Journal of Health Economics 21 (2002) 699–708

Should the US allow prescription drug reimports

from Canada?
Paul Pecorino∗
Department of Economics, Finance and Legal Studies, University of Alabama, P.O. Box 870224,
Tuscaloosa, AL 35487-0224, USA
Received 31 May 2001; accepted 14 February 2002

Abstract
As a result of public outrage over lower prescription drug prices in Canada, Congress passed
legislation that would allow these drugs to be imported into the US. The lower Canadian prices
reflect price regulation. Opponents of allowing these imports have argued that the US will import
Canadian price controls and that profits of pharmaceutical companies will be hurt. In this paper,
a model is developed in which a good sold in the foreign country is subject to a negotiated price
which is determined in a Nash bargaining game. When imports back into the home country are
allowed, this negotiated price also becomes the domestic price. This causes the home firm to make
fewer price concession in the Nash bargaining game. Home firm profits are found to rise under
the reimport regime for both of the demand functions analyzed in this paper. © 2002 Elsevier
Science B.V. All rights reserved.
JEL classification: I18; F1

Keywords: Prescription drugs; Reimports; Price controls; Canada

1. Introduction

As a result of public outrage over lower prescription drug prices in Canada, Congress
passed legislation, later signed into law, which would allow these drugs to be imported into
the US.1 Though the Secretary of Health and Human Services later declined to implement
the law, saying that it was unworkable as written, the issue of whether drug reimports from
∗ Tel.: +1-205-348-0379; fax: +1-205-348-0590.

E-mail address: ppecorin@cba.ua.edu (P. Pecorino).

1 See the Chemical Market Reporter, 30 October 2000, p. 15 for a report on the legislation. There have been

many articles in the popular press about prescription drug prices and the fact that they are higher in the US than
in other countries. See, among others, Newsweek, 25 September 2000.

0167-6296/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 7 - 6 2 9 6 ( 0 2 ) 0 0 0 3 5 - 8
700 P. Pecorino / Journal of Health Economics 21 (2002) 699–708

Canada should be allowed remains an important policy question.2 Drug reimports from
Canada have drawn the determined opposition of the pharmaceutical industry.3 The price
differential between Canada and the US is not due to price discrimination as that term is
commonly used. Rather, it reflects the single payer system in Canada under which provincial
governments can exercise monopsony power in setting prices for prescription drugs (see
Anis and Wen (1998, p. 22)). Thus, one argument against the import of drugs from Canada
is that it will, in essence, lead to the import of Canadian price controls, erode firm profits
and reduce incentives for innovation.
In this paper, a simple partial equilibrium model of trade is developed in which there
is a home country monopolist which can potentially sell in the home and foreign mar-
ket. In the foreign market, price determination is modeled as a Nash bargaining game
between the firm and the foreign government. In the home market, the monopolist is
free to charge its profit maximizing price. Under the first regime analyzed, reimport into
the home country of goods sold in the foreign country is not allowed. Under the second
regime, reimport is allowed, so that the negotiated foreign price becomes the domestic price
as well.
Allowing reimports back into the home country will cause the domestic monopolist to
bargain harder in the Nash bargaining game.4 As a result, profits earned by the monopolist
may not fall when reimports from the foreign country are allowed. Under linear demand,
it is shown analytically that the monopolist’s profits always rise when reimport is allowed
back into the home country. This result continues to hold for all of a large number of
parameterizations of a constant elasticity demand function. Allowing reimports always
raises consumer surplus in the home country. Thus, in all of the cases analyzed, home
welfare is higher under the reimport regime.

2. The no reimport (NR) regime

A firm in the home country produces a commodity Q which can be thought of as a

pharmaceutical product. Since the firm has patent protection on this product, it acts as a
monopolist. The development process is not modeled in this paper, but presumably the firm
bore large sunk costs of development in bringing this product to market. The marginal cost
of production is the constant C. The good may be sold in the home and foreign markets.
The exchange rate is constant and normalized to one. Assume that the two markets are at
a similar level of development, so that except for a scaling parameter k > 0, demand is
the same in the home and foreign market. Thus, in the home and foreign markets demand

2 In declining to implement the law, Donna Shalala cited safety concerns in the bill as written (Wall Street

Journal, 27 December 2000, p. A2). I will not address safety concerns in this paper, but will instead assume that
these could be addressed by appropriately formulated legislation.
3 As the bill was taking shape Jeff Trewhitt, a spokesman for the Pharmaceutical Research and Manufacturers

of America remarked, “this was a bad idea before and it’s still a bad idea.” (Washington Post, 5 October 2000,
p.A4).
4 While the mechanisms are not quite the same, the intuition developed in this paper bears some resemblance

to the manner in which most favored customer clauses can give a firm a credible commitment to bargain harder
with future customers. See Shapiro (1989, p. 394) and the references cited therein.
 P. Pecorino / Journal of Health Economics 21 (2002) 699–708 701

is given by Q(P) and kQ(P∗ ), respectively, where Q (P) < 0. Profits on domestic sales
are Π (P ) = (P − C)Q(P ). The assumptions on cost and demand ensure that profits on
foreign sales are given by kΠ (P ∗ ) = k(P ∗ − C)Q(P ∗ ). This is a partial equilibrium
analysis and any income effects on demand stemming from changes in the trade policy
regime are ignored.
The assumption that, except for a scaling factor, demand is the same in the home and
foreign countries can be justified by the fact that the legislation passed by Congress only al-
lowed for reimports from the EU, Canada, Japan and Australia (Chemical Market Reporter,
30 October 2000, p. 15). All of these are highly developed countries with similar levels of
income.
Since reimports of the good back into the home country are not allowed, perfect mar-
ket segmentation is possible under the NR regime. Because of the assumption made on
demand, the monopolist would choose to charge the same price in both markets if it
were free to set prices in the foreign country. Instead prices are negotiated in the for-
eign market in a Nash bargaining game. Thus, the firm earns monopoly profits Π M in
the home market and the profits kΠ (P∗ ) in the foreign market, where P∗ is a negotiated
price. In addition, the negotiated price determines consumer surplus in the foreign country,
kCS(P∗ ).
The negotiated price must exceed the constant marginal cost of production C, oth-
erwise the domestic firm is better off reverting to its threat point of no sales under
which it earns zero profits. We will also assume that the domestic firm commits to
meet foreign demand, whatever its level, at the negotiated price. Since P ≥ C, this
assumption plays no role in this section, but will be of some importance in Section 3,
where goods sold in the foreign country can flow back over the border into the home
country.5
The foreign government would like to maximize consumer surplus in its country, while
the monopolist would like to maximize profits from sales in the foreign market. In the
absence of an agreement, profits and consumer surplus are both zero.6 Thus, zero is the
threat point for both the domestic firm and the foreign government. The Nash bargained
price is found from the solution to7

Max
∗
[kCS(P ∗ )]α [kΠ (P ∗ )]1−α , (1)
P

5 Presumably Canadian politics would not allow US pharmaceutical firms to create shortages of prescription

drugs in Canada.
6 This assumes that the foreign country can prevent its citizens from purchasing at the monopoly price in the

home country in the event a bargain is not reached. If this were not so, the domestic monopolist could refuse to
bargain with the foreign government and simply sell at the monopoly price to foreign citizens who make purchases
in the home country. Since Canadian prices are below US prices for prescription drugs, the assumption seems
justified.
7 In the original Nash bargaining game, bargaining power is set to 1/2 for each party. Eq. (1) is the standard

generalization of Nash (1953). Maggi and Rodrı́guez-Clare (1998) use the Nash bargaining game to model nego-
tiations between the government and an industry lobby which seeks trade protection. Wes (2000) uses the Nash
bargaining game to model bargaining between an upstream and downstream firm in the context of an international
trade model.
702 P. Pecorino / Journal of Health Economics 21 (2002) 699–708

where α reflects the bargaining power of the foreign country. The first-order condition to
this problem implies
αΠ (P ∗ ) (1 − α)Π  (P ∗ )
∗
= , (2)
CS(P ) Q(P ∗ )

where the prime superscript denotes a derivative.8 Let the price which solves (2) be denoted
∗ .9
PNR
The results here depend in a very obvious way on α. If α = 1, then we must have P = C,
where Π (C) = 0. This reflects that fact that the foreign government has all the bargaining
power when α = 1. If α = 0, we must have Π  (P ∗ ) = 0, which means that the home firm
charges the monopoly price in the foreign market.
Under the NR regime, total profits for the monopolist are
∗
ΠNR
Total
= Π M + kΠ (PNR ). (3)

Later in the paper, the expressions in (2) and (3) will be used to help compare firm profits
under the reimport (R) and NR regimes.

3. The reimport regime

3.1. The Nash bargaining game

Under the R regime, the negotiated price in the foreign country becomes the price in the
home country as well, due to the ability to reimport the good. There are no transportation
costs, so we assume that the law of one price holds for the good in question. Price concessions
by the domestic monopoly in the Nash bargaining game are much more costly than in the
NR regime, because they affect the domestic market as well as the foreign market. As a
result, we should expect the domestic firm to drive a harder bargain in negotiations with the
foreign government. While this will tend to help firm profits, the lower price in the domestic
market under the R regime will tend to hurt firm profits. Thus, the overall effect on firm
profitability appears to be ambiguous.
The home firm’s surplus from being able to sell in the foreign market is (1 + k)Π (P ∗ ) −
Π , where (1 + k)Π (P ∗ ) reflects profits in both the home and foreign market when
M

reimports are allowed and Π M reflects the threat point of only selling in the home market.
Thus, the price derived from the Nash bargaining game solves
Max
∗
[kCS(P ∗ )]α [(1 + k)Π (P ∗ ) − Π M ]1−α . (4)
P

Eq. (4) is identical to Eq. (1), except that the −Π M term is implicit in Eq. (1). Without
reimports, profits from home sales are always Π M , whether or not a bargain is reached. As

8 Recall that CS (P ∗ ) = −Q(P ∗ ).

9 If the bargaining set is convex, it will ensure a unique solution to the bargaining problem. It can be shown
that the bargaining set is convex as long as demand is not too convex. The bargaining set is convex for both of the
specific demand functions considered in this paper.
 P. Pecorino / Journal of Health Economics 21 (2002) 699–708 703

a result, Π M is netted out of the expression in Eq. (1). When reimports are allowed, profits
on home sales depend upon the bargain reached in the Nash game. As a result, Π M is not
netted out of the expression in Eq. (4).
The first- and second-order conditions from the maximization problem in (4) imply,
respectively
α(Π (P ∗ ) − Π M /(1 + k)) (1 − α)Π  (P ∗ )
= , (5)
CS(P ∗ ) Q(P ∗ )


ΠM
− αQ (P ∗ ) Π (P ∗ ) − − Q(P ∗ )Π  (P ∗ )
(1 + k)
+ (1 − α)Π  (P ∗ )CS(P ∗ ) ≡ Γ < 0. (6)

The expression (6) will be denoted by Γ in the comparative statics which follow.
Denote the price determined from (5) as PR∗ . The profits of the domestic firm are

ΠRTotal = (1 + k)Π (PR∗ ). (7)

The price P∗ depends upon the bargaining power and size of the foreign country (α and k)
in an intuitive way. Total differentiation of (5) reveals10
dP ∗ Q(P ∗ )[Π (P ∗ ) − Π M /(1 + k)] + Π  (P )CS(P )
= < 0, (8)
dα Γ
dP ∗ αQ(P ∗ )Π M
= < 0. (9)
dk Γ (1 + k)2
The negotiated price is decreasing in both the bargaining power and size of the foreign
country. When the foreign market is large relative to the home market, the firm is hurt
relatively less in the home market by price concessions. As a result, the negotiated price is
lower.
Not only will an increase in foreign bargaining power lower the negotiated price, it will
lower home firm profits since dΠ/dα = Π  (P ∗ )dP ∗ /dα < 0. (Note: Π  (P ∗ ) > 0 if
P ∗ < P M .) An increase in the size of the foreign market has an ambiguous effect on the
profits of the home firm. Differentiate Eq. (7) to get

dΠRTotal dP ∗
= Π (P ∗ ) + (1 + k)Π  (P ∗ ) .
dk dk
The first term in this expression is positive, reflecting the greater potential sales in a larger
market, while the second is negative reflecting the fall in the negotiated price as k rises. The
overall effect is ambiguous.

10 The numerator of (8) is always positive. The first term in brackets is the surplus earned from sales in the foreign

market divided by 1 + k. This must be nonnegative and will be strictly positive for 0 < α < 1. In addition Π  (P∗ )
will be nonnegative over the relevant range and strictly positive for α < 1.
704 P. Pecorino / Journal of Health Economics 21 (2002) 699–708

3.2. A note on policy evaluation

Typically in a partial equilibrium setting such as this, the welfare effects of changes in
policy are evaluated by summing the changes in consumer surplus, producer surplus and
government revenue. This welfare standard may be problematic here, since the (unmodeled)
innovation of new drugs may be adversely affected if firm profits fall in the move to the R
regime. The effect on consumer surplus of moving to the R regime is unambiguous. Under
the NR regime, domestic consumers pay the monopoly price, while in the R regime, they
pay the monopoly price only in some limiting cases. In all other cases, they pay a strictly
lower price. As a result, consumer surplus either rises or is unchanged in the move to the R
regime. Thus, if firm profits rise in the move to the R regime, home welfare unambiguously
rises. If home profits fall in the move to the R regime, it is difficult to make a definitive
welfare statement in the absence of an explicit model of innovation. For all of the results
which follow, firm profits either rise or are unchanged in the move to the R regime. Thus,
an unambiguous welfare comparison is made in each case.

3.3. Limit results

A policy evaluation requires comparing profits under the R and NR regimes (compare
(7) with (3)). With a general demand function Q(P∗ ), we can only make this comparison
for limiting values of α. When α = 1, the foreign government has all the bargaining power,
and the domestic firm is unable to obtain surplus from sales in the foreign country under
either regime. From Eq. (2), under the NR regime, the domestic firm is forced into marginal
cost pricing in the foreign country and earns zero profits from sales there. Total profits are
ΠNRTotal = Π M , the monopoly profit from sales in the domestic market. From (5), under the

R regime with α = 1, we again have total firm profits ΠRTotal = (1 + k)Π (PR∗ ) = Π M .
Thus, firm profits are equal under the two regimes. However, the price to home consumers
is lower under the R regime, since they now pay the price negotiated in the foreign market.
As a result, consumer surplus and home welfare both rise in the move to the R regime. This
surprising outcome is summarized as Result 1.

Result 1. When all bargaining power resides with the foreign government (α = 1), home
welfare is higher under the R regime.

When α = 1, the domestic firm earns zero profits on foreign sales and so is already
pinned to its minimum profit level Π M . Under the R regime, the negotiated foreign price
must exceed C by enough to make up for the firm’s lost profits on domestic sales.
Now suppose that all bargaining power lies with the domestic firm so that α = 0. Eqs. (2)
and (5) both require that Π  (P ∗ ) = 0 in order to be satisfied when α = 0. This implies
that the domestic firm earns monopoly profits in both markets (= (1 + k)Π M ) under both
regimes. Thus, firm profits and consumer surplus are both unchanged when we move from
the NR regime to the R regime. This analysis is summarized as Result 2.

Result 2. When all the bargaining power resides with the domestic firm, home welfare is
the same under the NR and R regimes.
 P. Pecorino / Journal of Health Economics 21 (2002) 699–708 705

Thus, for extreme values of the bargaining parameter, the model gives unambiguous
answers about the effects from moving from an NR regime to an R regime. When the
foreign country has all the bargaining power (α = 0), home welfare is higher under the R
regime. When the home firm has all the bargaining power (α = 1), home welfare is the
same under the two regimes. We are unable to make general statements about the effects
on home welfare from moving to the R regime for intermediate values of α. To gain further
insight into this question we analyze specific demand functions in Section 4.

4. Comparing the R and NR regimes with specific demand functions

In order to further explore the properties of the model, we will conduct an analysis using
linear demand in Section 4.1 and constant elasticity demand in Section 4.2.

4.1. Linear demand

Assume that the demand function takes the form

QD = a − bP. (10)
When (10) is combined with Eqs. (2) and (3) it is possible to obtain the following solution
for profits under the NR regime.11
(a − bC)2
ΠNR
Total
= (1 + k − a 2 k) (11)
4b
Combining (10) with Eqs. (5) and (7) yields the following solution for profits under the R
regime.

(a − bC) 2
(1 + k)(1 − α 2 ) + 2α
ΠRTotal =
8b
 
(1 − α)2 + k(1 + α)2 
+ (1 − α)(1 + k) . (12)
(1 + k)

Compare (12) to (11) to see that when we have the extreme values α = 0 or 1, profits are
equal under the two regimes.
A move to the R regime from the NR regime raises firm profits if and only if ΠRTotal −
Total > 0. Subtract (11) from (12) and simplify to find that the sign of Π Total − Π Total is
ΠNR R NR
the same as the sign of
4α 2 (1 − α)2 k ≥ 0, (13)
where the strict inequality holds for 0 < α < 1. For all interior values of α, ΠRTotal −ΠNR
Total >

0, i.e. firm profits are always higher under the R regime than the NR regime.
11 Details of the calculations are available upon request.
706 P. Pecorino / Journal of Health Economics 21 (2002) 699–708

Allowing reimport back into the home country causes the domestic firm to bargain harder
in the Nash game which determines the price of the good it sells. As a result, profits rise
on foreign sales. However, profits fall on domestic sales, as allowing for reimports pushes
the domestic price below the monopoly price. In the case of linear demand, the first effect
outweighs the second and firm profits always rise for interior values of ␣. For α = 1, firm
profits are unchanged under the R regime, but domestic consumers pay a lower price. Thus
domestic welfare is higher in this case as well. This analysis is summarized as Result 3.

Result 3. In the case of linear demand, firm profits are higher under the R regime for
0 < α < 1, while home welfare is higher for 0 ≤ α < 1. For α = 1, home welfare is the
same under the R and NR regimes.

4.2. Constant elasticity demand

Let demand take the form

QD = zP−ε , (14)

where ε reflects the elasticity of demand and z is a scaling factor. Combining (14) with (2)
and (3) it is possible to derive an analytical solution for profits under the NR regime.
ΠNR
Total
= zC1−ε (ε − 1)ε−1 [ε −ε + k(1 − α)(ε − α)−ε ]. (15)

Combining (14) with (5), it is possible to obtain the following equation for the R regime:
αC 1−ε ε −ε (ε − 1)ε
P ∗−ε [P ∗ (ε − 1) − (ε − α)C] = . (16)
(1 + k)
It is not possible to solve for P∗ analytically from (16), so numerical methods have been
used instead. The solution for P∗ is then used to solve for ΠRTotal .
The baseline parameterization of the model is C = 5, α = 0.5, k = 0.1 and ε = 2.
The value of k = 0.1 makes the foreign market 10% of the size of the home market. This
seems approximately correct for the case of the US and Canada. In the 1st, 3rd and 5th
column of Table 1, one of the parameters is allowed to vary while the others are held at their
baseline values. Columns 2, 4 and 6 report %Π = [ΠRTotal − ΠNR Total ]/Π Total , which is
NR

Table 1
Numerical results under constant elasticity of demand
ε %Π α %Π k %Π

1.25 0.41 0.05 0.01 0.1 0.84

1.5 0.62 0.10 0.02 0.2 1.31
2.0 0.84 0.30 0.25 0.5 1.87
4.0 1.09 0.50 0.84 1 1.91
10.0 1.20 0.70 1.89 2 1.57
20.0 1.24 0.90 2.37 5 0.92
50.0 1.26 0.95 1.66 10 0.53
The baseline values of the parameters are C = 5, α = 0.5, k = 0.1 and ε = 2.
 P. Pecorino / Journal of Health Economics 21 (2002) 699–708 707

the associated percent change in profits as we move from the NR to the R regime. Though
the increase in profits in moving to the reimport regime are small in every case, they are
always positive. Thus, the results under constant elasticity of demand are consistent with
the results under linear demand, where it was shown that profits always rise in the move
to the reimport regime. Since consumer surplus always rises in the move to the R regime,
welfare is unambiguously higher under this regime.
The gain in profits from moving to the R regime is 0.41% when ε = 1.5 and rises to
1.26% when ε = 50. At low values of ε, the domestic monopolist negotiates a relatively
better deal under the NR regime than when ε takes on a high value. Thus, there is more
scope for gains in moving to the R regime when ε has a high value. Similarly, when α takes
on a low value (i.e. when most bargaining power lies with the home monopolist), very good
outcomes are negotiated under the NR regime and there is little scope for gains in moving to
the R regime. As shown in the table, %Π increases until α = 0.9. Past this point so little
bargaining surplus is gained under either regime that the gains from moving to the R regime
are diminished. As discussed earlier in the paper, when α = 1, profits are unchanged in the
move to the R regime. When k takes on a low value, there are only limited gains to be had
from tougher bargaining in the foreign market, because it is so small. Thus, %Π grows
larger as k rises from 0.1 to 1. However, for k > 1, the gains from moving to the R regime
fall. As the home market becomes relatively smaller, there is less incentive to bargain hard
in the R regime. For k > 1, this effect dominates.
In the movement from the NR to R regimes, prices rise abroad and fall at home. Why
is the rise in profits abroad greater than the fall in profits from domestic sales? First,
because the markets are identical, except for a scaling factor, the domestic firm would
ideally like to charge the same monopoly price in each market. In a world with unrestricted
markets, there are no motivations for the firm to engage in price discrimination. Second,
under the NR regime, the domestic price maximizes profits, while the foreign price is less
than its profit maximizing level as long as α > 0. As a result, a small reduction in the home
price will have no first-order effect on profits, while a small increase in the foreign price
will lead to a first-order increase in profits. This does not prove that a move to the R regime
raises firm profits for all demand functions, because the regime shift results in a discrete
rather than an infinitesimal change in prices. Still, once this intuition is understood, the
results of this section should not be viewed as surprising.

5. Conclusion

For both linear and constant elasticity demand, we find that profits of the domestic
monopolist rise when reimports into the home country are allowed. When reimports are
allowed, the foreign negotiated price becomes the domestic price as well. As a result, the
home firm more firmly resists price concessions in the Nash bargaining game. Why, in light
of the results of this paper, do pharmaceutical firms strongly resist allowing import of drugs
from Canada into the US? There are several issues which would need to be addressed in any
further policy analysis of this issue. First, pharmaceutical firms make sales in many foreign
markets, and in some of these markets, there are profit maximizing incentives to engage
in price discrimination. Allowing imports from these countries back into the US would
708 P. Pecorino / Journal of Health Economics 21 (2002) 699–708

destroy the ability of these firms to engage in such profit maximizing price discrimination.12
Second, there are an existing set of prices which have already been negotiated under the NR
regime. The pharmaceutical companies will take a capital loss if prices negotiated under the
NR regime were allowed to prevail under the R regime. Finally, it is possible that there are
some demand specifications for which profits would not rise in the move to the R regime.
These questions deserve further study. The framework established here should prove
valuable when addressing these issues. However, this paper removes the presumption that
pharmaceutical profits will fall if reimport of prescription drugs into the US is allowed.

Acknowledgements

I would like to thank John Conlon, Walter Enders, Sami Dakhlia, Akram Temimi,
Yongsheng Xu, Antoinette Criss, seminar participants at Georgia State University and an
anonymous referee for providing helpful comments on this paper.

References

Anis, A.H., Wen, Q., 1998. Price regulation of pharmaceuticals in Canada. Journal of Health Economics 17, 21–38.
Maggi, G., Rodrı́guez-Clare, A., 1998. The value of trade agreements in the presence of political pressures. Journal
of Political Economy 106, 574–601.
Nash, J., 1953. Two person cooperative games. Econometrica 21, 128–140.
Shapiro, C., 1989. Theories of oligopoly behavior. In: Schmalensee, R., Willig, R.D. (Eds.), Handbook of Industrial
Organization, Vol. 1. North-Holland, Amsterdam.
Wes, M., 2000. International trade, bargaining and efficiency: the holdup problem. Scandinavian Journal of
Economics 102, 151–162.

12 On humanitarian grounds, there are strong reasons why we might want to see pharmaceutical firms engage in

price discrimination by selling at a lower price in developing countries.