Trade and Empire 1870-1914

"Partly as a response to interest by policymakers in the effects of the introduction of the Euro, economists have recently taken an interest in quantifying the gains from joining currency unions. Rose (2004) surveys the literature and concludes that, on average, joining a currency union boosts trade by between 30% and 90%. Similar benefits have also been reported for historical periods. Using a panel data set consisting largely of European based trade for five-year intervals between 1870-1910, Lopez-Cordova and Meissner (2001) find that two countries on the gold standard, on average, traded 60 percent more than countries not on the same monetary standard. Flandreau and Maurel (2001) and Estevadeordal, Frantz, and Taylor (2003) have also used historical data to examine the effects of currency unions. Recent scholarship has also seen a renewed interest in understanding whether empires facilitate or hinder economic development (Lal 2000; Ferguson 2004). Although it has long been known that political or military power can be used to acquire natural resources or raw materials, create overseas markets for exports, and expand the investment opportunities of home-country investors, economists and economic historians have begun to re-examine the institutional legacies such empires leave and the longer-run impact on economic development (Acemoglu, Johnson, and Robinson 2001, 2002, Engerman and Sokoloff 1997). Contrary to a long historical tradition that has viewed empire in a negative light, Ferguson (2003a, b) has suggested that the British Empire was instrumental in fostering the development of institutions conducive for economic growth. He argues that Britain effectively provided rule of law, secure rights of personal liberty and private property and stable, responsive, honest, and relatively efficient governance to its Empire. Among the purported benefits of empire to growth, he suggests that British rule provided a common-law system with protections for shareholders and creditors that encouraged capital formation. He also stresses that Britain acted as the world’s banker during the 19th century, and that it promoted and “enforced” free trade while other countries moved towards more protectionist policies towards the end of the century. This paper explicitly links these two disparate literatures by analyzing the effects of empire on bilateral trade flows during the gold standard era. In particular, we test two propositions that emerge from this earlier research: (1) membership in a currency union boosts bilateral trade and (2) “membership” in an empire boosts bilateral trade. It is plausible that both channels are important in explaining trade flows. Quantifying the effects of empire on trade has largely remained an underdeveloped literature. Although Estevadeordal, Frantz, and Taylor (2003) included a dummy variable for colonial dependency in their study on the rise and fall of trade between 1870-1939, they did not systematically analyze the role of empire on trade. Indeed, their gravity model of trade only considers one year’s worth of data from our sample period, 1913, and empire receives little attention in their discussion. Lopez-Cordova and Meissner (2001) focused on estimating the impact of the gold standard on trade, but did not consider the effects of empire or colonial dependence on trade. Given the resurgence in demand for studies seeking to quantify the effects of empire, it seems quite appropriate to consider one of the more obvious channels through which empire has an impact. Historians of the British Empire have, in fact, made trade a centerpiece of the discussion of the economic effects of that empire. At least since 1870, British tariffs were lower than its European neighbors and its rivals in empire (Bairoch 1989), and it managed to keep tariffs low throughout much of its empire, thus promoting trade. Some have even argued that free trade spread during this period more as a result of Britain’s power than by example (Ferguson, 2003b). Edelstein (2004, p.205) suggests that Britain’s enforcement of free trade during the gold standard era benefited the home country by as little as 1.8% and by as much as 6.5% of GDP. Trade benefits may have been even greater for its colonies or dependencies. The colonial share of British imports rose from a quarter to a third from the 1870s to the 1920s (Cain and Hopkins 2001, p.432), and trade continued to expand even when other countries turned protectionist in the face of the 19th-century factor price convergence, in part because colonial administrators were able to keep protectionism in check (Williamson 2000). We also consider whether other European imperial powers (French, German, and Dutch) had any measurable impact on trade. This will allow us to further test whether the British Empire was particularly good at nurturing trade (perhaps due to its laissez faire policies and its commitment to lower tariff rates during the period), as Ferguson (2003a,b) has emphasized, or whether, in fact, trade patterns with empires were driven less by any institutional advantage and more by dependency or coercion. We also test whether previous estimates of joining a currency union are overstated due to the omitted variable of empire. Since countries were linking up to gold at the same time empires were expanding, the two results may be confounded. We estimate the effects of empire and the gold standard on trade by employing an augmented gravity model of trade, where distance between trade partners and the size of their economies are key determinants of trade. We construct a new database of bilateral trade flows using data from the Annual Abstract of British Statistics for the period 1870 to 1913. This new database contains approximately 286 country pairs and has roughly 12,000 bilateral trade pairs. Moreover, it is much better suited for assessing the relative importance of these two channels than previous databases on trade during the gold standard (Barbieri 1996) because it contains a far greater number of observations of dependencies that were part of empires during this period. Increasing the country coverage is particularly important because previous studies of the gold standard (Lopez-Cordova and Meissner 2001) could not have hoped to sort out the confounding effects since nearly all the countries that were colonial dependents in their sample were also on the gold standard. In contrast, our database improves the identifying variation by including dependents of empires that were also on silver and paper standards."