• A Introduction

  In early 2015, Brazilian Senator Fátima Bezerra submitted to Congress Bill No. 49/2015 (the ‘Brazilian FBP Bill’), a proposal to introduce a Fixed Book Price (FBP) policy in Brazil. An FBP policy is in essence a state-sanctioned, countrywide Resale Price Maintenance (“RPM”) arrangement.¹xThroughout this article, when referring to ‘RPM arrangements’ and/or ‘FBP regimes’, we are not drawing any conceptual distinction. They are synonyms for the purpose of this article. Indeed, from a competition policy standpoint, RPM arrangements and FBP regimes are different in only one main aspect: their extent. FBP arrangements are state-sanctioned and applicable to the whole relevant jurisdiction, whereas RPM arrangements are usually implemented by certain market players and assessed from a competition standpoint individually. In the case of a publisher-implemented FBP regime, this means the whole downstream market – i.e., bookstores, and all other retailers – are subject to the specific pricing policies of the upstream market. As noted in the following, we however do not focus on this aspect of FBP regimes. We treat FBP regimes as minimum RPM arrangements, and assess them from an antitrust standpoint based on that premise. The Brazilian FBP Bill is, to a large degree, inspired by France’s FBP law, in force since the 1980s. France’s FBP law came into force as a policy aimed at fostering access to books and, ultimately, at encouraging people to read more. At the time of writing, several countries had FBP policies in force, including Argentina, France, Germany, Greece, Italy, Japan, Lebanon, Mexico, the Netherlands, Norway, Portugal, South Korea, Spain, Sri Lanka and Thailand.²xBrazil’s current legal framework already contains a provision that expressly allows publishers to fix the purchase price of books (Article 60 of Law No. 9,610/1998). By way of this provision, publishers may fix the price of books with retailers via supply contracts (at least in theory). However, due to the increased risk of antitrust scrutiny by the Administrative Council for Economic Defense (CADE), Brazil’s competition enforcer, Brazilian publishers have refrained from fixing prices, choosing instead to simply recommend them to bookstores and other companies of the retail segment.
  In 2017, as Brazil’s political climate seems to be heading towards more stability, the Brazilian FBP Bill is once again entering the spotlight of Brazil’s social and cultural policy debates.³x See, e.g., a discussion of the Brazilian FBP Bill in Brazil’s most influential newspaper, Folha de São Paulo, available at: <www1.folha.uol.com.br/ilustrada/2017/04/1871549-autores-nunca-foram-tao-influentes-diz-ceo-da-editora-harpercollins.shtml>. On August 23, 2017, the Brazilian FBP Bill surpassed a significant milestone by winning the approval of Congress’ Commission on Constitution and Justice, and is now set to be voted in a plenary session.⁴x See the official press release reporting this development, which is available at: <http://www12.senado.leg.br/noticias/materias/2017/08/23/ccj-aprova-criacao-de-politica-nacional-para-precos-de-livros>.
  Behind FBP regimes is the idea that a book is a different type of good. Those that defend these regimes argue that it should not be commoditized and its sale should not be solely subject to the rules of free markets.⁵xIn Olczak’s words: “RPM has been tolerated in this market predominantly because of the cultural importance of the product (see Ringstad, 2004) and for example the judgement in the 1962 UK Restrictive Practice Court decision to allow RPM to continue emphasized that ‘books are different’” (M. Olczak, ‘The impact of recommended retail prices on bilateral bargaining and retail pricing’, October 2011, p. 2). Moreover, according to the International Publishers Association, “Supporters of FBP point out that books have special value; they are indispensable to our individual development, as well as to society. As objects of culture, they therefore deserve to be treated differently from other tradable commodities.” International Publishers Association, Global Fixed Price Report, available at: <www.internationalpublishers.org/images/news/2014/global-fixed-price-report.pdf>. The reason for this is that the value of books is arguably not merely material. They are tools for one to access knowledge, a new world, or to escape our own. They therefore have a cultural and educational value, and compatible with several provisions of modern constitutions. Thus, a main premise of an FBP regime is that a dense network of well-stocked bookstores is a necessary condition for a large variety of books to be published.⁶x See M. Canoy et al., ‘Chapter 21: The Economics of Books’, in Handbook of the Economics of Art and Culture, Vol. 1, Amsterdam, Elsevier, 2006, p. 742.
  Fostering access to books has been a policy priority for a number of governments ranging across the political spectrum. Generally, in these countries, the legal system establishes that governments must stimulate the book market’s supply chain, both at the retail and at the publishing level. This comes either from a specific interpretation of vague constitutional provisions dictating, for example, that access to culture is a social or individual right. They may also come in the form of more specific provisions, which leave less room for debate on the means (which policies?) that are needed to achieve certain ends established by law.
  Regardless of the manner in which the legal system encourages a country to put in place an FBP regime aimed at fostering access to books, the fact is that these policies have, at least since the mid-20th century, been subject to constant scrutiny of a mixed nature, based both on technical and ideological arguments. The problem is that scholars and policymakers face difficulties in trying to separate technical arguments from ideological ones when examining the pros and cons of adopting FBP regimes – especially since, in many cases, they tend to be interdependent, and come in the common genre of “violations to competition law.”
  A significant portion of regulators, academics and policymakers argue, with the support of standard antitrust economics applied to RPM arrangements, that FBP regimes are harmful to competition and reduce individual and social net welfare. Supporters of FBP, on the other hand, argue, also drawing on antitrust analysis, that absent an FBP regime, the book market tends to become more concentrated – which may also be detrimental to competition. The reasons for countries to adopt FBP policies are not necessarily compatible with the goals of standard antitrust policy. Indeed, they include an acknowledgement of the importance of (i) structuring conditions for bookstores to maintain themselves in the book industry, to raise the amount of books available in the market and stimulate reading habits; and (ii) encouraging literary production, creating incentives for bibliodiversity.
  Those that are pro-FBP argue that, without such a policy, publishers would not have the right incentives to invest in books of less commercial appeal, and that small, niche bookstores would not survive in the market. The book supply chain would end up comprising only large retail chains selling blockbuster books. This – the argument concludes – would undermine a country’s cultural and social development. On the other hand, those that do not agree with FBP policies argue that, in addition to there not being that much of an obvious link between fixed prices for books and an increase in the number of both book titles and bookstores, FBP means raising prices for consumers, to the detriment of their welfare. Indeed, they argue, the idea that bookstores would not have the freedom to give consumers discounts, or compete with rival firms on prices, would only serve to create obstacles on an individual’s access to books, rather than reduce them.
  In this confusing context, important questions emerge. How can policymakers be sure that these arguments that draw upon antitrust economics are applicable to their particular social and institutional environment? Are there specific needs in developing countries that an analysis based solely on standard antitrust economics overlooks? Is standard antitrust economics the appropriate tool to address all welfare issues associated with FBP regimes? Do FBP regimes make more sense in some countries (e.g., those where reading habits are low) than in others?
  There seems to be a rift between the reasons for implementing FBP regimes, and those used to scrutinize them. The reasons for their implementation involve values like culture, individual emancipation, distributivism, social development, and equal access to essential goods for human well-being. These values to some extent contribute to consumer welfare, which is arguably the ultimate goal pursued by antitrust policy (the argument is that if culture were not important to consumers there would ultimately be no private-owned bookstores, theatres, movies, museums, and other important cultural amenities around). However, they are – at least apparently – not entirely coincidental with the traditional notion of consumer welfare, as defined by antitrust economics, which focuses blatantly on price, output quantity, quality, and enhanced consumer choice.⁷xFor an in-depth compilation of what constitutes ‘standard antitrust analysis’, see: L.M. Kahn, ‘Amazon’s Antitrust Paradox’, Yale Law Journal, Vol. 126, No. 3, January 2017. Thus, while reasons associated with the goals of social regulation seem to be the backbone of FBP policies, arguments coming from the field of economic regulation are commonly used to contest them.⁸xA sophisticated distinction of the different values associated with ‘social regulation’ and ‘economic regulation’ can be found in: E. Windholz & G. Hodge, ‘Conceptualising Social and Economic Regulation: Implications for Modern Regulators and Regulatory Activity’, Monash University Law Review, Vol. 38, No. 2, 2012. In order for the debate surrounding the pros and cons of implementing FBP regimes to become more productive, it would be important to build a bridge between the reasons for designing and enforcing such policies, on one side, and standard antitrust analysis, on the other.
  There are many interesting arguments at the table that both corroborate and compromise the case for an FBP policy. However, what a cross-border examination of FBP regimes sheds light on goes a step further: throughout history, these policies have experimented cognizable trends. The objective FBP regimes pursue and their design have changed subtly, yet relevantly throughout history. In our view, the current academic and public policy debate surrounding FBP regimes, in both countries considering adopting or revoking them, would benefit from an enhanced awareness of these trends and their policy implications. Ultimately, so would the antitrust analysis of these policies. We argue that a better grasp of these trends could potentially result in a more sober examination of the welfare risks associated with FBP policies.
  Although for didactic purposes the ‘economic regulation’ and ‘social regulation’ distinction is useful, most public policies implemented in the real world embody a very complex mix between these two types of regulations. Our historical examination in the first part of this article suggests that FBP has shifted from a contractual agreement focused on advancing industry interests into what can be considered a textbook example of a mix between social and economic regulation. This finding can shed light on the debate surrounding Brazil’s FBP Bill. We will seek to analyse if, and to what extent, social and economic regulation (i.e., antitrust concerns) coexist in the Brazilian FBP Bill. Whether or not the values of social or economic regulation should prevail in the context of FBP is not the core question of our investigation. It is rather whether there is an explicit, acceptable balance between social values and economic values in the proposed Brazilian FBP Bill.⁹xFor the purpose of our analysis, we will rely on the theoretical framework set out by Windholz & Hodge, 2012, which we will discuss in details below.
  This article has three remaining sections. (i) We will first discuss the context that lead FBP policies in certain major jurisdictions to be implemented, elucidating and discussing specific trends. (ii) Informed by the conclusions of the first section, we will turn to examining the Brazilian FBP Bill, seeking to identify how social and economic values are balanced (if at all). In order to do so, we shall rely on antitrust literature, on examples of countries in which FBP regimes yielded positive results (i.e., results compatible with the goals set by the policy), and on the specific features of the Brazilian book market (from the publishing stage all the way to the profile of book readers). (iii) The last section concludes.

• B FBP Policies: A Historical Overview

  I The Problem at Hand: FBP Regimes and Their Implications for the Supply of Books

  For some years now, competition in the retail segment of the Brazilian book market has been experiencing significant changes, which began when megastores started entering and establishing themselves in the market in the second half of the 1990s. The subsequent entry of large online retail stores changed the market’s competitive standards. Because they operate on large scale and yield significant portfolio power (besides books, they sell other products that may interest consumers), these companies are capable of lowering prices to a point that is commercially infeasible for smaller bookstores. Moreover, the bargaining power of these larger online retailers is growing greater than that of smaller bookstores.¹⁰x See V. Vara, ‘Is Amazon Creating a Cultural Monopoly?’, The New Yorker, available at: <www.newyorker.com/business/currency/is-amazon-creating-a-cultural-monopoly>. See also ibid., p. 5. This at the same time squeezes publisher mark-ups and mitigates the ability of bookstores to compete with large online retailers, to the detriment of competition. Unsurprisingly (some might say), the amount of bookstores in the Brazilian market has been decreasing in the last few years. In 2011, Brazil had 3,481 bookstores; in 2014, this number had already dropped to 3,095.¹¹xAssociação Nacional de Livrarias, ‘A localização das livrarias brasileiras (2014)’. Available at: <http://anl.org.br/web/pdf/noticias/livrarias_brasileiras_2014.pdf>.
  This trend is not limited to the Brazilian market. US Census data reveal that, in the United States, the number of bookstores has dropped by 28% between 2007 and 2012. Borders, a large American bookstore chain, went bankrupt in 2011, and Barnes & Nobles – another large bookstore chain – has in the last three years been facing difficulties to maintain itself in the market. In the United Kingdom, a country that abolished its fixed book price regime in 1995, the number of independent bookstores dropped by 35.7% between 2005 and 2014.¹²x See: A. Singh, ‘Decline of the Independent Bookshops in the UK has Fallen to Just 987, with Surrey’s Oldest Bookshop the Latest Victim’, The Telegraph, 22 February 2014, available at: <www.telegraph.co.uk/culture/books/booknews/10654506/Decline-of-the-independent-bookshop-as-UK-figures-fall-below-1000-for-first-time.html>. According to data made available by Nielsen and Bowker, independent bookstores in the United Kingdom – a total of 987 – account for 3% of the book retail market. Bookstores in the country sell at an average retail price of 10 euros, and the share of large online retailers is far from negligible. Amazon represents 27% of the total print book segment, and 84% of the total e-book segment.
  Where do FBP regimes fit within what seems to be a growing, worldwide trend? What is the relevant correlation between FBP policies and the fast decrease in the number of bookstores in the above-mentioned countries? One way to answer this question for the time being is by comparing the data from the United Kingdom with that of Germany, a country where an FBP policy is currently in force. Data from Buch und Buchhandel in Zahlen 2013 and GfK Mediacontrol has revealed that there are currently 4,038 independent bookstores in the country – over four times as many as in the United Kingdom. Bookstores in Germany account for 31% of the book retail market. The average retail price of a book in Germany is 10.56 euros (very close to the price charged in the United Kingdom). Moreover, Amazon accounts for 14.9% of the print book retail market. This notwithstanding, data from the 1998-2002 period revealed that English consumers were spending more money on books after the Net Book Agreement was revoked in 1995, and that the number of book titles had increased exponentially during that period.¹³xThe authors admitted to some extent that while “the effect of abolition on book prices appears to have been downwards, as hoped,” “(…) problems of measurement and aggregation mean that we should admit some uncertainty on this count” (S. Davies et al., ‘The Benefits from Competition: Some Illustrative UK Cases’, DTI Economic Papers no 9, Centre for Competition Policy, University of East Anglia, 2014, p. 46, available at: <https://core.ac.uk/download/files/7/9633876.pdf>). The increase in the number of patterns has remained steady. Figures from the International Publishers Association in 2014 revealed that UK publishers released the largest number of titles in the world (second place was Taiwan, and third place was Slovenia).¹⁴xA. Flood, ‘UK Publishes More Books Per Capita Than Any Other Country, Report Shows’, The Guardian, 22 October 2014, available at: <https://www.theguardian.com/books/2014/oct/22/uk-publishes-more-books-per-capita-million-report>.

  Table I Production and print runs in France¹⁵xSource of data, available at: <www.snel.org.br/wp-content/themes/snel/docs/Palestras/apresentacaojeanguy.pdf>.

  	1990	1995	2000	2005	2010	2011	2012	2013
  Production in # of print runs
  New titles	20,252	21,998	25,832	33,740	40,021	41,902	44,678	46,619
  Reprints	18,162	20,999	26,045	32,988	39,287	39,366	41,616	48,865
  Total of titles published	38,414	42,997	51,877	66,728	79,308	81,268	86,294	95,484
  Production (in thousands of copies)
  New titles	212,311	220,497	242,951	350,539	386,439	379,489	387,389	353,393
  Reprints	173,854	165,665	179,910	208,393	245,489	240,579	243,524	218,640
  Total of # copies published	386,165	386,162	422,861	558,932	631,928	620,062	630,913	572,033
  Average print run
  New titles	10,483	10,024	9,405	10,389	9,656	9,056	8,671	7,580
  Reprints	9,572	7,889	6,908	6,317	6,249	6,111	5,852	4,474
  Global average print run	10,053	8,981	8,151	8,376	7,968	7,630	7,311	5,911

  A simple comparison between the two countries is illustrative. Indeed, prices are similar in both of them, but the price of books is regulated in one of them (Germany) and not the other (the United Kingdom). On the other hand, while independent bookstores thrive in Germany; the same does not hold true for English independent bookstores. Can we say this is a consistent pattern? Does not regulating price mean the structure of the book market will end up being more concentrated? Does a concentrated market mean the supply of book titles will necessarily become less diverse and consumers will end up reading less?
  Let us turn to the first two questions by going through the case of another European country that abolished its FBP policy not too long ago: Switzerland. Until 2007, an FBP policy was in place in Switzerland; the question of whether the country should keep its FBP policy was submitted to public debate, and in 2012, a referendum decided against the law that established the FBP regime. The Swiss book market nowadays is marked by high discounts for bestselling books, and overall higher prices of all other books compared to the pre-2007 period. Amazon books sold in the country come from Germany and all books are priced with at least a 20% discount. Currently, a substantial amount of independent bookstores are going out of business. At the same time, the two largest book chains in the country (Orell Füssli and Thalia.ch) merged in 2013. The market is becoming more concentrated.¹⁶xThese statements were obtained from a slide deck presentation entitled Fixed Book Price System in Germany, drafted in 2015 by Joachim Kaufmann, Vice-President of the Bonnier Books New Markets, for the International Forum on Fixed Price Law in Rio. Available at: <www.snel.org.br/wp-content/themes/snel/docs/Palestras/apresentacaoalemanha.pdf>.
  As to the third question, there seems to be evidence pointing in all directions. What we mean by this is that there does not seem to be a universal pattern along the following lines: “if you abolish an FBP regime, the number of book titles will necessarily decrease.” In some institutional contexts, and depending on a number of factors that are hard to pinpoint ex ante, the contrary can happen, as the UK example described above proves. That being said, there seems to be robust evidence pointing in the opposite direction: “if you introduce an FBP regime, the number of book titles will gradually increase.” France, a country that adopted an FBP policy in 1985, has experienced a stable and significant yearly increase in the total number of titles published in the country, as one can see in Table I.
  The pattern we are describing – i.e., the correlation between greater market concentration and less book title diversity as a result of not regulating book prices – has become popular in cultural public policy debates across the globe. Unsurprisingly, several countries have bought into the idea that FBP policies prevent market concentration, keeping independent bookstores in the market selling a greater number of book titles at reasonable prices (at least not at greater prices than those charged in countries without FBP policies). According to the Frankfurt Book Fair Business Club, nearly half of the largest book markets in the world (Brazil is one of them) currently have an FBP policy in place, as shown in Table II.
  Several other countries, in addition to the 11 found in Table II, currently have FBP regimes in place. Annex I, drafted by International Publishers Association and supplemented by us, lists all of these countries and contains key information on the design of each FBP policy. As Annex I shows, a total of 20 countries have FBP policies in force, usually to regulate the prices of print books, sold in bookstores and online. The FBP regimes of nearly all the countries listed above have been implemented by way of a law – Japan and Norway are the only exceptions. The fixed book price in most countries stays in force for a specific amount of time, generally no longer than two years. After that period, discounts can be applied freely. Some countries have established some exceptions to the no-discount period: under certain circumstances, a group of consumers (students, for example) or a retail channel (book fairs, for example) can benefit from discounts.

  Table II Main FBP regimes in the world

  Country	FBP in place?	Law or agreement?
  United States	No	N/A
  China	No	N/A
  Germany	Yes	Law
  Japan	Yes	Agreement
  France	Yes	Law
  United Kingdom	No	N/A
  Italy	Yes	Law
  Spain	Yes	Law
  Brazil	No	N/A
  South Korea	Yes	Law
  The Netherlands	Yes	Law
  Russia	No	N/A
  Norway	Yes	Agreement
  Turkey	No	N/A
  Poland	No	N/A
  Australia	No	N/A
  India	Yes	Law
  Canada	No	N/A
  Switzerland	No	N/A
  Belgium	No	N/A
  Sweden	No	N/A
  Mexico	Yes	Law
  Austria	Yes	Law

  II What Lead Countries to Implement FBP Policies in the Past? Identifying Trends

  A handful of FBP policies in the world (see Annex I) have been either revoked or repealed in the past, usually under the argument that fixed book prices hurt competition. In essence, the rationale applied by antitrust authorities is the one applied to RPM arrangements in general.¹⁷xRPM is a pricing arrangement in which an upstream firm – usually, a manufacturer – seeks to influence the price of its own products by downstream firms – usually, distributors. These arrangements are usually implemented with the assistance of coercive mechanisms, such as fines or the overall exclusion from the distributor network in the event of noncompliance. In antitrust, RPM arrangements are usually associated with two main anticompetitive effects, which may materialize under certain circumstances. First, it is argued that RPM can facilitate cartel formation among manufacturers, as well as among distributors.¹⁸x See M. Motta, Competition Policy: Theory and Practice, Cambridge, Cambridge University Press, 2004, p. 158 et seq. See also L. Peeperkorn, ‘Resale Price Maintenance and Its Alleged Efficiencies’, European Competition Journal, Vol. 4, 2008, p. 201 et seq. Second, it is held that RPM can increase a manufacturer’s unilateral market power, which may lead to both the elimination of its rivals and a decrease in dynamism and innovation among distributors.¹⁹x See Paragraph No. 224 of European Commission, ‘Guidelines on Vertical Restraints’, 10 May 2010, p. 80, available at: <http://eur-lex.europa.eu/legal-content/PT/TXT/PDF/?uri=CELEX:52010SC0411&from=PT>.
  Concerned with the potential anticompetitive effects of RPM arrangements, antitrust authorities in several jurisdictions have scrutinized and sought to revoke FBP arrangements, which were said to deprive booksellers of the incentives to compete over prices, ultimately injuring consumers who have to pay more for books. One example is that of the Swedish antitrust authorities, which in 1974 repealed the FBP regime in force in the country. Two years later, Finland and Australia also revoked their respective FBP regimes because their antitrust authorities raised competition concerns.²⁰xAvailable at: <www.snel.org.br/wp-content/themes/snel/docs/Material/global.pdf>. The United Kingdom’s FBP regime – known as the Net Book Agreement – followed a similar fate in 1995. It was scrutinized by the Office of Fair Trading, after which it lost the support of leading publishers and retailers; the Publisher’s Association ended up indefinitely suspending the agreement.²¹xEuropean Parliament, ‘European Parliament Resolution with Recommendations to the Commission on the Drawing-up of a Directive of the European Parliament and of the Council on the Fixing of Book Prices (2001/2061(INI))’, Strasbourg, 16 May 2002, available at: <www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+TA+P5-TA-2002-0244+0+DOC+XML+V0//EN>. In Ireland, where the book trade was regulated by the UK’s Net Book Agreement until 1989; the European Commission declared that cross-border fixed price arrangements were a violation of the law, and the agreement in Ireland was abandoned.²²x Ibid. Similarly, prior to being repealed via referendum in 2012, a Swiss FBP policy was in force until 1999, when it was sanctioned by the Swiss competition authorities.²³xConcurrences, ‘The Swiss Parliament is Discussing a Bill Destined to Reintroduce Fixed Book Prices, Once Sanctioned by Swiss Competition Authorities for Violating Competition Law’, Bulletin, News Issues, May 2009, available at: <www.concurrences.com/Bulletin/News-Issues/May-2009/The-Swiss-Parliament-is-discussing>. Hungary’s 100-year-old FBP regime was also abolished in 2007 due to competition concerns.
  In-depth antitrust scrutiny has therefore been something far from uncommon in the history of FBP policies. One reason for this may be the manner in which they were originally designed and implemented (the alternatives being a private, industry-wide agreement, or a law or decree). Another reason may concern their purposes at the time, i.e., the prevailing interest and aim with the implementation of the regime that existed at the time. A shift in the declared purposes of FBP policies, and in the manner in which they were implemented, may also explain why antitrust scrutiny has been less intense in recent years.

  1 The Pro-Industry FBP Regime Period: From Industry Agreements to Laws

  The first price-fixing arrangement concerning books was put in place in 1829, when publishers in the United Kingdom started feeling the need to combat excessive retailer discounts. Eight years later, Denmark’s publishers and booksellers established an agreement of their own. In 1888, Germany’s Börsenverein – which represented the interests of publishers, wholesalers and retailers – included provisions of fixed retail prices into their statutes. If members did not comply with the fixed prices, they risked facing sanctions. A year later, France’s Cercle de la Librairie standardized trade relations between the two types groups that of members of the association: publishers and booksellers. The Cercle de la Librairie introduced a mechanism where publishers set and disclosed recommended retail prices for books.
  In 1900, the United Kingdom adopted an agreement among publishers called the “Net Book Agreement.” The Net Book Agreement was approved by the Publishers’ Association and the Association of Booksellers of Great Britain and Ireland. According to Wilson, however, “Retailers had no active part in forming the agreement.”²⁴x Ibid., p. 11. By way of the agreement, publishers were free to name any of their products a ‘net book’ (the agreement covered not only books, but also maps, pamphlets and other materials). The publisher was authorized to fix a ‘net price’ for each of its ‘net books’. All retailers of the book had to comply with the net price, and could only sell books at a lower price in three different circumstances: (i) an annual two-week National Book Sale, during which retailers were allowed to sell their net books under the net price if they had not been ordered in the past 12 months; (ii) when the book was second-hand, and 6 months had passed since publication; (iii) retailers were allowed to grant 10% discounts in their sales to libraries.²⁵xDavies et al., 2014, p. 31. If a retailer failed to comply with the fixed price, it would be collectively denied any future supply of books from all publishers that were a part of the agreement.²⁶xInternational Publishers Association, Global Fixed Price Report, available at: <www.internationalpublishers.org/images/news/2014/global-fixed-price-report.pdf>. Importantly, if a book was marked by the publisher as a net book, it would be subject to the net price for as long as it remained a net book, except in the above-mentioned scenarios. The duration of the applicability of the net price was uncertain and unregulated.
  The United Kingdom’s Net Book Agreement was subject to antitrust scrutiny in 1962, just a few years after the 1956 Restrictive Practices Act was introduced, which prohibited any collective enforcement of restrictive practices. In 1957, the Net Book Price was rewritten and formalized with the purpose of becoming exempt from the Restrictive Practices Act. As a result, in 1962, the Restrictive Practices Court, in the context of analysing the illegality of the Net Book Agreement, exempted it from the scope of the Restrictive Practices Act. According to Mr. Justice Buckley, “Books are different,” and the Net Book Agreement was in the interest of society.²⁷xDavies et al., 2014, p. 31. Until the mid-1990s, the Net Book Agreement remained exempt from antitrust scrutiny. In 1997, the Court once again analysed the Net Book Agreement, after it had already lost the support of the Publishers’ Association in 1995. This time, the Court concluded that the Net Book Agreement should be revoked because “increased price competition would have beneficial effects on the retail price”²⁸x Ibid., p. 38. and the risk of falling numbers of independent stores, retail stockholdings and published titles was not a major concern.
  The FBP regimes implemented by the Dutch, the Germans, the English and the French had one thing in common: they all consisted of agreements, either between publishers alone or between publishers and retailers. None of them was the result of direct governmental intervention, by way of a law or any other state-produced norm. In 1924, however, this pattern started to change. France’s Poiret Decree established a regime of recommended prices for booksellers; the vast majority of booksellers complied with the suggested prices.²⁹x Ibid., p. 24. This was the first time a state directly regulated the price of books, instead of allowing private parties to reach agreements on their own. Notwithstanding, private industry-wide agreements were still adopted in many countries for a few decades after the Poiret Decree.

  2 The Pro-Culture/Pro-Reading Period: Laws Aimed at Stimulating the Book Industry

  Through time, FBP arrangements became more and more common around the world. In 1950, countries like Hungary, Italy, Sweden, Japan, the Netherlands, Australia, Finland, Ireland and Austria had all begun regulating the price of books. However, the real shift from FBP arrangements implemented via agreements to FBP regimes established by laws came in the 1980s. In 1981, France’s ‘Loi Lang’ (Law No. 81-766, of 10 August 1981, altered by Law No. 81-500, of 13 May 1985) came into force, and has since then been a source of inspiration for many other countries that have designed and implemented FBP policies (as well as those that are debating the topic).
  In the years that followed its coming into force, Austria, Greece, Luxembourg, Denmark, Italy and Portugal turned their FBP agreements into laws, mainly in response to the increased antitrust concerns raised by the European Commission.³⁰xM. Appelman, ‘Fixed Book Price’, in R. Towse (Ed.), A Handbook of Cultural Economics, Elgar Publishing, Cheltenham, 2003, p. 240. In 2002, Germany followed suit, and replaced its fixed book price agreement for a law governing the issue. The risk of facing scrutiny from competition enforcers, in some cases, and the concrete action of antitrust agencies, in others, led certain countries with industry-implemented FBP agreements in force to repeal them instead of switching to laws. The United Kingdom, Ireland and Switzerland, mentioned above, are some examples.
  Despite antitrust risks, FBP regimes are still being discussed and implemented in current days. Mexico passed a law establishing an FBP regime in 2008. In 2013, Israel also adopted an FBP regime via a law, although currently the country is debating whether to keep this regime in place. In 2014, Slovenia’s Parliament also approved a law establishing an FBP regime. Brazil, Quebec and Poland are some of the most prominent examples of political systems where the adoption of an FBP policy has been the subject of political debate in the past few years.

  3 Comparing the Two Periods

  There may be some relevant differences between the publisher-implemented agreements drafted and put into force in the past (the 19th and early to mid-20th century) and the laws or publisher-implemented agreements that have been designed and put into force in the last few decades. The main one, in our view, is that the political motivation to establish an FBP regime seems to have changed – at least to some degree. Indeed, what was before seen and designed as a means for the book industry to protect itself from more efficient rivals is now being viewed with different lenses. The self-imposed contractual arrangement that characterized the early years of FBP seems to be shifting to a version that, at least in theory, is designed with the purpose of promoting public interest.
  In policy debates, FBP regimes are increasingly being associated with enhanced bibliodiversity and a geographically spread-out supply of books, rather than with higher prices resulting from the industry’s power and ability to protect its own interests in spite of that of consumers. The point is not that publisher-implemented FBP arrangements today have the sole purpose of protecting industry interests. It is rather that the appearance of FBP regimes can be more accurately explained as a materialization of the motivation to protect supply-side interests, and that policymakers and the book industry as a whole began acknowledging that these arrangements resulted in positive externalities (bibliodiversity, increased supply, etc.), but that in order for them to flourish, FBP regimes needed some adjustments.
  Unsurprisingly, therefore, as of the 1980s many governments started drafting and introducing FBP regimes via laws, influenced by the Loi Lang. What was first primarily auto-regulation to maintain balance in the industry and avoid the ‘costs’ of ‘unfair competition’ turned into a public policy that incorporated other goals of public interest. More importantly, the FBP’s design changed in a significant manner. Prior to the Loi Lang, a standard FBP regime worked in a manner similar to the Net Book Agreement: it consisted of a publisher-imposed fixed price, with an indefinite duration of applicability. Retailers did not, as a rule, have the possibility of granting discounts, except under exceptional circumstances.
  The Loi Lang changed this scenario. Article 5 of the Law establishes that the fixed price’s applicability duration is two years counting from the day the book is released. In other words, two years after a book is released, retailers are free to grant the discounts they deem fit. Although there may be adverse effects associated with fixing the price of books, one might argue that they are increasingly becoming more limited in scope. Moreover, although incipiently, Loi Lang started the trend of exempting the sale of certain books, such as textbooks, to which the reasons for implementing an FBP regime were not applicable. Indeed, Article 3º of the Law provides a few exemptions, such as one for sales to institutions that purchase textbooks for their members.
  The trend we believe to have identified can be summarized as follows: (i) the justification for the introduction of an FBP regime has been changing; and (ii) as a result of the shift in the purpose of FBP policies, their actual design has been changing, and the scope of applicability (i.e., the duration and type of books covered) of the fixed book price is becoming smaller.
  The Spanish case is a good example of the trend we are suggesting. Indeed, while the primary justification of the first law that established an FBP regime in Spain (Law 9/75 of 12 March 1975) was “the avoidance of unfair competition,” a new FBP regulation of 1990 (Royal Decree 484/1990) acknowledges that the regime is “a fundamental means of promoting diversity in publishing and bookselling.”³¹xM.L. Palma Martos & L. Palma Martos, “Fixed Book Pricing in Spain: A Debate Between Economic Efficiency and Cultural Diversity’, 14th International Conference of the ACEI, Vienna, 6-9 July 2006, p. 6, available at: <www.fokus.or.at/fileadmin/fokus/user/downloads/acei_paper/Palma-Martos.pdf>. The argument to justify the FBP framework changed, and it would not be far-fetched to assume that the increasing intensity of antitrust scrutiny was also one of the causes. Notwithstanding, the Spanish FBP policy was subject to the scrutiny of the Spanish Antitrust Authority, which issued a policy paper recommending the abolishment of the regime and the establishment of a free market system in the book industry.³²xTribunal de Defensa de la Competencia, ‘Informe sobre el precio fijo de los libros’, September, 1997, available at: <https://www.cnmc.es/Portals/0/Ficheros/Promocion/Informes_y_Estudios_Sectoriales/1997/1.pdf>.
  In 2004, the argument changed slightly yet again, as Martos reports:

  the Ministry of Culture has presented to the Council of Ministers the draft of the new Law on reading, the book and libraries which are intended to supersede the 1975 Law. The fixed price, which according to the Ministry of Culture reflects a policy of “cultural exception”, is reaffirmed (…).³³xPalma Martos & Palma Martos, 2006, p. 7.

  The official message emphasizes the goal of the FBP regime: to foster reading and culture. To this end, the FBP regime was once again reviewed. Law 10/2007, of 22 June 2007, a new FBP regime, was introduced in Spain, but with one difference: textbooks were excluded from the FBP regime’s scope. According to the preamble of the document,

  the regulation of book sales is the result of the conviction that books are not just mere goods: they are the physical support of human thought, of science and of literary creation, and allow for mankind to undertake a unique and transcendental activity which is reading. The dissemination of these creations, their cultural value and diversity require a certain guarantee of both the quality of the pieces but also their commercialization, in order to make them accessible to a greater number of potential readers.³⁴xFree translation. Original text available at: <https://www.boe.es/buscar/act.php?id=BOE-A-2007-12351>.

  Moreover, the fixed price’s applicability of duration provided for in Law 9/75 – which preceded the Loi Lang – was not limited by law. Indeed, Article 33 of the Law simply established that all but a few types of books had to be sold at a price fixed by the publisher, and did not explicitly limit the duration of applicability of said price. It was not until the coming into force of Royal Decree 484/1990 that this changed. Pursuant to Article 2(f) of the rule, the fixed price’s duration of applicability lasted only two years. If one looks at Annex I, one will notice that, nowadays, the duration of applicability of the fixed book price is limited either by law or agreement in the vast majority of cases.³⁵xWe cannot say all cases, because there is no data available for some of the countries.
  Furthermore, a comparison between Royal Decree 484/1990 and Law 10/2007 reveals yet another difference in the rules’ pricing systems. While Royal Decree 484/1990 prohibited the granting of discounts during the period of applicability, Law 10/2007 establishes in Article 9 that, during the period of applicability, book prices may range from 95% to 100% of the price fixed by the publisher. This is still an incipient trend in other countries, and one that is not yet widely adopted (the newly introduced Israeli FBP system, for example, establishes a ‘no-discounts’ policy, as we will see in the following). However, gradually, FBP regimes seem to be providing retailers with a little more breathing room during the fixed price’s period of application. This could make sense in some cases where booksellers have information advantage vis-à-vis publishers, are quicker in noticing that a given book is a commercial failure, and thus will not sell at all at the price fixed by the publisher. In addition, progressive, or limited, discounts would have enabled retailers to compete over prices.

  III FBP Regimes Today: Some Differences between Developed and Developing Countries

  From 1980s onwards, FBP arrangements began to be marketed as a governmental solution for countries with populations that (i) do not have the habit of reading (more applicable to developing countries like Mexico); and/or (ii) where authors are not adequately compensated for their work, and therefore the incentives to produce literary works are low (more appealing in developed countries like Israel).
  In Mexico, the political and public debate prior to and after the coming into force of the 2008 law that established an FBP regime in the country was mainly centred on the social benefits of reading, and on how the arrangement would foster reading habits throughout the country. According to Barba, Mexicans read an average of 2.9 books a year, and 51% of the population does not remember the last book they read.³⁶xPalma Martos & Palma Martos, 2006, p. 7, at 3. Thus, one of the main reasons why Mexico passed the FBP law was that a social programme aimed at fostering reading would inevitably fail if it were not complemented by state-induced incentives to strengthen the supply chain of the book industry, so as to provide access to books under similar conditions “in every corner of the country.”³⁷xF.L. Barba, ‘El precio único del libro en México’, Tiempo, Apuntes, p. 84, available at: <www.uam.mx/difusion/casadeltiempo/19_iv_may_2009/casa_del_tiempo_eIV_num19_81_85.pdf>. Another important argument that backed the 2008 FBP law in Mexico was that books are not like other goods. Books are cultural goods and deserve the state’s protection due to their social, rather than merely commercial, nature.³⁸x“Un aspecto central en esta determinación fue la consideración de que cualquier programa de fomento a la lectura estará destinado al fracasso si no pasa por un proyecto integral que fomente la producción de libros y el acceso a ellos en todos los rincones del país en condiciones semejantes. Igualmente, el argumento de peso en los legisladores mexicanos fue que el libro no es una mercancía como las otras. El libro es un bien cultural que merece la protección del Estado debido a su función social más que mercantil” (Ibid.). Importantly, the President’s Cabinet was of the opinion that an FBP regime would end up adversely affecting efficient booksellers – capable of offering better prices – and consumers.³⁹x Ibid., p. 82.
  The Mexican law, albeit welcomed by the book industry at the outset, was also subject to significant legal and public scrutiny after it came into force. The Mexican Constitutional Tribunal reviewed the law to assess whether it harmed competition in the country, and concluded that it did not.⁴⁰xCNN, ‘La Corte avala la ley que fija precios únicos para los libros en México’, CNN, 1 September 2011, available at: <http://mexico.cnn.com/entretenimiento/2011/09/01/la-corte-avala-la-ley-que-fija-precios-vunicos-para-los-libros-en-mexico>. The problem with the law then shifted to its effectiveness. According to representatives of the Mexican book industry, the law has ‘no teeth’, i.e., it does not expressly establish sanctions for those who fail to comply with the relevant FBP provisions, and no governmental authority charged with enforcement. Reports state that the consequence of this is that while some booksellers do comply with the law, others do not.⁴¹xV. Juárez, ‘Reporte global sobre el precio fijo del libro’, Uvejota, 19 June 2014, available at: <http://uvejota.com/articles/3577/reporte-global-sobre-el-precio-fijo-del-libro/>. One can say that the booksellers that are obeying the law are, in commercial terms, being adversely affected by it. Thus, rather than strengthening the book chain, Mexico’s law seems to be harming it, and failing to fulfil its goals.
  Conversely, in 2013, Israel implemented an FBP policy by passing the Law for the Protection of Literature and Authors in Israel (Temporary Provision no 5773-2013). The law came into force on 6 February 2014, and introduced “new regulatory requirements with respect to publishing, distribution, sales and profit allocation across the book supply chain.”⁴²xHerzog Fox & Neeman, ‘Israel’s new Authors Act’, Lexocology, 28 January 2014, available at: <www.lexology.com/library/detail.aspx?g=00148cfe-eae1-455b-96fd-93698702da3e>. The law prohibited bookstores from selling books at a price that was different from the retail price fixed by the publishers. Booksellers were not allowed to grant discounts on new titles for 18 months (exceptions in Annex 1, below), nor to bundle offers (‘buy one, get one free’). According to Herzog Fox & Neeman,

  This “no discounts” rule and its “unbundling” effects are, together, the one single provision that drew the most fire from the Act’s critics and most of the public debate concentrated on this clause.⁴³x Ibid.

  Each year, publishers and bookstores drafted and entered into an agreement establishing the discount at which books were to be sold to stores during that year. Bookstores were not allowed to request any discounts beyond those sanctioned by the annual agreement; publishers were prohibited from granting additional discounts.
  During the period in which the book price set by the publisher had to be complied with, publishers were obligated to pay authors at least 8% of retail price for the first 6,000 copies sold, and 10% for all other copies sold (after the first 60,000 copies). If the book was the author’s first, she would be entitled only to payments corresponding to 80% of the percentages applicable to authors that had already written and published books in the past. After the 18 months FBP protection period provided for in the Israeli law, publisher must pay authors at least 16% of the actual payment received by the publisher for all books sold.
  A provision in Israel’s Law for the Protection of Literature and Authors established that the FBP regime was to be re-examined three years after it came into force. However, even before the expiry of the FBP regime’s trial period, the law is already under attack. Prominent figures such as the Minister of Culture Miri Regev and Yoav Kisch publicly challenged the law’s effectiveness.⁴⁴xKisch and Regev argued that: “The law created an absurd situation in which the legislature’s intervention harmed the competitive market and increased prices for the citizen… The legislature intervened too much in the free market economy, and the purpose of canceling the law is to bring back the former situation, and allow prices to drop and the spread of new books in the Israeli market to increase” (The Jerusalem Post, ‘Cancellation of Law Which Increased Book Prices Moves Forward’, The Jerusalem Post, 20 March 2016, available at: <www.jpost.com/Israel-News/Politics-And-Diplomacy/Cancellation-of-law-which-increased-book-prices-moves-forward-448574>). The purpose of the law was to “ensure Israeli authors proper pay for their creations, to promote literature in Israel, to preserve cultural diversity in publication and distribution of books.”⁴⁵x The Jerusalem Post, 2016. However, according to the law’s opponents, by limiting the ability for booksellers to discount on new titles, the FBP regime increased the price of books, reducing the amount of book sales.
  In June 2016, Israel’s FBP regime was re-examined and repealed.⁴⁶xIsrael National News, ‘Book Law’ Banning Discounts on Book Sales Repealed’, Israel National News, 31 May 2016, available at: <www.israelnationalnews.com/News/News.aspx/213037#.V1DhAeRyy9i>. The recent Israeli experience shows that while the rhetoric in support of FBP regimes may have changed and ‘improved’ throughout the last years, the arguments against FBP regimes have largely remained the same: FBP increases prices, compensates inefficient market players, harms efficient suppliers, and does not necessarily entail foster bibliodiversity. To better assess the argument against FBP, we must therefore enter the realm of antitrust law and economics.

• C An Analysis of the Brazilian FBP Bill

  I Reframing the Debate

  The history of FBP has been marked by constant antitrust scrutiny. While such scrutiny were evidently called for in a scenario where most FBP arrangements had the sole purpose of protecting and advancing industry interests (many of them were even imposed by the industry itself, with no government participation), one might ask if this is still the case in a scenario where most FBP regimes are shifting to a more moderate design. From mechanisms aimed at protecting the industry from unfair competition, or even agreements put in place with no justification at all (at least no declared, publicly-acceptable one), FBP regimes are now self-declared cultural and developmental policies: they are designed to provide individuals with more access to culture, across different regions of different, unequal countries (especially in developing countries), and to ensure authors – especially new ones – are adequately compensated for their work (see the Israeli example).
  What are the features of an FBP policy nowadays?

  1. First, the duration of applicability of the FBP is limited in time, and in most cases is no longer than two years.

  2. Second, some countries have widened the scope of exceptions to the FBP regime: textbooks, government purchases for schools, library purchases, etc.

  3. Third, some FBP regimes are, to a limited extent, allowing retailers to grant discounts during the period of applicability.

  Table III Social and economic regulation (Windholz & Hodge’s framework)

  	Economic regulation	Social regulation
  Primary values	Efficiency Competition Innovation Individualism Choice	Justice Fairness Equity Social Cohesion Trust
  Supporting (i.e. secondary) values	Justice Fairness Equity Social Cohesion Trust	Efficiency Competition Innovation Individualism Choice

  Source: Table elaborated by Windholz & Hodge (2012).

  These features differ significantly from those of the first laws and agreements aimed at regulating fixed prices. As noted above, one of the reasons for the shift was the intensity of antitrust scrutiny. FBP policies had to adapt to the pressure of economic regulators – they were correct in doing so, of course. As a result, what we see today embodied in FBP policies is a work in progress: a gradual reconciliation between the goals and values of economic regulation and social regulation. Windholz and Hodge isolated the values of economic regulation and compared them with those of social regulation.⁴⁷xWindholz & Hodge, 2012, p. 227. The result can be seen in Table III.
  The authors argue that, in many different scenarios subject to regulatory intervention, one cannot properly ascertain whether the regulator acted “as either purely economic or purely social in nature.”⁴⁸x Ibid., p. 233. Indeed, while ‘economic regulators’ pursue their objectives in accordance with their ‘primary values’, they also operate within an environment constrained by secondary values. The actual balance between social and economic values is, in this scenario, inevitably context specific, “with different balances being struck with respect to different policy areas and at different points of time in light of changing needs and circumstances.”⁴⁹x Ibid.
  These findings led Windholz and Hodge to conclude that regulatory regimes have more than one purpose; they can at the same time attain socially desirable outcomes and foster consumer choice, for example. After all, “(…) all regulators exist within a social context, and social regulators within an economic context.”⁵⁰x Ibid. Moreover, Windholz and Hodge found that the difference between social and economic regulations does not lie in the absence of one or more values, or the prevalence of one over the other; the values of economic and social regulation are not – or should not – be in competition with one another; the difference resides in “the primacy and supporting roles played by sometimes contradictory values.”⁵¹x Ibid., p. 234.
  Therefore, the question of whether economic regulation – such as antitrust law and economics, for example – should consider social values (and vice versa) is unproductive, beside the point and at odds with the real world where law meets actual institutions, individuals and problems. The question that one should instead ask is whether when economic regulators (inevitably) take social values into account, they do so “in a manner that is transparent and in which the balance that is struck is explicit and clear.” This balancing between the values inherent to social and economic regulation is key in order to enhance the credibility and overall legitimacy of a given public policy.⁵²x Ibid., p. 237. Moreover, social goals such as the promotion of culture, education, diversity, access to knowledge, freedom of speech, etc., also contribute to maximizing consumer welfare. Contemporary antitrust policy should, thus, not disregard them. This is something commentators have been noting.⁵³xStucke, for example, is of the view that: “now is the time to reconsider antitrust’s political, social, and moral concerns. In reconsidering the goals of competition as a means to secure political, economic, and individual freedoms, antitrust can be more responsive to citizens’ concerns about promoting well-being” (M.E. Stucke, ‘Reconsidering Antitrust’s Goals’, Boston College Review, Vol. 53, 2012, p. 624). Pitofsky, on the other hand, notes that: “(…) if the free-market sector of the economy is allowed to develop under antitrust rules that are blind to all but economic concerns, the likely result will be an economy so dominated by a few corporate giants that it will be impossible for the state not to play a more intrusive role in economic affairs” (R. Pitofsky, ‘The Political Content of Antitrust’, University of Pennsylvania Law Review, Vol. 127, 1979, p. 1051).
  As noted in the introduction of this Article, FBP is a textbook example of a very complex mix between social and economic regulation. Social regulation provides the reasons for FBP’s existence. However, FBP is attacked based on arguments that draw on economic regulation. Putting aside the question of whether the values of social or economic regulation should prevail, the question we will seek to address in the remainder of this article is whether there is an explicit, acceptable balance between social values and economic values embodied in the proposed Brazilian FBP Bill.
  We will advance our analysis based on the premise that the Brazilian FBP policy is essentially a body of social regulation. Thus, its primary values considered are those originating from social regulation, and its secondary values are those of economic regulation. Indeed, the current version of the FBP scheme found in the Brazilian FBP Bill:

  1. Encompasses all segments of the book industry (publishing, distribution and retail);

  2. Contemplates coercion mechanisms for cases of non-compliance with the fixed book price established, and therefore does not constitute a mere suggestion of prices, which is generally unproblematic from an antitrust standpoint⁵⁴xOn February 2001, CADE dismissed an RPM complaint against publishers and textbook distribution companies because of the fact that, under the relevant agreement, compliance with the prices was not mandatory, but rather merely suggested. As noted by the case’s Reporting Commissioner, “the prices suggestions the publishers are disclosing do not constitute antitrust violations under Brazilian law, since the Brazilian Copyright Law sanctions this conduct, obligating publishers to suggest book cover prices to booksellers. It is based on this price suggestion that authors are payed royalties for their manuscripts” (Case No. 08000.001191/95-64; Defendant: Empresas Editoras e Distribuidoras de Livros Didáticos; decision of 14 February 2001). On 27 July 2011, CADE once again dismissed na RPM complaint against several publishers. According to Reporting Commissioner Olavo Chinaglia “(…) what we see in this case is that publisher are drafting price lists – a conduct that is expressly sanctioned by law – which are used as a reference in the booksellers’ pricing policies. The case files lead one to conclude that, in addition to suggested prices being efficient in the market under investigation, compliance or not with the suggested prices is a strategic and circumstantial decision up to each bookseller, and there is apparently no threat or possibility of retaliation if bookstores choose to charge different prices” (Case No. 08012.001743/2002-40; Defendants: Editora Globo, et al.; decision of 27 July 2011).;

  3. Establishes a one-year applicability period, after which retailers are free to grant the discounts they wish on their products;

  4. Establishes a six-month applicability for books after their second edition;

  5. Exempts rare, antique, outdated, imported, and books destined for publicly funded institutions.⁵⁵xThe document is available at: <https://www.senado.gov.br/atividade/materia/getPDF.asp?t=160838&tp=1>.

  Points (i) and (ii) are the source of most of the antitrust concerns associated with FBP regimes. By contemplating a coercive mechanism to ensure compliance with the fixed prices, and by encompassing all segments of the book industry, the regime suggested by the proposed Brazilian FBP Bill is largely equivalent to that of a common minimum RPM arrangement. As noted above, minimum RPM arrangements risk giving rise to two main anticompetitive effects: (i) the facilitation of collusive practices between manufacturers, as well as between distributors; (ii) the increase in a manufacturer’s unilateral market power.⁵⁶xThe European Commission Guidelines on Vertical Restraints mentions additional anticompetitive effects resulting from RPM arrangements, such as preventing distributors from reducing their retail prices, potentially raising prices consumers will have to pay and leading distributors to favor the brand that adopts the RPM scheme (due to the mark-up guaranteed by the fixed price), therefore harming the competitors of the manufacturer that fixed prices. Points (iii), (iv) and (v) can be interpreted as factors that mitigate the extent and weight of the antitrust concerns raised by the FBP regime. As mentioned above, FBP regimes nowadays are seeking to mitigate antitrust concerns by limiting the scope of applicability – in terms of time and types of books covered – of the book’s price fixed by the publisher.
  In light of this, and given that the Brazilian FBP Bill seems to fit the trend described above, the question that needs to be answered is whether the primary social values associated with the bill are properly balanced with the secondary economic values. The primary social values can be described as:

  1. The promotion of bibliodiversity (a widened range of different titles from different authors, commercially appealing or not);

  2. Geographically widespread access to books;

  3. Adequate compensation for authors, so as to encourage the creation of knowledge and innovation;

  4. Facilitated access to knowledge, which is arguably enhanced by the experience one goes through in a bookstore.

  Justice, fairness and equity are the proposed Brazilian FBP Bill’s ultimate social values. These objectives can be compromised if bookstores have no incentives to stay in business as they face too much competition from online bookstores. They may also be compromised if the book market responds solely to the law of supply and demand, and therefore commercially unappealing, yet culturally valuable books are not offered.
  The supporting economic values of the Brazilian FBP Bill can be described as:

  1. Lower prices for consumers; and

  2. More output variety (subject to the laws of supply and demand).

  Economic efficiency and competition are the policy’s ultimate economic values. Both of these values can be significantly compromised by collusion and by a unilateral abuse of market power. Indeed, collusion and the unilateral abuse of market power gives rise to the risk of an increase in prices (which are not contested by competitors), and to a decrease in output. They also create incentives for companies to invest in customer services, in product differentiation and in innovation in general.
  In the remaining portion of the article, we will turn to the investigation of two main questions: (i) is there evidence that the supporting economic values are significantly compromised by the Brazilian FBP Bill?; and, more importantly, (ii) is there evidence that the primary social values that the Brazilian FBP Bill seeks to promote will indeed materialize in Brazil? In other words, does the Brazilian FBP Bill provide balance?

  II How Significant Are the Antitrust Risks Associated with the Proposed Brazilian FBP Bill?

  1 Risk of Collusion

  As mentioned above, one of the main anticompetitive effects generally associated with RPM arrangements – including FBP arrangements – is the risk of the conduct facilitating cartel formation. The antitrust literature explores two situations in which this risk can materialize in the upstream market. In the first one, suppliers of a given product facing difficulties to follow each other’s prices reciprocally would start fixing retailer prices in order to facilitate monitoring. Since RPM arrangements increase market price transparency, the suppliers would have more visibility over the behaviour of their peers, monitoring compliance with their agreement and swiftly detecting any departure from it. In the second situation, the suppliers, concerned with the possibility of a price war among distributors leading to a drop in upstream prices (harming the cartel), would uniformly adopt either minimum or fixed prices.
  In order to support the thesis that RPM may facilitate a cartel among distributors, some commentators claim that the suppliers are used by the distributors as an instrument for cartel formation and stability (in these cases, the supplier/manufacturer plays the role of a ‘maestro’). By standardizing distributor prices and monitoring compliance, the supplier contributes to the implementation of the agreement, and prevents distributors from having to constantly keep in touch for the sake of the stability of the agreement.
  There is some – albeit outdated – empirical evidence that distributors/dealers have, on some occasions, been capable of persuading manufacturers to impose RPM arrangements and act as ‘maestros’.⁵⁷xT. Overstreet, ‘Resale Price Maintenance: Economic Theories and Empirical Evidence’, USFTC Bureau of Economics Staff Report, 1980, available at: <https://www.ftc.gov/sites/default/files/documents/reports/resale-price-maintenance-economic-theories-and-empirical-evidence/233105.pdf>. However, the study that contains this empirical data also concluded that in over 91% of the RPM cases for which information on the retail market structure was available, there were over 100 retailers in the relevant market. This suggests that widespread collusion among retailers was unlikely, and that the RPM arrangements analysed may have been motivated by other factors (not collusion).
  Many commentators have been recently expressing great scepticism towards the claim that RPM arrangements are collusion facilitators, be it in the supplier’s segment, or in the distributor’s segment, being those cases limited to a very particular circumstances. According to these authors, the claim lacks economic robustness.⁵⁸xElzinga & Mills, for example, note that “(…) just as real world examples of retailer cartels stitched together by RPM are not common, real world examples of manufacturer cartels that used RPM to curtail cheating are uncommon. Indeed, the economic nexus between cartels and RPM is not robust.” See K. Elzinga & D. Mills, ‘The Economics of Resale Price Maintenance’, in Issues on Competition Law and Policy, ABA Section of Antitrust Law, 2008, p. 1847. See, in the identical sense, R. Blair & D. Kaserman, Antitrust Economics, 2nd ed., Oxford, Oxford University Press, 2009. Unsurprisingly, then, although there is an abundance of empirical research seeking to prove the hypothesis that RPM is an important piece in collusion strategies, the hypothesis is yet to be confirmed. To the contrary.⁵⁹xIn the words of Organization for Economic Cooperation’s (“OECD”) background note: “(…) the available empirical evidence is scant and dated, and it does not show that RPM always tends to restrict competition and reduce output. On the contrary, it points in the other direction. Perhaps most importantly, it indicates that RPM is not commonly used to facilitate cartels.” See OECD, Roundtable on resale price maintenance – Background note, 2008, p. 38, available at: <https://www.oecd.org/daf/competition/43835526.pdf>. Ornstein, for example, concluded that a mere 10% of the cases involving RPM investigated by the Federal Trade Commission (FTC) referred to cartels.⁶⁰x See S. Ornstein, ‘Resale Price Maintenance and Cartels’, The Antitrust Bulletin, No. 30, 1985, pp. 401-423. Similarly, another study shows that, among the 153 complaints involving RPM filed with the FTC between 1976 and 1986, only 5.9% of them made any reference to collusion. Moreover, Kleit, after analysing RPM cases in the United States before the conduct’s standard of illegality was clarified by courts, concluded that there is no clear evidence that RPM arrangements were actually used as cartel strategies.⁶¹xAdditionally, Kleit notes that “(…) the rise of RPM in the 1880s can be directly attributed to the rise in importance of advertised branded goods. Given this, and the fact that horizontal restraints predate RPM by at least several centuries, it seems unlikely that the primary motivation for RPM is cartel facilitation.” See A. Kleit, ‘Efficiencies without Economists: The Early Years of Resale Price Maintenance’, Southern Economic Journal, Vol. 59, No. 4, 1993, p. 617.
  In light of the fact that these empirical endeavours seem to have reached the same conclusion, the plausibility of the argument that RPM may be used to facilitate cartel formation in the book market seems negligible. Even if one were to contest the validity of this conclusion, it would still be unlikely that a publisher-instituted RPM regime would facilitate collusion. Indeed, commentators have been keen to mention that one can count with one’s fingers the amount of cases in which it is plausible to argue that RPM facilitated collusion in the upstream market.⁶²x See, e.g., H. Marvel, ‘Resale Price Maintenance and the Rule of Reason’, The Antitrust Source, 2008, p. 1. See also R. Blair, ‘The Demise of Dr. Miles: Some Troubling Consequences’, Antitrust Bulletin, No. 53, 2008, p. 137.
  Commentators are also sceptical about the claim that RPM facilitates cartel formation in the downstream market. Indeed, would it really make sense for suppliers in the upstream market to support a cartel in the downstream market? Naturally, the cartel would entail in a decrease in sales from distributors to end-consumers (readers). This necessarily means that upstream sales will also decrease, and therefore most likely so will profits.⁶³x See Blair & Kaserman, 2009. Additionally, RPM arrangements ensure that players in the downstream market will not compete over prices, but that they can certainly try to cheat the cartel by competing discretely over other factors (e.g., better customer services).⁶⁴x See Elzinga & Mills, 2008, p. 1846. Other factors can also be responsible for facilitating both upstream and downstream cartels, but there is in fact no substantial evidence showing that RPM is indeed an effective strategy for this specific purpose.
  Moreover, in order for the risk of collusion to be credible – with or without the assistance of an RPM arrangement –, the market at issue needs to have conducive features towards the reduction of costs of reaching and maintaining collusion, such as homogeneous products, similar cost structures, few market players with similar market shares.⁶⁵x See A. Jones & B. Sufrin, EU Competition Law, Oxford, Oxford University Press, 2014, p. 787. As emphasized by OECD: “market conditions do not always favour cartel formation. The well-known group of factors that affect the plausibility of a successful cartel in general (such as the number of firms in the market, the homogeneity of their products, the ease of entry, etc.) will make cartels too difficult to maintain in some markets, even with the assistance of RPM” (emphasis added). OECD, supra note 59, p. 30. Absent these features, the cartel would have trouble profiting above competitive levels (via quantity reduction and price increases) and would simply not be practiced, due to an absolute lack of economic rationality.
  There are no such features in the publishing segment. First, books are clearly not homogeneous products. It is impossible for one to find the perfect substitute for a book. Casa Grande e Senzala, a classic in Brazilian anthropology and sociology written by Gilberto Freyre, does not substitute Raízes do Brasil, another Brazilian sociology classic by Sérgio Buarque de Holanda, even though they both contributed significantly to the understanding of Brazil. Similarly, Vidas Secas, a classic novel by Graciliano Ramos, does not mean one can go without reading Quase-Memória, by Carlos Heitor Cony.
  An important consequence of this is that, at least in theory, publishers have some room to comfortably fix their prices; publishers are, indeed, price makers. When they purchase copyrights for books, they also purchase the right to price independently.⁶⁶x See F. Sá Earp & G. Kornis, A economia do livro: a crise atual e uma proposta de política, IE/UFRJ, Rio de Janeiro, Text for discussion no 04/2005, p. 24. Thus, if a publisher holds monopoly power over the commercial exploration of a given piece, there are no incentives for this publisher to collude with rivals, fixing prices: it would be very hard for publisher X, owner of the copyrights for a give book Y, to agree with publisher A that book Y will be sold at a certain price. Publishers actually have the incentive to compete for the best authors and best pieces. This is what will make readers prefer their books to those of competitors.
  Additionally, the publishing market is highly deconcentrated, both in terms of volume and revenues, as can be seen in Table IV. Indeed, not a single publisher accounts, alone, for over 10% of the market.
  When the supply structure of a market is highly deconcentrated, the risk of collusion decreases substantially.⁶⁷x See, e.g., R. Whish & D. Bailey, Competition Law, 7th ed., Oxford, Oxford University Press, 2012, p. 522. The reason for this is that, in a market with a large amount of players, it is hard to gather a significant number of competitors willing to collude. If only a small part of the market colluded, the stability of the cartel would be compromised, and the competitive harm of the cartel would be negligible, since all demand could just turn to other suppliers that are not taking part in the cartel.

  Table IV Publisher market shares

  Publisher	Market share (revenue)	Market share (volume)
  Saraiva	8.20%	5.64%
  Record	6.66%	7.64%
  Intrínseca	6.31%	9.07%
  Sextante	5.29%	8.53%
  Companhia das Letras	4.66%	5.04%
  Grupo Santillana	2.91%	1.78%
  Planeta do Brasil	2.77%	3.41%
  Rocco	2.64%	3.29%
  Leya Brasil	2.51%	2.54%
  Grupo Gen	2.50%	–
  Ediouro	2.24%	3.45%
  Globo	2.12%	2.54%
  Objetiva	2.05%	2.07%
  Novo Conceito	1.85%	2.82%
  Elsevier Brasil	1.65%	–
  Outras Editoras	45.65%	35.51%

  Source: Nielsen BookScan – Total 2014 (201401-201452)

  There is yet another feature of the book publishing market that hinders cartel formation: the barriers to entry in this market are not very significant.⁶⁸x Ibid., p. 64, at 4. Indeed, albeit it is true that experience in the publishing industry, a good network of professional relationships with authors, foreign publishers, translators, etc., and a consistent commercial strategy indeed contribute to the success of new publishing companies, recent developments in the Brazilian publishing market prove the likeliness and feasibility of successfully entering it.⁶⁹x See G. Mello, ‘Desafios para o setor editorial brasileiro de livros na era digital’, BNDES Setorial, Economia da Cultura, No. 36, p. 444. That is the case, for example, of publishers like Companhia das Letras, established in 1986, Sextante and Grupo Autêntica, in 1998, and Intrínseca e Novo Conceito, founded in 2003 and 2004, respectively. Foreign firms have also successfully entered the Brazilian market; some examples are the entry of Planeta, from Spain, in 2004, and Grupo Leya, from Portugal, in 2008.
  The stability of a cartel also depends on the barriers to entry of the market in which it is implemented.⁷⁰xAs well noted by OECD: “Generally, if barriers to entry are low or if substitute products exist, collusion will not be successful. If, on the contrary, high entry barriers protect market incumbents from competitive pressure from potential new entrants, it is more likely that the collusive outcome will be reached and maintained sufficiently stable over time.” OECD, Public Procurement, Competition Policy Roundtables, 2007, p. 21. Indeed, when barriers are high, the risk of a new player entering the market – sometimes compromising a highly sensitive, and hard-to-reach consensus between members – is low. Indeed, nothing guarantees that the new entrant will agree with the terms of the cartel agreement, and there is even the risk of him bringing the existence of the cartel to the attention of public authorities. It is also not impossible for this entrant to simply decide not to participate in the cartel. The entrant would therefore be able to sell its products at a lower price than the one charged by the cartel, absorbing a part of the demand of the cartel members. In this context, a cartel will rarely survive, even if assisted by an RPM arrangement.
  Finally, there is one more reason why it would be difficult for a cartel among publishers to hold together: the demand for books is always uncertain and, ultimately, unknown.⁷¹x See Sá Earp & Kornis, 2005, p. 4. The antitrust literature has been repeatedly pointing out that, when demand for a given product is seasonal, the probability of the cartel maintaining itself is low, since incentives to cheat the agreement become high.⁷²x See M. Levenstein, ‘What Determines Cartel Success?’, Journal of Economic Literature, Vol. 44. No. 1, 2006. Demand in the publishing market is not even seasonal (i.e., subject to some degree of certainty). Several pieces are virtually bets: who would have thought that Harry Potter was going to be so much of a success? If it were launched today, can one say for sure it would be as successful? As the demand for books is highly uncertain, the incentive to cheat the terms of a possible cartel agreement is, evidently, very high.

  2 Risk of a Unilateral Increase of Market Power

  We are intuitively inclined to believe that RPM arrangements are harmful to competition for a very simple reason: they supposedly increase the prices consumers end up paying. This argument is based on the premise that this rise in prices has a negative impact on consumer welfare. However, the underlying question that begs an answer is: does the welfare of Brazilian book consumers depend solely on prices?
  An analysis of the characteristics of the pent-up demand for books suggests the contrary. It consists of people who are usually in the job market or preparing themselves to enter it (high school and college students), with an average income, and that recognize the social value of books, but do not read due to a lack of interesting book titles and time.⁷³x See Sá Earp & Kornis, 2005, p. 9. Price, on the other hand, does not seem to be a very important factor for book readers. Only 4% of readers have stated that they have not been reading in the last three months because books are too expensive.⁷⁴x See data from “3^a Edição da Pesquisa Retratos da Leitura no Brasil,” funded and carried out by Instituto Pró-Livro e o Ibope.
  In the Brazilian book market, therefore, consumer welfare depends more on things like book diversity and amount of books available. Generally speaking, book consumers want more choice, which is also something antitrust policy seeks to promote.⁷⁵xIn Henry Ergas words: “as a general matter, competition policy aims at protecting, and where appropriate and efficient extending, the range of choices available to consumers” (H. Ergas, ‘The Interaction and Coordination of Competition Policy and Consumer Policy: Challenges and Possibilities’, in OECD, The Interface between Competition and Consumer Policies, OECD Policy Roundtables, 2008, p. 18. See also: N.W. Averitt & R.H. Lande, ‘Consumer Sovereignty: A Unified Theory of Antitrust and Consumer Protection Law’, Antitrust Law Journal, 2007. See also: L. Sylvan, Activating Competition: The Consumer Protection – Competition Interface, Trade Practices Workshop, University South Australia, 2004. More choice is not ensured by lower prices. It thus seems necessary to foster the development of variety in book supply, in order to have more diverse bookstores, specializing maybe in poetry, or African history, etc. This would give consumers more choice and, therefore, increase their general welfare.
  Even if prices indeed contributed significantly to consumer welfare – which does not seem to be the case, in light of the data available on the profile of Brazilian book consumers – the anticompetitive nature of FBP regimes still seems to a certain extent far-fetched. The reason for this is that it would not necessarily entail in an increase in book prices. The low concentration levels of the publishing segment contemplated in Table IV indicates that competition in the market is vigorous. In our view, an RPM arrangement would not alter this reality. RPM ensures publishers that book cover prices shall, as a rule, be the same regardless of the retail channel in which they are sold.
  Each publisher fixes cover prices for their books, over which, as a rule, he or she has exclusive commercial rights. Although price is not the main competitive factor in the book market, one cannot say that it is completely irrelevant. If the cover price fixed by the publisher is higher than the reserve price that a consumer confers on the book (i.e., the maximum amount the consumer is willing to pay), the consumer would simply opt to not buy it.⁷⁶xThe size and complexity of books are factors that influence the marginal utility consumers confer on the product. The reason for this is that these factors determine the opportunity costs associated with reading, and therefore affect the marginal utility. The opportunity costs associated with reading books that are less intellectually challenging are lower than the costs associated with reading dense, complicated books. See F. Van der Ploeg, ‘Beyond the Dogma of the Fixed Book Price Agreement’, Journal of Cultural Economics, No. 28, 2004, p. 6.
  It would therefore be irrational for the publisher to fix an excessively high price. In doing so, the publisher would risk losing customers to a competing publisher that also sold interesting books in the eyes of consumers, but at lower prices. Therefore, although it is reasonable to expect, at an initial stage, an increase in the prices of book in countries with a recently established FBP book regime, these prices would tend to readjust and decrease as a result of competition among publishers; this in fact already occurs, as the rate of reductions in publisher mark-ups in the past few years seem to indicate. Indeed, data from Fundação Instituto de Pesquisas Econômicas (FIPE) indicates that, between 2004 and 2013, the average prices charged by publishers have dropped 43.7% (discounting inflation).⁷⁷x See CBL/SNEL, Produção e Vendas do Setor Editorial Brasileiro – Produto 3: Histórico do setor com dados de 2013 e relatório sobre a nova metodologia, São Paulo, FIPE, 2014, pp. 9/10.

  Table V Average book prices in countries with FBP regimes

  	1998	1999	2000	2001	2002	1998-2002 (%)
  Germany	19.1	18.2	16.9	15.0	16.5	-13
  Austria	15.7	16.1	14.3	14.1	15.4	-2
  Netherlands	16.1	15.3	13.3	13.9	13.6	-15
  Spain	10.5	11.4	10.5	10.2	10.6	0
  France	7.1	6.8	6.0	5.7	5.4	-24
  Average	13.7	13.6	12.0	11.7	12.3	-10

  Source: Sá-Earp & Kornis (2005), based on data from the EuroMonitor (2003).

  Table VI Average book prices in countries without an FBP regime

  	1998	1999	2000	2001	2002	1998-2002 (%)
  Bélgica	18.4	18.0	15.9	15.8	15.8	-7
  EUA	12.0	12.1	12.9	13.3	13.7	14
  Itália	12.2	13.1	11.2	11.6	12.7	4
  Japão	7.5	8.7	9.1	8.1	7.8	4
  Reino Unido	12.2	11.8	11.2	10.8	11.5	-6
  Média	12.5	12.7	12.1	11.9	12.6	1

  Source: Sá-Earp & Kornis (2005), based on data from the EuroMonitor (2003).

  This explains why the average price of books in countries that adopted FBP arrangements is not lower than those of countries where book cover prices are unrestricted. It also clarifies why the average price of books in countries with FBP arrangements have experienced a drop. Data collected by Sá-Earp & Kornis⁷⁸xF. Sá-Earp & G. Kornis, Preço único do livro: solução frágil para um problema grave, p. 7, available at: <www.snel.org.br/wp-content/themes/snel/docs/Material/preco.pdf>. listed in Tables V and VI reveal just that.
  One can observe from a joint analysis of Tables V and VI that, between 1998 and 2002, book prices have not increased in countries that have adopted FBP regimes. On the contrary, they decreased or remained stable (Spain). The prices of books in countries without an FBP regime, on the other hand, increased in the vast majority of cases (they dropped only in the United Kingdom and Belgium). Recent data from France – a country where an FBP regime is currently in place – shows that books are still highly accessible to the population. The average price of books is 11 euros, which corresponds to the price of a movie ticket in Paris, the country’s capital. Moreover, between 1998 and 2008, book prices increased half as much as the consumer price index of that period.⁷⁹x See: <www.internationalpublishers.org/news/blog/entry/why-fixed-book-price-is-essential-for-real-competition>.
  On the other hand, recent data from the United Kingdom indicates that the price of best-sellers have indeed dropped 1% since 1995, the year in which the country put an end to its FBP regime. However, the prices of all other books have increased over 50% since 1995, significantly exceeding the consumer price index of that period (28%). Although this data does not confirm the hypothesis that FBP regimes reduce book prices, they significantly compromise the argument that FBP regimes necessarily lead to an increase in the prices of this product.

  III Can We Expect the Social Welfare Benefits of the Proposed FBP Bill to Materialize?

  The data and the literature we have reviewed in the preceding topic suggest that, in Brazil, publishers do not have enough incentives to engage in collusive or abuse of dominance practices. Indeed, publishers work with heterogeneous goods (books), in a highly deconcentrated and competitive market marked by low entry barriers. It is also worth noting that Brazilian book consumers do not view price as the obstacle to their consumption of books. Moreover, even if price were the most important competitive factor in the book industry, publishers lack the rationale to charge excessive prices for their books, since they would risk losing customers to a competing publisher that also sold interesting books in the eyes of consumers, but at lower prices. Thus, it does not seem that there is a substantial risk that the proposed FBP Bill would facilitate collusion at the retail or manufacturing segment of the Brazilian book industry. The same holds true for the risk that publishers would unilaterally increase abuse their market power.
  This being the case, half of the argument for an FBP regime in Brazil seems to be in place: the supporting values of the policy (i.e., the policy’s economic values) are not significantly compromised. Thus, the question that remains is whether there is reason to believe that the primary values of FBP regimes would materialize in the Brazilian case.

  1 Bibliodiversity

  FBP supporters assert that these regimes foster bibliodiversity. This argument is supported by two main points. First, it is argued that FBP regimes create incentives for entry in the retail market, expanding supply. In addition to implementing differentiation strategies – i.e., investments in product advertising/promotion, or in improving the quality of bookstore customer services – new retailers can choose to work in specific niches. In other words, new, specialized retailers can appear in the market, focusing on interior design books, or foreign literature, etc. This means more bibliodiversity and, consequently, more choices for consumers.⁸⁰xAs noted by Bittar, “independent libraries have a central role in strengthening the ‘book culture’ and the habit of reading. With specialized services, such as attentive care and a well-qualified staff, they end up being one of the most effective ways of introducing new, rising authors or challenging literature into the market. In many places, these bookstores are cultural centers, where literary events and workshops are organized. Instead of prioritizing a stock of new, bestselling books, independent bookstores usually make most of their money with books of with no commercial appeal, or books of specialized niches.” (A.C. Bittar, ‘Digital Rights Management, Concorrência e Acesso ao Conhecimento no Mercado e Livros Digitais’, Master’s dissertation presented to the Escola de Direito de São Paulo da Fundação Getulio Vargas, 2015, p. 122.) Additionally, it is argued that FBP regimes foster more bibliodiversity because they enable cross-subsidies, due to the excess of revenues fixed prices confer.⁸¹xVan Der Ploeg explains: “Since profits under monopoly pricing are evidently higher than profits in a competitive equilibrium (y + a − F > b + a − F), more book titles are profitable and thus more titles are published/sold under the FBP than in competitive equilibrium. It is possible to print and sell extra books at low and almost non-increasing marginal cost, so that the producer loss b is likely to be small. Also, the price elasticity of the demand for books ε is likely to be small as a large part of the full cost of reading is the opportunity cost of time. Hence, the monopoly mark-up is likely to be large and monopoly profits y are large. It thus seems likely that many more book titles are published under the FBP than under perfect competition.” (Van Der Ploeg, 2004). Thus, the revenues obtained from a publisher with high sales volumes of a best-seller can be invested in publishing and releasing pieces that are of social and cultural interest, but not necessarily of common taste and commercially appealing.⁸²x See Rouet, François, Le Livre, mutations d'une industrie culturelle, Les Études de la documentation française, 2 ed., Paris: La Documentation française, 2007.
  One point that compromises the argument for bibliodiversity is that publishers may have the incentive to ‘pocket’ the excess revenue earned via an FBP regime, instead of investing in a wider array of titles.⁸³xOn this argument, see Van der Ploeg, 2004. Is there evidence that this argument applies to Brazil?
  In our view, there is hardly an incentive for publishers to ‘pocket’ this excess revenue in the form of higher profits. Indeed, and as mentioned above, a careful examination of the pent-up demand for books in Brazil reveals that one of the reasons why consumers end up not reading books is that there is a lack of interesting titles in the market. Price, on the other hand, does not seem to be the key competitive factor in the publishing segment. This being the case, publishers’ commercial strategies depend on their capacity of producing a wide array of titles, in order to therefore give consumers more choice. Currently, however, the investment in the diversity of titles may be smaller than it could be because of the bargaining power that large retailers still possess (see Table VII), even though the use of technology has made it difficult to provide a reliable depiction of the structure of this market.
  In the absence of an FBP regime, publishers may tend to adopt a more conservative commercial approach, and are less inclined to ‘bet’ on completely new, incipient literary pieces. Authors that are already successful today may have even more chances to publish a new book than completely unknown, beginning-of-career authors. In this sense, FBP regimes may increase the chances of new authors, with new ideas, appearing in the form of their literary pieces. This also benefits consumers, which therefore have access to new pieces dealing with a greater variety of subjects, and can count on a larger number of titles at their disposal in the market.

  Table VII Structure of the Brazilian book retail market

  Retailer	Number of stores	Market-share
  Saraiva	104	25%
  Cultura	19	11%
  B2W	2	9%
  Leitura	51	8%
  Curitiba	21	7%
  FNAC	11	5%
  Nobel	200	4%
  Laselva	60	2.5%
  Travessa	8	2.5%
  Vila	8	2%
  Total	484	76%

  Source: Livrarias – Publishnews (2014).⁸⁴xData extracted from: <www.snel.org.br/wp-content/uploads/2015/07/apresentacao_MarcosdaVeigaPereira_.pdf>.

  2 Geographic Capillarity

  Creating incentives for the Brazilian network of bookstores to be highly dense and geographically well-distributed is very important. Books are typical examples of experience goods.⁸⁵x See Van Der Ploeg, 2004, p. 5. The concept of an experience good encompasses products whose quality cannot be accurately estimated by consumers when buying the product, regardless of the price charged (i.e., prices are not necessarily an indication of quality in these cases).⁸⁶x See P. Nelson, Philip, ‘Information and Consumer Behavior’, Journal of Political Economy, Vol. 78, No. 2, March/April 1970. Understanding how demand for books is created and grows is a key step in accepting the argument that FBP fosters an enhanced geographic capillarity of books.
  If books that are more expensive are not necessarily the best, how can a consumer know if a given book will be of interest? An indication from a friend with the same taste in books could be an option. Reading a synopsis in the back cover of a book can be an alternative. The ranking of a book within the best-sellers of the month can be another way of having an idea of the book’s quality and whether a reader will like it. Although these options help consumers make more informed decisions, the density and capillarity of bookstores is important for another reason: to create opportunities for consumers to get in touch with the product, since prior demand for books does not necessarily exist, and therefore daily exposition of these products is of essence in order to arouse the interest of consumers in purchasing them, creating and developing a reading habit.

  Table VIII Geographic distribution of Brazilian bookstores

  North	4%
  Northeast	16%
  Centre-West (including the Federal District)	6%
  Southeast	55%
  South	19%

  Source: ANL (2014).

  It is argued that FBP fosters more capillarity in the country’s network of bookstores, because by guaranteeing a certain profit margin for retailers, RPM contributes to the maintenance of smaller bookstores – characterized by high fixed costs, but stocked with a large variety of titles – in the market.⁸⁷xFixed costs must be paid by the company regardless of how productive the month/year was (in the case of booksellers, regardless of how many books were sold). Examples of fixed costs are rent and employee salaries. This is particularly important in Brazil, a country in which bookstores are scarce and terribly geographically distributed.⁸⁸xThat is the scenario noted by Lemos: “[I]f one considers the existence of bookstores as a proxy for access to books, the market in Brazil is very small. The country as a whole has approximately 2000 bookstores or an average of just one bookstore for every 84,400 Brazilians. The absolute majority of Brazilian cities do not have any bookstores. Until roughly 10 years ago, entire states such as Roraima, Tocantins, and Amapá have only two bookstores each. In the overall Northern region of the country, the average is one bookstore for every 215,300 inhabitants.” (R. Lemos, From Legal Commons to Social Commons, Centre for Brazilian Studies, University of Oxford, Working Paper no 80, 2006, p. 13.) There are currently 3,095 bookstores in Brazil, 69% of which have only one store, 12% of which are equipped with two to ten stores, and 19% of which have more than ten stores.⁸⁹x See supra note 1. These bookstores are highly concentrated in the south-eastern part of the country, as one can understand from the data set out in Table VIII.
  Table VIII also revealed that there is a significant scarcity of books in the centre-west and north of the country. An RPM regime for books could, by stimulating the entry and maintenance of new market players in the retail segment, help alter this socially undesirable scenario.
  It is also socially undesirable that the number of Brazilian bookstores remain stable. Currently, there is one bookstore for every 64,954 inhabitants in the country. This number is significantly below the international average. In the United States, for example, there is one bookstore for every 24,053 inhabitants.⁹⁰x See: <www.publishersweekly.com/pw/by-topic/industry-news/bookselling/article/57631-where-the-stores-are.html>. Spain currently has one bookstore for every 10,335 inhabitants⁹¹xThere are currently 4,500 bookstores in the country according to data made available by El País, a prominent local newspaper. See: <http://elpais.com/diario/2011/04/22/sociedad/1303423201_850215.html>. (there is a total of 4,500 bookstores in the country). In Germany, on the other hand, there was one bookstore for every 15,000 inhabitants in 2006.⁹²x See M. Uribe, ‘El acceso al libro y el precio único’, Pensar El Libro, No. 4, 2006. In Argentina’s capital alone, Buenos Aires, there is one bookstore for every 4,000 inhabitants; Buenos Aires is known for being the city with most bookstores in the world.⁹³x See: <www.aljazeera.com/news/2015/07/argentine-capital-world-centre-bookstores-150716143741268.html>. In this context, book supply incentives like an RPM regime may change the habits of Brazilian consumers. An increase in the number of bookstores may lead Brazilians to read more and have more access to culture and individual education tools. In the current scenario, it is common knowledge that the number of readers in Brazil is decreasing: between 2007 and 2011 alone there was a 9.1% drop.⁹⁴xAccording to data made available by the “3^a Edição da Pesquisa Retratos da Leitura no Brasil,” a report prepared by the Instituto Pró-Livro and IBOPE. In 2007, there were 95.6 million readers in Brazil. In 2011, the figure fell to 88.2 million. One must bear in mind that the report considers ‘readers’ to be individuals that read, completely or in part, at least one book in the last 3 months. The number of readers who have the habit of reading every day, or at least more than once a week, is much lower. An RPM regime could contribute to reversing this trend.

  3 Adequate Compensation for Authors

  The argument that FBP regimes create the correct incentives for authors to be adequately compensated is structured as follows: FBP aligns the interests of the publishing segment with those of the retail segment, thus avoiding depreciations in the supply of copyrights and, therefore, a reduction in the supply of books in the long run. The relevant question is how.
  Nowadays, selling books is not necessarily the ultimate goal of the aggressive discount strategies of online retailers. The purpose of these strategies is actually to simply attract consumers to the online platform, encouraging them to disclose subscription information, thus allowing online retailers to more accurately target future purchases of products that are more expensive, such as computers, fridges, microwaves, etc. By treating books almost as if they were commoditized and cheap products, standardizing pricing based on unaligned priorities which may be incompatible with the interests of publishers, online platforms signal to the population that books do not have the potential to change individuals.
  Consumer perception will, over time, be that books are not worth paying much for. In this context, incentives to invest in literary production – be it individuals (authors) or firms (publishers, etc.) – will tend to diminish: in the eyes of society it will be deemed a second-tier, unprestigious and uncompensating activity. In this context, and under a dynamic rather than merely static analysis, one would expect that the number of people willing to produce quality literary content would decrease.
  Consequently, the supply of copyright will drop and, unavoidably, the total amount of books available in the market will too. This risk increases when the retail market has significant bargaining power. This is the case when the retail market is highly concentrated, thus giving the few players operating in it a lot of purchasing power. Naturally, these few players would be capable of requiring lower prices. Nevertheless, if they abuse this power, the retailers risk making upstream market players conclude that it is simply not worth supplying a product already so depreciated by the downstream market. With this, consumers are left worse off in the long run.

  4 The Bookstore Experience

  FBP regimes may contribute to protecting and improving the bookstore experience in two main ways: (i) by increasing retail mark-ups and at the same time dampening intrabrand competition; (ii) by avoiding free riding.
  Increased retail mark-ups . Increasing retail mark-ups does not create an incentive for retailers to invest in improvements in the quality of the retailer’s establishment,⁹⁵x See supra note 48. and contributes to incrementing retailer sales efforts, which have a “naturally greater incentive to advertise such products of its line of business.”⁹⁶x See C. Mattos, ‘Fixação do Preço de Revenda Mínima no CADE: o caso SKF’, Revista do Ibrac, Vol. 24, 2013, p. 82. The reason for this is simple: if an FBP arrangement necessarily entails in a greater mark-up for the retailer, and if the retailer cannot compete for prices,⁹⁷xOne of the main efficiencies associated with RPM arrangements generally is that they contribute to strengthening interbrand competition (See Elzinga & Mills, 2008, p. 1844). Thus, one might argue that even with an FBP regime in force, retailers would still compete, but rather than over prices, (i.e., who is capable of offering the best discounts), over investments on the improvement of customer services. Elzinga & Mills explain: “The motivation of a manufacturer using RPM is not to enable retailers to raise prices and gouge consumers. The economic logic of using RPM to induce efficient retail service is for the manufacturer to impose a price floor to restrain downstream price competition in order to foster service competition. A manufacturer has no incentive to overcompensate its retailers; doing so would reduce the manufacturer’s profits and would make the manufacturer’s product line less competitive in the marketplace. Instead, the manufacturer wants its retailers to make just enough money to market its brand effectively, and no more.” Elzinga & Mills, Ibid. By definition, RPM increases retailer profit margins, allowing retailers to invest in the improvements of customer services (Mattos, 2013, p. 82. See also W.L. Holahan, ‘A Theoretical Analysis of Resale Price Maintenance’, Journal of Economic Theory, Vol. 21, 1979, pp. 411-420 and F. Mathewson & R. Winter, ‘The Law and Economics of Resale Price Maintenance’, Review of Industrial Economics, Vol. 13, 1998, pp. 57-84 and R. Deneckere, H.P. Marvel, & J. Peck, ‘Demand Uncertainty and Price Maintenance: Markdowns as Destructive Competition’, American Economic Review, Vol. 87, 1997, pp. 619-641). In order to differentiate themselves from rivals, bookstores would indeed be more encouraged to invest in, for example, training their staff, cleaning their facilities, and putting together a diverse stock of books, all to the benefit of consumers. in what other way would he or she be able to expand market share, absorbing portions of its competitor’s demand? By advertising and otherwise promoting and diversifying its products, as well as by improving the quality of customer service. This is what will attract customers in a scenario in which price is not the main competitive advantage.
  In this context, bookstores with the best cafés, nicest and most helpful salesmen, most comfortable facilities and most interesting cultural attractions (such as jazz shows, for instance), will probably sell more products than their competitors that simply focus on selling books and offering no ancillary services. This already seems to be an incipient differentiation strategy for small Brazilian bookstores.⁹⁸x See G. Porcidonio, Letras e diversões: para atrair público e driblar crise, livrarias e sebos investem em shows, festas e peças, RioShow, 2015, available at: <http://rioshow.oglobo.globo.com/eventos/eventos/eventos-em-livrarias-parte-2-13489.aspx>. An FBP regime would help amplify this trend. The investment in product advertising and promotion, as well as in the improvement of customer services does not necessarily entail a direct loss of profits. Indeed, one could expect that this investment be compensated by an overall increase in sales. Since demand shapes supply, one could reasonably expect that a widespread increase in the total amount of books sold will lead to a general increase in the total amount of books manufactured in Brazil.
  According to Scherer and Ross, customer services are irrelevant to consumers that already know what they are looking for in advance, and how to use it. They are also irrelevant for consumers who are being inexpensive goods.⁹⁹xF.M. Scherer & D. Ross, Industrial Market Structure and Economic Performance, Houghton Mifflin Company, Boston, 3rd ed., 1990, p. 552. This however is not, and should not, be the case of book consumers. Books are not trivial, inexpensive books, nor do consumers necessarily know (or should know) in advance which one to buy. An important part of the consumer’s purchase experience is precisely entering a bookstore and discovering something new.¹⁰⁰xThe need to invert the consumer’s purchasing process – leading consumers to stop behaving on the lines of ‘I need/want this and will look for it’, and start behaving more like ‘I found it! This is what I wanted’” – in order to stimulate the habit of reading in Brazil was the topic of a presentation Júlio Cesar A.S. da Cruz gave in the Public Hearing of the Fixed Book Price Policy in the Brazilian Senate, on 30 July 2015. Available at: <http://legis.senado.leg.br/comissoes/reuniao;jsessionid=4D59CB1DDF4690CC8D8A49FB2BAD8D84?0&reuniao=3547&codcol=47>. And, at that moment, customer services are crucial.
  The mitigation of free-riding . In the retail book market, RPM arrangements are especially important because in the absence of a fixed book price, consumers tend to free ride on services available at bookstores that take customer services investments seriously, and then use the information available there to purchase books from other retailers who do not invest in customer services, and therefore are capable of offering lower prices. In this scenario, bookstore owners have less incentive to invest in services that improve the purchase experience of customers. Indeed, the risk of them being incapable of profiting from this investment is enormous.
  This problem (or risk) is even greater nowadays, in which consumer behaviour is highly influenced by the internet. Imagine the following: a consumer enters a bookstore, goes through several new books while sitting in a very comfortable and cosy chair, has access to a specialized salesman capable of clarifying important questions on the products, etc., but when it becomes time to buy the book, the consumer logs on an online platform and finds it at a much lower price. And then buys the book at the best price. It would not be absurd to say this happens often. The bookstore’s investment in services is, thus, unduly captured by large online retailers, which benefit from bookstore services and sell their products with high discounts, or simply at a lower price because they operate with much lower fixed costs than bookstores. This risk increases as new technologies develop: with only a few clicks on your smartphone, you can now search for prices and purchase books from online retailers while still sitting in a bookstore.
  In this context, FBP regimes can benefit consumers – indeed, by fixing the book’s sale price (the cover price), free riding is inhibited, since the prices charged by different bookstores become the same; in this scenario, retailers start having to compete over other factors rather than price alone, like for example customer services.

• D Conclusion

  In a country like Brazil, in which reading habits are so incipient, an FBP policy could, by creating incentives for new bookstores to enter the market, improve the shopping experience of consumers, encouraging them to read more. In this article, we have presented arguments in favour of the introduction of the proposed Brazilian FBP Bill.
  We credit the relevance of our analysis (beyond the Brazilian context) to two main reasons. First, contrary to common knowledge, minimum RPM arrangements do not predominantly produce anticompetitive effects in the publishing market (speaking in purely ‘economic regulation’ terms). Second, there are markets in which minimum RPM arrangements can contribute to promoting social objectives that should be treated as compatible with antitrust law – the FBP regime Brazil is discussing whether to introduce or not enhances individual welfare in both dimensions (social and economic).
  The social objectives pursued by a potential FBP regime in Brazil are legitimate – especially in developing countries lacking equal access to books across different regions and where the habit of reading is not disseminated. Thus, they should not fall within the same foxhole as the original objectives of FBP policies (which were to protect industry interests, to the detriment of consumers). More importantly, these objectives (or values) must not be obfuscated by purely economic values. An FBP policy is by nature not purely the product of economic regulation, but rather a very subtle and balanced mix between social and economic regulation (where social regulation produces the predominant values).
  As shown throughout this article, there is little to no evidence that the proposed Brazilian FBP Bill will increase the risk of collusion in the publishing segment. A lot of theoretical and empirical work has been showing that the idea that RPM arrangements can be used as cartel facilitators is rather implausible in markets such as the one at hand. The structural characteristics of the Brazilian book market are not favourable to the formation of price cartels; collusion is thus unsustainable in the long run. The same conclusion would apply to incentives to unilateral effects, as competition in the market upstream (the publishing market), as well as the choosing process of a book by consumers would prevent any such endeavour.
  Moreover, there is evidence that the proposed Brazilian FBP Bill has the potential to promote the social values attached to FBP policies, such as (i) creating incentives for the expansion of the supply of book titles in the country (bibliodiversity); (ii) increasing the denseness and capillarity of the network of bookstores, which contributes to an overall increase in the demand for books, since consumers are more exposed to the products (and, thus, more inclined to purchase them); (iii) providing incentives for authors to produce new, cutting-edge content by creating the expectation that she/he will be adequately compensated; (iv) protecting and improving the bookstore experience by encouraging the retail market to compete vigorously over other relevant competitive factors in the book market, such as the quality of customer services and the diversity their stock of books.
  Human creativity is our best bet to deal with the complex and serious problems of contemporary society. To be creative, however, one must necessarily experiment: read new, different and provocative pieces of literary work, in order to thus think outside pre-established lines. Vargas notes that, in a dynamic, collaborative, decentralized society such as the one we now live in, institutional leaps – which dictate the rhythm of social development – are not only cumulative, but also unpredictable. For this reason, “creativity is inherent to collective dynamics and historical evolution.”¹⁰¹x See D. Vargas, Jurista Criativo, Cadernos FGV Direito Rio no 11, 2014, p. 93. By stimulating entrepreneurship and the risk inherent in releasing new titles and new content, FBP policies free the market from the slavery of what we already know, of best-sellers alone, of common taste, and encourage us to seek new horizons.

• Annex I

  Key details of FBP regimes around the world

  Country	Paper fixed price?	Paper books fixed price system details			E-book fixed price?	If yes for both, same system?	Textbooks included?
  		Legal basis	Duration of applicability	Maximum discount
  Argentina	Yes	Law	18 months after publicationa	10% during book fairs, book festivals; 10% for schools and libraries, 50% for public entities	Yes	Yes
  Austria	Yes	Law	2 years after publication	5%; 10% for libraries	Yes		No
  France	Yes	Law	2 years after publication and 6 months from last bookstore purchase	5% (consumers), 9% (libraries)	Yes	No
  Germany	Yes	Law	Minimum 18 months	As a rule, no discountsb	Yes	Yes
  Greece	Yes	Law	2 years from publication	10%	Yes	Yes	No
  Italy	Yes	Law	20 months after publication and 6 months from last bookstore purchase	15%; 20% for books sold during Book fairs or trades fairsc	No		Yes
  Israel	Nod	Law	18 months after publication of the first edition	Discounts of up to 20% allowed during the Hebrew Book Week; 10% discounts can be granted by online bookstores; up to 20% discount for a sales of over 100 copies to one individual	Yes	No, the retail price for printed books and e-books can be different	Yes
  India	No, but India’s Maximum Retail Price Act provides that a maximum price must be printed on every book	Law		Discounts are not regulated, only pricing over the maximum retail price
  Japan	Yes	Resale Price Maintenance Agreements (voluntary agreements between publishers and wholesalers or between booksellers and wholesalers)	It depends on the agreement	It depends on the agreement	No
  Lebanon	Yes. Only applicable to textbook	Decree of Ministry of Economy		20% to schools, distributors and booksellers	No		Yes
  Luxembourg	Yes. Applies to domestic books only	Yes			N/A
  Mexico	Yes, but there are no sanctions for those who fail to comply with the law	Law	Applies to the first 18 months after the books is either published or imported	As a rule, no discounts	No		No
  Netherlands	Yes	Law	Publishers can reduce prices every six months	5-10% discount for students	No		No
  Norway	Yes	Agreement between publishers and booksellers’ associations	Until 30th April the year after publication	12.5% for book shops, 20% for libraries	Yes	Yes	No
  Portugal	Yes	Law	18 months	10%; 20% for libraries and during book fairs	No	No	No
  Slovenia	Yes	Law	6 months	0%; 20% only for book fairs, book series and subscriptions	Yes	Yes	Yes
  South Korea	Yes	Law	18 months	19% (10% + n9% through special discounts on online bookstores)	Yes	Yes	No
  Spain	Yes	Law	2 years after publication, 6 months after distribution in bookstores	5%; 10% for World Book Day and Book fairs, 15% for public entities	Yes	Yes	No. Textbooks for compulsory education are excluded
  Sri Lanka	Yes	Law	No data	No data	No data	No data	No data
  Thailand	Yes	Law	No data	No data	No data	No data	No data

  ^a The price can be lower if the publisher takes the book out of the catalogue. In this case, the books must be retired from the bookshops or the publishers have to wait 180 days so that out-of-print books can be sold as remainders.
  ^b Discounts of up to 5% may be granted to research libraries, and up to 10% for municipal, state and school libraries, as well as churches and libraries belonging to the armed forces/police. For school textbook orders, discounts are based on the volume and value of an order, ranging from 8% to 15%.
  ^c Publishers can realize promotional sales and offer discounts varying from 15% to 25%.
  ^d Repealed during the preparation of this article. Given that the regime is discussed above, we have decided to keep the information on it in this table.

Noten

• 1 Throughout this article, when referring to ‘RPM arrangements’ and/or ‘FBP regimes’, we are not drawing any conceptual distinction. They are synonyms for the purpose of this article. Indeed, from a competition policy standpoint, RPM arrangements and FBP regimes are different in only one main aspect: their extent. FBP arrangements are state-sanctioned and applicable to the whole relevant jurisdiction, whereas RPM arrangements are usually implemented by certain market players and assessed from a competition standpoint individually. In the case of a publisher-implemented FBP regime, this means the whole downstream market – i.e., bookstores, and all other retailers – are subject to the specific pricing policies of the upstream market. As noted in the following, we however do not focus on this aspect of FBP regimes. We treat FBP regimes as minimum RPM arrangements, and assess them from an antitrust standpoint based on that premise.

• 2 Brazil’s current legal framework already contains a provision that expressly allows publishers to fix the purchase price of books (Article 60 of Law No. 9,610/1998). By way of this provision, publishers may fix the price of books with retailers via supply contracts (at least in theory). However, due to the increased risk of antitrust scrutiny by the Administrative Council for Economic Defense (CADE), Brazil’s competition enforcer, Brazilian publishers have refrained from fixing prices, choosing instead to simply recommend them to bookstores and other companies of the retail segment.

• 3 See, e.g., a discussion of the Brazilian FBP Bill in Brazil’s most influential newspaper, Folha de São Paulo, available at: <www1.folha.uol.com.br/ilustrada/2017/04/1871549-autores-nunca-foram-tao-influentes-diz-ceo-da-editora-harpercollins.shtml>.

• 4 See the official press release reporting this development, which is available at: <http://www12.senado.leg.br/noticias/materias/2017/08/23/ccj-aprova-criacao-de-politica-nacional-para-precos-de-livros>.

• 5 In Olczak’s words: “RPM has been tolerated in this market predominantly because of the cultural importance of the product (see Ringstad, 2004) and for example the judgement in the 1962 UK Restrictive Practice Court decision to allow RPM to continue emphasized that ‘books are different’” (M. Olczak, ‘The impact of recommended retail prices on bilateral bargaining and retail pricing’, October 2011, p. 2). Moreover, according to the International Publishers Association, “Supporters of FBP point out that books have special value; they are indispensable to our individual development, as well as to society. As objects of culture, they therefore deserve to be treated differently from other tradable commodities.” International Publishers Association, Global Fixed Price Report, available at: <www.internationalpublishers.org/images/news/2014/global-fixed-price-report.pdf>.

• 6 See M. Canoy et al., ‘Chapter 21: The Economics of Books’, in Handbook of the Economics of Art and Culture, Vol. 1, Amsterdam, Elsevier, 2006, p. 742.

• 7 For an in-depth compilation of what constitutes ‘standard antitrust analysis’, see: L.M. Kahn, ‘Amazon’s Antitrust Paradox’, Yale Law Journal, Vol. 126, No. 3, January 2017.

• 8 A sophisticated distinction of the different values associated with ‘social regulation’ and ‘economic regulation’ can be found in: E. Windholz & G. Hodge, ‘Conceptualising Social and Economic Regulation: Implications for Modern Regulators and Regulatory Activity’, Monash University Law Review, Vol. 38, No. 2, 2012.

• 9 For the purpose of our analysis, we will rely on the theoretical framework set out by Windholz & Hodge, 2012, which we will discuss in details below.

• 10 See V. Vara, ‘Is Amazon Creating a Cultural Monopoly?’, The New Yorker, available at: <www.newyorker.com/business/currency/is-amazon-creating-a-cultural-monopoly>. See also ibid., p. 5.

• 11 Associação Nacional de Livrarias, ‘A localização das livrarias brasileiras (2014)’. Available at: <http://anl.org.br/web/pdf/noticias/livrarias_brasileiras_2014.pdf>.

• 12 See: A. Singh, ‘Decline of the Independent Bookshops in the UK has Fallen to Just 987, with Surrey’s Oldest Bookshop the Latest Victim’, The Telegraph, 22 February 2014, available at: <www.telegraph.co.uk/culture/books/booknews/10654506/Decline-of-the-independent-bookshop-as-UK-figures-fall-below-1000-for-first-time.html>.

• 13 The authors admitted to some extent that while “the effect of abolition on book prices appears to have been downwards, as hoped,” “(…) problems of measurement and aggregation mean that we should admit some uncertainty on this count” (S. Davies et al., ‘The Benefits from Competition: Some Illustrative UK Cases’, DTI Economic Papers no 9, Centre for Competition Policy, University of East Anglia, 2014, p. 46, available at: <https://core.ac.uk/download/files/7/9633876.pdf>).

• 14 A. Flood, ‘UK Publishes More Books Per Capita Than Any Other Country, Report Shows’, The Guardian, 22 October 2014, available at: <https://www.theguardian.com/books/2014/oct/22/uk-publishes-more-books-per-capita-million-report>.

• 15 Source of data, available at: <www.snel.org.br/wp-content/themes/snel/docs/Palestras/apresentacaojeanguy.pdf>.

• 16 These statements were obtained from a slide deck presentation entitled Fixed Book Price System in Germany, drafted in 2015 by Joachim Kaufmann, Vice-President of the Bonnier Books New Markets, for the International Forum on Fixed Price Law in Rio. Available at: <www.snel.org.br/wp-content/themes/snel/docs/Palestras/apresentacaoalemanha.pdf>.

• 17 RPM is a pricing arrangement in which an upstream firm – usually, a manufacturer – seeks to influence the price of its own products by downstream firms – usually, distributors. These arrangements are usually implemented with the assistance of coercive mechanisms, such as fines or the overall exclusion from the distributor network in the event of noncompliance.

• 18 See M. Motta, Competition Policy: Theory and Practice, Cambridge, Cambridge University Press, 2004, p. 158 et seq. See also L. Peeperkorn, ‘Resale Price Maintenance and Its Alleged Efficiencies’, European Competition Journal, Vol. 4, 2008, p. 201 et seq.

• 19 See Paragraph No. 224 of European Commission, ‘Guidelines on Vertical Restraints’, 10 May 2010, p. 80, available at: <http://eur-lex.europa.eu/legal-content/PT/TXT/PDF/?uri=CELEX:52010SC0411&from=PT>.

• 20 Available at: <www.snel.org.br/wp-content/themes/snel/docs/Material/global.pdf>.

• 21 European Parliament, ‘European Parliament Resolution with Recommendations to the Commission on the Drawing-up of a Directive of the European Parliament and of the Council on the Fixing of Book Prices (2001/2061(INI))’, Strasbourg, 16 May 2002, available at: <www.europarl.europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+TA+P5-TA-2002-0244+0+DOC+XML+V0//EN>.

• 22 Ibid.

• 23 Concurrences, ‘The Swiss Parliament is Discussing a Bill Destined to Reintroduce Fixed Book Prices, Once Sanctioned by Swiss Competition Authorities for Violating Competition Law’, Bulletin, News Issues, May 2009, available at: <www.concurrences.com/Bulletin/News-Issues/May-2009/The-Swiss-Parliament-is-discussing>.

• 24 Ibid., p. 11.

• 25 Davies et al., 2014, p. 31.

• 26 International Publishers Association, Global Fixed Price Report, available at: <www.internationalpublishers.org/images/news/2014/global-fixed-price-report.pdf>.

• 27 Davies et al., 2014, p. 31.

• 28 Ibid., p. 38.

• 29 Ibid., p. 24.

• 30 M. Appelman, ‘Fixed Book Price’, in R. Towse (Ed.), A Handbook of Cultural Economics, Elgar Publishing, Cheltenham, 2003, p. 240.

• 31 M.L. Palma Martos & L. Palma Martos, “Fixed Book Pricing in Spain: A Debate Between Economic Efficiency and Cultural Diversity’, 14th International Conference of the ACEI, Vienna, 6-9 July 2006, p. 6, available at: <www.fokus.or.at/fileadmin/fokus/user/downloads/acei_paper/Palma-Martos.pdf>.

• 32 Tribunal de Defensa de la Competencia, ‘Informe sobre el precio fijo de los libros’, September, 1997, available at: <https://www.cnmc.es/Portals/0/Ficheros/Promocion/Informes_y_Estudios_Sectoriales/1997/1.pdf>.

• 33 Palma Martos & Palma Martos, 2006, p. 7.

• 34 Free translation. Original text available at: <https://www.boe.es/buscar/act.php?id=BOE-A-2007-12351>.

• 35 We cannot say all cases, because there is no data available for some of the countries.

• 36 Palma Martos & Palma Martos, 2006, p. 7, at 3.

• 37 F.L. Barba, ‘El precio único del libro en México’, Tiempo, Apuntes, p. 84, available at: <www.uam.mx/difusion/casadeltiempo/19_iv_may_2009/casa_del_tiempo_eIV_num19_81_85.pdf>.

• 38 “Un aspecto central en esta determinación fue la consideración de que cualquier programa de fomento a la lectura estará destinado al fracasso si no pasa por un proyecto integral que fomente la producción de libros y el acceso a ellos en todos los rincones del país en condiciones semejantes. Igualmente, el argumento de peso en los legisladores mexicanos fue que el libro no es una mercancía como las otras. El libro es un bien cultural que merece la protección del Estado debido a su función social más que mercantil” (Ibid.).

• 39 Ibid., p. 82.

• 40 CNN, ‘La Corte avala la ley que fija precios únicos para los libros en México’, CNN, 1 September 2011, available at: <http://mexico.cnn.com/entretenimiento/2011/09/01/la-corte-avala-la-ley-que-fija-precios-vunicos-para-los-libros-en-mexico>.

• 41 V. Juárez, ‘Reporte global sobre el precio fijo del libro’, Uvejota, 19 June 2014, available at: <http://uvejota.com/articles/3577/reporte-global-sobre-el-precio-fijo-del-libro/>.

• 42 Herzog Fox & Neeman, ‘Israel’s new Authors Act’, Lexocology, 28 January 2014, available at: <www.lexology.com/library/detail.aspx?g=00148cfe-eae1-455b-96fd-93698702da3e>.

• 43 Ibid.

• 44 Kisch and Regev argued that: “The law created an absurd situation in which the legislature’s intervention harmed the competitive market and increased prices for the citizen… The legislature intervened too much in the free market economy, and the purpose of canceling the law is to bring back the former situation, and allow prices to drop and the spread of new books in the Israeli market to increase” (The Jerusalem Post, ‘Cancellation of Law Which Increased Book Prices Moves Forward’, The Jerusalem Post, 20 March 2016, available at: <www.jpost.com/Israel-News/Politics-And-Diplomacy/Cancellation-of-law-which-increased-book-prices-moves-forward-448574>).

• 45 The Jerusalem Post, 2016.

• 46 Israel National News, ‘Book Law’ Banning Discounts on Book Sales Repealed’, Israel National News, 31 May 2016, available at: <www.israelnationalnews.com/News/News.aspx/213037#.V1DhAeRyy9i>.

• 47 Windholz & Hodge, 2012, p. 227.

• 48 Ibid., p. 233.

• 49 Ibid.

• 50 Ibid.

• 51 Ibid., p. 234.

• 52 Ibid., p. 237.

• 53 Stucke, for example, is of the view that: “now is the time to reconsider antitrust’s political, social, and moral concerns. In reconsidering the goals of competition as a means to secure political, economic, and individual freedoms, antitrust can be more responsive to citizens’ concerns about promoting well-being” (M.E. Stucke, ‘Reconsidering Antitrust’s Goals’, Boston College Review, Vol. 53, 2012, p. 624). Pitofsky, on the other hand, notes that: “(…) if the free-market sector of the economy is allowed to develop under antitrust rules that are blind to all but economic concerns, the likely result will be an economy so dominated by a few corporate giants that it will be impossible for the state not to play a more intrusive role in economic affairs” (R. Pitofsky, ‘The Political Content of Antitrust’, University of Pennsylvania Law Review, Vol. 127, 1979, p. 1051).

• 54 On February 2001, CADE dismissed an RPM complaint against publishers and textbook distribution companies because of the fact that, under the relevant agreement, compliance with the prices was not mandatory, but rather merely suggested. As noted by the case’s Reporting Commissioner, “the prices suggestions the publishers are disclosing do not constitute antitrust violations under Brazilian law, since the Brazilian Copyright Law sanctions this conduct, obligating publishers to suggest book cover prices to booksellers. It is based on this price suggestion that authors are payed royalties for their manuscripts” (Case No. 08000.001191/95-64; Defendant: Empresas Editoras e Distribuidoras de Livros Didáticos; decision of 14 February 2001). On 27 July 2011, CADE once again dismissed na RPM complaint against several publishers. According to Reporting Commissioner Olavo Chinaglia “(…) what we see in this case is that publisher are drafting price lists – a conduct that is expressly sanctioned by law – which are used as a reference in the booksellers’ pricing policies. The case files lead one to conclude that, in addition to suggested prices being efficient in the market under investigation, compliance or not with the suggested prices is a strategic and circumstantial decision up to each bookseller, and there is apparently no threat or possibility of retaliation if bookstores choose to charge different prices” (Case No. 08012.001743/2002-40; Defendants: Editora Globo, et al.; decision of 27 July 2011).

• 55 The document is available at: <https://www.senado.gov.br/atividade/materia/getPDF.asp?t=160838&tp=1>.

• 56 The European Commission Guidelines on Vertical Restraints mentions additional anticompetitive effects resulting from RPM arrangements, such as preventing distributors from reducing their retail prices, potentially raising prices consumers will have to pay and leading distributors to favor the brand that adopts the RPM scheme (due to the mark-up guaranteed by the fixed price), therefore harming the competitors of the manufacturer that fixed prices.

• 57 T. Overstreet, ‘Resale Price Maintenance: Economic Theories and Empirical Evidence’, USFTC Bureau of Economics Staff Report, 1980, available at: <https://www.ftc.gov/sites/default/files/documents/reports/resale-price-maintenance-economic-theories-and-empirical-evidence/233105.pdf>.

• 58 Elzinga & Mills, for example, note that “(…) just as real world examples of retailer cartels stitched together by RPM are not common, real world examples of manufacturer cartels that used RPM to curtail cheating are uncommon. Indeed, the economic nexus between cartels and RPM is not robust.” See K. Elzinga & D. Mills, ‘The Economics of Resale Price Maintenance’, in Issues on Competition Law and Policy, ABA Section of Antitrust Law, 2008, p. 1847. See, in the identical sense, R. Blair & D. Kaserman, Antitrust Economics, 2nd ed., Oxford, Oxford University Press, 2009.

• 59 In the words of Organization for Economic Cooperation’s (“OECD”) background note: “(…) the available empirical evidence is scant and dated, and it does not show that RPM always tends to restrict competition and reduce output. On the contrary, it points in the other direction. Perhaps most importantly, it indicates that RPM is not commonly used to facilitate cartels.” See OECD, Roundtable on resale price maintenance – Background note, 2008, p. 38, available at: <https://www.oecd.org/daf/competition/43835526.pdf>.

• 60 See S. Ornstein, ‘Resale Price Maintenance and Cartels’, The Antitrust Bulletin, No. 30, 1985, pp. 401-423.

• 61 Additionally, Kleit notes that “(…) the rise of RPM in the 1880s can be directly attributed to the rise in importance of advertised branded goods. Given this, and the fact that horizontal restraints predate RPM by at least several centuries, it seems unlikely that the primary motivation for RPM is cartel facilitation.” See A. Kleit, ‘Efficiencies without Economists: The Early Years of Resale Price Maintenance’, Southern Economic Journal, Vol. 59, No. 4, 1993, p. 617.

• 62 See, e.g., H. Marvel, ‘Resale Price Maintenance and the Rule of Reason’, The Antitrust Source, 2008, p. 1. See also R. Blair, ‘The Demise of Dr. Miles: Some Troubling Consequences’, Antitrust Bulletin, No. 53, 2008, p. 137.

• 63 See Blair & Kaserman, 2009.

• 64 See Elzinga & Mills, 2008, p. 1846.

• 65 See A. Jones & B. Sufrin, EU Competition Law, Oxford, Oxford University Press, 2014, p. 787. As emphasized by OECD: “market conditions do not always favour cartel formation. The well-known group of factors that affect the plausibility of a successful cartel in general (such as the number of firms in the market, the homogeneity of their products, the ease of entry, etc.) will make cartels too difficult to maintain in some markets, even with the assistance of RPM” (emphasis added). OECD, supra note 59, p. 30.

• 66 See F. Sá Earp & G. Kornis, A economia do livro: a crise atual e uma proposta de política, IE/UFRJ, Rio de Janeiro, Text for discussion no 04/2005, p. 24.

• 67 See, e.g., R. Whish & D. Bailey, Competition Law, 7th ed., Oxford, Oxford University Press, 2012, p. 522.

• 68 Ibid., p. 64, at 4.

• 69 See G. Mello, ‘Desafios para o setor editorial brasileiro de livros na era digital’, BNDES Setorial, Economia da Cultura, No. 36, p. 444.

• 70 As well noted by OECD: “Generally, if barriers to entry are low or if substitute products exist, collusion will not be successful. If, on the contrary, high entry barriers protect market incumbents from competitive pressure from potential new entrants, it is more likely that the collusive outcome will be reached and maintained sufficiently stable over time.” OECD, Public Procurement, Competition Policy Roundtables, 2007, p. 21.

• 71 See Sá Earp & Kornis, 2005, p. 4.

• 72 See M. Levenstein, ‘What Determines Cartel Success?’, Journal of Economic Literature, Vol. 44. No. 1, 2006.

• 73 See Sá Earp & Kornis, 2005, p. 9.

• 74 See data from “3^a Edição da Pesquisa Retratos da Leitura no Brasil,” funded and carried out by Instituto Pró-Livro e o Ibope.

• 75 In Henry Ergas words: “as a general matter, competition policy aims at protecting, and where appropriate and efficient extending, the range of choices available to consumers” (H. Ergas, ‘The Interaction and Coordination of Competition Policy and Consumer Policy: Challenges and Possibilities’, in OECD, The Interface between Competition and Consumer Policies, OECD Policy Roundtables, 2008, p. 18. See also: N.W. Averitt & R.H. Lande, ‘Consumer Sovereignty: A Unified Theory of Antitrust and Consumer Protection Law’, Antitrust Law Journal, 2007. See also: L. Sylvan, Activating Competition: The Consumer Protection – Competition Interface, Trade Practices Workshop, University South Australia, 2004.

• 76 The size and complexity of books are factors that influence the marginal utility consumers confer on the product. The reason for this is that these factors determine the opportunity costs associated with reading, and therefore affect the marginal utility. The opportunity costs associated with reading books that are less intellectually challenging are lower than the costs associated with reading dense, complicated books. See F. Van der Ploeg, ‘Beyond the Dogma of the Fixed Book Price Agreement’, Journal of Cultural Economics, No. 28, 2004, p. 6.

• 77 See CBL/SNEL, Produção e Vendas do Setor Editorial Brasileiro – Produto 3: Histórico do setor com dados de 2013 e relatório sobre a nova metodologia, São Paulo, FIPE, 2014, pp. 9/10.

• 78 F. Sá-Earp & G. Kornis, Preço único do livro: solução frágil para um problema grave, p. 7, available at: <www.snel.org.br/wp-content/themes/snel/docs/Material/preco.pdf>.

• 79 See: <www.internationalpublishers.org/news/blog/entry/why-fixed-book-price-is-essential-for-real-competition>.

• 80 As noted by Bittar, “independent libraries have a central role in strengthening the ‘book culture’ and the habit of reading. With specialized services, such as attentive care and a well-qualified staff, they end up being one of the most effective ways of introducing new, rising authors or challenging literature into the market. In many places, these bookstores are cultural centers, where literary events and workshops are organized. Instead of prioritizing a stock of new, bestselling books, independent bookstores usually make most of their money with books of with no commercial appeal, or books of specialized niches.” (A.C. Bittar, ‘Digital Rights Management, Concorrência e Acesso ao Conhecimento no Mercado e Livros Digitais’, Master’s dissertation presented to the Escola de Direito de São Paulo da Fundação Getulio Vargas, 2015, p. 122.)

• 81 Van Der Ploeg explains: “Since profits under monopoly pricing are evidently higher than profits in a competitive equilibrium (y + a − F > b + a − F), more book titles are profitable and thus more titles are published/sold under the FBP than in competitive equilibrium. It is possible to print and sell extra books at low and almost non-increasing marginal cost, so that the producer loss b is likely to be small. Also, the price elasticity of the demand for books ε is likely to be small as a large part of the full cost of reading is the opportunity cost of time. Hence, the monopoly mark-up is likely to be large and monopoly profits y are large. It thus seems likely that many more book titles are published under the FBP than under perfect competition.” (Van Der Ploeg, 2004).

• 82 See Rouet, François, Le Livre, mutations d'une industrie culturelle, Les Études de la documentation française, 2 ed., Paris: La Documentation française, 2007.

• 83 On this argument, see Van der Ploeg, 2004.

• 84 Data extracted from: <www.snel.org.br/wp-content/uploads/2015/07/apresentacao_MarcosdaVeigaPereira_.pdf>.

• 85 See Van Der Ploeg, 2004, p. 5.

• 86 See P. Nelson, Philip, ‘Information and Consumer Behavior’, Journal of Political Economy, Vol. 78, No. 2, March/April 1970.

• 87 Fixed costs must be paid by the company regardless of how productive the month/year was (in the case of booksellers, regardless of how many books were sold). Examples of fixed costs are rent and employee salaries.

• 88 That is the scenario noted by Lemos: “[I]f one considers the existence of bookstores as a proxy for access to books, the market in Brazil is very small. The country as a whole has approximately 2000 bookstores or an average of just one bookstore for every 84,400 Brazilians. The absolute majority of Brazilian cities do not have any bookstores. Until roughly 10 years ago, entire states such as Roraima, Tocantins, and Amapá have only two bookstores each. In the overall Northern region of the country, the average is one bookstore for every 215,300 inhabitants.” (R. Lemos, From Legal Commons to Social Commons, Centre for Brazilian Studies, University of Oxford, Working Paper no 80, 2006, p. 13.)

• 89 See supra note 1.

• 90 See: <www.publishersweekly.com/pw/by-topic/industry-news/bookselling/article/57631-where-the-stores-are.html>.

• 91 There are currently 4,500 bookstores in the country according to data made available by El País, a prominent local newspaper. See: <http://elpais.com/diario/2011/04/22/sociedad/1303423201_850215.html>.

• 92 See M. Uribe, ‘El acceso al libro y el precio único’, Pensar El Libro, No. 4, 2006.

• 93 See: <www.aljazeera.com/news/2015/07/argentine-capital-world-centre-bookstores-150716143741268.html>.

• 94 According to data made available by the “3^a Edição da Pesquisa Retratos da Leitura no Brasil,” a report prepared by the Instituto Pró-Livro and IBOPE. In 2007, there were 95.6 million readers in Brazil. In 2011, the figure fell to 88.2 million. One must bear in mind that the report considers ‘readers’ to be individuals that read, completely or in part, at least one book in the last 3 months. The number of readers who have the habit of reading every day, or at least more than once a week, is much lower.

• 95 See supra note 48.

• 96 See C. Mattos, ‘Fixação do Preço de Revenda Mínima no CADE: o caso SKF’, Revista do Ibrac, Vol. 24, 2013, p. 82.

• 97 One of the main efficiencies associated with RPM arrangements generally is that they contribute to strengthening interbrand competition (See Elzinga & Mills, 2008, p. 1844). Thus, one might argue that even with an FBP regime in force, retailers would still compete, but rather than over prices, (i.e., who is capable of offering the best discounts), over investments on the improvement of customer services. Elzinga & Mills explain: “The motivation of a manufacturer using RPM is not to enable retailers to raise prices and gouge consumers. The economic logic of using RPM to induce efficient retail service is for the manufacturer to impose a price floor to restrain downstream price competition in order to foster service competition. A manufacturer has no incentive to overcompensate its retailers; doing so would reduce the manufacturer’s profits and would make the manufacturer’s product line less competitive in the marketplace. Instead, the manufacturer wants its retailers to make just enough money to market its brand effectively, and no more.” Elzinga & Mills, Ibid. By definition, RPM increases retailer profit margins, allowing retailers to invest in the improvements of customer services (Mattos, 2013, p. 82. See also W.L. Holahan, ‘A Theoretical Analysis of Resale Price Maintenance’, Journal of Economic Theory, Vol. 21, 1979, pp. 411-420 and F. Mathewson & R. Winter, ‘The Law and Economics of Resale Price Maintenance’, Review of Industrial Economics, Vol. 13, 1998, pp. 57-84 and R. Deneckere, H.P. Marvel, & J. Peck, ‘Demand Uncertainty and Price Maintenance: Markdowns as Destructive Competition’, American Economic Review, Vol. 87, 1997, pp. 619-641). In order to differentiate themselves from rivals, bookstores would indeed be more encouraged to invest in, for example, training their staff, cleaning their facilities, and putting together a diverse stock of books, all to the benefit of consumers.

• 98 See G. Porcidonio, Letras e diversões: para atrair público e driblar crise, livrarias e sebos investem em shows, festas e peças, RioShow, 2015, available at: <http://rioshow.oglobo.globo.com/eventos/eventos/eventos-em-livrarias-parte-2-13489.aspx>.

• 99 F.M. Scherer & D. Ross, Industrial Market Structure and Economic Performance, Houghton Mifflin Company, Boston, 3rd ed., 1990, p. 552.

• 100 The need to invert the consumer’s purchasing process – leading consumers to stop behaving on the lines of ‘I need/want this and will look for it’, and start behaving more like ‘I found it! This is what I wanted’” – in order to stimulate the habit of reading in Brazil was the topic of a presentation Júlio Cesar A.S. da Cruz gave in the Public Hearing of the Fixed Book Price Policy in the Brazilian Senate, on 30 July 2015. Available at: <http://legis.senado.leg.br/comissoes/reuniao;jsessionid=4D59CB1DDF4690CC8D8A49FB2BAD8D84?0&reuniao=3547&codcol=47>.

• 101 See D. Vargas, Jurista Criativo, Cadernos FGV Direito Rio no 11, 2014, p. 93.