Statistical analysis of trade policy-making with formal models

The general idea of statistical analysis of trade policy-making is to explain trade policy patterns or voting behavior on trade policy in parliament by measurable variables that can be linked to specific constituents. This method has found its major application in the analysis of votes on trade-related bills in the US Congress, most notably on the transfer of fast track authority for international trade negotiations to the President and on the ratification of international trade agreements.

In these models, the general public always favors free trade based on their consumer interests. Their interests as recipients of factor rewards depend on the mobility of factors across sectors. If they can be transferred from one sector to another without cost, the interests of an individual correspond to its relative endowment with capital and labor in comparison with the national relative factor endowment. Countries that are relatively rich in labor compared to capital, tend to import goods and services whose production is relatively capital-intensive. Thus, labor becomes less abundant compared to capital and marginal labor productivity will increase. Since marginal productivity tends to influence factor rewards, wages stand to rise. Accordingly, an individual in a labor-rich country will favor free trade if it is relatively well endowed with labor, while it will prefer protectionism if it is relatively well endowed with capital.Footnote ³ If factors are sector-specific, and shifting them to another sector reduces their returns, an individual with relatively high stakes in a sector strongly competing with imports prefers protectionism because this increases sector-specific factor rewards.

The effort of protectionist or liberal special interest groups to influence trade policies is commonly equated with their financial contributions to the election campaign of their parliamentary representative. Contributions can shape trade policy in two ways. First, they may enhance the chances of politicians or parties that endorse those trade policies that the special interest group prefers.Footnote ⁴ The underlying idea is that special interest groups finance campaigning expenses that influence voters. Alternatively, contributions may directly ‘buy’ sympathetic policies.Footnote ⁵ This implies that either legislators who have received contributions feel obliged to reciprocate the favors, or that legislators pass trade policies in anticipation of subsequent rewards. Empirical evidence shows that special interest groups give to like-minded candidates in order to improve their election chances, as well as to those legislators whose party tenure and committee responsibilities suggest they can directly deliver desired policies.Footnote ⁶

The independent causal role of a parliament follows from the ideological rating of its members. Special interest groups attempting to guide the electorate typically publish this rating. This ideological variable shows when parliamentarians vote consistently with their worldview and convictions – derived in particular from their previous voting patterns – rather than according to the interests and campaign contributions of their district's constituents.

The main contribution of this statistical approach based on simplified formal models has been to show that campaign contributions work well in influencing policies (at least in the case of the US Congress),Footnote ⁷ and to examine the circumstances determining the efforts and chances of special interest groups to attain or reverse protection.Footnote ⁸ Regardless, a number of shortcomings concern the attribution of interests to domestic constituents, their channels for influencing politics, and the spectrum of domestic constituents considered.

A first problem is that individuals' trade policy preferences are difficult to establish as they are shaped by a wide range of factors. In the statistical models, interests are framed as effects of trade policy on individual income. In reality, the perception of economic effects of trade policy is not determined exclusively by objectively discernible economic facts, but is dependent on subjective beliefs and education levels.Footnote ⁹ Furthermore, trade policy preferences also depend on altruism, social status, and ideology.Footnote ¹⁰ Poverty alleviation through trade inclines individuals toward free trade. Based on actual income or the individual's subjective evaluation, an advantageous position in domestic society is also associated with pro-free-trade preferences, whereas community attachment and nationalism dispose people towards protectionism.

A second fundamental limitation to the statistical approach that results from its rationalist commitment is its focus on voting and campaign contributions as channels for influencing trade policy-makers. Unfortunately, this ignores the information needs of the legislative and executive branches regarding the evaluation of likely material and political effects of international agreements. Private actors can help them to recognize how domestic regulation is working and how domestic regulatory changes implemented in accordance with WTO agreements would affect the home market. They also provide information about which foreign regulation is impeding trade, which changes would be recommendable, and how much trade might expand in response. Hence, relevant expert knowledge gives access to policy-makers.Footnote ¹¹

Furthermore, the rationalist focus on a logic of consequences excludes the logic of appropriateness.Footnote ¹² In a logic of consequences, actors select their means in order to arrive at certain ends. Actors who follow a logic of appropriateness, choose a certain behavior because they consider this behavior in itself to be right. They follow norms that prescribe appropriate behavior depending on the situation and the identity of the actor. Actors do not exclusively choose between the logic of consequences and the logic of appropriateness, but interests and norms compete in the final balance. The two logics are furthermore intertwined at an unconscious level as they influence perception. Actors tend to interpret their normative obligations in a way that fits their interests.Footnote ¹³ In the context of the WTO, the logic of appropriateness has two, potentially conflicting, implications. On the one hand, governments and delegates feel obliged to further national welfare. This moderates the influence of elections and contributions where they run afoul to the governments' perception of the public interest. Conversely, bureaucrats and delegates tend to believe that it is their duty to serve their country's special interest groups with whom they deal (or have dealt during their work in the capital) on a daily basis. Thus, they consider not only citizens' political representatives, but also special interest groups as legitimate stakeholders in trade policy formation.

Finally, the formal models exclude a number of domestic constituents active in trade policy-making, such as ministries, regulatory agencies, sub-federal units, and civil society organizations with non-economic objectives. Yet, these organizations do not simply reflect the interests of voters and economically oriented special interest groups in a symmetric way. Instead, they amplify the impact of certain groups while keeping independent organizational interests. A similar shortcoming is the way in which formal models simplify the functioning of parliaments.Footnote ¹⁴

For these reasons, statistical analysis drawing on formal models necessitating strong simplifying assumptions alone does not lead to a satisfactory assessment of the respective influence of domestic constituents. For instance, a non-governmental organization supporting environmental protection would, according to the standard assumptions of these models, have no constituency among the self-interested population. Politicians would be deaf to the information it provides and the legitimacy claims it makes, and consequently its existence could be safely disregarded. Since this contradicts common-sense observation of trade policy-making, existing statistical analysis alone is insufficient.

Verbal analysis of trade policy-making in a WTO context

An alternative to the formal study of trade policy-making is the verbal analyses contained in case studies, as well as in expert observations on general tendencies in trade policy-making practice. A comprehensive contribution to this approach is the collection of case studies presented by the WTO (2006).Footnote ¹⁵ Several circumstances and developments that many developing countries have in common can be inferred from this collection. The broader WTO agenda has been (1) reflected in a greater involvement of ministries and regulatory agencies beyond the trade, economic, finance, and foreign policy ministries that have traditionally reigned over these matters.Footnote ¹⁶ Special interest groups and civil society organizations show (2) a growing interest in participating in the formation of a national negotiating position. At the same time, governments have become more accommodating towards private actors because they believe in the legitimacy and effectiveness of an inclusive trade policy-making approach.

These two developments have led to the establishment of national councils (3) in which various ministries, regulatory agencies, special interest groups, civil society organizations, and academic institutions discuss trade policy.Footnote ¹⁷ What these councils have in common is that they do not produce policies that are governmentally binding, but they create expectations that their recommendations should be respected.Footnote ¹⁸ This structured coordination gives private actors a stronger incentive to organize and to develop a negotiating position, as well as a considerably improved capacity to make their opinion heard, especially when compared to the ad-hoc exchanges with governments as practiced in the past. Additionally, private actors often accompany or form part of national delegations at Ministerial conferences.Footnote ¹⁹

Furthermore (4), governments, private actors, and even academic institutions are frequently overtaxed by the complexity of WTO issues.Footnote ²⁰ This lack in governmental capacity opens the possibility to private actors to influence their country's negotiating position provided that they manage to develop expertise.Footnote ²¹ Public awareness of trade issues varies (5) across countries, issues, and time. Whereas discontent with liberal trade policies played a central role in some elections,Footnote ²² it remained peripheral in others.

The tenet of these case studies is mirrored in other contributions to the literature, examining trade policy-making both in industrialized and developing countries. Common themes in a collection of case studies edited by Jackson and Sykes (Reference Jackson and Sykes1997) are the substantial and generally increasing influence of various ministries and regulatory agencies, of special interest groups, and of the sub-federal level particularly in the US and in Canada. The case studies compiled by Macrory, Appleton, and Plummer (Reference Macrory, Appleton and Plummer2005) show a similar picture of greater heterogeneity of domestic constituents that participate in trade policy-making and of advancing institutionalization of state–society relations.

Esty (Reference Esty2002) and Keohane and Nye (Reference Keohane, Nye, Porter, Sauvé, Subramanian and Zampetti2001) see the ‘club model’ of multilateral cooperation coming to an end. This model isolated the trade ministry from involvement by other ministries and the general public, thus immunizing agreements from disaggregating. Stein (Reference Stein2001) notes that following the Uruguay Round, state actors in several countries re-claimed their constitutional right to have a say on the negotiation of WTO treaties. In the EU, this concerned the relationship between the Commission and the member states, in Canada between the federal and the provincial level, in Japan between the executive and the parliament. Page (Reference Page2001) and Raghavan (Reference Raghavan2000) observe that domestic constituents have become more involved in the establishment of negotiating positions in developing countries after the Uruguay Round. Charnovitz (Reference Charnovitz2000), Deslauriers and Kotschwar (Reference Deslauriers, Kotschwar, Doh and Teegen2003) and Florini (Reference Florini and Siebert2003) point to the mounting activity of civil society organizations on WTO issues.

In conclusion, it is possible to derive certain power relations between domestic constituents and trends therein from the verbal analysis of trade policy-making in the WTO context. Within one piece of work, however, it is already difficult to infer precise information about the relative influence of various constituents because each actor's influence is loosely qualified but not ranked. By the same token, changes in their relative influence are noted, but the net effect of several developments is frequently not established or only vaguely characterized. In addition, the comparison across works from this verbal strand of analysis is notoriously problematic.

Level of Influence

Table 1 shows a ranking of the average influence of domestic constituents for industrialized and developing countries. Respondents could choose between five answers (very low, low, medium, high, very high). Averages were constructed assigning weights to the responses, ranging from 1 (very low) to 5 (very high). Where a country returned two or three questionnaires, a country-specific average was formed first, thus ensuring that all participating countries have the same weight regardless of the number of questionnaires they sent back. Averages are presented on the second-digit level in order to keep the figures from distortion through rounding. It is necessary to emphasize that the subjective evaluation of delegates and the small sample size allow only to discern approximate patterns.

Table 1.

Ranking of domestic constituents by their influence

Several aspects of these results shown in Table 1 are noteworthy:

1. 1. Overall, the difference in the political landscape between industrialized and developing countries is minor. There is no divide between industrialized countries – characterized by a complex inter-institutional process with strong participation of economic and non-economic societal actors and advanced public accountability – and developing countries – where one or few ministries could determine trade policy to their liking. One unsurprising difference is that national parliaments are more influential in industrialized than in developing countries (3.63 vs. 2.96). Interestingly, additional ministries and regulatory agencies wield greater power over negotiating positions in developing than in industrialized countries (3.57 vs. 2.93).

2. 2. The influence of additional ministries and regulatory agencies, as well as of parliaments, is remarkably high (2.93 and 3.63 in industrialized countries, 3.57 and 2.96 in developing countries).Footnote ²³ This being the case, trade-policies cannot be explained solely by regarding national interests derived from material economic circumstances, or from the interests and power of sector-based special interest groups or owners of factors of production, as some formal models imply. Executive and legislative institutions matter, whether as spaces where particularistic interests are aggregated or where national interests are defined.

3. 3. Sub-federal units remain among the least important actors (2.50 in industrialized countries, 1.90 in developing countries). While they still rank above the general public in industrialized countries, they are even less influential in developing countries.

4. 4. Special interest groups from import-competing and export-oriented sectors are the most powerful actors (3.87 and 3.67 in industrialized countries, 3.62 and 3.45 in developing countries).Footnote ²⁴ The emphasis put on them in the formal models of trade policy-making is thus warranted in principle. The small difference to other constituents suggests, however, that this focus has been exaggerated. Furthermore, one can see that import-competing and export-oriented interest groups wield approximately equal influence, both in industrialized and developing countries. This contradicts the strongly held traditional notion that protectionist interests groups dominate in domestic politics, which used to justify the need to insulate the GATT regime from domestic influences.

5. 5. On average, civil society organizations are considered to possess moderate power (3.20 in industrialized countries, 2.76 in developing countries). Interviews with the delegates and pertinent literature show that the influence of civil society organizations is highly asymmetric across issues and time. For example, while the lobby from consumer organizations to lower tariffs on goods is weak, public campaigns can exert considerable pressure to reduce subsidies that harm poor farmers abroad or contribute to environmental damage.

6. 6. The general public has only a minor role to play regarding power influence. The small difference between election and non-election years is surprising (2.40 vs. 2.27 in industrialized countries, 2.10 vs. 1.90 in developing countries).Footnote ²⁵ In the interviews, delegates expressed the view that the general public matters significantly only when it is incited and directed by special interest groups or civil society organizations. But governments seem not to actively analyze public opinion in order to anticipate how the general public will receive possible agreements.

A further comparison can be established between countries with different political regimes. Table 2 shows the regime type of developing countries participating in the survey which are accounted for in the Polity IV database for the year 2004 (excepting only Antigua, Barbuda and Hong Kong). The regime type value is constructed by subtracting a country's value on an autocracy index (Auto) from its value on a democracy index (Demo).

Table 2.

Regime type of developing countries participating in the survey

A regression analysis shows no significant correlation between the participation of domestic constituents and regime type. The reason for this astonishing result is that in countries with ‘decent’ political conditions (that is, regime type values of 3 or higher), which make up for the majority of responses, no such correlation exists. When those five countries with the lowest regime type values (−10 to −2) are juxtaposed with the remaining set, however, the disparity is striking. Table 3 presents the average values of domestic constituent participation for the 18 ‘democratic’ and the 5 ‘autocratic’ developing countries. It can be seen that only additional ministries and regulatory agencies wield about equal influence in ‘autocratic’ as in ‘democratic’ countries. Pressure groups with economic interests still fare relatively well even in ‘autocratic’ countries, whereas the national parliament, sub-federal units, civil society organizations, and the general public have only low to very low influence.

Table 3.

Influence of domestic constituents according to countries' regime type

Changes in influence

The second important aspect of the questionnaire dealt with the development of the domestic policy-making process. Answers could again be marked on a scale ranging from 1 to 5, with 1 standing for strongly decreased, 3 for unchanged, and 5 for strongly increased. Table 4 displays a ranking of increases in the influence of domestic constituents for industrialized and developing countries. These values result when the value of 3, or ‘unchanged’, is subtracted from the average response values. A value of 0 thus signifies that there has been no change in relative influence of this constituent.

Table 4.

Ranking of domestic constituents by increases in their influence

It should be noted that the responses are likely to substantially underestimate real changes since the experience of many delegates is limited to a few years; most delegates who participated in the Uruguay Round are no longer at their missions and would not be filling out this questionnaire.

Upon analysis of these tables, the following points can be elucidated:

1. 1. With regards to developing countries, an overall tendency of increasing domestic politicization of trade policy-making is confirmed by the responses. This comes at the cost of those ministries traditionally in charge of trade policy-making in capitals, while also decreasing the discretion of delegates in Geneva. In industrialized countries, patterns of influence among stakeholders have also changed. Those stakeholders that have strengthened their position even in industrialized countries are the national parliament, civil society organizations, and additional ministries and regulatory agencies. The fact that these are all actors with strong non-economic concerns is of interest.

2. 2. The tendency noted during the Uruguay Round that sub-federal units are becoming more involved in the development of WTO negotiating positions has not lasted. A large majority of delegates perceive their influence as unchanged.

3. 3. Import-competing and export-oriented interest groups have grown more influential in developing countries. In industrialized countries, by contrast, the sway of import-competing interests has remained stable, while the power of export-oriented sectors and companies has even diminished. This is in line with recurring complaints that multinational companies do not back the current Doha Round with the same strength evidenced during past multilateral trade rounds.

4. 4. Interestingly, the influence of the general public, whether in election or non-election years, has stayed largely unchanged. The increasing domestic politicization of WTO issues appears thus to stem less from a greater interest in trade issues shown by society in general, but rather from a more efficient organization of societal non-economic interests (as reflected in the relatively strong gains in influence by civil society organizations).

More protectionist policy preferences

In the interviews, delegates expressed the view that the greater involvement of additional ministries, such as those for agriculture, fisheries, education, or culture, made their negotiating positions on average more protectionist. One possible reason for this is that additional ministries are more concerned about the non-economic regulatory objectives that might be endangered by liberalization. They tend to consider these objectives as more important and more threatened by liberalization than economic ministries. The latter emphasize the central role of economic growth and are more likely to believe in the market forces as the most effective way to attaining this growth. This negative attitude towards the WTO appears to be connected to a lesser degree of knowledge about likely effects. Exposed to substantial uncertainty about the legal functioning of the WTO and the likely effects of liberalization, additional ministries that are only recently or occasionally concerned with the WTO tend to believe in pessimistic scenarios.

A second reason for the protectionist inclination of additional ministries is that they perceive a stake in the existing policies that are being questioned by multilateral liberalization as they have themselves developed and implemented them. At issue is thus not only a rational evaluation of how to weigh competing objectives and to select optimal policies to attain them, but also bureaucratic resistance to changes in established regulatory approaches. This is especially pertinent if these changes are induced from the outside.

Connected to this is the fact that additional ministries have a different relationship to the WTO. As they are less acquainted with the reciprocal nature of the WTO, they do not accept the need to make concessions on what they consider optimal policies. Being far removed from the actual negotiations, they are less aware of the problems at the international level that an overly protectionist position implies for their state. They do not directly witness the pressure brought to bear on their country and they are less upset by the idea of tarnishing their country's reputation as a cooperative member of the international community.

This being the case, these ministries are generally classified as protectionist forces within their countries' trade policy-making process. They are seen as committed to regulatory objectives and approaches that are potentially in conflict with WTO rules, while remaining less informed or concerned about their overall beneficial effects. They are often confronted with manifold warnings and complaints about the WTO from their stakeholders, and not acquainted with the big picture of concession trading between member states.Footnote ²⁷

The second point on which delegates agreed was that the greater average participation of parliaments tended to decrease their countries' readiness to open up their markets.Footnote ²⁸ One reason for this is that parliamentarians have narrower constituencies, and thus more specific interests, than the executive bodies. They enlist support for their specific protectionist needs by endorsing the claims for protection by other parliamentarians. In addition to this log-rolling procedure, parliamentarians usually lack a thorough understanding of the WTO. As the selective signals they receive from their constituents are biased towards a negative perception of the WTO, they tend to be weakly informed, but skeptical.Footnote ²⁹

Delegates perceived civil society as a strongly protectionist force. Most non-governmental organizations attempt to preserve a non-economic good, such as a healthy environment, safe food, and public education, against a perceived infringement driven by national and multinational economic interests. Consumer interest groups that traditionally appreciate free trade for its price-decreasing and variety-enhancing effect, are relatively weak and have been partly aligned to the WTO-skeptical mainstream, particularly regarding instances of health concerns arising from genetically modified organisms.

Furthermore, in developing countries, civil society organizations assume the defense of certain import-competing interests. In doing this, they do not only engage in surprising alliances with ailing industries, but they genuinely represent economic and non-economic interests. For example, rural development organizations equally struggle for the incomes of poor farmers, basic and agriculture-specific public services, and environmentally friendly local production and consumption patterns. This link to small producers with minor immediate export interests, fear of foreign competition, and an aversion against domestic regulatory liberalization that might accompany the opening to international trade makes civil society organizations all the more protectionist.

A further difference to the classic case is that international civil society organizations engaged in a worldwide struggle against ‘neo-liberalism’ are often preponderant in developing countries. They have the necessary expertise, the international connections, and the financial resources to shape the opinion amongst national civil society organizations on the WTO and to multiply civil society activity that is in line with their beliefs.

While most civil society organizations that are active on trade issues in industrialized countries are also positioned against the WTO, an increasing share subscribes to the ‘fair trade’ concept espoused in particular by organizations like OXFAM. Rather than unconditionally fighting against the WTO as an instrument of ruthless capitalism and imperialism, they exert political pressure on industrialized countries' governments to reduce tariffs and subsidies judged as harmful to the global poor or the environment.

Stronger discouraging effects of risky WTO agreements

Common sense and scientific evidence converge in the observation that governments, and human beings more generally, are risk-averse. This being the case, governments would prefer an international agreement with a guaranteed gain rather than agreements with several possible results if both offer the same average gain. More generally, the value of an agreement for governments decreases as the standard variation of expected outcomes increases. The important point is that a discouraging effect of risk can exist even if WTO agreements do not decrease governments' utility below the status quo without an agreement. Even if all expected outcomes involve net gains, uncertainty about the level of gains reduces the value of an agreement. Assuming that negotiations are efficient, governments expand their cooperation until the marginal benefits of further cooperation equal its marginal costs. By reducing the value of possible agreements, the risk involved in WTO agreements thus exerts a discouraging effect on the progress of multilateral trade liberalization. In other words, the greater the risks involved, the earlier the marginal gains from additional cooperation become zero.

The discouraging effect of risk is relevant in the context of domestic trade politics because governments find it difficult to assess how strongly domestic constituents will react to agreements and how much pressure they will be able to exert on the government if they are dissatisfied. Many domestic constituents are slow to inform themselves about possible WTO agreements, to adopt a position, and to mobilize their forces – with the intention of shaping the negotiations, of hindering implementation, or of ‘punishing’ their government with an eye on preventing further agreements of the same sort. Anticipating future political consequences is additionally complicated by domestic constituents' incentive to misrepresent their resolve and strength in order to impress governments. This means that governments are less willing to make concessions in WTO negotiations as the influence of domestic constituents increases because they fear internal disputes and public protest at the negotiating and implementation stage.

Reduced negotiating flexibility

If negotiations are (Pareto-) inefficient, gains from multilateral liberalization, which would be attainable despite governments' politically biased negotiating positions, remain unexploited. Thus, there would be an alternative, more liberal agreement that would improve the utility of at least one government without reducing the utility of another government. One reason for such inefficiency is a lack of flexibility in negotiating positions. This problem is discussed in the present section. States' excessive claiming on the gains from cooperation, another cause of inefficiency, will be treated subsequently.

Flexibility can be understood as the speed and intensity with which states react to market access offers and requests, as well as to suggestions for rules, submitted by other states in WTO negotiations. Member states' flexibility significantly influences the success of WTO negotiations. Since states benefit asymmetrically from any concession, it is not sufficient that a large number of states be willing to make concessions of substantial total value provided that others do the same. It is also necessary that these concessions benefit those states that are prepared to reciprocate. In other words, the exchange of concessions is an intricate matching process that requires a constant assessment of trade-offs across sectors and issue areas (e.g. market access in textiles against concessions on intellectual property rights). This is complicated by the fact that states are uncertain about the preferences of other states. The evaluation of how much a concession costs one government and benefits others depends on complex repercussions on economic and non-economic values, as well as on domestic political considerations. Consequently, states need to learn about each other's preferences during negotiations and to adapt their offers and demands. Moreover, windows of opportunity may open during one major meeting and close immediately afterwards. Ministerial meetings, the most important events in WTO negotiations, are held every two years. In the meantime, circumstances may change and thwart progress towards an agreement that would have been within reach before.Footnote ³⁰ Governments may change, pre-election periods may immobilize main players, economic conditions and public opinion may turn, and protectionist interest groups may strengthen their positions. Therefore, member states need both the flexibility to recalibrate their negotiating positions between the rare Ministerial meetings and to achieve this within the days or even hours of critical negotiations.

Delegates confirmed that the flexibility of negotiating positions has been diminishing as more domestic actors have acquired a greater influence on trade policy-making.Footnote ³¹ First and foremost, consulting constituents with a recognized say on trade policy-making and achieving sufficient support for new negotiating positions takes time. Second, the trend towards institutionalization of domestic trade policy-making processes additionally recognizes and legitimizes the claims of private actors. Domestic constituents come to expect that compromises developed in these participatory policy-making systems weigh heavily on the country's negotiating position. This perception restrains governments that want to avoid political conflict at home, binding them to their original negotiating positions if no alternative consensus can be established. Lastly, governments need additional time to learn about the likely domestic political consequences of possible WTO agreements. The negotiating period thus serves as a test phase to gauge the clout of special interest groups and civil society organizations and the slow opinion building within the parliament and general public. The more domestic constituents care about the actions their government takes in the WTO, the longer governments will wait before making a move in order to avoid miscalculations.

Tougher fighting for the gallery

The bargaining over how the gains from cooperation in the WTO should be shared between governments constitutes a major obstacle to multilateral liberalization.Footnote ³² Since each government is uncertain about how other governments evaluate possible WTO agreements and their best alternatives to a negotiated agreement, it has an incentive to engage in aggressive gain-claiming, hoping to shape the resulting agreement in its own favor. By contrast, governments are enticed to moderate their initial negotiating positions and soften their gain-claiming over time. Doing so avoids delays and reductions in the ambition of the negotiations, prevents deadlocks that might damage the WTO in the long-term, and helps to avoid peer pressure.

Greater domestic politicization of trade policy-making shifts the balance of these competing factors in favor of tougher gain-claiming strategies. Economic interest groups have become more active in many member states since the Uruguay Round, and the greater transparency of WTO negotiations facilitates the monitoring of how governments defend particular interests. Therefore, governments face greater incentives to be tough in their demands to attain market access abroad and in their refusal to grant access to their domestic markets. A similar case can be made regarding non-economic values. Here, the increasing influence of civil society organizations frightens governments. This keeps them from making concessions without being able to argue that they have been struggling as hard as they reasonably could to further non-economic values. In both cases, governments benefit not only from what they attain, but also from what they are seen to be fighting for.

This effect is mitigated by the fact that other domestic constituents desire their government to be more willing to compromise. This is predominantly a consequence of the fact that making compromises facilitates the conclusion of an agreement and allows a state to claim concessions from other countries in return. Nevertheless, delegates in the interviews felt that greater domestic constituent participation toughens gain-claiming in the final balance. An explanation for this is that export-oriented interests use their influence first in compelling their governments to insist on concessions from other countries. The benefits of such lobbying are immediately visible. The benefits of convincing their governments to improve market access for foreign producers with the expectation that this might favor their own export interests are more uncertain. Own market opening may bring few foreign concessions. And even if market access abroad may be improved in exchange, it is not clear which export-oriented industries will be concerned. The export lobby thus suffers from a free-rider problem when attempting to promote own market opening. In contrast, import-competing pressure groups can focus their clout on preventing specific liberalization commitments that would harm their interests. Furthermore, the additional ministries and parliaments acquiring a greater say in trade politics tend also to be skeptical of own market opening but do little to restrain demands for market opening abroad.Footnote ³³ This explains why overall domestic constituent involvement is associated with tougher gain-claiming.