trade policy of a free trade agreement in the presence of foreign lobbying

Journal of International Economics 77 (2009) 37–49

Contents lists available at ScienceDirect

Journal of International Economics

j ourna l homepage: www.e lsev ie r.com/ locate /econbase

Trade policy of a free trade agreement in the presence of foreign lobbying☆

Andrey Stoyanov ⁎Economics, Atkinson Faculty of Liberal Arts and Professional Studies, York University, 2009-4700 Keele Str., Toronto, ON, Canada

☆ I am grateful to my advisors Werner Antweiler, MCopeland for continuous support and guidance. I amBarrington-Leigh, Kishore Gawande, Vadim Marmer,discussions on the topic. I also would like to thankUniversity of British Columbia, McMaster University, NUniversity, Syracuse University, University of Toronto,Victoria, 2007 European Econometrics Society SummEconomic Association conference for their comments an⁎ Tel.: +1 416 736 2100x22833; fax: +1 416 736 5188.

E-mail address: [email protected]

0022-1996/$ – see front matter © 2008 Elsevier B.V. Aldoi:10.1016/j.jinteco.2008.10.001

a b s t r a c t
a r t i c l e i n f o
Article history:
This paper studies the effect Received 14 April 2008Received in revised form 7 October 2008Accepted 8 October 2008
Keywords:Trade policyFree Trade AreasForeign lobbies

JEL classification:F12F13F14D72

of foreign lobbies on trade policy of a country which is a member of a Free TradeAgreement (FTA). It uses a monopolistically competitive political economy model in which the governmentdetermines external tariffs endogenously. The effect of foreign lobbying under the FTA is examinedempirically using Canadian industry-level trade data that allow differentiating of lobby groups by the countryof origin. The analysis suggests that the presence of foreign lobbying has a significant effect on the domestictrade policy. The heterogeneity of foreign lobbies is also important: the presence of an organized lobbyinggroup in an FTA partner country tends to raise trade barriers while an organized lobbying group of exportersfrom outside of the FTA is associated with less protection.

© 2008 Elsevier B.V. All rights reserved.

1. Introduction

In the political economy literature a growing number of studiesview trade policy as an endogenous outcome of lobbying activity byspecial interest groups. Several authors (Goldberg and Maggi, 1999,Gawande and Bandyopadhyay, 2000) have confirmed that lobbyingintensity by domestic firms is one of the main determinants of thecross-industry pattern of protection. More recently, Gawande et al.(2006) also find that lobbying by foreign firms for trade barriersreduction has a significant effect on the structure of tariffs acrossindustries. However, if a country is a member of a regional free tradeagreement (FTA) and foreign firms can affect the government'sdecision regarding trade policy, it becomes necessary to distinguishforeign lobbying from within and outside of the FTA. Organizedforeign interests with preferential market access will lobby for moreprotection against other foreign firms, and the trade agreement maybecome more protectionist with a strong lobby group in a prospectiveFTA partner country. Active foreign lobbying under the preferential

atilde Bombardini and Brianalso grateful to Christopher

and Daniel Trefler for usefulseminar participants of the

ew Economic School, RyersonYork University, University ofer Meetings, and Canadiand suggestions.

.

l rights reserved.

trade agreement may not only lead to an increase in trade barriers, butalso make welfare-reducing trade agreements politically feasible.

In this paper I analyze the effect of foreign lobbying on domestictrade policy when the country is a member of a preferential tradeagreement using Canadian post-NAFTA trade data. This analysisreveals two main results. First, the activity of foreign lobbyists inCanada is a significant determinant of the Canadian trade policy, andsectors in which foreign firms without preferential market access arepolitically organized tend to receive less protection. This resultsupports the previous finding by Gawande et al. (2006) for the US.Second, NAFTA has an important effect on the structure of foreignlobbies. The data confirm that foreign firms with preferential marketaccess lobby for more protection just as domestic firms do. This resulthas important implications for the effect of an FTA on a country's tradepolicy. It implies that prior to NAFTA, US firms lobbied for Canadiantrade barriers reduction like all other foreign firms, but once NAFTAwas signed, they switch to lobbying for trade barriers increase. As aresult, an FTA with a large and politically strong partner country mayraise trade barriers and increase trade distortions, making trade policyof regional trading blocks more protectionist.1

1 Two considerations should be taken into account when partner country lobbyingfor more protection is considered. First, the WTO tariff binding constrains the lobbyingopportunities by the partner country firms; however, they may still play an importantrole in the future and impede multilateral trade liberalization in those sectors. Second,the WTO precludes countries from raising their tariffs once the FTA is signed. Yetforeign lobbies may oppose further tariff reduction as described above and use anti-dumping and countervailing measures to gain protection.

mailto:[email protected]

http://dx.doi.org/10.1016/j.jinteco.2008.10.001

http://www.sciencedirect.com/science/journal/00221996

2 “No person or party shall accept or use contributions from a person who is not aCanadian citizen…, corporation or association that does not carry on business inCanada, foreign political party, or foreign government,” (S.217).

38 A. Stoyanov / Journal of International Economics 77 (2009) 37–49

This paper is the first one that analyzes from theoretical andempirical points of view the effect of foreign lobbying on domestictrade policy in the presence of an FTA. The political economy modelpresented in this paper incorporates a monopolistically competitivemarket structure into the Grossman and Helpman (1994) ‘protectionfor sale’ setup (henceforth GH) to analyze the role of foreign lobbyingin the making of national trade policy. The framework is furtherextended by allowing for two types of foreign interest groups, namely,lobbying groups formed by firms from an FTA partner country and byfirms from countries outside of the FTA, aggregated into the rest of theworld (ROW). This differentiation of foreign lobbies by market accessis important in the presence of an FTA. When two countries join aregional trade agreement, granting zero import tariffs to each other,firms from countries with preferential market access will lobby formore protection under the FTA to lock it from competition from theROW firms, while firms from outside of the FTAwill continue to lobbyfor trade liberalization. This differentiation of foreign lobbyingobjectives, that follows from the presence of the FTA, implies that ina world where almost every country is a part of at least one regionaltrade agreement, a complete theory of the effect of foreign lobbing onthe national trade system should take this differentiation into account.

The model of foreign lobbying in the presence of the FTA is testedusing Canadian post-NAFTA trade data. The empirical analysissuggests a strong and statistically significant effect of domestic andforeign lobbying on Canadian trade policy, and points to theimportance of distinguishing partner country lobbying from ROWlobbying. Using data on lobbying intensity by sector and by country oforigin, this paper verifies that the main predictions of the model areconsistent with the data. First, the importance of both partner countryand ROW foreign lobbying under the FTA is confirmed in the Canadiandata: while the presence of an organized domestic lobbying groupraises industry import tariffs by 3–5% relative to an unorganizedindustry, organized partner-country lobbying would raise it by 1–2%,and foreign lobbying from the ROW would lower it by 2–3%. Theseresults are economically meaningful and confirm the main predictionof themodel that foreign firmswith preferential market access behavejust as domestic firms do, which introduces an additional distortion inthe policy making process.

Second, the empirical evidence shows that even politicallyunorganized sectors receive a positive level of protection from thegovernment of 2–4% import tariff, providing support for the imper-fectly competitive structure of the model. Additionally, an importantcontribution of this work to the political economy literature of trade isits determination of more plausible values for the government'svaluation of political contributions. In this paper, the government isestimated to value political contributions more than national welfare,which is in sharp contrast with previous tests of the benchmark GHspecification in which governments were found to have strongerpreferences for welfare (Goldberg and Maggi, 1999; Gawande andBandyopadhyay, 2000). This result explains why relatively smallpolitical contributions may have strong policy effects.

My empirical results provide a new perspective on the effect ofregional trade blocks on trade policies of member countries. Most ofthe literature on endogenous trade policy concludes that a country ismore likely to reduce its external tariff when it enters an FTA.Richardson (1993), Bagwell and Staiger (1997), and Bohara et al.(2004), Bond et al. (2004) show that a welfare-maximizing govern-ment will lower external import tariffs once an FTA is formed. Bydoing this, the government restores part of the tariff revenue lost dueto the shift in import demand from the ROW to the partner countryfirms.

Ornelas (2005a,b) examined the political economy of an FTAwithout foreign lobbying using an oligopolisticmarket structure in theGH model. He shows that FTA formation weakens the lobbying powerof domestic firms because the elimination of tariffs between FTAmember countries shifts part of the tariff rent from domestic firms

towards firms from a partner country. The model in this paper,however, allows for cross-border lobbying activity, which lessens theeffect introduced by Ornelas (2005a,b) since the reduction in politicalactivity by domestic firms is coupled with an increase in contributionsfor protection by the FTA partner country firms. As a result, in thepresence of foreign lobbying a country's government may in fact wantto raise external tariffs under the FTAwhen strong lobbying by an FTApartner country puts extra pressure on the government for highertrade barriers.

Activities of foreign interest groups in national policy-making havereceived a growing attention in the political economy literature. Manyscholars have evaluated the intensity of foreign lobbying in the US andargue that it has high potential for policy influence. Mitchell (1995)found that foreign affiliates in the US contributed 5.6% of totalcorporate political contributions in 1987–88, and 42% of them hiredprofessional lobbyists to promote their interests in Washington.Hansen and Mitchell (2000) claim that, although foreign corporationsmake lower political contributions than domestic ones due to theexisting legal restrictions, they are just as intensive as domesticcorporations with respect to lobbying activity and lobbyingexpenditures.

While it has been argued that foreign corporations are quiteinfluential in US politics, very little research has been done on theireffect on trade policy outcomes. In their pioneering work, Gawandeet al. (2006) demonstrate that foreign agents lobbying expenditure inthe US is even greater than political contributions by domesticcorporations, and that the elasticity of the US import tariff withrespect to foreign lobbying is almost as big as with respect to thedomestic one. Their paperwas thefirst to show that foreign lobbying isan important factor in the formation of national trade policy and arguethat foreign lobbyingmay be beneficial to the country's trade policy asa counter-pressure to domestic interests, helping to reduce distor-tional trade barriers. In contrast, my paper suggests that the aboveargument is invalid in the presence of preferential trade agreements,when firms from a partner country prefer to maintain highdiscriminatory tariffs for third-country imports. Since most of thecountries are members of at least one preferential trade agreement,this paper demonstrates that cross-border lobbying may stimulatemore protectionist trade policies of FTAs and disruptmultilateral tradeliberalization.

The paper is structured as follows. Section 2 provides some evidenceon the lobbyingactivities of foreignfirms inCanada. Section3 introducesthe modified GH version of the model that allows imperfectlycompetitive market structure and two groups of foreign lobbies, andmotivates the empirical methodology. Section 4 describes the data.Estimation results are presented in Sections 5 and 6 concludes.

2. The role of foreign lobbies in Canada

The first law that regulates the activity of domestic and foreignlobbyists in Canada is the Canada Elections Act, introduced in 1960.This law regulates the amount of political contributions by Canadiannationals to political parties and explicitly bans the use of politicalcontributions from foreigners.2

However, this law suffered from a lack of transparency in thelobbying process and on September 30, 1989, a Lobbying RegistrationAct (LRA) came into force. This piece of legislation introduced adefinition of a lobbyist and a requirement for lobbyists to register withthe Lobbyists Registrar. But more importantly, it requires lobbyists toprovide information about the name and business address of theorganization “that has a direct interest in the outcome of the lobbyists'activities on behalf of the client,” all of its subsidiaries and corporate

4 The model developed in this section is a generalization of the GH model withmonopolistic competition by Chang (2005) that allows for country bias in consumers

39A. Stoyanov / Journal of International Economics 77 (2009) 37–49

headquarters, if there is one. An amendment to the LRA, introduced in1996, made this information publicly available. It also introduced astrict disclosure of funds policy applied to political parties. Togetherwith the Canada Elections Act, the LRA made it difficult for foreignfirms to lobby their interests in Canada directly.

Nevertheless, politically active foreign firms can still influencetrade policy outcomes in at least two ways. First, they can hireCanadian agents and consultants to lobby the executive branch ontheir behalf and affect policy outcomes in away that suits the interestsof foreign firms. Second, subsidiaries of foreign enterprises can makelegal political contributions with their own funds to defeat legislatorswho are unfriendly to their interests. Since there are no restrictions onthe share of foreign capital in the assets of a company that makespolitical contributions, any local subsidiary of a foreign corporationcan make political donations from its own funds if “it carries businessin Canada.”Moreover, almost any big foreign company that exports toCanada has an independent local sales department, which is legallyallowed to lobby for a reduction in trade barriers on the productsimported by its parent company into Canada. Lobbying efforts of suchsubsidiaries will be counter to the efforts of domestic firms and,therefore, pooling all Canadian firms together regardless of theirownership may lead to misleading results and estimation problems.

In the trade policy literature, corporate political activity is typicallymeasured by financial contributions to candidates and political parties,while very little attentionhas beenpaid to othermeansof affectingpolicyoutcomes such as direct lobbying. However, earlier research on the effectof foreign companies on national policy (Hansen and Mitchell, 2000)suggests that foreign corporations prefer direct lobbying to politicalcontributions not only because of legal restrictions on contributions, butalso because of informal legitimacy questions for politicians with respectto accepting money from corporate sources with foreign ownership.Hansen and Mitchell found that because lobbying is less visible thancontributions, foreigners use it as intensively (and effectively) asdomestic firms do. For these reasons, lobbying expenditures seem to bea better measure of foreign political involvement, especially in countrieswith legal restrictions on political contributions by foreigners.

Unfortunately, data on lobbying expenditures by domestic andforeign corporations are unavailable for Canada. To illustrate the roleof this channel in foreign political activity in Canada, I collected theinformation on the number of lobbyists officially registered with theOffice of the Registrar of Lobbyists, as is required by the LRA.3 Thelobbyists registration data is publicly available and is discussed inmore detail in Section 4.2, but the following figures demonstrate therelative importance of foreign lobbying in Canada. In 1996–97, therewere 1032 officially registered lobbyists representing interests ofmanufacturing firms regarding Canadian trade policy, with 47%, 26%and 27% of them acting on behalf of Canadian, US and ROW firms,respectively. These figures highlight the potential strength of foreignlobbyists in Canada and suggest that the number of contacts of foreignfirms with Canadian policymakers was at least not smaller than that ofdomestic firms. Then, when almost every country in a world tradesunder preferential agreements, it is important to consider the effect offoreign lobbying under the FTA when a group of foreign firms withpreferential market access may lobby for more protection andreinforce trade diversion effects.

3. The model

The theoretical model is based on the Grossman and Helpman(1994) political economy model and presents several modifications

3 The Act defines a lobbyist as “an individual who, for payment, undertakes to lobbyon behalf of a client” and represents an organization in arranging meetings with publicoffice holders, or “communicate with a public office holder in an attempt to influencethe development of any legislative proposal, …the making or amendment of anyregulation, …the development or amendment of any program or policy.”

that allow for the presence of foreign lobbying and facilitateeconometric estimation. In their original formulation, Grossmanand Helpman considered a small open economy, which leaves noroom for foreign companies to lobby because pre-tariff prices areexogenously fixed. As such, with perfectly competitive globalmarkets and free entry, an increase in the domestic tariff rate willnot affect profits of foreign firm from export operations, leavingthem no incentives to participate in trade-policy game. In this work Idevelop and build into the GH setup a model of monopolisticcompetitionwith differentiated goods4 to allow foreign firms to gainor lose from import tariffs.5

There is 1 industry in the model and three countries: Canada(Home country), the US (FTA Partner country) and the ROW, denotedby H, P and ROW, respectively. Industries are denoted by indexi∈ {1,…, N} and countries by i∈ {H, P, ROW}There are ni

jfirms in

country j and industry i. These firms are assumed to be symmetricwithin the same country and industry, i.e. they share the same coststructure and hence face the same demand functions and charge thesame prices. In total, there are (niH+niP+niROW) different varieties ofeach product i.

A representative consumer maximizes a quasilinear utility func-tion with a constant elasticity of substitution index nested into aCobb–Douglas function:

U = X0 + ∑N

i = 1ωilnXi

Xi = nHi d

H1σ i

i xHσ i−1σ i

i + nPi d

P1σ i

i xPσ i−1σ i

i + nROWi dROW

1σ i

i xROWσ i−1σ i

i

! σ iσ i−1

ð1Þ

where Xi is an aggregate consumption index for product i, ωi is theshare of product i in the total consumer's expenditure, xij is thedemand for product i produced in country j, σiN1 is the elasticity ofsubstitution between varieties of product i, and di

j is a country-widetaste (or quality) parameter for product i imported from country j.Maximizing Eq. (1) subject to the standard budget constraint, weobtain the demand functions and an aggregate price index forproduct i:

Xi =ωi Pið Þ−1; iz1 ð2Þ

xji =ωid

ji

p ji

p ji

Pi

!1−σ i

ð3Þ

Pi = nHi d

Hi pHi� �1−σ i + nP

i dPi pPi� �1−σ i + nROW

i dROWi pROWi

� �1−σ i� � 1

1−σ i ð4Þ

Firms within one country and sector are assumed to have the sameconstant marginal cost. This allows us to consider the Canadianmarket independently from other markets, i.e. prices in the Canadianmarket depend only on the demand elasticity, the (fixed) number offirms and the fixed marginal cost structure. Denoting a specific importtariff set by the home country government on imports of product i

preferences, country-specific productivity and specific import tariff.5 Similar result can be obtained under other forms of imperfect competition, e.g.

oligopoly, but the advantage of monopolistic competition is that trade policy isdetermined by the shares of domestic and foreign firms on the home country market.In general, market shares are more stable than inverse import penetration ratiosrequired by the model of perfect competition (Grossman and Helpman, 1994) oroligopoly (Gawande et al., 2006).


from country j as τij, we can write the profit of a country j firm that

sells product i in the Canada as:

πji = pj

i −cji−τ

ji

� �qji ð5Þ

where qij is the quantity supplied and ci

j is the marginal costs. I assumethat the number of firms is large enough to ignore the effect of theirindividual pricing decisions on the industry price index Pi, i.e. eachfirm takes the price index as given. Knowing product demandfunctions (3), each firm sets the profit-maximizing price as a markupover its marginal costs:

pHi = σ i

σ i−1

� �cHi ;p

Pi =

σ i

σ i−1

� �cPi + τ

Pi

� �;pROWi = σ i

σ i−1

� �cROWi + τROWi

� � ð6Þ

For convenience, isolate costs from Eq. (6) and write downequilibrium profits (5) as:

πji = σ

−1i p j

i qji ; 8j ð7Þ

The government chooses import tariffs to maximize a weightedsum of national welfare W and political contributions C:

G τj;Cj� �

= ∑iCHi + aW + b∑

iCPi + c∑

iCROWi ð8Þ

where CiH, Ci

P and CiROW are industry-wide political contributions from

each country. Coefficient a is a weight that the government assigns tonational welfare relative to political contributions. The government isallowed to value domestic and foreign contributions differently asreflected by parameters b and c that show the government'spreferences for the US and ROW contributions, respectively, over thecontributions by domestic firms. As long as accepting contributionsfrom foreign firms involves risk of reputation loss or law infraction,politicians may prefer domestic contributions to overseas donationsthus both coefficients are presumably less or equal to one.6

Firms in industry i can organize themselves and form a group tolobby the local government for a change in trade policy.7 Firms withinthe FTA pay no import tariffs and hence lobby for more protection,while firms from other countries lobby for lower tariffs for theopposite reason. The lobby representing industry i of country jmaximizes its welfare from obtaining protection net of politicalcontribution: (Wi

j−Cij). As in Grossman and Helpman (1994), theequilibrium trade policy is a solution to a two-stage game. In the firststage, knowing the government's objective function, each organizedlobbying group provides the government with a schedule of politicalcontributions as a function of import tariff. In the second stage,observing contribution schedules, the government sets trade policythat maximizes its objective function (8). Grossman and Helpman(1994) show that for truthful contribution schedules the optimal tradepolicy is the one that maximizes joint surplus of the government andorganized lobbying groups. Let αi denote the share of the homecountry population entitled to the domestic industry i profits, and Ii

j

denote an index variable that takes the value of onewhen industry i in

6 In this simple model it is assumed that the government only values nationalwelfare and political contributions. However, politicians are also concerned aboutpolitical risk from being involved in relationships with foreign agents. If politicianswant to minimize the risk of being disclosed in protecting interests of foreign residentsand this risk is proportional to the amount of foreign contributions, the model cangenerate different valuation of political contributions from different sources bypolicymakers. Therefore, without modeling political risk explicitly, we can assume thatthe government may have a political bias against foreign monetary contributions.

7 With the number of firms in the sector being limited by the endowment of sector-specific capital, firms in each industry have an incentive to form a lobby group andseek for protection from foreign competition. Here I ignore the free-riding problemwithin each sector. See Bombardini (2008) for an extensive discussion of firm-levelcontribution decision.

country j is politically organized and zero otherwise. The joint welfarefunction then takes the form:

X = ∑N

i = 1IHi W

Hi + aW + b ∑

N

i = 1IPi W

Pi + c ∑

N

i = 1IROWi WROW

i ð9Þ

where WiH=ni

H τiH+αi (TR+CS) is welfare of the domestic industry i

gross of political contributions, TR and CS are total tariff revenue andconsumer surplus, respectively, Wi

j=nijτij j∈ (P, ROW) is gross welfare

of foreign industries i from exports to the home county market, andW=∑i (niHτiH)+TR+CS is national welfare. Taking the first ordercondition of the joint welfare functionwith respect to the ROW importtariff rate and rearranging it, one obtains the expression for theequilibrium trade policy:

eiτROWi

pROWi

= −1σ i

+ σ i−1ð Þ τPi

pPisPi +

aa + α

σ i−1σ i

sHi +1

a + ασ i−1σ i

IHi sHi +

+bIPia + α

σ i−1σ i

sPi +cIROWi

a + ασ i−1σ i

sROWi −1� � ð10Þ

where sji =njip

jix

ji

PiXidenotes the share of country j firms on the Canadian

market for product i at the tax-included prices. On the left-hand sideof Eq. (10), (τiROW/piROW) is the ad-valorem tariff on the ROW imports,which is multiplied by the price elasticity of demand for the ROWimports εi. Therefore, as in the benchmark GHmodel, trade protectionis inversely related to the import demand elasticity.8 The first term onthe right-hand side is negative: the model predicts that with moredifferentiated varieties will receive import subsidy. This result is adirect consequence of monopolistic competition model with specificimport tariff.9

The second element on the right-hand side shows the positiverelation between the FTA external and internal tariffs and, followingBagwell and Staiger (1997), reflects a tariff complementarity effect: ifthe tariff rate for the partner country is high, it is optimal for thegovernment to raise the external tariff as well. Intuitively, an increasein the within-FTA tariff rate causes a decline in imports from the ROW,and tariff revenue collected on the ROW imports is higher for higherτROW. This, in turn, raises imports from the partner country, thatgenerates more tariff revenue for higher partner country tariff rate.The tariff complementarity effect is proportional to the market shareof the partner country firms siP and is stronger if the partner countryand the ROW exports are close substitutes.

In contrast to the benchmark case, even for unorganized industriesprotection may still be positive due to the imperfectly competitivemarket structure, as emphasized by the third term, since thecoefficient a

a + αis positive. Because the share of domestic firms on the

market reflects their ability to capture protection benefits, and thetariff is more effective in re-distributing consumers expendituretowards domestic varieties when substitution elasticity is high, thetariff level is proportional to si

H and increasing with σ. The fourth termis similar to the benchmark GH model: a politically organizeddomestic industry receives more protection from the government.Moreover, the level of protection is higher if domestic and importedvarieties are close substitutes and if the domestic sector is relativelylarge, as the domestic lobby has more to gain from protection in thiscase.

8 It should be noted that without the MFN rule, a set of equilibrium tariffs for allimporters would be determined by a system of simultaneous equations with thenumber of equations being equal to the number of importing countries. With the MFNand the FTA, the number of equations goes down to two. However, under completetrade liberalization agreement, a within-FTA tariff is exogenously set to zero and thesecond term on the right-hand side of (10) vanishes.

9 In the model of monopolistic competition, the effect of a specific tariff on price isamplified by producer's markup. Therefore, for low σ the price elasticity with respectto tariff is high and the gain in consumer surplus from a subsidy outweighs theincrease in government's expenditure.

11


The fifth term reflects the effect of political activity by partnercountry firms on the national trade policy. Iip enters the equationpositively, making protection more likely in those sectors wherepartner country exporters are organized into lobbying groups andwhere product varieties are closer substitutes. Similarly to thedomestic lobby, the effect of partner country firms lobbying on theimport tariff is proportional to their market share.

The last term is negative and reflects the effect of lobbying effortsby the ROW firms to reduce protection. As before, the scaling factor σ i−1

σ i

reflects higher motivation by the ROW firms to lobby for tradeliberalization when the degree of substitution between varietieswithin a given industry is high, but unlike domestic and partnercountry lobbying, the ROW lobbying intensity declines with themarket share. The intuition behind this result is an increased damagefrom protection for small ROW industries, and as a consequence theseindustries will resist tariff increase more intensively.10

As in the GH model, domestic and partner country's lobbyingresults in overprotection and welfare reduction relative to the first-best outcome. The presence of an organized foreign lobby from theROW may help to (partially) restore the optimal level of import tariffsand raise national welfare as a counter pressure to domestic or USlobbying for protection. However, the presence of the ROW lobbyingalone causes underprotection and is thus welfare-reducing. Given thatthe number of politically organized domestic or US industries is onaverage twice of the number of organized ROW industries, the overallnet effect from the presence of the partner country lobbying is likely tolead to welfare reduction, whereas the overall welfare effect of theROW lobbying activity is unambiguous: the effect is positive if ROWfirms counter-lobby against the effort of domestic and partner countryfirms to raise protection and negative if ROW firms form a singleorganized lobbying group in the sector.

Eq. (10) motivates the following form of the estimation equation:

Yi = β0 + β1sHi + β2IHi sHi + β3IPi s

Pi + β4IROWi 1−sROWi

� �Yi =

σ i

σ i−1

� eiτROWi

pROWi

+1σ i

! ð11Þ

β1 =a

a + α; β2 =

1a + α

; β3 =b

a + α; β4 = −

ca + α

ð12Þ

In Eq. (10) the inverse elasticity was taken on the left-hand sideand both sides weremultiplied by σ i

σ i−1because substitution elasticity is

likely to be measured with error. Using Eq. (12), the four coefficientestimates of the reduced form (11) can be used to derive fourstructural parameters of the model.

4. The data

The empirical section of this paper estimates the effects ofdomestic, partner country and ROW lobbying activity on the Canadianpost-NAFTA trade policy. Given the relative size of the US andCanadian economies, the effect of US lobbying in Canada will beconsiderably larger than the effect of Canadian lobbying in the US.Therefore, focusing on Canadian data is particularly advantageous forthe empirical analysis of foreign lobbying under the FTA. This study isconducted for 249 Canadian 6-digit NAICS manufacturing sectors(NAICS 31–33) for the period of 1996–97. The US was treated as a

10 This result follows from the Cobb-Douglas utility function. Fixed productexpenditure shares imply that the import tariff imposed on one variety will raiseconsumer's expenditure on all varieties through aggregate price index proportionallyto their market shares because large market share is indicative of the number ofvarieties produced in a given country, productivity advantage, and consumerspreferences toward varieties produced in that country. Therefore, the higher is theROW market share (and the lower is the share of other varieties), the less harmful isthe import tariff for the ROW exporters.

Canadian FTA partner country, while all other countries that have nopreferential trade agreements with Canada were aggregated intoROW.11 The estimation of Eq. (10) requires the following data: themeasure for trade protection, imports by the country of origin and bysector, domestic output by sectors, substitution and price elasticities,political organization dummies, and three sets of instruments formarket shares.

4.1. Protection measures and market shares

Domestic manufacturing shipments data for 249 NAICS-6 indus-tries are provided by Industry Canada. The values of Canadian imports,as well as customs duties collected, were obtained from StatisticsCanada at the HTS-10 level and aggregated to NAICS-6 using theconcordances tables from the International Trade Division of StatisticsCanada.

Primarily, I use two measures for trade barriers: import tariffs andthe share of imports that is subject to non-tariff trade restrictions.Although the original GH model was meant to analyze the politicaleconomy of import tariff formation, tariff rates are often argued to bean imperfect measure of trade protection for the analysis ofendogenous trade policy formation in the presence of WTO tariffregulation. With limitations on the magnitude of tariffs imposed bythe WTO, organized interests would seek non-tariff protection fromimport competition that are adopted unilaterally by differentcountries, as opposed to tariffs that are set cooperatively in WTOnegotiations. Nevertheless, I use tariffs as a second measure ofprotection. Even in the presence of WTO tariff regulation, tariffs canstill reflect lobbying activity of Canadian and US interest groupsthrough smaller (larger for ROW lobbying) tariff reductions duringWTO negotiations that take place after trade agreement is signed. Ifthe government takes interests of domestic and partner country(ROW) lobby groups into account during negotiations, then organizedindustries are likely to receive smaller (larger) tariff reductions duringWTO negotiations.

In light of this, I used tariff, Non-Tariff Barriers (NTBs) andprotection coverage share as a measure of protection. Ad-valoremtariff rates were obtained as the ratio of aggregated duty collected bycustoms over the value of imports.12 NTBs for Canadian imports wereobtained from the TRAINS database maintained by UNCTAD, whichshows the proportion of imports that is covered by one or morequalitative restrictions. These datawere available at the HS-6 level andwere aggregated into NAICS-6 groups. In addition, the protection sharevariable was constructed as the share of Canadian imports that issubject either to the positive import tariff or NTBs.

Descriptive statistics for protection measures and market sharesare presented in Table 1. In 1997 the average tariff rate, NTBs andprotection coverage ratios for the ROW imports were 4.8%, 18.2% and77.5%, respectively. Tariffs and NTBs are highly correlated both withinand outside of the FTA, which implies that different measures ofprotection are still highly complementary.

4.2. Political organization dummies

Many previous studies that have tested the GH model empiricallyused firm-level political contributions to assign the value for thepolitical organization dummy variable.13 Although these data are

Mexico also have preferential market access in Canada in 1997 but since there is noevidence on the presence of Mexican lobby in Canada in the data. Israel is anothercountry that had trade agreement with Canada at that time and there are Israeli firmslobbying trade interest in Canada. But since this FTA came into force on January 1, 1997,it is hard to classify these firms according to market access using 1996-97 data.Therefore, I treat US as a single Canadian FTA partner country.12 Therefore, tariff measure controls for some non-tariff distortions as well, such asantidumping or countervailing duties.13 For example, Goldberg and Maggi (1999), Gawande and Bandyopadhyay (2000),Facchini et al. (2006), Bombardini (2008).

Table 1Descriptive statistics for protection measures and market shares, 1997

US ROW US ROW Canadian US ROWUS ROW NTB NTB Protection Protection Market Market MarketTariff Tariff Coverage Coverage Share Share Share Share Share

Mean 0.013 0.048 0.172 0.182 0.563 0.775 0.661 0.223 0.116Median 0.003 0.034 0 0 0.646 0.986 0.683 0.205 0.075Standard Deviation 0.076 0.053 0.314 0.307 0.419 0.328 0.222 0.162 0.134Minimum 0 0 0 0 0 0 0 0 0Maximum 1.17 0.54 1 1 1 1 1 0.70 0.64Corr. with tariff 1 1 0.46 0.68 0.14 0.45Corr. with NTB 0.46 0.68 1 1 0.45 0.35Corr. with protection share 0.14 0.45 0.45 0.35 1 1No. of observations 248 248 248 248 248 248 248 248 248


available for Canada for 1997 and afterwards, this paper uses adifferent approach. As was previously mentioned, foreign corpora-tions prefer direct lobbying to political contributions becausetransparency of political contributions may raise concerns aboutforeign interference into political processes. Furthermore, sincedifferent means of political involvement are highly correlated (Hansenand Mitchell, 2000), direct lobbying seems to be an appropriatemeasure for domestic political activity as well.

In this work, the degree of political activity in an industry ismeasured by the number of lobbyists representing the corporateinterests of that industry. The Lobbyists Registration Act (LRA) requiresevery individual to register at the Lobby Registrar Canada if the personseeks a meeting or a phone call to any public office holder regardingthe development, modification or cancellation of legislative proposals,regulations, public policies and programs. The assumption thatpolitical contributions will be ineffective for the determination oftrade policy without such contact seems to be reasonable and,therefore, political contributions should be followed up by a personalcontact with a policymaker. For this reason, the number of registeredlobbyists is used to measure firm-level lobbying intensity within anindustry.

The main advantage of this data set is the large amount of detailedinformation lobbyists are required to submit. This includes informa-tion on the business address of a corporation that benefits fromlobbying, its subsidiaries and headquarters, and the objective of themeeting with a public office holder. This information is very helpful indetermining the “nationality” and industrial affiliation of lobbyistsrepresenting interests of multi-product multinational corporations.Another advantage of this data set is that it gives a very narrowdefinition of a lobbyist. Any person representing his or her owninterests, and who is not being paid for arranging the meeting withthe public office holder, is not obliged to register. This removesinformation on the very small firms. Large firms, which have highlobbying power and can effectively influence the decisions of policy-makers, typically use the service of professional consultants orcorporate lobbyists, who are required to register.

Firms were assigned a NAICS-6 industry code using the CanadianCompany Capability database maintained by Industry Canada. As a

Table 2Descriptive statistics for the number of lobbyists, 1996–97

Canada US ROW

Average number of lobbyists per sector 1.94 1.08 1.15Median number of lobbyists 1 0 0Standard Deviation 3.30 2.06 2.32Minimum 0 0 0Maximum 18 12 14Total number of lobbyists 481 267 284% of sectors with at least one lobbyist 0.50 0.32 0.35% of sectors with at least two lobbyists 0.38 0.26 0.26% of sectors with at least three lobbyists 0.23 0.19 0.15

result, each lobbyist is matched to at least one industry, and forcompanies that report multi-industry activity the number of lobbyistswas replicated over all sectors of operation. Assigning an industrycode to some multi-product firms still involves some degree ofdiscretion. For example, some firms in the automobile sector operatein more than ten NAICS-6 industries. Since the number of such firms isrelatively small, I assigned each firms to no more than three mainNAICS codes using different information sources: the CanadianCompany Capability database, the Federal Corporations Registry andthe North America Compustat database. The databases listed aboveallow assigning industry codes to US and ROW firms. The LRA alsorequires lobbyists to declare a “subject-matter in respect to which anindividual undertakes to communicate with a public office holder.” Inmany cases, information on the purpose of lobbying activity reportedin the lobbyist registration form allowed me to attribute a multi-product firm to a single (or a small number of ) NAICS codes wherefirm's lobbying activity is mostly focused on.

National affiliation of each firm that a particular lobbyist isrepresenting was determined from location of its manufacturingfacilities14 and constructed from two sources. First, the lobbyist regis-tration form requires registrants to provide “the name and businessaddress of the parent corporation and those subsidiaries which directlybenefit from the lobbying.” Sometimes, lobbyists provide incompleteinformation and do not include information on domestic and/or foreignsubsidiaries. In this case the datawas complemented by the informationfrom other databases mentioned previously. Again, quite often the“nationality”of thefirmwasdeterminedby the “objective” sectionof thelobbyist registration form.15

There are two more advantages of using lobbyists' registry dataover using political contribution data. First, all lobbyists are requiredto identify the broad subject matter of their lobbying activity and inthis paper I restrict the sample to only lobbyists concerned with tradepolicy issues. It allows one to isolate effectively firms lobbyingparticularly for a change in trade policy from those lobbying for otherbroad policy issues such as tax policy, environment, etc. This isespecially a problem for domestic lobbyists: on average, only one outof eight lobbyists, representing the domestic manufacturing sector, isconcerned with trade policy. Therefore, pooling political contributionsby all domestic firmsmay cause serious measurement problems in thepolitical organization variable for trade policy analysis.

Second, the main channel used by foreign firms to lobby theirinterest in Canada is through local subsidiaries, which distribute the

14 For example, a US company that has manufacturing facilities in Canada is treatedas Canadian firm.15 For instance, a lobbyist of Toyota Canada Inc, Ontario-based subsidiary of ToyotaMotor Corp., was attributed to the ROW on the basis of the meeting purpose to “secureinternational trade for automobile engines” as these engines are manufactured andshipped from Japan. Moreover, this lobbying objective statement allows to pin downthe lobbyist to one specific NAICS6 industry.

Table 3Descriptive statistics for the elasticity of substitution and price elasticity

Elasticity of substitution Elasticity of import demand

Mean 5.83 5.32Standard Deviation 3.06 2.97Median 4.98 4.52Minimum 1.43 0.17Maximum 21.47 21.36No. of observations 248 248


imported goods within Canada. Formally, these firms should beassigned to the service sector (NAICS 41–45) and dropped from thesample, but this would substantially underestimate lobbying effortsby foreign firms. For each service firm concerned with internationaltrade policy issues I used the company profile and lobbying objectivesinformation to assign an appropriate manufacturing industry codeand country if the firm has a manufacturing headquarter. Therefore,lobbying data allows to control more effectively for foreign lobbyingactivity in Canada.

The amendment to the LRA announced in 1995 introduced severalimportant refinements that made it more desirable to use post-1995lobbying data. First, for the purpose of transparency, the lobbyists'registry database became available for research purposes. Second, thisamendment extended the amount of information that must bereported. But most importantly, it made disclosed information morecomplete and reliable. For the first time lobbyists were obliged toprovide all the information, and effective enforcement devices wereintroduced to encourage better compliance. It extended the power ofthe Lobbyists Registrar, which was authorized to seek clarification ofinformation submitted. The registrar was allowed to conduct an auditof provided information and, when necessary, investigate theprovided information and impose sanctions for violating the LRA.16

For these reasons, the data for political activity by firms werecollected for the 1997 election cycle and complemented with the 1996lobbying data to take into account a possible small lag in trade policyresponse to lobbying efforts. Therefore, for each industry I calculated thetotal number of Canadian, US and ROW lobbyists and, similarly to otherstudies, several thresholds for the number of lobbyists in an industrywere set to determine the values of political organization dummies.Descriptive statistics for the number of lobbyists is provided in Table 2.

4.3. The elasticities of substitution

To my knowledge, there are no studies to date that estimatesubstitution elasticities for Canadian NAICS-6 industries, especiallywithin a framework of monopolistic competition. In this study,substitution elasticities were estimated using the approach byFeenstra (1994), recently applied by Broda and Weinstein (2006) toa large set of US imported commodities.

Table 3 presents the summary statistics.17 As a robustness check, Iestimated the substitution elasticities for NAICS industries at variouslevel of aggregation and verified that more aggregated commoditiesare more differentiated: the average value of σ decreased from 5.85 to5.34 and 4.56 while moving respectively from six to five and fourdigits NAICS. As another robustness check, I estimated US elasticity ofsubstitution using the same estimation procedure, time period, andindustry classification. Presumably, consumers in Canada and the USshould have similar tastes, and varieties that are close substitutes inCanada should be close substitutes in the US as well. This suggestion issupported by 0.69 correlation coefficient.

4.4. Instrumental variables

In Eq. (10), market shares are likely to be determined simulta-neously with the tariff rates and should be properly instrumented.Trefler (1993) proposed to instrument the import penetration ratiowith industry factor endowments as the measure of comparative

16 In 2001 an independent study of compliance to the LRA was conducted by KPMGConsulting Inc. (2001). The already-registered lobbyists were asked if they were awareof any non-compliance behavior. Reported results indicate that 50% of consultantlobbyists and 15% of corporate lobbyists were aware of non-registered lobbying, whilethey evaluated the aggregate compliance rate at 70% and 100%, respectively. In general,compliance was perceived to be high, although non-compliance behavior is still animportant issue.17 Price elasticities of the ROW import demand were calculated from consumersdemand function: ei =

AqROWiApROW

i

pROWiqROWi

= σ i− σ i−1ð ÞsROWi

� �.

advantage independent of the level of protection. Following thisapproach, a list of instruments for the Canadian market shareincludes: the share of production to non- production workers, thecapital stock in machinery and construction, inventories, and theconsumption of fuel and electricity. All of these data are provided byStatistics Canada. The same list of instruments was constructed for theUS market share in Canada using the US Census data.

To instrument the ROW share in the Canadian market, the“gravity”-type distance measure between Canada and the averageexporter was constructed. For every product, the pair-wise log-distance between Canada and the exporting country was weighted bythe share of this country in the global export of the product.18 The totalexports by country and by sector were constructed using the UNCTADdatabase. The data on geographic distance, weighted by populationdensity and economic activity within each country, were taken fromthe Centre d'Etudes Prospectives et d'Informations Internationales.The rationale for using this “distance” variable is the following: if mainproducers of a particular good are located far from Canada,transportation costs are high and the ROW share in the Canadianmarket is likely to be small regardless of Canadian trade policy.

Political organization dummies are likely to bemeasuredwith errorand are potentially endogenous. To instrument the US and Canadianpolitical organization dummies, I use the information industrialconcentration, such as shares of big and medium firms and the CR-4concentration ratio. Industrial concentration is mostly technologicallydetermined and at the same time it is easier for firms in moreconcentrated industries to overcome free-riding problem and form alobby group. The ROW lobbying intensity is instrumented with thecountry-level CR-4 concentration ratio on Canadian imports market.

The right-hand side variables of Eq. (11) include non-linearcombinations of endogenous variables and to consistently estimateits coefficients I use LIML estimator proposed by Kelejian (1971) withBekker (1994) standard error correction.19 Hansen et al. (2006)demonstrated that this approach has better small sample propertiesthan 2SLS and is asymptotically correct in the presence of manyinstruments and many weak instruments.

5. Results

5.1. Test of a benchmark GH model

As a starting point, I will present the results on a GH version of themodel with homogeneous goods to test how well the new data onCanada can fit the benchmark model and compare its performancewith the results of previous empirical studies. Since in the benchmarkmodel markets are perfectly competitive and import supply isinfinitely elastic, there is no reason for foreign firms to participate inlobbying, and in the benchmark case I will consider only the effect ofdomestic lobbying groups on the home country trade policy.

18 Since the exporter's share on the Canadian market is endogenous, I use the shareon the global market excluding Canada, which is unaffected by Canadian tariff rate.19 To account for non-linearity of the endogenous variables, the list of instrumentsalso includes quadratic terms and cross products of exogenous variables, selected onthe basis of their correlation with endogenous variables.

20 The government's objective function C+aW is equivalent to a1C+a2(W-C ), inwhich a = a2

a1−a2, a1 is the weight on political contributions and a2 is the weight on a

welfare net of political contributions. Therefore when a is much greater than one,a1 = a + 1

a

� �a2≈a2.

Table 4Estimation results for the benchmark GH model with different protection measures

Dependent variable

Tariffs NTBs Protection share

(1) (2) (3) (1) (2) (3) (1) (2) (3)XHi

Mi

−0.0059⁎⁎ −0.0057⁎⁎ −0.006⁎⁎ −0.0349⁎⁎ −0.0348⁎⁎ −0.0354⁎⁎ −0.0163 −0.0164 −0.0177(0.0024) (0.0024) (0.0024) (0.0139) (0.0136) (0.0142) (0.0200) (0.0198) (0.0204)

XHi

MiIHi 0.0097 0.0111 0.0102 0.0271 0.0313 0.0249 0.0826 0.0864 0.0831

(0.0081) (0.0087) (0.0088) (0.0448) (0.0486) (0.0484) (0.0709) (0.0784) (0.0768)

Structural parameters

α 0.61 0.51 0.59 1.29 1.11 1.42 0.20 0.19 0.21(1.279) (0.946) (1.221) (3.708) (2.897) (4.629) (0.928) (0.833) (0.928)

a 102.5 89.3 97.8 35.6 30.9 38.8 11.9 11.4 11.8(168.7) (125.3) (161.9) (95.05) (73.88) (119.4) (25.12) (23.00) (25.88)

N 248 248 248 248 248 248 248 248 248Log-likelihood −191.6 −192.0 −194.8 −637.8 −635.9 −640.7 −688.2 −693.1 −689.4AIC 1.569 1.573 1.596 5.168 5.152 5.191 5.574 5.614 5.584

Notes: ⁎, ⁎⁎ and ⁎⁎⁎ denote statistical significance at 10%, 5% and 1%, respectively. Standard errors in parenthesis. In columns (1), (2) and (3) industry is considered as politicallyorganized if it is represented by 1, 2 and 3 lobbyists, respectively.


Following Grossman andHelpman (1994), the optimal import tarifffor a small open economy with politically organized sector-specificfactors of production takes the following form:

eiτROWi

pROWi

= −α

a + αXHi

Mi+

1a + α

XHi

MiIHi ð13Þ

where XiH is a domestic value of shipments and Mi is a total value of

imports from the ROW. The model predicts that the inverse importpenetration ratio enters the equation negatively, while the coefficienton its interactionwith political organization dummy is positive. Table 4represents estimation results for Eq. (13) using tariffs, NTBs andprotection share data as a measure of Canadian trade barriers.

As in Goldberg and Maggi (1999), several thresholds were used forthe number of lobbyists in the construction of political organizationdummies to verify that the results are not driven by the way thesedummies are assigned. In the first column of Table 4 an industry isconsidered to be politically organized if it is represented by at leastone lobbyist. For the second and third columns the threshold is twoand three lobbyists, respectively.

First, consider the estimation with tariffs as a protection measure.The estimates of the regression model (13) are of correct signs acrossall specifications: politically organized sectors receive more protec-tion, while protection in unorganized sectors is negative and increaseswith import penetration. The latter result is statistically significant at a5% confidence level. In organized sectors protection declines with theimport penetration, but this result is not statistically significant. Themodel estimates are very robust to the way political organizationdummies are constructed. When trade barriers are measured withNTBs and protection share, results correspond closely to thoseobtained for the tariff equation: the parameter estimates preservecorrect signs, but are estimated with less precision when tradebarriers are measured with protection share. Overall, the estimates ofthe structural model (13) using Canadian data are generally in the linewith the results of the studies by Goldberg and Maggi (1999) andGawande and Bandyopadhyay (2000).

The estimates of the structural parameters of the model varyconsiderably across different measures of protection. However, thevariances of these parameters are very high and one cannot rejecthypotheses that both α and a are the same across all specificationsconsidered. The fraction of the population represented by a lobby, α, isestimated to be around 0.6 for the specifications with tariffs, 0.2 forspecifications with protection shares, and greater than one for specifica-tions with NTBs. In general, 95 confidence interval for α includes thewhole [0; 1] interval and themodel does not allowone to obtain a precisemeasure for α. Nevertheless, the obtained results do not contradict the

previous estimates of by Goldberg and Maggi (1999) and Gawande andBandyopadhyay (2000), who estimated the share of the populationrepresented by interest groups to be around 0.85 and unity, respectively.

The estimates of the government's political bias vary from 10 in thespecificationwith shares to 100 in specificationwith tariffs. The valuesof the parameter a greater then ten imply that the government assignsapproximately equal weights to political contributions and to anational welfare net of political contributions, which supports theresults of studies that use US data.20

Overall, the results of the GH model with Canadian data arebroadly consistent with those obtained by Goldberg and Maggi (1999)and Gawande and Bandyopadhyay (2000) and other studies for theUS. These results will serve as a benchmark against which the resultsof the monopolistic competition model with foreign lobbying and FTAparticipation will be compared in the next section.

5.2. Estimation results for the monopolistic competition model withforeign lobby

In this section I present the estimation results for the politicaleconomy model of trade with monopolistic competition, FTAmembership and two groups of foreign lobbies. The results from theEq. (11) appear in Table 5, in which several measures are used tomeasure trade barriers. Columns with different numbers denotedifferent threshold levels for construction of the political organizationvariable. Columns (1) and (2) report the results when an industry isassumed to be politically organized if it has at least one and threelobbyists, respectively. In column (3) a country's j industry isorganized if it is represented by at least three lobbyists and accountsfor strictly more than one third of a total number of lobbyists in thatindustry. With this measure industries with positive but relativelysmall number of lobbyists relative to thewhole industry are still codedas not organized.

First, note that for any measure of protection and politicalorganization, the coefficient β1 is positive and almost alwaysstatistically significant, implying that domestic industries receive apositive level of protection regardless of political economy factors.This is consistent with the assumption of imperfect competition andprediction of the model that the welfare-maximizing governmentalways finds it worthwhile to protect home country producers againstcompeting importers when domestic and foreign products are close

Table 5Estimation results for the monopolistically competitive model (10) with foreign lobbying and different protection measures

Dependent variable

Tariffs NTBs Protection share

(1) (2) (3) (1) (2) (3) (1) (2) (3)

SiH 0.072 0.179⁎⁎ 0.210⁎⁎ 0.828 0.939⁎ 0.868⁎ 0.915 1.757⁎⁎ 1.741⁎⁎

(0.093) (0.085) (0.084) (0.636) (0.570) (0.524) (0.848) (0.802) (0.792)IiHSi

H 0.472⁎⁎⁎ 0.557⁎⁎⁎ 0.528⁎⁎⁎ 2.320⁎⁎⁎ 3.512⁎⁎⁎ 2.895⁎⁎⁎ 2.966⁎⁎⁎ 3.091⁎⁎⁎ 3.086⁎⁎⁎(0.059) (0.070) (0.069) (0.403) (0.468) (0.429) (0.537) (0.655) (0.648)

IiPSi

P 0.150 0.156 0.345⁎ 3.047⁎⁎⁎ 2.152⁎ 3.033⁎⁎⁎ 4.064⁎⁎⁎ 5.041⁎⁎⁎ 5.515⁎⁎⁎(0.158) (0.167) (0.174) (1.079) (1.119) (1.082) (1.438) (1.568) (1.634)

IiROW (SiROW−1) −0.109⁎⁎ −0.264⁎⁎⁎ −0.246⁎⁎⁎ 0.005 −0.914⁎⁎ −0.715 −1.910⁎⁎⁎ −2.542⁎⁎⁎ −2.491⁎⁎⁎

(0.051) (0.067) (0.073) (0.345) (0.447) (0.453) (0.460) (0.626) (0.684)


α 1.97 1.47 1.50 0.07 0.02 0.05 0.03 −0.25 −0.24(0.270) (0.228) (0.244) (0.313) (0.188) (0.208) (0.330) (0.317) (0.310)

a 0.15 0.32 0.40 0.36 0.27 0.30 0.31 0.57 0.56(0.238) (0.191) (0.201) (0.351) (0.200) (0.222) (0.360) (0.350) (0.340)

b 0.32 0.28 0.65 1.31 0.61 1.05 1.37 1.63 1.79(0.356) (0.314) (0.355) (0.591) (0.352) (0.428) (0.623) (0.691) (0.698)

c 0.23 0.47 0.47 0.00 0.26 0.25 0.64 0.82 0.81(0.112) (0.131) (0.156) (0.153) (0.131) (0.164) (0.193) (0.261) (0.290)

N 248 248 248 248 248 248 248 248 248Log-likelihood −74.2 −70.1 −68.9 −550.7 −542.1 −522.1 −622.0 −625.7 −624.3AIC 0.64 0.61 0.60 4.48 4.41 4.25 5.06 5.09 5.08

Notes: ⁎, ⁎⁎ and ⁎⁎⁎ denote statistical significance at 10%, 5% and 1%, respectively. Standard errors in parenthesis. In columns (1) and (2) an industry is politically organized if it isrepresented by 1 and 3 lobbyists, respectively. In column (3) an industry is politically organized if it is represented by at least 3 lobbyists and accounts for strictly more than one thirdof the total number of lobbyists in that industry.

Table 7


substitutes and markets are imperfectly competitive. The positivelevel of protection for unorganized sectors is in contrast to thebenchmark GH model and finds strong support in the data. As thetheory predicts, among politically unorganized sectors protectionincreases with the share of domestic firms on the market. Whenprotection is measured with tariffs, the point estimate for β1 in themost preferred specification in terms of the log-likelihood function(column (3)) is 0.21, which converts to the welfare-maximizing ad-valorem import tariff of 2.2% for an average Canadian industry, giventhe average Canadian market share, price and substitution elasticitiesof 0.66, 5.32 and 5.83, respectively. In terms of the optimal level ofNTBs and protection share, the welfare-maximizing NTB coverage forthe average industry is estimated to be 8.9% of total imports fromoutside of the FTA, while the welfare-maximizing share of importssubject to any trade restriction is 17.9%.

The effect of a politically organized domestic lobbying (coefficientβ2) is always estimated to be positive and very significant, indepen-dently of the construction of the political organization dummy and themeasure of trade distortion. Everything else being equal, activedomestic lobbying in the industry leads to a higher level of protectionand this effect is significant and robust across all specifications. Table 6shows the change in the tariff rate and NTB had the average domestic

Table 6The effect of an active lobbying group presence on the level of protection in an averageindustry

Protection measure (industry average, %)

Tariffs (4.81) NTB (18.19) Protection share (77.53)

(1) (2) (3) (1) (2) (3) (1) (2) (3)

Activedomesticlobbying

4.85 5.72 5.42 23.83 36.07 29.73 30.46 31.74 33.49

Activepartnercountrylobbying

0.52 0.54 1.19 10.54 7.44 10.49 14.06 17.44 25.65

Active ROWlobbying

−1.5 −3.62 −3.38 −0.68 −12.55 −9.81 −26.22 −34.9 −42.77

or foreign industry changed its status from unorganized to organized.The results imply that the presence of a politically organized domesticlobby tends to increase import tariffs by 5.4% for the average industry,the NTB coverage ratio by 29.7%, and the protection share by 31.7%.

The novel results of this section are the estimates of coefficients β3

and β4. The coefficient β3 measures the effect of the FTA partnercountry's lobbying and is always estimated to be positive, althoughwith tariffs as a measure of protection it is insignificant in twospecifications. As expected, the effect of US lobbying is much strongeron NTBs than on tariffs. This is consistent with the initial hypothesisthat in the presence of tariff regulation by the WTO, domestic andpartner country lobbyists can affect tariffs only by resisting tomultilateral trade liberalization during post-NAFTA WTO rounds oftrade negotiations. In the short run, the effect of US lobbying onCanadian import tariff may be not as pronounced as the effect onNTBs, adopted unilaterally by Canadian government.

The effect of the ROW lobbying (coefficient β4) is always negativeand significant at 5% except for two specifications with NTBs,presumably because foreign lobbies are more effective in lobbying

The effect of domestic and foreign lobbying on Canadian average levels of protection

Protection measure (industry average, %)

Tariffs (4.81) NTB (18.19) Protection share (77.53)

(1) (2) (3) (1) (2) (3) (1) (2) (3)

Averageeffect ofdomesticlobbying

2.65 1.41 1.33 13.04 8.88 7.31 16.67 7.81 7.8

Averageeffect ofpartnercountrylobbying

0.22 0.13 0.21 4.53 1.82 1.87 6.04 4.26 3.4

Averageeffectof ROWlobbying

−0.06 −0.65 −0.48 0 −2.24 −1.38 −10.51 −6.24 −4.81

Table 8Estimation results for the monopolistically competitive model (10) with additional controls

Dependent variable

Tariffs NTB Protection share

(1) (2) (3) (1) (2) (3) (1) (2) (3)

SiH 0.210⁎⁎ 0.307⁎⁎⁎ 0.210⁎⁎⁎ 1.503⁎⁎⁎ 1.555⁎⁎⁎ 1.247⁎ 1.999⁎⁎⁎ 2.727⁎⁎⁎ 1.941⁎⁎⁎

(0.099) (0.089) (0.091) (0.686) (0.610) (0.658) (0.909) (0.857) (0.856)IiHSi

H 0.412⁎⁎⁎ 0.506⁎⁎⁎ 0.393⁎⁎⁎ 2.024⁎⁎⁎ 3.302⁎⁎⁎ 2.165⁎⁎⁎ 2.506⁎⁎⁎ 2.630⁎⁎⁎ 2.498⁎⁎⁎(0.058) (0.067) (0.056) (0.406) (0.460) (0.405) (0.538) (0.646) (0.527)

IiPSi

P 0.194 0.224 0.307⁎⁎ 3.254⁎⁎⁎ 2.636⁎⁎⁎ 3.432⁎⁎⁎ 4.401⁎⁎⁎ 5.165⁎⁎⁎ 5.727⁎⁎⁎(0.159) (0.163) (0.156) (1.104) (1.116) (1.127) (1.463) (1.568) (1.467)

IiROW (SiROW−1) −0.100⁎⁎⁎ −0.260⁎⁎⁎ −0.202⁎⁎⁎ 0.044 −0.878⁎⁎ −0.458 −1.849⁎⁎⁎ −2.563⁎⁎⁎ −1.989⁎⁎⁎

(0.051) (0.065) (0.055) (0.355) (0.449) (0.397) (0.470) (0.631) (0.517)CR4 0.029⁎⁎⁎ 0.038⁎⁎⁎ 0.032⁎⁎⁎ 0.017⁎⁎⁎ 0.022⁎⁎⁎ 0.021⁎⁎⁎ 0.014⁎ 0.020⁎⁎ 0.009⁎

(0.011) (0.011) (0.010) (0.007) (0.007) (0.007) (0.008) (0.010) (0.005)K/L −0.075⁎ −0.067 −0.075⁎ 0.013 0.198 −0.051 −0.018⁎⁎⁎ −0.017⁎⁎⁎ −0.018⁎⁎⁎

(0.045) (0.043) (0.042) (0.310) (0.296) (0.306) (0.004) (0.004) (0.004)Wage −0.094⁎⁎⁎ −0.106⁎⁎⁎ −0.087⁎⁎⁎ −0.060⁎⁎⁎ −0.068⁎⁎⁎ −0.054⁎⁎⁎ −0.263⁎⁎⁎ −0.397⁎⁎⁎ −0.202⁎⁎⁎

(0.0061) (0.0059) (0.0058) (0.0042) (0.0040) (0.0042) (0.056) (0.057) (0.054)N 248 248 248 248 248 248 248 248 248Log-likelihood −61.4 −53.0 −48.9 −542.3 −530.3 −538.7 −612.1 −614.7 −603.9AIC 0.56 0.49 0.46 4.44 4.34 4.41 5.00 5.02 4.93

Notes: ⁎, ⁎⁎ and ⁎⁎⁎ denote statistical significance at 10%, 5% and 1%, respectively. Standard errors in parenthesis. In columns (1) and (2) an industry is politically organized if it isrepresented by 1 and 3 lobbyists, respectively. In column (3) an industry is politically organized if it is represented by at least 3 lobbyists and accounts for strictly more than one thirdof a total number of lobbyists in that industry.

21 High volatility in the estimates of a across different measures of trade protectioncan be a scale effect as α appears in denominator of all reduced form coefficients. Italso should be pointed out that measuring α is a standard problem in the literatureand its estimation proved to be difficult.


for tariff using WTO mechanisms. According to these results, thepresence of politically organized US industries in Canada leads tohigher Canadian import barriers and the ROW lobbying effort isnegatively correlated with the Canadian protection measures. Thus, itseems safe to conclude that, while our results do not allow researchersto get a precise estimate of β3, they strongly support the hypothesis offoreign lobby differentiation with respect to market access implied bythe FTA and predicted by the theoretical model of Section 3. The pointestimates of coefficient β3 in specifications with the best data fit are0.345 (column (3) for tariffs), 3.033 (column (3) for NTBs) and 5.041(columns (2) for protection share). Table 6 shows the implication ofthese estimates for the observed levels of protection. A politicallyorganized industry in the FTA partner country tends to increase acountry's average import tariff for that industry by 1.2%, average NTBcoverage by 10.5% and average protection coverage by 17.4%. Similarly,the presence of a politically organized group of exporters from outsideof the FTA tends to decrease the average Canadian import tariff by3.4%, the average NTB by 9.8% and the average protection share by34.9% . These numbers are economically plausible and provideadditional support to the estimates of the model.

Estimates of Eq. (11) allow one to derive the values of the structuralparameters of the model. As expected, absolute magnitudes ofcoefficient β4 are always less than β2. Since the strength of domesticlobbying is normalized to one, cb1 implies that ROW lobbying isrelatively less important for Canadian policymakers than lobbying bydomestic firms.With the values of c between 0.25 and 0.5, foreign firmsneed to spend 2–4 time more on lobbying than domestic firms to havethe sameeffectonpolicyoutcomes. This result is not surprisinggiven thelegislative restrictions on contributions by foreign firms in Canada. Inspite of that, policymakers' valuationof contributions fromabroad is stillpositive and significantly different from zero. The estimates of thegovernment's preferences towards political contributions from the FTApartner country (parameter b) depend on the protection measure thatUS firms are trying to affect. In lobbying for tariffs, partner countrycontributions are estimated to be 35–80% less effective than domesticcontributions, while in lobbying for NTBs the effectiveness of domesticand partner country lobbying is approximately equal. Finally, whentrade distortions are measured by protection share, the government'svaluation of political contributions from the FTApartner country relativeto domestic ones rises to 1.63. However, in neither case the hypothesisthatb=1 can be rejected at 95% confidence level, suggestingeither lowerlevels of political risk associated with protecting interests of US firms in

Canada, or high interest and ownership interconnections amongstCanadian and US firms.

At the same time, US contributions seem to be more importantthan contributions from the ROW. Depending on the specification, thegovernment's valuation of US contributions is estimated to be twotimes the valuation of contributions from other countries. Amongpossible reasons are greater proximity of the US and Canadianfinancial systems and a more sophisticated mixture of asset structurethat makes it more difficult to distinguish between Canadian and USfirms (relative to the distinction between Canadian and ROW firms).

The estimates of the population share organized into lobbying varysubstantially across different specifications. In tariff equations, theparameter o is always greater than one, but the standard error is verylarge and 95% confidence intervals almost always overlap with the[0;1] interval. In equations with NTBs and protection shares, pointestimates for o are also very imprecise and in all specifications onecannot rule out the possibility that α is negative.21

As for the relative importance of national welfare for the government,the parameter a is estimated to be 0.3–0.5 with reasonable degree ofprecision as compared to the benchmark GHmodel. It implies that whenorganized interests lobby for a change in tariff policy, the government'svaluation of political contributions is three to five times higher than thevaluation of national welfare net of contribution. The result thatpolicymakers are driven mostly by political contributions they receivefrom different lobby groups sharply contrasts with results obtainedpreviously inaperfectlycompetitive setup:GawandeandBandyopadhyay(2000) estimated a to be over 3, 000; in Goldberg and Maggi (1999) itsvalue is around 70. My own estimates for Canada fall in the range 30–100(see Section 5.1). However, Gawande and Bandyopadhyay (2000)recognized that high estimates of a contradict the empirical evidencethat welfare loss from protection is always greater than the amount ofpolitical contributions policymakers receive in exchange for protection,andonlywith lowvalues ofawewouldobserve enoughvariation in tariffsrelative to welfare-maximizing rates.

Estimates of the reduced from Eq. (11) shown in Table 5 have ameaningful economic interpretation for an average Canadian industry.Given the average market shares of the three groups of firms, price

Table 9Estimation results for the monopolistically competitive model (10) with additional controls and without elasticity adjustment of the dependent variable

Dependent variable


(1) (2) (3) (1) (2) (3) (1) (2) (3)

SiH −0.009 0.006 −0.008 0.065 0.070 0.031 −0.355⁎⁎⁎ −0.239⁎⁎⁎ −0.357⁎⁎⁎

(0.014) (0.012) (0.013) (0.084) (0.075) (0.081) (0.081) (0.079) (0.076)IiHSi

H 0.057⁎⁎⁎ 0.067⁎⁎⁎ 0.051⁎⁎⁎ 0.261⁎⁎⁎ 0.408⁎⁎⁎ 0.267⁎⁎⁎ 0.301⁎⁎⁎ 0.233⁎⁎⁎ 0.277⁎⁎⁎(0.008) (0.009) (0.008) (0.050) (0.057) (0.050) (0.048) (0.060) (0.047)

IiPSi

P 0.002 0.003 0.006 0.336⁎⁎⁎ 0.243⁎ 0.327⁎⁎⁎ 0.231⁎ 0.280⁎ 0.275⁎⁎(0.022) (0.023) (0.022) (0.136) (0.137) (0.139) (0.131) (0.145) (0.130)

IiROW (SiROW−1) −0.012⁎ −0.039⁎⁎⁎ −0.033⁎⁎⁎ 0.036 −0.138⁎⁎⁎ −0.075 −0.208⁎⁎⁎ −0.292⁎⁎⁎ −0.252⁎⁎⁎

(0.007) (0.009) (0.008) (0.044) (0.055) (0.049) (0.042) (0.058) (0.046)CR4 0.031⁎⁎ 0.044⁎⁎⁎ 0.037⁎⁎⁎ 0.019⁎⁎⁎ 0.028⁎⁎⁎ 0.026⁎⁎⁎ −0.052 0.037 −0.079

(0.015) (0.015) (0.014) (0.009) (0.009) (0.009) (0.089) (0.093) (0.086)K/L −0.003 −0.007 0.000 0.002 0.004 0.001 −0.024⁎⁎⁎ −0.025⁎⁎⁎ −0.023⁎⁎⁎

(0.006) (0.0609) (0.059) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004)Wage −0.225⁎⁎⁎ −0.237⁎⁎⁎ −0.216⁎⁎⁎ −0.107⁎⁎⁎ −0.114⁎⁎⁎ −0.098⁎⁎⁎ −0.324⁎⁎⁎ −0.428⁎⁎⁎ −0.272⁎⁎⁎

(0.085) (0.082) (0.080) (0.005) (0.005) (0.005) (0.050) (0.052) (0.048)N 248 248 248 248 248 248 248 248 248Log-likelihood 428.7 436.8 441.7 −22.2 −10.4 −19.0 −13.3 −24.0 −2.5AIC −3.39 −3.46 −3.50 0.24 0.15 0.22 0.17 0.26 0.08


Table 10Estimation results for the monopolistically competitive model (10) when multinationals with manufacturing facilities in Canada being treated as foreign firms

Dependent variable


(1) (2) (3) (1) (2) (3) (1) (2) (3)

SiH 0.079 0.198⁎⁎ 0.267⁎⁎⁎ 0.912 1.079⁎ 1.202⁎⁎ 1.183 1.899⁎⁎ 2.037⁎⁎

(0.093) (0.085) (0.088) (0.641) (0.564) (0.555) (0.858) (0.791) (0.792)IiHSi

H 0.481⁎⁎⁎ 0.598⁎⁎⁎ 0.467⁎⁎⁎ 2.352⁎⁎⁎ 3.618⁎⁎⁎ 3.965⁎⁎⁎ 2.901⁎⁎⁎ 3.427⁎⁎⁎ 3.178⁎⁎⁎(0.059) (0.072) (0.078) (0.403) (0.475) (0.491) (0.540) (0.666) (0.701)

IiPSi

P 0.112 0.219 0.393⁎⁎ 2.619⁎⁎ 3.143⁎⁎⁎ 3.468⁎⁎⁎ 3.903⁎⁎⁎ 5.807⁎⁎⁎ 5.865⁎⁎⁎(0.150) (0.165) (0.176) (1.030) (1.096) (1.101) (1.377) (1.537) (1.573)

IiROW(SiROW−1) −0.093⁎ −0.200⁎⁎⁎ −0.169⁎⁎ −0.080 −1.043⁎⁎ −0.614 −1.822⁎⁎⁎ −2.452⁎⁎⁎ −2.363⁎⁎⁎

(0.050) (0.063) (0.073) (0.343) (0.421) (0.460) (0.459) (0.590) (0.657)


α 1.91 1.34 1.57 0.04 −0.02 −0.05 −0.06 −0.26 −0.33(0.264) (0.205) (0.317) (0.315) (0.181) (0.162) (0.349) (0.283) (0.307)

a 0.16 0.33 0.57 0.39 0.30 0.30 0.41 0.55 0.64(0.234) (0.178) (0.252) (0.350) (0.193) (0.171) (0.381) (0.310) (0.340)

b 0.23 0.37 0.84 1.11 0.87 0.87 1.34 1.69 1.85(0.326) (0.291) (0.412) (0.530) (0.348) (0.306) (0.600) (0.615) (0.664)

c 0.19 0.33 0.36 0.03 0.29 0.15 0.63 0.72 0.74(0.108) (0.110) (0.176) (0.149) (0.119) (0.120) (0.200) (0.211) (0.279)

N 248 248 248 248 248 248 248 248 248Log-Likelihood −74.0 −70.1 −82.0 −552.3 −539.5 −537.4 −624.5 −623.4 −625.8AIC 0.64 0.61 0.70 4.49 4.39 4.37 5.08 5.07 5.09



and substitution elasticities, one can back up the welfare maximizinglevel of protection and the average effect of each lobby group onCanadian trade policy. Table 7 shows the fraction of Canada's averagetariff and NTB that is explained by domestic and foreign lobbying.22 Asa result, the average Canadian industry receives 1.33% import tariffs,7.31% NTB coverage and 7.81% protection share from lobbying bydomestic firms; 0.21% tariffs, 1.87% NTBs and 4.26% protection sharefrom lobbying by partner country firms −0.48% tariffs, −1.38% NTBsand −6.24% protection share from lobbying by ROW firms. Theseresults are economically reasonable and statistically significant.

22 i.e. by calculating a simple average of the fitted values of the dependent variableimplied by the model estimates. Therefore, by construction, the contribution of eachlobby group to the average import tariff and NTB goes down to zero as the threshold forthe number of lobbyists in the definition of political organization dummies goes up.

5.3. Robustness tests

Up to now, we tested structural relationship between trade policyoutcomes and lobbying activity by domestic and foreign interestgroups. However, the main message of the previous section stillholds in a more general setup. Table 8 presents extended regressionsthat include other explanatory variables of trade policy predicted byearlier literature. We expect that more concentrated industries aremore easily get organized and are more effective in lobbying forprotections labor-intensive industries receive more protection fromthe government because in a capital-abundant country these aremore likely to be import-competing industries; and industries withgreater numbers of unskilled low-paid workers are more protectedby the government for social justice reasons. Table 8 indicates thatall additional variables have expected signs and mostly significant.

Table 11J−test for the benchmark and monopolistically competitive GH models

P-value

H0 H1 1 2 3

Dependent variable: tariffsMC GH with foreign lobby is true Benchmark GH is true 0.50 0.55 0.59Benchmark GH is true MC GH with foreign lobby is true 0.25 0.26 0.11MC GH w/o foreign lobby is true Benchmark GH is true 0.42 0.45 0.43Benchmark GH is true MC GH w/o foreign lobby is true 0.37 0.34 0.19

Dependent variable: NTBsMC GH with foreign lobby is true Benchmark GH is true 0.70 0.43 0.52Benchmark GH is true MC GH with foreign lobby is true 0.22 0.07 0.04MC GH w/o foreign lobby is true Benchmark GH is true 0.79 0.47 0.61Benchmark GH is true MC GH w/o foreign lobby is true 0.49 0.15 0.52

Dependent variable: Protection shareMC GH with foreign lobby is true Benchmark GH is true 0.01 0.15 0.00Benchmark GH is true MC GH with foreign lobby is true 0.18 0.18 0.00MC GH w/o foreign lobby is true Benchmark GH is true 0.01 0.21 0.00Benchmark GH is true MC GH w/o foreign lobby is true 0.12 0.30 0.17

Notes: P-values denote the confidence level for rejecting H0. MC stays for “Monopolisticcompetition” version of GH model. In columns (1) and (2) an industry is politicallyorganized if it is represented by 1 and 3 lobbyists, respectively. In column (3) anindustry is politically organized if it is represented by at least 3 lobbyists and accountsfor strictly more than one third of the total number of lobbyists in that industry.


But more importantly, all coefficients of the central concerns, β2, β3

and β4, preserve their signs and significance. Table 9 reports resultsfor specification (11) with additional control variables and withoutadjustment of the dependent variable by elasticity terms. Again, USlobbying activity in Canada is always found to lead to moreprotection, while ROW lobbying is associated with lower levels ofprotection.

To examine the robustness of the results with respect to thedefinition of a foreign lobbyist, Eq. (10) was estimated using the datain which nationality of a lobbyist was determined on the basis of thelocation of its client's headquarter rather than manufacturingfacilities. Since many US multinationals lobbying in Canada operatelocal affiliates, this data transformation affects primarily the numberof US lobbyists raising it to 1.31 per industry. The results are reportedin Table 10. Once again, they indicate positive and significant effect ofthe US lobbying on the level of trade protection in Canada, which isnot surprising given common lobbying interests of Canadian and USmultinational firms. On the other hand, the effect of the ROW lobbyingfor tariffs gets weaker, whichmay reflect conflicting interest of foreignimporters and multinationals serving Canadian market through theirlocal subsidiaries.

5.4. The comparison of the benchmark with monopolistic competition GHmodels

This paper is the first one that estimates empirically the GH modelwith monopolistically competitive market structure. To test thisversion of the model versus the benchmark model with perfect com-petition, I used a J-test for non-nested hypothesis testing (Davidsonand MacKinnon, 1993, p. 388). The testing procedure requires thesame dependent variable, so for the purpose of this section bothsides of Eq. (11) were multiplied by σ i

σ i−1, and 1

σ iwas moved to the right-

hand side. The alternative benchmark specification (13) remainsunchanged.

As in Davidson and MacKinnon (1993), let f(sij, Iij, σi) be a LIMLregression model for the monopolistic competition specification (11),and let g (sih, Iih) be a LIML regression model for the benchmarkspecification (13). The validity of the benchmark GH model can betested using an augmented LIML regression g =g(sih, Iih)+λ1f where fis constructed from fitted values of the LIML estimation of Eq. (11). The

benchmark GH model is rejected in favor of the monopolisticcompetition model if the test for λ1=0 is rejected. In a similar way, themonopolistic competition model is rejected in favor of the benchmarkGH if in the regression model f = f(si, Iij, σi)+λ2ĝ the hypothesis λ2=0 isrejected.

The J-test results for three measures of protection are presented inTable 11. For each measure of protection the first two rows of eachpanel of Table 11 compare the benchmark versus the full monopolisticcompetition model with three lobby groups. However, the original GHmodel does not allow to analyze the behavior of foreign lobbies andthe last two rows of each panel compare GH versus monopolisticcompetition model without foreign lobbying to isolate the effect ofimperfectly competitive market structure on explanatory power of“protection for sale” model. The results of Table 11 suggest thatneither model can be rejected as a correct one when protection ismeasured with tariffs. At the same time, the benchmark GH model isalways weakly rejected in favor of the model with monopolisticcompetition as the monopolistically competitive specification of themodel adds more explanatory power. The second part of each sectionof the table displays J-test results for the benchmark model andmonopolistic competition with only domestic lobbying, using thesame set of instruments. Again, neither model can be rejected, butthese results suggest that simply taking into account the imperfectlycompetitive structure of the market improves the explanatory powerof the “Protection for sale” model.

When protection is measured with NTBs, J-test results indicatethat the benchmark model is rejected in favor of the monopolisticcompetition model with foreign lobbies and weakly rejected in favorof the model without foreign lobbies, while the monopolisticcompetition version is never rejected. This implies that the mono-polistically competitive model developed in this paper captures moreinformation than the model with perfect competition and betterexplains cross-industry patterns of protection.

6. Conclusion

The main objective of this paper is to study the effect of politicaleconomy factors on trade policy of an FTA in the presence of foreignlobbying. This paper accomplishes this goal by modifying the originalGH model through the introduction of a monopolistically competitivemarket structure to allow for the presence of two types of foreignlobbies: lobbies from the FTA partner country and from outside of theagreement. The paper shows that signing a free trade agreement witha politically active partner country leads to a welfare-reducingincrease in import tariffs in the presence of foreign lobbying.

The empirical results suggest that foreign lobbying has asignificant and differentiated effect on Canadian trade policy as aresult of Canada's membership in NAFTA. This is the first paper thatanalyzes differentiation of foreign lobbying objectives, that followsfrom the presence of the FTA, and I show empirically that thisdifferentiation does matter for Canadian trade policy. I quantifydomestic and foreign post-NAFTA lobbying activity in Canada andshow that US firms successfully lobby the Canadian government forincreases in trade barriers, while foreign firms from outside of NAFTAeffectively lobby in Canada for trade barrier reductions. This impliesthat active lobbying by an FTA partner country is an additionalpolitical economy factor that was not considered in previousresearch and that may lead to more protectionist trade policies ofregional trading blocks and impede multilateral trade liberalization.Moreover, in the environment where almost every country in theworld is a member of at least one regional trade agreement, for acomplete understanding of a country's trade policy we should notonly consider the effect of foreign lobbying but also take into accountits differentiated effect. The empirical results also reveal that the GHmodel with monopolistic competition has more explanatory powerthan its original version.


References

Bagwell, K., Staiger, R.W., 1997. Multilateral tariff cooperation during the formation offree trade areas. International Economic Review 38 (2), 291–319.

Bekker, P.A., 1994. Alternative approximations to the distributions of instrumentalvariable estimators. Econometrica 62 (3), 657–681.

Bohara, A.K., Gawande, K., Sanguinetti, P., 2004. Trade diversion and declining tariffs:evidence from Mercosur. Journal of International Economics 64 (1), 65–88.

Bombardini, M. (2008): “Firm Heterogeneity and Lobby Participation,” Forthcoming inthe Journal of International Economics.

Bond, E.W., Riezman, R.G., Syropoulos, C., 2004. A strategic and welfare theoreticanalysis of free trade areas. Journal of International Economics 64 (1), 1–27.

Broda, C., Weinstein, D.E., 2006. Globalization and the gains from variety. QuarterlyJournal of Economics 121 (2), 541–585.

Chang, P.-L., 2005. Protection for sale under monopolistic competition. Journal ofInternational Economics 66 (2), 509–526.

Davidson, R., MacKinnon, J., 1993. Estimation and Inference in Eeconometrics. OxfordUniversity Press, New York.

Facchini, G., Biesebroeck, J.V., Willmann, G., 2006. Protection for sale with imperfectrent capturing. Canadian Journal of Economics 39 (3), 845–873.

Feenstra, R.C., 1994. New product varieties and themeasurement of international prices.American Economic Review 84 (1), 157–177.

Gawande, K., Bandyopadhyay, U., 2000. Is protection for sale? Evidence on theGrossman–Helpman theory of endogenous protection. Review of Economics andStatistics 82 (1), 139–152.

Gawande, K., Krishna, P., Robbins, M.J., 2006. Foreign lobbies and US trade policy.Review of Economics and Statistics 88 (3), 563–571.

Goldberg, P.K., Maggi, G., 1999. Protection for sale: an empirical investigation. AmericanEconomic Review 89 (5), 1135–1155.

Grossman, G.M., Helpman, E., 1994. Protection for sale. American Economic Review 84(4), 833–850.

Hansen, W., Mitchell, N., 2000. Disaggregating and explaining corporate politicalactivity: domestic and foreign corporations in national politics. American PoliticalScience Review 94 (4), 891–903.

Hansen, C., Hausman, J., Newey,W., 2006. Estimationwithmany instrumental variables.Discussion paper, working paper.

Kelejian, H., 1971. Two-stage least squares and econometric systems linear inparameters but nonlinear in the endogenous variables. Journal of the AmericanStatistical Association 66 (334), 373–374.

KPMG Consulting Inc., 2001. Study on Compliance Under the Lobbyists Registration Act.Prepared for Office of the Ethics Counsellor. Ottawa. September 14, 2001.

Mitchell, N., 1995. The global polity: foreign firms' political activity in the United States.Polity 27 (3), 447–463.

Ornelas, E., 2005a. Endogenous free trade agreements and the multilateral tradingsystem. Journal of International Economics 67 (2), 471–497.

Ornelas, E., 2005b. Rent destruction and the political viability of free trade agreements.Quarterly Journal of Economics 120 (4), 1475–1506.

Richardson, M., 1993. Endogenous protection and trade diversion. Journal of Interna-tional Economics 34 (3–4), 309–324.

Trefler, D., 1993. Trade liberalization and the theory of endogenous protection: aneconometric study of U.S. import policy. Journal of Political Economy 101 (1),138–160.

trade policy of a free trade agreement in the presence of foreign lobbying

Documents