeted investigations that had yielded successful results previously.

In light of the story told in this essay, the 1960s and early 1970s emerge as a crucial and somewhat underappreciated period for attempts to regulate the practices of private economic actors. It is important to note that such efforts predated the end of the Bretton Woods system, fixed exchange rates, and the removal of capital controls in many countries. Parallel efforts at the OECD and emerging ones at the European Economic Community on global tax governance suggests that it was not only the end of the Bretton Woods system and the abolition of currency controls that ushered in new initiatives and efforts to supervise and regulate flows of capital and attendant business practices. Banking, and more specifically, the gradual establishment of global banking supervision in the second half of the 1970s and throughout the 1980s, was in some ways the laggard. ⁶ In a related vein, social scientists—economists in particular—who write on tax havens and the offshore economy tend to see tax havens as a phenomenon that only began to harm economies seriously from the 1980s and ’90s and onward. These were, according to one set of authors, the “golden years” for tax havens. ⁷ The absence of hard data for earlier periods is mistaken for the absence of a more serious problem. As this essay argues, governments begged to differ. Officials were keenly aware of the growing use of tax havens among citizens and corporations, certainly in the 1960s and even during the 1920s and ’30s. In an age when the state relied increasingly on tax collection to raise revenue, this was viewed as a serious problem. Starting in the late 1960s, the loss of revenue and the distributional consequences of tax dodging were considered severe enough to warrant mustering considerable resources at the IRS and similar institutions in other countries to be devoted to numerous meetings per year, where generally scarce staff cooperated in the detection of tax dodging. This effort and the material turned up by the investigations are a reminder that tax havens and tax avoidance and evasion flourished long before the 1980s.

Attempts to coordinate tax laws and regulate taxation practices at the international level date back to the early twentieth century. Due to unprecedented budget pressures, World War I and its aftermath gave rise to the adoption of mass-based income taxes as well as taxes on corporate profits. While evidence increasingly points to people evading inheritance and other taxes in the nineteenth century, the Great War marked an important caesura in the history of taxation and tax evasion. ⁸ Wealthy individuals and corporations immediately refused to be part of the social contract that was the new progressive taxation and enlisted lawyers and accountants to help them dodge taxes. ⁹ For companies operating across multiple tax jurisdictions, it soon became clear that the absence of tax coordination at the international level offered opportunities to cheat. At the same time, with the adoption of income taxes in many countries, it was possible in theory and increasingly in practice for the same corporate income to be taxed twice: once “at home” at the seat of the enterprise and once abroad wherever business activity was taking place. Corporate interests thus began to advocate for improved international rules that would avoid such double taxation. Governments, on the other hand, feared the loss of much-needed revenue and pressed for the closing of loopholes that had opened up in the inevitable spaces between national tax jurisdictions, where individual national laws did not coherently sync up. Tax laws and fiscal systems were thus important factors in shaping global capital flows and, increasingly, the investment strategies of multinational corporations. ¹⁰

As long as income and estate taxes were low and not widely applied, lost revenue and potential double charges mattered less. With the general rise in tax rates in many countries during World War I, this changed. ¹¹ Now that fiscal revenues made up more significant portions of national budgets, governments were eager to catch tax evaders. At the same time, chambers of commerce soon became the most vocal supporters of measures to alleviate the evils of double taxation. The 1922 Genoa economic conference, preoccupied primarily with the economic rebuilding of East Central Europe after the war, took up the matter of evasion and invited the newly formed League of Nations to address the issue in conjunction with double taxation. The League’s early discussions of tax evasion centered on the possibility of information exchange among member countries, and, in this context, on whether certain aspects of bank secrecy could be lifted to help identify tax evasion. Swiss banking interests, supported by the Swiss Federal Council and British officials, made sure that such proposals came to naught. British fears evolved not so much around bank secrecy as around London’s role as the world’s leading financial capital. British officials described the concern as “we could not afford to exchange information […] without frightening foreign capital away.” ¹²

Consequently, the League’s work pivoted to the problem of double taxation in the second half of the 1920s. As part of its activities in the Economic and Financial Organization, the League created various committees of experts devoted to double taxation. By the late 1920s, the prevailing opinion was that countries’ considerable differences in fiscal systems made a general, multilateral tax convention impracticable and that the best solution would be a series of bilateral conventions. ¹³ During these years, a model treaty for the avoidance of double taxation between two countries emerged from the League’s work, which was widely adopted bilaterally in the 1930s. This network of double taxation treaties grew considerably after World War II.

During the 1930s, another crucial principle of taxation found its way into the model tax treaty prepared by the League’s experts. Double taxation created problems by potentially taxing the same income twice in different countries. But it could also lead to conflicts between two such places as to which got to claim the bulk of tax revenue, and in extension how profits ought to be allocated in corporate groups and among entities situated in multiple locations. The League turned to this question in the early 1930s. It considered two main approaches: the first, the “separate entities” approach, held that the entire company group consisted of separate enterprises, with a parent company in the “home” country and subsidiaries in other jurisdictions where the company operated. According to this understanding, tax obligations were allocated separately to each entity in the group according to the profits earned by such a hypothetically separate enterprise. The second, the “formula apportionment” approach, established the integrated profits of the entire group and then allocated taxes based on certain factors and benchmarks. The League came out in favor of the first approach, and thus the “separate entities” standard was adopted in the model tax treaty convention that would form the basis for bilateral tax agreements. At the stroke of a pen, therefore, multinational companies with subsidiaries in other countries could now treat such establishments as separate entities for tax purposes. To ascertain the taxable profits that each entity in a multinational group would be able to claim, the entity was treated as if it were operating at arm’s length, and thus independently. The “arm’s length principle” offered ample opportunity for accounting manipulations and the kind of artificial profit shifting that multinationals such as Google and Amazon continue to practice to this day. Instead of being booked in high-tax countries where they are earned, profits are transferred to other entities in the company group, conveniently located in tax havens, where they incur minimal or zero tax rates. The door had swung open wide to transfer pricing abuses. ¹⁴ The so-called separate entity approach continues to dominate international business taxation and is frequently criticized by advocates of multinational tax reform. ¹⁵

Efforts to improve the coordination of global taxation continued after World War II. The Organization of European Economic Cooperation (predecessor to the OECD), the United Nations, and the European Economic Community all at different points discussed matters of international taxation. ¹⁶ From the second half of the 1960s, governments grew increasingly concerned about tax havens and their impact on revenue levels. Lower-income countries, on the other hand, vented their frustration with what they viewed as undue interference and other transgressions of Western multinationals. The revelations about corporate meddling in the internal affairs of countries such as Chile (by the International Telephone and Telegraph Corporation, a US multinational) and general frustrations about multinational companies’ activities in newly independent former colonies led in 1976 to the establishment of the United Nation’s Centre for Transnational Corporations. All of this contributed to an atmosphere of open critique of the practices of multinational companies. ¹⁷

It was in this moment that French officials launched their initiative to create an intergovernmental forum for cooperation and information exchange in tax matters. The first contact that eventually led to the establishment of the group occurred at the French Embassy in London in December 1969, where a French official reached out to a UK counterpart to suggest a meeting over tax matters. In February 1970, an exploratory meeting was held between French and UK officials. It was then decided to involve the US and Germany as well, and the first full meeting of representatives of the four countries took place in Paris in July 1970. The Group adopted the name “Group of Four” and later also referred to itself officially as Interfisc.

The official line on what had motivated the establishment of the Group is that, at the time, the OECD’s Fiscal Committee was studying the avoidance of taxation through the exploitation of tax treaties. But the OECD work was slow, so to achieve progress more quickly, France launched the initiative that would become the Group of Four. The unspoken other part of the truth, however, was that France had always been especially galled by the Swiss tax haven at its doorstep. The fraught relationship between France and Switzerland in tax matters dates to the interwar period when French citizens pursued Swiss tax evasion opportunities en masse. By the 1960s, there was no longer just a French but also a broader international outcry at the Swiss tax haven for harboring the funds of criminals, brutal dictators, and wealthy citizens and companies all over the world. Moreover, in the United States, there was lingering resentment over Switzerland’s role in World War II and its extremely slow and tepid aid in the restitution of Jewish assets moved to safety in Switzerland during the Nazi period. Swiss multinationals, many of them in the pharmaceutical industry but also those in food production, like Nestlé, soon found themselves at the center of the Group’s interest. A clear indication that there was, in fact, an anti-Swiss thrust to the Group’s initial formation was the repeated efforts by Group representatives to deny any such attitude. Officials bent over backward at their meetings to devise strategies that would make their actions less likely to appear directed against Switzerland. One British participant said, “We do not wish to appear to be attacking the Swiss with their banking secrecy.” ¹⁸

At the center of the Group of Four was a so-called working group and a policy group, made up of members of national revenue authorities such as the US IRS and the UK Inland Revenue; delegates from countries’ Finance ministries; and sometimes diplomats, stationed at embassies in Group of Four countries, who specialized in international economic matters and taxation. At a typical meeting, each country appears to have been represented by four to six people, often men. Representatives are named (sometimes only with last names) on the participant lists included in the materials for some meetings, but it is difficult to unearth substantial information on these civil servants unless they were in exceptionally prominent positions, such as Sir Anthony Battishill (chairman of the British Inland Revenue from 1986 to 1997), Fred Goldberg (IRS commissioner from 1989 to 1992), or Dominique de la Martinière (French directeur général des impôts [Director General of Taxes] from 1967 until 1973). During the 1970s, Anne H. McNicol, undersecretary, board of the UK Inland Revenue, was the lone female participant in the Group. Only in the late 1980s and early 1990s did female delegates from the US and Germany (but not France) make a more regular appearance at meetings. One of the longest-serving Group participants was Thomas Menck, of the German Finance Ministry, who led the German delegation from its early meetings and into the late 1980s. The policy group normally met twice a year, following a preparatory meeting of the working group, in which the working group put together the materials and reports that the policy group would discuss in more detail. At each set of meetings, representatives normally focused on a topic set at the previous meeting. Delegates were tasked with preparing memoranda on certain discussion topics, which were circulated in advance of upcoming meetings. In 1982, it was decided to hold additional meetings of tax inspectors at three-year intervals. ¹⁹ Along with holding regular meetings, the Group also organized training programs in the United States and Germany for field agents who, at their respective revenue agency, specialized in international taxation. The idea was to provide these agents with a closer look at and with hands-on experience of taxation practices in the Group’s countries and of international taxation broadly. At the initial meeting of the Group in Paris, one of the French participants, the head of the French revenue authority, went so far as to suggest that in addition to exchanging information and possibly deploying tax agents to the other countries, the proposed working group could establish a tax enforcement organization that was the “equivalent to Interpol,” but such plans did not materialize. ²⁰

The Group’s existence was to be kept a secret to avoid spooking multinationals and other subjects of investigation and preventing companies from staying one step ahead of investigators. Another reason for remaining confidential was the fear of offending other OECD members who might feel excluded from such a “club.” ²¹ Soon after it began its work, questions arose about the confidentiality around the Group’s existence as well as the legality of passing on information acquired by one tax authority to another. Ultimately, it was decided to frame the cooperation as falling within the scope of the information exchange and mutual legal assistance clauses in fiscal matters that formed part of many bilateral tax treaties. While the treaties only specified bilateral exchange, the Group argued that there was nothing in the treaty framework that prevented the sharing of information among Group members. Such information had to be considered necessary for carrying out the provisions of such agreements in levying taxes in both countries and for preventing fraud and tax avoidance. As an example, information exchange often included the routine sharing of names of persons in one country who received income payments in another. ²²

At the early meetings, a decision had to be made as to the Group’s priority in the coming years. It was decided quickly that multinational corporations, and especially the fraught problem of transfer pricing and the allocation of profits in multinational groups of companies, would make for a good start. Transfer prices were hard to scrutinize in almost all cases. As one official described the problem, “the use of artificial transfer prices—unless very flagrant indeed—is difficult to spot and even more difficult to prove. A considerable amount of research into the methods of production, management, marketing and even technical details of the product concerned may be required.” ²³ Transfer prices in the pharmaceutical industry stood out and were nearly impossible to penetrate.

Pharmaceutical companies were notorious for their tax avoidance and evasion strategies. Part of it was the sheer complexity of the research and development (R&D) and production process of drugs. Manufacturing could typically range from very simple to extremely complex. Part of the manufacturing process might entail patented operations or trade secrets intended to prevent competitors from duplicating a process. The required raw materials could come from domestic or foreign sources. Some were extremely expensive and rare; some cheap and abundantly available. Some manufacturing operations thus added much more value to a product than others. What is more, it was often difficult to determine the cost of R&D that went into the creation of a drug. Last, the use of patents or trademarks for many drugs also opened doors to abuse. To assess a royalty agreement and determine if the terms were justifiable, it was necessary to factor in the general value of the discovery of the drug or process behind it, the drug’s novelty and whether it was likely to be superseded in the foreseeable future, and the length of time before the patent expired, among other considerations. ²⁴

Pharmaceutical companies had become extremely adept at exploiting these complexities. Profit shifting among different member entities of the bigger group in different countries could be achieved by paying excessive royalties to a company within the group for the use of patents or trademarks, making excessive payments for shared expenses of a common good, excessive payments for research expenditures accrued by the whole group, and making purchases and sales of ingredients at artificially increased or decreased prices among group companies. In the latter case, the actual market price for some of the more complex ingredient components was almost impossible to determine for revenue authorities who were, of course, not trained in chemistry or any other science involved in the manufacturing process. ²⁵ For all these reasons, the pharmaceutical industry was identified as a suitable target, as any possible inconsistencies as well as information on the prices of ingredients provided to authorities in the participating countries could potentially be revealed by the exchange of information that the Group was envisioning.

It so happened that many big pharmaceutical companies were Swiss. ²⁶ The Group of Four soon focused its activities on one such entity, the Swiss pharmaceutical company Hoffmann LaRoche, today commonly known as Roche. The company, founded in 1896 in Basel, Switzerland, was notorious for shrouding its activities in secrecy. The company did not publish annual data on turnover and profits, as Swiss law did not insist on such niceties. Despite the absence of such data, Roche was thought to be one of the biggest pharmaceutical companies in the world, after Merck in the United States. ²⁷ Roche’s initial focus had been on the synthetic manufacture of vitamins. It was the first to produce synthetically manufactured vitamin C. In the 1950s, the company added a promising line to its existing portfolio: a new type of tranquilizer. The success of these drugs propelled Roche’s growth in the 1960s. Two drugs were central to its expansion: Librium, introduced in 1960; and a few years later, the very popular Valium, branded “Mother’s Little Helper” in the eponymous Rolling Stones song. ²⁸ The two drugs would be at the center of international investigations into the company’s pricing policies.

Slowly but doggedly, the Group of Four began building its case against Roche. In France, Roche’s operations were based on a licensing agreement. The Swiss parent company granted its French subsidiary the right to manufacture, package, and sell certain drugs in France, but the brand names and formulas continued to belong to the Swiss company. French investigators were convinced that the royalties that the French subsidiary paid to the Swiss parent were “particularly high and absorb a substantial part of the profits of the French company,” thus reducing its tax liability in France. ²⁹ Behind this lay the problem of ascertaining what constituted a fair allocation of the R&D costs expended on drugs. Roche’s R&D activities took place mostly in Switzerland and in the United States. But a fair allocation of R&D costs would arguably have to map on to a geography of profits and be proportionate to sales levels in different countries. At one meeting, German representatives reported to have found evidence of double charging of basic R&D costs. ³⁰ A comparison of Roche’s tax filings in multiple member countries was expected to yield a clearer picture of the prices charged for raw materials, active ingredients, partly processed goods, and other products.

Roche had in fact simultaneously appeared on the radar of the UK’s Monopolies and Mergers Commission. In 1971, the Roche case had been referred to the Commission by the Department of Health and Social Security. ³¹ It was estimated that Roche held at least 85 percent share of the British market in tranquilizers. In the United Kingdom, these tranquilizers were sold by prescription and thus paid for by the country’s National Health Service (NHS). In 1967, the UK government and the pharmaceutical industry entered into what was termed the Voluntary Price Regulation scheme, under which the government could demand data on sales and profits from pharmaceuticals, thus allowing it to determine whether the prices it was paying the companies for drugs were generating reasonable but not excessive profit margins. ³² Roche was refusing to provide such information. ³³ In 1973, the Commission published a damning report criticizing the “excessive” profits that Swiss pharmaceutical giant Roche was making from selling the new and extremely popular Valium and Librium tranquilizers. According to the report, Roche was charging its affiliate in the UK far more than the market price for the two main active ingredients. The transfer price was £370 for a kilogram of Librium and £922 for valium. In Italy, where Roche’s drugs were not under patent protection, the same active ingredients could be bought from Italian companies for £9 and £20 per kilogram, respectively. ³⁴ The report also objected to the level of R&D costs the UK subsidiary paid to the Swiss parent company, thus lowering its profits in the UK. ³⁵

Upon publication of the Monopolies Commission’s report, the UK government ordered Roche to cut its prices by half. The company initially obliged but subsequently threatened to raise prices again, claiming it had been treated unfairly. After moving through the courts, the case was eventually settled in 1975. ³⁶ The company also had to pay a fee to the NHS to recover some of the “excessive” profits it had made by charging its inflated prices. The political uproar that followed the revelations about Roche led to the establishment of a special transfer pricing unit at the UK Inland Revenue. Several countries in Europe, as well the US and Australia, began their own investigations into Roche’s pricing policies. ³⁷

The Group’s investigation of Roche continued after its court battle in the UK. Other aspects of the company’s transfer pricing arrangements continued to arouse suspicions. It is unclear from the archival record whether the investigation by the UK Department of Health and Social Services and subsequently by the UK Monopolies Commission was in any way triggered by findings from the Group’s activities. Given the near simultaneous unfolding of the investigations, it is hard to see pure coincidence at work. In describing any kind of result of their activities, the paper trail produced by Group members is extremely circumspect and vague. Documents refer to the extremely useful information that Group of Four investigations were producing and to the success of its activities. But details are not provided. In the case of Roche, the Group’s documents mentioned that its activities caused Roche to adjust its numbers for several years, resulting in the payment of back taxes. The United Kingdom received £1.8 million in back taxes as a result of the coordinated Group’s investigation up through 1972. ³⁸ But other than such occasional general statements, there appears to have been a deliberate attempt to keep the details out of the preserved record. In the US and UK cases, such a practice was likely not least motivated by the fear of being “FOIA-ed,” a concern that member representatives frequently voiced after it was suspected that information about the Group’s existence might have leaked. Such a Freedom of Information Act request threatened to disclose not just potentially embarrassing details about companies’ practices but also the extent of information sharing as well as investigative practices. Hence, while it cannot be determined with certainty that the Group’s investigation and the referral to the Monopoly Commission were linked in any way, it is easy to assume that a tip-off might have taken place.

While the Roche investigation appears to have taken up much of the Group’s energies in the early years, it simultaneously drew up a list of other companies to be scrutinized, not least to avoid the accusation of singling out Roche. The other entities investigated simultaneously and in following years were the Swiss pharmaceutical company CIBA-Geigy, the Welcome Trust (UK), Roussel-Uclaf Group (France), and Pfizer (US), as well as Nestlé and another Swiss case in the food processing industry. ³⁹ Nestlé had numerous subsidiaries in France, Germany, and the US as well as a Bahamian company that, according to UK representatives, was “acting as a clearing house for the international trading operations of the group.” The company was refusing to supply any information on the role and activities of the Bahamas entity. Overall, the Group stated its investigation of the pharmaceutical industry and its transfer pricing practices had equipped members with a much more detailed and comprehensive understanding of transfer pricing that could be applied equally to other industries. ⁴⁰

Transfer pricing remained the main focus of the Group’s activities throughout the mid-1970s. In all of its investigations, Group discussions made clear that tax avoidance by multinational corporations was not a new phenomenon at all. A paper by the Group of Four called the formation of foreign companies for the purpose of avoiding taxes a “distinct post-World War II development,” and stated there could be little doubt that the manipulation of intercompany pricing, transfers of patent and other property rights, and shifting of management fees resulted in an “appreciable revenue loss.” ⁴¹ At subsequent meetings, the US representatives shared materials prepared by the IRS in the context of America’s own attempts to clamp down on tax evasion and avoidance. This effort, too, long predated the 1980s and ’90s. In the early 1960s, US President Kennedy was preparing for a major piece of tax legislation that was to curb tax dodging by US individuals and corporations via foreign companies. IRS investigations had confirmed what many suspected—the revival of American business activities in Europe as well as increasingly in the non-Western world in the 1950s and ’60s had led to a major increase in tax evasion and avoidance. The information shared with the Group painted a detailed picture of the emerging geography of transfer pricing and tax evasion in these crucial decades, some of which continues today. It also detailed the practice of what transfer pricing can achieve, and why the arm’s-length principle and separate entity accounting continues to be a gift to tax-avoiding multinationals.

In the 1950s, Panama and Venezuela were favored tax haven destinations for setting up companies that would allow North American businesses to avoid taxes. As one example of a common practice, in 1947 a Canadian citizen, owner of a Canadian manufacturing plant, and a US distributing company, together with a US citizen, organized a Panamanian corporation. Prior to 1947, the Canadian manufacturing company sold its products to US customers through the related US company. After setting up the Panamanian company, the Canadian product was first sold to the Panamanian company, which in turn sold to US customers. Sales prices were arranged so as to keep the profits of the US company artificially low. This was done by setting prices slightly above cost on the sales from the Canadian company to the Panamanian company. The Panamanian company then sold at abnormally high prices to the US corporation; thus, almost all profits accrued to the Panamanian company instead of the US one. As Panama did not tax profits from business activity that had taken place outside of the country, the profits went tax-free. ⁴²

Even among these earlier cases, pharmaceutical companies featured prominently. In 1955, a US pharmaceutical company had set up a Panamanian subsidiary that it used to keep profits out of the United States. That company was likely Pfizer. ⁴³ The original domestic US company formed a new Panamanian holding company, which itself formed foreign trading companies in Panama. The foreign trading companies took over the foreign sales from the domestic US pharmaceutical company, thereby shifting profits to the foreign trading companies. These companies then passed the untaxed profits to the Panamanian holding company as nontaxable dividends. The holding company subsequently used the funds to expand manufacturing operations abroad. Shortly afterward in 1956, the US pharmaceutical company also formed a Swiss company for the purpose of holding patents. It transferred foreign rights to patents and licensing agreements with foreign manufacturers and distributors to the new Swiss entity. New licensing agreements for the foreign manufacturing of a drug invented and produced initially by the US company were then granted by the Swiss patent-holding company. The following illustrates why such arrangements can be characterized as artificial rather than representing real sites of business and management activity. The employees of the domestic US pharmaceutical company handled all the negotiations of licensing agreements with foreign manufacturers and provided the technical know-how and services. The Swiss patent-holding company was therefore reimbursing the domestic US pharmaceutical company and funneling some of the money back to the US. The income from foreign licensing agreements for drugs was made possible not least from a very substantial R&D program that the US pharmaceutical company maintained in the United States. Yet as a result of all these artificially construed sales and compensation arrangements, the foreign licensing income was realized by the Swiss patent-holding company alone and remained untaxed in the US and “subject to minimal taxes abroad.” ⁴⁴

Since the 1950s, American-based Pfizer Inc. also had a flourishing operation in Panama at the Colón Free Zone as one of the first US corporations to use the newly established site. Gillette and the North American Tobacco Company (maker of Lucky Strike cigarettes) had already done so or were in the process of establishing manufacturing capacity. The Colón Free Zone was a tax haven in the tax haven of Panama, set up in 1948 and designed to attract light manufacturing under a regime of tax and tariff incentives, one of a number of such zones that would proliferate in so-called developing countries in the coming decades. ⁴⁵ It was difficult to determine whether the profits made by the Panamanian company on goods purchased from the US company were unduly high, since 90 percent of the goods were further processed by Pfizer in the Colón Free Zone. US tax authorities were asking Group members for information that could help them corroborate the prices charged by Pfizer to its Panama subsidiary. ⁴⁶

Transfer pricing was particularly attractive when it combined tax havens with tax treaties. Early on in the proliferation of tax treaties, companies had engaged in what is termed treaty shopping, which refers to the practice of arranging business in such a way that made it possible to take advantage of the benefits offered in tax treaties without being entitled to such benefits through genuine business activity. As an example, a US parent company paid interest to an Italian subsidiary but routed said interest through a Swiss company to take advantage of the low 5 percent tax rate on interest offered through the US–Swiss treaty. If the US company had paid interest directly to the Italian one, it would have been taxed at the rate of 30 percent. “Treaty shopping” applied here because the Swiss intermediary was set up solely for the purpose of accessing the US–Swiss tax treaty and not as a reflection of genuine business activity in Switzerland. ⁴⁷

The history of the Netherlands Antilles offers the most interesting example of a tax haven that crafted a deliberate strategy to offer treaty benefits. Here too, it is evident that the Antilles were facilitating corporate tax avoidance from the late 1950s and certainly through the 1960s and 1970s. The Group of Four extensively and repeatedly investigated the “Antilles Route,” as it came to be known, among other examples of treaty abuse through Switzerland. The Netherlands Antilles was a set of island territories located in the Caribbean Sea off the coast of Venezuela, forming the Dutch colony of Curaçao and Dependencies. In 1954 the islands became known as the autonomous (but not independent) Netherlands Antilles. Curaçao attained economic importance in the Dutch Empire when oil was discovered in Venezuela in the early twentieth century. Royal Dutch Shell built a refinery on Curaçao, quickly becoming the company’s largest, and the world’s third largest, by 1938. Shell’s Curaçao connection suddenly acquired new meaning when Nazi Germany’s drive toward expansion and war threatened its European business based in the Netherlands. The oil company moved its legal domicile to Curaçao. In February 1940, Shell established an office in Curaçao. Following the German attack on the Netherlands in May of that year, the formal deed moving Shell’s domicile from The Hague to Curaçao was signed immediately. Two years later, some 140 companies had established similar offices. While the purpose of these company registrations was different, they certainly resembled later shell companies set up with the goal to avoid taxes. The boards and most business operations of these companies remained abroad, with merely administrative formalities being carried out in Curaçao. When the war ended, and after the sequestration of Dutch assets by the Allies was no longer a threat, these companies wound up their Curaçao activities and returned to the Netherlands. ⁴⁸

The wartime move to Curaçao had unintentionally put the island on the corporate map. When tax rates increased in many countries after the war, bankers recalled the escape that multinationals like Shell had made. One such bank was Nederlandsche Handel-Maatschappij (NHM). Its intentions were clear. “Wealth taxes” were scaring capitalists, and the solution was to protect investors from what the bank dubbed “fiscal difficulties” by placing their assets in a trust or holding company in a low-tax jurisdiction like Curaçao. The NHM, after successfully lobbying the Dutch government in the Antilles, achieved its goals: companies deriving their profits from business activity outside the Antilles had their tax rate lowered to between 2.4 percent and 3 percent. ⁴⁹

Initially, business was slow. The Antillean government soon realized it would have to do more to get its tax-haven aspirations off the ground. It therefore launched an effort to use tax treaties to encourage the flow of investments to Curaçao shell companies. But Curaçao could not conclude such treaties by itself; it needed the Netherlands to take care of that. In 1951, the Antillean government asked to join the tax treaty that had been concluded between the Netherlands and the United States. Dutch accounting firms, as well as the Dutch multinational company Unilever, closely watched the negotiations between the Dutch Ministry of Foreign Affairs and the US side in Washington, DC. The agreement reached in 1955 exempted royalties and interest payments to Curaçao shell companies from the 30 percent tax normally withheld at a source in the US and also lowered the tax on dividend payments to 15 percent. The effect was swift: shell company registrations increased from 180 in 1956 to 400 the following year. The results quickly became so significant that the US explored its options to terminate and subsequently renegotiate a different treaty agreement because of the loss of tax revenue it was experiencing through the Antilles route. ⁵⁰

In 1952, the Antilles also successfully sought to join the tax treaty between the Netherlands and the UK. Signed five years later, it came into force retroactively in 1953. In 1965, the Antilles government moreover rearranged its own fiscal relations with the Netherlands. It got The Hague to exempt dividends sent from a Dutch company to Curaçao from taxation at the Dutch source. ⁵¹ The Netherlands thus began morphing into the corporate tax haven that it continues to be today. Large Dutch multinationals such as Unilever, Royal Dutch, and Philips were paying close attention to the process and, likely, sought to influence the negotiations. The following illustrates the opportunities for tax avoidance. A Dutch company earning patent royalties from Sweden would normally pay 48 percent corporate taxes on such income, if received directly. But if the company set up a subsidiary in Switzerland that owned the patent and receives royalty payments there, the royalties would be taxed at 8 percent. The Swiss subsidiary could then pass the tax-free proceeds to a Dutch shell company on the basis of the tax treaty between the Netherlands and Switzerland. The shell company would then pass the royalty income to a Curaçao shell, where it would be taxed at the rate of 3 percent. ⁵² The Netherlands thus began to build one of the largest networks of tax treaties with other countries, offering low taxes on dividends and royalties on both ends. The treaties propelled the movement of funds via the Netherlands to Curaçao and generated revenue for Dutch legal, financial, and accounting services.

When not focusing on transfer pricing, the Group sometimes investigated cases involving individuals, often celebrities. Here, too, the hope was that other Group members might furnish useful information that would complement the efforts of individual countries. One such case was Mr. X, or “Monsieur X,” who had not filed a tax return in France since 1962, despite working as a journalist and eventually editor in chief of a “women’s magazine.” Mr. X claimed to be resident in Ireland but, according to French authorities, he never spent more than three weeks per year in that country. Mr. X used an apartment of 250 square meters in Paris’s 16th Arrondissement, where he lived with his spouse, who was a famous actress. The apartment was technically owned by a Swiss company, a “screen designed to conceal his presence in France.” For the entire period investigated, he owed taxes and penalties amounting to almost 3.8 million French francs. ⁵³

Another such case concerned The Church of Scientology. After its arrival in the UK, it successfully requested to be classified as an educational establishment, which allowed it to claim a “charity exemption” in tax returns. During informal talks among Group members, UK authorities learned that US officials had carried out a widespread investigation into the affairs of Ron Hubbard, founder and head of the Scientology movement. The US voluntarily shared this information with the UK. It must have suspected that a large share of the “royalties” paid to Hubbard by various Scientology organizations and subgroups in the US were being directed to accounts at a Swiss bank in Zurich. The US provided the UK with information on these bank accounts and payment schedules, most of which passed through the Provincial Bank at East Grinstead, where Scientology UK was headquartered. ⁵⁴

The Group occasionally ventured into the world of international entertainers, artists, and athletes as well as celebrities leading highly international lives. Under so-called rent-a-star arrangements, artists would receive their payments at companies set up in Liechtenstein and often claim to not fall under the minimum residency rules of any country to qualify for tax status there. Rudolf Nureyev, the famous ballet dancer, was one of several artists under contract with a company called Interart Establishment in Vaduz in the tax haven of Liechtenstein. The company functioned as an artist’s employer and thus received that person’s income in such a way as to escape taxes where the artist resided and that also avoided withholding taxes in the country where the payment originated and the artist had performed. The exchange of information among Group members led to significant back taxes and penalties for Nureyev. The Rolling Stones band was investigated for similar practices in Germany. ⁵⁵ France requested information from other Group members on the actors Catherine Deneuve and Jean-Paul Belmondo and on the composer Francis Lai. In turn, France provided data on the Rolling Stones, singer-songwriter Joan Baez, film producer Victor Pahlen, and film director Anatole Litvak to other Group members. ⁵⁶

By the late 1970s and beginning of the 1980s, the Group’s discussions changed. As soon as the Group had begun its work, questions arose about its confidentiality as well as the legality of passing on information gathered by tax authorities in one country to counterparts in another. With the existence of the Group to be kept out of the public eye, members worried about arousing the suspicion of companies under investigation. The reason was Roche. After its company lawyers were presented with questions from member countries’ revenue authorities that clearly suggested cooperation, they demanded an explanation. German officials soon thereafter received a letter from the Swiss tax authorities (which never failed to support Swiss tax subjects in their tax avoidance and evasion as long as the damage occurred abroad), alleging a “witch hunt” was being directed at Swiss companies. ⁵⁷ German authorities subsequently had to withdraw from the Group’s Roche investigation, fearing a battle in court that the German Ministry of Finance was likely to lose. In coming years, German delegates frequently warned the Group of the strong legal protections enjoyed by German taxpayers, often making German officials hesitant to undertake investigations that might prompt lawsuits.

By 1976, the existence of the Group was widely suspected, but it was agreed that no public announcement should be made just yet. ⁵⁸ Different national authorities had made vague references in statements about the international exchange of information on tax matters but had so far refrained from explicitly revealing the Group’s existence. French representatives, most vehemently opposed to acknowledging the Group’s existence, suggested they encourage the formation of similar groupings among other OECD countries. Once created, they could provide cover for the Group of Four’s activities by allowing the Group to claim it was merely doing what other OECD countries were doing. ⁵⁹ The Group’s activities were first leaked to the German press, seemingly via a member of the German Ministry of Finance who spoke out of turn. In April 1974, after statements appeared in the German press (and reported in other countries, including the UK), the German Finance Ministry confirmed that the four countries were cooperating in tax matters but did not detail or confirm the existence of the Group. ⁶⁰ The official draft announcement, collectively disclosing the existence of the Group to the press, included that a study group named Interfisc had been set up by the IRS in the United States and finance ministries of the UK, Germany, and France to study profit allocation at the international level, “in particular the intercompany pricing policy of multinational companies.” Membership personnel consisted of the heads of treaty sections at the US agency and European ministries who were experts in the tax audit divisions. The Group’s goal, moreover, was to share experiences in auditing multinational companies. ⁶¹ The US, UK, and Germany did not make their official announcements until late in 1977. By the next meeting, France still had not done so. ⁶²

With the Group’s existence now widely known, at least in professional and corporate circles, it appears that its work gradually began to change tack. Investigations into select industries like pharmaceuticals and, even more so, individual enterprises, mostly disappeared from the agenda. What survived from the Group’s early days were in-depth studies of certain common patterns and techniques of avoidance (often involving specific sites such as the Netherland Antilles), in which all group members compiled memos on their experience with said schemes. But, by and large, the Group’s work veered away from the investigation of specific cases and toward more technical matters, such as comparisons of rules and definitions of known critical issues in member countries, ranging from thin capitalization rules (e.g., how much interest on corporate debt can be deducted from taxes) to definitions of domicile and residence and other concepts that differed across national tax traditions. Early on, the Group also explicitly dealt with tax havens and their role in facilitating tax dodging. The topic remained on the agenda for the period covered here. It also became common for Group members to share and discuss descriptions of new anti-avoidance legislation that individual countries passed.

The reasons for the Group’s changes in agenda and approach are likely multifold. For one, member countries feared lawsuits brought by companies questioning the legal basis for information exchange. Something like the Group’s meetings and its early simultaneously targeted audits of individual companies was not explicitly what the rather general mutual assistance clauses in tax treaties had envisioned. Group members from the US and UK were mostly confident that national courts would decide in their favor if a case was brought. French Group members saw no chance for defeat of its revenue authority in court. But German members, because of the country’s strong post-1945 tradition of privacy laws and defense of individual freedoms, were concerned. As noted earlier, during the Roche investigation, the company’s lawyers threatened to go to court to prevent German officials from handing over evidence to France as a Group member, leading German Group representatives to withdraw from the Roche investigation over fears that Roche would very likely be successful in a lawsuit. In the Group’s exchanges, it shows that Roche, true to form, had threatened France to move a planned plant to Germany instead if investigations into the French Roche subsidiary continued, after previously making the same threat to German officials. ⁶³ Later, German plans to implement the European Community Directive on mutual assistance in tax matters caused “a significant degree of controversy” in that country for identical reasons. ⁶⁴

Concerns about data gathering and privacy grew more pronounced in all Group member countries in the 1980s. Fears over computerization and the resulting longtime storage of personal information and the threats thus posed to privacy meant that revenue authorities had to tread lightly on information sharing. Tax data was one area in the organization of information that was changing dramatically because of the use of more powerful computers. At a series of meetings in the late 1970s and early 1980s, German tax officials presented a new computer system, using information it had collected and subsequently digitized on companies set up in Switzerland by German citizens, likely for the purpose of tax evasion. The information had to be acquired old-school style by looking through a list of new company registrations published weekly in Switzerland in the Handelsregister (companies register). Once manually entered into the new computer, the German members were able to quickly sort and visualize information in ways that appear quotidian and that we would take for granted today but that were nothing short of groundbreaking at the time. For example, data could be sorted to quickly show all companies registered at one particular address in Zurich (usually a lawyer’s office). It could also be sorted to show street-level data on clustered company addresses in certain parts of Zurich. Finally, the data could be arranged by the name of the person(s) serving as director(s), revealing the names of lawyers primarily involved in a company’s registration business. ⁶⁵ But such advances simultaneously led to calls for closer monitoring of revenue authorities’ practices. As one Group member put it, “The increasing use of computers for the storage and processing of information about individuals and enterprises has given rise to great anxieties about the provenance, relevance, accuracy and up-to-dateness of the information held, its confidentiality and the use made of it.” ⁶⁶ In reflecting on the Group’s past work, an internal memo from a UK member spoke of “problems of confidentiality as well as the natural development of the Group’s work” that led to a “change of emphasis” over time away from investigations of individual industries. ⁶⁷

It is clear, moreover, that the changing political circumstances of the 1980s did not go without effect on the Group’s ability to carry out its work. One British representative at the September 1979 meeting in Cambridge, Massachusetts, referred to the UK’s new government under Prime Minister Margaret Thatcher, saying it clearly “wanted to cut the Inland Revenue down in size and also down to size,” and that “there was some threat to reduce the quality of enforcement procedures.” The UK’s Group members would therefore have to proceed with great care. ⁶⁸ In the same context, another representative from the UK cautioned the Group to steer clear from any initiative against tax havens. Reporting on discussions among British revenue officials, the representative explained: “It was said we had to wait for the right moment and that was not now. The thrust of political activity in the UK was the other way. […] The present political climate was not ripe.” ⁶⁹

All in all, by the mid-1980s, the tone of conversation among Group representatives had changed considerably. Documents now discussed at meetings included a British “Taxpayer’s Charter” [of rights] of July 1986 that informed taxpayers that they were entitled to “help and information,” “courtesy and consideration,” “fairness,” “privacy and confidentiality,” and “independent appeal and review,” among other descriptions. In France, President Jacques Chirac set up the Committee on Taxpayers’ Rights in 1986. In the United States, according to one document, the US was considering legislation at the time that would have made IRS employees personally liable if a taxpayer’s rights were violated in the process of an audit. These were, of course, not new features of the relationship between British and French taxpayers and their respective tax authorities. What is remarkable is that it now appeared necessary to reassure taxpayers, the public, and likely those holding political power of such entitlements. ⁷⁰ As one British report put it, “in part these concerns reflect a general shift which is taking place in public attitudes to the powers of tax authorities, with more emphasis being given to taxpayers’ rights.” Such opposition to the exchange of information coexisted with more mundane corporate greed and concerns in the business world about how exceedingly effective cooperative measures generally were. Some multinationals openly threatened “to move operations to other territories where they are more readily able to conceal profits” if information exchange proved too effective. ⁷¹

There had also been a generational change among those attending meetings. Some of the initial representatives who had forged working relations with their colleagues over some fifteen or so years appear to have retired and were replaced by different officials. In some instances, representatives themselves, whether seasoned or new, appear to have become cynical about the work of the Group, deriding other countries’ efforts and sneering at the behavior of other meeting attendees. ⁷² One UK participant described a “particularly frustrating” meeting in 1985, “largely because the United States tended to confine itself to statements about its own internal organization […] leaving the floor to the Germans who were much concerned with trivialities and the French who were not much concerned with anything at all.” At the same moment, others were suggesting changes to the meeting format and frequency—but after diverging from established practices, the Group returned to its previous routines. Despite such occasional wobbles, meetings continued. ⁷³

Moreover, similar programs slowly developed to complement the Group of Four’s activities. Arguably, the Group of Four had set a precedent for such cooperative work. On multiple occasions, other countries approached the Group with requests to join, but all of them—Japan, Canada, Italy—were turned down. Keeping the Group small was said to likely prove more efficient. ⁷⁴ It was only the Nordic countries—Denmark, Sweden, Norway, and Finland—that had established a similar group for exchanging information on tax matters earlier than the Group of Four. But it was the Group of Four that became a template for similar multilateral groupings and bilateral spin-offs. In 1980, the United States entered into a Pacific equivalent group with Canada, Japan, and Australia for fostering tax cooperation, and it met for the first time in Honolulu in 1981. The Pacific Association of Tax Administrators apparently still exists. ⁷⁵ The United States and other initial Group member countries also formally established so-called Simultaneous Examination Programs, under which concurrent investigations into individual enterprises could be carried out. The first such program was established in 1977 between the US and Canada, with following programs with the UK (1978), Germany and France (1979), and Norway (1981). ⁷⁶ By 1985, such multilateral and bilateral forms of information sharing had become so widespread that the “leading international accounting firm” Arthur Andersen, as British officials characterized it, could publish a booklet on exchanges of information in tax matters, “in which it advised companies to proceed on the assumption that data available to one tax authority would be communicated to other authorities with a potential interest.” ⁷⁷

Apart from initial announcements about its existence, the Group does not appear to have attracted much publicity. It is therefore difficult to ascertain whether it still exists, and in what form. The archival trail on the Group’s meetings runs dry in the British archives in 1988 and ends in the German archives with a 1991 meeting in San Francisco. The San Francisco meeting materials reference plans for one or more meetings in 1992. ⁷⁸ Given that these years come up on the 30-year embargo policy in many archives, it might be that subsequent materials have simply not yet been released into the archives. Other multilateral initiatives to establish global tax regulation picked up in the 1990s due to concerns about money laundering and other illegal activities conducted through offshore jurisdiction. The events of 9/11 added concerns about terrorism funding and led to further steps toward sharing information and cooperation, to ultimately culminate in the current efforts underway to institutionalize a global minimum tax. ⁷⁹ The long history of efforts to regulate global taxation makes abundantly clear that the success or failure of regulatory efforts depend not only on their design and implementation but also on the historical circumstances in which they operate. The changing attitudes toward revenue authorities in the 1980s and the resulting shift in focus and practices among the Group of Four illustrate such a dependence.

The global minimum tax initiative is novel in that it lets go of the idea that lower tax rates will reduce incentives for avoidance and evasion, a notion peddled frequently by both politicians and economists. In theory, the logic of the race to the bottom is taken out of global tax competition under current reform plans. Companies paying a tax rate lower than the new global minimum tax (set admittedly at a shamefully low level of 15 percent) in any jurisdiction—including tax havens—will simply have to top up to the minimum level. Previously, when the Trump administration in the United States passed its so-called Tax Cuts and Jobs Act (TCJA) in 2017, including a reduction in the corporate tax rate, it was touted as an instrument that would “bring home” multinational companies that had stashed away cash in tax havens. With the top corporate tax rate lowered from 35 percent to 21 percent, multinationals simply would no longer be incentivized to rout profits through tax havens as they had done for the past decades. Economists have now begun to present tentative evidence of what actually happened after the 2017 reform, and their findings are not encouraging. The TCJA did not result in a large-scale repatriation of funds, and worse does not appear to have reversed multinationals’ use of tax havens for profit booking except in some exceptional individual cases. If anything, the amount of profits routed through tax havens appears to have plateaued at its highest level. ⁸⁰ The history of intergovernmental efforts to curb tax avoidance in the twentieth century, including at the Group of Four, underscores that the policy logic underlying the TCJA and similar arguments is deeply flawed. Tax avoidance and evasion were serious enough in the 1950s and ’60s to compel governments into action. For the following decades, officials deemed it necessary to revert to unusual measures—among them setting up a covert working group—to stem the flow of money offshore. During these years, corporate and individual tax rates in high-income countries reached their highest level, as is well known. But when rates were slashed in the 1980s, this did not lead to a decrease in tax avoidance. To the contrary, in the face of lowered tax rates, avoidance and evasion increased significantly. The driving factor was the ease with which capital could move across borders after the final removal of capital controls in many countries in the late 1970s, not the rate of tax. Multinationals as well as individuals continue to avoid and evade taxes at record levels long after the era of high midcentury tax rates faded Acknowedgement: I wish to thank the anonymous reviewers and editors of the BHR for their helpful feedback and guidance. ⁸¹