1

Rationales for offering subsidies

This opening chapter briefly introduces governments’ incentives to offer

subsidies. The question on governments’ rationales to subsidize is
approached from both a positive and normative perspective.1 First,
why does a government de facto offer subsidies (positive theory)?
Second, why should a government offer subsidies (normative theory)?2
Some important terminological specifications could be helpful in set-
ting the stage. In general terms, the concept of ‘subsidy’ in the SCM
Agreement ‘captures situations in which something of economic value
is transferred by a government to the advantage of a recipient’.3 In this
introductory chapter, the thorny issue of which situations exactly are
covered by the term ‘subsidy’ is left aside. A subsidy is considered an
antonym to a tax, and thus connotes a transfer of money from the
government to a private actor.4 Taxation and subsidization are hereby
considered as two alternative fiscal instruments by which a government
could intervene in the market and are on this basis distinguished from
non-fiscal or ‘regulatory’ government interventions (e.g., technical regu-
lations). Puzzling observations that a tax or regulatory burden on some
private actors might very well be considered as a subsidy to other private
actors or that a subsidy might simply compensate for another tax or
regulatory burden are thus neglected in this introduction. Depending
on the direct recipient, ‘consumer subsidies’ are primarily distinguished

1
The rationales for imposing CVDs is integrated into the normative analysis (see below
Chapter 18).
2
Such a distinction between a ‘positive’ and ‘normative’ theory of trade policy is made, for
example, by A. K. Dixit, ‘Trade Policy: An Agenda for Research’, in P. Krugman (ed.),
Strategic Trade Policy and the New International Economics (Cambridge, MA: MIT Press,
1986), 283–304, at 296.
3
Appellate Body Report, US – Softwood Lumber IV, para. 51.
4
WTO Secretariat, World Trade Report 2006: Exploring the Links between Subsidies, Trade
and the WTO (Geneva: WTO Publications, 2006), at 47.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 6 rational es for off ering subsid ies

from ‘producer subsidies’ in the literature. Depending on the conditions

for receiving producer subsidies, different types are further distinguished.
First of all, a fundamental distinction is drawn between ‘export subsidies’
and all other types of subsidy, often labelled ‘domestic subsidies’. In the
economic literature, an export subsidy is only paid on production that is
exported and, as it thus directly stimulates exportation, is often catego-
rized as a ‘trade instrument’. In contrast, ‘a production or output subsidy’
is a form of domestic subsidy that is granted on all production or output,
regardless of whether it is exported or not. Other types of domestic
subsidy are those offered to specific inputs or activities in the production
process, such as subsidies for research and development (R&D subsidies)
or labour subsidies. ‘Import substitution subsidies’, also named ‘local
content subsidies’, are a specific sort of domestic subsidy as they are offered
on condition of the use of domestic over imported inputs. Whereas
production subsidies are mostly recurring, governments sometimes
offer non-recurring, one-off subsidies, for instance for the acquisition
of fixed assets (e.g., equipment, plant). In short, producer subsidies are
distinguished on the basis of the conditions attached thereto and thus on
the specific activity (e.g., exports, production, R&D, acquisition of fixed
assets) they directly or indirectly aim at stimulating. Accordingly, the
concept of ‘subsidies’ is occasionally distinguished from so-called
‘transfers’, as the latter are not conditioned on any specific use and are
therefore considered to leave the allocation of resources unaffected.
With these broad descriptions in mind, the rationales for subsidization,
or the absence thereof, are introduced.

1.1 The absence of a rationale for subsidization

In a world of complete and perfectly competitive markets, mere interaction
between supply and demand results in an efficient allocation of resources
and a level of output produced at the lowest possible price, which equals
the marginal cost of production and the socially optimal price.5 Welfare
is maximized under market forces (Pareto optimum6) and government

5
In a perfectly competitive market, firms are price takers and can enter or exit freely, and
products are homogeneous. As a result, price will equal marginal costs of production.
Complete markets are characterized by full information and the absence of externalities,
resulting in a price which also equals the socially optimal price.
6
Marginal costs to producers equal marginal benefits to consumers, implying that no one
can be made better off without someone else being made worse off.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 rationales for offering subsidies 7

interventions only distort efficient resource allocation by creating a wedge

between the marginal cost price and the socially optimal price.7
In those markets, there is no reason why a static welfare-maximizing
government would offer a specific subsidy to an industry. In the case where
this subsidy is provided by a so-called ‘small’ country (i.e., a country which
cannot affect the world price and thus its terms of trade8), a welfare loss arises
to this country because the increase in producer welfare does not cover the
cost to the government, although the welfare of other countries is, by defi-
nition, not affected.9 A subsidy offered by a so-called ‘large’ country lowers the
world price and, therefore, can affect the welfare of third countries. As a
consequence, because it negatively affects its terms of trade, welfare in the
subsidizing country deteriorates even more when a subsidy is given by a ‘large’
country to its export-competing industry.10 Overall, the rest of the world is
better off in welfare terms as a result of the depressed world price, but net
exporting countries (and all foreign producers) are adversely impacted,11
whereas net-importing countries (and all foreign consumers) are benefiting
from the lower price. From the perspective of the subsidizing country,
production and export subsidies both benefit domestic producers as they
increase their sales, but they have a very different welfare effect on domestic
consumers: these benefit from the lower price induced by domestic subsidies
(if given by a large country), whereas they are hurt by export subsidies, as such
subsidies give an incentive to domestic producers to export rather than to
produce for the domestic market, thereby creating a wedge between the world
price and the higher domestic price. Thus export subsidies are seen as more
distortive than domestic subsidies, both from the perspective of the subsidiz-
ing country (as they also negatively affect domestic consumers) as well as

7
Welfare is commonly defined as the sum of consumer surplus (i.e., the difference
between the price consumers have to pay and are willing to pay or their ‘marginal
benefit’), producer surplus (i.e., the difference between the price at which producers sell
and are willing to sell or their ‘marginal cost’) and government revenue.
8
Terms of trade is defined as the ratio of export prices to import prices.
9
Any displacement of foreign producers in volume terms is also considered too marginal
to be noticeable.
10
This is illustrated below in figure 5.2 (Chapter 5, section 5.1.3.2.2), which illustrates the
welfare effects of the US production subsidies for cotton that were challenged in the US –
Upland Cotton case. A welfare maximizing large country is rather advised to tax exports.
A production subsidy to its import-competing industry could, in theory, be welfare-
improving for the subsidizing country, but an optimal tariff would be a more direct, and
thus efficient, instrument to exploit its terms of trade to maximize its welfare. Y.-H. Yeh,
‘On Subsidies vs. Tariffs’, 38 Southern Economic Journal (1971), 89–92.
11
The are displaced by subsidized exports and have to accept the depressed world price for
their remaining sales.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 8 rational es for off ering subsid ies

from the perspective of net exporting countries (as they have a more direct
effect on the terms of trade).12
Hence, to understand why countries offer subsidies, the complete and
perfectly competitive markets assumption has to be relaxed or subsidi-
zation has to be explained by reasons other than maximizing welfare. For
the purpose of the present study, four explanations for subsidization are
distinguished in which these assumptions are relaxed.

1.2 The market failure rationale for subsidization

The presence of subsidization can be explained if the perfect and complete
market assumption does not hold. Even if the market price equals the
marginal cost of production, this market price may not reflect all benefits
or costs to society and thus deviate from the ‘socially optimal price’. Positive
or negative externalities (also called ‘spillovers’) are, respectively, benefits or
costs resulting from consumer or producer actions that are not reflected in
the market price and, thus, external to the market. Such marginal external
benefits or costs can be internalized by government intervention in a way
that the new market price equals the socially optimal price.
Hence, in the presence of market failures, the Pareto-efficient outcome
does not result from market forces but requires government intervention. Of
course, government intervention does not guarantee that a Pareto optimal
outcome will be reached. Governments should intervene in an effective way
and tackle the market failure as directly as possible and choose the appro-
priate instrument (e.g., some type of subsidy, tax, regulation) therefor
(targeting principle).13 In the case where the optimal instrument directly
targeting the market failure is unavailable for political or other reasons, the
theory of the second-best applies: governments have to have recourse to a
second-best option to solve the market failure by intervening in other seg-
ments of the economy, but only insofar as the benefits of correcting the
market failure still outweigh the costs that result from the creation of new
distortions in those other segments (cost–benefit analysis).14

12
One reason why these subsidies are usually criticized by third countries could be found in the
fact that the adversely affected producers are better organized than benefiting consumers.
13
J. N. Bhagwati, ‘The Generalized Theory of Distortions and Welfare’, in J. N. Bhagwati (ed.),
International Trade: Selected Readings, 2nd edn (Cambridge, MA: MIT Press, 1987), 265–86.
14
For an overview of papers dealing with second best interventions, see P. Krishna and
A. Panagariya, ‘A Unification of Second Best Results in International Trade’, 52 Journal
of International Economics (2000), 235–57.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 rationales for offering subsidies 9

If a domestic market failure occurs, a trade policy response is, at most,

a second-best option, because a new distortion is created. Domestic
distortions should in principle be corrected by domestic instruments
(e.g., taxes or subsidies on domestic consumption, production, or input
factors) and not by trade instruments (e.g., tariffs, export taxes, or export
subsidies).15 In general, if domestic production is too low because of a
domestic market failure, a production subsidy is superior to an import
barrier (or export subsidy), as the latter also negatively affects consumers
in the domestic market.16,17 Equally, a production or output subsidy is
warranted only where the externality is directly linked (or fixed) to the
level of production.18 Even in that case, a consumer subsidy might be
more efficient than a production subsidy, because a consumer subsidy
boosts production without discriminating between different producers.
A consumer subsidy does not discriminate against foreign producers. If
the market failure is not directly linked to production, but rather to an
activity (e.g., R&D), so-called functional or horizontal subsidies inducing
these activities are more targeted than selective subsidies to particular
sectors (e.g., high-tech industry).
From a national and world welfare perspective, all countries in which
distortions are present are advised to adopt corrective measures (e.g.,
corrective subsidies). Conversely, countries in which such distortions are
not displayed should not intervene, even though they might be con-
fronted with corrective measures (e.g., subsidies) abroad and might thus
claim that the playing field is not level.19

15
J. N. Bhagwati and V. K. Ramaswami, ‘Domestic Distortions, Tariffs and the Theory of
Optimum Subsidy’, 71:1 Journal of Political Economy (1963), 44–50; H. G. Johnson,
‘Optimal Trade Intervention in the presence of Domestic Distortions’, in J. N. Bhagwati
(ed.), International Trade: Selected Readings, 2nd edn (Cambridge, MA: MIT Press,
1987), 235–63.
16
See also J. J. Barceló, ‘Subsidies and Countervailing Duties – Analysis and Proposal’, in
R. Howse (ed.), The World Trading System: Critical Perspective on the World Economy,
vol. 3, Administered Protection (London: Routledge, 1997), 252–314, at 259; K. Bagwell,
‘Remedies in the WTO: An Economic Perspective’, working paper, 9 January 2007, at 25.
17
This statement has to be nuanced where taxes required to finance subsidies are themselves
distortionary. In that case, a combination of tariffs and subsidies may be optimal to correct
the domestic distortion. See D. Brou and M. Ruta, ‘A Commitment Theory of Subsidy
Agreements’, WTO Staff Working Paper ERSD-2012–15 (25 September 2012), 33 pp.
18
See also G. M. Grossman, ‘Promoting New Industrial Activities: A Survey of Recent
Arguments and Evidence’, 14 OECD Economic Studies (Spring 1990), 87–125, at 118.
19
See also A. V. Deardoff, ‘Economic Effects of “Levelling the Playing Field” in
International Trade’, RSIE Discussion Paper No. 289, July 2009, at 20.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 10 rational es for off ering subsidies

1.3 The profit-shifting rationale for subsidization

In some markets, barriers to entry (e.g., large fixed costs) result in only a
limited number of firms present because a high level of production is
required to recover initial costs. In such oligopolistic markets, individual
firms are not price takers as under perfect competition, but set output
(Cournot model) or price levels (Bertrand model) taking into account
output and price decisions by competitors. Hence profits of one firm are
directly affected by strategic decisions of its competitors. The central
insight of strategic trade theory, which has developed since the 1980s,
is that governments’ trade policy can alter this strategic interaction
between firms in such a way as national welfare is optimized. Here,
governments do not intervene to correct market failures, but rather to
shift profit strategically from foreign competitors to domestic firms.
In the market equilibrium, firms are making true profit (price is higher
than average costs) and total output is less than under perfect competition
(oligopolistic distortion). Brander and Spencer have shown that this feature
of oligopolistic markets has important consequences for the design of trade
policy.20 A country could shift a larger share of profitable output from
foreign competitors to domestic firms by subsidizing domestic exports.
The export subsidy commits domestic firms to a higher level of exports,
resulting in a reaction by foreign competitors to contract their output. Given
that the profit gain to domestic firms (expanded output and market share at a
price above average costs) is larger than the subsidy amount (and the
negative terms of trade effect), net welfare of the subsidizing country
increases.21 In addition, total output also increases, resulting in lower
world prices to the benefit of importing countries. On the other hand,
exporting countries are hurt as profitable output is shifted away from their
firms. Accordingly, such strategic trade policy has a beggar-thy-neighbour
element: the subsidizing country’s welfare increases, but at the expense of
other exporting countries’ welfare. The change in total world welfare as a
result of this profit-shifting subsidy is nonetheless positive because the

20
J. A. Brander and B. J. Spencer, ‘Export Subsidies and International Market Share
Rivalry’, 18 Journal of International Economics (1985), 83–100.
21
The contribution of the subsidy to the profit of the domestic firm is offset by the subsidy cost
to the government (transfer). Yet the subsidy has an additional indirect positive effect on
domestic firms’ profit by lowering the output level of foreign firms. The subsidy’s strategic
effect on foreign firms’ behaviour precisely explains the positive welfare effect in the
subsidizing country. See also P. Krugman, ‘The US Response to Foreign Industrial
Targeting’, 1984:1 Brookings Papers on Economic Activity (1984), 77–131, at 98–9.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 rationales for offering subsidies 11

oligopolistic distortion shrinks: the world price lowers and output increases,
both coming closer to the competitive equilibrium.22 This Brander-Spencer
model thus provides an economic rationale for welfare maximizing countries
to subsidize output (as such or only exports23) of firms operating in oligopo-
listic markets.
However, this theory is not considered to be a robust policy recom-
mendation for governments, because the model is sensitive to specific
assumptions related to, for instance, the mode of competition (not
optimal under Bertrand competition),24 the effect on other oligopolistic
industries,25 the credibility of the government in pre-committing itself,26

22
This is the case unless the subsidy drives foreign competitors out of the market (or deters
entrance), turning the domestic competitor into a monopolist. D. R. Collie, ‘A Rationale for
the WTO Prohibition on Export Subsidies: Strategic Export Subsidies and World Welfare’,
11 Open Economies Review (2000), 229–45, at 230; D. Leahy and J. P. Neary, ‘Multilateral
Subsidy Games’, 41:1 Economic Theory (October 2009), 41–66.
23
The analysis is similar in the case of output or production subsidies not dependent on
exportation, as the model assumes that there is no domestic consumption in both
exporting countries.
24
If firms compete on price (Bertrand competition), the optimal policy would be an export
tax rather than an export subsidy. The export tax commits the domestic firm to a higher
price, thereby giving an incentive to foreign firms to set a higher price. In contrast to
Cournot competition, foreign firms therefore also benefit from the export tax and total
global welfare is reduced. See J. Eaton and G. M. Grossman, ‘Optimal Trade and
Industrial Policy under Oligopoly’, 101:2 Quarterly Journal of Economics (May 1986),
383–406, at 392–4; J. A. Brander, ‘Strategic Trade Policy’, in G. M. Grossman and
K. Rogoff (eds.), Handbook of International Economics: Volume 3 (Amsterdam:
North-Holland, 1995), 1395–455, at 1416. Yet R&D subsidies remain optimal under
both the Cournot and the Bertrand options. K. Bagwell and R. W. Staiger, ‘The
Sensitivity of Strategic and Corrective R&D Policy in Oligopolistic Industries’, 36
Journal of International Economics (1994), 133–50; D. Leahy and J. P. Neary, ‘Robust
Rules for Industrial Policy in Open Economies’, 10 Journal of International Trade &
Economic Development (2001), 393–409. Further, if both firms are unionized, export
subsidies become optimal under Bertrand competition. S. Bandyopadhyay,
S. C. Bandyopadhyay, and E. Park, ‘Unionized Bertrand Duopoly and Strategic Export
Policy’, 8:1 Review of International Economics (2000), 164–74.
25
Profit-shifting subsidies to one oligopolistic industry might cause a greater profit-
extracting loss in other oligopolistic industries where they compete for the same scarce
production factor. A. K. Dixit and G. M. Grossman, ‘Targeted Export Promotion with
Several Oligopolistic Industries’, 21 Journal of International Economics (1986), 233–49;
G. M. Grossman, ‘Strategic Export Promotion: A Critique’, in P. Krugman (ed.),
Strategic Trade Policy and the New International Economics (Cambridge, MA: MIT
Press, 1986), 47–68, at 58–60.
26
J. P. Neary and Dermot Leahy, ‘Strategic Trade and Industrial Policy towards Dynamic
Oligopolies’, 110 Economic Journal (April 2000), 484–508.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 12 rational es for off ering subsidies

the opportunity cost of public funds,27 and the number of firms.28

Further, governments might lack sufficient knowledge of the oligopolis-
tic market in question to set an efficient level of subsidization.29
Krugman therefore concludes that ‘surprisingly few of the new trade
theorists themselves are strategists’.30 Still, Brander underlines the stra-
tegic trade theory’s robust general finding that an oligopoly almost
always creates an incentive for intervention, although another robust
finding is that competing countries have an incentive to conclude an
agreement limiting such profit-shifting subsidization.31

1.4 The redistribution rationale for subsidization

Subsidies are regularly used as government instruments to redistribute
income among regions, as this is seen as imperative not only from an
equity point of view but equally to lower potential social or political
tensions between regions.32 Such in-country disparities in income levels
might be the result of differences in resources (e.g., natural resources)
between regions. Alternatively, the mere presence of external economies
of scale could explain why production might cluster in some regions and
not in others. Whereas from an economic viewpoint governments might
27
The Brander–Spencer model gives equal weight to the welfare of producers and govern-
ments. Yet the opportunity cost of public funds likely exceeds unity, as these are often
collected on the basis of distortionary taxation. Government welfare could also be given
greater weight than producer profit because producers could be partly foreign-owned
(this profit would not be incorporated in national welfare) or because of income
distributional reasons. If so, the case for profit-shifting subsidies becomes less evident.
Subsidies only remain optimal for surprisingly low levels of additional costs attached to
public funds. J. P. Neary, ‘Cost Asymmetries in International Subsidy Games: Should
Governments Help Winners or Losers?’, 37 Journal of International Economics (1994),
197–218; H. K. Gruenspecht, ‘Export Subsidies for Differentiated Products’, 24 Journal
of International Economics (1988), 331–44; Brander, above n. 24, at 1410; J. Ma, ‘Is an
Export Subsidy a Robust Trade Policy Recommendation toward a Unionized Duopoly?’,
20:2 Economics & Politics (2008), 141–55.
28
The Brander–Spencer model also holds when it is extended to more than two firms, but
only if the number of domestic firms is not too large. A. Dixit, ‘International Trade
Policy for Oligopolistic Industries’, 94 (supplement) Economic Journal (1984), 1–16, at
11–12; see also P. A. G. van Bergeijk and D. L. Kabel, ‘Strategic Trade Theories and Trade
Policy’, 27:6 Journal of World Trade (December 1993), 175–86, at 182.
29
P. Krugman, ‘The Narrow and Broad Arguments for Free Trade’, 83:2 American
Economic Review (1993), 362–6, at 363–4.
30
Ibid., at 363–4.
31
Brander, above n. 24, at 1447. See also below Chapter 16, section 16.2.
32
Different reasons explaining why societies redistribute income are listed in the World
Trade Report 2006, above n. 4, at 89–90 and 95.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 rationales for offering subsidies 13

rather opt for stimulating further clustering, they could choose to sub-
sidize production in backward regions for redistributive reasons. Thus
subsidies on this latter ground do not aim at correcting a certain market
failure, but at altering the market outcome.33 Redistribution is preferably
reached through domestic subsidies rather than through export subsidies
and/or import duties, as the last two options negatively affect domestic
consumers.34 Still, under the assumption of complete and perfectly com-
petitive markets, such subsidies stimulating production in backward regions
might not be first best instruments to achieve redistribution.35 Indeed, redis-
tribution of income could in principle be better obtained by using so-called
lump-sum taxes and transfers, which by definition do not affect the efficient
allocation of resources (no deadweight losses). In that case, income would
simply be transferred from inhabitants of advanced regions to backward
regions through a system of direct taxes on high incomes and transfers (i.e.,
not conditioned on any use) to low incomes.36 Of course, transfers are not
costless as they could have an adverse effect on incentives and incur an
administrative cost.37 Moreover, if regional labour distribution generates
external benefits that are not reflected in wages (e.g., reducing congestion in
cities), stimulating employment (wage subsidy) or production in backward
regions could become appropriate.38

1.5 The political-economy rationale for subsidization

The political-economy literature does not start from the assumption that
decision-makers aim at maximizing the welfare of their constituencies,
but instead assumes that politicians aim at maximizing their own welfare
(self-interest), which is often modelled in terms of maximizing their
chances of (re-)election. The outcome of the decision-making process,
for example on offering subsidies, therefore, depends on its effect in the

33
Ibid., at 107.
34
J. E. Meade, The Theory of International Economic Policy: Volume II: Trade and Welfare
(London: Oxford University Press, 1955, repr. 1966), at 314.
35
IMF, Fuel and Food Price Subsidies: Issues and Reform Options (Washington, DC: IMF,
2008), 35.
36
For instance, this was already described by Meade, above n. 34, at 314.
37
World Trade Report 2006, above n. 4, at 90.
38
Here, again, a market failure would be present, which could be targeted directly by a
wage subsidy.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 14 rational es for off ering subsidies

political ‘marketplace’.39 In the standard model, this ‘marketplace’ is

constructed as a democracy, in which politicians have to obtain the support
of the majority to be elected. Under simple majority voting, the outcome of
a decision depends on whether it has a positive or negative effect on
the median voter. Subsidies benefit their recipients, while other voters
merely bear a cost in the form of taxes raised to pay for the subsidy.
Here, a subsidy would only be offered if the median voter is one of
those beneficiaries. Hence this model would predict that sector-specific
subsidies are unlikely to be offered as only a few benefit.40 In reality,
however, the opposite situation seems to happen more frequently:
the more benefits are concentrated and the more costs are diffused, the
more likely that promotion or protection is given. This insight was
offered and modelled by Olson, who explained the conditions for the
emergence of ‘special interest groups’. Promotion (e.g., subsidies)
might offer large individual gains to individual producers, and thus
give these beneficiaries an incentive to lobby for such an outcome,
whereas the costs of subsidies are highly diffuse and relatively small
for each taxpayer, implying that they have no incentive to oppose such
action.
Grossman and Helpman constructed a formal model41 in which (i)
‘special interest groups’ offer campaign contributions to politicians
running for re-election so as to influence their choice of trade policy
(e.g., export subsidies), and (ii) politicians aim at maximizing their
chances of re-election (self-interest), which depends on the amount of
such contributions as well as on the overall economic welfare.42 Under
the small country hypothesis, free trade would maximize overall welfare.
Yet because campaign contributions also influence politicians’ behav-
iour, the equilibrium deviates from this welfare optimum: export sub-
sidies are offered to all sectors that are organized as special interest

39
A. O. Sykes, ‘Protectionism as a “Safeguard”: A Positive Analysis of the GATT “Escape
Clause” with Normative Speculations’, 58:1 University of Chicago Law Review (1991),
255–305, at 275.
40
World Trade Report 2006, above n. 4, at 64.
41
G. M. Grossman and E. Helpman, ‘Protection for Sale’, 84 American Economic Review
(1994), 833–50. See also B. L. Gardner, ‘The Political Economy of US Export Subsidies
for Wheat’, in A. O. Krueger (ed.), The Political Economy of American Trade Policy
(University of Chicago Press, 1996), 291–331.
42
Politicians’ (potential) concern for overall welfare thus only arises because of self-
interest, as it might increase their chances of re-election.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 rationales for offering subsidies 15

groups, implying higher prices on those products in the domestic market

than in the free trade situation.43
Two interesting extensions of the model are presented. First, different
special interest groups might have opposite interests as producers. In
particular, downstream producers who use intermediate inputs are very
often organized as a special interest group and would oppose export
promotion as it pushes the price upwards. Facing opposition of down-
stream producers, producers of intermediate goods are less likely to be
rewarded export promotion than producers of final goods. Second,
the introduction of a large country assumption not only alters the
(unilateral) equilibrium of the country in question but could also result
in co-operation (trade negotiations) with third countries, because this
trade policy has an impact on those third countries. Under the non-co-
operative equilibrium (‘trade war’), producers are more likely to get tariff
protection (e.g., countervailing duties – CVDs) but are less likely to
receive export subsidies than under the small country hypothesis,
given that large countries’ overall welfare could be maximized by impos-
ing an optimal tariff and/or export tax.44 Grossman and Helpman show
why politicians motivated by self-interest would nonetheless gain from
concluding a trade agreement (‘trade talk’).45,46

43
In the equilibrium, the joint welfare of politicians (which is a function of campaign
contributions and overall welfare) and special interest groups is maximized. Grossman
and Helpman, above n. 41, at 833–50.
44
The ‘concern’ for overall welfare pushes the government towards an export tax, whereas
export industries bid for export subsidies. Hence the equilibrium might be either an
export tax or an export subsidy (or free trade). Compared with the small country
situation, export industries bidding for export subsidies not only compete with other
producers who have the opposite interest (e.g., downstream producers), but also face
opposition because of the fact that the subsidy has a negative effect on overall welfare.
45
Compared with the non-co-operative equilibrium, politicians’ benefits that result from
concluding an agreement flow from the extra market access generated in third countries
and from the reduction in government support for foreign firms competing in the
import market, for which they will be rewarded by export industries and import-
competing industries respectively. G. M. Grossman and E. Helpman, ‘Trade Wars and
Trade Talks’, NBER Working Paper Series No. 4280 (February 1993).
46
Ethier criticizes this model because it incorporates terms of trade motivations on the part of
the government. The central premise of the ‘received theory’ elaborated in this model is that
trade agreements are concluded because, at least to some degree, governments are concerned
with terms of trade considerations. According to Ethier, such models, however, do not
correspond with reality. For example, the Grossman and Helpman model would suggest
that, under certain conditions (see above n. 41), countries would impose an export tax to
manipulate terms of trade (non-co-operative equilibrium). This model would thus predict
that countries would multilaterally agree to limit such terms of trade manipulations (co-
operative equilibrium). In reality, however, countries almost never impose export taxes (and

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 16 rational es for off ering subsidies

In sum, the political-economy literature explains why countries would be

willing to offer subsidies even though this would not benefit their overall
welfare. Subsidies are hereby not offered to correct market failures, and thus
to improve overall welfare, but because its beneficiaries (i.e., special interest
groups) are successful in lobbying. In political-economy terms, an export
subsidy may very well be perceived as a gain to an exporting country (e.g.,
rewarded by exporters) and a cost to an importing country (e.g., harm to
import-competing industries). Here, a ‘government failure’ is present, not
because the government lacks the administrative capacity or information to
correct market failures efficiently, but because it is influenced by special
interest groups.47

1.6 Conclusion
Four rationales have been introduced to help explain why governments
in reality offer subsidies (positive theory): subsidies might be employed
to correct market failures, to shift profits in oligopolistic markets, to
redistribute income, and/or for political-economy reasons. To be sure,
these four rationales do not cover the wide spectrum of why govern-
ments offer subsidies, although other rationales could often be traced
back to one or more of these four types.48
From a normative perspective, the different rationales clearly work in
different directions. Whereas the political-economy rationale suggests
that a country (and the world as a whole) would be better off in welfare
terms if multilateral disciplines prevent subsidization (‘tie-their-own-
hands’ argument), the market failure rationale may suggest that policy
space should be preserved to offer certain subsidies to maximize its own

certainly not for terms of trade reasons). Moreover, these taxes are in principle not restricted
under the WTO unless these are bound in a Member’s Schedule. As Ethier argues, the
Grossman and Helpman model only fits with reality if political-economy motivations
completely dominate terms of trade motivations. W. J. Ethier, ‘The Theory of Trade Policy
and Trade Agreements: A Critique’, 23 European Journal of Political Economy (2007),
605–23.
47
A variant is that governments do aim at maximizing global welfare, but to this end have
to rely on special interest groups for information and thus anyway fail to reach the
welfare optimum. K. Baylis, ‘Unfair Subsidies and Countervailing Duties’, in W. Kerr
and J. D. Gaisford (eds.), Handbook on International Trade Policy (Cheltenham: Edward
Elgar, 2007), 347–60, at 351–2.
48
For instance, subsidies to safeguard jobs in declining industries might, depending on the
facts, be inspired by political-economy, market failure (e.g., congestion in the labour
market), and/or redistribution rationales.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

 rationales for offering subsidies 17

and world welfare and the redistribution rationale equally suggests that
some policy space might be needed to redistribute income. Finally, the
profit-shifting rationale suggests that an agreement on limiting subsidi-
zation is welfare-improving for both subsidizing countries, but could
depress overall world welfare.
Hence, when examining the regulatory framework explained in Part I,
the long-understood but still often neglected lesson – that ‘policy space’
is not something inherently valuable in the sense of being conducive to
maximizing welfare or, more broadly, sustainable development – should
be recalled. Since the rejection of mercantilism centuries ago by the
founding fathers of the comparative advantage theory, it is (or should
be) common knowledge that curtailing its own ‘policy space’ on trade
policy (i.e. tie-their-own-hands) might be costly in political terms but
beneficial in welfare terms.

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press

https://doi.org/10.1017/CBO9781139046589.004 Published online by Cambridge University Press