Transcript Slide 1

Intellectual Property Rights Protection in
Developing Countries: The Case of
Pharmaceuticals.
Marshall Lecture
European Economic Association Congress,
Barcelona
August 27, 2009
Penny Goldberg
Princeton University,
NBER and BREAD
Background and Motivation
• IPR protection has emerged as a principal issue in domestic competition
policy.
• In the international context: TRIPS & WTO (1995).
• Member countries had to recognize and enforce product patents in all
fields of technology, including pharmaceuticals.
• Developing countries had an extension of 10 years. Had to have patent
protection by 2005.
• Here: Focus on pharmaceuticals because of the public policy importance
of this sector.
• Question: Is there a case for harmonization of patent policies across
countries at different stages of development?
Why is it a highly contentious issue?
One side of the argument...
•
Patent enforcement in developing countries will lead to higher prices.
•
“The idea of a better-ordered world is one in which medical discoveries
will be free of patents and there will be no profiteering from life and
death.” [Indira Gandhi (1982)]
Why is it a highly contentious issue?
(continued)
The other side of the argument...
•
“‘High priced medicines after change in patent laws is a misconception.”
[Organization of Pharmaceutical Producers of India].
• TRIPS compliant patent laws attract foreign R&D investment in
developing countries, and promote technology transfer.
• Patents may provide incentives for research on developing-country
specific diseases
Thesis of this talk
• Existing arguments reflect the trade-off between static efficiency loss
(higher prices) and potential dynamic gains (research, new products).
• But: Higher prices are unlikely to materialize because of price regulation
• Dynamic gains also unlikely. Developing country markets too small to
change research priorities of pharmaceutical concerns.
• Potentially largest effect on distribution and availability of new products in
the developing world
 Distribution over time (launch of new drugs)
 Distribution over space (rural versus urban)
• Then what is the fuss about?
 Global reference pricing
 Long-run concerns about exports from developing markets
Road Map
• Theoretical arguments regarding IPR in developing countries
• The challenges on the empirical side
 What makes developing countries special?
• Lessons from a Case Study: India and Quinolones (newer generation of
antibiotics)
 Analyze patterns in the data
 Impose structure to estimate welfare effects
• Do the results generalize? Some cross-country evidence
• Policy implications
• What has happened since new patent legislation was signed in 2005?
Theoretical Arguments
Standard analysis of patents in the closed economy:
 static (short-term) pricing distortions
 dynamic (longer-term) innovation gains
 optimal policy equates the marginal static efficiency loss to the marginal
dynamic benefit (Nordhaus 1969).
Complications in multi-country setting:
Note: Most trade economists not in favor of TRIPS
Theoretical Arguments (continued)
 Fundamental Externality: Benefits of innovation are spread beyond national
boundaries. Countries differ in their capacity for innovation because of:
o Skill endowment and technical know-how
o relative size of domestic market
 Is Patent Length Harmonization (PLH) a necessary condition for global
efficiency? Global Efficiency: Regime that provides the optimal aggregate
incentives for innovation throughout the world
Answer: NO (Grossman and Lai [2003])
However: PLH has important implication for the distribution of
welfare between developed and developing countries.
 Global Reference Pricing
 Further complications with differentiated products such as pharmaceuticals
On the Empirical Side …
• Very few empirical studies on drugs in developing countries.
• Arguments based on assumptions, rather than estimated parameters.
Main limitations of this approach:
○ Domestic and foreign products are assumed to be perfect
substitutes. Any welfare losses associated with patent enforcement
stem from price increases of foreign products alone.
○ Substitution towards other drugs and therapeutic segments ignored.
• Work on pharmaceutical markets in developed countries (plenty of
papers) not pertinent to the TRIPS debate, because developing countries
differ from developed countries in five critical respects:
(Some) Important Differences between Developed
and Developing Countries
1.
2.
3.
4.
5.
Per-capita health expenditures lower
No health insurance coverage
Different diseases
Storability, transportation and administration of drugs different.
People may not purchase full dosage. As a result:
- long-run elasticity may be smaller than short-run elasticity
- externalities
Comparing the health sector
in low-income and developed economies
India
Pakistan
Canada
U.S.A.
4.9
23
17.8
82.2
82.2
4.1
18
22.9
77.1
77.1
9.1
2058
72.0
28.0
15.5
13.0
4499
44.3
55.7
15.3
Information on health expenditures
Total health expenditures as % of GDP
Per-capita total health expenditures ($)
Public health expenditures as % of total
Private health expenditures as % of total
Out-of-pocket expenditures as % of total
Top ten leading causes of burden of disease: all ages
India
U.S.A. and Canada
Cause
Cause
DALYs (000)
DALYs (000)
Acute lower respiratory infection
Ischaemic heart disease
24,806
2,955
Perinatal conditions
Diarrhoeal diseases
Ischaemic heart disease
23,316
22,005
11,697
Unipolar major depression
Alcohol dependence
Road traffic injuries
2,511
1,736
1,670
Falls
Unipolar major depression
10,897
9,679
Cerebrovascular disease
Osteoarthritis
1,651
1,029
1,017
996
Tuberculosis
Congenital abnormalities
7,578
7,454
Diabetes mellitus
Trachea/bronchus/lung cancers
Road traffic injuries
Measles
7,204
6,474
Dementias
Self-inflicted injuries
940
858
Shares of Major Therapeutic Segments in Retail Sales:
India versus the World Market
Share of retail sales (%)
Therapeutic segment
World: 2001
India: 2000
Rank
1
Share(%)
19.6
Rank
4
Share(%)
8.0
Central nervous system (CNS)
2
16.9
6
6.7
Alimentary tract and metabolism
3
15.3
1
23.6
Respiratory system
4
9.5
3
10.4
Anti-infectives
5
9.0
2
23.0
Musculo-skeletal
6
6.1
5
7.3
Genito-urinary
7
5.7
9
3.1
Cytostatics and immunosuppressants
8
4.0
13
0.1
Dermatologicals
9
3.3
7
5.6
Blood and blood-forming agents
10
3.1
8
3.9
Sensory organs
11
2.1
10
1.6
Diagnostic agents
12
1.8
12
0.1
Systemic hormonal products
13
1.6
11
1.5
Others including parasitology
.
2.3
.
5.4
Cardiovascular system
What we have done in the context of India
• Use detail product-level data from India to estimate key demand and
supply parameters: own and cross-price elasticities, expenditure
elasticities, marginal costs
• In the period covered by our data, India did not recognize
pharmaceutical product patents; hence many products available that
were under patent in the U.S.
• Carry out counterfactual analyses of what prices, consumer welfare, firm
profits would have been had patents been in effect
The Basic Thought Experiment
• Had patents been in effect, domestic products that were in violation of
patents would have to be withdrawn from the market
• Estimates of key demand and supply parameters can then be used to
simulate the effects of product withdrawal on prices, consumer welfare
and profits
Limitation of Analysis:
TRIPS applies only post-1995. Most of the products that will be affected
are still undergoing clinical trials in the U.S. and hence not in our data
Does not tell us what will actually happen, but only what would have
happened if patent laws had been enforced earlier…
Case Study: Indian Pharmaceuticals
Quinolone sub-segment of Systemic Anti-Bacterials
• Why India?
• Leading example of a low-income developing country that had not
recognized patents.
• Strong opposition to TRIPS.
• Disease profile of Indian population mirrors that of many low-income
countries.
• Domestic Indian pharmaceutical industry was as of 2002 the largest
producer of “generic” drugs (followed by Brazil). Market structure:
many small and medium-sized domestic firms selling drugs that are
patented elsewhere.
India’ s pharmaceutical sector
• Indian Patents Act (1970): excluded pharmaceutical product patents,
recognized process patents for a term of 7 years
• Leftward tilt in policy during 1970s: drug price controls, restrictions on
foreign equity shares, capacity expansion
• Liberalization since early 1990s
• Dramatic growth of Indian pharmaceutical industry in the last 30 years
 Indian firms are major exporters
 India the largest producer of formulations, by volume, and a leading
bulk drug producers
Production, exports, imports and domestic sales of
pharmaceutical formulations (Rs. Billions)
\s
(Rs. billions)
• \s
200
180
formulations
\s
(Rs. billions)
160
140
120
100
80
60
40
20
Output
Exports
Imports
Domestic sales
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
0
Production, exports, imports and domestic sales of
bulk drugs(Rs. Billions)
\s
(Rs. billions)
• \s
50
45
formulations
\s
(Rs. billions)
40
35
30
25
20
15
10
5
Output
Exports
Imports
Domestic sales
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
0
India’ s pharmaceutical sector
• Market structure also changed:
•
declining share of multinational subsidiaries from about 80% to 90%
in 1970 to roughly 30% in 2000
•
increase in number of firms in organized sector of the industry
(roughly 400 firms)
•
mushrooming of very small-scale units
Top 20 firms by domestic retail pharmaceutical sales in India
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
1971
Company
Sarabhai
Glaxo
Pfizer
Alembic
Hoechst
Lederle
Ciba
May & Baker
Parke Davis
Abbott
Sharp & Dome
Sudrid Geigy
Unichem
East India
Sandoz
Deys
Boots
T.C.F.
Warner Hindustan
John Wyeth
Origin
Dom
For
For
Dom
For
For
For
For
For
For
For
For
Dom
Dom
For
Dom
For
Dom
For
For
Year
1981
Company
Glaxo
Hoechst
Pfizer
Alembic
Geoffrey Manner
Burroughs Wellcome
Ranbaxy
Boots
German Remedies
Richardson Hindustan
Parke Davis
Warner-Hindustan
Roche
Merck, Sharp & Dome
Cynamid
Unichem
Cadilla
Standard
E. Merck
East India
Year
Foreign subsidiaries’ share of domestic retail
sales (%)
1970
75-90
Origin
For
For
For
Dom
For
For
Dom
For
For
For
For
For
For
For
For
Dom
Dom
Dom
For
Dom
2001
Company
Glaxo SKB
Ranbaxy
Cipla
Nicholas Piramal
Aventis
Sun
Dr. Reddy’s
Zydus Cadila
Knoll
Pfizer
Wockhardt
Alkem
Lupin
Novartis
Aristo
Pharma Marketing
Torrent
Alembic
Cadila Pharmaceutical
USV
1981
60-75
1991
49-55
Origin
For
Dom
Dom
Dom
For
Dom
Dom
Dom
For
For
Dom
Dom
Dom
For
Dom
Dom
Dom
Dom
Dom
Dom
2000
28-35
Systemic Anti-Bacterials (Antibiotics)
• Focus on the systemic anti-bacterials segment of the market, and within
that on the quinolones sub-segment
• Systemic anti-bacterials include all the original miracle drugs (for
treatment of bacterial infections) that sparked the development of the
research-based pharmaceutical industry as well as later generations of
molecules
• Systemic anti-bacterials accounted for about 20% of retail
pharmaceutical sales in India in 2000
• Systemic anti-bacterials segment divided into several sub-segments, each
representing a family of related molecules
Therapeutic
Categorization
All Pharmaceutical
Products
Cardiovascular
system
Tetracycline
Central
nervous
system
Macrolides
Ampicillin
Anti-infectives
Quinolones
Respiratory …
system
Cephalosporin
Others
Penicillin
Trimethoprim
Why Quinolones?
• Systemic anti-bacterials (i.e., antibiotics) important in a country where
infections a major cause of disease.
(life quality enhancing drugs, such as anti-depressants, Viagra, etc. will
presumably be affected more by patents; however, less important from a
public health policy perspective)
• Antibiotics important in terms of revenue share in Indian market (#2 in
revenues). Within antibiotics, quinolones one of the largest with 20.8
revenue share.
• Quinolones belong to the latest generation of antibiotics. Drug of choice
for most infections  there should be many substitutes available.
• Several quinolone products still under patent protection in the U.S.
• The idea behind our exercise: Imagine that patents had been enforced
in India as they were enforced in the U.S. Simulate market outcome
(we are not deriving the effects of actual TRIPS enforcement in 2005,
since most patents had expired by then). But can get a sense of what
the effects will be when important life-saving drugs are introduced in
the future….
Methodological Approach
• Explicitly model consumer and firm behavior.
• Detailed estimation of the demand and supply structures: own-price and
cross-price elasticities and bounds on marginal costs.
• Counter-factual analysis: price changes, consumer welfare losses, firm
profit gains/losses.
• Decompose consumer welfare losses into three components:
Loss of Variety Effect
-
Expenditure Switching Effect
+
Reduced Competition Effect
=
Consumer Welfare Loss of TRIPS
Key: Demand Estimation
Two-Stage Budgeting and AIDS:
•
•
•
•
Natural therapeutic categorization.
Flexible functional form and very general demand patterns.
Can be applied to both household-level data and market-level data.
Finite virtual prices
How we Define a “Drug”
• Data is highly disaggregate: information at the SKU level
• Differentiate along two dimensions: molecule and nationality.
• Aggregate over dosage forms: syrup, capsule, tablet; strength: 100 mg,
500 mg; packet size: 10 tablets, 24 tablets…
• Aggregate over domestic / foreign firms: Ranbaxy, Cipla; Bayer, Glaxo…
Two-stage Budgeting Approach
Anti-biotics
Penicillins
Macrolides
Ampicillin
Quinolones
Cephalosporins
Trimetho
-prim
Dom
Cipro
For
Cipro
Others
Dom
Spar
Dom
Norflo
For
Norflo
Dom
Oflo
For
Oflo
Some Interesting Patterns in the Data
• Four main molecules.
• Domestic share in each case SUBSTANTIALLY larger than share of
foreign subsidiaries.
• Domestic CIPRO by far the biggest (53% revenue share).
• Yet, prices of domestic products higher by ca. 10% (last 2 rows of the
table).
• Note also:
○ VERY large number of domestic firms.
○ Foreign firms violate patent laws in India while respecting them in
their own countries.
○ Products often introduced in India by Indian firms. Foreign firms
follow years later.
• General impression from these tables: Domestic drugs sell at a premium.
Distribution networks and ease of access a possible explanation.
The Quinolone Sub-Segment
Molecule
Share (%) of sales of
quinolones
Domestic
Foreign
firms
subsidiaries
Sales (Rs. millions): 2000
Domestic
firms
Foreign
subsidiaries
Ciprofloxacin
53.0
2.7
3,030
156
Norfloxacin
11.2
0.1
640
3
Ofloxacin
11.6
3.1
665
177
Sparfloxacin
10.8
0.1
620
4
U.S. or European
patent-holder
Year of U.S. patent
expiry
Year of US-FDA
approval
Year first introduced
in India
No. of domestic
Indian firms
No. of foreign
subsidiaries
No. of products of
domestic firms
No. of products of
foreign subsidiaries
Ciprofloxacin
Norfloxacin
Ofloxacin
Sparfloxacin
Bayer
Merck
Ortho-McNeil
Rhone-Poulenc
2003
1998
2003
2010
1987
1986
1990
1996
1989
1988
1990
1996
75
40
17
25
8
2
2
1
90
48
21
30
10
2
2
1
Sales weighted average price per-unit API of products produced by:
Domestic Indian
firms
Foreign subsidiaries
11.23
9.04
88.73
78.11
10.29
4.99
108.15
.
Main Demand Side Results
(drive counterfactual simulation results and conclusions)
• VERY large cross-price elasticities between domestic products with
different molecules. In fact, some domestic products appear to be closer
substitutes to one another, than domestic/foreign products containing
the same molecule
•  Domestic Products closer substitutes to one another than to foreign
products with the same chemical composition. Genuine empirical result.
• Possible Interpretation: Differences between domestic and foreign
products in the distribution networks. The retail coverage of domestic
firms (as a group) is much more comprehensive than that of
multinational subsidiaries.
• Quite possible therefore that local pharmacies more likely to have in
stock domestic products containing different molecules than they are
domestic and foreign versions of the same molecule.
In Sum:
• Consumers seem to prefer domestic to foreign products (as indicated by
the higher prices and higher market shares)
• Domestic Products close substitutes to one another
• Likely reasons for these patterns:
 Delay in launch of new products by multinational patent-holders in
India, possibly due to reference pricing.
 Lack of well developed distribution and marketing network by
multinationals.
• Important policy implication: loss of variety as a result of patent protection
likely to be substantial.
 New drugs will be available to Indian consumers with a delay
 Even after their launch, their distribution especially in rural areas may
be lacking
Evidence from other countries
• How general are the previous findings?
• No direct evidence on consumer preferences in other countries.
• However, evidence on firm/product entry from cross-country studies
• Two studies particularly relevant:


Lanjouw (2005): 68 countries of all income levels and drug launches
between 1982-2002
Danzon and Epstein (2008): 15 countries and drug launches in 12
different therapeutic classes between 1992-2003.
Evidence from other countries (continued)
• Main findings:



Launch timing is influenced by price regulation. If price regulation
reduces prices, it contributes to launch delay.
Global reference pricing is particularly important here: Manufacturers
delay launch in low-price countries to avoid undermining higher
prices in other countries. Hence, referencing policies adopted in highprice countries impose welfare costs on low-price countries.
Effect of IPR is ambiguous.
Implications for Welfare
• Consumer welfare losses are substantial: $400M when we remove all
domestic quinolones from the market -- 65% of the sales of the entire
anti-bacterial market in India in 2000.
• Consumer welfare loss not that large as long as some domestic
competition remains. Loss largest when all domestic products disappear
from the market.
 Driven by large cross-price elasticities across domestic products.
• A substantial fraction of the loss is due to the loss of product variety.
Again, driven by large cross-price elasticities for domestic products 
Corresponds to the case where strict price regulation would keep the
prices at their pre-TRIPS level.
• Conclusion: Price regulation alone not sufficient in order to mitigate the
loss suffered by Indian consumers.
Implications for Welfare (continued)
• Prices increase between 100% to 400%.
• Losses to domestic firms pale in comparison: $50M per year  May
explain why domestic firms have been divided regarding TRIPS.
• Moderate gains to foreign patent owners:
○ Without price regulation: $53M per year
o With price regulation:
$19.6 per year
o “developing a new drug costs $802M on average”, Mednews (2001).
• What is the fuss about? Global reference pricing?
Total Welfare Loss
Loss of variety:
compulsory licensing
Consumer
Welfare Loss
Substantial
price increase:
substitutes
within segment
important
+
Firm
Profit
Loss
Total effect
bigger than sum
of components
Increase in R&D unlikely
Foreign
Profit
Increase
Policy Implications
• Substantial loss of consumer welfare from reduction in product variety—
more than 50% of overall losses under most scenarios
• Suggests possible role for compulsory licensing
• Counterargument: If the value of product variety derives from coverage
of distribution networks and associated ease of access, this component
of consumer welfare loss may be transitory

with India recognizing patents, MNC subsidiaries may invest in
expanding distribution networks
• But: Incentives to undertake such investments will be small as long as
the market remains small. Price controls will further reduce incentives.
On Incentives…
• Sales of Bayer’s ciprofloxacin:
• Profit of Bayer’s ciprofloxacin:
• Total revenue in the ENTIRE
anti-bacterials segment in India:
$1.6 billion in 2000
$640 million in 2000
$610 million in 2000
Note: The profit calculation assumes as 40% markup, which is standard in
this industry.
Summary
Recent debate has focused on:
- prices
- R&D
Our research points to another issue  distribution and marketing of
EXISTING products
- over time
- over space (rural versus urban in particular)
Big Question:
Will patent enforcement change the supply side of the market?
•
•
Incentives of multinationals to invest in distribution networks and
marketing in developing country markets?
Joint Ventures with domestic firms? Licensing?
So, what has happened since 2005?
• Evidence is VERY preliminary
• Recent work by Arora, Branstetter and Chatterjee(2008):
 “Striking Increase in R&D intensity of Indian pharmaceutical firms”
 In particular:
 Increase in absolute R&D expenditures
 Increase in R&D intensity
 Increase in measures of research output
 Increase in stock market valuation of Indian firms’ R&D
investment.
So, what has happened since 2005?
• But:
 No evidence that the above developments are driven by “independent
innovations”
 No new products
 No evidence of R&D collaboration between Indian and Western firms
(with few exceptions)
• Question: Where did the R&D expenditures go?
• Answer:
 Process and not product innovations
 Sales of generic products abroad: Explosion of export activity
 Contract manufacturing
So, what has happened since 2005?
Big benefit of TRIPS:
It opened up to Indian firms foreign markets for TRIPS-legal imitations –
the generics market!
At the same time Oxford Analytica reports:
•
•
•
•
•
Complaints that price controls undermine the growth strategy of the
pharmaceutical industry.
In 2005 number of new drug products introduced to the domestic
market fell by 50%. Their contribution to total sales was just 1%.
Industry confined to re-processing of international drugs patented
before 1995.
Because of TRIPS, no new generic drugs.
Foreign firms are slow to introduce new products; little interest in its
small scale opportunities
Conclusion
• Indian firms seem to have benefited from TRIPS, but mainly through
exports
• Risks to Indian consumers real.
• More research into the distribution of new drugs in India and other
developing countries necessary.
 How soon?
 Do they reach poor rural areas?
 Perhaps the hardest question: Welfare implications of distribution:
Do we want new drugs to reach uninformed consumers?
THANK YOU!