Transcript Regulating Drug Prices with Ramsey Pricing Principles

Regulating Drug Prices with
Ramsey Pricing Principles
Robert Kemp, PhD (Econ)
Associate Professor
College of Pharmacy, Touro University
College of Pharmacy
Touro University-California
1. Regulating Drug Prices with
Ramsey Pricing Principles
Introduction
Regulating Drug Prices with
Ramsey Pricing Principles
1.
2.
3.
4.
5.
6.
Introduction
Regulation of pharmaceuticals
Ramsey optimality
Ramsey principles for price setting
Conclusion
Appendix (Cost-effectiveness analysis
cannot and should not be used to set
prices.)
II. Regulation of Pharmaceuticals
• Regulation of which drugs can be
prescribed; marketing regulation.
• Regulation of prescribing practice.
• Determination of expenditures in a
therapeutic area, i.e. anti-diabetic
agents.
• Determination of price ceilings.
• Determination of level of reimbursement or
subsidy.
Institutional settings for health
economics decisions affecting drugs
1. Allocation of resources among providers
1.
2.
3.
4.
Insurance
Medical care
Hospitals
Pharmaceuticals
2. Allocation of resources among therapeutic
areas or disease states.
3. Allocation of resources by controlling access to
drugs in a therapeutic area.
4. Allocation of resources by controlling price for
pharmaceuticals within a therapeutic area.
Regulating Drug Prices
• Prices of drugs are regulated in many countries,
including the USA. The methods include
– Reference pricing, ceiling prices, subsidization, rate of
return, “AMP+FUL”, “Best price”, etc.
• There are also restrictions that affect level of
reimbursement or access
– Clinical recommendations, prior authorization, and
other administrative rules.
– Cost-effectiveness analysis or another form of
economic evaluation (programme evaluation)
threshold
Economic efficiency
• None of the rules of administration seem to be
based on economic efficiency criteria, or
microeconomic principles.
• The relationship between economic evaluation
and pricing is not stately explicitly.
• Ramsey pricing, or the inverse elasticity rule,
has been suggested (and sometimes
institutionalized) as the rationale for pricing in
many industries including the pharmaceutical
industry.
Ramsey Price Regulation
• As a guiding principle or “Rule of Reason”
– Microeconomics replaces regulatory precedent as
foundation of regulation. In 1970’s economists
showed the inefficiency of existing price regulations in
transportation, energy, public utilities, and other
industries.
– New insights into sustainability of monopoly, optimal
departures from marginal cost pricing, optimal
commodity taxation, and multi-part pricing, etc.
– Principles of marginalist approach were widely
accepted but few attempts to base positive regulatory
approach to assure performance under regulation.
“Informational constraints” the usual reason cited for
not pursuing Ramsey optimality as new “rule of
reason”.
Regulating Drug Prices with
Ramsey Pricing Principles I
• Hypothesis I:
– If there was a monopsonist who wanted to set prices
on drugs for a population within a therapeutic area,
then using Ramsey principles would be an
improvement over economic evaluation techniques
and over capricious administrative decisions based
on some notion of cost-containment or solely on the
basis of clinical effectiveness.
Regulating Drug Prices with
Ramsey Pricing Principles II
• Hypothesis II:
– If there was a monopsonist who wanted to set prices
on drugs for a population across therapeutic areas,
then using Ramsey principles would be an
improvement over economic evaluation techniques,
and over capricious administrative decisions based
on some notion of cost-containment, or solely on the
basis of clinical effectiveness.
Price regulation of pharmaceuticals
• Relevant market to be considered for
pricing
– Population in a Geographic area
– Therapeutic area (examples diabetes,
hypertension) or disease state.
• Should regulate relative prices within the
market and not worry about so called “silo
effects”.
Normative goals through pricing
• For pharmaceutical companies
– Innovation: New drugs that are “priority
NMEs” are given higher prices (or higher
subsidies).
– Reward for performance: A drug that
produces better outcomes in a therapeutic
area is given higher prices. (Is this reflected in
elasticity of demand?)
– Costs can never be used as sole justification
of prices, however prices need to reflect the
long-run average costs of drug companies.
Normative goals through pricing
• For health authority or third-party payer
– Deadweight loss due to monopoly is minimized:
agencies assign second-best prices on drug
commodities.
– Deadweight loss due to monopsony is minimized:
agencies assign optimal taxes on drug commodities
(subsidies are negative taxes).
– Payers act as financial agent of the patient insuring
that patients are given access to the drugs that
improve their health. (Play volume game.)
Normative goals through pricing
• Preserve the clinical imperative: Medical
expert (clinical agents) look after patients
insuring that patients are given access to
the drugs that improve their health.
• Relate the evaluation of clinical
effectiveness and the differential prices of
drugs in a therapeutic area.
• (Consumers get the drugs they are willing
to pay for?)
III. Ramsey optimality
Ramsey Pricing
For simplicity, suppose that the demand functions for the
n produced drugs in a therapeutic area (q1, q2, ..., qn) are
independent:
qi = Di(pi), for i = 1, 2, ..., n.
The total revenue available for the drugs is
R(q1, ..., qn) = p1(q1)q1 + ... + pn(qn)qn.
Ramsey Pricing
The cost of producing these goods is C(q1, ..., qn).
The regulator wants to choose outputs so as to
maximize consumer surplus,
qi
pi(t)dt - C(q1, ..., qn),
0
Ramsey Pricing
subject to the constraint that revenue exactly equals
cost (or that profit is a given constant). The firstorder conditions are
pi - Ci = ‫( ג‬Ri - Ci)
for i = 1,... ,n, where Ci and Ri are the partial
derivatives of C and R with respect to qi and ‫ ג‬is the
Lagrangian multiplier on the constraint.
Ramsey Pricing
This condition may be rewritten as
pi – Ci
pi
=
ki
εi
where k = ‫ ג‬/(1 + ‫ )ג‬and εi is the elasticity of demand for qi.
That is, the price markup over marginal cost, (pi - Ci)/pi, is
inversely proportional to the price elasticity of demand for
that good. If k = 1, this condition is the standard monopoly
price-discrimination condition. If k = 0, this condition is the
same as in competition.
Ramsey Pricing
• Ramsey Pricing is "second-best" pricing,
whereby drugs for patients with inelastic
demands pay a higher markup over
marginal cost than those with more elastic
demands.
• The basic goal of Ramsey Pricing is to
recoup the fixed costs from those patients
who have the fewest alternatives, while
minimizing the distortion associated with
prices in excess of marginal costs.
Ramsey Prices and Taxes
• In order for the rules of optimal relative
prices and optimal taxes to be applied, we
must set out a price for one good as a
“numeraire” upon which all other relative
prices are set.
• This is the major normative decision to
make.
IV. Ramsey principles for price
setting
Where do you get the εi ?
• Based on clinical imperative?
– from prescribers preference and derivation of
an index based the order preference of
expert medical doctors?
– based on a combination of epidemiology, best
practice, evidence-based medicine, clinical
effectiveness?
– Utility values?
– Based on consumers willingness to pay?
• Based on expenditure cap in a therapeutic
area?
Price Regulation of Drugs
• Where do you get the Ci?
– In other industries prescriptive accounting
rules are used with an attempt to find long-run
MC (Kahn, Baumol).
– Cost should capture relevant R&D costs, so
that priority NME’s are rewarded.
– Majority of marginal costs in USA are
marketing costs.
2004 Revenue Allocation for Top 7 US Pharmaceutical Cos
Marketing, Advertising
and Administration
Other
32%
36%
18%
14%
Research & Development
Profits (net income)
Source: Families USA, The Choice: Health Care for People or Drug Industry Profits, 2005
Price Regulation of Drugs
• What about the budget (or profit)
constraint?
– Set it arbitrarily to company at industry
average in fully allocated model (like in UK).
– Effectively, the amount allocated to a
therapeutic area from a monopsonist will
determine the revenue that can be taken from
the market by a particular company for a
particular drug.
PROFIT CURVE
Initial pricing
• The most important use of Ramsey
principals maybe in setting the initial price
upon entry.
– There is evidence of inherent inverse
elasticity in pricing.
US launch price determined by
demand-side characteristics
FDA level of Ratio of Median Price of
therapeutic Entrants to Existing Drugs
advance
Acute
Chronic
Important
2.97
2.29
Modest
1.72
1.19
Little
1.22
0.94
Source: Lu and Comanor
“Strategic Pricing of New Pharmaceuticals”
Me-toos
• Entry of a me-too into a market would
increase ε for both the incumbent product
and the rival entrant.
Set prices on a yearly basis
• Ex ante price for new drugs based on clinical
outcomes and doctor’s preferences.
• Ex post review of existing drugs (like France).
• Econometric test for Ramsey optimality in
existing pricing structure.
• (Ramsey-like monopolistic price discrimination
was found in rail prices in sustainably
monopolized markets prior to rail price
regulation (Dumas, “Weak Invisible Hand”). In
postal prices prior to entry by other carriers.)
What about generics?
• In theory, Ramsey optimality conditions
hold as monopoly goes to rivalry, or
competition through entry. (see
Breutigam,1984, or Miller, 2007).
– (As k becomes zero, we approach a
competitive solution where p = MC.)
Subsidy or price regulation?
• What about supply induced demand?
– With a monopsonist and a patented monopoly
it doesn’t matter whether you use Ramsey
optimal taxes (subsidies) or Ramsey optimal
prices. There is no difference between a price
regulation and a tax in this case (Baumol and
Bradford, 1970).
Ramsey optimal pricing across
therapeutic areas
• Ramsey optimality is sustained across a
class of firms with different elasticities and
cost structures that provide the same
output, known as the “intra-modal case”
taken from transportation pricing.
• Once one has a numeraire price in one
market, relative prices (optimal) within a
market and relative prices (optimal) across
modes can be found.
Conclusion
• Ramsey optimality can be used to set a “rule of
reason”; or the theoretical foundation for drug
prices in a therapeutic area.
• “Informational problems” will exist in an attempt
to institutionalize the rule of reason, however,
having an efficiency criteria guiding pricing may
lead to improvements in allocation over existing
administrative rules.
Ramsey Optimality
• Ramsey optimality can be set up to give the
essential relationship between the value of the
drug therapy to the patient and the optimal
prices. Such a valuation of output can be based
on expert clinical imperative, community-based
values (utilities), or WTP.
• Costs can be explicitly assigned by the
monopsonist, and can take into account the
effect of the adaptation of new technologies on
capital cost.
Ramsey Optimality
• Programme evaluation techniques assume that prices
are exogenous to the evaluation and should not be used
to set prices.
• Clinical imperative should be preserved: Outcome
afforded by the technology is substantiated on clinical
grounds and built into the pricing regulation.
• Ramsey optimality does not have the artificial disjoint
between a “budget impact” analysis and costeffectiveness analysis.
• The “margin” for which “incremental CEA” is targeted is
often undefined. The Ramsey optimal formulation
begins with a budgetary constraint and so defines the
optimality conditions.
Ramsey optimal commodity taxes
• Model of single buyer (monopsonist)
• Goal is to find the least-distorting pattern of
taxes
• Assumes infinite supply elasticity
• Results:
– optimal taxes vary between goods such that tax rates
are proportional to the elasticity of demand
– Reduces compensated quantity demanded for each
good by the same percentage relative to pre-tax
demands
• Baumol and Bradford (1970) show that Ramsey
rule is equivalent to inverse elasticity rule for
monopolist if there are no cross price effects
Ramsey Optimal Taxes
Average Investment per Stage of
Development
Per Compound Including Failures
250
250
175
200
150
100
51
33
50
31
0
Discovery
PC-Devel
Phase I
Phase II
Phase III
Note: Does not include Capital Investments, Time Value of Money, or Risk Free Return on Capital.
Source: IMS Internal Analysis
Average Investment for Stage of
Development
Per Compound Excluding Failures
33
35
30
25
20
14
15
10
11
5
5
5
0
Discovery
PC-Devel
Phase I
Phase II
Phase III
Note: Does not include Capital Investments, Time Value of Money, or Risk Free Return on Capital.
Source: IMS Internal Analysis
Revenue from Blockbuster RX
Millions
US$
USA
Ex-USA
800
700
600
500
400
300
200
100
0
LAUNCH
2
4
6
Patent years after launch
8
OP2
Years off patent
Prescriptions of Blockbuster RX
Prescriptions
1,000
180
160
140
120
100
USA
Ex-USA
80
60
40
20
0
LAUNCH
2
4
6
Patent years after launch
8
OP2
Years off patent
NCE Lifecycle Sales Analysis
Average Percent of Peak Sales Potential
96%
100%
Hatch
Waxman
Exclusivity
80%
96%
95%
100%
83%
75%
76%
71%
64%
60%
53%
41%
40%
46%
32%
21%
20%
6%
0%
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15
Years After Launch
Source: IMS (Based on 816 NCE’s launched since 1983)
Cost-effectiveness analysis cannot
and should not be used to set
prices .
Economic evaluation as prescriptive
regulatory methodology
• NICE, PBAC, COHHTA, (various rather
poorly defined attempts in the USA)
• A “clinical recommendation” is based on
cost-effectiveness or cost-utility analysis
based upon an acceptable level of ICER,
and, maybe
• A pricing negotiation follows this
recommendation. What criteria are used in
pricing?
Economic evaluation as prescriptive
regulatory methodology
• Clinical imperative: Outcome afforded by
the technology is substantiated on clinical
grounds.
• Can one method of evaluation be used in
all cases? History of public policy analysis
says probably not!
Economic evaluation as prescriptive
regulatory methodology
• Major effect of programme evaluation
techniques in the marketplace for dugs is
to set restrictions on access to drugs for
specific populations.
– This is due to the way that the analysis of
health outcomes is structured and what
method of evaluation are employed.
Basing pricing on economic
evaluation?
• Programme evaluation techniques assume
that prices are exogenous to the
evaluation. Cannot use these techniques
to set prices in any non-arbitrary way.
– Common strategy for the drug company is to
fully specify modeled analysis, except for the
price of the drug, set ICER = 0 and solve for
the price of their drug.
Basing pricing on economic
evaluation?
• Economic evaluation includes costs that
are irrelevant to the differential pricing of a
product in a therapeutic area.
– A “cost-effective” drug may be one that lowers
costs in another therapeutic area or lowers
cost in another form of service (keeps you
away from the doctor).
Basing pricing on economic
evaluation?
• What should be done about priority NMEs?
What is the comparator if any (“best medical
practice”)?
• How did prices get at the current level?
Grandfathering? Cross-subsidization?
• How do we price a priority NME?
• How do we make pair wise ICERs give us an
unambiguous answer when there are more than
two drugs in the therapeuitc area?
Prices and inter-temporal
allocations
• Current prices and interest rates may be
rather poor parameters to use in allocating
resources over time.
• Using Markov models or other sorts of
modeling to set current prices of drugs
requires considerable imagination.
Australia
• Guidelines for submission to the PBAC is
a cookbook for the methodology to be
used.
• After PBAC recommendation, then to the
Pharmaceutical Pricing Authority:
– One clear pricing rule: Cost-minimisation
analysis probably will require new drug to be
priced at lower level than incumbent drugs in
the therapeutic class.
– Otherwise there is no clear rationale except
for historical precedent of pricing
France
• The French system with the Transparency
Commission and Pricing Authority has
sought to make the link between price and
the value of therapeutic improvement from
a drug.
• May have some Ramsey optimality
principles built into it.
France
Budget constraint
Reimbursement level
• Pi(Q)i = Expenditure
• set by Transparency
Commission based
• Government sets
on level of therapeutic
price ceiling for
improvement
ambulatory care
– 100%
• Volume agreement
– 75%
set by epidemiological
– 50%
estimate.
– 25%
• If Qi is exceeded then
– 0%
Pi is decreased.