From GATT to the World Trade Organization

Twenty-three governments, dominated by the United States, agreed tariff reductions that reflected the negotiating power of the state with the only functioning economy. Trade rules written at a moment of extreme asymmetry encode that asymmetry, and the exceptions negotiated then still favour their authors.

The charter covered employment, commodity agreements, restrictive business practices and investment, and the American Senate would not accept it. The ambitious instrument died and the modest provisional one survived for forty-seven years, which is a common fate for comprehensive designs.

The round was launched partly to respond to European integration, since a common external tariff around a growing market threatened American exports. Trade negotiations are conducted about market access, and the driver here was a shift in the relative position of two blocs.

With tariffs already low, the obstacles were subsidies, standards, procurement and customs valuation, all of which are domestic regulation. Trade negotiation moved inside the state, and every subsequent round has therefore been about sovereignty as much as about commerce.

The United States pushed services, intellectual property and investment onto the agenda because they were where its comparative advantage lay, and developing countries resisted for the same reason. Agendas are bargained, and what gets negotiated reflects who wants what.

The secretariat produced a complete draft because the members could not, filling gaps with its own compromises. A secretariat that drafts rather than records acquires real influence, and Dunkel used it deliberately to force the parties to react to a text.

Two parties settled the issue that had blocked a hundred and seventeen others, in a bilateral deal presented to the rest afterwards. Multilateral systems are frequently unblocked by bilateral bargains between the members whose disagreement is decisive.

The new organisation had binding dispute settlement with automatic adoption of rulings and authorised retaliation, which is real enforcement. It works between members of comparable weight and is far less useful to a small state whose retaliation nobody would notice.

An open trading order requires a state willing to keep its market open when others do not, and in 1947 only one could. Hegemonic provision of that public good is what made the system work, and the arguments about what happens when it withdraws began in the 1970s.

The provisions the developing and European members most wanted were exactly those the United States would not ratify. A charter negotiated to accommodate many interests could not pass the legislature of the state whose participation was indispensable.

By 1967 Europe negotiated as one and Japan had recovered, so the United States faced counterparts rather than supplicants. The distribution of economic power had changed within twenty years, and the terms of subsequent rounds reflected it.

The round concluded amid the oil shocks, currency instability and rising protectionism of the 1970s, and it held the system together at a moment of real strain. Institutions prove their value in bad conditions rather than good ones.

Agriculture had been excluded from every previous round because the European Community and the United States could not agree, and including it made this round both more valuable and nearly impossible. Deadlock between the two largest members determines what a system can attempt.

The agricultural provisions were rejected by France and the round stalled for another year, because one large member's domestic politics could block a global agreement. In a consensus system the most constrained party sets the pace.

France rejected the accord and eventually extracted modifications, showing that even within a bloc the internally constrained member retains leverage. Two-level bargaining, between states and within them, is where trade negotiations are actually decided.

The system was built when the United States was unchallenged, and China's accession in 2001 introduced a member whose economic model the rules had not anticipated. Institutions encode the distribution at their founding, and this one has been strained by its change.

The agreement rested on the judgment that the trade wars of the 1930s had contributed to the political catastrophe that followed. Liberalising commerce was pursued as a security policy, and its architects said so explicitly and repeatedly.

Truman declined to resubmit it after judging the votes were not there, and the executive's international commitment was defeated by domestic politics. This is the same constitutional mechanism that killed the League, thirty years later and with less notice.

American negotiating authority came from legislation delegating tariff power to the president for a fixed period, which is what made a multi-year round possible. Domestic institutional design determines what an executive can credibly negotiate abroad.

Non-tariff barriers are enacted by parliaments for domestic reasons, so disciplining them means constraining democratic legislatures. That tension, between regulation chosen at home and rules agreed abroad, is the origin of every later dispute about trade and democracy.

Agricultural protection is sustained by concentrated rural constituencies with disproportionate political weight in both Europe and the United States. The hardest problems in trade negotiation are always the ones where domestic politics is most resistant.

French farmers demonstrated in large numbers and the government maintained a position it could not abandon without serious domestic cost. Trade agreements have to be ratified at home, and negotiators bargain within limits set by constituencies they neither control nor can afford to ignore, which is why agriculture blocked the round for years.

The European side could accept limits on subsidised exports because its own reform of the common agricultural policy had already begun for budgetary reasons. Domestic reform undertaken for internal reasons created the space for an international agreement.

The single undertaking required developing countries to accept intellectual property and services obligations to keep textile and agricultural access. Whether that bargain was fair has been argued ever since, and the argument stalled the following round permanently.

Most-favoured-nation treatment, national treatment, binding of tariff concessions and a consultation procedure: four rules that turned bilateral bargains into a multilateral system. The design is elegant, and it has been extended rather than replaced for seventy-five years.

GATT survived as a provisional agreement administered by a secretariat that formally did not exist, applying rules to the extent they did not conflict with existing legislation. Institutional improvisation of that kind governed world trade for two generations.

Negotiating across the board by formula rather than product by product transformed what a round could achieve, cutting industrial tariffs by a third. A change in procedure produced a change in outcome, which is the recurring lesson of these negotiations.

The separate codes were optional, producing a system in which different members had different obligations. This à la carte design fragmented the regime, and its replacement by a single undertaking was the Uruguay Round's principal structural reform.

Punta del Este established a single undertaking, meaning nothing is agreed until everything is, which forced trade-offs across sectors. That procedural decision made the round enormous and made it possible, and it has since made further rounds unworkable.

Putting a single consolidated text on the table changes a negotiation from many bilateral disputes into one document to be amended. The single negotiating text is a well-established mediation technique, and its use here is why the round eventually closed.

The accord converted agricultural protection into tariffs that could then be reduced, and set separate disciplines for domestic support and export subsidies. That architecture is the basis of the agriculture agreement and remains in place.

Compulsory jurisdiction, an appellate body and authorised countermeasures made this the most legalised regime in international economic relations. Its subsequent paralysis, when appointments to the appellate body were blocked, showed how much depended on one member's consent.

States accepted binding limits on a core sovereign power, the setting of their own tariffs, in exchange for reciprocal limits by others. That exchange is the basic form of most international economic law, and it began at scale here.

The organisation international society designed for trade never existed, and its functions were performed by an agreement never intended to be permanent. The society often ends up with the institution it can ratify rather than the one it planned.

Developing countries participated with little influence, and the round's results largely concerned trade among industrial states. The society's economic institutions reflected the interests of their most active members, which produced the demands that led to UNCTAD.

Developing countries largely declined the codes and were granted what the system calls special and differential treatment, a formal recognition that identical obligations may not be appropriate for members at very different levels of development. The principle was conceded here and has been contested at every negotiation since.

Intellectual property brought patent and copyright standards into the trade system, requiring states to legislate domestically to a common standard. The reach of international economic governance extended dramatically, and the legitimacy of that reach has been contested since.

An international official proposed terms to sovereign states on his own authority, which they could have rejected outright and did not. That the members treated the draft as a legitimate basis says something about the standing the secretariat had accumulated.

The two largest traders settled a matter affecting every member and then sought endorsement, which is how great powers have always operated within multilateral institutions. Formal equality coexists with a practice in which some members decide first.

A permanent organisation with near-universal membership and binding adjudication is a considerable extension of what states will accept. That it was achieved for trade and nowhere else suggests where the society's tolerance for legal constraint actually lies.

Reciprocity and non-discrimination were established as the normal way to conduct trade, so that protection became an exception requiring justification. Reversing the default is the deepest change the system made, and it survived even where the rules were broken.

Havana's commitments on full employment and development were dropped, so the trade system was constructed around liberalisation alone. What the regime was understood to be for was narrowed by the failure, and the excluded subjects have been argued about ever since.

The round was named after a president assassinated before it began, and completing it carried a political meaning beyond the tariffs. Trade rounds acquire significance from their context, and this one was pursued partly as an act of continuity.

Trade began to be understood as extending beyond borders into how countries regulate themselves, which is a much larger claim. Once that understanding was accepted, the agenda expanded continuously and the political resistance grew with it.

Trade came to be understood as covering services, ideas and investment rather than goods crossing borders, an enormous expansion of the concept. Redefining what trade means was the round's most consequential move, and it happened at the launch rather than in the negotiation.

The Dunkel text became a focal point that everyone argued about, which is exactly what a stalled negotiation needs. Once a document exists, the question changes from what should be agreed to what should be changed, and that is a much easier question.

Agriculture had been treated as exceptional, a matter of food security and rural society rather than commerce, for the whole postwar period. Bringing it within the trade rules required reclassifying it, and the resistance was about that reclassification.

The organisation became the symbol of globalisation and the target of the protests that began at Seattle four years later. What the institution was understood to represent, far more than what its agreements said, determined the politics that followed.