VoxEU Column Economic history International trade

The history of spaghetti

As WTO negotiations have failed, pragmatists have embraced preferential trade agreements. This column presents new evidence on the Anglo-French treaty of 1860 and argues that, contrary to all conventional wisdom, history’s most celebrated bilateral deal did not stimulate trade. That lesson should give pause to modern advocates of preferential trade agreements.

The late July 2008 failure of the ministerial meeting in Geneva to reach a multilateral agreement for further trade liberalisation in the Doha round has led some observers to question the adequacy of the WTO framework. The recent review of Jagdish Bhagwati’s book Termites in the Trading System by Alan Beattie, the Financial Times’ world trade editor, is an illustration.

Did multilateral trade talks die with Doha?

Alan Beattie suggests that Bhagwati’s charges against the spread of “spaghetti bowl” arrangements (all these bilateral or regional preferential trade agreements – or PTAs) while unobjectionable academically speaking, really lack practicality. He mentions that the world is in need of something more feasible than the behemoth deals that are at the heart of those failed Doha negotiations. He does not forecast any success for multilateral deals in the near future and concludes that we need something faster. In any case pointing to the shortcomings of PTAs has never prevented them from spreading.

The long history of “spaghetti bowl” trade agreements

True enough. In fact, historians would argue, “spaghetti” has a long history in the world of trade arrangements. Until recently, historians have seen bilateral treaties as a catalyst of, rather than an impediment to, trade liberalisation. Remember, this was the early 19th century, and the world was a gigantic field experiment for protectionism of all kinds.

Then came Sir Robert Peel and his remarkable political skills, social unrest, troubles in Ireland, and, of course, terrible weather. Then the scout of free trade – the repeal of the Corn Laws – came in 1846 and Britain became the sole guardian of the temple of sound economic principles. But protectionism was entrenched on the Continent. Britain was left to place its hopes in pedagogy and a bit of evangelisation.

People on the Continent were slow learners, and when they spoke free trade they stammered. Not much happened during the late 1840s and 1850s, nothing, in effect until the Anglo-French Treaty of 1860. This was an awakening.

The Cobden-Chevalier treaty

Negotiated secretly by Richard Cobden and Michel Chevalier in late 1859, its advent was announced in a letter by Napoléon III, published in the French official newspaper on 5 January 1860. Most importantly, the treaty contained the Most Favoured Nation clause.

Contagion ensued, with its characteristic sigmoid pattern, as countries rushed in to sign bilateral deals (Figure 1). As a result, the “Cobden-Chevalier” agreement has been conventionally portrayed as a milestone in 19th century liberalisation – in effect as the mother of all free trade. For instance, Bairoch’s work of reference argues that its signature opened the “phase of European free trade” (Bairoch 1989, p. 36) and more recent scholars have endorsed this view.

Figure 1. A free trade epidemic: Number of European MFN clause treaties as percentage of potential, 1855 - 1875

Source: Accominotti and Flandreau (2008)

Central in this notion is the role attributed to interaction between bilateral negotiations and the MFN clause, and the precise way they worked in the historical setting where they were implemented. Economic historians have suggested that the MFN clause worked to put the free-riding problem on its head: bilateralism cum MFN would have operated by encouraging further concessions from countries fearful of trade diversion and as a result, the MFN clause would have proved to be a remarkably efficient instrument that encouraged other countries to join and also receive MFN treatment. Bilateralism is really multilateralism lite, isn’t it?

Well, not quite. This we know from Ricardo, who emphasised that unilateral free trade is the best possible policy. But we know it as well from “RICardo”, the name for a large historical database that I have put together with colleagues that addresses such questions as the evolution of bilateral trade flows since the early 19th century.

The Cobden-Chavalier treaties did not stimulate trade

RICardo enables us to ask simple questions such as: did trade respond positively to the alleged massive stimuli provided by the Cobden-Chevalier treaty and subsequent network of Most-Favoured-Nation clauses? To our own amazement, the short answer to this question is: not a penny.

Despite torturing the data in many ways that are described in my article with Olivier Accominotti (Accominotti and Flandreau 2008) and in many more ways than we can print, we could not find a single effect of the Cobden-Chevalier agreement, or of any of its copycat treaties, on trade flows.

Unilateralism not bilateralism stimulated 19th century trade

This is because, we found, a lot of liberalisation had occurred before any treaty was signed (the way Ricardo recommended). And the Continent, on this account, was just like Britain (only, liberalisation received less publicity). What happened in 1860 was that policy makers, faced with growing resistance to further trade liberalisation, decided to change the instruments of liberalisation. They decided to push the agenda out of parliament, towards the diplomatic arena. If the goal was to get visibility, this was a stroke of genius. But if the objective was to foster trade expansion, they were very ill advised. For while the first round effect was indeed an outburst of trade treaties, the second round effect of trade creation was never to come.

The backlash at home

Instead, what they saw was, first, a backlash against earlier efforts as protectionists used political rivalries as a coordinating device. Second, once in place, MFN clauses turned out to have the perverse influence posited by theorists. They deterred countries from making additional concessions to one another, since these would be extended automatically across the board. In effect, the bilateral treaties of the 1860s and early 1870s were almost never followed by subsequent agreements aimed at deepening liberalisation.


All in all, I am inclined to argue that the bilateral, practical, approach advocated by the Cobden-Chevalier treaty really heralded a retreat from laissez-faire. If there is a lesson from history here, it is the following: Spaghetti is undoubtedly a delicacy. But better leave it to the restaurant.


Accominotti, Olivier and Marc Flandreau, 2008, “Bilateral Treaties and the Most-Favored-Nation Clause. The Myth of Trade Liberalization in the 19th Century”, World Politics, 60, 147-88. An early version available as CEPR DP 5423.

Bairoch, Paul, 1989, “European trade policy, 1815-1914”, in P. M. and S. Pollard (eds), The Cambridge Economic History of Europe, Volume VIII: The industrial economies: the development of economic and social policies, pp. 1-160, Cambridge: Cambridge University Press.

3,779 Reads