Paper to be presented at the Summer Conference 2010 on "Opening Up Innovation: Strategy, Organization and Technology" at Imperial College London Business School, June 16 - 18, 2010
BRAZILIAN PHARMACEUTICAL INDUSTRY AND GENERIC DRUGS: THE POLITICAL INTENTIONS AND THE UNEXPECTED CHANGES Thiago Caliari
[email protected] Ricardo Ruiz CEDEPLAR-UFMG
[email protected]
Abstract: The main goal in introducing generic medicines into the Brazilian market was to decrease prices. However, in addition to this effect, such policy brought about unintended changes in the industrial structure. In this paper we evaluate the intentional impacts as well as the unplanned impacts from the introduction of generic medicines, particularly with respect to scale and innovation. Our main results show: (a) a negative impact on price indexes, which is nonetheless attenuated by sectoral regulation: (b) a large increase in scale in favor of domestic companies, but with just a small increase in R&D expenses and innovation within those companies. Such increase in scale should be seen as a window of opportunity for sectoral industrial policies aiming at larger scale and profitability, associated with commitment to innovative investments, which is certainly a difficult policy to make.
JEL - codes: O30, L11, I18
Brazilian Pharmaceutical Industry and Generic Drugs: The Political Intentions and the Unexpected Changes
1
1. Introduction The importance of pharmaceutical industry is undisputable and for this reason many sectoral industrial policies highlight it as a key sector for the course of action. In Brazil, such approach could be noticed in the Industrial, Technological and Foreign Trade Policy (PITCE) of 2003, in which the pharmaceutical industry was one of the four major sectors for intervention. The sector is also highlighted in the current Policy for Productive Development (PDP), of 2008, which places the pharmaceutical industry within the so-called Health Industrial Complex as one of the strategic areas for sectoral investment. Both PITCE and PDP differ from the policies adopted during the nineties, which were characterized by horizontal strategies with little concern for sectoral specificities. Such policies had profound impacts on the domestic pharmaceutical sector and their final result was the prevalence of multinational corporations in terms of market share, which has only changed in the current decade, after the introduction of generic drugs in the Brazilian market. The policy of generic drugs was driven mainly by social and budgetary motivations, and not so much by an industrial motivation, given that its main goals were the decline of prices for the final consumer and the reduction of public expenses on health care. However, this policy caused significant changes in the domestic productive structure. The goal of this paper is to evaluate the policy’s objectives of increasing welfare, along with the unintended results in the industrial structure. The remainder of the paper is organized as follows. In section 2, we present the domestic industry and the generic drugs policy, focusing on the specificities of the Brazilian case. In section 3, we discuss the intended effects of the policy, namely: decrease in prices. In section 4, we present the studies and results on the unintended impacts of the policy on the productive structure, grouped as descriptive analyses of scale and profitability, technological effort and, lastly, econometric analyses of R&D determinants. Section 5 concludes the paper, and it is followed by references in section 6 and appendices in section 7. 2. Domestic Industry and the Generic Drugs The structure of the Brazilian pharmaceutical industry in the nineties was closely related to the horizontal industrial policies implemented in the beginning of that decade. Among these policies, the ones with greatest impact for the sector were: (a) decline in import tariffs, specially after 1994, (b) exchange rate fluctuations, (c) price liberalization after 1992, (d) implementation of a new patent legislation in 1996 after signing the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) in 1995, and (e) strong decrease in the financial instruments for supporting sectoral investment (e.g. funds from the National Bank for Economic and Social Development and from the Financing Agency of Industrial Projects and Research1) (Queiroz & Gonzáles, 2001; Romano, 2005). Such policies brought about important changes in the industrial structure. Some of the well-known results are: (a) increase in imports along with a low growth of exports, (b) acquisition of domestic companies by foreign enterprises, (c) significant increase in prices, (d) deverticalization and productive specialization, (e) stagnation of domestic production or even contraction of production in some segments (Frenkel, 2002; Gadelha, 2002). Furthermore, the incipient national productive structure was hit by a strong foreign competition during the nineties due to the decline in import tariffs. According to Oliveira (2005), 1096 industrial firms were closed and 355 projects were canceled, only in the area of fine chemistry. Trade openness and price liberalization did not bring about an increase in consumption or a decline in prices, as expected. Actually, the opposite was true: increasing
1
In Portuguese: Banco Nacional de Desenvolvimento Econômico e Social and Financiadora de Projetos Industriais e de Pesquisa, respectively. 2
prices and stagnation in consumption. Such price increases were contained by the Plano Real, the economic stabilization plan, but only for a short period of time. Figure 1 compares the prices of pharmaceutical products with the general market price index (IGP-M). From 1995 to 1999, there is a new process of sectoral inflation. The relative price increase during this period can be partly explained by price liberalization (end of price controls), by the limits for imitation due to the TRIPS and by the end of import restrictions. The combination of these effects allowed for an increase in the price of drugs along with external dependence in terms of trade and technology. In sum, by the end of the nineties, the medicine industry presented important structural problems: trade deficits, technological dependence and price increases. The policy of generic drugs is introduced within such context, and its main goal was to minimize public and private expenditures. It was not strictly an industrial policy, since its only goal was to cheapen public purchases in order to reduce expenses on drugs and to ease the access by the population. The implementation of the policy in February 1999, via the law number 97872, was based mainly on such propositions3. Figure 1 Evolution of Prices of Pharmaceutical Products (1989-1999) 130 125 120 115
Exchange Depreciation
Generic Drugs Law CAMED Price Adjustment
110 105 CMED Price Adjustment 100 95 90 85
ago/94 fev/95 ago/95 fev/96 ago/96 fev/97 ago/97 fev/98 ago/98 fev/99 ago/99 fev/00 ago/00 fev/01 ago/01 fev/02 ago/02 fev/03 ago/03 fev/04 ago/04 fev/05 ago/05 fev/06 ago/06 fev/07 ago/07 fev/08 ago/08 fev/09 ago/09
80
INPC Produtos Farmacêuticos / IGP-M IGP-M / IGP-M Source: Author’s calculations based on data from IBGE and IPEAData.
It should be noted that the introduction of generic drugs in Brazil is late. In the USA, the official implementation of generic drugs was in 1984, with the Hatch-Waxman Act (Berndt,
2
The generic medicine – according to the Brazilian Common Denomination (DCB), which specifies the name of the pharmacologically active principles – is a medicine similar to a reference or innovative product, to which it aims to be a substitute, and it is in general produced after the expiration or renunciation of patent protection or other rights of exclusivity, provided that its efficacy, safety and quality have been proved, and it is designated by DCB, or in its absence, by the International Common Denomination (DCI) (Brasil, 2003). 3 The Minister of Health at that time, Jose Serra, would often speak about the need of wider and easier access to drugs by the Brazilian population as a crucial precondition for the implementation of a health policy in the country. 3
2002)4. In India, the institution of the Indian Patents Act in 1970 created patents protection only for processes – and only for three years – which allowed for reverse engineering and technological learning for the production of existing drugs (Fink, 2000; Kremer, 2002; Grace, 2004). Such late entry in the production of generic drugs in Brazil is nowadays still reflected in the low share of generic drugs in the domestic market. According to data from Pró-Genéricos (2009), the market share of generic drugs in Brazil in 2008 was only 17%, in comparison with 60% in the USA and in the UK, for instance5. Although the main objective was to reduce prices, the unintended effects of the policy were strong and brought about a change in the Brazilian productive structure. We could say for sure that the domestic companies were only able to achieve a competitive scale after this policy was implemented. Table 1 shows the change in market share in benefit of the domestic companies. Table 1 Market share of pharmaceutical companies in the Brazilian market (1998 and 2007) Firms
Country of Origin
Market Share (1998)
Firms
Country of Origin
Market Share (2007) 7.1%
Novartis
Switzerland
6,3%
EMS Sigma Pharma
Brazil
Roche
Switzerland
5,5%
Sanofi-Aventis
France
6.4%
USA
5,4%
Ache
Brazil
5.6%
Germany
5,2%
Medley
Brazil
5.5%
Brazil
4,7%
Novartis
Switzerland
4.4%
Janssen Cilag
Belgium
3,7%
Eurofarma
Brazil
3.5%
Boehringer Ing.
Germany
3,7%
Pfizer
Glaxo Wellcome
United Kingdom
3,5%
Bayer Schering Plough
Schering Plough
Germany
3,2%
Bristol-Meyers Squibb Hoechst Marion Roussel Aché/Prodome
Eli Lilly Demais empresas
USA
3.4%
Germany
3.2%
Boehringer Ing.
Germany
2.6%
Denmark
2.4%
-
55.9%
USA
3,0%
Nycomed
-
55,8%
Demais empresas
Source: Callegari (2000) for 1998 e IMS Health, MIDAS (2007) for 2007. Note: Hoechst Marion Roussel was purchased by Sanofi-Aventis in 1999; Schering Plough merged with Merck (EUA) in 2009; Janssen Cilag is part of the Johnson & Johnson group since 1961.
The table presents the largest companies in the domestic market in 1998 and 2007. In 1998 only one firm among the ten largest had domestic capital: Aché. In 2007, in turn, four of the ten largest companies had participation of domestic capital, namely: EMS Sigma Pharma, Aché, Medley and Eurofarma, all producing generic drugs. Such increase in the participation of national firms has been accompanied by an increase in the share of generic drugs in the domestic market. According to the website Pró-genéricos (2009), 88% of the supply of generic drugs in the domestic market comes from Brazilian firms. Therefore, it is clear that, despite other policies aiming to stimulate the domestic pharmaceutical sector, the policy of generic drugs has been the most successful in terms of scale of production, allowing some domestic firms to appear among the largest ones in the national market. However, there are still some doubts regarding technological capacities. The key question is: what is the position of domestic firms after the policy has been implemented in terms of technological capacity?
4
The Hatch-Waxman Act was signed in 1984, but generic drugs have been sold in the USA since 1973. The world market for generic drugs grows approximately 17% per year and represents approximately US$ 55 billion (Pró-Genéricos, 2009). The main highlight is the American market, which represents approximately 40% of the total. 4 5
As mentioned before, the generic drugs use diffused technology, i.e. they are copies of products which have been in the market for a long time. In this case, can the entry of domestic firms in the market for generic drugs generate virtuous cycles of innovation, and allow for technological catching up? Will these firms be able to appear among the leader firms, not only in scale but also in terms of innovative content? Or, has the policy of domestic drugs been able to only promote national leadership in low-tech products? Before being able to evaluate the unintended effects of the policy of generic drugs (changes in market share, in scale and in technological capacities of the firms), it is important to examine the intended effects of the policy, namely, price reductions. 3. Policy Intentions: decrease in prices 3.1. The prices of generic and reference drugs The effect of generic drugs on consumer prices is controversial. There are few studies on the magnitude of the impact of generic drugs on health care costs, but there is a fair amount of research that shows the impact of generic drugs on the brand-name drugs to which they refer to. In this perspective, some international studies find interesting and diffuse results. Statman (1981) investigates the price behavior of twelve brand-name products before and after the expiration of their patents. The study shows that the market share of these products in terms of value has declined to 96% of its initial level and that only one of the twelve products showed significant decrease in prices. The small change in market share suggests low generic-price elasticity of the demand of brand-name products. Grabowski & Vernon (1992) have examined the effects of the entry of generic drugs on eighteen products which had their patents expired between 1983 and 1987. By estimating the impact of the number of generic medicine suppliers in the market and the impact of generic prices on brand-name prices, they found that the price of brand-name products increase relatively to generic medicine prices after their entrance in the market. Franck & Salkever (1997) have come to a similar conclusion. They have investigated a sample of 32 drugs whose patents expired during the middle 1980s and their results show the increase in prices of brand-name products after the entry of generic drugs in the market, along with a steep decline in average prices due to the low prices of generic drugs. Some studies have come to opposite results, such as Caves et al (1991) and Griliches & Cockburn (1993). Caves et al (1991) suggest that simply comparing the market before and after generic drugs may bring about inconsistent results because it omits factors which may cause changes in brand-name prices over time. Their results show that the entry of one generic medicine reduces the price of brand-name products in 2%, whereas the entry of twenty generic products reduces prices in 17%. Also, they have estimated a regression along the lines of Grabowski & Vernon (1992) and found that the effect of generic drugs on the price of brandname products would decrease with the increase in the number of generic medicine suppliers. Griliches & Cockburn (1993) discuss the measurement problems of the official indicators in the USA, which do not treat generic and brand-name products as substitutes. They suggest a reformulation of the indicators, stating that the entrance of generic drugs in the market comes with a significant period of declining prices in the early life cycle. Their results show that whereas the inflation indexes show increasing prices of reference drugs after the entrance of generic drugs, the reformulation of these indexes show declining prices of brand-name products. For the Brazilian case, the contributions of Fiúza and Lisboa (2001), Hasenclever (2002) and Nishijima (2008) should be mentioned. Fiúza and Lisboa (2001) suggest that, after the entry of generic drugs, the companies producing reference drugs adopt a “Ramsey pricing” strategy. They differentiate their products increasing selling prices even more, in order to extract consumer surplus in segments with low price-elasticity of demand, leaving the segments with higher price-elasticity of demand for the generic drugs. 5
Hasenclever (2002) and Nishijima (2008) come to different results. Both studies found a negative bias in the prices of reference drugs due to the increasing supply of generic drugs. However, according to Hasenclever (2002), the effect is indirect, through the impact of generic drugs on market concentration. For this author, the increasing competition in the market, either due to significant presence of substitutes or due to the introduction of generic drugs, seems to generate declines in prices of reference drugs. It means that this effect is mostly caused by the competitive interaction between firms and not as much by to substitution effects from the consumer. Despite the diverging results found in the literature, the indication of the Brazilian Ministry of Health and of CMED is that the entry price of generic drugs should be 35% lower than the price of reference drugs. However, what is the influence of this indication and of the market share of generic drugs on the sectoral price index? Is this policy able to bring about price reductions – and therefore increasing welfare – or CMED’s regulation has stronger power to influence final consumer prices? This is the question we aim to answer in this paper. Therefore, differently to the studies mentioned in this section, which focused on the substitution effects of the generic drugs, our main question is whether the policy of generic drugs have had a significant impact on the price level of the medicine industry – which assumes the substitution effects and a massive participation in the industrial supply. 3.2. The Choice of the Price Index and an Evaluation of the Price Evolution Since our main goal is to capture the effects of generic drugs on the final consumer, we chose to analyze the National Consumer Price Index – Pharmaceutical Products (INPC) 6. The reason is the range of household income included in calculation of this index. INPC measures the cost of living for families with monthly income from 1 to 6 times the minimum wage in Brazil, which is closer to the consumer basket of low-income populations. Given that INPC measures a narrow income range, it is more likely that the influence of generic drugs would be captured by this index. This is because there is a higher elasticity of demand and thus higher consumer substitutibility effect on the drugs in this level of income. In addition, the calculation of the price index considers the bestselling products, through a unweighted geometric average of the variations in the prices of the collected products (IBGE, 2007). Thus, the index may be composed of any combination of drugs (generic, similar, and reference drugs), in any number. The INPC Pharmaceutical was chosen given its greatest likelihood to capture the increase in the market share of generic drugs within the income range included in the index7. The model used in the econometric analysis replicates an oligopoly structure with adaptive adjustment a la Bertrand, in which producers adjust prices according to their market share. This is a reasonable assumption for the pharmaceutical market, given that it is highly oligopolized, specially if we consider the high degree of substitution between generic and brand-name drugs, which amplifies the possibility of price wars. For a better illustration,
6
There are basically three different price indexes in Brazil which include specific data and allow for the analysis of the pharmaceutical sector separately: the Wholesale Price Index – Pharmaceutical Products (IPA-OG Pharmaceutical), the Broad Consumer Price Index – Pharmaceutical Products (IPCA Pharmaceutical), and the National Consumer Price Index – Pharmaceutical Products (INPC Pharmaceutical). 7 According to the website Pró-Genéricos (2009), income groups under four times the minimum wage present a limitation of access to drugs of around 65% due to income constraints. It is expected that such income groups would consume lower price drugs, precisely the generic drugs. However, the consumption of similar drugs may also be high within these groups, also due to the low costs of this line of products. This may be a limitation for the use of the indexes. Nevertheless, our model is able to isolate the generic medicine in order to overcome these limitations. 6
assume that the total supply in the sector may be divided in two parts: one related to the supply of generic drugs ( Q G ) and another one related to other drugs ( Q OM ), so that:
QtS = QtG + QtOM
(1)
The producers of generic drugs ( Q G ) fix the prices of generic drugs in the current period ( Pt G ) and the producers of the other drugs ( Q OM ) fix the prices of other drugs in the current period ( Pt DM ). Considering the model a la Bertrand, price competition occurs such that firms decide their prices adaptively, based on past prices, on past market performance (both their own performance and their rivals’), and on the exchange rate ( et )8, so that
Pt G+1 = f ( Pt G , QtG , QtOM , et )
(2)
OM Pt OM , QtG , QtOM , et ) +1 = f ( Pt
(3)
However, the use of different prices for both lines of products is not feasible due to data constraints. Thus the prices are aggregated according to the overall index by sector. Prices Pt G and Pt OM are aggregated in one single index Pt * , which evolves according to the market share of both lines of drugs9. Thus Pt * would be a weighted average of the prices of each line of drugs, namely
Pt* = αPt G + (1 − α ) Pt OM ,
0