¹ World Trade Organization, Regionalism and the World Trading System (Geneva: World Trade Organization, 1995)Google Scholar.

² Note that concessions became effectively multilateral because the unconditional (as compared with the conditional) MFN clause specifies that concessions made by the signatories to third parties must be granted to signatories of the treaty for no additional compensation; see Hornbeck, S. K., “The Most-Favored-Nation Clause in Commercial Treaties,” Bulletin of the University of Wisconsin (February 1910), 343Google Scholar. States did develop means to keep concessions bilateral, for example, through minute subdivisions in tariff categories; see, e.g., Platt, D. C. M., Finance, Trade, and Politics in British Foreign Policy, 1825–1914 (Oxford: Clarendon Press, 1971), 93Google Scholar.

³ This matches the argument of hegemonic stability theory. For example, see Kindleberger, C. P., The World in Depression, 1929–1939 (Berkeley: University of California Press, 1973)Google Scholar; and Gilpin, Robert, War and Change in World Politics (Cambridge: Cambridge University Press, 1981)CrossRefGoogle Scholar. Also, for a valuable critique of hegemonic stability theory, see Lake, David, “Leadership, Hegemony, and the International Economy: Naked Emperor or Tattered Monarch with Potential?” International Studies Quarterly 37 (December 1993)CrossRefGoogle Scholar.

⁴ Notably, the tariff levels of European states in the nineteenth century were not correlated with the degree of concentration of power in the international system. See McKeown, Timothy J., “Hegemonic Stability and the Nineteenth Century Tariff Levels in Europe,” International Organization 37 (Winter 1983)CrossRefGoogle Scholar.

⁵ The first major work in the study of emergent systems was Schelling, Thomas, Micromotives and Macrobehavior (New York: W. W. Norton, 1978). Recent efforts in this direction include Robert Axelrod, The Complexity of Cooperation: Agent-Based Models of Competition and Collaboration (Princeton: Princeton University Press, 1997)Google Scholar; and Epstein, Joshua M. and Axtell, Robert A., Growing Artificial Societies: Social Science from the Bottom Up (Washington, D.C.: Brookings Institution, 1996)Google Scholar. For economic models of this genre, see Krugman, Paul, The Self-Organizing Economy (Cambridge: Blackwell Publishers, 1996)Google Scholar; and Scheinkman, J. A. and Woodford, M., “Self-Organized Criticality and Economic Fluctuations,” American Economic Review 84 (May 1994), 417–21Google Scholar.

⁶ Oye, Kenneth, Economic Discrimination and Political Exchange: World Political Economy in the 1830s and 1980s (Princeton: Princeton University Press, 1992)Google Scholar; Gilligan, Michael J., Empowering Exporters (Ann Arbor: University of Michigan Press, 1997)CrossRefGoogle Scholar; and Baldwin, Richard, “A Domino Theory of Regionalism,” in Bhagwati, Jagdish, Krishna, Pravin, and Panagariya, Arvind, Trading Blocs: Alternative Approaches to Analyzing Preferential Trade Agreements (Cambridge: MIT Press, 1999)Google Scholar.

⁷ There is a large literature on the economics of customs unions, with a welter of findings (often contradictory) regarding exactly when customs unions may be harmful to the welfare of members. The starting point of modern research on customs unions is Viner, Jacob, The Customs Union Issue (New York: Carnegie Endowment for International Peace, 1950)Google Scholar. More recent landmarks include Riezman, Raymond, “A 3 × 3 Model of Customs Unions,” Journal of International Economics 9 (August 1979)CrossRefGoogle Scholar; P. Lloyd, J., “3 × 3 Theory of Customs Unions,” Journal of International Economics 12 (February 1982)CrossRefGoogle Scholar; and Wooton, Ian, “Preferential Trading Agreements: An Investigation,” Journal of International Economics 21 (August 1986)CrossRefGoogle Scholar. For a striking illustration of the fact that a world dominated by a few customs unions may be associated with lower world welfare than a world of no trade blocs (even holding tariffs constant), see Krugman, Paul, “The Move toward Free Trade Zones,” in Policy Implications of Trade and Currency Zones: A Symposium Sponsored by the Federal Reserve Bank of Kansas City (Kansas City, Mo.: Federal Reserve Bank of Kansas City, 1991)Google Scholar; and Bond, Eric and Syropoulos, Constantinos, “The Size of Trading Blocs: Market Power and World Welfare Effects,” Journal of International Economics 40 (May 1996)CrossRefGoogle Scholar. Kemp and Wan demonstrate, however, that there is always a level of postunion tariffs that is Pareto-improving for both PTA members and nonmembers; see Kemp, Murray C. and Wan, Henry Y. Jr., “An Elementary Proposition Concerning the Formation of Customs Unions,” Journal of International Economics 6 (February 1976), 95–9CrossRefGoogle Scholar.

⁸ In principle, dyadic in this context incorporates PTAs with more than two members. None of the treaties signed during the 1860s had more than two parties, however.

⁹ See Rogowski, Ronald, Commerce and Coalitions: How Trade Affects Domestic Political Alignments (Princeton: Princeton University Press, 1989)Google Scholar. Notably, however, Rogowski demonstrated how longrun coalitions form around trade issues, whereas a coalition resulting from a particular PTA would presumably reflect a short-run congruence of interests.

¹⁰ For example, Trefler, Daniel, “Trade Liberalization and the Theory of Endogenous Protection,” Journal of Political Economy 101 (February 1993)CrossRefGoogle Scholar; Finger, M. and Harrison, A., “The MFA Paradox: More Protection and More Trade?” (Manuscript, World Bank, 1995)Google Scholar; and Milner, Helen, Resisting Protectionism: Global Industries and the Politics of International Trade (Princeton: Princeton University Press, 1988)Google Scholar.

¹¹ There is a third alternative—multilateral reductions in tariffs, as has occurred under the auspices of GATT in the post-World War II period. There was no equivalent forum to pursue multilateral reductions in tariffs in the 1860s. Why there was no equivalent of GATT during the 1860s is an interesting question but one not pursued here.

¹² Oye (fn. 6); and Gilligan (fn. 6). Closely related are the arguments of Destler, I. M. and Odell, John, Anti-Protection: Changing Forces in the United States Trade Politics (Washington, D.C.: Institute for International Economics, 1987)Google Scholar. While they do not directly discuss PTAs, they do point out that potential retaliation from country A in response to protectionist measures in country B may have a very focused distributional impact on exporters in B.

¹³ Political economy models illustrating the possibility of a prisoners' dilemma structure of payoffs of a PTA include Pahre, Robert, “Reactions, Reciprocity, and Retaliation: A Model of Tariff Policy Linking State and Domestic Interests” (Manuscript, 1994)Google Scholar; and Grossman, Gene and Helpman, Elhanan, “Trade Wars and Trade Talks,” Journal of Political Economy 103 (August 1995)CrossRefGoogle Scholar. See also Putnam, Robert, “Diplomacy and Domestic Politics: The Logic of Two-Level Games,” International Organization 42 (Summer 1988)CrossRefGoogle Scholar. Closely related are Johnson, H. G., “Optimum Tariffs and Retaliation,” Review of Economic Studies 21, no. 55 (1954)Google Scholar; Riezman, Raymond, “Tariff Retaliation from a Strategic Viewpoint,” Southern Economic Journal 48 (January 1978)Google Scholar; and Conybeare, John, Trade Wars: The Theory and Practice of International Commercial Rivalry (New York: Columbia University Press, 1987)Google Scholar. These last three assert that states may face a prisoners' dilemma with respect to the imposition of optimal tariffs to improve terms of trade.

¹⁴ Oye (fn. 6) labels these “divertable externalities,” in that the costs to a particular third party may be diverted to fourth parties by extending the market access to that third party.

¹⁵ For example, Grossman and Helpman showed, using a median voter model, that free-trade areas are more likely to garner political support when trade diversion outweighs trade creation; see Grossman, Gene and Helpman, Elhanan, The Politics of Free Trade Agreements, Working Paper 4597 (Cambridge: National Bureau of Economic Research, 1993)CrossRefGoogle Scholar.

¹⁶ For an insightful discussion of this point, see Rodrik, Dani, “Political Economy of Trade Policy,” in Grossman, Gene and Rogoff, Kenneth, eds., Handbook of International Economics (New York: Elsevier, 1995)Google Scholar.

¹⁷ Baldwin (fn. 6).

¹⁸ Ibid.

¹⁹ Milner, Helen, “Industries, Governments, and the Creation of Regional Trade Blocs,” in Ruggie, John G., ed., The Political Economy of Regionalism (New York: Columbia University Press, 1997)Google Scholar.

²⁰ For an early and insightful discussion of the interrelationship between preference/discrimination and product differentiation, see Benham, Frederic, Great Britain under Protection (New York: Macmillan, 1941)Google Scholar. For a discussion of why political mobilization of sectors is more likely when the goods produced are differentiated, see also Gilligan, Michael, “Lobbying as a Private Good with Intra-Industry Trade,” International Studies Quarterly 41 (September 1997)CrossRefGoogle Scholar.

²¹ Ashley, Percy, Modern Tariff History: Germany-United States-France (New York: E. P. Dutton, 1926), 326Google Scholar.

²² Ibid., .327.

²³ For an intriguing analysis that applies coalition theory to customs-union formation, see Yi, Sang-Seung, “Endogenous Formation of Customs Unions under Imperfect Competition: Open Regionalism Is Good,” Journal of International Economics 41 (August 1996)CrossRefGoogle Scholar.

²⁴ See Valente, Thomas, Network Models of the Diffusion of Innovations (Cresskill, N.Y.: Hampton Press, 1995)Google ScholarPubMed. For a discussion of diffusion processes and a consideration of diffusion through some central source versus diffusion via contagion, see Lave, Charles and March, James, An Introduction to Models in the Social Sciences (New York: Harper and Row, 1975)Google Scholar. For a classic analysis of diffusion processes, see Coleman, James, Katz, E., and Menzel, H., Medical Innovation: A Diffusion Study (Indianapolis, Ind.: Bobbs-Merrill, 1966)Google Scholar. For a recent discussion, see Rogers, Everett, Diffusion of Innovations (New York: Free Press, 1995)Google ScholarPubMed.

²⁵ This is a relatively reasonable assumption in the domain of trade, where iteration is high, fostering mutual cooperation. See Axelrod, Robert, The Evolution of Cooperation (New York: Basic Books, 1984)Google Scholar. Of course, there may be substantial conflict over the distribution of the gains of cooperation, sometimes leading to mutual defection, as the occasional trade wars of the 1880s and 1890s attest (e.g., France-Italy, Germany-Russia).

²⁶ Simon, Herbert, Models of Man (New York: Wiley, 1957)Google Scholar.

²⁷ Lohmann, Susanne, “The Dynamics of Informational Cascades: The Monday Demonstrations in Leipzig, East Germany, 1989–91,” World Politics 47 (October 1994)CrossRefGoogle Scholar; Granovetter, Mark, “Threshold Models of Collective Behavior,” American Journal of Sociology 83 (May 1978)CrossRefGoogle Scholar; Granovetter, Mark and Soong, R., “Threshold Models of Diversity: Chinese Restaurants, Residential Segregation, and the Spiral of Silence,” in Clogg, C., ed., Sociological Methodology, vol. 18 (Washington, D.C.: American Sociological Association, 1988)Google Scholar.

²⁸ Valente (fn. 24).

²⁹ As discussed below, this assumption regarding the distribution of g_ij greatly facilitates interpretation of the model. All thresholds are also assumed to be independent of each other. Note that in reality this is not likely to be the case; e.g., some states might be especially likely to sign treaties, and some especially unlikely, so one might expect g_ij to be correlated with g_ik.

³⁰ A lattice also has the advantage of making it easier to visualize the process of contagion, as the example below illustrates.

³¹ The exact amount the thresholds should be lowered is arbitrary. Lowering thresholds by 1 has the benefit in terms of ease of interpretation of the simulations in that when a state has δ partners, it will definitely sign treaties with its remaining neighbors.

³² The sum of the thresholds = 40 (number of dyads) × 1.5 (average threshold) = 60.

³³ The probability that three treaties will be signed is

for δ>2 and otherwise

³⁴ Note that there are six distinct degrees of centrality in a 5 × 5 lattice. Additional simulations indicate that centrality of the first mover is monotonically related to the average number of treaties signed in equilibrium.

³⁵ The probability that A will sign a treaty with either B or F in the first round is 1 −(2/3)² ≈ .56. The proportion of treaties signed in Figure 12 reaches a maximum of approximately .5, which means that if the first mover were large and signed a treaty with at least one other state in the first round, subsequently 90 percent of potential treaties would be signed (on average).

³⁶ Note that this increases the number of potential treaties from forty to seventy-two; the analysis below, of course, examines the proportion of potential treaties that are signed.

³⁷ If the initial actor to sign a treaty is in the corner, the analogous numbers are 29 percent and 69 percent, respectively.

³⁸ While it is prohibitively difficult in this model to analytically derive the probability that a certain number of treaties will be reached in equilibrium, it is simple to calculate the probability that zero treaties will be reached under a given set of assumptions. Specifically, that number is

where p is the number of treaty-eligible neighbors the first mover has. Thus, for = 3, in the highly connected scenario (p = 8), this is 3.9 percent; in the less connected scenario (p = 4) this is 20 percent; in the totally connected scenario (p = 24), .006 percent.

³⁹ Note this is different from the preceding assumptions regarding interconnectedness. The reason for this last assumption was that this way all actors in the system have the same level of interactions, thus eliminating geographic centrality as a factor in determining centrality in the network of treaties.

⁴⁰ There were 0 treaties 82 times (.9²⁴ × 1000 = 80), and 1 treaty once.

⁴¹ At t = 0, g_ij is drawn from a uniform distribution between 0 and 10. After the system reaches an equilibrium where every state has a treaty with every other state, the member of each dyad has 23 treaties with nondyad members, so g_ij = the initial g_ij − (2 × 23). Thus, to increase the postequilibrium g_ij. to a positive number, one would need to add between 36 and 46. To put this in more general terms, in a system which has reached an equilibrium where every state has signed a treaty with every other state, and every state has M trading partners, and the initial g_ij is drawn from a uniform distribution between 0 and, in order to eliminate the treaty between i and j, g_ij will need to be increased by (2 × M)—the initial g_ij. Thus, notably, the more interconnected the world (the larger M), the more robust the equilibrium where every state has signed a treaty with every other state.

⁴² If all treaties with one state, I, were removed, g_kl would equal the initial g_kl (drawn from between 0 and 10) − (2 × 22), which would be between −34 and −44.

⁴³ For a perceptive discussion of the MFN network of treaties from 1860 to 1929, see Irwin, Douglas, “Multilateral and Bilateral Trade Policies,” in de Melo, Jaime and Panagariya, Arvind, eds., New Dimensions in Regional Integration (Cambridge: Cambridge University Press, 1993)Google Scholar.

⁴⁴ Landes, David, “Technological Change and Development in Western Europe, 1750—1914,” in Clapham, J. H., ed., The Cambridge Economic History of Europe: The Industrial Revolution and After, pt. 1 (Cambridge: Cambridge University Press, 1965), 363Google Scholar.

⁴⁵ Ibid., 450.

⁴⁶ Ibid., 429.

⁴⁷ See Bogart, Ernest, Economic History of Europe, 1760–1939 (London: Longmans, Green, 1942), 109, 118Google Scholar.

⁴⁸ Bale, Per and Chen, K., “Self-Organizing Criticality,” Scientific American 264 (January 1991)Google Scholar.

⁴⁹ Lord Napier, British ambassador to Berlin, to Lord Russell, British foreign secretary, March 24, 1865; originally cited in Peter Marsh, Bargaining on Europe: Britain and the First Approach to a European Economic Community, 1860—1892 (New Haven: Yale University Press, forthcoming), chap. 3.

⁵⁰ An important antecedent condition to this treaty was the general increase in power of Napoleon III and a French constitution that gave the executive (Napoleon III) power to set tariffs in treaties that did not require ratification by the legislature. This institutional setting of executive autonomy determined what Nelson labeled the “supply” conditions of protection and would be exogenous in the model above; see Nelson, Douglas, “Endogenous Tariff Theory: A Critical Survey,” American Journal of Political Science 32 (August 1988)CrossRefGoogle Scholar. A second factor that increased Napoleon's desire to strike a deal with Britain was his aim to acquire Britain's cooperation with respect to his objectives in Italy. See McKeown (fn. 4); and Ratcliffe, Barrie M., “The Origins of the Anglo-French Commercial Treaty: A Reassessment,” in Ratcliffe, , ed., Great Britain and Her World, 1759–1914: Essays in Honour of W. O. Henderson (Manchester: Manchester University Press, 1975)Google Scholar. As noted above, for each dyad there will be a variety of idiosyncratic factors that affect the tendency of that pair to sign a treaty.

⁵¹ Dunham, Arthur Louis, Anglo-French Treaty of Commerce of 1860 (Norwood: Plimpton Press, 1930)Google Scholar.

⁵² Schippel, Max, Grundzuge der Handelspolitik (Berlin: Akademischer Verlag fur sociale Wissenschaften, 1902), 171Google Scholar, as cited in Ashley (fn. 21), 32.

⁵³ Bairoch, Paul, “European Trade Policy, 1815–1914,” in Clapham, J. H. et al. , The Cambridge Economic History of Europe (Cambridge: Cambridge University Press, 1989)Google Scholar.

⁵⁴ More than a third of Belgian exports went to France at this time—162m francs out of total exports of 470m francs; see Mitchell (1975), 489, 506. It exported about as much to France as to its next two largest trading partners combined; e.g., in 1860 its exports to France were valued at 162m francs, and to Germany and the U.K., at 169m francs.

⁵⁵ Ashley (fn. 21), 32.

⁵⁶ Memo, Joseph Crowe (consul-general for the Kingdom of Lower Saxony) to Lord Russell, May 15, 1861, as originally cited in Marsh (fn. 49), chap. 3.

⁵⁷ Lambi, Ivo Nikolai, Free Trade and Protection in Germany, 1869—1879 (Wiesbaden: Franz Steiner, 1963), 35Google Scholar.

⁵⁸ von Delbrixck, Rudolf, Lebenserinnerungen, 1817—1867, vol. 2 (Leipzig: Duncker und Humblot, 1905), 201–3Google Scholar, as cited in Lambi (fn. 57), 5, emphasis added.

⁵⁹ Bairoch (fn. 53), 40; see also Figure 15.

⁶⁰ This period was chosen because many of the treaties stated that they would come up for renewal in ten to twelve years, and the data on the renewal of particular treaties are incomplete. It is at least possible to say with some reliability that the treaties captured in Figure 15 were actually in force during this time.

⁶⁰ This period was chosen because many of the treaties stated that they would come up for renewal in ten to twelve years, and the data on the renewal of particular treaties are incomplete. It is at least possible to say with some reliability that the treaties captured in Figure 15 were actually in force during this time.

⁶³ Of course, this is less puzzling if one views Britain as the benevolent hegemon rather than as an exploitative hegemon; see Snidal, Duncan, “The Limits of Hegemonic Stability Theory,” International Organization 39 (Autumn 1985)CrossRefGoogle Scholar. One might hypothesize that Britain encouraged other states to sign treaties as part of its effort to decrease trade barriers in the international system. There is no evidence to support this hypothesis, and some that contradicts it, e.g., Fuchs (fn. 62); and Marsh (fn. 49).

⁶⁴ Fuchs (fn. 62), 29.

⁶⁵ Marsh (fn. 49), chap. 2.

⁶⁶ Ibid., chap. 3.

⁶⁷ Ibid.

⁶⁸ Ibid.

⁶⁹ Landes (fn. 44), 440.

⁷⁰ Thus, for example: “The particular blend of worsted wear in which Bradford specialized included warps of cotton, which remained cheaper than wool, to produce a hard and lustrous cloth which went well over the crinolines of the day, better than the more expensive all-woolen worsteds in which the French specialized”; Marsh (fn. 49), chap. 3.

⁷¹ Dunham (fn. 51), 88.

⁷² Fuchs (fn. 62), 31–32.

⁷³ Ibid., 59.

⁷⁴ Ibid., 30–31.

⁷⁵ Irwin (fn. 43).

⁷⁶ Fuchs (fn. 62), 64.

⁷⁷ Ibid., 65.

⁷⁸ Ibid., 66.

⁷⁹ Ibid., 70.

⁸⁰ Irwin (fn. 43), 105.

⁸¹ Note that these levels were significantly higher than pre-World War I levels.

⁸² Haight, Frank A., A History of French Commercial Policies (New York: Macmillan, 1941)Google Scholar.

⁸³ There are, of course, other processes at work that might confound contagion. For example, in the case of the EC/EU, members at any given time will need to evaluate the policy preferences of potential members. If those preferences differ substantially from those of current members, current members may delay membership of potential members until those differences matter less. See Downs, George, Rocke, David, and Barsoom, Peter, “Managing the Evolution of Multilateralism,” International Organization 52 (Spring 1998)CrossRefGoogle Scholar. Also, since the EU allows free movement of factors, some potential members raise the possibility of large (and politically costly to current members) migrations to current EU members. The EU has developed “second class” citizenship for Turkey and many of the countries in Central Europe, offering, that is, much of the access to goods but not to people. From a research design perspective, the system of bilateral treaties conveniently eliminates some of the messiness involved in the process of adhesion to multilateral PTAs.

⁸⁴ See also Yi (fn. 23).

⁸⁵ For an intriguing analysis of the Zollverein and other multilateral PTAs, see also Mattli, Walter, The Logic of Regional Integration: Europe and Beyond (Cambridge: Cambridge University Press, 1999)CrossRefGoogle Scholar. Mattli's argument is that (1) customs unions represent an attempt to internalize externalities that cross borders within the customs union; and (2) decreased transportation costs increase externalities across borders (increasing the payoffs to a customs union) and decrease the costs of administering a larger geographic unit.

⁸⁶ Bowring, John, Report on the Prussian Commercial Union (London: W. Clowes, 1840), 33–34Google Scholar, as cited in Ashley (fn. 21), 11.

⁸⁷ Bowring (fn. 86), 37, as cited in Ashley (fn. 21), 11.

⁸⁸ Henderson, William Otto, The Zollverein (London: Frank Cass, 1984), 107Google Scholar.

⁸⁹ Ibid., 103–49.

⁹⁰ Moravcsik, Andrew, “Negotiating the Single European Act: National Interests and Conventional Statecraft in the European Community,” International Organization 45 (Winter 1991)CrossRefGoogle Scholar.

⁹¹ Cowles, Maria Green, “Setting the Agenda for a New Europe: The ERT and EC, 1992,” Journal of Common Market Studies 33 (December 1995), 507CrossRefGoogle Scholar.

⁹² Organization for Economic Cooperation and Development, Regionalism and Its Place in the Multilateral Trading System (Paris: OECD, 1996), 76Google Scholar.

⁹³ Baldwin (fn. 6), 489.

⁹⁴ Fioretos, Karl-Orfeo, “The Anatomy of Autonomy: Interdependence, Domestic Balances of Power, and European Integration,” Review of International Studies 23 (July 1997), 313–14CrossRefGoogle Scholar.

⁹⁵ Oye (fn. 6), 164. Note that customs unions-also have the analytically distinct property that because there is a common exterior tariff, the collective might be able to reap gains through the imposition of an optimal tariff. Further, it is clear that at least in the case of the EC, a major objective of members was to allow EC firms to achieve economies of scale that would allow them to compete with U.S. firms; see Balassa, Bela, The Theory of Economic Integration (Homewood, II.: Richard Irwin, 1961)Google Scholar.

⁹⁶ Lawrence, Robert Z., Regionalism, Multilateralism, and Deeper Integration (Washington, D.C.: Brookings Institution, 1996), 77Google Scholar.

⁹⁷ Yeats finds that trade within Mercosur has grown fastest in those products that do not have a comparative advantage and concludes that Mercosur is therefore trade diverting; see Yeats, Alexander, Does Mercosur's Trade Performance Raise Concerns about the Effects of Regional Trade Arrangements, Policy Research Working Paper, no. 1729 (Washington, D.C.: World Bank, 1997)Google Scholar.

⁹⁸ Yi finds that if agreements are open to new members, trade blocs will snowball so that eventually all states will belong to one PTA (i.e., there will be multilateral free trade); see Yi (fn. 23).

⁹⁹ Young, Oran, “Regime Dynamics: The Rise and Fall of International Regimes,” in Krasner, Stephen, ed., International Regimes (Ithaca, N.Y.: Cornell University Press, 1983)Google Scholar.

¹⁰⁰ Ibid., 99.