قیمت گذاری داروها با پوشش بیمه سلامتی ناهمگن

In this paper, we examine the role of insurance coverage in explaining the generic competition paradox in a two-stage game involving a single producer of brand-name drugs and n quantity-competing producers of generic drugs. Independently of brand loyalty, which some studies rely upon to explain the paradox, we show that heterogeneity in insurance coverage may result in higher prices of brand-name drugs following generic entry. With market segmentation based on insurance coverage present in both the pre- and post-entry stages, the paradox can arise when the two types of drugs are highly substitutable and the market is quite profitable but does not have to arise when the two types of drugs are highly differentiated. However, with market segmentation occurring only after generic entry, the paradox can arise when the two types of drugs are weakly substitutable, provided, however, that the industry is not very profitable. In both cases, that is, when market segmentation is present in the pre-entry stage and when it is not, the paradox becomes more likely to arise as the market expands and/or insurance companies decrease deductibles applied on the purchase of generic drugs.

The development and growth of the generic pharmaceutical industry over the past 25 years has come in response to rising healthcare costs. In 2004, healthcare costs represented 15.3 percent of GDP in the United States, the highest share among OECD countries, followed by Switzerland (11.6 percent), Germany (10.9 percent), and France (10.5 percent), all above the OECD average of 8.9 percent (OECD, 2007). Private expenditures per capita were also the highest in the United States, more than double the private expenditures per capita of any other OECD country. For all OECD countries, the rate of growth of health spending per capita increased over the period 1999–2004 by more than 5 percent per year. With pharmaceutical expenditures representing 10–25 percent of health expenditures (OECD), many countries have attempted to promote the use of generic drugs in a variety of ways in order to keep healthcare costs down while maintaining or even increasing accessibility to pharmaceuticals and retaining incentives to invest in innovation and research and development. In 1984, the Drug Price Competition and Patent Term Restoration Act, also known as the Waxman–Hatch Act, was introduced in the U.S. in order to improve generic competition by lowering barriers to entry for generic drugs and to increase patent terms for new drugs delayed by complicated and time-consuming approval procedures of the U.S. Food and Drug Administration (FDA), the agency responsible for the safety and efficacy of drugs. Under this legislation, (duplicative) testing for generic drugs was eliminated and replaced by the requirement that an Abbreviated New Drug Application (ANDA) be submitted by generic entrants demonstrating the equivalence between their products and the original (brand-name) drugs. Not surprisingly, the entry of generic drugs into the pharmaceutical market intensified dramatically following the introduction of the Waxman–Hatch Act, and as a result of the expiration of patents on many high-sales-volume brand-name drugs (Frank and Salkever, 1997). In response to the increased market share of generic drugs and lower prices of pharmaceuticals overall, the prices of brand-name drugs did not fall consistently with the predictions of traditional market entry models; instead, they were often observed to increase. This phenomenon is often referred to as the generic competition paradox, or GCP (Scherer, 1993). The first evidence of the Wagner and Duffy (1988) who found substantial price increases associated with entry despite significant decreases in generic prices among top selling name brand drugs. Several other instances of support for the paradox were provided in subsequent studies, including those by Grabowski and Vernon (1992), Frank and Salkever (1997), Perloff et al. (1995), and, to a lesser extent, Caves et al. (1991). Grabowski and Vernon, using data on 18 major drugs in the mid-1980s, showed that prices for name brands increased on average by 7 percent after entry and continued to increase in the following year. Frank and Salkever utilized data on the patent expiration of 45 drugs facing competition for the first time between 1984 and 1987, and found that, although significant market share was lost upon entry by brand-name producers, the price of their product generally increased. Similarly, Perloff et al. showed brand-name price increases using data from the mid-1980s from the U.S. anti-ulcer drug market. Although Caves et al. did not observe price increases in their study of 30 drugs between 1976 and 1987, they found prices to decrease by only small amounts following entry (2 percent on patent loss, although this loss increased with the number of entrants) and by far less than the price decrease experienced by entrants. In other studies, including the work by Wiggins and Maness (2004) on 98 anti-infectives from 1984 to 1990, no evidence in support of the paradox was detected. Traditional oligopolistic models of entry suggest that the increased competition caused by generic entry should drive prices down for all firms, as illustrated by a move from monopoly to duopoly with homogeneous goods. While most, if not all, models attempting to explain the paradox would suggest that average prices fall after entry, the standard models do not explain why the price charged by the incumbent firm could increase after entry. Several explanations have been proposed to theoretically support the empirical finding that prices of brand-name drugs increase after entry. For example, but outside of the realm of the pharmaceutical industry, models of entry-induced price increases in oligopolistic or monopolistically competitive markets, including those by Satterthwaite (1979), Salop (1979), Rosenthal (1980), suggest that economies of scale or specific demand curve changes can lead to post-entry price increases. 1 Due to the nature of pharmaceutical production, economies of scale are not typically significant and therefore this is not a likely explanation for the paradox. However, several papers employ changes in the elasticity of demand to explain the paradox, including the brand-loyalty models of Caves et al. (1991), Grabowski and Vernon (1992), Frank and Salkever (1997), and Kamien and Zang (1999). 2 In these models, exogenous segmentation of the market occurs upon entry, as one group of consumers is price sensitive while another is not. 3 The segmentation is exogenous in the sense that the individual groups exist separately in the market prior to entry but the brand-name producer is not permitted to choose whether or not it wants to serve only one group prior to entry, and the size of each group remains constant before and after entry. Other studies attempting to explain the paradox introduce price stickiness into models due to imperfectly informed doctors (Bhattacharya and Vogt, 2003), product differentiation with collusion or price competition (Perloff et al., 1995), or quality differences (Berndt et al., 1993 and Griliches and Cockbrun, 1994). Bhattacharya and Vogt argue that doctors’ stock of knowledge about the presence and efficacy of new generics evolves slowly and is manipulated by producers through advertising. Perloff et al. show that the paradox can occur when products are significantly differentiated in product space. Pre-entry, the firm lowers its price to serve segments of the market located far away from its product in its characteristics, but increases its price once entry occurs and those consumers switch to the entrant's product. Product location is exogenous, and the paradox is not possible for cases in which the entrant's product and the incumbent's product are closely related (little product differentiation). Berndt et al. and Griliches and Cockbrun have that price increases are generated by quality improvements, and that prices increase over the life of a product (although increase more slowly post-entry). In this paper, we combine some of the features of the models of previous studies, including product differentiation and brand loyalty, but focus on the role of insurance coverage in the segmentation of the market. Specifically, we examine the endogenous segmentation of the market by the brand-name producer, both before and after entry, to determine whether or not insurance coverage can explain the paradox, and its relation to the other theories of the literature. The relevance of insurance in pricing decisions has empirical support. Hellerstein (1994), for example, using prescription data from the eight largest therapeutic drug classes, describes how most individual doctors prescribe both brand-name and generic drugs (suggesting that a lack of awareness or knowledge is not driving the prescription decision). Furthermore, doctors with higher fractions of Medicaid, Medicare, HMO, and privately insured patients are more likely to prescribe generics, although the links are not particularly strong (or even negative) for certain drug classes. Pavcnik (2002) suggests that brand-name pricing is very sensitive to out-of-pocket expenditures, and estimates, using data from Germany, that the price adjustment to an exogenous change in insurance coverage ranges between 10 percent and 26 percent. In a cross-country study, Danzon and Chao (2000) show that the effect of generic competition on brand-name prices depends on the insurance coverage and pricing regime, and conclude that countries with fee pricing (like the U.S.) tend to experience large decreases but countries with strict reimbursement regulation and insurance (like France, Italy, and Japan) tend to experience price increases.4 In the present analysis, we construct a model in which consumers differ on the basis of coverage, and the brand-name producer can choose, through its price for the brand-name drug and taking into account the impact of its own decisions on generic pricing, which consumers it wants to target and which consumers it wants to leave out of the market (pre-entry) or to the generic producers (post-entry).5 The inclusion of a parameter, q, into the utility function describing consumers’ preferences, which captures the perceived quality differential between brand-name and generic drugs, allows for a separation between the price effects of brand loyalty as reflected in q and the price effects of segmentation induced by insurance coverage heterogeneity. We thus derive conditions under which the GCP occurs in instances in which brand loyalty (or the perceived quality differential) alone does not give rise to price increases for brand-name drugs when generic drugs are introduced so that the paradox can only be attributed to insurance coverage considerations. In the context of the model, we revisit the brand loyalty argument to explain the paradox and show that brand loyalty does not always yield increases in the price of brand-name drugs after generic entry; whether brand loyalty leads to price increases depends upon other parameters of the model such as the degree of product differentiation, production costs, and the willingness to pay for pharmaceuticals. When the brand loyalty argument does not apply and the market is segmented on the basis of insurance coverage (at least after the introduction of generic drugs), we show that the price of brand-name drugs can increase following generic entry at low levels of product differentiation and can decrease at high levels of product differentiation. The extent of substitutability between brand-name and generic drugs is not by itself a key factor in determining whether the GCP arises and does not necessarily involve unambiguous effects on the likelihood of the paradox, as in the study by Perloff et al. (1995); instead, a combination of product differentiation and market profitability, along with market size and preferential treatment of purchases of generic drugs by insurance companies, is a stronger determinant of when the paradox is likely to emerge so that lower product differentiation can support the paradox and higher product differentiation can yield no paradox. When brand-name and generic drugs are close substitutes, the producer of brand-name drugs responds to generic entry by supplying to fewer consumers (those who have better insurance coverage) and can thus charge a higher price; such a strategy becomes less appealing as the market becomes less profitable so that the GCP is more likely to occur when the two types of drugs are not very differentiated but the market is very profitable. What happens if brand-name and generic drugs are not close substitutes depends on whether segmentation occurs prior to generic entry: if it does, the producer of brand-name drugs responds to generic entry by supplying to more consumers and must thus charge a lower price so that the paradox does not occur when the two types of drugs are highly differentiated and is even less likely as the market becomes more profitable; if it does not, it responds to generic entry by supplying to fewer consumers (moving away from providing drugs to the entire market) especially when the market is less profitable so that the GCP is more likely to occur at high levels of product differentiation but low levels of market profitability. The larger the market and/or the lower the deductible insurance companies apply on the purchase of generic drugs relative to the deductible on the purchase of brand-name drugs, the stronger the incentive to segment is and the more likely it is for the GCP to arise. We structure the remainder of the paper as follows: in Section 2, we introduce the model and derive the equilibrium before and after generic entry both in the absence of health insurance coverage considerations (that is, when consumers are homogenous) in Section 2.1 and when consumers differ in their health insurance coverage in Section 2.2; in Section 3, we provide a numerical example; in Section 4, we give concluding remarks.

In this paper, we examine the relevance of insurance coverage heterogeneity in explaining the GCP, that is, the observation that the price of brand-name drugs increases following the introduction of generic drugs, in a two-stage game involving a single producer of brand-name drugs and a fixed number of quantity-competing producers of generic drugs. Past theoretical work attempting to explain the paradox has relied on product differentiation, exogenous market segmentation, and brand loyalty. Here we suggest that, even in cases where differentiation and brand loyalty alone would not result in brand-name price increases, endogenous segmentation on the basis of insurance coverage by brand-name producers can reverse the decline in prices caused by generic competition. While empirical work has indicated that insurance may play an important role in the paradox, no theoretical study to date has examined the role of endogenous segmentation by insurance coverage. In our model, when consumers are homogenous in their insurance coverage, the GCP can arise provided that consumers’ perceived quality differential between brand-name and generic drugs exceeds some threshold level which depends positively upon the degree of substitutability between the two types of drugs and the willingness to pay for drugs, and negatively upon marginal production costs. With the perceived quality differential at a level which does not give rise to price increases following the introduction of generic drugs when consumers have similar coverages (that is, with the brand loyalty argument assumed away), we show that heterogeneity in insurance coverage leads to the paradox not only when the market is fully covered in the pre-entry stage, so that we can expect a price increase as a result of the market shrinkage due to the segmentation, but also when the market is partially covered prior to generic entry, although the paradox is less likely to arise in the latter case. In particular, and independently of market coverage, a price increase for brand-name drugs is likely to result as generic drugs become available when the two types of drugs are highly substitutable and the market is profitable. Under full market coverage, the paradox is also likely to arise when the two types of drugs are highly differentiated and the market is not particularly profitable. While the extent of substitutability between the two types of drugs is an important factor in the determination of how the price of brand-name drugs adjusts as generic drugs make their way into the pharmaceutical market, although its effects turn counter with respect to previous studies (e.g., Perloff et al., 1995), market profitability is equally important. Within the context of our model with only interior solutions being considered, it is then possible for the paradox to arise when the two types of drugs are highly substitutable in some cases (consistent with other studies which rely on product differentiation) but not in other cases, depending on how profitable the industry is. Furthermore, under full market coverage, the paradox may arise when the two types of drugs are either highly substitutable (and the industry highly profitable) or highly differentiated (and the industry highly unprofitable). The set of feasible combinations of product substitutability and industry profitability levels at which the paradox is observable is also dependent upon the size of the market and the additional coverage on the purchase of generic drugs for a given insurance package. Specifically, as the market expands (this is equivalent to an improvement in the best insurance package available) and/or insurance companies decrease deductibles applied on the purchase of generic drugs, the GCP becomes more likely to arise. In essence, at low levels of product differentiation, the producer of brand-name drugs responds to generic entry by providing its product to fewer consumers (those with better insurance coverage) and can thus supply it at a higher price; as the market becomes less profitable and/or the best insurance package involves a lower coverage and/or the difference in deductibles between the two types of drugs decreases, the incentive to shrink the segment of the market buying brand-name drugs weakens making the GCP less likely to occur. On the other hand, at high levels of product differentiation, the producer of brand-name drugs responds to generic entry by offering its product to more consumers, and must thus charge a lower price, when it does not cover the entire market prior to generic entry, but to fewer consumers, and can thus charge a higher price, when it covers the entire market prior to generic entry. In the latter case, as the market shrinks in terms of insurance packages available (with the best coverage decreasing) and/or insurance companies offer a lower deductible on the purchase of generic drugs, the incentive to supply to fewer consumers weakens so that the GCP becomes less likely to occur or, equivalently, the GCP is more likely to arise when the market is not profitable. As a final note, in this paper, we do not attempt to cover every factor influencing the pricing decisions of drug producers. Many healthcare systems of the world include variations of more heavily regulated reference pricing. We leave consideration of the effects of insurance heterogeneity in these frameworks to future research.