Week 10 Summaries

Kenneth A. Oye, ed., Cooperation Under Anarchy, Princeton University Press, 1986.

Kenneth A. Oye, "Explaining Cooperation Under Anarchy: Hypotheses and Strategies"

Oye’s theme is why, given the anarchic condition between states, cooperation emerges in some cases and not others.  He asks two main questions: “What circumstances favor the emergence of cooperation under anarchy?” and “what strategies can states adopt to foster the emergence of cooperation by altering the circumstances they confront?” (1-2).  His answers are an early attempt to link insights from game theory and macroeconomics, security studies and international political economy.  He discusses three “circumstantial dimensions” that “serve both as proximate explanations of cooperation and as targets of longer-term strategies to promote cooperation”: payoff structure, shadow of the future (Axelrod’s term), and number of players.

First, Oye discusses how payoff structure affect prospects for cooperation (“conscious policy coordination”) and how strategies to alter it can improve these prospects.  He stresses that this analysis applies only to the class of games where cooperation is necessary to realize mutual interests, and suggests that analyses of cooperation in IR should only use stag hung, chicken, and PD as games of last resort after harmony and deadlock are ruled out.  Oye claims that the determinants of payoff structure are often subject to “willful modification through unilateral, bilateral, and multilateral strategies.”  For example, a state can use troops as hostages to diminish payoff from own defection or choose a procurement strategy preferring defensive to offensive weapons.  In longer-term situations it can seek to alter other’s perceptions of interest (e.g. US telling Soviets about MAD during SALT I; more generally the socialization or internalization of new norms by states).  (10-11)

Second, Oye describes how strategies to increase the ‘shadow of the future’, as well as strategies to increase the capacity for recognition (distinguishing between cooperation and defection) and control (flexibility and ability to shift policy in a tit-for-tat strategy), can increase the prospect of continuing interaction and the likelihood of cooperation in iterated games.  Reciprocity can be encouraged through enhancement of recognition capacities: the explicit codification and clarification of norms, as well as surveillance and verification mechanisms, serve to reduce ambiguity and increase transparency.  The shadow of the future can be lengthened and single-play situations turned into iterated oes by decomposing interactions over time (e.g. breaking down a huge trade deal into a long-term series of deliveries and payments) and by linking issues (making cooperation on one issue contingent on cooperation in a separate, future issue).

Finally, Oye discusses how cooperation becomes more difficult in N-person games, thanks to increasing complexity, rising transaction and information costs, the failure of conditional cooperation and the possibility of free-riding, etc.  Some of these problems can be alleviated through conventions (to diminish transaction and information costs), regime creation, collective enforcement mechanisms, and so on.  Alternatively, another way to promote cooperation is to reduce the number of actors necessary to realize common interests.  Bilateral and regional decomposition through discriminatory strategies can do this, albeit at the cost of lowered gains from cooperation and the risk of causing negative externalities.  

George W. Downs, David M. Rocke, and Randolph M. Siverson, "Arms Races and Cooperation."  

Downs, Rocke, and Siverson (DRS) discuss the conditions that determine when unilateral action, tacit bargaining, and negotiation are likely to reduce intensity of arms races.  They approach this question through a formal analysis of arms races, informed by a historical analysis of nineteenth and twentieth-century arms races that ended up peacefully.  They conclude that most of the arms races that ended in cooperation “have done so not because one side adopted a particular cooperative strategy, but because the basic character of the race was altered by events that were not directly connected with it.  Such events, which collectively might be called the ‘tilt of the board,’ spring largely from economic circumstances and changes in the larger political order…” (119).  

DRS define arms race as continual mutual defection (DD).  In this setting, assuming perfect information and rational actors, it must be true that DD > CD and not both (CC > DC and CC > DD).  These condition produce nine orderings, listed on p. 121.  Of these, four are related to Deadlock (i.e., one’s two most preferred outcomes require one to defect).  DRS claim that these preferences and situations are not as pathological as others have assumed, and can explain the Anglo-German naval race before WW1, and the preferences of the Germans in the 30s and Soviets in the 50s.  In such conditions, unilateral strategies (eg defensive rather than offensive weapons) may seem irrelevant, but DRS claim that states can “nonetheless chose [them] from other motives with similar effect.” (125).  They concede is thoroughly irrelevant, but conclude negotiation might work even if both prefer defection, if issue-linking can alter the payoff structure.  In a PD as opposed to Deadlock: unilateral strategies must deal with whether the assumption that security rather than influence is the adversary’s primary motive, and whether defensive weapons are a viable option.  Tacit bargaining can work in the form of Axelrod’s tit-for-tat strategy, assuming the prospect of iteration.  Once again, negotiation works through issue-linkage, and DRS point to 3 historical cases where this de-escalated arms races: the Washington Naval Treaty, SALT I, and SALT II.

In the second half of their article DRS discuss information and control problems and how they “expand the range of games that can lead to arms races” (133).  Strategic misrepresentation is when one side tries to extract concessions in a bargaining situation by pretending to have preferences it does not hold.  However such a bluff can be called, and in order to maintain credibility, the preference must be followed (e.g. a weapons system must actually be built, such as MX missiles or ABMs).  Imperfect intelligence (e.g. British overestimates of French shipbuilding in the 1850s) or worst-case analysis can create the impression of defection to the opponent (CD) even though it hasn’t happened.  This yields further games that could lead to arms races, including the stag hunt (135).  Problems of interpretation (the result of experience or ideology) can have the same effect as poor information.  Control problems can cause arms races to continue if decision makers cannot cooperate (or more subtly, signal their desire to cooperate but cannot actually do so) or defect to punish defection, because their governments resist change.  DRS claim that the ability of bureaucracies to kill policy changes through leaks to the press explains why leaders use summits and personal envoys to move through important stages of negotiations.  Unilateral strategies are vulnerable to pressures of uncertainty, especially given the subjective variance among interpretations of what constitutes a defensive rather than offensive weapons system (e.g. fortifications in France and Germany, aircraft carriers in the pacific, alliances).  Finally, information and control problems diminish the effectiveness of Tit-for-Tat, though it remains superior to other strategies, especially when modified so as to retaliate less frequently in order to account for possible misinformation and lack of control.

John Conybeare, "Trade Wars: A Comparative Study of the Anglo-Hanse, Frnaco-Italian, and Hawley-Smoots Conflicts."

Conybeare compares three cases of trade wars (Anglo-Hanse trade wars of C14-17, Franco-Italian tariff wars in the 1880-90s, and the tariff wars that followed from the Hawley-Smoot act of 1929), which he defines as attempts to achieve “economic objectives directly related to the traded-goods sector of the economy” by means of “restrictions on the flow of goods and services” (147).  Conybeare analyzes what promotes or inhibits cooperation in these three wars along the three dimensions discussed by Oye: payoff structure, iteration, and number of actors.  

Payoff structure in trade wars is a function of four factors: (1) “relative size, trade dependence, or stages in a business cycle”; (2) “the dynamics of bargaining, including transaction costs, side payments, linkage to other issues, coalitions, and diplomatic skill”; (3) domestic structure, especially the role of rent-seeking interest groups; and (4) cognitive factors (ideology, misperception, risk, and time preferences) (149-50).  
In his study of the Franco-Italian and other late C19 trade wars, Conybeare examines how asymmetry in payoff structure can affect cooperation.  When Italy, a small developing economy, initiated a trade war with France, a large and major trading partner, it didn’t calculate the asymmetry in the costs for mutual defection.  Moreover the French could continue to raise the cost of defection for Italy, possibly to the extent that it would turn the game into Chicken for Italy while remaining in a PD itself (Conybeare calls this “Called Bluff”).  Eventually the costs of DD for Italy were so high that it accepted major concessions to return to a higher level of cooperation – which was still less than the original state of CC.

Iteration obviously is central to trade wars; the question is why an extended ‘shadow of the future’ does not produce or sustain cooperation.  Conybeare points to escalating patterns of uncooperative interaction developed over time, changing perceptions over time, and the isolation of trade issues by institutional structures  (e.g. GATT, which was created to arbitrate trade disputes, ends up constraining the ability to retaliate in the future).  The related case is the Anglo-Hanse wars, which Conybeare describes as a prisoner’s dilemma.  He explains that the two sides failed to sustain cooperation over several centuries of iterated conflict thanks to very high transaction costs, the influence of rent seeking interest groups on negotiations, and a sustained economic recession in England which raised the current marginal value of defection in the form of predatory income transfers, and lowered the shadow of the future.

Finally the free-rider problem enters trade wars in the form of the MFN treaty clause, which creates a public good by specifying that states cannot grant a concession to one state without granting it to all states.  Conybeare takes this large-number public-good problem to be the crux of the Hawley-Smoot wars.  While the public-good problem had existed earlier, after World War 1 there were many more countries involved, as well as further complications like war industries and reparations.  The Hawley-Smoot act raised American tariffs while the US still sought MFN privileges from trading partners.  As a result others retaliated by attempting to remove the public good of MFN privileges via country quotas, cartel arrangements, and other preferential mechanisms.  Without such exclusionary devices, or else a coercive regime such as GATT, Conybeare concludes, there is no shadow of the future.

Robert Axelrod and Robert O. Keohane, "Achieving Cooperation Under Anarchy: Strategies and Institutions."

Axelrod and Keohane (AK) begin by comparing the analyses of IPE and security studies in the above case studies along the three dimensions of mutuality of interest, shadow of future, and number of players.  They emphasize the importance of actors’ shifting perceptions and interpretations of interests in the determination of payoff structure, and in changes to that structure (e.g. from PD to Deadlock).  Their key point (231) is that IPE and security cooperation can be analyzed with the same framework.  However, AK concede, this claim seems less true when it comes to the ‘shadow of the future’: IPE generally seems to exhibit more iteration and thus offers a better chance of successful retaliation for defection and more potential for cooperation, while military interactions may include the possibility of pre-emptive war and the ability to change the nature of the game at a single stroke.  Sanctioning problems occur in both the security and IPE contexts and are best explained by “the conditions that determine whether defection can be prevented through decentralized retaliation: the ease of identifying sources of action, the ability of governments to focus retaliation or reward on particular targets, and the incentives that exist for members of a group to punish defectors” (236).  

AK conclude by arguing that a focus on the broader contexts of international interactions –namely issue linkages, multi-level games, reciprocity in complex situations, and the role of international institutions – is crucial to understand how actors can adopt strategies to alter payoff structures, lengthen the shadow of the future, and break down N-player games, all of which can promote cooperation (as distinct from harmony of interests) under conditions of anarchy.  

James D. Fearon, “Bargaining, Enforcement, and International Cooperation,” International Organization 52:2 (Spring 1998), pp. 269-306.

Fearon argues against scholars in both the neo-realist and neo-liberal camps who use various 2x2 representations to characterize the strategic setting in international relations.  Such scholars distinguish various international contexts into categories such as P.D., Stag Hunt, Chicken, etc.  Much of the debate between neo-realist and neo-liberals has occurred, for example, over issues such as whether cooperation can be sustained under a P.D. or how Battle of Sexes-type situations might create conditions unfavorable for cooperation.

Fearon proposes that it is more accurate and theoretically fruitful to see most international cooperation as occurring under a common strategic structure.  This structure has two stages:
        1. Bargaining – A Coordination Game / War of Attrition over the terms and form of a cooperative agreement.  This type of game cannot easily be represented in 2x2 form and is better characterized as an n-round interaction with discounting.  Both countries incur costs if no agreement is reached, and whichever state “caves in” first tends to get a worse deal.
        2. Enforcement – A repeated P.D. in which {C,C} represents the bargaining outcome agreed to in Stage 1.

Several interesting implications arise from this analysis:
        1. A state’s relative power in a given situation is their willingness to hold out during the bargaining phase.  This contrasts to traditional measures such as state size and capability.
        2. States will only seriously bargain over issues in which enforcement is reasonably feasible.  Therefore scholars studying cooperation will face selection bias if they only look at cases where such negotiation takes place.  
        3. Contrary to common neo-liberal claims, an extended shadow of the future can be inimical for cooperation.  Neo-liberals often assert that a long shadow of the future is good for cooperation since it dissuades players from pursuing single-shot defection in a repeated P.D.  However, in Fearon’s two-stage model, this benefit only arises in Stage 2.  In Stage 1, states might actually choose to hold out longer during the bargaining process if they have a long time horizon.  This can lead to fewer international agreements being struck to begin with.  Hence, agreements with big consequences or a long duration will tend to be struck only after protracted negotiation if at all.

Charles P. Kindleberger, The World in Depression, 1929-1939.  Berkeley: University of California Press, 1973. pp. 288-305.

Ch. 14: An Explanation of the 1929 Depression
    Kindleberger tries to answer the cause for the widespread, long-lasted depression of 1929. His explanation is that the international economic system became unstable by British inability and U.S. unwillingness to assume responsibility for stabilization by neglecting five functions: maintenance of open market for trade, provision of stable long-term lending, policing of stable exchange rate system, ensuring macroeconomic policy coordination, and acting as a lender of last resort (providing liquidity in financial crisis). The author argues that these functions should be organized and executed by a single country that assumes responsibility for the system. The difficulty of stabilization lay in considerable latent instability in the system and the absence of a stabilizer (seriousness of the economic shock to the system does not matter much).
    The author compares British cooperative measures with the U.S. policies prior to the world depression, emphasizing the difference between the U.S. and U.K policies. Britain maintained free trade policy by opening their market (while the U.S. passed the Smoo-Hawley Tariff Act in 1930 which strengthened import restrictions), lend abroad on a countercyclical basis (lend to foreign nations when domestic economy is in recession, and draw foreign investment when domestic economy is in a boom), adhered to fixed exchange rate based on gold standard, and served as the lender of last resort. By 1931, however, it became clear the Britain could not exercise such leadership in international economy because its capacity for leadership had gone. But the U.S. declined its leader role offered by European states after the World War I from domestic interest.
    Kindleberger suggests three stable and three unstable scenarios on the leadership of the world economy. The stable three are: 1) continued U.S. leadership 2) new responsible leadership role by Europe, Japan, or some other country 3) effective cession of economic sovereignty to international institutions. The last is the most attractive but least likely. The unstable three are: 1) contention for leadership between major economies 2) one unable to lead and the others unwilling as was in 1929 to 1933 3) each [actors] retaining a veto over programs of stability or strengthening of the system without seeking to secure positive programs of its own.

David Lake, "Leadership, Hegemony, and the International Economy: Naked Emperor or Tattered Monarch with Potential?" International Studies Quarterly 37:4 (December 1993), pp. 459-489.

Hegemonic stability theory is the exploration of connections between hegemony and economic liberalization, and between hegemony and peace.  Lake's article is on hegemony theory as it relates to economic liberalization only and is primarily a review article.

Hegemonic stability theory was first noted regarding the British Empire and the rise of free trade by E.H. Carr (1939). Gilpin (1971) revived the subject in the language of modern political science, and ensuing discussion of Kindleberger (1973) and others ended in the mid-1980s. After a short pause in the discussion of hegemonic stability theory, Lake attempts to bring hegemonic stability theory from the "museum" of graduate student syllabi back onto the active research program of political science research, regardless of the "catcalls" of his peers that he received when proposing the revival at a conference. Lake says that a productive direction for leadership theory is to identify the necessary and sufficient conditions of international economic infrastructure, and explicating when leadership is likely to be benevolent or coercive.

Lake criticizes the field for theoretical sloppiness and divides the theory into "Leadership Theory", which "builds upon public goods models and seeks to explain the production of economic infrastructure" and "Hegemony Theory". which "focuses on the different structurally-derived trade policy preferences of states and attempts to explain international economic openness" (459).

Leadership theory, exemplified by Kindleberger (1973), holds that the public good of economic infrastructure requires a hegemon to be provided.  Lake defines economic infrastructure as providing monetary liquidity to avert freezes and panics, managing foreign exchange rates, and coordinating domestic monetary policies because a stable international economy requires a secure medium of exchange and stable store of value. In addition to international monetary coordination, a hegemon provides protection for international property rights. All of these functions are public goods. Critics outside the economic mainstream argue that these attributes of monetary and property stability are necessary for free trade, which is not a public good because free trade is good for hegemons but not necessarily good for smaller powers.

In Lake's primary theoretical contribution (he departs for a page or two from the review style), he argues that Kindleberger is wrong in thinking that there necessarily needs to be only one stabilizer (one hegemon). At times the public good cannot be provided with one state because of the high costs of provision, but can be provided by more than one state (what Lake defines as the "k-group") through Keohanian (1984) coordination. A single leader is therefore not a "necessary condition" of the provision of the public good of international monetary and property infrastructure.

Lake's "hegemony theory" assumes that states have different trade preferences -- some prefer free trade, while others prefer protectionism.  The hegemon prefers economic openness, while other states are split between those that prefer openness and those that prefer protectionist strategies, depending on their mix of interest groups, discount rates, and the size of their economy. Lake notes that hegemony theory goes against mainstream economics, in that it assumes that economic liberalization is not good for all states.

Todd Sandler, “The Economic Theory of Alliances: A Survey,” Journal of Conflict Resolution 37:3 (Sep., 1993), pp. 446-483.

Sandler discusses two models of alliances: the pure public good model and the joint product model.  In the pure public good model, because the benefits of defense are nonrival and nonexcludable, defense is regarded a public good.  It is nonrival because a unit of defense can be consumed by one ally without detracting from the consumption opportunities for other allies.  It is nonexcludable because the benefits of defense are available to all, once it is provided.  In this model, states seek to maximize their utility functions subject to their own resource constraint (see Equation 5).

Due to the public nature of the good, the collective action problem arises.  First, as the number of allies increases, the extent of suboptimal provision of the good is expected to worsen.  Secondly, the large, wealthy allies shoulder the defense burdens of the smaller allies.  And thirdly, the impact of endowment asymmetry among potential members on the collective action is important.  Unequal endowments are conducive to collective action, since it is then more likely that at least one potential participant will get sufficient benefits as to assume the entire cost of providing the good.

The joint product model differs from the public good model in “allowing the defense activitiy to produce a variety of outputs that may be purely public, purely private, and impurely public.” (459)  In this model, defense burden sharing is more apt to be associated with benefits received when private benefits are a significant proportion of defense benefits received by allies.  Factors such as strategic doctrine and military technology can determine the mix of public and private outputs derived from military activities.  The upshot of this model is that nations would want to consume more of the complementary good, and this is only possible if they increase their own defense activities.  Hence an increase in spillins no longer decreases the nation’s defense activities, provided that complementarity of joint products is sufficiently strong.

Sandler then introduces a number of theoretical extensions such as cost differences among allies, the endogeneity of threat, the variation of the underlying game structure, and the use of different technologies of public supply.  Lastly, the author tests the theory, and draws a number of conclusions:
    •    The joint product model tends to outperform the pure good model
    •    Allies tend to respond positively to spillins, thus supporting the notion of complementarity between defense joint products.
    •    Defense demands did not always respond positively to threat.  
    •    Military doctrine can affect the demand for defense.

Andrew Moravcsik, “Negotiating the Single European Act: National Interests and Conventional Statecraft in the European Community,” International Organization 45:1 (Winter 1991), pp. 19-56.

The Single European Act links liberalization of the European market with procedural reform. The first half of this reform package aims to create “an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured”. The second half consists of procedural reforms designed to streamline decision making in the governing body of the (then) EC, the Council of Ministers. The author seeks to explain the timing and content of the reform package, and presents two explanations. The first (which he rejects) stresses the independent activism of international or transnational actors and the second emphasizes bargaining between the leaders of the powerful states.

According to the “supranational institutionalism” account, the reform was guided by actors and institutions acting “above” the nation-state. The European Parliament advocated European federalism and a broad expansion in the scope of EC activities. Business interest groups, at times working together with EC officials, hoped to bolster the competitiveness of European firms by calling for a more liberal EC market. The Commission, under the leadership of Delors, also pushed toward the construction of a single European market. The core of the supranationalist explanation is thus an elite alliance between transnationally organized big business groups and EC officials, led by Delors.

The alternative approach, intergovernmental institutionalism, includes the following principles:
1. Intergovernmentalism - The EC is based on interstate bargains between its leading member states. Heads of government with a small group of ministers and advisers initiate and negotiate the major initiatives.
 2. Lowest-common-denominator bargaining – The bargains struck in the EC reflect the relative power positions of the member states. Larger states hold a de facto veto power over fundamental changes in the scope or rules of economic liberalization. Thus bargaining tends to converge toward the lowest common denominator of large state interests.
 3. Protection of sovereignty - Policymakers safeguard their countries against the future erosion of sovereignty by demanding the unanimous consent of regime members to sovereignty-related reforms. They also avoid granting open-ended authority to central institutions that might infringe on their sovereignty and prefer to work through intergovernmental institutions such as the Council of Ministers.

According to the intergovernmental approach, an analysis of the 1992 initiative must begin by examining the underlying preferences of Germany, France and Britain. France supported strengthening monetary coordination and political and defense cooperation, but was opposed by Britain and German in both cases. All three were inclined towards procedural reform in the EC decision making institutions and liberalization of the internal market.

After reviewing the history of the 1992 initiative the author concludes that the historical record does not confirm the importance of international and transnational actors. Representatives of the Parliament were excluded from decisive forums. The Parliament actually passed in the end a resolution protesting that the SEA “in no way represent[s] the real reform of the Community that our peoples need”. Most transnational business lobbies got involved late. While logistical support from the Commission may indeed have hastened a final agreement, there is little evidence that it altered its substance. The evidence, however, points to the importance of the three elements of intergovernmental institutionalism. Heads of government and their representatives carried out the negotiations. The result represents the convergence of domestic policy preferences in the largest member states. The steady narrowing of the institutional reform to a “minimalist” position in which majority voting is restricted to internal market policy, the power of Parliament is limited, and the future spillover to areas such as monetary policy is blocked confirms the enduring preoccupation of all three major states with maintaining sovereignty and control over future changed in the scope of EC activities.

Another question that should be asked is why did underlying national policy preferences converge at this point in time. Part of the answer can be found in the domestic politics of France, Germany and Britain. The author suggests four explanations, none of which is entirely satisfactory.
1. The convergence of policy preferences reflected the views, either European or neoliberal, of the three major European leaders of the time – Mitterand, Kohl and Thatcher.
2. Domestic bureaucracies internalized norms of cooperation and changed the views of heads of government.
3. Over the years, centrist parties, particularly those of the center-right have tended to support EC reforms, while the strongest opposition to further integration has been located on the extremes of the ideological spectrum. In the mid-1980s, the dominance of centrists in ruling coalitions created a rare opening for reform.
4. In the 1970s and early 1980s the three states faced economic difficulties. The convergence of policy preferences may have resulted from the failure of purely national strategies of economic policy, which created or legitimated pressure for coordinated liberalization at the European level.        

James Alt and Michael Gilligan, “The Political Economy of Trading States: Factor Specificity, Collective Action Problems and Domestic Political Institutions,” Journal of Political Philosophy 2:2 (1994), pp. 165-192.

The authors proceed in three steps.  1) Given the stakes in trade issues, which groups will overcome collective action problem and form coalitions? 2) What are the stakes in trade issues? These are determined by factor endowment and factor mobility. 3) What is the effect of domestic political institutions on the composition and strength of the coalitions.

I.  Collective Action

Pareto observed that “a protectionist measure provides large benefits to a small number of people, and causes a very great number of consumers a slight loss. This circumstance makes it easier to put a protectionist measure in practice” (168).  Why is this so?
            Collective political action is a public good, and, therefore, its provision has the problem of excludability.  People will only contribute to the point where their marginal benefit from the good equals the marginal cost to them – in other words, they don’t internalize the benefits derived by the whole group from their individual contributions.  This leads to a sub-optimal provision of the public good.  Since each member of the group can consume contributions of others, each person receives less benefit from their own contribution than it cost to make that contribution, and so they consume less of the good than would be the case if the good were excludable. 
            The relationship between the size of the group and the amount of public good provided is ambiguous.  Though members of larger groups contribute smaller individual amounts, the total amount provided increases with numbers.  Within groups, those who have larger stakes in the issue will provide a disproportionately larger share of the good (e.g. lobbying). 
            How then do we explain Pareto’s observation?  The are two reasons for why this happens: 1) there may be per person transaction costs in organizing groups.  Then it would cost more to organize large groups; 2) “if policy outcomes are probabilistic, members of large groups with small per person stakes and contributions may suppose that their own contributions will be insignificant to the political outcome and therefore not make them” (170).

II. Preference formation and coalitions based on preferences.
A.     Hecksher-Ohlin (HO) Stolper-Samuelson (SS) – Mobile factors
            The SS theorem states that an increase in a price of a product (e.g. caused by protection) will cause an unambiguous increase of real incomes of owners of factors (e.g. land, labor, capital) used intensively in the production of that good.  Because factors are assumed to be mobile across sectors, owners of the same factor will experience the same changes to their returns, regardless of which industry they were in.
            HO theorem states that a country will export goods that use relatively abundant factors of production and import goods that use relatively scarce factors.  Combining the two theorems, we see that free trade will increase the returns to the abundant factors and decrease the returns to the scarce factors.  Thus, the conflict is between the factors, with abundant factors favoring free trade and scarce factors favoring protection (Rogowski’s arguments).   

B.     Ricardo-Viner (RV) – Specific factors
            Specificity corresponds to the loss of value in moving an asset from its current to its next best use.  Highly specialized and differentiated economies, high transaction costs, ethnic rivalries, centrally planned economies, and other barriers to entry/exit to/from industries increase specificity.
            RV assumes that some factors are specific, and one (e.g. labor) is mobile.  Let’s say that freer trade decreased the price of import-competing good.  Labor will flow out of import-competing and into exporting industries.  But the specific factors remain in the import competing industry and suffer a productivity decline because of labor outflow.  The income of that specific factor will fall with respect to both the imported and the exported goods.  Similarly, the income of factors specific to the exporting industry will unambiguously rise. 
            The effect on the mobile factor is ambiguous.  First, it depends on the intensity of use.  If labor is used more intensively in the exporting industry, its income will increase; if it’s used more intensively in the import-competing industry, its income will decrease.  Second, it depends on the consumption patens.  If labor heavily consumes the imported good, whose price went down, its real income is likely to increase; if it heavily consumes the exported good, its income is likely to decrease.  The tradeoffs are illustrated in Figure 1 (177).  Thus, in a specific factor model, the conflict will be between the exporting and import-competing sectors, with exporting sectors favoring free trade and import-competers favoring protection. 
            If the mobile factor is used intensively in the exporting sector and heavily consumes the imported good, it will join the free-trade coalition.  On the other hand, if the mobile factor is used intensively in the import-competing sector and heavily consumes the exported good, it will join the protectionist coalition.  When the combination of intensity of use and consumption patterns is such that the effect on the mobile factor is ambiguous, the specific factors will offer side-payments to the mobile factor to induce its support. 

III.  Coalitions: Factor Mobility, Institutions, and Collective Action
           High factor mobility means that benefits of trade policy will be highly dispersed across sectors, which produces non-excludability and raises collective action costs.  Converse is true for when factors are specific. 
            Majoritarian institutions make necessary formation of large interest groups that can better affect policy outcome.  SS thm is more consistent here.  Conversely, when decision makers are insulated, access to the decision-maker becomes critical, and support of large groups unnecessary.  RV thm is more consistent here.
            Thus, “[o]ther things being equal, Majoritarian political systems and factor mobility will both mean that benefits will be more dispersed and, therefore that it will be harder to organize a successful interest group” (183).
            The possible coalition patterns based on costs of collective action and factor specificity are represented in Figure 3 (184).  SS-HO/Rogowski’s theories of broad, factor-based coalitions, work only when factors are mobile and collective action cost are low.  Raise the collective action costs, and no coalitions will form, since free-riding will be rampant.  Make factors specific while keeping collective action cost low, and many interest groups will form, including consumers who are inherently pro-free-trade.  Then coalition of import-competing industries will have to form to counter consumers and exporters.  Such coalition would be unstable, since it will produce policies of protection for all import-competing industries, and with that level of protection, everyone would be better off under free trade. 
            When factors are specific and collective action costs are high, individual industries will request and often receive protection and trade policy will be dominated by special interest, as described by Pareto and Olson and consistent with RV thm.
            By adding institutions (as described above) to the mix of dimensions affecting trade policy coalitions, we get a wide array of possible outcomes, with SS and RV models as only two (out of 8).  Figure 4 (186).
            In their conclusion, the authors note that it is sometimes stated that since factors are mobile in the long run, the SS/Rogowski model applies in the long run, even though RV applies in the short run.  This is not necessarily true, since owners of specific factors may undertake many actions to increase the specificity of their assets in order to keep their returns from protection high in the future.  Thus, RV-Frieden could apply in the long run, as well as in the short run.