Pharmaceutical markets in Japan and the United States

Although the United States and Japan are the two largest markets for branded pharmaceuticals, their structures and market performance are quite different. Like its European counterparts, a Japanese government ministry sets the prices paid to national and international drug companies. In contrast, US prices are set largely through negotiations between private health insurers and drug companies. Related to these different structures are quite different results. In particular, for pharmaceuticals sold in both countries, US prices are about twice what they are in Japan. Since many of the same companies sell pharmaceuticals in both countries, these differences require an explanation. A frequent one offered is that Japanese authorities exercise their inherent monopsony power to obtain the low prices, which private insurers in the United States are unable to impose. While correct on the surface, this explanation fails to confront deeper policy issue concerned with promoting public health. Because pharmaceutical research and development is a global public good, there are strong incentives for buyers to “free ride” on the outlays of others. Under those circumstances, global R&D outlays, based on country-specific incentives, would be set below optimal levels, and the supply price of substantial pharmaceutical innovation would not be covered. Because the US market is more than half the world total, its incentive structure is fundamentally different from that of Japan. Indeed, higher US drug prices, which drive worldwide incentives, are needed to promote current numbers of new pharmaceuticals introduced in both countries.


Introduction
The United States and Japan are the two largest markets for branded pharmaceuticals in the world. In a sample of 33 OECD developed countries, these two countries represented 45% of the total volume of prescription drugs dispensed in 2018: with the US accounting for 24% and Japan 21% of the 33-nation total [21, p 19]. Pharmaceutical markets in other OECD countries are much smaller in terms of the physical volume of pharmaceuticals consumed.
Comparing dollars spent on pharmaceuticals, the comparisons are different. In that case, the US pharmaceutical market is again the largest, now by an even greater magnitude. It accounts for fully 58% of OECD sales while Japan lies in second place with just over 9% [21, p. 19]. On a per capita basis, Japan spends $580 on pharmaceuticals per person while the US spends more than double that amount at $1420 per capita [21, p 19]. Other countries spend much less, with Germany at just under $40 billion in aggregate sales and $481 per capita [21, p. 19].
What these data indicate is that US pharmaceutical expenditures are much higher than elsewhere even while the physical volume consumed per capita is not unusually high. The clear reason for these differences are higher branded drug prices.
The question explored here is how to account for the substantial price differences between the two countries. In the discussion below, we first describe the unique features of both US and Japanese markets along with their different regulatory regimes. However, rather than looking at these regimes as exogenous factors resulting from political forces, the second part employs a recognized economic model to explain regulatory and legislative decisions. Indeed, the discussion there suggests that actions taken by regulatory regimes are endogenously determined and result from underlying economic conditions.

Prices and quantities in Japan and the United States
The Japanese pricing structure Consider the following summary from the Japanese National Institute of Public Health: Similar to the systems in France and Germany, the Japanese healthcare system is a multi-payer system. Private payers are not allowed, and all payers are public insurers.
[Furthermore], the Ministry of Health, Labour and Welfare … is responsible for determining the price of medicines [26, p. 270].
In this structure, prices are set through procedures determined by the Ministry.
For drugs with alternatives available for the same medical indication, "the daily price of the new medicine is to be equal to that of the comparable drug, to secure fair competition in the market" [34, pp. [13][14][15]. Price competition is then intentionally absent as between drugs. In contrast, for drugs without available alternatives, the "cost calculation method" is used, which rests on manufacturing, R&D and marketing costs along with provision for an "operating profit of 14.7%" [34, pp. [13][14][15]. Furthermore, drug prices calculated through either method are "adjusted in case of large disparity between average foreign price, [as determined by the] average of US, UK, Germany and France." Pharmaceuticals priced at more than 25% higher than the average foreign price are lowered, while those priced more than 25% below the average foreign price are increased [34, p. 16]. The price setting schedule for generic drugs is different: for newly listed drugs "it shall be 50% of the original product; 40% if the number of brands exceeds 10" [34, p. 19].
Despite regulatory controls and the generally moderate prices set for branded dugs, the Health Ministry has confronted increasing pharmaceutical expenditures. These increased outlays result from both an increasingly aged population needing more drugs and also an increasing number of "expensive new technologies" [8, p. 43] Under these pressures, and after some years in trial and discussion, the Ministry adopted in April 2019 a new drug evaluation system resting specifically on costeffectiveness analysis. However, not all drugs would be subject to this type of evaluation [8, pp. 43-45].
The ostensible purpose of this additional evaluation was to adjust prices downward for drugs that did not meet defined cost-effectiveness criteria. From available data, incremental cost-effectiveness ratios (ICERs) would be derived. If the ICER is less than Y5 million ($ 43,800) per Quality Adjusted Life Year (QALY), its price would not be changed, although at higher ratios, a step-wise function would be used to adjust price downward [8, pp. 46-47].

The US pricing structure
The US pharmaceutical pricing structure is fundamentally different from that found in Japan or elsewhere. As Professor Rice points out in his recent book, "nearly all of the countries either set pharmaceutical price levels or engage in explicit negotiations with manufacturers." However, "none of these activities are carried out by the US federal government" [22, p. 209]. Instead, "current U.S. policy … prohibits government negotiation and relies instead on competition" [22, p. 210]. That US policy rests on Competition rather than Regulation is both correct and long-standing. As far back as 1984, with the passage of the Hatch-Waxman Act, that policy direction was established.
This policy direction, however, does not mean that government resources are absent. As reported in Table 1, more than 40% of all US retail prescriptions are supported by the two largest government programs: Medicare Part D for those over 65 and Medicaid for low-income citizens. While these programs inject a large volume of government funding for pharmaceuticals, they lie in total below that of private insurers and other self-funding entities like unions and large corporations in terms of numbers of prescriptions dispensed. Because of substantial US tax advantages from doing so, employee compensation packages commonly involve "benefits" which include a substantial share of health insurance premiums. Through this indirect process, the effective buyers of most dispensed pharmaceuticals are private insurers financed largely through employee benefit packages.
The impact of this structure is that US consumers increasingly do not pay directly for their pharmaceuticals. See the data provided in Fig. 1. While in 1960, pharmaceuticals were purchased like most commodities from their own funds, that share has fallen to less than 20% currently. Indeed, only 5% of all retail prescriptions were 1 3 paid in cash in 2019. Public and private entities are the predominant agents on the buying side of US pharmaceutical markets.
The prices charged for branded pharmaceuticals are set in negotiations on individual products between drug manufacturers as sellers and private insurers as buyers. An important feature of this marketplace is that by legislative mandate, the large government programs do not negotiate prices separately with the drug companies. This controversial provision means in effect that the leading government agencies cannot exercise monopsonistic power over the prices paid but instead must accept those resulting from private negotiations. Both sides in these pricing negotiations are major companies and both have strategic actions available to them. Drug companies as sellers can bring an approved product to market at an announced "list" price. In response, insurance companies like most buyers can determine the quantities to be purchased. Critically, the fact that a drug is prescribed by a physician does not necessarily mean that an insurer will pay for it.
Among the actions frequently taken by insurers in response to high-priced drugs are the following: to require detailed physician statements as to medical necessity, to increase co-pays or other consumer charges, or to impose step-therapy requirements that require lower priced drugs are used first before a higher priced alternative is approved. Through such means, insurance companies can limit the quantities purchased of expensive pharmaceuticals.
In these circumstances, a frequent outcome is that insurance companies expand coverage and thereby increase the quantities of a drug sold in return for reduced prices, whether in the form of lower list prices or increased rebates off list prices. The essential point here is that in these price negotiations, both drug companies and insurance companies have important roles to play in setting final prices.

Branded and generic pharmaceuticals
Another important difference between the US and Japanese markets is the differing roles played by generic pharmaceuticals in the two countries. In the United States, the share by volume of Unbranded Generics is fully 84% of the 2018 total, while in Japan generic drugs account for only 34% [21, p. 20]. Unbranded generics represent a dominant share of the US market but only about one third of the Japanese market. A likely explanation for this difference is that for newly listed generics, the Japanese price will be "50% of the original product, [but] for oral medicine, 40% if the number of brands exceeds 10" [34, p. 19]. In contrast, US generic prices depend on the number of competing sellers. See the data provided in Fig. 2. With lower US generic prices, one finds much larger shares of generic pharmaceuticals.
Although American consumers use more generic drugs than their Japanese counterparts, their revenue shares in the two countries are similar. In the United States, generics account for 12% of pharmaceutical revenues and in Japan only 13% [21, p. 20]. Apparently, the lower Japanese volume of generic pharmaceuticals is largely balanced by the higher prices charged for those drugs. In both countries, revenues flow predominantly to the leading branded drug companies, as indicated in Table 2.
The expansion in US generic sales was the direct result of government intervention into the pharmaceutical marketplace. Enacted in 1984, the Hatch Waxman Act substantially altered the structure of drug approval regulations as employed by the Food and Drug Administration (FDA). While new chemical entities still require proof of both safety and efficacy before market entry is permitted, US producers need only demonstrate bio-equivalency with the original pharmaceutical to launch a generic product. Under the new law, generic entry has expanded sharply, and generic quantities expanded from about 14% of new prescriptions in 1984 to over 90% by 2019.
As anticipated, increased generic penetration has had an impact on pharmaceutical spending per capita in the United States. Between 1984, when the new law was enacted, and 2016, generic sales reduced the growth rate of real spending by an estimated 2.26% annually [13]. In the same period, however, enhanced insurance payments led to a higher spending growth followed by higher prices charged for branded pharmaceuticals. As generic substitutes stood ready to compete with the original branded drugs upon patent expiration, the major branded companies raised prices more rapidly before reaching their expected patent cliffs [4, pp. 12-20]. The increasing prominence of generic pharmaceuticals has led to lower prices on generics but appear also to have led to higher prices for patented, branded pharmaceuticals.
When the 1984 Hatch Waxman law was passed, large molecule biologics were relatively unimportant factors in US pharmaceutical markets and were not addressed specifically. More recently, however, those medications have become more common such that by 2018, fully 17 out of 59 new drugs approved by the FDA were biologics. A critical feature of those medications is that manufacturing processes are far more difficult and costly than with small molecule drugs. Furthermore, the final product invariably varies in structure and effect as between manufacturers, and even among batches produced by the same manufacturer. As a result, it was inherently more difficult to demonstrate "bioequivalence" as required by the Hatch Waxman Act. Indeed, the FDA has acknowledged that the current law did not establish a "pathway" for successor versions of approved biologics. A new law, passed in 2010, addressed these issues, 1 and effectively created the concept of a "Biosimilar." To gain FDA approval, successor biologics were required instead to have no "meaningful" differences from the innovative product. After a slow start, increasing numbers of biosimilar drugs have been approved. There are already eleven branded molecules with biosimilar competition, and many others are in process [10]. While adding an element of price competition to selected pharmaceutical markets, their high manufacturing costs make it unlikely that highly competitive generic markets will arise in these circumstances as they did with small molecule generic drugs.

The Japanese pharmaceutical industry
The ten largest Japanese pharmaceutical companies are listed below, based on Japanese 2019 sales [  The Japanese industry includes both domestic and international companies. One anticipates therefore similar menus of pharmaceuticals in the two countries. Between 2008 and 2019, 447 new drugs were approved by the FDA in the United States while 400 were approved in Japan. Among the new drugs approved in Japan, 202 were first approved in the US, 82 in Europe, 80 in Japan and 36 elsewhere [29, pp. 1268-1274]. In addition, the US is Japan's largest pharmaceutical trading partner: 22% of Japan's pharmaceutical exports are shipped to the US and 23.5% of Japan's pharmaceutical imports are received from the US [12].

Demand conditions in Japan and the United States
Although government support of pharmaceutical purchases is substantial in both countries, there are important differences. While Japanese demand is government-driven with little or no private intervention, just over half of all dispensed prescriptions in the United States are paid by private entities. Of net payments made in 2019 by health insurers and other payers in the US of $356 billion, patient costs represented $82 billion, or 23% [9]. On average, these costs came to $19.44 per prescription. In contrast, consumer costs per prescription in Japan ranged from 10 to 30% of total pharmaceutical costs, depending on the patient's age and health insurance plan [14, pp. 1-11] These data suggest that consumer costs as a proportion of manufacturer charges are broadly similar in the two countries.
While Japanese prices are set according to established government policies, US prices are determined in a decentralized manner. A relevant question is whether outcomes are likely to be fundamentally different as a result; and in particular, whether the 2019 adoption of cost effectiveness (CE) evaluation procedures might lead Japanese and American branded prices to converge.
Because US prices paid by health insurers and other payers reflect a drug's therapeutic value, the new CE procedures do not necessarily affect relative prices in the two countries. Indeed, there is evidence that relative prices established through British government imposed CE requirements do not vary significantly on average from those set under the market-based US system [3]. While international differences remain, relative prices are not closely tied to regulatory decisions on individual products' cost effectiveness.

The willingness to pay for branded pharmaceuticals
Despite their supply-side similarities, Japanese and American branded pharmaceutical companies face quite different demand-side factors, which have led to striking price differences. See Table 3 that provides pharmaceutical price indices for US branded and generic drugs relative to those in Japan, Germany and the U.K. As indicated there, US branded originator prices are much higher than those reported for the other three countries with the Japanese price difference lying between the other two. In contrast, for the case of unbranded generics, US prices are lower than in Japan and elsewhere.
The conventional explanation for these observed price differences is that government price controls are employed in Japan and elsewhere while market processes determine prices in the United States, and simply that markets are less effective in restraining high branded prices. While correct on a procedural basis, that explanation is not sufficient for it presumes that regulatory decisions are made arbitrarily and without regard to market realities. In effect, the conventional explanation presumes that regulatory decisions in the US and elsewhere are made for budgetary reasons related to payments for the current menu of products; and with little concern for supporting further innovation.
Final prices for branded drugs depend critically on the prospective buyers' Willingness to Pay for the drug at issue. Whether the buyer is a government Ministry as in Japan, or a private insurance company as in the United States may be less important. In both cases, the seller cannot charge more than a buyer is willing to pay. Particularly in the case where single sellers confront buyers who act on behalf of large segments of effective demand, the underlying reservation prices of both buyers and sellers are relevant for determining final prices.
In Japan, the Ministry determines prices under public guidelines. As was recently pointed out: Japan has faced rapidly rising healthcare expenditures, mainly owing to … launch of some expensive health technologies [that]… has exacerbated This impending situation increased the pressure to incorporate economic evaluation into the healthcare decision-making process [8, p. 43].
In addition, A price-raising scheme for highly cost-effective products has been introduced for the purpose of evaluating innovation … for an ICER below Y2 million per QALY ($17,300) and … is highly innovative [8, p. 47].
While the latter provision is an important component of the new cost-effective program, minimizing expenditures for existing pharmaceuticals may have been its principal regulatory purpose. In contrast, the US demand structure rests on a different set of presumptions. Not only are government outlays on pharmaceuticals less than half of total expenditures but also there are strong pressures to promote pharmaceutical innovation. In this realm, US incentives are different from those present in Japan and elsewhere, and it is such incentives that have led to much higher US drug prices.
As far back as 1984, US legislation in the form of the Hatch-Waxman Act effectively created the US Generic Pharmaceutical Industry as distinct from the corresponding branded industry. In effect, there would now be two US pharmaceutical industries. While the purpose of the new generic industry was to set low prices on products whose patent protection had expired, the acknowledged task of the established branded industry was specifically to engage in extensive research and development efforts designed to find new therapeutically advanced pharmaceuticals. When the new law was debated, it was recognized that branded drug prices might thereby be higher, but that result was considered an acceptable cost for the resulting health benefits in the US and throughout the world.

Pharmaceutical industry R&D as a global public good
Pharmaceutical Research and Development is an economic activity that rests fundamentally on what output is being produced and what revenues are anticipated. On this point, there is considerable evidence that pharmaceutical price regulation designed to lower drug prices leads directly to reduced R&D outlays [7,11,15,25]. In contrast, higher prices incentivize increased R&D outlays and promotes the development of new pharmaceuticals [31]. As stated in a recent US government report "the gains from global sales of innovative products drive incentives for research and development" [32, p. 8]. Moreover, a critical feature of the information generated through R&D, which deals with the safety and efficacy of new drugs, is its availability internationally. It has the trappings of a global public good.
Such goods, often referred to as Collective Goods, are defined by requirements of both non-exclusivity and non-rivalry [17, p. 695]. The first requirement turns on whether users can be excluded from reaping its benefits because the good is employed elsewhere; with Public Goods, exclusion is not feasible. The second requirement is that once provided, "additional units can be consumed at zero social marginal costs" [17, p. 695]. This second condition is sometimes considered that of "joint supply" such that its use by one party does not detract from its use by others [2, p. 5]. In its most common presentation, each party cannot prevent others from benefitting from his or her efforts nor can each diminish what is available to others. With collective goods, the amount supplied is necessarily the same for all.
To be sure, this usage requires that the information generated is recognized and valued internationally. To the extent that regulatory authorities require, for example, that the clinical testing be carried out in the same country, the "public good" feature of this information is diminished, but not entirely so. Even where trials are conducted elsewhere, they still denote what has been done and likely can be repeated.
A critical feature of the classic public goods discussion is that independent decision-making, commonly represented as actions leading to a Nash equilibrium, fails to provide sufficient quantities of the good in question. Because each party recognizes the presence of others such that their individual efforts would contribute little to the total available quantity, he or she becomes a "free rider hoping to benefit from the expenditures of others. If every person adopts this strategy, then no resources will be allocated to public goods" [17, p. 697]. Even though all parties would benefit collectively from the good's availability, each party's efforts individually to minimize costs leads it not to be supplied.
That conclusion applies to a great variety of collective actions, and is most prominently explored in Olson's classic volume [18]. Applying the same concept to a nation, he writes that "a state is first of all an organization that provides public goods for its members, the citizens" [18, p. 15]. However, it commonly does so by imposing sanctions or other forms of coercion precisely to avoid "free rider problems." And among the most prominent of such public goods is Public Health.
However, Olson acknowledges an important exception to the general rule that an element of coercion is required: [Where] one or more members get such a large fraction of the total benefit that they find it worthwhile to see that the collective good is provided, even if they have to pay the entire cost -may get along (i.e. the collective good is provided) without any group agreement or organization [18, p. 46] Not only does Olson suggest that conclusion applies to citizens within a country, but it does also to countries within a broader alliance. What is proposed here is that US policies facilitating global pharmaceutical R&D can be explained by this exception.
In a succeeding paper, Olson and Zeckhouser apply these concepts to the functioning of alliances formed for a specific purpose, whose objective they characterize as a Global Public Good [19]. Starting from the premise that "individuals (nations) acting independently do not have an incentive to provide optimal amounts of such goods," (19) pp. 267-268) alliances arise specifically to provide for their common objective in sufficient quantities. Indeed, the authors suggest that since common interests bind the members of the group, "it may well be that most alliances are never embodied in any formal agreement" [19, p. 273].
Olson and Zeckhauser reach the following conclusion: There will also be a tendency for the 'larger' members, -those that place a higher absolute value on the public good -to bear a disproportionate share of the burden [19, p. 268].
That paper spawned a large literature that sought to apply the theory of alliances to various public concerns, including specifically public health [23]. But the underlying premise of alliances remained: that independent decision-making leads to sub-optimal outcomes such that collective actions are needed. This conclusion, however, recognizes that "burdens are anticipated to be shared in a disproportionate fashion, leading a large rich ally to allocate a greater share of its GDP to [the common purpose] than a small, poor ally" [23, p. 875] This unequal sharing of burdens has become known as "the exploitation hypothesis" even when fully acknowledged by the impacted party.

Global R&D expenditures as alliance funding
There are insights from the economic literatures on both Public Goods and Alliances that apply specifically to the prospective health gains resulting from pharmaceutical innovation. That these health benefits are substantial is reported in a recent study applicable to the United States. Between 1990 and 2015, US life expectancy increased by 3.3 years, and this research apportioned that improvement among various contributory factors [1]. The authors recognized the twelve most significant factors which together contributed 85%, or 2.9 years, to the aggregate gain. Among these factors, Pharmaceuticals was the second most important, after specific Public Health measures. It represented a net gain of 35% of the twelve leading effects. These findings, the authors point out, "underscore the central role of medications in explaining reduced mortality" [1, pp. 1551-1554]. Similar effects should apply to Japan as well with the highest life expectancy in the world particularly since the beneficial effects of pharmaceuticals are most pronounced in older people [6, p. 42].
While sales of innovative products can be limited by patent protection and related policies to authorized companies, the related product information is available worldwide. Although this information is not limited internationally, it is costly to create. In particular, Phase 3 clinical trials can cost upwards of $100 million each [5, 24, p. 553]. In a sample of ten large US companies, which introduced 106 new drugs between 2009 and 2018, the average cost per introduction was $2.8 billion in 2018 dollars,while for a larger sample of 47 somewhat smaller companies which introduced 63 new agents, costs were lower at $1.3 billion per introduction [33].
Providing sufficient demand-side revenues and incentives are required for these investments to be made. However, once these costs are covered, all receipts exceeding the direct costs of manufacturing add to a firm's profitability. In effect, the high margins received from branded pharmaceutical sales in the United States are those supporting the global public good resulting from pharmaceutical industry R&D.
A recent US government report emphasized that pharmaceutical R&D supports a global public good: Worldwide profits drive innovation incentives, but when worldwide profits are partially determined by centralized pricing of governments, this induces unique free-riding issues…. It is in each country's interest to have other countries provide the returns to fund innovation by generous reimbursement…. Put simply, providing innovative returns is a global public goods problem that leads to classic under-provision through government free-riding [31, pp. 10-11] That judgment reflects the point suggested earlier; that pharmaceutical R&D supports a global public good, for which most countries understandingly contribute relatively little to its development cost. As noted above, that conduct represents rational independent conduct. However, as also consistent with the economics of alliances, the largest member, the United States, contributes an estimated 78% of OECD profits to those earned from pharmaceuticals [31, p. 11], and thereby provides most of the incentives needed for its creation. Since prospective profits drive R&D investments, the United States predominantly finances this global public good thorough its much higher prices.

Policy conclusions
Pharmaceutical markets in Japan and the United States are configured differently but with important similarities. In both countries, large international drug companies play major roles, and presumably sell many of the same products. These international companies are also those engaged in substantial R&D efforts intended to develop new medications which are then introduced world-wide. However, there is often a substantial drug lag with most new pharmaceuticals introduced first in the United States.
Where conditions between the two countries diverge most is in the prices charged for the same drugs: branded prices are much higher and generic prices much lower in the United States. And it is these pricing differences which foment the strong policy debate in the United States, where there are frequent proposals to introduce government regulations designed to limit US prices to international levels [30].
One possible explanation is that countries with higher values of statistical lives spend more heavily on health care including pharmaceuticals. However, limited to comparisons between Japan and the United States, that hypothesis is not supported by available data. Revealed values of statistical lives (VSL) in Japan exceed $10 million, which is more than twice comparable estimates for the United States [27, p. 257].
The issues raised by high US branded prices are those explored theoretically in the economic literatures on global public goods and the formation of global alliances. Because research findings regarding new pharmaceuticals have world-wide implications, rational government agencies maximize their individual effectiveness by relying on the efforts of others. While such actions are commonly derided as "free-riding," they merely follow from good governmental practice, and are expected in the presence of global public goods.
A common solution to such issues is the formation of explicit or implicit alliances where each member agrees to fund a share of the costs of the public good. However, in many such circumstances, "the exploitation hypothesis" arises in which the largest members bear a disproportionate share of joint costs. 3 In the limit, and under certain conditions, the largest members find it advantageous to provide individually for the public good even while other members will benefit at little cost to themselves. Given the prospective benefits to the United States from pharmaceutical R&D, bearing the full costs through higher drug prices has been its preferred policy choice.
To be sure, that conclusion turns on the presence of a direct connection between higher drug prices and higher levels of pharmaceutical innovation. From available data, the trade-off between branded prices and pharmaceutical innovation appears unassailable. While the United States bears most of these costs, enhanced world health is the beneficiary.