Prescription Drugs: Spending Controls in Four European Countries (Chapter
Report, 05/17/94, GAO/HEHS-94-30).

Although the United States has traditionally allowed drug prices to be
determined by the free market, other countries use a variety of policies
to control prescription drug costs. This report analyzes the effects of
pharmaceutical policies in the following four European countries:
France, Germany, Sweden, and the United Kingdom. GAO (1) describes the
strategies these countries use to control prescription drug prices and
limit pharmaceutical spending, (2) reviews the effects of the policies
on drug prices and spending, and (3) analyzes the effects of these
policies on pharmaceutical research and development. GAO found that
controls on pharmaceuticals have succeeded in restraining prices in the
four countries but have not prevented increases in drug spending, owing
to higher consumption and newer, more expensive drugs.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-94-30
     TITLE:  Prescription Drugs: Spending Controls in Four European 
             Countries
      DATE:  05/17/94
   SUBJECT:  Pharmaceutical industry
             Price regulation
             Foreign governments
             Pharmacological research
             Comparative analysis
             Drugs
             Research and development costs
             Price indexes
             Health care cost control
             Price fixing
IDENTIFIER:  France
             Germany
             Sweden
             United Kingdom
             Medicaid Program
             
**************************************************************************
* This file contains an ASCII representation of the text of a GAO        *
* report.  Delineations within the text indicating chapter titles,       *
* headings, and bullets are preserved.  Major divisions and subdivisions *
* of the text, such as Chapters, Sections, and Appendixes, are           *
* identified by double and single lines.  The numbers on the right end   *
* of these lines indicate the position of each of the subsections in the *
* document outline.  These numbers do NOT correspond with the page       *
* numbers of the printed product.                                        *
*                                                                        *
* No attempt has been made to display graphic images, although figure    *
* captions are reproduced. Tables are included, but may not resemble     *
* those in the printed version.                                          *
*                                                                        *
* A printed copy of this report may be obtained from the GAO Document    *
* Distribution Facility by calling (202) 512-6000, by faxing your        *
* request to (301) 258-4066, or by writing to P.O. Box 6015,             *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time.                                    *
**************************************************************************


Cover
================================================================ COVER


Report to the Chairman, Special Committee on Aging, U.S.  Senate

May 1994

PRESCRIPTION DRUGS - SPENDING
CONTROLS IN FOUR EUROPEAN
COUNTRIES

GAO/HEHS-94-30

Prescription Drug Spending Controls


Abbreviations
=============================================================== ABBREV

  AGI - annual gross income
  AMM - Autorisation de mise sur le marchï¿½
  DM - Deutsche Mark
  EC - European Community
  GDP - gross domestic product
  GP - General Medical Practitioner
  IPS - Indicative Prescribing Scheme
  NHS - National Health Service
  NSIB - National Social Insurance Board
  OECD - Organization for Economic Cooperation and Development
  PACT - Prescribing Analyses and Cost
  PPRS - Pharmaceutical Price Regulation Scheme
  R&D - research and development
  RPS - reference price system

Letter
=============================================================== LETTER


B-251111

May 13, 1994

The Honorable David H.  Pryor
Chairman, Special Committee on Aging
United States Senate

Dear Mr.  Chairman: 

This report, prepared at your request, examines how other countries
regulate prescription drug prices, how those policies affect drug
prices, and how they affect pharmaceutical research and development. 

As arranged with your office, unless you publicly announce the
contents of the report earlier, we plan no further distribution until
30 days from its issue date.  At that time we will send copies of the
report to interested congressional committees and other interested
parties.  We will also make copies available to others upon request. 

This report was prepared under the direction of Jonathan Ratner,
Assistant Director, Health Financing and Policy Issues.  If you have
any questions, please call Scott Smith, Assistant Director, who may
be reached on (202) 512-7119.  Major contributors to this report are
listed in appendix VII. 

Sincerely yours,

Sarah F.  Jaggar
Director, Health Financing
 and Policy Issues


EXECUTIVE SUMMARY
============================================================ Chapter 0

While the United States has traditionally let drug prices be
determined by the free market, other countries use a variety of
policies to control prescription drug costs.\1 However, the rising
cost of health care, and increasing prescription drug prices in
particular, has increased the financial burden on vulnerable segments
of the U.S.  population.  In addition, widely reported disparities in
prescription drug prices between the United States and other
industrialized countries have heightened congressional interest in
policies to control pharmaceutical prices.\2

The Chairman of the Senate Special Committee on Aging asked GAO to
study the range of policies to contain prescription drug costs in
other industrialized countries.  The Chairman was particularly
interested in the pharmaceutical cost containment efforts of
countries that--like the United States--are home to strong
research-based pharmaceutical industries.  In response to this
request, GAO analyzed the effects of pharmaceutical policies in four
European countries--France, Germany, Sweden, and the United Kingdom. 
Specifically, this report has three objectives:  (1) to describe the
strategies used in these countries to control prescription drug
prices and limit pharmaceutical spending; (2) to review the effects
of these policies on pharmaceutical prices and spending; and (3) to
analyze the effects of these policies on pharmaceutical research and
development (R&D). 


--------------------
\1 The Omnibus Budget Reconciliation Act of 1990, which requires that
drug manufacturers give Medicaid programs rebates for outpatient
drugs based on the lowest prices available to any purchaser, is an
exception to this rule.  Prior to the passage of this bill, there
were no government controls on drug prices in the United States. 

\2 See, for example, Prescription Drugs:  Companies Typically Charge
More in the United States Than in Canada (GAO/HRD-92-110, Sept.  20,
1992); Prescription Drugs:  Companies Typically Charge More in the
United States Than in the United Kingdom (GAO/HEHS-94-29, Jan.  12,
1994); Association Belge des Consommateurs, Statement Prepared for
the United States Senate Special Committee on Aging, (Nov.  16,
1989). 


   BACKGROUND
---------------------------------------------------------- Chapter 0:1

In the United States, some prescription drugs are purchased by
consumers, some are financed by insurers, and some are paid for by
government programs such as Medicaid.  In contrast to the United
States, prescription drugs in many other countries are financed
entirely through a national health insurance system.  Consequently,
the financial viability of these national health insurance systems
depends on restraining prescription drug costs.  To control
pharmaceutical spending and reduce the fiscal pressure on their
national health insurance systems, governments in France, Germany,
Sweden, and the United Kingdom have adopted a range of national
pharmaceutical policies. 

In each of these countries, however, this need for cost containment
has been tempered by attention to how price restraint might affect
pharmaceutical firms.  Country officials must weigh the concerns of a
strong, research-based pharmaceutical industry with the national
interest in pharmaceutical spending restraint.  In addition, national
authorities remain concerned that their cost containment policies
could diminish the development of new drug products.  In the United
States, this view has been expressed not only by the pharmaceutical
industry but also by some consumer activists and independent
analysts. 


   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:2

To reduce the growth of pharmaceutical costs, the four countries we
studied have employed a variety of national policies.  These policies
have largely--though not exclusively--targeted drug manufacturers. 
For example, France, and until 1993, Sweden, imposed
product-by-product price controls.\3 Germany and Sweden have limited
manufacturers' prices more indirectly, by imposing limits on insurer
reimbursement levels, and the United Kingdom has imposed profit
controls.\4 However, in recent years, these countries have extended
pharmaceutical cost containment policies to other players in the
market, like consumers and physicians.  For example, governments in
Germany and the United Kingdom have introduced incentives for
physicians to prescribe more cost- effective medications. 

As a group, these policies appear to have been effective at
restraining drug prices, but they have been unable to prevent
continued increases in drug spending.  Despite modest increases in
drug prices compared to the United States, between 1985 and 1991
these countries did not achieve the degree of pharmaceutical spending
restraint sought by country officials.  Instead, these countries
experienced increases in pharmaceutical spending comparable to that
in the United States.  In these European countries, higher
pharmaceutical spending has been driven largely by two
factors--higher pharmaceutical consumption and the use of newer, more
expensive drugs.  Government policies have not controlled these
forces entirely, although they have likely kept drug spending from
rising even more rapidly. 

Pressures to reduce this growth in prescription drug expenditures
have spurred efforts to make patients and physicians more aware of
drug prices and more financially responsible for drug spending.  For
example, in the last 5 years, consumers in all four countries have
been asked to pay a greater share of prescription costs.  In Germany
and the United Kingdom, physicians have been given drug spending
budgets or targets.\5 In addition, France, Germany, and the United
Kingdom have stiffened regulation of manufacturers by implementing
across-the-board price cuts.  While it is generally too early to
evaluate the success of these policies, country officials expect that
they will help restrain spending by reducing consumption and
over-prescribing. 

In pursuit of pharmaceutical cost containment, each
country--regardless of its specific policies--has encountered a
tension between low drug prices and pharmaceutical research. 
Although the presence of pharmaceutical price regulation does not
preclude the existence of an innovative industry, GAO's analysis
supports the conclusion that higher drug prices strengthen the
incentives for pharmaceutical R&D.  However, the significance of this
effect for public policy was difficult to evaluate, for two reasons. 
First, estimates of the size of the price-R&D relationship are
imprecise.  Moreover, the impact of declines in R&D spending for the
production of new drugs, especially for the more significant
innovations, is uncertain. 

Although government regulation has restrained drug prices in the four
countries we examined, the implications--and the desirability--of
similar intervention in the U.S.  pharmaceutical market are unclear. 
More specifically, determining the potential impact of a change in
U.S.  policy is complicated by existing institutional differences
between the U.S.  and other countries.  In addition, the U.S. 
pharmaceutical market is appreciably larger than the market in any
one of the other four countries.  In any event, any gains from
regulation of drug prices or spending must be weighed against the
consequences of such regulations for pharmaceutical research and
development. 


--------------------
\3 In January 1993, Sweden changed its strategy from direct price
controls to reimbursement controls. 

\4 Other spending control policies used in these countries include
consumer cost sharing and limits on which drugs are eligible for
reimbursement. 

\5 In January 1994, pharmaceutical industry representatives and
government officials in France adopted an informal agreement that,
among other things, would allow drug manufacturers greater
flexibility in pricing within a target growth rate for pharmaceutical
expenditures. 


   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:3


      COUNTRIES CONTROL PAYMENTS
      MADE TO MANUFACTURERS,
      WHOLESALERS, AND PHARMACISTS
-------------------------------------------------------- Chapter 0:3.1

In all countries studied, the principal policy to control spending
focuses on the price that manufacturers can charge.  In France (and
until 1993, in Sweden), the government sets prices paid to
manufacturers.  In Germany and in Sweden, manufacturers are largely
free to set prices, but for many drugs there are limits on the amount
insurers can pay.  In the United Kingdom, the government limits the
profits that manufacturers can earn from sales to the national health
care system; manufacturers largely can set introductory prices within
that constraint, but generally cannot increase drug prices.  In
addition to these policies, all four countries limit payments to drug
wholesalers and pharmacists by setting wholesale and retail
margins.\6

Each country has imposed additional controls on consumers and
physicians.  All four countries have, to varying degrees, increased
the consumers' share of drug costs.  France, Germany, and the United
Kingdom have also established drug lists that specify which drugs
will not be reimbursed by the national insurance system.  In
addition, Germany has imposed drug budgets that make physicians
financially responsible for over-prescribing.  The United Kingdom has
also placed more responsibility on physicians by giving each
physician a drug spending target, and by providing physicians
information on drug costs, efficacy, and prescribing patterns. 


--------------------
\6 The exception to this is in Sweden, where wholesaler fees are not
subject to government regulation, but are negotiated between
wholesalers and manufacturers. 


      POLICIES HAVE LIMITED DRUG
      PRICES, BUT DRUG SPENDING
      HAS CONTINUED TO INCREASE
-------------------------------------------------------- Chapter 0:3.2

Generally, these policies seem to have been successful in achieving
each country's pricing goals.  Drug price increases between 1985 and
1991 were less than the overall inflation rate in all four countries
we reviewed; by contrast, in the United States, drug prices rose at
over twice the rate of inflation.\7 The lowest drug price increases
were in France and Sweden, which had the tightest form of drug price
controls.  But even in the United Kingdom, which has the least
restrictive form of pricing restraint, prices rose at only half the
comparable U.S.  rate.  (See fig.  1.)

   Figure 1:  Pharmaceutical
   Prices Have Risen Slower Than
   the Inflation Rate, 1985-91

   (See figure in printed
   edition.)

Note:  Swedish data are for 1990. 

Source:  Organization for Economic Cooperation and Development. 

However, while these countries' policies may have slowed the growth
in drug prices, they have not completely contained the rise in drug
spending.  Even in countries with low prices, spending continues to
rise because of increases in drug consumption, increases in the
volume of prescriptions, and the higher relative prices of new drugs. 
Despite lower increases in drug prices, total drug spending in two of
the four countries rose about as rapidly between 1985 and 1990 as did
drug spending in the United States.  (See fig.  2.)

   Figure 2:  Inflation-Adjusted
   Pharmaceutical Spending Growth,
   1985-90

   (See figure in printed
   edition.)

Note:  United Kingdom data are for 1985-89. 

Source:  Organization for Economic Cooperation and Development. 


--------------------
\7 For each country, the inflation rate was measured by the growth in
the gross domestic product (GDP) deflator. 


      COUNTRIES ARE ADOPTING NEW
      POLICIES TO FURTHER CONTROL
      DRUG SPENDING
-------------------------------------------------------- Chapter 0:3.3

The menu of spending controls these countries have applied to
manufacturers, wholesalers, and retailers has not achieved the degree
of spending restraint sought by health financing officials.  As a
result, the governments in these countries are supplementing their
existing policies in order to better control utilization and the mix
between high- and low-priced drugs.  These additional new policies
are shifting the financial burden of drug spending from the
government and insurance systems to consumers, physicians, and
manufacturers. 

For example, since 1993, all four countries have increased the
patients' share of drug costs, and France, Germany, and the United
Kingdom are limiting the types of drugs that will be reimbursed by
the insurance system.  France, Germany, and the United Kingdom have
also imposed global cost reductions on pharmaceuticals.  Germany has
instituted a global budget for pharmaceutical spending, with the cost
of budget overruns to be borne by physicians.\8 France has also
considered a global budget on pharmaceutical products which would
give manufacturers more flexibility in setting drug prices but would
make them accountable for drug budget overruns.  In Sweden, the
government implemented a system that will lower the amount that the
insurance system will pay for many drugs. 


--------------------
\8 Overruns during 1993 would have also been borne by drug
manufacturers; however, the budget was not exceeded. 


      REDUCTIONS IN DRUG PRICES
      LEAD TO LOWER R&D
      EXPENDITURES
-------------------------------------------------------- Chapter 0:3.4

Transcending the specifics of each country's pharmaceutical policies
is a tension between low drug prices and pharmaceutical research. 
GAO's analysis indicates that higher drug prices contribute to the
development of new drugs by encouraging firms to devote more
resources to R&D.  However, the effect of prices on R&D is subject to
several significant qualifications.  First, the size of the effect is
difficult to measure precisely.  Second, the impact of an R&D
spending decline on the production of new drugs is uncertain--both
for breakthrough drugs and for more modest therapeutic improvements. 
Third, drug prices are only one of many factors that influence
pharmaceutical R&D.  Therefore, pharmaceutical spending control
policies can coexist with a strong research-based industry, even
though by themselves such policies would decrease R&D spending. 


   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:4

GAO is not making recommendations in this report. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:5

GAO obtained comments on this report from academic experts in the
economics of the pharmaceutical industry and from selected officials
in each country studied.  Their suggested revisions were
incorporated, as appropriate, into this report. 


INTRODUCTION
============================================================ Chapter 1

Research efforts by the pharmaceutical industry worldwide, and in
particular by companies based in the United States, have resulted in
the development of many new drugs to treat diseases and prolong or
save lives--benefits often viewed as priceless.  Nonetheless,
prescription drugs do have a price.  Throughout the industrialized
world, but especially in the United States, that price is considered
by many citizens to be too high. 

The increasing burden of paying for prescription drugs has led some
members of Congress to propose federal regulations limiting
prescription drug prices.  However, critics of such regulations,
within and outside the industry, have asserted, among other things,
that regulations that reduce drug prices would cripple U.S. 
pharmaceutical companies' ability to develop life-saving and
life-improving drugs. 

Because the United States has not regulated drug prices in the past,
our country's experience does not provide the evidence necessary to
resolve this debate.  European countries, however, have employed
policies for several decades to control pharmaceutical prices and,
indirectly, expenditures.  The nature of the choices facing the
United States can be illuminated by studying the European experience
with these policies. 

This report undertakes such a study, directed at analyzing both the
ability of these policies to control costs and the potential tension
between pharmaceutical innovation and cost containment.  The report
focuses on the pharmaceutical prices and spending control policies
that have been adopted by four of these countries:  France, Germany,
Sweden, and the United Kingdom. 


   RISING DRUG PRICES CREATE
   FINANCIAL BURDEN FOR MANY
   CONSUMERS
---------------------------------------------------------- Chapter 1:1

Continuing increases in prescription drug spending have placed
increasing financial burdens on those Americans who depend on
prescription drugs to maintain good health.  Total outpatient
expenditures on prescription drugs in the United States nearly
doubled between 1980 and 1991 (from $15.8 billion to $29.2 billion),
even after adjusting for inflation.\1 Much of this increase was
driven by increases in prescription drug prices, which rose by more
than twice the rate of inflation between 1980 and 1991.\2

Health care consumers, in general, are particularly sensitive to
these increases because of the high proportion of drug expenditures
that are not covered by health insurance.  While outpatient
prescription drugs are a relatively small amount of total health care
costs--less than 5 percent in 1991--over half of this amount is paid
out of pocket (compared to 18.1 percent of spending for physician
services and 3.4 percent for hospital care).  The greatest burden of
these out-of-pocket costs is likely to fall on the elderly, who as a
group both use more drugs and are less likely to have insurance
coverage for those drugs, because the federal Medicare program does
not offer outpatient prescription drug coverage. 


--------------------
\1 Some portion of this increase may be attributable to a general
movement of treatment from inpatient to outpatient settings over this
period. 

\2 Price indexes provide some indication of the rate of prescription
drug price increases as compared with price inflation in the general
economy.  But some research indicates that prescription drug indexes
may over-sample medium-aged drugs that undergo above-average price
increases, and under-sample younger products that experience
less-than-average price increases, thereby overstating annual average
drug price inflation.  (See Ernst R.  Berndt and others, "Auditing
the Producer Price Index:  Micro Evidence From Prescription
Pharmaceutical Preparations," Working Paper No.  4009, National
Bureau of Economic Research (Washington, D.C.:  Mar.  1992). 
Alternatively, indexes may understate annual changes in average drug
prices because they generally do not measure the impact of new drugs,
many of which enter the market at relatively high prices. 


   PHARMACEUTICAL INDUSTRY FEARS
   PRICE REGULATION WOULD HINDER
   DRUG DEVELOPMENT
---------------------------------------------------------- Chapter 1:2

Recent developments by the pharmaceutical industry have led to
important advances in medical treatment.  Drugs that were not
available prior to the 1980s are now commonly used to treat ulcers,
cardiac disease, high blood pressure, Acquired Immune Deficiency
Syndrome (AIDS), and many other ailments.  Ongoing research,
including the development of biotech drugs, may offer promising
improvements in the types of medicines available both to prolong life
and to improve the quality of life for people suffering from chronic
illnesses. 

Many such new drugs have been developed by pharmaceutical firms based
in the United States.  Among the world's top 15 companies in the
innovative drug industry in 1991, 8 were U.S.-based; these companies
had combined 1991 revenues of $36 billion.  U.S.-based pharmaceutical
firms developed over 40 percent of the new major global drugs
discovered between 1970 and May 1992.\3

According to pharmaceutical industry representatives, as well as some
independent observers, the threat of drug price regulation in the
United States could threaten the continuation of this record of
innovation.  They contend that drug price regulation would severely
decrease the rate of new drug development.  They maintain that high
profits are required to support the high costs associated with new
drug development, estimated to be between $140 million and $194
million (in 1990 dollars) for each new chemical entity.\4


--------------------
\3 Heinz Redwood, Price Regulation and Pharmaceutical Research:  The
Limits of Co-Existence (Suffolk, England:  Oldwicks Press Limited,
1993).  Redwood defines major global drugs as those drugs that have
been marketed or reached the post-clinical stage in at least six of
the world's seven leading pharmaceutical markets--the United States,
Japan, Germany, France, Italy, the United Kingdom, and Spain. 

\4 This figure is net of tax preferences given to pharmaceutical R&D. 
See U.S.  Congress, Office of Technology Assessment, Pharmaceutical
R&D:  Costs, Risks and Rewards, OTA-H-522 (Feb.  1993), pp.  67-69. 


   DRUG PRICES, RESEARCH AND
   DEVELOPMENT, AND AFFORDABILITY
   IN OTHER COUNTRIES
---------------------------------------------------------- Chapter 1:3

While the United States is a leader in new drug development, it is
also a leader in drug prices.  As several recent studies show,
prescription drug prices in other countries are generally lower than
in the United States.\5 Some of these countries have relatively
little drug research and development, but others have relatively
strong innovative drug industries.  For example, France, Germany,
Sweden, and the United Kingdom are home to firms that developed over
25 percent of new drug entities between 1970 and May 1992.\6

Affordability of drugs to individual consumers is not as much of a
problem in these other industrialized countries as it is in the
United States.  In this regard, many of these countries have
universal health insurance systems that provide pharmaceutical drug
coverage at little or no out-of-pocket cost to consumers.\7

Universal drug coverage, however, has shifted the burden of paying
for drugs from the individual to the insurance system, thereby
creating an incentive for the government to restrain spending growth
and to maintain the fiscal stability of the health insurance system. 
In addition, the relatively high level of drug spending in several
European countries has increased the importance to government
officials of restraining drug spending growth.  For example, while
pharmaceutical spending in 1990 composed about 8 percent of total
health spending in the United States (as well as in Sweden), it
accounted for almost 11 percent of health care costs in the United
Kingdom, about 17 percent in France, and over 21 percent in Germany. 

In response to the chronic pressure of rising health costs in
general, and drug spending in particular, on their health insurance
systems, these four countries (among others) have employed a variety
of policies designed to restrain the growth in drug prices and
spending.  In implementing these policies, each country confronts two
conflicting goals:  the reduction of the costs of pharmaceuticals to
the national health insurance system; and the maintenance of
incentives to encourage pharmaceutical manufacturers to continue
developing new drug products and attract industrial investment from
the international pharmaceutical industry. 


--------------------
\5 See, for example, Prescription Drugs:  Companies Typically Charge
More in the United States Than in Canada (GAO/HRD-92-110, Sept.  30,
1992); Prescription Drugs:  Companies Typically Charge More in the
United States Than in the United Kingdom (GAO/HEHS-94-29, Jan.  12,
1994); Association Belge des Consommateurs, Statement Prepared For
the United States Senate Special Committee on Aging, (Nov.  16,
1989); and W.  Duncan Reekie, "Drug Prices in the UK, USA, Europe,
and Australia," Australian Economic Papers (June 1984), pp.  71-78. 

\6 See Heinz Redwood, Price Regulation and Pharmaceutical Research: 
The Limits of Co-Existence (Suffolk, England:  Oldwicks Press
Limited, 1993), p.  22. 

\7 There are also fewer networks for buying prescription drugs in
other countries than in the United States.  For example, in the
countries we studied, consumers generally purchase their
pharmaceuticals from retail pharmacists.  By contrast, while most
Americans buy their pharmaceuticals at retail pharmacies, many
purchase through mail order houses and managed care organizations. 
See Stephen W.  Schondelmeyer and Joseph Thomas III, "Trends in
Retail Prescription Expenditures," Health Affairs 9:3 (Fall 1990),
pp.  131-145. 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:4

The Chairman of the Senate Special Committee on Aging asked our
office to report on how other countries regulate prescription drug
prices, how those policies affect drug prices, and how they affect
pharmaceutical R&D.  Our first report on this subject examined
Canada's approach to drug price regulation.\8 In this second report,
we focus on countries that, unlike Canada but like the United States,
have strong innovative drug industries. 

Specifically, the objectives of this study were to

  describe the methods used in France, Germany, Sweden, and the
     United Kingdom to control prices of outpatient prescription
     drugs and to limit pharmaceutical spending;

  review the effects of these measures on pharmaceutical prices and
     spending; and

  analyze the effect of pharmaceutical prices and price regulations
     on R&D. 

We reviewed the pharmaceutical price and spending control measures
used by France, Germany, Sweden, and the United Kingdom (see apps. 
I-IV).  We selected these industrialized democracies for their
variety in the policies used to influence prescription drug prices
and because they are home to companies engaging in pharmaceutical
R&D.  Moreover, the governments in all study countries, like that of
the United States, seek to contain rising health care costs and to
reform elements of their health care systems. 

In the course of our review, we collected and reviewed technical
literature and government documents that describe pharmaceutical
price control measures used in these countries and analyzed the
effects of these measures.  We interviewed officials in each study
country representing the national government and the pharmaceutical
industry.  We also interviewed other officials, such as
representatives of consumer groups, academia, and the health
insurance systems about these issues.  In addition, we developed a
multivariate statistical model to estimate the effects of various
factors on pharmaceutical R&D in European countries and in the United
States (see app.  V). 

Our review was conducted from March 1992 through January 1994.\9
Because this report describes prescription drug spending controls in
foreign countries, we did not obtain comments from the Department of
Health and Human Services on this report.  However, pertinent
portions of this report were reviewed by academic experts in the
economics of the pharmaceutical industry and by selected officials in
each country.  Based on the comments received, we made technical
revisions to this report as appropriate. 


--------------------
\8 Prescription Drug Prices:  Analysis of Canada's Patented Medicine
Prices Review Board (GAO/HRD-93-51, Feb.  17, 1993). 

\9 Although this report contains information through January 1994 on
the price and spending control measures employed by the study
countries, changes in these measures are frequent and ongoing. 


DRUG COST CONTROLS HAVE MIXED
SUCCESS AT RESTRAINING
PHARMACEUTICAL PRICES AND SPENDING
============================================================ Chapter 2

As part of their national health insurance systems, France, Germany,
Sweden, and the United Kingdom each covers prescription drugs, and
all face a continuing challenge to restrain national spending on
pharmaceuticals.  In this persistent struggle, each country has
developed spending control strategies consistent with two premises: 
first, that drug manufacturers can, if left unchecked by regulation,
charge prices substantially above their costs, because patents and
marketing efforts protect them from competitors; and second, that
insurance coverage and physician responsibility for prescribing
discourages comparison shopping by consumers, who lack incentives to
seek out the most cost-effective drugs and have limited knowledge
about alternative medications.  In designing approaches to dampen
pharmaceutical spending, governments have tended to rely more on
regulations and sanctions than on policies to strengthen competition
and incentives. 

Currently, the scope of pharmaceutical cost containment strategies is
diverse, targeting not only price but other determinants of drug
spending.  At least until the late 1980s, however, efforts to
restrain drug prices had focused largely on controls at the point of
sale--that is, at the prices charged, for example, by drug
manufacturers to drug wholesalers, or by pharmacists to consumers. 
These traditional policies seem to have restrained prices, but
increases in drug utilization and higher prices for new drugs have
pushed up drug spending.  Faced with this further stress on their
national health care budgets, government officials in the countries
we studied have concluded that, as a tool for restraining
pharmaceutical spending, controls on prices alone are not sufficient. 

As a result, each country has introduced or is developing a
distinctive set of policies.  They are designed to reduce the growth
in prescriptions written, encourage the use of drugs that are more
cost-effective, and shift some of the burden of higher drug spending
from the national health insurance system to consumers, physicians,
and drug manufacturers. 


   FISCAL PRESSURES ON HEALTH
   INSURANCE SYSTEMS UNDERLIE
   GOVERNMENT EFFORTS TO CONTROL
   DRUG SPENDING
---------------------------------------------------------- Chapter 2:1

Each of the four countries we reviewed has a national health
insurance system that offers universal access to health care,
including prescription drug products.\1 These systems pay for most or
all of the costs of prescription drugs.  Consequently, the insurance
systems bear the financial burden of prescriptions most heavily and
directly, while consumers pay relatively little. 

In these countries, pharmaceutical outlays are a significant part of
health care spending.  In the period 1989 through 1990, the last
years for which comparative data are available, pharmaceutical
expenditures ranged from 8.2 percent of total health spending in
Sweden to over 20 percent in Germany (see fig.  2.1).  Given the
fiscal weight of the pharmaceutical sector, each of these countries
has looked to this sector for a significant contribution to the
national effort at slowing the growth of overall health care
spending. 

   Figure 2.1:  Pharmaceutical
   Expenditures as a Share of
   Total Health Spending, 1990

   (See figure in printed
   edition.)

Note:  United Kingdom data are for 1989. 

Source:  Organization for Economic Cooperation and Development. 


--------------------
\1 The particular type of health insurance system varies by country. 
Sweden and the United Kingdom have single payer systems in which the
government provides health insurance for the entire population. 
Germany and France have multiple payer insurance systems, where
workplace-based insurers provide coverage for most or all of the
population.  For a further description of these systems, see apps. 
I-IV; also, see Health Care Spending Control:  The Experience of
France, Germany, and Japan (GAO/HRD-92-9 Nov.  15, 1991) and Marilynn
M.  Rosenthal and Marcel Frenkel, eds., Health Care Systems and Their
Patients:  An International Perspective (Boulder, CO:  Westview
Press, 1992). 


   EACH COUNTRY LIMITS DRUG
   SELLERS' ABILITY TO SET PRICES
   FREELY
---------------------------------------------------------- Chapter 2:2

Each country we reviewed has sought, as part of its efforts to manage
its health care budget, to contain pharmaceutical spending with
several different types of policies.  These have included consumer
cost sharing and restrictions on which drugs will be reimbursed, but
the most prominent policies have been ones that limit drug
manufacturers' ability to set their prices freely.  That is, these
countries have, until recent years, centered their pharmaceutical
cost containment on regulations that limit drug prices directly or,
by limiting insurance reimbursement, do so indirectly.  These
regulations are found at various points in the distribution chain for
pharmaceuticals:  the sale from manufacturer to wholesaler, from
wholesaler to pharmacy, and from pharmacy to consumer. 

Regulations targeted at drug manufacturers' prices in the four
countries we studied embody one of three mechanisms: 

  product-by-product price controls,

  limits on insurers' reimbursement levels, or

  profit controls. 

Prices are also regulated at subsequent links in the distribution
chain.  The fees charged by wholesalers and pharmacists typically are
not allowed to exceed a set ceiling.\2 These fees can be calculated
as a fixed amount per prescription or as a percentage of price.  (See
table 2.1.)



                          Table 2.1
           
                Wholesaler and Pharmacist Fees

Country       Payment Policy
------------  ----------------------------------------------
France        Wholesale margin is set by law at 10.74
              percent of the manufacturer price (exclusive
              of value-added tax).

              Pharmacist margin is calculated according to a
              sliding scale that decreases in proportion to
              the drug's price.

Germany       Allowable wholesale markups range between 12.0
              to 21.0 percent of the manufacturer price,
              depending on the price of the product.

              Allowable pharmacy markups range from 30.0 to
              68.0 percent of the wholesale price (exclusive
              of the value-added tax), depending on the
              price of the product.

Sweden        Wholesaler markups are negotiated between
              wholesalers and manufacturers, and average 4.2
              percent of the manufacturer price (this is
              equivalent to 2.8 percent of the retail
              price).

              Pharmacies, which are run by an agency that is
              two-thirds owned by the government, add 41
              percent to the wholesale price (this margin is
              equivalent to 29 percent of the retail price).

United        Wholesale and pharmacist margin together
Kingdom       cannot exceed 12.5 percent of the retail list
              price.

              Pharmacists also receive a dispensing fee of
              ï¿½1.512 per prescription for the first 1,500
              prescriptions per month, and ï¿½0.715 for each
              prescription thereafter. In May 1992, this fee
              averaged ï¿½1.08 per prescription.

              Pharmacist fees are reduced by a rate intended
              to capture discounts they receive from
              wholesalers.
------------------------------------------------------------

--------------------
\2 The exception to this is in Sweden, where wholesaler fees are not
subject to government regulation, but are negotiated between
wholesalers and manufacturers. 


      REGULATIONS ON MANUFACTURER
      PRICES DIFFER IN DEGREE OF
      PRICING FREEDOM
-------------------------------------------------------- Chapter 2:2.1

Each country has regulations that are designed to limit--either
directly or indirectly--the price that drug manufacturers charge to
wholesalers (or to retailers that buy directly from the
manufacturer).  As described below, these policies differ in the
extent that manufacturers are free to set launch prices for new
products as well as to increase prices on existing products.  (Apps. 
I-IV describe these policies in greater detail.)

Product-by-product price controls are the most direct form of price
regulation, in that they largely bar manufacturers from selling their
drug products at prices above those approved by the government (or
other paying authority).  In the two countries we studied where
product-by-product price controls have been used for outpatient
prescription drugs--France and, until 1993, Sweden--both new product
prices and price increases were regulated by the government.  New
product prices emerge from negotiations between the government and
each drug manufacturer.  The criteria for setting these prices
include the therapeutic value of the drug and the price of comparable
treatments.\3 Price increases in both countries are allowed only with
prior government approval.\4

Limits on insurer reimbursement prices set an upper limit--or
reference price--on the amount the insurer can pay for groups of
identical or equivalent drugs.  Drug manufacturers are free to set
any launch price or price increase that they choose, but consumers
must pay the difference between that price and the reference price. 
Manufacturers' ability to charge a price that is higher than the
reference price is limited by consumers' willingness to incur
out-of-pocket costs for pharmaceuticals. 

Germany and Sweden illustrate different ways that reimbursement
prices can be calculated.  In Germany, a drug's reference price is
computed essentially as the average of the prices of that drug and
similar products.\5 In Sweden, the reference price for a drug is set
at 10 percent above the price of the least expensive generic
equivalent.  In Germany, drugs are not covered under the reference
price system (RPS) if they do not have a sufficient number of
comparable products, while in Sweden, only one generic equivalent is
needed to set a reference price.  In Germany, the statutory health
insurers (known as sickness funds) pay the price that manufacturers
set for drugs without a reference price (less the required patient
copayment of Deutsche Mark (DM) 3 to DM 7).\6 By contrast, in Sweden
the government negotiates with manufacturers the prices that can be
charged for these drugs.\7

Profit controls, used in the United Kingdom, are a more indirect form
of drug spending control.  A manufacturer that introduces a drug
product into the U.K.  market may freely set its launch price at any
level, as long as company profits do not exceed a negotiated target. 
More precisely, the National Health Service (NHS), which in effect is
the national health insurer, negotiates a profit ceiling with most
drug manufacturers.\8 Through this process, the government relates
each manufacturer's profits and hence, indirectly, their prices, to
the level of investment in pharmaceutical production and R&D in the
country for the purpose of supplying drugs to NHS.\9 However, even
under this profit control scheme, drug manufacturers are still
subject to drug price regulations.  While manufacturers freely set
prices when introducing new drugs--so long as profits do not exceed
the target level--they cannot increase drug prices without prior
government approval. 


--------------------
\3 In France, as in Sweden, the allowable price may also be
influenced by the contribution of the drug's sales to the national
economy.  In addition, Sweden based its allowable price on the price
charged for the drug in other countries, and in particular, on the
price in the manufacturer's home country. 

\4 In France, the government prohibits price increases for drugs that
have been on the market less than 2-1/2 years.  After that time,
prices can only be increased through a global pricing directive,
which raises or lowers the prices of all drugs on the market by a set
percentage.  In Sweden, the government tries to keep drug price
increases within the rate of inflation. 

\5 Three different categories are used to define sets of similar
drugs:  (1) drugs with the same active ingredients (for example,
brand name drugs and their generic equivalents); (2) drugs with
therapeutically comparable active ingredients (for example,
beta-blockers or H-2 antagonists); and (3) drugs with therapeutically
comparable effects (for example, different aspirin combinations). 
The reference price for a particular drug is adjusted for variations
from the average product's strength and package size. 

\6 In Germany, many single source products that lack comparable
products cannot be assigned reference prices.  Furthermore, other
products do not yet have reference prices because of the technical
difficulties in ascertaining which products have comparable
therapeutic ingredients or actions.  As of July 1993, about half of
pharmaceutical products in Germany had reference prices.  In 1993,
the German government simplified the way that drugs are put into
comparable groups.  The government hopes that this simplification
will allow for the eventual inclusion of 70 percent of drugs into the
reference price system. 

\7 These negotiations are performed for patented drugs that do not
have generic substitutes and for over-the-counter drugs that the
manufacturer wants included under the reimbursement system.  Factors
going into the negotiations include the basis of the drug's
therapeutic value, the price of comparable products in other
countries, the price of the drug in other countries, and the extent
to which the drug's usage substituted for more expensive treatments. 
No negotiations take place for nonreimbursable drugs (for example,
drugs sold in hospitals); instead, manufacturers are able to price
these drugs freely. 

\8 The United Kingdom's profit control scheme applies to all firms
with sales to NHS of over ï¿½0.5 million (or about $740,000) per year. 

\9 Under the United Kingdom's profit control scheme, which excludes
generic drugs, manufacturers' profits are regulated in two ways,
depending on their capital investment in the country.  Manufacturers
with sizeable capital investment are permitted to price drugs in line
with target profit levels, based on their return on capital--current
profit levels on sales to the NHS are set at 17 to 21 percent of the
capital invested in the country, and devoted to supplying brand-name
(that is, nongeneric) prescription drugs to NHS.  Other manufacturers
selling in the U.K.'s drug market also have target profit levels, but
these are based on their return on sales.  Manufacturers can justify
keeping additional profits (up to 25 percent over their target level)
if the additional profits are attributable to new products or to
increased operating efficiency.  (See app.  IV.)


   REGULATIONS ARE MORE EFFECTIVE
   AT LIMITING DRUG PRICES THAN AT
   RESTRAINING DRUG SPENDING
---------------------------------------------------------- Chapter 2:3

The drug spending controls applied in these four countries have had
mixed success at restraining the level of pharmaceutical
expenditures.  On the one hand, drug prices in these countries have
grown relatively slowly under the price and profit controls--less
than the rate of general inflation.\10 But while price restraint
probably has kept total drug spending lower than it would have been
otherwise, total drug spending--which is affected by the quantity of
drugs sold as well as their prices--has continued to rise faster than
the countries' governments are willing to accept. 

Between 1985 and 1991, the countries with the most direct types of
price controls--France and Sweden--had the lowest average rates of
increase in drug prices (see fig.  2.2).\11 \12 In the United
Kingdom, which has the most indirect type of price control, nominal
drug price increases were the highest of the countries we reviewed;
nonetheless, even U.K.  drug prices rose relatively slowly--at about
half the general rate of inflation.  By contrast, during the same
period (1985-91), pharmaceutical prices in the United States
increased at an average annual rate that was over twice the general
inflation rate. 

   Figure 2.2:  Changes in Nominal
   and Inflation-Adjusted
   Pharmaceutical Prices, 1985-91

   (See figure in printed
   edition.)

Notes:  Inflation is measured by the growth in the GDP deflator. 
Swedish data are for 1985-90. 

Source:  GAO calculations, based on Organization for Economic
Cooperation and Development data. 

While the price restraint may have helped achieve some moderation in
the growth of drug spending, the countries we examined had limited
success in restraining the growth in total pharmaceutical
expenditures during the same time period (see fig.  2.3).  The
relative increases in pharmaceutical spending were greater for
countries with direct price controls than for those with more
indirect approaches.  In France and Sweden, the countries that
employed direct price controls, the average annual growth in
pharmaceutical spending between 1985 and 1990 was comparable to that
in the United States.  In Germany and in the United Kingdom,
pharmaceutical spending grew at a slightly slower rate than in the
United States.  However, pharmaceutical spending in Germany and the
United Kingdom grew more rapidly than overall inflation.\13

   Figure 2.3:  Pharmaceutical
   Spending Growth, 1985-90

   (See figure in printed
   edition.)

Note:  United Kingdom data are for 1985-89. 

Source:  Organization for Economic Cooperation and Development. 


--------------------
\10 The general inflation rate is measured by the growth in the price
deflator for Gross Domestic Product in each country. 

\11 Swedish data are for the period 1985-90. 

\12 Drug price inflation can occur even under regulatory regimes,
such as those in France and the United Kingdom, which largely
restrict drug price increases.  This is because the pharmaceutical
price index, on which drug price inflation is based, is composed of a
market basket of drugs that changes over time.  As new drugs become
part of this market basket, the cost of this basket can increase if
the price of those new drugs exceeds the average cost of the other
drugs in the previous market basket. 

\13 Data on pharmaceutical spending in the United Kingdom are for the
period 1985 through 1989. 


      SPENDING GROWTH IS LARGELY
      ATTRIBUTABLE TO FACTORS
      BEYOND THE REACH OF DRUG
      PRICE AND PROFIT CONTROLS
-------------------------------------------------------- Chapter 2:3.1

The increase in pharmaceutical spending does not necessarily imply
that the controls were ineffective at restraining drug spending. 
Indeed, these policies may have kept drug expenditures from rising
higher than they would have otherwise.\14 However, the rise in drug
spending suggests that factors outside the purview of these
regulations outweighed any restraining impact that price and profit
controls may have had on drug expenditures. 


--------------------
\14 Analyses of the effects of Germany's reference price system
suggest that drug prices and spending were lower after the imposition
of reference pricing than they would have been otherwise.  We were
not able to identify any formal studies on how the policies used in
France, Sweden, or the United Kingdom affected drug spending, nor
were there sufficient data for doing a before-and-after analysis on
the policies' effects. 


         INCREASES IN DRUG
         UTILIZATION
------------------------------------------------------ Chapter 2:3.1.1

Increases in drug utilization likely provide one source of these
spending increases.  As figure 2.4 shows, drug utilization grew more
rapidly than drug prices in the four countries we reviewed,
suggesting that greater utilization accounted for a large amount of
the growth in drug spending.  By contrast, in the United States drug
utilization grew far less rapidly than drug prices, thereby
suggesting a greater role for drug price increases in explaining
spending growth.  (See fig.  2.4.)

   Figure 2.4:  Growth in
   Inflation-Adjusted
   Pharmaceutical Prices and
   Utilization, 1985-90

   (See figure in printed
   edition.)

Note:  United Kingdom data are for 1985-89. 

Source:  Organization for Economic Cooperation and Development. 

Increases in utilization can come from population growth and from
increases in the elderly's share of the population--both of which
occur independently of price and profit controls.  The increases in
the elderly can be of particular importance in explaining higher
spending levels, since elderly people are likely to have higher per
capita drug use than are the nonelderly.  Each of the countries we
reviewed has experienced increases in the elderly's share of the
population, especially in persons over the age of 75.  (See table
2.2.)



                          Table 2.2
           
              Growth in the Share of the Elderly
                     Population, 1985-91

                          (Percent)

                                              United
Share of elderly in                   German  Kingdo  United
total population      France  Sweden       y       m  States
--------------------  ------  ------  ------  ------  ------
Age 65 and over
------------------------------------------------------------
1985                   12.8%   17.4%   14.8%   15.1%   11.9%
1991                    14.1    17.7    15.4    15.8    12.7

Age 75 and over
------------------------------------------------------------
1985                     6.3     7.4     6.9     6.4     4.8
1991                     7.0     8.1     7.2     7.0     5.2
------------------------------------------------------------
Source:  Organization for Economic Cooperation and Development. 


         HIGHER PRICES FOR NEWER
         DRUGS
------------------------------------------------------ Chapter 2:3.1.2

Increases in drug spending may also be caused by the use of newer,
more expensive drugs.  Despite the control mechanisms in place in
these four countries, new drugs tend to have higher average prices
than the drugs they replace, increasing the pressure on drug budgets
even when consumption levels remain constant.  These new products,
which can range from innovative treatments to modest improvements
over existing products, can strain drug budgets when they replace
less expensive medications.\15 Higher new drug prices have been cited
as a particular problem in the United Kingdom, where companies are
free to set new drug prices so long as their profits remain within
the target range. 

The price and profit controls used in these countries generally do
not provide patients and physicians with an incentive to choose
products that are less expensive.  Of the systems that we reviewed,
only the reference price systems, used in Germany and Sweden, create
incentives for consumers to choose lower-priced products.  Under this
system, a single reimbursement rate applies to drugs that are
considered therapeutically equivalent or comparable to one another;
if the price exceeds this level, then the consumer pays the
remainder.  By contrast, neither direct price controls nor profit
controls create incentives for consumers or physicians to choose a
less expensive medication. 


--------------------
\15 Even when use of these medications replaces more expensive
nondrug treatments, they can increase the pharmaceutical budget. 
Consider the hypothetical example of a new medication that costs
$1,000, but reduces the need for surgery that would cost $25,000. 
Each time that the medication is prescribed in lieu of surgery,
hospital costs would be reduced by $25,000, but prescription drug
spending--
accounted for in another budget--would be increased by $1,000. 


   INCREASED SPENDING HAS SPURRED
   ADOPTION OF POLICIES THAT SHIFT
   COSTS AND ENCOURAGE
   COST-EFFECTIVE PRESCRIBING
---------------------------------------------------------- Chapter 2:4

The health financing systems in the countries we reviewed have been
strained by the pattern of increases in pharmaceutical spending that
approach or outstrip the growth of GDP.  These strains have resulted
in the adoption of major changes in the drug reimbursement policy in
Germany and Sweden, proposals for major changes in such policy in
France, and modifications in both Germany and the United Kingdom that
are intended to make physicians more aware of drug costs.  These new
policies--sometimes working within the context of existing price and
profit controls, and sometimes not--are designed to meet two
objectives: 

  first, to shift the burden of increased pharmaceutical spending
     from government to consumers, physicians, and drug
     manufacturers; and

  second, to stimulate price competition in the pharmaceutical sector
     by encouraging consumers and physicians to choose more
     cost-effective medications.\16


--------------------
\16 Sweden's recent payment reform was imposed, to some extent, for
an additional reason--to respond to a European Community directive
that requires member countries to publicly disclose the rules
governing pricing of prescription drugs.  The directive does not
interfere with the right of countries to control prices or
reimbursement by any method they choose, provided the method used is
"transparent" and does not discriminate between foreign and domestic
drug manufacturers.  Sweden is not a European Community member, but
has applied for membership. 


      INCREASES IN CONSUMER COST
      SHARING
-------------------------------------------------------- Chapter 2:4.1

One approach used to reduce drug spending is to increase consumers'
financial responsibility for prescription drugs.  From 1989 through
1993, all four countries have increased the patient's share of drug
costs.  (See table 2.3.)



                                    Table 2.3
                     
                        Patient's Share of Drug Costs Has
                      Increased in France, Germany, Sweden,
                         and the United Kingdom, 1989-93

Country      1989                   1991                   1993
-----------  ---------------------  ---------------------  ---------------------
France       Copayment of 0, 30,    Copayment of 0, 30,    Copayment of 0, 35,
             60, or 100 percent of  60, or 100 percent of  65, or 100 percent of
             drug cost, depending   drug cost, depending   drug cost, depending
             on the particular      on the particular      on the particular
             drug.                  drug.                  drug (effective
                                                           summer 1993).

Germany      Copayment of DM 3 per  Drugs under the        Copayment of DM 3-DM
             prescription.          reference price        7, depending on the
                                    system: Patients pay   price of the drug.\a
             Starting June 1,       the amount by which
             drugs under            the retail price       In addition, the
             the reference price    exceeds the reference  consumer pays any
             system: Patients pay   price.                 amount by which the
             the amount by which                           retail price exceeds
             retail price exceeds   Drugs not under the    the reference price.
             the reference price.   reference price
                                    system: DM 3 per
             Drugs not under the    prescription
             reference price
             system: DM 3 per
             prescription.

Sweden       Flat copayment of SEK  Flat copayment of SEK  Copayment of SEK 120
             90 for up to 10 drugs  90 for up to 10 drugs  for first
             written on same        written on same        prescription and SEK
             prescription form.     prescription form,     10 for additional
                                    for a maximum          prescriptions
                                    prescribing period of  obtained from the
                                    90 days.               pharmacy at the same
                                                           time, for a maximum
                                                           prescribing period of
                                                           90 days.

                                                           In addition, the
                                                           consumer pays any
                                                           amount by which the
                                                           drug's price exceeds
                                                           the reference price.

United       Flat copayment of      Flat copayment of      Flat copayment of
Kingdom\b    ï¿½2.80 for drugs        ï¿½3.40 for drugs        ï¿½4.25 for drugs
             covered by NHS.\c      covered by NHS.\c      covered by NHS.\c
--------------------------------------------------------------------------------
\a As of January 1994, the copayment in Germany is based on the size
of the prescription rather than on the price of the drug. 

\b Table lists copayment levels as of April 1 of each year cited.  In
addition, patients in the United Kingdom receiving frequent
prescriptions may buy a season ticket covering the costs of all
prescriptions for either 4 months or 12 months.  In April 1989, the
4-month season ticket cost ï¿½14.50, and the 12-month season ticket
cost ï¿½40.  By April 1993, these costs increased to ï¿½22 for the
4-month ticket and ï¿½60 for the 12-month ticket. 

\c Because of exemptions to cost sharing, about 80 percent of drugs
dispensed in the United Kingdom have no consumer copayment. 

The higher copayments may have the dual purpose of (1) shifting some
of the financial burden of pharmaceuticals away from the national
health insurance system and toward consumers, and (2) raising
consumer cost-consciousness about their prescriptions, thereby
reducing alleged overutilization of drugs. 

Certain features of some copayment policies can be expected to limit
their effectiveness at restraining drug spending.  First, copayments
that cover only certain drugs or certain segments of the population
will reduce spending less than would more comprehensive cost sharing. 
For example, until 1993, there were no copayments for German
pharmaceuticals covered under the reference price system (so long as
the drug's price did not exceed the reference price).  Therefore,
consumers had no incentive to reduce consumption of those items.  In
the United Kingdom, copayment exemptions for the elderly, the poor,
children, and pregnant women (among others) eliminates all cost
sharing for about 80 percent of prescriptions written. 

Second, copayments that are the same amount for every prescription
cannot affect the choice between more and less expensive medications. 
If the consumer's copayment is identical for an expensive drug and
for a cheaper substitute, the consumer has no reason to choose the
less expensive medication. 

Third, the small size of the copayments may also limit their ability
to reduce the number of prescriptions filled.  However, raising the
copayment could present a financial barrier to poor households or to
people who need to use a high volume of pharmaceuticals. 


      ENCOURAGING PHYSICIANS TO
      PRESCRIBE LESS EXPENSIVE
      MEDICINES
-------------------------------------------------------- Chapter 2:4.2

To an increasing extent, pharmaceutical payment reforms in the
countries we reviewed--particularly in the United Kingdom and
Germany--are designed to encourage economical prescribing by
physicians and to emphasize the use of less expensive drugs.  These
policies recognize the vital role of the physician as the primary
decisionmaker regarding choice of pharmaceuticals and, to varying
degrees, tie financial incentives for physicians to the prescribing
choices that they make. 

The United Kingdom uses a two-pronged strategy for encouraging
physicians to be agents for lower pharmaceutical spending: 

First, the government provides information to individual physicians
about their prescribing habits (relative to those of their
colleagues).  Physicians receive a periodic report on the number and
cost of the drugs they prescribed, compared to norms for physicians
in their area.  The government also provides physicians with
information on the safety and cost-effectiveness of alternative drug
products.  This information is intended to allow the physicians to
make more responsible choices about prescribing. 

Second, physician spending targets are used to restrict
pharmaceutical sales.  Since 1991, physicians in the United Kingdom
have been subject to the Indicative Prescribing Scheme (IPS), which
sets financial targets for physician prescribing.  Under IPS, doctors
are given a financial benchmark, referred to as an indicative amount
of prescribing.  Physicians' indicative targets are based on several
factors, including historical expenditures, demographic composition
of their patients, and drug price inflation.  These targets are not
binding caps, although physicians who consistently prescribe more
than their targeted amounts can be targeted for advice and detailed
monitoring, and in a last resort, cases of gross over-prescribing can
be penalized.\17 \18

Germany also instituted pharmaceutical budgets on physicians, but
these controls--implemented in 1993 as part of a comprehensive health
financing reform--place more stringent financial controls on
physicians than do the United Kingdom's policies.  As of January
1993, Germany has had a global budget for pharmaceuticals, which, if
exceeded, will be offset by a reduction in the ambulatory care
physician budget.  In 1993, the total pharmaceutical budget for
office-based physicians was set at about DM 24 billion, or about $15
billion.  While 1993's spending did not exceed this level, any cost
overrun up to DM 280 million (about $175 million) would have been
offset by a reduction in the 1994 ambulatory care physician budget. 
(The cost overrun would also be borne by pharmaceutical manufacturers
if it reached DM 280 million, up to another DM 280 million.) For most
regions, the 1994 budget is set at the 1993 level, and all cost
overruns will be borne by reductions in the ambulatory care physician
budgets.\19

The global budgets in Germany appear to have had an impact in the
short time that they have been in effect.\20 Total prescription drug
costs for sickness funds declined by about 20 percent in the first
half of 1993, compared to the same period in 1992, and total 1993
drug spending was actually less than the budgeted amount and,
therefore, less than 1991's total.  In addition, in the first half of
1993, physician prescribing fell by about 17 percent below the 1992
level. 

Several reasons have been suggested for the drop in drug spending in
Germany.  First, physicians substituted cheaper generic drugs for
more expensive, brand-name drugs.  As a result, sales of the cheapest
generic drugs increased in some cases by as much as 250 percent. 
Second, many patients--especially those with long-term
illnesses--obtained their prescriptions in December 1992 (before the
law took effect) and thus did not need to acquire their drugs in the
first few months of 1993.  Third, physicians have been less willing
to prescribe drugs with doubtful efficacy (e.g., anti- varicosis
drugs) or drugs for conditions that can be treated in different ways
(e.g., drugs for diets).\21

Citizens and officials in both countries have been concerned about
whether the budgets are reducing access to pharmaceuticals.  In the
United Kingdom, some observers believe that the budgets are
constraining physicians' ability to prescribe the most effective
drugs and respond to special patient needs, such as those of the
elderly.  However, government officials believe that the physician
budgets could, instead, increase the quality and cost-effectiveness
of prescribing, and so improve patient care.  In Germany, some
officials have expressed concern that the older drugs that physicians
are prescribing in order to save costs may be less effective than
newer, more innovative products.  However, there is no firm evidence
either to support or contradict this contention. 


--------------------
\17 The provisions requiring physicians to justify this prescribing
behavior are separate from and predate IPS. 

\18 Some physicians in the United Kingdom--25 percent as of April
1993--are subject to an alternative budgeting scheme, known as the GP
fundholding scheme.  Under this scheme, which is voluntary,
physicians who are in relatively large group practices are given a
practice budget, which is intended to cover all prescribing costs for
patients as well as the cost of some hospital services, outpatient
services, administrative services, and visiting and district nurse
services. 

\19 Most regional physicians' associations chose to accept the 1994
budget set at the 1993 level rather than negotiate a budget based on
real 1993 expenditures. 

\20 No systematic evidence exists on the effects of IPS in the United
Kingdom. 

\21 There was a disproportionate decrease in the prescription of
drugs that are considered to be therapeutically controversial and
drugs that are considered to be therapeutically meaningful.  For
example, drugs in the former group include circulatory drugs and
vitamins (which declined 29.9 percent and 29.1 percent,
respectively).  Drugs in the latter group include antibiotics and
anti-diabetic drugs (which declined 5.2 percent and 0.7 percent,
respectively). 


      MORE STRINGENT CONTROLS
      BEING APPLIED TO DRUG
      MANUFACTURERS
-------------------------------------------------------- Chapter 2:4.3

While many of the recent policy changes in the countries reviewed
have applied to patient and physician practices, France, Germany, and
the United Kingdom--to differing degrees--have also made efforts at
reducing payments to manufacturers.  These efforts have taken three
forms:  first, across-the-board price cuts; second, limits on total
manufacturers' sales; and third, limits on the types of drugs
eligible for reimbursement. 


         ACROSS-THE-BOARD PRICE
         CUTS
------------------------------------------------------ Chapter 2:4.3.1

One method used to reduce pharmaceutical spending is across-the-board
cuts in payments to drug manufacturers.  France, Germany, and the
United Kingdom have used this measure in recent years.  France's most
recent price reductions occurred in 1991, when the government ordered
that pharmaceutical prices be cut by 2.5 percent.  Germany
implemented across-the-board price cuts in 1993, when the government
ordered a 5-percent reduction in the price of drugs not covered by
the reference price system, and a reduction in over-the-counter
(nonprescription) drug prices to 2 percent below the May 1992 price
level.  The government also mandated a price freeze on these drugs
that will be in effect through 1994.  The United Kingdom also
implemented global price cuts in 1993, ordering a 2.5-percent price
cut on all products, which is to be followed by a 3-year price
freeze. 


         BUDGETS
------------------------------------------------------ Chapter 2:4.3.2

Of the countries we reviewed, only Germany has imposed budgets that
apply to manufacturers.  As described in the previous section,
Germany's 1993 global budget sets total limits on annual
pharmaceutical spending.  While physicians were to bear part of the
budget overrun--the first $175 million in 1993--subsequent overruns
(up to $175 million) would have come from the pharmaceutical
manufacturers.  However, under the 1994 budget, manufacturers will
not have to bear the financial burdens of overruns if physicians
exceed the budget. 

France may adopt drug budgets for manufacturers.  In 1991, the French
government proposed a drug payment system in which manufacturers
would each have a budget for total drug sales to the social insurance
system.  Under this framework, manufacturers could have been able to
set prices freely, as long as their total revenues from sales to the
national health system did not exceed the budget.  This proposal was
never enacted, due to political opposition.  However, in January
1994, representatives of the pharmaceutical industry and French
government reached an informal agreement that, if implemented, would
include many aspects of this 1991 proposal. 


         LIMITING DRUGS ELIGIBLE
         FOR REIMBURSEMENT
------------------------------------------------------ Chapter 2:4.3.3

Governments can limit the drugs eligible for reimbursement through
lists that explicitly identify specific drugs as ineligible for
reimbursement.  Drugs may be excluded from the payment system because
they (1) offer questionable therapeutic value or (2) have prices that
are high relative to alternative medications of similar or equal
therapeutic value.\22

Three of the countries we studied are either establishing or
expanding negative drug lists in an attempt to limit prescription
drug dispensing.  In January 1994, France established a list of 24
drugs and procedures which will not be reimbursed.  The United
Kingdom is in the process of excluding additional drugs from its
reimbursable lists.  Germany currently has a nonreimbursable drug
list, but after 1995 plans to replace this with a list of drugs that
are eligible for reimbursement. 


--------------------
\22 In Germany, drugs will be excluded from reimbursement only if
they have questionable therapeutic value; in France, Sweden, and the
United Kingdom, reimbursement decisions take into account a drug's
price as well as its therapeutic value. 


THE EFFECT OF PRESCRIPTION DRUG
PRICES ON PHARMACEUTICAL RESEARCH
AND DEVELOPMENT EXPENDITURES
============================================================ Chapter 3

As we described in chapter 2, France, Germany, Sweden, and the United
Kingdom have employed a variety of policies to control pharmaceutical
spending and prescription drug prices.  Despite the differences in
their specific policies, each country confronts a similar
dilemma--preserving a strong domestic pharmaceutical industry while
controlling national spending on pharmaceuticals.  Specifically, the
concern has centered on the potential trade-off between low drug
prices and pharmaceutical firms' spending on R&D.  Although other
factors are also important, economic analysis confirms that higher
drug prices strengthen the incentives for firms to invest in
pharmaceutical R&D.  Nonetheless, empirical estimates of the size of
this price-R&D relationship are imprecise, and the significance of
drug price decreases for the development of new drugs is uncertain. 


   TENSION BETWEEN LOW DRUG PRICES
   AND R&D INCENTIVES TRANSCENDS
   THE SPECIFICS OF EACH COUNTRY'S
   POLICIES
---------------------------------------------------------- Chapter 3:1

France, Germany, Sweden, and the United Kingdom have adopted
strikingly different price restraint mechanisms--from the
product-by-product price controls in France to the United Kingdom's
profit control scheme.  Despite the differences in their specific
policies, these countries' measures have had a common result--lower
prescription drug prices.\1 In each of the countries we studied, the
national authorities face a potential conflict between their interest
in containing prescription drug costs and their concern that
reductions in drug prices and spending may hurt the domestic
pharmaceutical industry.  In particular, this concern has focused on
the potential depressing effect of lower drug prices on
pharmaceutical R&D.  Analysis of this relationship between drug
prices and R&D reveals a tension that transcends the specifics of
each country's pharmaceutical policies. 

Although the specific form of pharmaceutical regulation will be
important to pharmaceutical companies, these policies' impact on R&D
stems primarily from their influence on prescription drug prices.  A
reduction in prescription drug prices can be expected to reduce
companies' spending on pharmaceutical R&D, because firms will have
less incentive to invest in R&D when they expect to receive lower
prices for their products.  Moreover, a reduction in drug prices can
stem from any source--either a government regulation or other factors
in the market.  In this respect, the tension between drug prices and
R&D transcends policy specifics. 

The conflict between cost containment and R&D is not confined to
countries like France, Germany, Sweden, and the United Kingdom--that
is, countries that have adopted explicit pharmaceutical spending
control policies--because the drug price-R&D connection does not rest
on specific policies adopted in specific countries.  Consequently,
countries with less regulated pharmaceutical markets, like the United
States, also contend with potential trade-offs between low drug
prices and high spending on research.  As a result, the general
relationship between drug prices and R&D can be estimated by
analyzing data from a wide range of countries, from the highly
regulated to the more market-oriented. 


--------------------
\1 These regulations reduce drug prices relative to their level
without regulations.  Even if regulated drug prices increase, as they
often do, they usually rise less rapidly than in the absence of
regulations. 


   HIGHER DRUG PRICES STRENGTHEN
   THE INCENTIVE FOR R&D SPENDING,
   BUT OTHER FACTORS ALSO MATTER
---------------------------------------------------------- Chapter 3:2

When prescription drug prices decline, pharmaceutical companies are
faced with a potential loss of revenue--for both the drugs they
currently produce and especially for their future product line. 
Pharmaceutical companies have less incentive to invest in costly R&D
if the resulting products will bring in lower profits.  However, a
number of factors--
including both government policies and market forces--influence
firms' expectations of future profits, and are therefore important to
firms' R&D decisions. 


      REGULATIONS THAT LOWER DRUG
      PRICES REDUCE INCENTIVES FOR
      R&D
-------------------------------------------------------- Chapter 3:2.1

As we described in chapter 2, governments in France, Germany, Sweden,
and the United Kingdom have employed a range of strategies to control
pharmaceutical prices and spending.  In each country, these
regulations have not only reduced the prices of today's
pharmaceutical products, but also will undoubtedly put downward
pressure on the prices of tomorrow's prescription drugs.  For
example, managers of pharmaceutical firms can expect French price
controls to continue exerting downward pressure on drug prices in
France.  In addition, firms in the United Kingdom are restricted from
increasing drug prices without government approval.  In general, both
current laws and prudent business judgment lead firms to expect that
in the future, prescription drug prices will be lower with government
regulation than they would have been otherwise. 

These future prices are central to companies' R&D decisions.  Firms
invest in R&D today in order to discover new pharmaceutical products,
which will earn profits in the future.  According to the Office of
Technology Assessment, a typical new drug is introduced to the market
only after an average of 12 years of research and testing.\2 If firms
foresee lower earnings potential for future products, R&D becomes
less attractive. 


--------------------
\2 In this context, a "new drug" refers to a drug based on a new
chemical entity or compound, rather than, for example, an
extended-release form of an existing drug. 


      COUNTRIES WITH HIGHER DRUG
      PRICES ARE OFTEN DRUG
      INNOVATORS, BUT EXCEPTIONS
      UNDERLINE THE IMPORTANCE OF
      NONPRICE FACTORS
-------------------------------------------------------- Chapter 3:2.2

The pattern of world pharmaceutical R&D generally confirms that high
drug prices create greater incentives for R&D.  Among major
industrialized countries, a pattern prevails--countries with higher
drug prices tend to be associated with more pharmaceutical R&D. 
These countries (including the United States and Germany) have high
prescription drug prices, high R&D, and many new drugs.  Conversely,
in low-price countries like Spain and Australia, R&D spending is low,
and very few new drugs are developed. 

However, despite this general link between drug prices and R&D,
significant exceptions exist--a few low-price countries are
pharmaceutical innovators, while several high-price countries lack
strong industries.  For example, in Sweden and the United Kingdom, an
innovative industry coexists with price regulation, while in Canada,
average or high prices have not resulted in significant R&D.  France
represents an intermediate case, neither fully conforming to the
general pattern nor sharply deviating.  France has been able to
produce some innovative drugs, despite low domestic prices.  However,
with these low drug prices, France has experienced a decline in new
drug development by French firms, and French products have not been
widely adopted overseas.\3

Countries with apparently weaker connections between drug prices and
R&D--like the United Kingdom and Canada--reveal the importance of
nonprice determinants of R&D.  Although prescription drug prices can
influence pharmaceutical R&D, drug prices are clearly not the only
factor affecting research decisions, nor are they necessarily the
most powerful.  For example, while Canada has relatively high drug
prices compared to many European countries, Canada's compulsory
licensing policy and weakened patent protections appear to have
limited the Canadian industry's research spending.  Government
policies, from tax credits to patent laws, can stimulate or deter
pharmaceutical R&D investment.  Likewise, market forces can encourage
or discourage firms from spending more on research. 


--------------------
\3 For example, in the period 1975 through 1989, France produced 12.2
percent of the world's new pharmaceutical products, but only 3.1
percent of "globalized" products--that is, those products available
in six of the world's seven major pharmaceutical markets.  (See P. 
Etienne Barral, Fifteen Years of Pharmaceutical Research Results
Throughout the World (1991).)


         GOVERNMENT POLICIES
         INFLUENCE PHARMACEUTICAL
         R&D
------------------------------------------------------ Chapter 3:2.2.1

The government's impact on pharmaceutical research does not arise
solely from drug price regulation, but also from other arenas such as
patent policy and tax law.  R&D decisions also hinge on these other
government policies, which are described below: 

  The effective patent life for new products is the period of time
     for which a firm has the exclusive right to market a new drug. 
     The longer a firm is protected from competition, the greater the
     profits the firm can expect to earn from a new drug, and the
     greater incentive for R&D.  However, firms must apply for
     patents as soon as a compound is discovered, before the drug is
     reviewed for safety and efficacy by the national authority. 
     While the drug is being reviewed, some of the drug's patent term
     is "used up" before the product reaches the market.  The longer
     the approval process takes, the shorter the firm's "effective"
     patent life is. 

  Tax policy can create additional incentives for R&D.  Some
     countries (including the United States) try to encourage firms'
     R&D efforts by giving firms special tax credits for each dollar
     they spend on R&D.  These tax credits reduce the firm's R&D
     costs; therefore, tax credits may provide firms with an
     incentive to increase their R&D expenditures. 

  Public funding (subsidies or outright grants for scientific
     research) may stimulate firms to do more applied research. 

  Product liability law may deter firms from R&D projects
     (particularly in certain therapeutic categories) if
     pharmaceutical companies cannot protect themselves against the
     risk of costly suits related to new products. 


         MARKET FORCES ALSO AFFECT
         PHARMACEUTICAL R&D
------------------------------------------------------ Chapter 3:2.2.2

Although the pharmaceutical market is heavily influenced by
government policy, market forces also play an important role.  The
choices made by consumers and their physicians, together with
government policy, create the market environment on which firms must
base their R&D decisions.  As described below, these market
forces--which differ across countries--are also important factors in
R&D decisions. 

  The size of the market (both domestic and foreign) for a
     pharmaceutical product will influence the amount of revenue a
     firm can expect to receive for a product, and thereby affect its
     R&D.  For example, countries with universal insurance coverage
     for prescription drugs may have higher consumption per capita,
     as consumers have access to pharmaceutical products regardless
     of their ability to pay.  Lifestyle choice, cultural norms, and
     household incomes may also influence the use of prescription
     drugs. 

  Wage and equipment costs form part of the out-of-pocket expenses
     involved in pharmaceutical R&D.  An increase in these costs
     would make it more expensive for firms to conduct R&D. 

  The "scientific infrastructure" of a region or country helps
     determine the pharmaceutical industry's access to qualified
     personnel.  For example, a strong, nearby academic community may
     make it easier to hire qualified people and may enhance research
     output. 


   REDUCTIONS IN DRUG PRICES LEAD
   TO LOWER R&D EXPENDITURES
---------------------------------------------------------- Chapter 3:3

In our statistical analysis, we found a positive relationship between
drug prices and R&D, reinforcing the reasoning that higher drug
prices strengthen the incentive for R&D.  We estimated this
relationship several different ways, and the size of the estimated
R&D response to prices varied accordingly.  According to a
representative estimate, a 1-percent decline in drug prices leads to
a 0.68-percent decline in R&D spending.  However, while statistically
distinguishable from zero, this estimate is statistically imprecise. 
The data are consistent with a response as high as 1.2 percent or as
low as 0.1 percent.  (See app.  V for a more detailed discussion of
these results.)

We obtained these estimates using a multiple regression model that
relates changes in pharmaceutical R&D to changes in drug prices,
controlling for the influence of other factors.  The results pertain
to data on the pharmaceutical R&D spending of 87 companies, for the
years 1988 to 1991.  This group of firms covers 12 countries,
including France, Germany, Sweden, the United Kingdom, and the United
States.  We used the national index of drug prices to characterize
the price levels facing a given company in a particular country.\4
Our results are consistent with previous economic analyses of the
pharmaceutical industry, in which other measures of the incentive for
R&D--for example, firms' profit rates and market shares--were
positively related to R&D.\5


--------------------
\4 This is a proxy for the company's expectation of the pricing
environment that it will face when the results of its R&D--its new
products in the future--reach the market. 

\5 For example, see William S.  Comanor, "The Political Economy of
the Pharmaceutical Industry," Journal of Economic Literature, 24 (3)
(Sept.  1986), pp.  1178-1217; and Global Competitiveness of U.S. 
Advanced-Technology Manufacturing Industries:  Pharmaceuticals, U.S. 
International Trade Commission, Report to the Committee on Finance,
U.S.  Senate.  US ITC Pub.  2437 (Washington, D.C.:  1991). 


   INFORMATION IS LIMITED ABOUT
   SIZE AND SIGNIFICANCE OF
   POTENTIAL R&D REDUCTION
---------------------------------------------------------- Chapter 3:4

Although our analysis reaches the general conclusion that higher drug
prices encourage pharmaceutical R&D, we have more limited evidence
about the effects of the specific policies adopted in specific
countries.  We urge a cautious interpretation of our results, for
three major reasons:  (1) we expect that the strength of the
price-R&D relationship will differ across countries; (2) we do not
know whether a decrease in R&D spending by firms would bear more on
innovative or imitative drug products; and (3) these results suggest
that drug prices are negatively related to R&D, but convey more
limited information about the relationship between any specific
policy of price regulation and R&D.  A regulation's impact may depend
not only on the resulting changes in prices but also on other
factors, such as the size of the market on which these policies are
imposed.\6 Therefore, the size of the effect of drug price regulation
on new drug development remains an open question. 


--------------------
\6 In addition, price regulation could conceivably be implemented in
conjunction with other policies--such as expanded insurance coverage
for prescription drugs--that might be expected to encourage R&D. 


      THE STRENGTH OF THE
      PRICE-R&D RELATIONSHIP IS
      EXPECTED TO VARY ACROSS
      COUNTRIES
-------------------------------------------------------- Chapter 3:4.1

Our statistical analysis estimates an average price-R&D linkage
across national boundaries.  However, we expect that the importance
of domestic prices to R&D will differ from country to country. 
Economic theory suggests that two factors, which vary across
countries, will affect the strength of the relationship between price
regulation and R&D spending.  First, the price-R&D connection can
vary with the size of the pharmaceutical market in the manufacturer's
home country.  If a small country (such as Sweden) exhibited falling
domestic prices, the impact on drug company revenues and R&D would be
limited because this country's consumers account for only a small
share of the global pharmaceutical market.  By contrast, a loss of
revenue in a larger market would likely have more far-reaching
effects on domestic and foreign pharmaceutical firms.  Second,
domestic firms that are export-oriented will be less concerned with
prices in the home country.  For example, for firms in the United
Kingdom, which earn much of their revenues and profits from exports,
prices in the home market are less important.  By contrast, when
firms are more heavily oriented toward their domestic market, then
domestic prices are likely to have a stronger impact on the R&D
decision. 


      REGULATION'S IMPACT ON THE
      PRODUCTION OF NEW
      INNOVATIONS IS UNCERTAIN
-------------------------------------------------------- Chapter 3:4.2

Although R&D spending undoubtedly leads to the discovery of new
drugs, we cannot tell how a decrease in R&D spending would affect the
distribution of new drug discoveries.  Not all drugs are equally
valuable to physicians and their patients.  So-called "breakthrough"
drugs are based on new compounds and represent a substantial
improvement over existing therapies.  These drugs are of considerable
value in helping people get well.  By contrast, "me-too" drugs
represent little or no improvement over current treatments. 
Clinically, these "me-too" drugs are generally less valuable to
patients than breakthrough products.\7 We cannot determine whether a
reduction in R&D spending would manifest itself in "breakthrough" or
"me-too" drug projects. 

The effects of price regulation on the quality of R&D will depend on
both the average price level for new drugs and the relationship
between the price of a new drug and its therapeutic value.  For
example, if regulators allow higher introductory prices for
breakthrough drugs than for me-too drugs, this may create additional
incentives for manufacturers to produce more innovative drugs.  In
addition, for drug manufacturers to have sufficient incentive to
undertake high-risk projects, the firm must be able to pay the costs
of the research projects that did not prove successful.  In general,
the lower the price given to new drugs, the less likely that the firm
can bear the cost of failed R&D.  Therefore, other things being
equal, lower introductory prices can create a greater incentive for
the firm to concentrate its R&D in projects with a higher probability
of eventual success.  Data are presently inadequate to estimate the
extent to which these incentives may change the mix of breakthrough
and me-too drugs. 


--------------------
\7 However, the clinical value of a new drug may not fully reflect
that drug's value to society.  For example, a "me-too" drug may,
through price competition, lower the prices of competing therapies. 


      REGULATION'S IMPACT ON
      QUANTITY AND QUALITY OF R&D
      WILL DEPEND ON REGULATORY
      DESIGN AND IMPLEMENTATION
-------------------------------------------------------- Chapter 3:4.3

While the major R&D impact of spending control strategies comes from
their effect on prescription drug prices, the important though
secondary effects of these policies may stem from their design. 
However, evidence of these potential effects is limited.



         CONTROLLING THE
         INTRODUCTORY PRICE OR
         CONTROLLING ONLY THE RATE
         OF INCREASE? 
------------------------------------------------------ Chapter 3:4.3.1

As described in chapter 2, the government in France controls both the
price at which a drug is introduced to the market and any subsequent
price increases.  The United Kingdom, by contrast, allows
manufacturers to freely set introductory prices (subject to the
profit constraint) but largely controls future price increases.  The
United Kingdom's policy may create an incentive for new drug
development, as companies can increase the average revenue of their
product line only by putting new drugs on the market.  The quality of
new drugs, however, would not be assured.  Companies may have an
increased incentive to tinker with delivery mechanisms and dosages or
produce imitative drugs in order to send "new" products to the market
and command higher prices. 


         ARE FIRMS GIVEN HIGHER
         PRICES TO REWARD R&D? 
------------------------------------------------------ Chapter 3:4.3.2

In addition to the United Kingdom, some countries may allow companies
to charge higher introductory prices for innovative products than for
less significant drugs.  For example, in France, regulators will
allow a higher price for a more innovative product.  However, critics
of the French system contend that this premium is not large enough to
create sufficient incentive for innovation. 


         ARE ALL FIRMS TREATED
         EQUALLY, OR DO SOME FIRMS
         HAVE AN ADVANTAGE? 
------------------------------------------------------ Chapter 3:4.3.3

Some countries' price regulatory authorities may treat firms
differently depending on their national origin.  In the United
Kingdom, for example, a company's allowed profit is calculated on the
basis of the capital it has invested in the United Kingdom; the more
capital the firm invests in the United Kingdom, the higher the
allowed profits.  This policy gives an advantage to British firms and
foreign firms that locate their European offices in Britain, and may
encourage R&D in that country. 


CONCLUSIONS AND POLICY
IMPLICATIONS
============================================================ Chapter 4

While for decades many European countries have intervened in the
pharmaceutical market to restrain prices and spending, the United
States has allowed drug manufacturers to set prices freely. 
Recently, however, public dissatisfaction with the rising cost of
prescription drugs has prompted new congressional proposals to
restrict pharmaceutical prices.  Indeed, as efforts at redesigning
the U.S.  health care system have accelerated, interest in drug price
regulation has been heightened in both the legislative and executive
branches.  Lacking firsthand experience with pharmaceutical price and
spending controls, the United States can learn from its European
counterparts' attempts to contain drug prices and spending.  The
findings of this report suggest three lessons that should be
considered: 

1. Price controls for individual products are only one of a large
number of pharmaceutical spending control policies, ranging from
strict regulatory approaches to more market-based solutions. 

Though price controls on prescription drugs have been prominent in
Europe, they do not exhaust the variety of techniques and
philosophical orientations that U.S.  decision makers can consider. 
To control pharmaceutical expenditures, France, Germany, Sweden, and
the United Kingdom each employs an array of policies, some regulatory
and some market-based.  The balance struck varies from country to
country--ranging from controlling corporate and physician actions by
legal and administrative sanctions to strengthening competition by
reshaping incentives.  For example, France has emphasized the
regulatory approach by imposing stringent product-by-product price
controls.  By contrast, the United Kingdom has evolved a more
eclectic strategy:  profit controls--a relatively flexible regulatory
approach that allows companies considerable pricing freedom--are
coupled with policies to sharpen competition among drug companies by
encouraging physicians to prescribe less expensive medicines. 

A government's use of price or profit controls is sometimes confused
with its preference for high or low drug prices.  Though countries
with unrestrained pricing tend to have high prices, the introduction
in such countries of policies to strengthen competition might well
reduce drug prices significantly.  Likewise, though price controls
tend to be found in countries (like France) with drug prices that are
low, price controls and other types of price regulation could be used
to achieve reductions in drug prices that are more modest.  In short,
it is important to distinguish the goals for drug prices from the
means available to achieve those goals. 

2. An effective approach to reducing pharmaceutical expenditures is
likely to be multipronged, because policies exclusively aimed at drug
prices are insufficient to control pharmaceutical spending. 

Despite the existence in these countries of price and profit
controls, total spending on pharmaceuticals has continued to rise. 
Contributing to this upward trend in spending is growth in both drug
prices and the quantity of drugs sold.  In turn, the quantity sold
reflects the actions of drug companies as well as of consumers and
their physicians.  In line with this analysis, these countries have
augmented their traditional controls, which primarily targeted the
pharmaceutical industry, with additional strategies aimed at
consumers and physicians.  These policies--such as the increased use
of cost-sharing and the adoption of physician drug budgets--are
intended to encourage consumers and physicians to more carefully
evaluate whether a prescription should be written, or whether a
lower-priced drug could be substituted for a higher-priced product. 

3. The presence of pharmaceutical price regulation does not preclude
the existence of an innovative drug industry, but lower drug prices
can discourage pharmaceutical research and development.  However, it
cannot yet be determined the extent to which less R&D translates into
fewer new drugs that offer substantial therapeutic improvements. 

In France, Germany, Sweden, and the United Kingdom, innovative
industries coexist with drug price regulation.  However, our analysis
indicates that higher drug prices contribute to the development of
new drugs by encouraging firms to devote more resources to R&D. 
Therefore, a decline in drug prices, from whatever cause--regulation,
pro-competitive policy, or other market forces--can be expected to
lead to a decline in firms' expenditures on R&D for new drugs. 

The significance of less R&D is clouded by several factors.  First,
the extent of the response of R&D to lower drug prices has not been
established precisely.  Second, the significance to society of a
reduction in R&D would be greater if only breakthrough drugs were
affected, and much less if only the development of me-too drugs were
slowed. 

Although government regulation has restrained drug prices in the four
countries we examined, the implications--and the desirability--of
similar intervention in the U.S.  pharmaceutical market are unclear. 
More specifically, the effects of a price reduction in any of these
countries may differ from the effects of a similar price reduction in
the United States, because each country represents a relatively
smaller share of the global pharmaceutical market.  In addition, the
particular price and spending control policies used in these
countries may not be readily transferrable to the United States
because of institutional differences across countries.  In any case,
any gains from regulation of drug prices or spending must be weighed
against the consequences of such regulations for pharmaceutical
research and development. 


FRANCE'S DRUG SPENDING CONTROL
POLICIES
=========================================================== Appendix I

As a result of France's strategies to control pharmaceutical
spending, drug prices in France are among the lowest in Europe.  In
conjunction with its regulation of the health insurance system, the
French government has imposed a variety of controls on pharmaceutical
prices that apply to participants throughout the pharmaceutical
market:  drug manufacturers, drug wholesalers and pharmacists,
consumers, and physicians. 

Although France's low prices have kept pharmaceutical expenditures
from rising faster than they would have without its price controls,
persistent rapid growth in spending has led the French government to
consider supplementing price controls with significant new measures. 
France's low prices are viewed by some government officials and
academic experts as encouraging over-consumption while discouraging
drug companies from investing more in research and development.  In
1991, the Socialist government then in power proposed a reform that
called for two principal budgets:  (1) a global budget for total
pharmaceutical expenditures, composed of budgets for each
manufacturer; and (2) budgets for certain innovative drugs.  The
reform was designed to mitigate the unwanted side effects of price
controls by limiting increases in drug costs while encouraging
expenditures on R&D the government withdrew this plan from
consideration after extensive debate.  However, in January 1994,
pharmaceutical industry representatives and government officials
reached an agreement that adopts many aspects of this proposal. 


   OVERVIEW OF THE FRENCH HEALTH
   INSURANCE SYSTEM
--------------------------------------------------------- Appendix I:1

The current French health care financing system, established in 1945,
is part of the Social Security system (Sï¿½curitï¿½ Sociale).  In keeping
with a tradition dating back to 1893, when free medical assistance
was first granted to the poor, it is designed to provide universal
access to health care. 

Three main national health insurance funds provide comprehensive
health insurance to over 98 percent of the population.  Most people--
approximately 75 percent--belong to a single national "sickness
fund"--in effect, a highly regulated nonprofit insurer; all other
insured are covered by the other two national health insurance funds
or one of 15 special (occupation-based) funds.  Consumers do not pay
deductibles for health care services, but copayments for physician
and hospital services can reach 30 percent of the regulated fees. 
Most people (about 72 percent) also have complementary nonprofit and
private insurance to pay the consumers' share of the costs not
covered by their standard benefit package. 

In 1991, health care represented 9.1 percent of the French Gross
Domestic Product.  The Social Security system financed about 73
percent of these expenditures.  The government and local authorities
contributed another 1.1 percent through earmarked taxes; private and
nonprofit complementary insurance contributed 6.2 percent; and
consumers paid the remainder out-of-pocket (about 19.7 percent). 

Employers and employees both share in the cost of health insurance
based on a government formula.  The employers' and employees'
contribution as a share of gross wages and salaries average about
12.8 and 6.8 percent, respectively.  In 1991, total health care
expenditures, which have slowly risen in real terms, were 498,130
million francs (or about $88.3 billion in 1991 dollars). 


      PHARMACEUTICAL COVERAGE IN
      THE FRENCH HEALTH INSURANCE
      SYSTEM
------------------------------------------------------- Appendix I:1.1

The Social Security system is the principal purchaser of
pharmaceuticals, which accounted for approximately 14 percent of the
French health care budget in 1991.  The system, together with French
private and nonprofit insurance funds, provides nearly complete
coverage for pharmaceutical products in France.  Usually, consumers
pay the full cost of the prescription and complete a form requesting
reimbursement from the Social Security system or health insurer. 
However, consumers in France sometimes ask pharmacists to request the
reimbursable amount directly from the Social Security system or the
insurer and pay only the copayment amount. 

About 4,200 different drug products are available on the French
market.  In contrast to other European countries, France has
virtually no generic drug market.\1 The small size of the generic
drug market is attributed to (1) low prices for brand name drugs and
(2) French laws prohibiting pharmacists from substituting a generic
drug for a brand-name drug.\2


--------------------
\1 In 1991, generic drugs comprised less than 5 percent of total drug
sales. 

\2 Parallel imports--identical products imported from countries where
drug prices are lower--are also discouraged for these reasons. 
However, because of France's low drug prices, it has benefited from
"parallel exporting" and has been able to improve its pharmaceutical
trade balance by acting as a form of drug discount warehouse to other
countries. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG MANUFACTURERS
--------------------------------------------------------- Appendix I:2

France uses a three-step process to set drug prices and closely
monitors the increase in prices of new and existing drugs.  Recently,
government officials and pharmaceutical industry representatives
supplemented the existing price setting process with an informal
agreement that, among other things, provides drug companies with
greater flexibility in setting drug prices. 


      THE PRICE SETTING PROCESS
------------------------------------------------------- Appendix I:2.1

In theory, drug manufacturers in France are free to set drug prices. 
In practice, France has price controls on most drug products,
representing about 80 percent of the drugs sold in France.\3 Since
Social Security is the largest payer of pharmaceuticals, the
government is able to induce manufacturers to offer most products at
the government-set level. 

Introductory drug prices are determined in a three-step process.\4
First, the AMM (Autorisation de mise sur le marchï¿½) Commission,
similar to the U.S.  Food and Drug Administration, reviews each drug
for quality, safety, and efficacy to determine whether to issue a
marketing license. 

Second, the Transparency Commission, composed in part of
representatives from industry, medical universities, the national
sickness funds, and the Ministries of Health and Social Affairs,
reviews each drug to determine whether, compared to the existing drug
for the same indication, it will be more cost-effective and produce
better clinical results or fewer side effects.  The Transparency
Commission recommends a "technical price," based in part on whether
the drug represents a major or minor advance in therapy and on the
number of drugs in the same therapeutic class.  If no other drugs are
in the class, the Commission evaluates the new drug by comparing it
to the average cost of treating the disease without this drug.  If
the Commission finds that the new drug offers additional therapeutic
value relative to currently available drugs or treatments, a higher
price is granted to the drug. 

Third, the Economic Committee, primarily composed of representatives
from four government ministries (the Ministries of Finance,
Economics, Health, and Social Affairs), reviews the Transparency
Commission's recommended technical price and the manufacturer's
suggested price.  Following this review, the Economic Committee
proposes to the manufacturer an "economic price" for the drug.  This
economic price may be higher than the technical price if the new drug
is expected to offer benefits to the national economy, such as
increased exports, job creation, increased investments, or more R&D. 

The economic price becomes the basis of negotiations between
representatives of the Economic Committee and the manufacturer over
the launch price of the drug and, in effect, becomes the drug price
after the negotiations are completed.  The negotiations between the
Economic Committee and the manufacturer may last 6 to 12 months.  A
manufacturer can accept or reject the proposed economic price, which
is fixed for a period of 2-1/2 years.\5 However, most manufacturers
eventually agree with the Committee's proposed price because sales
volumes would otherwise be drastically reduced.\6

Drugs not reimbursed by the Social Security system are subject to
review by the AMM Commission, but not subject to reviews by the
Transparency Commission or Economic Committee.  Manufacturers of
these drugs, which account for about 10 percent of the drug market,
can set prices freely.\7


--------------------
\3 About 11 percent are sold in hospitals, which negotiate drug
prices on their own.  The remaining 9 percent are sold privately,
without any price constraints. 

\4 During this three-step process, the government also sets the
reimbursement rate of outpatient prescription drugs paid for by the
Social Security system. 

\5 Price increases for new drugs--drugs on the market for less than
2-1/2 years--are seldom granted. 

\6 There are no examples of important drugs that are not reimbursed
by the government. 

\7 These items are generally over-the-counter drugs.  In some cases,
they are products which are put on the market despite the lack of an
agreed upon price between the manufacturer and the government. 
Examples of this latter group of products include third-generation
contraceptives and nicotine patches. 


         DETERMINATION OF DRUG
         PRICE INCREASES
----------------------------------------------------- Appendix I:2.1.1

For drugs marketed longer than 2-1/2 years, price increases are
strictly controlled.  The government limits the price changes for
these drugs through blanket pricing directives, which raise or lower
the price of all drugs on the market by a set percentage.\8 The most
recent directive, issued in 1991, mandated a price decrease of 2.5
percent on all drugs.  The term "blanket pricing directive" is
somewhat misleading, because it does not apply uniformly to all
drugs.  Rather, the price change for individual drug products may
exceed or fall short of the "blanket" price change, as the firm's
average increase or decrease is equal to the mandated "blanket
change."


--------------------
\8 One industry official told us that the government has increasingly
opted to reexamine and lower the prices of individual drugs marketed
after 30 months, especially if the quantities sold are significant. 


      PROPOSAL CALLING FOR GLOBAL
      AND INDIVIDUAL DRUG-SPECIFIC
      BUDGETS
------------------------------------------------------- Appendix I:2.2

A 1991 reform proposed by the government to the National Assembly
would have provided manufacturers with incentives to increase
research and capital expenditures, limited increases in drug prices,
and promoted the more cost-effective use of drugs without reducing
France's high level of social benefit coverage.  The reform would
also have limited the amount spent by manufacturers on advertising,
which some believe is excessive. 

To discourage consumption, the government proposed setting a global
limit on reimbursement for prescription drugs through the Social
Security system.  This global limit would have been subdivided into
separate budgets that would apply to total sales, manufacturer by
manufacturer.  Within each manufacturer's sales budget, the
manufacturer could freely set each drug's price.  Innovative drugs
would have been given separate drug-specific budgets for a period of
5 years as a means of encouraging investment.  A manufacturer
exceeding its budget would have been required to explain why and to
refund most of all of the excess to the Social Security system.  The
proposal was controversial and was withdrawn on December 31, 1992.\9
However, in January 1994, pharmaceutical industry representatives and
government officials supplemented the existing price setting process
with an informal agreement that incorporates many aspects of the 1991
proposal.  It would establish a target growth rate for pharmaceutical
expenditures and provide greater flexibility to drug companies in
setting prices.  In addition, the agreement may result in a reduction
in reimbursement rates for older products.  Details of this agreement
were not available at the end of January 1994. 


--------------------
\9 The government did, however, increase the tax on sales promotion
costs from 7 percent to 9 percent.  This increase, which took effect
at the end of 1992, is viewed by the pharmaceutical industry as an
indirect means of limiting expenditures on sales promotion. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG WHOLESALERS AND
   RETAILERS
--------------------------------------------------------- Appendix I:3

In France, the government regulates wholesale and retail margins to
help control retail prices of prescription drugs.\10 These margins
are regulated by decree and expressed as a percentage of the
manufacturer's price.\11 According to a government official,
modifying the value-added tax and pharmacists' and wholesalers'
margins has produced substantial savings in pharmaceutical costs over
the past 20 years.  These modifications over the last two decades
made it possible to lower the retail price of drugs by more than an
estimated 25 percent; these reductions were accomplished without
altering manufacturers' prices.  In 1991, the numerous successive
reductions of margins undertaken since 1967 resulted, according to a
government estimate, in savings of nearly 20 billion francs (or about
$3.55 billion in 1991 dollars)--roughly one-fifth of that year's
pharmaceutical expenditures by the government.\12


--------------------
\10 Wholesale and retail margins are not controlled by the government
for nonreimbursable drugs; however, these margins are only slightly
higher than the margins for reimbursable drugs. 

\11 The normal wholesalers' margin is set at 10.74 percent of the
manufacturer's price before tax.  The pharmacist's before tax gross
profit is calculated according to a sliding scale, which decreases in
proportion to the price of the drug. 

\12 We did not obtain sufficient information to validate these
government estimates. 


   SPENDING CONTROL STRATEGIES
   AIMED AT CONSUMERS
--------------------------------------------------------- Appendix I:4

In France, the consumer's share of drug costs represents the
proportion of the drug cost not reimbursed by the Social Security
system.  Each drug's reimbursement rate is set at either 100, 65, 35,
or 0 percent of the drug's cost, and the consumer pays the
remainder.\13 The government sets the reimbursement rate of
outpatient prescription drugs paid for by the Social Security system
during the three-step price setting process described earlier. 

Specifically, the Transparency Commission recommends one of three
reimbursement rate categories.  The reimbursement rate represents the
proportion of the drug cost covered by the Social Security system. 
The reimbursement rate is set at 100 percent for 128 vital medicines
and for all drugs used to treat patients with any one of over 30
diseases (defined as "long and costly"), such as Parkinson's disease
and AIDS.  The reimbursement rate declines to 35 percent for drugs
used to treat disorders or ailments that are not normally severe
(e.g., antiseptics and laxatives).  Prescription drugs not reimbursed
at 35 or 100 percent are reimbursed at 65 percent.  The Economic
Committee reviews the Transparency Commission's recommended
reimbursement rate and determines a final reimbursement rate. 

Certain groups of people are exempt from copayment.  These include
the chronically ill, the poor, the handicapped, and expectant
mothers.  Table I.1 shows that a large portion of the government's
expenditure for prescription drugs is for drugs that do not require
consumer copayment (i.e., drugs reimbursed at 100 percent). 



                          Table I.1
           
               Percentage of Prescription Drug
           Expenditures by Reimbursement Categories
                           in 1991

Reimbursement         Percent of          Percent of total
categories\a          expenditures        market
--------------------  ------------------  ------------------
100 percent           41.0 percent        1.0 percent

70 percent            48.0 percent        64.0 percent

40 percent            11.0 percent        21.0 percent
------------------------------------------------------------
\a Prior to July 1993, the reimbursement rates were set at 100, 70,
40, or 0 percent of drug costs. 

Source:  P.  Etienne Barral. 


--------------------
\13 In actual practice, most patients have supplementary insurance
for prescription drugs.  This insurance picks up most, if not all, of
the copayment cost. 


      PROPOSAL TO LIMIT FULL DRUG
      REIMBURSEMENT COST
------------------------------------------------------- Appendix I:4.1

A 1987 plan reduced the number of drugs reimbursed at 100 percent by
limiting it to specific drugs for specific illnesses, rather than all
drugs for specific illnesses, but this plan was controversial and was
rescinded after 1 year. 


   SPENDING CONTROL STRATEGIES
   AIMED AT PHYSICIANS
--------------------------------------------------------- Appendix I:5

In France, the government uses three measures aimed at providing
physicians with prescribing information.  The government also
periodically reviews its drug reimbursement list and recently
established a drug agency, which will focus on scientific and
technical matters. 


      PROVIDE PHYSICIANS WITH
      PRESCRIBING INFORMATION
------------------------------------------------------- Appendix I:5.1

Pharmaceutical consumption in France is high compared to other
European countries.  As a result, the government has instituted three
measures designed to reduce drug consumption.  First, medical-social
commissions examine each physician's prescribing pattern and ask
physicians prescribing more than the average to limit their
prescribing. 

Second, the Ministry of Health distributes transparency sheets
designed to provide summary information to physicians--for example,
data on the costs of daily treatment, dosage forms, and drug
interactions--which helps them select the appropriate drugs.  As of
December 1991, 30 transparency sheets had been issued; however, these
have been criticized for being out-of-date, incomplete, and overly
complex. 

Third, the national convention of physicians tracks prescription drug
costs in relation to the number of physicians' office and home
consultations on a quarterly basis.  These data are supposed to
enable physicians to prescribe more cost-effectively by
self-monitoring their prescribing patterns; however, this does not
appear to be the case, as physicians frequently ignore such data.\14


--------------------
\14 In December 1992, the National Assembly passed a law calling for
tighter physician controls.  One of the controls resulting from this
law is described below. 


      PERIODIC REVIEW OF THE
      REIMBURSABLE DRUG LIST
------------------------------------------------------- Appendix I:5.2

France also uses a list to define all drugs eligible for
reimbursement at the 100-percent, 65-percent, and 35-percent levels. 
The government periodically evaluates a drug's reimbursement status
and sometimes lowers the reimbursement category for particular drugs
(for example, from 65 percent to 35 percent). 

In January 1994, the French government adopted a body of guidelines
restricting 24 drugs or procedures.  These guidelines, which are
based on an agreement between the insurance industry and the
principal French physicians' unions, include restrictions on
treatments for conditions such as pregnancy, hypertension, and
hypoglycemia.  Physicians consistently practicing outside the
guidelines will be asked to explain their actions before a local
committee of physicians and insurance representatives and can be
assessed a financial penalty if they fail to provide a justifiable
explanation for these practices. 


      CREATION OF NEW DRUG AGENCY
------------------------------------------------------- Appendix I:5.3

In December 1992, the National Assembly approved the creation of a
new drug agency.  The new agency does not involve itself in pricing
and reimbursement decisions; rather, it confines itself to scientific
and technical issues and is similar in function to the U.S.  Food and
Drug Administration.  The agency provides a stronger, more modern
structure for the registration of drugs and for guaranteeing their
quality, safety, and efficacy. 


GERMANY'S DRUG SPENDING CONTROL
POLICIES
========================================================== Appendix II

Pharmaceutical spending control strategies in Germany are designed to
restrain the growth of drug expenditures without directly controlling
drug prices.  As part of its regulation of the health insurance
system, the German government has imposed a variety of controls on
pharmaceutical payment that apply to various participants in the
pharmaceutical market:  drug manufacturers, drug wholesalers and
pharmacists, consumers, and physicians. 

Increasing drug expenditures have led to several changes in the
German regulations over the last 5 years.  For instance, in 1989, the
government adopted its first regulations limiting payments to drug
manufacturers.  The government also increased consumers' copayment
levels at that time.  Continuing increases in pharmaceutical spending
led to further constraints in 1993 whereby the government mandated
global budgets for pharmaceuticals, reduced prescription drug prices,
increased consumers' copayment levels, and tightened controls over
physicians' prescribing patterns. 


   OVERVIEW OF THE GERMAN HEALTH
   INSURANCE SYSTEM
-------------------------------------------------------- Appendix II:1

The foundation of the modern German health care system was laid by
Bismarck in 1883.  Today, this system guarantees universal health
care coverage to all German residents by requiring that working
persons, regardless of income, have health insurance and by providing
coverage for the unemployed.  Germany has a multipayer system
comprised of over 1,200 sickness funds.  Approximately 90 percent of
the total population is insured by statutory sickness funds and
almost all of the remainder obtain private health insurance. 

Health care, comprising about 8.5 percent of Germany's gross domestic
product in 1992, is financed primarily through government-mandated
contributions shared equally by workers and their employers.  The
required premium contribution operates much like a payroll tax--a
fixed percentage of the employee's gross compensation is deducted
from each paycheck and transferred directly to the private nonprofit
sickness funds.  The current contribution rate averages about 13.4
percent of wages up to the statutory income ceiling, with the
employer and employee each paying half of this premium.  The
contribution rates for individual sickness funds, which are revised
annually, range from 8.5 to 16 percent. 

Under the sickness fund system, premiums reflect the incomes of the
members as opposed to their actuarial risk.  About 60 percent of
health expenditures were derived from the sickness funds, about 21
percent from general taxation, about 7 percent from private
insurance, and about 11 percent from unreimbursed out-of-pocket
expenditures.  In 1992, total sickness fund expenditures were
Deutsche Mark 167.29 billion (or about $107.24 billion in 1992
dollars). 


      PHARMACEUTICAL COVERAGE IN
      THE GERMAN HEALTH INSURANCE
      SYSTEM
------------------------------------------------------ Appendix II:1.1

Sickness funds provide nearly complete coverage for pharmaceuticals. 
Consumers pay about $1.84 to $4.29 per drug prescribed, depending on
the cost of the drug, and the sickness funds pay the difference.\1

Almost 10,000 different drug products are available on the German
market.  These include both innovative drug products and generic
drugs.  Generic drugs are widely used in Germany, accounting for
about 27 percent of prescriptions and about 18 percent of sales in
1992.  In contrast to other European countries, there is little use
in Germany of parallel imports--identical drug products imported from
countries where drug prices are lower.\2


--------------------
\1 In addition, for drugs having a fixed reimbursable price,
consumers also pay the difference between the drug's price and the
fixed amount, if the drug exceeds that fixed amount. 

\2 However, as of mid-1993, parallel imports must be sold when they
are (1) legally on the market (i.e., importers have to obtain a
marketing permit from the government) and (2) at least 10 percent and
DM 1 cheaper than the drug produced in Germany. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG MANUFACTURERS
-------------------------------------------------------- Appendix II:2

Germany's principal spending control strategy aimed at drug
manufacturers is the reference price system.  This system does not
set drug prices; rather, it sets the reimbursement levels at which
sickness funds pay for each prescription drug (consumers pay the
amount by which the product prices exceed the reimbursement levels). 
RPS has two primary functions:  first, to lower the prices of drugs
by inducing price competition, and second, to encourage greater use
of generic drugs by making consumers pay a greater share of the cost
of higher-priced, brand-name drugs. 

In addition to RPS, Germany recently instituted two additional
policies to restrict sickness funds' pharmaceutical expenditures: 
global drug budgets, and an order for manufacturers to lower their
drug prices. 


      THE REFERENCE PRICE SYSTEM
------------------------------------------------------ Appendix II:2.1

Reference prices for outpatient prescription drugs are determined in
a two-step process.  First, a commission comprised of physician and
sickness fund representatives meets with scientists, manufacturers,
and pharmacists to group drugs into three classes.  The three
classes, established by the 1989 Health Care Reform Act, are the
following: 

  Class 1--drugs with the same active ingredients (that is, generic
     substitutes), which account for about 35 percent of Germany's
     total pharmaceutical market;

  Class 2--drugs with therapeutically comparable active ingredients
     (for example, different beta-blockers or H-2 antagonists); and

  Class 3--drugs with therapeutically comparable effects (e.g.,
     aspirin combinations).\3

The commission considers several specific factors when grouping drugs
into classes.  For example, differences in bioavailability of Class 1
drugs must be considered if they are relevant in the therapy.\4
Further, the grouping of drugs into Classes 2 and 3 must not restrict
the availability of any medically necessary alternative therapy. 

Second, representatives from the sickness funds propose a reference
price for each grouping of drugs.  A statistical methodology is used
to calculate what is, in effect, an average price of drugs within a
similar group; this average price varies with the strength and
package size of a drug product, within each group of drugs.  The
reference price is set below the price of the most expensive drug in
the group (typically the leading brand-name drug) and above the price
of the least expensive drug (typically a generic drug).  This price
is finalized after conferring with drug manufacturers and
pharmacists.  The reference price is adjusted at least annually,
taking into account inflation and other factors. 

All prescription drugs available in Germany are covered by RPS, with
the exception of (1) specified prescription drugs defined in the
German Drug Act (e.g., vaccines), (2) pharmacy-made drugs, and (3)
patented prescription drugs with a new active ingredient representing
a therapeutic improvement or having fewer side effects than existing
drugs.\5 \6


--------------------
\3 Together, drugs in Classes 2 and 3 account for about 14 percent of
Germany's pharmaceutical market. 

\4 Bioavailability refers to the speed and extent to which a
substance reaches the circulatory system. 

\5 The 1993 reforms expanded the number of patented drugs exempt from
RPS by clarifying what constitutes an innovative drug. 

\6 In Germany, many single-source products that lack comparable
products cannot be assigned reference prices.  Furthermore, other
products do not yet have reference prices because of the technical
difficulties in ascertaining which products have comparable
therapeutic ingredients or actions.  In 1993, the German government
simplified the way that drugs are put into comparable groups.  The
government hopes that this simplification will allow for the eventual
inclusion of 70 percent of drugs into the reference price system. 


      GLOBAL DRUG BUDGETS
------------------------------------------------------ Appendix II:2.2

As part of its 1993 health financing reforms, the German government
established an annual budget for drug spending by the sickness funds. 
The 1993 budget set the pharmaceutical drug budget for office-based
physicians at the 1991 expenditure level (approximately DM 24
billion, or about $14.7 billion in 1993 dollars).  Expenditures above
DM 24 million, up to DM 280 million (or about $175 million), would
have been offset by a reduction in the 1994 ambulatory care physician
budget.  Additional overruns between DM 280 million and DM 560
million (or about $343.6 million) would have been paid for by the
pharmaceutical industry through a reduction in drug prices.  However,
total drug spending for 1993 stayed within the budget. 

For the 1994 budget, regional physicians' associations were given two
options:  (1) a 1994 global budget set at the 1993 level, but with
cost overruns borne solely by reductions in the ambulatory care
budget (rather than having the reductions capped at DM 280 million,
as in 1993), or (2) negotiating a budget based on 1993 expenditures
for the region.  Most of the regional physicians' associations have
chosen the first option. 

The government does not consider global budgets to be an adequate
long-term solution to structural health care problems.  As such, the
fixed budget for prescription drugs is considered to be only a
temporary remedy for a period of 3 years to curb drug expenditures. 
It will remain in effect until the regional physicians' associations
and the sickness funds agree on indicative prescribing amounts. 


      MANDATED REDUCTIONS IN DRUG
      PRICES
------------------------------------------------------ Appendix II:2.3

Also under the 1993 reforms, Germany undertook two new efforts aimed
at drug manufacturers.  First, it required manufacturers to reduce
their non-RPS drug prices by 5 percent and reduce the prices of their
over-the-counter products 2 percent below the May 1992 level. 
Second, the reform requires a price freeze for these drugs during
1993 and 1994.  It was the first time that the German government had
taken such steps. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG WHOLESALERS AND
   RETAILERS
-------------------------------------------------------- Appendix II:3

In Germany, the government sets allowable markups for drug
wholesalers and retailers.  These margins are in inverse proportion
to drug prices; however, drug sellers have an incentive to increase
their revenues by selling higher-priced drugs because they receive
more revenues from selling a higher-priced drug than they do from
selling a lower-priced drug. 

Wholesale margins vary between 12 and 21 percent of the
manufacturer's price, with the rate decreasing as the manufacturer's
price increases.\7 Retail margins vary between 30 and 68 percent of
the wholesale price (exclusive of the 15 percent value-added tax),
with the rate decreasing as the wholesale price increases.  In
addition, German pharmacies are required by law to give the sickness
funds a 5-percent discount off the drug's retail price.  Table II.1
provides an example of the pricing structure for a drug, starting
with the manufacturer's price and ending with its effective retail
price. 



                          Table II.1
           
           A Typical Example of Pharmacy Pricing in
                           Germany

Item/action                                    Amount
-------------  ------------------------------  -------------
               Manufacturer's price            DM 15.00

+              Wholesaler's markup (18         DM 2.70
               percent)

=              Wholesaler's selling price      DM 17.70

+              Pharmacy's markup (48 percent)  DM 8.50

=              Net pharmacy retail price       DM 26.20

+              Value-added tax (15 percent)    DM 3.93

=              Gross pharmacy retail price     DM 30.13

-              5-percent discount to sickness  DM 1.51
               funds

=              Effective retail price          DM 28.62
------------------------------------------------------------
Source:  Bundesvereinigung Deutscher Apothekerverbï¿½nde (ABDA). 


--------------------
\7 Wholesalers may reduce their margins to some extent; however, the
savings must, by law, be passed on to pharmacists, not to consumers
(that is, retail prices may not be reduced). 


   SPENDING CONTROL STRATEGIES
   AIMED AT CONSUMERS
-------------------------------------------------------- Appendix II:4

The German government, at different times, instituted varying levels
of consumer cost sharing.  These policies have been implemented for
several reasons:  to shift some of the financial burden of
pharmaceuticals from the sickness funds to consumers; to decrease
utilization by making consumers aware of the costs of drugs; and to
encourage consumers to choose less expensive drugs (such as generic
substitutes) by having lower copayments on less expensive drugs. 

The 1993 health reforms require consumers to make copayments on all
drugs, regardless of whether the drugs have a reference price
(previously, copayments were required only for drugs not under RPS or
for drugs where the sales price exceeded the reference price).  Since
January 1994, copayment levels have been linked to the quantity of
drugs purchased.\8 Consumers are required to pay DM 3 (or about
$1.84) for small quantities, DM 5 (or about $3.07) for medium
quantities, and DM 7 (or about $4.29) for large quantities.\9 Upper
limits on consumer cost sharing are based on gross income and family
size.\10 Exemptions are given to drugs administered during pregnancy
or directly related to pregnancy, children under 18 years old, and
persons with low income.  In addition, consumers pay the amount by
which a product's price exceeds the reference price. 


--------------------
\8 In 1993, copayment levels were linked to the prices of drugs
purchased. 

\9 The determination of what is a small, medium, or large quantity
depends on the illness and the dosage.  For example, for diabetic
medications, a 20-tablet prescription is considered a small quantity,
while for sedatives, a 20-tablet prescription is a medium
prescription. 

\10 The upper limit is based on the following formula:  limit = (AGI
- deduction) x percentage, where AGI is the annual gross income;
deduction equals DM 0 for family of 1, DM 7,056 for family of 2, DM
4,700 for each additional family member; and percentage equals 2
percent for AGI below DM 68,400 or 4 percent for AGI above DM 68,400. 


   SPENDING CONTROL STRATEGIES
   AIMED AT PHYSICIANS
-------------------------------------------------------- Appendix II:5

In an effort to spread the burden of rising pharmaceutical costs, the
government has tried to persuade physicians to prescribe more
cost-effectively through the use of transparency lists and price
lists; however, until recently, it had no direct means of persuading
physicians to prescribe more cost-effectively.  The 1993 reforms gave
the government more leverage on affecting physician prescribing
habits by instituting global drug budgets and through the future
development of a streamlined list of drugs eligible for
reimbursement. 


      PHYSICIAN DRUG BUDGETS AND
      INCREASED MONITORING OF
      PHYSICIANS' PRESCRIBING
      PATTERNS
------------------------------------------------------ Appendix II:5.1

Office-based physicians in Germany bear financial responsibility for
drug spending levels that exceed the annual budget.  As discussed
above, the 1993 pharmaceutical drug budget for office-based
physicians was frozen at approximately DM 24 billion (or about $14.7
billion).  Expenditures above DM 24 billion, up to DM 280 million (or
about $175 million), would have been offset by a reduction in the
ambulatory care physician budget; however, total drug spending for
1993 stayed within the budget.  In 1994, physicians will bear sole
financial responsibility for exceeding the 1994 drug budget. 

As an additional measure to promote drug spending control, regional
associations of office-based physicians and sickness funds are
responsible for monitoring physicians' prescribing patterns and for
establishing ceilings for the volume of drugs prescribed.  Under the
1993 reforms, the regional associations of office-based physicians
and sickness funds are now required to conduct an inquiry if
physicians exceed standard prescribing amounts for the region and
their specialty by 15 percent and to seek redress if this amount is
exceeded by 25 percent.\11

Evidence suggests that these two actions have significantly affected
physicians' prescribing patterns.  In the first half of 1993, the
number of drugs prescribed was about 17 percent below the 1992 level,
and total sickness fund prescription drug expenditures declined by
about 22 percent compared to the same period in 1992.  There was a
greater decrease in the use of drugs for which the therapeutic effect
is less widely accepted than for more therapeutically meaningful
drugs.  For example, the prescriptions for vitamins, mineral
preparations, and vein therapeutics fell by about 30 percent, while
the decrease in prescriptions for antibiotics fell by about 5
percent, for beta-blockers by about 10 percent, and for anti-diabetes
drugs by less than 1 percent. 

Several reasons have been suggested for the drop in drug sales. 
First, physicians substituted cheaper generic drugs for more
expensive, brand-name drugs.  As a result, sales of the cheapest
generic drugs have increased in some cases by as much as 250 percent. 
Second, physicians increasingly prescribed older products instead of
newer, more innovative drugs.  Third, patients--especially those with
chronic illnesses (for example, diabetes)--obtained their
prescriptions in December 1992 (just before the new regulations took
effect) and thus did not need to acquire their drugs in the first few
months of 1993.  Fourth, physicians seem less willing to prescribe
drugs with doubtful efficacy (such as anti-varicosis drugs) or for
conditions that can be treated without drugs (such as diets for
obesity).  Finally, uncertainty and misinformation about how to
manage within a drug budget caused physicians to curtail their
prescribing more often than necessary. 


--------------------
\11 Until the beginning of 1993, these regional associations notified
physicians surpassing their ceiling by more than 20 percent and those
exceeding their ceiling by 40 percent were required to justify their
actions.  However, few physicians were penalized for
over-prescribing, even though approximately 10 percent of the
physicians surpassed their ceilings by more than 20 percent annually. 


      ESTABLISHMENT OF A
      REIMBURSABLE DRUG LIST
------------------------------------------------------ Appendix II:5.2

The 1993 reforms also call for increased use of drug lists to define
eligibility for reimbursement by the sickness funds.  The German
government will establish a new Pharmaceutical Institute to develop a
detailed list of the drugs for which the sickness funds will provide
reimbursement after 1995.  This list will replace the list of
nonreimbursable drugs currently in use.  This list covers medicines
that have more than three active ingredients or those for which the
effect of the active ingredients has not been therapeutically proven. 
Currently, about 2,200 drugs are on the nonreimbursable list,
accounting in 1992 for DM 140 million (or about $89 million in 1992
dollars) in sales. 


SWEDEN'S DRUG SPENDING CONTROL
POLICIES
========================================================= Appendix III

As a result of Sweden's pharmaceutical spending control strategies,
drug prices in Sweden are at about the European average.  Through its
regulation over the health insurance system, the Swedish government
has imposed a variety of controls on outpatient prescription drug
prices that apply to various participants in the pharmaceutical
market:  drug manufacturers, drug wholesalers and pharmacists,
consumers, and physicians. 

In 1993, the Swedish government implemented important changes to the
drug pricing regulations designed to reduce pharmaceutical costs,
provide incentives for R&D, and comply with the European Community's
directive on transparency of pharmaceutical pricing.\1 These reforms
reassigned responsibility for negotiating the prices of reimbursable
drugs, introduced RPS for reimbursing brand-name drugs where
equivalent generic drugs exist, and increased the patient copayment
level for outpatient prescription drugs.  The reforms are expected to
save the government up to SEK 1 billion (about $133.69 million):  SEK
600 million (or about $80.21 million) from increasing the patient
copayment and SEK 400 million (or about $53.48 million) from the
introduction of RPS.\2


--------------------
\1 The European Community's transparency directive requires its
member countries to publicly disclose the rules governing pricing of
prescription drugs.  It does not interfere with the right of
countries to control prices or reimbursement by any method they
choose, provided the method used is based on objective and verifiable
criteria and does not discriminate between foreign and domestic drug
manufacturers.  Although Sweden is not a European Community member,
it is seeking membership. 

\2 All dollar figures cited in this appendix were calculated using
the average exchange rate for the first quarter of 1993. 


   OVERVIEW OF THE SWEDISH HEALTH
   INSURANCE SYSTEM
------------------------------------------------------- Appendix III:1

The fundamental principle of Sweden's social welfare system is that
all citizens are entitled to good health and equal access to health
care, regardless of where they live and their social and economic
circumstances.  In line with this principle, health care is seen as a
public sector responsibility supported by a national health insurance
system and by other social welfare services.  In 1955, Sweden
expanded social insurance benefits by establishing a comprehensive
national health insurance system that provides health care, sickness,
and maternity and parental benefits to Swedish citizens and alien
residents.  Today, health insurance pays for part of the cost of
outpatient medical and dental care, and most of the cost of
prescription drugs and hospital care, in addition to other services. 

The National Social Insurance Board is the government agency that
oversees these benefits in Sweden.  It centrally administers and
regulates the activities of 26 regional offices, which manage local
social insurance programs. 

In 1989, health care comprised about 8.6 percent of the gross
domestic product.  The system is primarily financed through employer
contributions, with additional funding coming from state grants. 


      PHARMACEUTICAL COVERAGE IN
      THE SWEDISH HEALTH INSURANCE
      SYSTEM
----------------------------------------------------- Appendix III:1.1

The national health insurance system provides nearly complete
coverage for outpatient prescription drugs.  NSIB sets the prices at
which eligible drugs will be reimbursed.  NSIB reimburses at these
prices, less the patient copayment rate (if such rate applies).  NSIB
pays the balance of the cost directly to Apoteksbolaget (the National
Corporation of Swedish Pharmacies), which reimburses the
pharmacies.\3

Over 3,000 different drug products are available on the Swedish
market.  Generic drugs are currently used in Sweden and the
establishment of Sweden's reimbursement rate setting system is
expected to further increase generic drugs' share of the market.\4 In
contrast to other European countries, there is no use of parallel
imports--identical products imported from other countries where drug
prices are lower--because this practice is forbidden under Swedish
law.\5


--------------------
\3 Apoteksbolaget's legal foundation is based on an agreement with
the government, which owns two-thirds of the shares. 

\4 Currently, generic drugs account for about 13 percent of the
pharmaceutical market in Sweden.  Some officials expect that generic
drugs' share of the market could increase to 25 percent under the
reforms implemented in 1993 and that the government can save an
additional SEK 800 million (or about $107 million) per year by using
more generic drugs. 

\5 Parallel imports and exports may be permitted after 1994, when an
agreement between the European Community and the European Free Trade
Association is ratified. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG MANUFACTURERS
------------------------------------------------------- Appendix III:2

Since January 1993, Sweden's principal strategy for controlling
prescription drug prices has been through a reimbursement system that
sets the prices that the national health insurance system will pay
for prescription drugs.\6 The government determines the reimbursement
level in one of two ways:  (1) through RPS, which sets the
reimbursement price for brand-name drugs where equivalent generic
drugs exist, or (2) by directly setting the reimbursement price for
those drugs not under RPS. 


--------------------
\6 Prior to implementation of RPS, Sweden directly set drug prices. 
The agency responsible for setting prices, Apoteksbolaget, negotiated
with manufacturers the wholesale prices of all pharmaceutical drugs
sold in Sweden--prescribed and over-the-counter--with the aim of
setting Sweden's drug prices no more than 5 to 10 percent higher than
the average of other European countries.  These price negotiations
were required prior to registering and approving drugs for marketing. 
When Sweden implemented RPS, it transferred the price setting
responsibility from Apoteksbolaget to NSIB (for only those drugs
subject to reimbursement by the government). 


      THE REFERENCE PRICE SYSTEM
----------------------------------------------------- Appendix III:2.1

Reference prices are set for nonpatented reimbursable drugs that have
at least one generic competitor on the Swedish market.\7 Under RPS,
the reimbursable rate for a prescription drug--known as the reference
price--is set at the price of the lowest generic drug equivalent plus
10 percent.  Manufacturers are free to set drug prices exceeding the
reference price; however, the amount of this excess must be paid by
consumers. 

NSIB publishes its reference price list every 3 months. 
Manufacturers objecting to a reference price can appeal to the
government for a price change, but these appeals must be filed within
3 weeks after the publication of the reference price.\8


--------------------
\7 To be included under RPS, drugs must meet several other criteria,
including having at least 20 percent of the drug's sales for
outpatient use and being on the market for more than 6 months. 

\8 During 1993, only one appeal was filed.  This appeal did not
concern the reference price per se but whether a certain product
should appear on the price list. 


      REIMBURSEMENT PRICE FOR
      DRUGS NOT UNDER THE RPS
----------------------------------------------------- Appendix III:2.2

Manufacturers of drugs not under RPS--new and patent-protected
brand-name drugs and over-the-counter products--must negotiate and
agree on a price with NSIB if they want to be included under the
reimbursement system; otherwise, the drug will not be eligible for
reimbursement.  Manufacturers choosing not to seek reimbursement for
drugs can price these drugs freely. 

NSIB conducts negotiations using the same criteria formerly used by
Apoteksbolaget in setting prices for all reimbursable drugs.\9 NSIB
emphasizes a drug's therapeutic value and its estimated contribution
in reducing overall health care cost.  It also considers (1) the
price of similar products sold in other countries; (2) the price of
the same product sold in other countries; and (3) the price of the
product in its home market.\10 \11 Finally, NSIB considers a drug's
projected sales volumes, R&D costs, manufacturing costs, and the
manufacturer's legal fees. 


--------------------
\9 Until April 1993, Apoteksbolaget acted as a consultant to NSIB on
price negotiation matters.  At that time, NSIB added personnel to
perform the price negotiations.  NSIB strives, as did Apoteksbolaget,
to set a reasonable wholesale price for each drug. 

\10 Apoteksbolaget used the same criteria to set drug prices for
domestic and international firms, but ensured that foreign firms did
not receive higher prices for products sold in Sweden than in the
firms' home markets.  Further, Apoteksbolaget ensured that Swedish
firms selling outside the country received high prices for their
drugs in Sweden. 

\11 Apoteksbolaget obtained information on price differentials from
11 Western European countries--agreed upon by pharmaceutical
manufacturers in Sweden as representing a fair set of
comparisons--from agencies within those countries and through
discussion with manufacturers in Sweden about drug prices in foreign
markets.  In addition, Apoteksbolaget determined mean and median
price differentials between Sweden and these other countries and
tracked price increases and decreases from year to year. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG WHOLESALERS AND
   RETAILERS
------------------------------------------------------- Appendix III:3

In Sweden, retail margins are controlled by the government. 
Wholesale margins are not regulated but result from negotiations
between manufacturers and wholesalers.  The manufacturers' prices to
the wholesalers account for 68.2 percent of the pharmacy selling
price.  Wholesalers add 4.2 percent to the manufacturers' prices
(their share accounts for 2.8 percent of the pharmacy selling price). 
Pharmacies add 41 percent to their purchasing price, which gives them
a margin of 29 percent.\12


--------------------
\12 The figures represent averages for the sale of outpatient,
prescribed pharmaceuticals. 


   SPENDING CONTROL STRATEGIES
   AIMED AT CONSUMERS
------------------------------------------------------- Appendix III:4

In an effort to reduce health care costs to the government, patient
copayment rules and levels for outpatient prescription drugs were
changed.  Effective July 1992, the copayment level rose to SEK 120
(or about $16.04) for the first drug on the prescription and SEK 10
(or about $1.34) for each additional drug.\13 Effective January 1993,
the upper spending limit per 12-month period was raised from SEK
1,500 (or about $200) to SEK 1,600 (or about $214).\14 As under the
previous system, patients only paid the actual drug price if drugs
were less than the patient copayment. 

For drugs included in RPS, which are priced below the copayment limit
and at, below, or above the reference price, a patient's
out-of-pocket expense equals the actual drug price.  For drugs
included in RPS, which are priced above the copayment limit and above
the reference price, a patient's out-of-pocket expense equals the
difference between the price of the drug and the reference price plus
the copayment limit.  For reimbursable drugs not included in RPS,
patients pay the actual drug price of drugs priced below the
copayment limit and pay the copayment limit for drugs priced above
the copayment limit.  For nonreimbursable drugs, patients pay the
full drug cost.  For certain drugs, which treat 32 chronic illnesses
or disorders (e.g., insulin for diabetics, drugs for epilepsy), there
is no out-of-pocket expense for the patients.\15 \16


--------------------
\13 Until July 1992, the patient copayment for prescription drugs (up
to 10 drugs at a time) was set at SEK 90 (or about $12.03). 

\14 Once patients reach their upper spending limit on medical
treatments and/or prescription drugs, all subsequent treatments/drugs
are provided free of charge. 

\15 Brand-name drugs having a lower-priced generic competitor cannot
be provided free of charge.  However, exceptions can be made for
special patients on established therapies. 

\16 The drugs treating the 32 chronic illnesses or disorders account
for 20 percent of the total cost of prescription drugs in Sweden,
reimbursable drugs account for 70 percent, and free aids (for
example, testing kits for diabetics) and foods (for example,
lactose-free dairy products) account for 10 percent. 


   SPENDING CONTROL STRATEGIES
   AIMED AT PHYSICIANS
------------------------------------------------------- Appendix III:5

Prior to January 1993, Sweden did not have in place any cost control
strategies aimed at physicians.  Now, by law, physicians must inform
patients about lower-cost generic prescription drugs.  Quarterly
information sheets issued jointly by the Medical Products Agency and
NSIB provide physicians with data on lower-cost generic alternatives. 
Prescribing doctors can choose not to substitute a generic drug for a
brand-name drug for medical reasons; if so, the prescribing doctor
must inform the patient and provide a written notice to the
pharmacist to this effect.\17


--------------------
\17 Substitution by the pharmacist is permitted in cases where the
doctor is unavailable and a delay can be of serious detriment to the
patient. 


      USE OF DRUG LISTS
----------------------------------------------------- Appendix III:5.1

Except for six drugs, all prescription drugs available in Sweden can
be subject to government reimbursement at a fixed retail price, less
the patient copayment rate (if such rate applies).  In addition,
over-the-counter drugs can be subject to reimbursement if the drug
has a generic competitor and if the manufacturer negotiates the price
of the drug with NSIB. 

A proposed reform would have changed the reimbursement rules so that
only prescription drugs--and not any over-the-counter drugs--were
covered under the reimbursement system; however, this proposal was
dropped because of strong opposition from consumers.  According to a
government official, there are no plans to develop a negative list in
the near future. 


THE UNITED KINGDOM'S DRUG SPENDING
CONTROL POLICIES
========================================================== Appendix IV

Pharmaceutical spending control strategies in the United Kingdom are
designed to restrain the growth of drug expenditures within the
U.K.'s health care system while encouraging pharmaceutical industry
investment and promoting pharmaceutical R&D.  This dual purpose has
led to a policy that gives manufacturers considerable pricing freedom
but ties allowable profits to a firm's capital in the United Kingdom,
thereby awarding higher profits to firms with more investment in the
country.\1 The U.K.  government also places strong emphasis on
policies aimed at physicians, with the intent of encouraging the
rational prescribing of drugs.  In addition to these strategies, the
U.K.  government has imposed a variety of additional controls that
apply to various other participants in the pharmaceutical market,
including drug wholesalers and pharmacists, and consumers. 

The government periodically reevaluates these strategies in an
attempt to ensure that the range of strategies adopted balance each
other and produce a coherent overall system.  Among the most recent
changes, adopted in an effort to restrain rising drug expenditures,
is a 2.5-percent price cut on all drug products, followed by a 3-year
price freeze.  In addition, the government has announced that it will
further limit the number of drugs NHS can offer.\2


--------------------
\1 Specifically, allowable profits are tied to that portion of the
firm's capital in the United Kingdom that is devoted to sales to the
National Health Service. 

\2 NHS' drug expenditures have risen about 12 percent per year during
the last 2 years. 


   OVERVIEW OF THE UNITED
   KINGDOM'S HEALTH INSURANCE
   SYSTEM
-------------------------------------------------------- Appendix IV:1

NHS, which falls under the Department of Health, operates the United
Kingdom's health care policy.  NHS offers comprehensive medical
services, including basic primary, hospital, and community care
services to all residents of the United Kingdom, at little or no
charge.  NHS also reimburses the cost of most prescribed drugs. 
While some people have private insurance, over 95 percent of the
patients treated in the United Kingdom receive their treatment from
NHS. 

Health care, roughly 6.6 percent of the gross domestic product in
1991, is largely financed from general tax revenues.  Additional
funds are derived from payroll and local taxes; charges for
prescriptions and other services such as dental treatment; and
payments by private patients in public hospitals.  In 1991, total NHS
expenditures, which have slowly risen in real terms, were 28.2
billion pounds (ï¿½) (or about $49.5 billion in 1991 dollars).\3 Of
this, ï¿½18.1 billion (or about $31.8 billion) went for hospital and
community care services and about ï¿½7.7 billion (or about $13.5
billion) for primary care (including about ï¿½3.5 billion (or about
$6.1 billion) for pharmaceuticals covered by NHS).\4


--------------------
\3 Except where otherwise noted, all dollar figures cited in this
appendix were calculated using the average exchange rate for the
first quarter of 1993. 

\4 U.K.  pharmacies carry two types of pharmaceuticals:  (1)
prescription drugs and (2) over-the-counter drugs.  This latter
category also includes some drugs with "pharmacy only" status; that
is, drugs that do not require a physician's prescription but require
a pharmacist's permission to be purchased.  The government provides
reimbursement for both prescription-only drugs and over-the-counter
drugs unless they have been specifically prohibited from NHS use
under the Selected List Scheme (which limits NHS reimbursement to
specific drugs). 


      PHARMACEUTICAL COVERAGE IN
      THE UNITED KINGDOM'S HEALTH
      INSURANCE SYSTEM
------------------------------------------------------ Appendix IV:1.1

NHS provides nearly complete coverage for prescription drugs in the
United Kingdom.  It is, in effect, the principal purchaser of
pharmaceuticals because it buys more than 90 percent of all
prescription drugs in the United Kingdom.  Consumers simply pay a
flat copayment of ï¿½4.25 (or about $6.25) to the pharmacist,
regardless of the price of the drug, and the government pays the
difference.  However, many consumers are exempt from copayment,
including the poor, the elderly, children under 16, expectant and
nursing mothers, and people suffering from certain long-term
illnesses.  As a result of these exemptions, only about 20 percent of
prescribed items were dispensed with a patient copayment in 1991. 

Over 4,000 different drug products are available on the U.K. 
market.\5 These include both innovative drug products and generic
drugs.\6 In recent years, the government has promoted generic
prescribing, which has led to an increase in the proportion of
generic drugs that were prescribed and dispensed.  Between 1987 and
1989, the percentage of generic drugs dispensed increased from 29 to
37 percent of all prescriptions. 


--------------------
\5 Currently, parallel imports--identical products imported from
other countries where drug prices are lower--make up about 8 percent
of the pharmaceutical market in the United Kingdom.  However, U.K. 
pharmacists are moving toward greater use of these products,
primarily because of the higher profits derived from this practice. 

\6 There are over 40 generic manufacturers in the United Kingdom. 
However, one manufacturer has 50 percent of the market, and two
others dominate the other half of the market. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG MANUFACTURERS
-------------------------------------------------------- Appendix IV:2

The principal strategy for controlling brand-name drug prices in the
United Kingdom is a profit control measure known as the
Pharmaceutical Price Regulation Scheme (PPRS).  Aimed at drug
manufacturers, PPRS is designed both to ensure reasonable drug prices
and to promote a strong and profitable pharmaceutical industry.  The
U.K.  government has a separate strategy for encouraging price
competition among generic drug prices. 


      THE EVOLUTION OF PPRS
------------------------------------------------------ Appendix IV:2.1

PPRS has evolved from a series of voluntary agreements between the
Department of Health and the pharmaceutical industry.\7 The first
agreement, dating back to 1957, was actually a price regulation
scheme.  By 1969, it had emerged as a scheme focusing on overall
profits rather than individual drug prices.\8 Subsequent versions
essentially retained the scheme established in the 1969 agreement.\9
The most recent PPRS agreement took effect in October 1993, replacing
the 1986 agreement. 


--------------------
\7 PPRS is not governed by specific law.  Rather, the agreement
provides flexibility so that arrangements with individual
manufacturers reflect the varying commercial practices within the
industry. 

\8 In the 1969 version, individual manufacturers' profits and costs
became the focus of the price regulatory arrangements.  This version
required manufacturers to produce an annual financial return showing
sales and their associated costs. 

\9 The 1978 version renamed the scheme "Pharmaceutical Price
Regulation Scheme." In 1986, generic drugs were removed from the
scheme, leaving only brand-name drugs under PPRS. 


      PPRS REGULATIONS ON PROFITS
      AND PRICE INCREASES
------------------------------------------------------ Appendix IV:2.2

PPRS is an indirect means of controlling brand-name drug prices by
regulating the overall profitability of manufacturers from their
pharmaceutical sales to NHS.\10 Under PPRS, manufacturers are free to
set prices for new drugs, taking into account the impact of their
pricing decisions on their overall profit targets, but are required
to obtain government approval before increasing prices of existing
drugs. 

Most pharmaceutical manufacturers are subject to profit controls
through PPRS.\11 A majority of these manufacturers' profits are
limited to 17 to 21 percent of capital invested in the United Kingdom
devoted to sales to NHS, with the exact rate negotiated between the
manufacturer and NHS.  Other manufacturers--those with relatively low
levels of capital invested in the United Kingdom--have allowable
profits set at 4.5 percent of their total sales to NHS.\12 In
addition, manufacturers may be allowed to keep additional profits--up
to 25 percent over their target level--if, for example, these higher
profits are attributable to new products or increased
efficiencies.\13 Manufacturers exceeding their approved targets
(including the permitted excess) must repay the government or agree
to lower the prices of their existing drugs. 

A three-step process is used to enforce PPRS.  First, each
manufacturer is required to provide NHS with sales, investment, and
cost data, which allow the government to determine the manufacturer's
profits.\14 Second, the government conducts an assessment of the
manufacturer's capital investments in the country.\15 Third, the
government assesses the manufacturer's costs, including the cost of
goods, distribution, promotion, and R&D.\16

Even under the profit control scheme, drug manufacturers are still
subject to drug pricing regulations.  While manufacturers freely set
prices when first introducing new drugs--so long as total profits do
not exceed the target level--they cannot increase drug prices without
prior government approval.  Only manufacturers not achieving their
basic target profits may apply for price increases.  In any one year,
the government receives applications from between 15 to 20 companies
for price increases.  Increases granted in accordance with PPRS are
generally below the rate of inflation and only enough to bring
manufacturers up to their targets.  In recent years, the drug price
increases have increased the drug bill by less than 2 percent. 

The most recent PPRS agreement imposed even tighter price controls on
manufacturers than had been in place previously.  This agreement
required firms to reduce prices on all products by 2.5 percent, and
ordered a 3-year price freeze as of October 1, 1993.\17 It also
established a threshold for price increases after that time,
requiring company profits to fall to less than 75 percent of the
company's target level before price increases could be granted. 


--------------------
\10 PPRS regulations apply to all firms with sales to NHS of over
ï¿½0.5 million (or about $740,000) per year. 

\11 As of October 1993, only manufacturers with sales of ï¿½20 million
(about $29 million) or more are required to provide detailed
financial information.  There are currently about 43 such
manufacturers.  Under the previous version of the PPRS agreement,
manufacturers with sales of ï¿½4 million (about $5.9 million) or more
were required to provide detailed financial information.  Less
detailed controls apply to the 46 manufacturers with annual sales
between ï¿½1 million (about $1.5 million) and ï¿½20.0 million. 

\12 A firm's allowable profit is based on sales, rather than on
capital invested in the United Kingdom, if its annual sales are
greater than 3.75 times its U.K.  capital investment.  Of the 43
manufacturers with sales exceeding ï¿½20 million, 7 are assessed on a
return-on-sales basis. 

\13 Prior to the current version of the PPRS agreement, manufacturers
were allowed to keep additional profits--up to 50 percent over the
target level.  On an annual basis, about 30 manufacturers were
required to justify having profits that exceed their target by up to
50 percent.  Under the current version, there is no longer any
requirement to justify retaining the additional profit (i.e., profits
up to 25 percent over their target level). 

\14 Manufacturers are expected to submit these reports within 6
months after the end of the manufacturer's financial year. 

\15 This is calculated by taking the fixed assets (land, buildings,
and manufacturing plants), adding the current assets (cash, debt, and
stock), and deducting the current liabilities (creditors and current
taxation).  If fixed assets are used to produce drugs that are both
sold to NHS and exported, the proportion of those assets that is
taken into account in determining a manufacturer's profit is based on
the ratio of NHS sales to export sales.  (Under the current version
of the PPRS agreement, the government will change its PPRS
calculation method, effective October 1996.  It will recognize all
fixed costs associated with manufacturing in the United Kingdom, to
avoid any disincentive to exports.) Borrowings that are part of a
manufacturer's normal annual trading activities are included in the
calculations, while borrowings of a long-term or structural nature
are excluded. 

\16 The maximum allowable amount for promotion is based on a mixture
of a flat rate allowance for all manufacturers, a percentage of
sales, and the number of significant drugs the company has on the
market.  The limit for the industry as a whole is about 8.2 percent
of sales, although smaller manufacturers are allowed to deduct up to
18 percent of sales as marketing expenses. 

\17 Companies are offered two alternatives in lieu of the global
price cut.  The first alternative allows a company to decrease prices
by different amounts for different products (that is, cutting prices
on some products by more than 2.5 percent, and others less than 2.5
percent), so long as the overall effect on receipts is the same as it
would be under the global price reduction.  The second option is to
leave prices unaltered but to make cash payments to the government in
lieu of the price reduction.  NHS expects most companies to institute
a global price cut rather than take any of these alternatives. 


      REIMBURSEMENT PRICE FOR
      GENERIC DRUGS
------------------------------------------------------ Appendix IV:2.3

As with brand-name drugs, manufacturers freely set generic drug
prices; however, the government sets out the level at which
pharmacists are reimbursed for the cost of generic drugs sold to
customers.\18 This price is calculated in one of three ways.  Where
there is considerable competition in the market, an average weighted
amount is assessed on the basis of the list prices of the main
manufacturers.  Where there is limited competition in the market, an
average weighted amount is assessed on the basis of the prices
offered by the main wholesalers.  Where there is effectively one
product, that price becomes the tariff price.  Therefore, the price
listed in the Drug Tariff reflects the level of prices set
competitively in the market. 


--------------------
\18 In a monthly government publication called the "Drug Tariff," the
government lists the prices at which it will reimburse pharmacists
for drugs dispensed. 


   SPENDING CONTROL STRATEGIES
   AIMED AT DRUG WHOLESALERS AND
   RETAILERS
-------------------------------------------------------- Appendix IV:3

Wholesale margins in the United Kingdom are regulated by the
government.  These margins are set at 12.5 percent of the retail (or
list) price, which is the same in all pharmacies.\19

In practice, there is no fixed retail margin in the United Kingdom
for drugs dispensed under NHS.  Retail margins vary according to what
wholesalers are prepared to offer pharmacists in particular
circumstances from within their 12.5-percent margins.  In addition to
the retail margin, pharmacists also receive a dispensing fee for each
prescription.\20

Pharmacists are also subject to a reduction in their total payment. 
The basis for the reduction is a government attempt to recapture
volume discounts that it believes many pharmacists receive from
wholesalers and parallel importers.  The reduction is also meant to
ensure that the amounts pharmacists are reimbursed by the government
reflect the discounts given in the market and to encourage
pharmacists to seek these discounts.  The reduction is applied at the
same rate to all pharmacists, being linked to the value of the
prescriptions dispensed each month rather than to the actual
discounts received.  The amount of these reductions is determined in
periodic surveys conducted by the government. 


--------------------
\19 The wholesaler margin consists of a 10-percent discount off the
retail price of the drug plus an additional 2.5 percent from the
manufacturer for prompt payment. 

\20 The dispensing fee is negotiated annually and is unrelated to the
price of the prescribed drug.  Currently, pharmacists receive ï¿½1.512
(or about $2.23) for each of the first 1,500 prescriptions dispensed
each month, and ï¿½0.715 (or about $1.06) for each prescription
dispensed thereafter for the rest of the month, averaging ï¿½1.08 (or
about $1.59) in May 1992.  Prior to April 1993, pharmacists also
received a payment of 2.5 percent of the cost of the drug dispensed. 


   SPENDING CONTROL STRATEGIES
   AIMED AT CONSUMERS
-------------------------------------------------------- Appendix IV:4

Patient cost sharing is the primary spending control strategy aimed
at consumers, but a large fraction of the population is exempt. 
Among those exempt from copayment are children under 16 (under 19 if
they are full-time students); the elderly; poor people; expectant and
nursing mothers; people with certain long-term illnesses; and war or
service pensioners who require prescription drugs for the accepted
disablement.  Together, these groups account for about 80 percent of
the prescription drugs dispensed in the United Kingdom in 1991. 

Consumers who are not otherwise exempt are required to make a
copayment of ï¿½4.25 (or about $6.25) for each prescription drug.\21
Patients requiring a great deal of medication, but not exempt from
copayment, can purchase a "season ticket," paying ï¿½60.00 (or about
$88.00) per year for an unlimited number of prescribed items, rather
than ï¿½4.25 for each item.\22


--------------------
\21 Copayments have increased gradually over the last several years,
from ï¿½0.45 (or about $0.66) in 1979 to the current ï¿½4.25. 

\22 Where the price of the drug is less than the copayment, patients
still pay the ï¿½4.25 copayment.  Approximately 53 percent of the
prescribed items dispensed cost less than the copayment. 


   SPENDING CONTROL STRATEGIES
   AIMED AT PHYSICIANS
-------------------------------------------------------- Appendix IV:5

Over the last several years, the United Kingdom has implemented
various spending control strategies aimed at physicians in an attempt
to influence their prescribing patterns.  The government first
introduced the Selected List Scheme, which limits NHS reimbursement
to specific drugs.  It then introduced the Prescribing Analyses and
Cost (PACT) system, which provides physicians with information on
their prescribing patterns, followed by the Indicative Prescribing
Scheme, which sets prescribing targets for physicians.  Most
recently, the government introduced the General Medical Practitioner
Fundholding Scheme, which provides physicians with a financial
incentive to prescribe effectively.  In addition, the government has
implemented several educational measures to further improve the
quality and cost-effectiveness of prescribing. 


      SELECTED LIST SCHEME
------------------------------------------------------ Appendix IV:5.1

The Selected List Scheme was introduced in 1985 to control NHS drug
expenditures and to promote the use of generic drugs.  The list
limits the number of drugs NHS can offer in certain categories,
including analgesics for mild to moderate pain, indigestion remedies,
laxatives, vitamins, cough and cold remedies, and benzodiazepine
sedatives and tranquilizers.  According to NHS officials, in 1986,
the Scheme saved ï¿½75 million (or about $110.3 million in 1986
dollars).\23

To help control rising expenditures on drugs, the government is
planning to expand the Selected List Scheme into 10 additional
therapeutic categories:  anti-diarrheals, allergic disorders,
hypnotic and anxiolytics, appetite suppressants, vaginal and vulval
conditions, contraceptives, anemia, topical antirheumatics, ear and
nose conditions, and skin conditions.  The government has not said
that NHS will not reimburse all products in these categories. 
Rather, an independent committee of experts will review all products
in these categories to determine which ones are cost-effective and
meet genuine patient needs.  Final decisions on the individual
products in all 10 categories will probably be made before mid-1994. 


--------------------
\23 Savings figures are only available for the first year the scheme
was in effect. 


      THE PRESCRIBING ANALYSES AND
      COST SYSTEM
------------------------------------------------------ Appendix IV:5.2

PACT, introduced in August 1988, enables physicians to compare their
prescribing patterns against the patterns of other physicians.  PACT
provides physicians with information about their prescribing patterns
within their practices.  It also allows physicians to compare their
prescribing patterns with local and national averages, and with the
average for six major therapeutic groups in the country.  Further,
PACT provides additional information if physicians exceed the
national averages by a set percentage.  Government officials told us
that in 1989, PACT is believed to have saved about ï¿½80 million (or
about $131.2 million in 1989 dollars). 


      INDICATIVE PRESCRIBING
      SCHEME
------------------------------------------------------ Appendix IV:5.3

IPS, introduced in 1991, is an attempt to improve the quality and
cost-effectiveness of prescribing by setting prescribing targets on
pharmaceutical expenditures.  A target is set at the regional level
and broken down to the district and individual physician level.\24
The use of indicative monetary targets for prescribing is intended to
make physicians more aware of the monetary implications of their
prescribing choices, and to encourage them to use less expensive
medications when feasible. 

Indicative budget targets are reflective of the individual
circumstances of each individual physician, depending on the
demographics of his or her caseload.  All physicians have an
opportunity to discuss their indicative amounts with their district
health authority.  Prior to 1993, indicative amounts were set
according to historical expenditures for the practices and were
adjusted for comparable averages and special factors (e.g., the
number of high-cost patients, and general increases to allow for the
rise in drug costs).  Since 1993, indicative amounts have also
reflected regional expenditure patterns, projected requirements, and
projected expenditures.  Local factors to be considered in
constructing these projected expenditures include the relative
incidence of expensive patients in a practice; the incidence of
specific local illnesses (e.g., industrial diseases); and the
relative local morbidity as assessed by district health authorities. 

While IPS does not place binding cash limits on physicians, it sets
targets against which performance can be monitored.  There are
separate provisions that require physicians to justify their
prescribing behavior where there is clear evidence of
over-prescribing.\25


--------------------
\24 In this report, the term "district" refers to the Family Health
Service Authorities. 

\25 These provisions are separate from and predate IPS.  There is no
direct relationship between these provisions and the indicative
budget targets, and exceeding the indicative amount would not in
itself be sufficient to trigger these provisions. 


      GP FUNDHOLDING SCHEME
------------------------------------------------------ Appendix IV:5.4

The GP Fundholding Scheme is a modified version of IPS that is used
for physicians who are in group practices.\26 Under this scheme,
which was expected to cover 25 percent of the physicians by April
1993, GPs in large group practices (of at least 7,000 patients) are
awarded a budget that meets the cost of some hospital services,
administrative office costs, and visiting and district nurse
services; and all prescribing for their patients.  Physicians can use
their funds for any of these services as they see fit.  For example,
they can also apply savings in any one year over the next 4 years
either to improve their facilities or to buy more prescription drugs
for patients. 


--------------------
\26 The GP Fundholding Scheme is voluntary and NHS can refuse
applications to join it.  The main criterion for eligibility is the
size of the practice. 


      EDUCATIONAL MEASURES AIMED
      AT IMPROVING QUALITY AND THE
      COST-EFFECTIVENESS OF
      PRESCRIBING
------------------------------------------------------ Appendix IV:5.5

Other educational measures used to improve the quality of physicians'
prescribing practices include (1) the Medicines Resource Center,
which issues monthly bulletins on prescribing issues to all
physicians; (2) the Prescribing Research Unit, which researches
variations in prescribing practices and provides NHS with information
about the normal range of prescribing practices for certain types of
drugs; and (3) the Medical Advisers' Support Center, which trains and
educates medical advisers who, in turn, work with physicians to
improve prescribing practices. 


A STATISTICAL ANALYSIS OF THE
EFFECT OF DRUG PRICES ON
PHARMACEUTICAL RESEARCH AND
DEVELOPMENT EXPENDITURES
=========================================================== Appendix V

Pharmaceutical firms' R&D decisions are influenced both by market
forces and government policies.  Governments in many countries,
including France, Germany, Sweden, and the United Kingdom, have
imposed pharmaceutical price and spending control regulations.  By
lowering drug prices in these countries, such regulations can affect
firms' R&D spending by changing the return the firm receives from its
R&D investment. 

We use statistical analysis to examine the impact of changes in drug
prices on pharmaceutical R&D.  Our results support the hypothesis
that decreases in the average level of drug prices tend to reduce
pharmaceutical firms' R&D spending; conversely, increases in average
drug prices tend to increase pharmaceutical R&D expenditures.  These
results must be interpreted cautiously, as our estimates of the
magnitude of this effect on R&D are imprecise and are also sensitive
to statistical modeling choices.  In addition, our information does
not extend to how price changes affect the mix of innovative versus
imitative new drugs. 


   PREVIOUS STUDIES HELPFUL, BUT
   ADDITIONAL RESEARCH IS NEEDED
--------------------------------------------------------- Appendix V:1

Previous studies have not directly examined the relationship between
pharmaceutical prices and R&D expenditures.  Some authors suggest
that drug prices and profits are positively related to pharmaceutical
R&D, but they do not test this proposition empirically.\1 An
International Trade Commission report presents an analysis suggesting
that R&D is positively related to companies' global market shares.\2
An older study of the pharmaceutical industry found that firms with
relatively high profits in one time period tended to spend more on
R&D in subsequent years.\3 More recent research has estimated the
average costs of new drug development--for example, the Office of
Technology Assessment has estimated that the average after-tax cost
of developing a new chemical entity lies between $140 million and
$194 million 1990 dollars.\4 \5 OTA estimated that even after
accounting for the risk of failure in new drug development, the
average revenues received from these new drugs exceeded their high
development costs.\6

Even on a theoretical level, however, considerable uncertainty
remains as to the appropriate level of pharmaceutical R&D.  By
advancing the state of scientific and medical knowledge, R&D can
create benefits to society above and beyond the payments the firm
receives for its discovery.  This line of reasoning suggests that
private companies will likely invest too little in R&D.  However,
some private R&D spending may be wasteful.  Multiple firms may "race"
each other to create a viable product using the same basic chemical
substances.  While not all R&D spending in these races is wasteful to
society as a whole, the effort of the "losing" firm may not produce
much additional technological advance.  In practice as well as in
theory, little agreement exists on the desired amount of
pharmaceutical R&D, nor has consensus been reached on what (if any)
policy should be used to reach this desired level. 


--------------------
\1 See Henry G.  Grabowski, "An Analysis of U.S.  International
Competitiveness in Pharmaceuticals," Managerial and Decision
Economics, 27, Special Issue (1989); and L.G.  Thomas, "Industrial
Policy and International Competitiveness in the Pharmaceutical
Industry," presented at the American Enterprise Institute (Oct. 
1993). 

\2 See U.S.  International Trade Commission, Global Competitiveness
of U.S.  Advanced-Technology Industries:  Pharmaceuticals, USITC
Publication 2437 (Sept.  1991). 

\3 Henry G.  Grabowski and John M.  Vernon, "The Determinants of
Research and Development Expenditures in the Pharmaceutical
Industry," in Robert B.  Helms, editor, Drugs and Health:  Issues and
Policy Objectives (Washington, D.C.:  American Enterprise Institute,
1981). 

\4 For example, see U.S.  Congress, Office of Technology Assessment,
Pharmaceutical R&D:  Costs, Risks, and Rewards, OTA-H-522,
(Washington, D.C.:  U.S.  Government Printing Office, Feb.  1993);
J.A.  DiMasi, et al., "The Cost of Innovation in the Pharmaceutical
Industry," Journal of Health Economics, 10, 107-142, 1991; Steven
Wiggins, The Cost of Developing a New Drug (Washington, D.C.: 
Pharmaceutical Manufacturers' Association, 1987); and R.  Hansen,
"The Pharmaceutical Development Process:  Estimates of Development
Costs and Times and the Effect of Proposed Regulatory Changes,"
Issues in Pharmaceutical Economics, 1979. 

\5 These figures include, as they should, expenses for failed R&D
projects as well as successful ones.  These figures estimate the
average cost of new drug development, not the marginal cost of new
drugs.  While the average cost of a new drug would equal the total
R&D spending across firms divided by the number of drugs, the
marginal cost of a new drug would equal the change in R&D spending
necessary to produce one more new drug.  Therefore, it would be
inappropriate to use these average cost figures to describe how new
drug development would respond to a change in R&D expenditures. 

\6 For a more comprehensive review of the economic literature on the
pharmaceutical industry, see William Comanor, "The Political Economy
of the Pharmaceutical Industry," Journal of Economic Literature, Vol. 
XXIV (Sept.  1986), pp.  1178-1217, and U.S.  International Trade
Commission, Global Competitiveness of U.S.  Advanced-Technology
Manufacturing Industries:  Pharmaceuticals, USITC Publication 2437
(Sept.  1991). 


   A DECREASE IN DRUG PRICES WILL
   LOWER THE FIRM'S PAYOFF TO R&D
--------------------------------------------------------- Appendix V:2

Pharmaceutical firms, whose primary concern must be to maximize
profits, fund R&D in order to discover new products.  Their reward
for discovering a new drug is the profit they can earn from selling
this new drug in the marketplace.  That higher drug prices provide an
incentive for firms to undertake R&D may seem obvious, but
nonetheless this has been challenged. 

To maximize its profits, a firm must make its R&D choices by
comparing expected costs and benefits of each particular R&D project. 
If the expected benefit of a new project--the revenues expected from
the resulting product--exceeds the cost of research funding, the firm
will increase its overall expected profit by funding the project.  If
the expected cost of the project is greater than the profit a firm
can reasonably expect for its efforts, the firm will be better off
not to fund the project. 

Price regulation will have a direct impact on the expected benefit of
an R&D project.  A permanent and effective price regulation policy
will reduce the revenues a firm will receive from tomorrow's
prescription drugs.  The firm's reward for spending money on R&D--the
revenue the firm will receive from its future products--will decline
with price regulation, compared to what revenue would have been in an
unregulated market.  With a reduction in the expected benefit from a
successful R&D project, the firm has less incentive to invest in R&D. 


      PROFITS, MARKETING EXPENSES
      ARE UNLIKELY TO FULLY
      CUSHION THE IMPACT OF DRUG
      PRICES ON R&D
------------------------------------------------------- Appendix V:2.1

Recently, the profitability-R&D nexus has been explored from a
different perspective.  Attention has been directed to the question
of whether the pharmaceutical industry has been earning "excess
profits." Pharmaceutical industry critics have claimed that
pharmaceutical firms' high profits imply that prescription drug
prices could be lowered without a sacrifice in R&D.  Similarly, some
have suggested that pharmaceutical firms' marketing expenses could be
reduced, instead of R&D, in the event of a decline in drug prices. 

Economic theory suggests that the presence of significant industry
profits or marketing expenses would be insufficient to break the link
between drug prices and pharmaceutical R&D, for several reasons. 
First, drug companies, like other private corporations, generally
seek to earn high returns; company profits provide the incentive for
these firms to pursue pharmaceutical R&D.  If the profits from
selling new drugs are reduced, firms have less reason to commit to
costly long-term R&D projects.  A firm's high profit level may
encourage other firms, who hope for similar results, to invest
additional resources in R&D.  Pharmaceutical firms must market the
products generated by their R&D in order to realize the profit
potential in their new drugs. 

Second, because considerable uncertainty surrounds the payoffs to
R&D, observed profits may not be an accurate indicator of the
compensation necessary to induce R&D investment.  R&D dollars must be
committed long before the outcomes are known.  Therefore, the stream
of profits observed after the product is marketed will differ from
the stream of profits the firm expected to earn from the product when
the initial R&D decision was made.  Under these circumstances, firms'
errors in forecasting future revenues would be indistinguishable from
excess profits. 

In addition, although there may well be "excess profits" on average
over a group of drugs, R&D decisions are made for each individual
drug.  That is, the firm makes its R&D decisions by comparing the
expected benefits and costs of each individual R&D project, not by
looking at overall benefits and costs for a large number of projects
taken together.  For example, suppose that the firm has the choice of
funding any or all of 10 research projects.  The firm can rank these
projects according to their expected payoff, as in figure V.1--if
funded, project A can be expected to bring in about $100 million,
project B has an expected value of $90 million, and so on.\7 If each
project cost $50 million, the firm would maximize its profits by
funding projects A through F.  On average, the firm could expect to
earn $150 million in profits from this set of drugs. 

   Figure V.1:  Relationship
   Between Drug Prices and R&D in
   the Presence of "Excess
   Profits"

   (See figure in printed
   edition.)

If drug prices were decreased, lowering the expected benefit of each
drug by 10 percent, then the benefits of funding projects A through E
would still exceed their costs.  However, project F would be
unprofitable.  The firm would maximize its profits by funding only
projects A through E, earning a profit of $100 million.  Despite the
presence of high industry profits, a reduction in price would
nonetheless result in a decline in R&D. 


--------------------
\7 These figures would take into account the uncertainty surrounding
the success of each project.  For example, if project A had a
25-percent chance of producing a $400 million product, its expected
value would be $100 million. 


      A DECREASE IN DRUG PRICES
      MAY INCREASE THE COST OF
      FINANCING R&D
------------------------------------------------------- Appendix V:2.2

In addition to the role of drug prices in creating incentives for
investment in R&D, drug prices may influence the cost of financing
R&D.  Industry representatives point out that pharmaceutical
companies finance their R&D expenditures largely from current
profits.  From this fact, industry representatives conclude that if
profits were reduced, firms would be unable to undertake R&D. 

This argument is partially--but only partially--correct.  A reduction
in profits may force pharmaceutical firms to obtain more of their R&D
financing from external sources.  If it is cheaper for firms to
finance R&D from current profits, then by limiting these profits,
price regulation could increase the cost of financing R&D projects. 
However, although pharmaceutical firms may choose to finance R&D
through their current corporate profits, these firms do have other
options.\8 Therefore, industry profits may facilitate R&D spending,
but it would be incorrect to say that R&D would not be possible
without high industry profits. 


--------------------
\8 These other options include debt financing and equity financing. 
If a firm chooses debt financing, the firm borrows money to pay for
its current R&D costs, and repays the debt once the revenue from the
project is received.  If a firm chooses equity financing, the firm
issues additional stock shares and uses the proceeds to pay for its
current R&D costs.  The firm will know, better than potential outside
investors, the expected costs and benefits of the project.  If it
wants to obtain the funds at the cheapest possible cost, the firm has
every incentive to exaggerate the potential benefits of the project
and minimize its potential costs.  Investors will need impartial
information to make their decisions, and gathering such information
is likely to be very costly, if not impossible.  Given the high
degree of risk in pharmaceutical R&D and this lack of information,
outside investors will likely demand very high interest rates for
debt financing and low prices for new stock issues.  Therefore, it is
probably cheaper for pharmaceutical firms to finance their R&D
through retained earnings.  Indeed, this is what most of the major
pharmaceutical firms do most of the time. 


   OTHER GOVERNMENT POLICIES CAN
   ALSO INFLUENCE R&D
--------------------------------------------------------- Appendix V:3

In each of the countries we studied, as well as in the United States,
the government plays a crucial role in the pharmaceutical market--as
regulator of drug prices and product approval; as granter of patents
and tax credits; as creator and enforcer of product liability laws;
as the provider of public funding for pharmaceutical and biomedical
R&D and (in the countries we studied) as sole or dominant purchaser
of pharmaceuticals under the national health insurance system.  In
each of these roles, the government exercises considerable influence
over pharmaceutical firms' R&D decisions. 

For example, government safety regulations can affect the incentives
for new drug development.  The cost of clinical trials contributes to
the cost of bringing a new drug to market; the more extensive these
requirements are, the higher the expected cost of R&D is.  As granter
of patents, the government sets the period of market exclusivity. 
The longer this period is, the more profitable are new drugs and
hence, the more incentives for R&D.  Tax credits and public funding
can spur R&D by reducing a firm's R&D costs, while strict product
liability can increase the risk-related costs of developing new
drugs. 


      GOVERNMENT INFLUENCE DEPENDS
      ON THE SIZE AND IMPORTANCE
      OF THE DOMESTIC
      PHARMACEUTICAL MARKET
------------------------------------------------------- Appendix V:3.1

We expect the strength of government influence on R&D expenditures to
depend critically on the size and importance of the pharmaceutical
market in the manufacturer's home country.  For example, if a
relatively small country (such as the Netherlands) were to see a fall
in drug prices, the impact on drug company revenues and profits would
be limited because this country's consumers account for a relatively
small share of the market.  However, a loss of revenue in a larger
country (such as the United States) would be expected to have more
significant effects on all firms--domestic and foreign.  Similarly,
firms that derive a larger portion of their sales from exports will
be less affected by the policies of their home governments. 


   MARKET FORCES AND R&D COSTS CAN
   ALSO AFFECT PHARMACEUTICAL R&D
--------------------------------------------------------- Appendix V:4

Despite widespread government influence in the pharmaceutical market,
market forces remain important.  Without some demand for new
pharmaceutical products from patients and physicians, R&D projects
will not be profitable.  Marketing practices, such as advertising to
physicians or (where allowed) directly to consumers, can affect
physician and consumer demand.  Cultural factors, demographics, and
local prescribing and practice patterns will affect the consumption
of prescription drugs and the acceptance of new products. 

The costs of conducting R&D will depend on market forces as well. 
Wages for scientists and other skilled workers can vary across local
labor markets.  In addition, a strong university system and easy
access to the scientific community may encourage research and
development. 


   EMPIRICAL ANALYSIS OF DRUG
   PRICES' EFFECT ON R&D
--------------------------------------------------------- Appendix V:5

Because so many factors combine to affect firms' R&D decisions,
isolating the impact of any one variable is difficult.  In addition,
some of these variables are difficult to quantify or measure
reliably.  However, many of the variables that influence
pharmaceutical R&D do not change over time, or change over time only
very slowly.  Patent and tax laws, for example, are not changed very
frequently.  By looking at changes in R&D across countries over time,
we can control for the influence of some important confounding
variables. 

Data were insufficient to estimate a complete structural model for
the pharmaceutical industry in each country.  However, we were able
to exploit the variation in drug price growth across countries to
test the hypothesis that drug price levels affect R&D.  In addition,
we used a simulation model to explore how government policy and
domestic market conditions may influence the strength of the
price-R&D relationship. 

Economic theory suggested several alternative specifications for
modeling the price-R&D relationship.  We varied the form of the
regression models we used to see how sensitive the results were to
the model specification.  The regression results suggest that any
policy--regulatory or competition-enhancing--that reduces drug prices
will decrease R&D spending.  The effect was statistically significant
and appeared consistently throughout the different empirical models,
although the size of the estimated effect was sensitive to the
changes in specification. 


      REGRESSION MODELS
------------------------------------------------------- Appendix V:5.1

We obtained data on reported pharmaceutical R&D expenditures for 87
firms in 12 countries for 1988 and 1989 from the Scrip Pharmaceutical
League Tables.\9 \10 These figures represent only that portion of the
company's total R&D expenditures devoted to pharmaceuticals.  Figures
for major foreign subsidiary companies were generally reported
separately from the parent company. 

For data on the average drug price level in each country, our measure
was the Organization for Economic Cooperation and Development (OECD)
pharmaceutical price index for each country; we used this index to
compare the growth in drug prices over time across countries.\11 \12
We also obtained international economic statistics, such as the GDP
and the GDP deflator (a measure of the general inflation rate), from
the OECD Health Data Base. 

These data enabled us to use regression analysis to estimate how
pharmaceutical prices affect R&D.\13 We hypothesize that the growth
in the firm's R&D spending would be affected by changes in drug
prices and in real GDP.\14 We measured these changes both in absolute
terms and as growth rates.  We estimated a linear and a nonlinear
(including squared terms) version of each model. 


--------------------
\9 Many pharmaceutical firms also produce other products.  We
included in our sample only those firms that specifically reported
pharmaceutical R&D, rather than total company R&D.  These firms
represented Belgium, Finland, France, Germany, Italy, Japan,
Portugal, Spain, Sweden, Switzerland, the United Kingdom, and the
United States. 

\10 We obtained similar data for 1990 and 1991.  However, most firms
in Germany did not report data for 1990 and 1991.  We estimated our
models on the 1988-1989 data, and on a pooled data set that included
all 4 years of data.  The results in each case were qualitatively
similar.  In this appendix, we report results based only on the
1988-1989 data, which include the German pharmaceutical firms. 

\11 The Organization for Economic Cooperation and Development (OECD)
is an international organization of 24 industrialized countries in
Europe, North America, and the Pacific. 

\12 The problems with price indexes for pharmaceuticals are well
known; see Fredrik Andersson and Peter McMenamin, "International
Price Comparisons of Pharmaceuticals--A Review of Methodological
Issues," Battelle Medical Technology and Policy Research Centre (Aug. 
1992).  However, we believe that these indexes provide the best
currently available measurement of pharmaceutical price changes over
time in each country. 

\13 This technique allows us to examine the impact of each
characteristic on R&D, holding other factors constant. 

\14 As discussed in chapter 3, economic theory suggests many
potential determinants of R&D.  Specifying a model of R&D in levels
demands, therefore, a rich data set.  However, data on R&D
determinants are patchy.  This lack of data necessitates an
alternative model that both is operational and accounts for the
(missing) confounding variables. 


         A MODEL SPECIFIED IN
         LEVELS
----------------------------------------------------- Appendix V:5.1.1

A straightforward way to estimate the effect of prices on R&D would
be to regress the firm's pharmaceutical R&D expenditure on the home
country pharmaceutical price level plus a set of control variables. 
This approach, relating the level of R&D to the level of each
determinant of R&D, is captured by equation 1.  The variables in
equation 1 are defined in table V.1. 

1. 





                          Table V.1
           
            Definition of Variables in Equations 1
                            and 2

Variable  Description
--------  --------------------------------------------------
Rijt      Research and development expenditures for firm i
          in country j at time t

t

Pjt       Pharmaceutical price level in country j at time t

Gjt       Measure of economic activity (GDP) in country j at
          time t

Xj        Vector of other variables (patent law, tax law,
          etc.) that influence R&D and are specific to
          country j

et        Random error term
------------------------------------------------------------
This model has the virtue of, at least potentially, including the
complex set of determinants affecting R&D.  However, the particular
features of the pharmaceutical market make this specification
inappropriate.  For example, variation across countries in the
accounting definitions of R&D makes it difficult to compare R&D data
from different countries.  Also, the levels approach would require
the researcher to account for firm size on the right-hand side of
equation 1, but doing so poses statistical problems.  Larger firms,
naturally, tend to spend more on R&D.  Most measures of firm size,
including the number of employees and total sales, are endogenous;
consequently, an instrumental variables or systems estimator would be
required to estimate this model.  We lacked the data to construct
such an estimator. 

Finally, we were unable to obtain data on a number of important
control variables.  Specific measures of marginal tax rates, for
example, are difficult to obtain.  It is very difficult to quantify
variables like "access to scientific infrastructure." Our inability
to account for these variables would subject the estimation to
omitted variable bias.  In particular, the estimated coefficient on
the drug price variable would be biased. 


      AN ALTERNATIVE, OPERATIONAL
      MODEL
------------------------------------------------------- Appendix V:5.2

To surmount these difficulties, we specified empirical models in
differences and in growth rates.  These models helped us minimize
omitted variable bias because we are required to account for only
those variables that change over time.  Instead of estimating the
effect of this year's pharmaceutical price level on this year's level
of R&D expenditure (as in equation 1), we looked at how changes in
the pharmaceutical price level lead to changes in R&D spending from
one year to another.  Most of the key variables in the R&D decision,
including patent lives and the scientific infrastructure, are fixed
from one year to the next.  Therefore, such variables will influence
the level of R&D at a point in time, but should not influence the
growth in R&D over time.  We assumed that the coefficients of these
variables are also constant over time; therefore, the variables can
be eliminated by first differencing. 

For example, we lagged equation 1 and subtracted it from the original
equation; the result is given below.  (Regression results based on
this specification are given in table V.4.)

2. 



Another way to estimate the effect of price regulation on R&D in
terms of changes in variables is to formulate a statistical model in
terms of growth rates.  We hypothesized that the growth rate in
prices and GDP from one year to the next will influence the growth
rate in R&D spending from one year to the next.  This specification
is given in the equation below.  (Regression results based on this
specification are given in table V.2.)

3. 



We also estimated a model based on equation 3, but including squared
terms.  This specification is given in the equation below. 
(Regression results based on this specification are given in table
V.3). 

4. 

 The results of these various regression models are
given in tables V.2 through V.4.  (For the specification in terms of
differences, we calculated the elasticity at the point of means.) The
elasticity estimate in table V.3, for example, implies that a
1-percent decrease in the pharmaceutical price level would lead to a
0.68-percent decrease in the average firm's R&D expenditures.  The
model with the smallest estimate would imply that a 1-percent
decrease in drug prices would cause a 0.3 percent drop in R&D
spending, while the model with the largest estimate would imply that
a 1-percent decrease in drug prices would cause a 0.7-percent drop in
R&D spending. 

Not surprisingly, we found heteroskedastic errors in several of these
equations.  Where heteroskedasticity was found, the standard errors
presented are based on consistent estimates of the covariance matrix
for each regression, using White's test. 

In each regression, our right-hand side variables explain a portion
of the cross-country variation in R&D, but we are not attempting to
explain the considerable variation among individual firms within a
single country.  In addition, these regressions express R&D in terms
of differences, rather than levels.  Therefore, our regressions
explain a relatively small portion of the cross-sectional sample
variation. 



                          Table V.2
           
                     Regression Results I

              Coefficient/estimated elasticity    t
Variable      (standard error)                    statistic
------------  ----------------------------------  ----------
Constant      21.512                              4.218\a
              (0.556)

Pharmaceutic  0.556                               1.80\b,c
al price      (0.318)
growth

GDP growth    -4.347                              -2.121\a
              (2.050)
------------------------------------------------------------
R\2 = 0.046
n = 87

Dependent variable:  Percentage growth rate in R&D expenditures in
1988-1989 for firm i, which is located in country j, corrected for
inflation with the GDP deflator. 

Independent variables:  Percentage growth in pharmaceutical prices
for 1988-1989 for country j, corrected for inflation with the
consumption deflator; and percentage growth rate in GDP for 1988-1989
for country j, corrected for inflation with the GDP deflator. 

\a Significant at the 0.05 level (two-tailed test). 

\b Significant at the 0.10 level (two-tailed test). 

\c Significant at the 0.05 level (one-tailed test). 



                          Table V.3
           
                    Regression Results II

              Coefficient/estimated elasticity    t
Variable      (standard error)                    Statistic
------------  ----------------------------------  ----------
Constant      -8.162                              -0.277
              (29.430)

Pharmaceutic  0.676                               1.914\a,b
al price      (0.353)
growth

Pharmaceutic  -0.015                              -0.267
al price      (0.055)
growth
squared

GDP growth    21.373                              0.856
              (24.978)

GDP growth    -5.047                              -1.036
squared       (4.873)
------------------------------------------------------------
R\2 = 0.073
n = 87

Dependent variable:  Percentage growth rate in R&D expenditures in
1988-1989 for firm i, which is located in country j, corrected for
inflation with the GDP deflator. 

Independent variables:  Percentage growth in pharmaceutical prices
for 1988-1989 for country j, corrected for inflation with the
consumption deflator; percentage growth rate in GDP for 1988-1989 for
country j, corrected for inflation with the GDP deflator; real price
growth squared; real GDP growth squared. 

\a Significant at the 0.10 level (two-tailed test). 

\b Significant at the 0.05 level (one-tailed test). 



                          Table V.4
           
                    Regression Results III

              Coefficient/estimated elasticity    t
Variable      (standard error)                    Statistic
------------  ----------------------------------  ----------
Constant      29.517                              0.856
              (34.465)

Pharmaceutic  0.494/0.306                         2.864\a
al price      (0.173)
difference

GDP           71133.4/0.00455                     1.22
difference    (58270.0)
------------------------------------------------------------
R\2 = 0.126
n = 87

Dependent variable:  Difference (in millions of dollars) between R&D
expenditures in 1988 and 1989 for firm i, which is located in country
j, corrected for inflation with the GDP deflator. 

Independent variables:  Percentage growth in pharmaceutical prices
for 1988-1989 for country j, corrected for inflation with the
consumption deflator; percentage growth rate in GDP for 1988-1989 for
country j, corrected for inflation with the GDP deflator. 

\a Significant at the 0.05 level (two-tailed test). 

The estimated impact of GDP varies among the models specified.  This
inconsistency may suggest that current GDP is a poor proxy for the
GDP that will prevail when new drugs are marketed.  In addition, a
rise in GDP may be correlated with a rise in interest rates in the
home country, which may depress R&D if capital markets are imperfect. 
However, prices have a consistently positive impact on R&D in each of
the models tested. 


   RESULTS SHOULD BE INTERPRETED
   CAUTIOUSLY
--------------------------------------------------------- Appendix V:6

Caution is required in interpreting these results, because the
magnitude of the effect was difficult to estimate and is subject to a
number of qualifications.  First, our sample of firms was limited to
those firms that reported figures, and therefore this group
constitutes a nonrandom sample.  This nonrandomness will affect the
results only if the reporting firms were to differ systematically
from the nonreporting firms; we have no evidence that this is likely. 

The simplicity of our specification may be considered a drawback of
the analysis.  Current price and GDP trends are only proxies for the
trends a firm expects as it makes its R&D decision.  Although most of
the other variables that we expect to impact R&D are generally
time-invariant in the short run, we are unable to account for
long-run variation in these variables.  Even if data were available,
the small number of countries with research-based pharmaceutical
industries restricts the number of explanatory variables we could
potentially include.  In addition, to identify the empirical model,
we assumed that the marginal benefit of additional R&D spending for
firm i in country j depends on pharmaceutical prices in the home
country j and on the "world" pharmaceutical price (which would be the
same for all firms).  In fact, firms may have differing marketing and
distribution advantages in markets other than in their home country,
and so the available "world" market may differ across firms.  We
lacked the data to give our model this degree of complexity.  The
simplicity of our model also prevented us from exploring other
possible forces behind the link between drug prices and R&D.  For
example, authorities in countries with weak industries may have felt
free to impose low rates of increase in drug prices because they did
not have a major domestic industry to protect.  Similarly, in
countries with strong industries, country authorities may have felt
constrained to impose more moderate pricing policies.  Data were
unavailable to test this hypothesis. 

In addition, we do not know how changes in drug prices may affect the
social value of pharmaceutical R&D.  The negative effect of a
decrease in drug prices may fall largely on either innovative or
imitative drug products.  The loss of innovative "blockbuster" drugs
would be a bigger blow to patients than the loss of line extensions
or "me-too" products. 

The size of the effect of prices on R&D was difficult to estimate
precisely, and the magnitude of the estimate was sensitive to the
empirical specification.  The statistical significance of the effect
was generally unaffected by the specification adopted, but the
estimate of the size of the effect differed across specifications. 
For all these reasons, the size of the effect of drug prices on
pharmaceutical R&D must remain an open question. 


   SIZE OF R&D EFFECT MAY BE
   INFLUENCED BY OTHER VARIABLES
--------------------------------------------------------- Appendix V:7

Unfortunately, data considerations prevented us from estimating a
complete structural model of the R&D decision for firms in each
country.  As the major pharmaceutical countries differ in the size of
their domestic markets, we expect that the strength of the price-R&D
relationship will vary across countries in ways our regression model
cannot capture.  To explore how government policy and market
conditions might affect the relationship between regulated prices and
R&D, we used a computer simulation model.  This model is designed not
to test whether drug prices affect R&D across countries in general,
but to illuminate those factors that may determine the size of the
potential R&D effect in a given country. 


      SIMULATION MODEL
------------------------------------------------------- Appendix V:7.1

Although different countries have established different policy
regimes with respect to the pharmaceutical industry, their historical
experience with pharmaceutical price regulation is too limited for us
to make empirical conclusions about the effects of government policy
on R&D.  One way to gain some insight into these issues is by using a
computer simulation model to combine theory and experience. 

A simulation combines a theoretical equation with actual data and
with reasonable conjectures about the size of parameters.  In a
simulation model, the researcher first derives a theoretical equation
that can be solved for the desired effect.  Then he or she
substitutes plausible values, or values derived from data, for the
other variables in the equation and solves it.  By varying the values
used in the equation, the researcher can see how sensitive the effect
is to changes in these parameters. 

Our simulation model analyzes hypothetical industry situations. 
However, the model was formulated to take into account several
important features of R&D in the pharmaceutical industry.  Firms
invest in R&D to maximize their profits.  The probability of success
for each individual project is low, but successful projects generate
considerable returns.  In the baseline case for the model, we
incorporate parameters representative of current industry experience
(with respect to the probability of success, the tax credit rate,
interest rates, and so forth).  This model is not nearly
sophisticated enough to provide a planning or forecasting device. 

This simulation model follows the work of Grabowski and Vernon.\15
Grabowski and Vernon used a simulation model to explore the potential
effects of lengthening pharmaceutical patent lives on the rate of new
drug discovery and the concentration of the pharmaceutical industry. 
We are examining a different dependent variable than Grabowski and
Vernon did--while they were interested in the structure of the
industry and the rate of new drug development, we are examining the
elasticity of R&D spending with respect to regulated drug prices. 
However, the two simulation models share several common
parameters--the probability of technical success and the R&D tax
credit being the most important examples. 

To develop the simulation, we used a simplified two-period model of
R&D.  The firm maximizes its profits with respect to R&D by investing
in an R&D project if the marginal benefit of the project exceeds its
marginal cost.  At the equilibrium R&D choice, then, the net benefit
of the marginal R&D project is equal to zero. 

In period one, the firm sells product A in its home (H) and foreign
(F) markets at the regulated prices P\AH and P\AF .  The firm must
pay production costs C\AH and C\AF , plus the capital, labor, and
financing costs of the R&D project, less any tax credits.  In period
two, the firm succeeds in generating new product B with probability
of success s\B , and fails to generate a new product with probability
[1 - s\B ].  If the firm's R&D is successful, the firm will sell
product B in its home and foreign markets at the regulated prices
P\BH and P\BF and pay production costs C\BH and C\BF . 

We assume that information is asymmetric--that s\B is known to the
firm only, and the firm cannot credibly reveal s\B to outside
investors.  Outside investors will therefore demand a higher rate of
return than the firm's opportunity cost of retained earnings.  The
firm will spend its profits on R&D before it seeks to borrow funds
from the outside market.  Therefore, if the cost of the firm's R&D
project exceeds its current profits, it will finance the project
using current profits and borrow the remaining sum.  If the cost of
the R&D project is less than current profits, the firm will finance
R&D solely through retained earnings. 

We consider the more general case in which the firm's internal funds
are insufficient to completely finance the marginal R&D project.  The
first order condition sets the net benefit of the marginal R&D
project at

5. 



The variables in equation 5 are defined in table V.5. 

Price regulation can affect pharmaceutical R&D in two ways--through
current profits and through expected future profits.  Price
regulation will reduce expected future profits and thereby lessen the
incentives for firms to conduct R&D.  Under price regulation, the
firm will achieve lower profit levels than it would receive if it
were allowed to set prices freely.  The more stringent the price
regulation (that is, the lower the regulated price), the lower the
expected benefits to conducting R&D. 

Price regulation can also influence the firm's R&D decision through
its current profits.  Price regulation will reduce the profits on the
firm's current product line.  If borrowing and lending rates differ,
a decline in current profits means that there has been a decline in
the availability of funds to finance new R&D.  The firm may need to
borrow additional funds for the external capital market, which
represents an increase in the cost of financing R&D.  If borrowing
and lending rates are equal, this financing effect drops out, and
price regulation's impact is confined to its effect on future prices. 

We will define the effect of price regulation on R&D expenditure as
follows: 

6. 



R&D expenditures are represented in this model by wL + rKK.  Solving
the first order condition for R&D expenditures and taking the
derivatives with respect to prices yields

7. 



and

8. 





where ï¿½i equals the price elasticity of marginal cost for product i. 

Evaluating the elasticity of R&D with respect to price at the point
of means yields

9. 



and

10. 





A major issue in any simulation is how to "initialize" the model. 
For several of these parameters, plausible proxies were available. 
We used the prime lending rate as a proxy for the borrowing rate
(r2), and the 10-year T bill rate as a proxy for the opportunity cost
of internal funds (r1).  As starting values, we used estimates
reported by the Office of Technology Assessment for the tax savings
rate  and the probability of success s\B ; we then varied
these parameters to test for sensitivity.\16 We varied the discount
rate from 0.1 to 0.25.  For several other parameters, we used
plausible conjectures to narrow the range of values we considered. 
For example, in the pharmaceutical industry, production costs are
likely to be relatively small compared to the total cost of any drug. 
The monopoly power granted to pharmaceutical firms through patents
and pharmaceutical companies' relatively high profit margins also
point to a low ratio of marginal cost to price.  We varied the
marginal cost-to-price ratio between 0.01 and 0.20.  To identify the
equation, we used the average pharmaceutical sales/pharmaceutical R&D
ratio of 87 firms in 1989. 

The simulation model is detailed in tables V.5 and V.6.  The
parameters and their values are given in table V.5; the results of
the simulation are given in table V.6.  The simulation results in
table V.6 confirm the belief that the sensitivity of R&D to changes
in drug prices may depend on the firm's environment.  For instance,
the variable that made the most difference in determining the
response of R&D to changes in regulated prices was the R&D tax
credit.  The larger this credit was, the more responsive the firm was
to changes in the regulated price.  When firms receive large R&D tax
credits, their cost of R&D is reduced, making it easier for them to
exploit new R&D projects in response to a price increase. 



                          Table V.5
           
              Variables Used in Simulation Model

                                              Source of
Variable            Definition      Value     value
------------------  --------------  --------  --------------
s\B                 Probability of  0.10 -    The Office of
                    success of      0.30      Technology
                    marginal R&D              Assessment
                    project in                reports
                    producing a               success rates
                    new drug\a                of between
                                              12.5 percent
                                              and 23.0
                                              percent. We
                                              used these
                                              estimates to
                                              construct our
                                              boundaries.

           Discount rate   0.05 -    Grabowski and
                                    0.25      Vernon varied
                                              their discount
                                              rate from 0.05
                                              to 0.2.

           Rate of tax     0.4 -     The Office of
                    savings for     0.8       Technology
                    every $1 the              Assessment
                    firm spends on            reports an
                    R&D                       estimate of
                                              0.54. We
                                              varied
                                              
                                              around this
                                              estimate.

[P\BH Q\BH]/R+D     Ratio of        6.4842    We computed
                    revenues to               the average
                    R&D for                   revenues/R&D
                    product B                 ratio for 87
                    in market H               firms in 1989.

[ C\BH/Q\BH]/P\BH   Ratio of        0.05 -    Industry
                    marginal cost   0.20      officials
                    to price for              indicated that
                    product B in              production
                    market H                  costs were a
                                              relatively
                                              small portion
                                              of total
                                              costs.

ï¿½\B                 Elasticity      0.01 -    Marginal costs
                    marginal cost   1.5       are thought to
                    with respect              be small
                    to price for              compared to
                    product B                 price;
                                              however,
                                              without strong
                                              priors on the
                                              magnitude of
                                              the elasticity
                                              of marginal
                                              cost, we
                                              selected a
                                              wide range.

r1                  Borrowing rate  10.50     Prime lending
                    for external    percent   rate as of
                    funds                     January 1989\b

r2                  Opportunity     9.09      10-year T bill
                    cost of         percent   rate as of
                    internal funds            January 1989
------------------------------------------------------------
\a This probability is conditional on the R&D having reached a point
where a go/no go decision could be made on a specific product. 

\b We used these interest rates because they belonged to the same
period as our data on the sales/R&D ratios of pharmaceutical
companies. 



                          Table V.6
           
                 Results of Simulation Model

            Value or
            range of    Range of values for estimate of
            values for  elasticity of R&D with respect to
Parameter   parameter   drug prices
----------  ----------  ------------------------------------
Elasticity  0.1-1.5     0.2803-0.6231
of
marginal
cost with
respect to
price

Discount    0.1-0.25    0.5325-0.8918
rate

Probabilit  0.1-0.3     0.2465-0.7082
y of
success

Tax credit  0.1-0.8     0.1246-1.094\
rate
------------------------------------------------------------
All parameters were initialized at the baseline rates, and then
varied as noted in the table. 

Baseline rates: 

Discount rate = 0.1
Probability of success = 0.2
Tax credit rate = 0.54
Elasticity of marginal cost
 with respect to price = 1.0

In addition, firms that face more elastic marginal cost curves in
their current and future product lines are more sensitive to the
impact of price regulation.  This conforms to our intuition; the more
elastic marginal cost is, the more the firm can exploit an increase
in the regulated price by raising output.  This effect, like the
effects of all the varied parameters, was large, indicating that the
effect of price regulation on R&D is responsive to the firm's
environment.  The probability of success in an R&D project was also
an important variable.  Firms are more likely to respond to price
increases by increasing R&D when success is surer.  Again, the range
of values was quite wide. 

A simulation model can be helpful in developing a range of plausible
values, but it is of limited value in testing whether the effect
truly exists.  A simulation model is, after all, based largely on
plausible conjectures rather than actual experience.  While our
regression model supports the contention that drug prices influence
pharmaceutical R&D, our simulation model illustrates that policy
parameters may influence the size of this effect. 


--------------------
\15 See Henry Grabowski and John M.  Vernon, "A Computer Simulation
Model of Pharmaceutical Innovation," in Arne Ryde Symposium on
Pharmaceutical Economics (1984), p.  165. 

\16 See Pharmaceutical R&D:  Costs, Risks and Rewards, U.S. 
Congress, Office of Technology Assessment, OTA-H-522 (Feb.  1933), p. 
196. 


ACKNOWLEDGMENTS
========================================================== Appendix VI

GAO would like to acknowledge the assistance of the following
experts.  These individuals provided valuable insights on the issues
discussed in this report; however, they do not necessarily endorse
the positions taken in the report. 

P.  ï¿½tienne Barral, Managing Director, Rhï¿½ne-Poulenc Rorer S.A.,
France
Joseph A.  DiMasi, Senior Research Fellow, Center for the Study of
Drug  Development, Tufts University
Jim Furniss, Head of Pharmaceutical Industry Branch, Department of
 Health - NHS, The United Kingdom
Michael Gluck, Senior Analyst, U.S.  Office of Technology Assessment
Henry G.  Grabowski, Professor of Economics, Duke University
Bernard Harrison, Private Consultant, The United Kingdom
Wolfgang Kaesbach, Head of Pharmaceutical Department, Bundesverband
 der Betriebskrankenkassen, Germany
Christoph H.R.  Lankers, Research Fellow, Wissenschaftliches Institut
der  Ortskrankenkassen, Germany
Louis Lasagna, Dean, Sackler School of Graduate Biomedical Sciences,
 Tufts University
Hï¿½kan Mandahl, Executive Vice President, The Association of Swedish
 Pharmaceutical Industries (Lï¿½kemedelsindustrifï¿½reningen), Sweden
Heinz Redwood, Private Consultant, The United Kingdom
Lise Rochaix, Professor of Economics, Universitï¿½ de Paris Dauphine,
 France
Stephen W.  Schondelmeyer, Director, PRIME Institute, University of
 Minnesota
Gisbert W.  Selke, Research Fellow, Wissenschaftliches Institut der
 Ortskrankenkassen, Germany
Edwin Smigielski, Business Director, Medizinisch-Pharmazeutische
 Studiengesellschaft e.V.  (MPS), Germany
Kevin Sullivan, International Affairs Manager,  Medizinisch-
Pharmazeutische Studiengesellschaft e.V.  (MPS), Germany
Judith L.  Wagner, Senior Associate, U.S.  Office of Technology
Assessment
Lars Wetterstrï¿½m, Senior Administrative Officer, National Social
Insurance  Board, Sweden


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix VII

Jonathan Ratner, Assistant Director, (202) 512-7119
Sarah Glavin, Economist, (202) 512-7180
David Gross, Assignment Manager
James Perez, Evaluator-in-Charge
Claude Hayeck, Evaluator
Thomas Laetz, Senior Evaluator
Josï¿½ Peï¿½a, Evaluator
Christopher Conrad, Evaluator
