Market Integration in the Age of Economic Liberalism:

Egypt, 1850-1939

 

May 1999

Tarik M. Yousef

Department of Economics

Center for Contemporary Arab Studies

Georgetown University

Washington, D.C. 20431

 

 

I Introduction

Over the past quarter century, economic historians have made a concerted effort in examining the role of technological advances and institutional change in promoting the economic transformation of the North Atlantic economies in the late 18th and 19th century. With a few exceptions, the countries of the Southern Mediterranean Basin including the Middle East have yet to be fully integrated into this research agenda even though the engines of global economic integration in the late 19th century were operating with force in much of the region. The expansion of trade with Europe and North America necessitated investments in modernizing transport infrastructure which in turn required capital inflows from abroad. The commercialization of society which accompanied integration into the global economy often induced fundamental changes in the social and economic institutions governing society.

These dynamics were clearly present in the economic history of Egypt from the mid-19th century through the inter-war period. While becoming gradually integrated into the global economy, Egypt was increasingly governed by a liberal economic ideology that fostered the growth of free markets and narrowed the role of the state to building legal institutions and transport infrastructure for the market. Did Egypt’s age of economic liberalism coincide with the rise of unified market economy? Did the revolution in transport infrastructure and legal systems promote efficiency in local and regional trade? These questions remain largely unanswered despite their centrality to evaluating the achievements of Egypt’s experiment with economic liberalism for over a century prior to WWII. This paper presents empirical evidence on the integration of commodity markets during this unique period in Egyptian economic history.

II The Evolution of Egyptian Commodity Markets, 1800-1939

Organized commodity markets have operated in Egypt for a very long time, mainly in the form in the form of weekly gatherings for local trade in consumer goods. The importance and structure of these markets has undergone rapid change since the early 19th century in response to the needs and policies of the state and the openness of Egyptian markets to international trade. Accordingly, one could conveniently classify the historical development of commerce on the basis of the division of production and distribution activities between the state and the market. The past two centuries have witnessed the full spectrum of this division of labor. Beginning with Mohamed Ali’s (1802-42) centralization of domestic and external commerce in state monopolies, the balance gradually shifted in favor of the market especially during the period of British occupation until the post-WWII period when the state once again controlled production and marketing. The second half of the 19th century saw a dramatic decline in the domination of the state in trade and an increased commercialization of the country side. This process was sparked to a large extent by the rapid expansion of cotton and grain exports and the invasion of European imports into Egypt (Issawi 1947). The growing, processing, and marketing of cotton and other cash crops involved extensive marketing arrangements between rural peasants, interior merchants, city buyers and export agents in Cairo and Alexandria (Owen 1969). In particular, it necessitated a revolution in the transportation system that would give the state an active role in the creation and integration of domestic markets. The irrigation projects of the second half of the 19th century provided additional navigable waterways in the Delta and Nile Valley, a development that was exploited with the introduction of steamboats that sailed as farther south as the Sudan. But the greatest of transport developments took place with the introduction of the railway in 1851. Thereafter, the construction of rails expanded on an impressive scale; the length of tracks reached 1,338 kilometers in 1869 and tripled to 5,606 kilometers by 1939. The transportation revolution was simultaneously affecting the communications system. Between 1850 and 1880, over 6,000 kilometers of telegraph were built and a further 9,000 added by 1939. Finally, the mail service expanded with the addition of 7,000 post offices around the country.

The marketization of the economy received further stimulus with the introduction of reforms affecting the institutions of private property, administration and regulation, and foreign participation in the economy. With the Land Law of 1858, private ownership of land was instituted and peasants were assessed taxes directly, eliminating tax farmers and collective village responsibility for tax payments. The reform of the legal system gained momentum in the 1870s with the promulgation of Western-inspired legal systems affecting commerce and civil life. The incentives for foreign participation in the domestic multiplied. The suppression of internal customs, the elimination of the state monopoly on transportation and the de facto lifting of the ban on direct purchases by foreign merchants from native growers in the interior in the 1850s, secured the freedom of commerce and stimulated greater foreign presence in agricultural marketing and exports. The formalization of the Mixed Courts in 1876, with jurisdiction in cases involving foreign and native subjects, provided a considerable measure of judicial order along with the protection of property and enforcement of contracts under the newly passed laws. The increased stability of Egypt together with the high prices commanded by grains and cotton during Crimean War and the American Civil War, respectively, attracted greater foreign presence and investment in Egyptian agriculture, infrastructure, and commerce (Owen 1969). As a result, the 18th century monopolistic structure where rural markets primarily served as bulking centers for Cairo and Alexandria, was gradually replaced with one structure that fostered horizontal links between villages, cities and export centers (Larson 1985). The share of commerce in total employment increased from 1.3% in 1880 to 4.6% in 1907 and 7.5% in 1937. Finally, trading activities benefited from the gradual monetization of the economy and the eventual replacement of gold and silver coins with paper currency in the early 20th century.

Should the transportation revolution and institutional reforms of the late 19th century have facilitated the creation of an integrated national market economy? Surprisingly, with a few exceptions, economic historians by and large have been skeptical of the efficient working of Egyptian commodity markets, even while acknowledging the importance of the noted transportation revolution and legal reforms in stimulating the growth of a dynamic national economy. For example, Hansen (1993), citing a study by Levin and Martin (1910), conjectured that, with the exception of intermediate wholesalers and cotton markets, consumer markets were noncompetitive and inefficient. Similarly, nationalist writings in the 1930s tended to reflect structuralist thinking on the failure of the price mechanism in the traditional sector in developing economies (Schultz 1964). With no concrete empirical evidence to support these conflicting readings of the historical record, the debate on market integration, and its relevance for understanding the impact of the transportation improvements and institutional changes in late19th and early 20th century Egypt remains open.

III Geography, Markets and Commodities

The topography of Egypt is practically flat, offering no natural obstacles other than waterways to any form of transportation. Since ancient times, economic life and urban centers have been concentrated along banks of Nile in the Valley (Upper Egypt) and the Delta (Lower Egypt). Our sample of 34 cities spans the entire administrative and geographical map of Egypt: 13 are in the Delta, 17 are in the Nile Valley, and the remaining 4 are coastal cities on the Mediterranean and Red Seas. We provide a rich test of price integration in the early 20th century, using a high-frequency panel of prices for as large a basket of commodities as possible. We focus on the inter-war years, a period of steep price deflation in Egypt that followed the high inflation episode during the Great War. The sample of 26 essential commodities including 24 food items may be conveniently classified on the basis of their perishability as goods since it helps illuminate the important role of storability in price integration. The list of 11 perishable goods includes bread, butter, cheese, eggs, milk, wet beans, fish, chicken, mutton, veal and beef. The 15 nonperishable goods are wheat, maize, flour, dry beans, lentils, rice, onions, potatoes, coffee, sugar, oil, alcohol, soap, petroleum and fissikh. We observe retail prices on a quarterly basis for each market for the period 1921-39; thus, for every commodity we have a panel of 76 price observations across 34 markets.

Traditional tests of market integration have utilized simple regressions of spatial prices as the tool of analysis (Metzer 1974; Latham and Neal 1983). Consider the prices of a commodity k in two localities i and j, Pi and Pj. Market integration then is examined by running the OLS regression of the two prices, for each commodity and pair of cities, or markets may be pooled together and the equation estimated for each commodity. Estimates of the regression coefficient close to one are interpreted as support for the hypothesis of integration. However, the results using such an approach would be misleading if the price series contained trends, i.e., were nonstationary. Alternatively, we follow the approach adopted in the recent literature, by focusing on a panel of intercity price differentials: that is, for commodity k we consider , the price differential expressed as percentage difference in prices between cities i and j. With perfect arbitrage, the price differential should be zero. This is the essence of the law of one price since abstracting from trade barriers, taxation and transportation costs, prices across markets should be equal (Kindleberger 1989). Fortunately, our data should enable us to abstract from all these factors except transportation costs which we approximate by distance in rail kilometers.

IV Empirical Results

We examine the hypothesis of market integration using the panel of cities on a commodity-by-commodity basis and provide estimates of the rates of convergence to the law of one price. We do this by testing the data for the stationarity (mean reversion) of intercity price differentials, Qij,k,t (Frankel and Rose 1996). Economically speaking, we ask whether intercity price differentials fluctuate around and return to their hypothetical mean level of zero or if they follow a random walk, i.e., meander without a tendency to revert to this long-run mean. For nonperishables, we are able to reject the random walk null hypothesis for all but one commodity at the 95% level and all but two commodities at the 99% - in percentage terms, 95% and 85% of the time, respectively. Although less overwhelming, we find similar evidence for perishables in 8 out of 11 commodities at the 95% and 99% levels. Furthermore, we find that all commodities that are internationally traded, perishable and non perishable, exhibit the strongest mean reversion behavior; we interpret this as reflecting a general efficiency of domestic commerce in internationally traded goods and not just in cotton as Hansen (1991) has suggested.

We then augment the panel empirical specification with city-specific effects. In other words, we allow for the possibility that the true mean of Qij,k,t has a city specific component that is not zero. One justification for this is the possibility that trade in commodities may entail systematic or sunk costs that may vary across cities (Obstfeld and Taylor 1997). In historical and contemporary context of Egypt, such considerations are plausible given the wide variance in the size, development and accessibility of markets in the Delta and Nile Valley. The implications of the augmented specification are surprising given that, in previous studies, the inclusion of market effects tended to weaken results considerably (Frankel and Rose 1995). In contrast, we are able to reject the random walk hypothesis as often as previously and with wider margins of significance.

What about the rates of convergence to the law of one price (LOP)? That is, what is the length of time it takes for deviations from market integration to disappear? In time series or cross section studies, the typical half-life of deviations from LOP fall in the range of 3 to 5 years. In Parsely and Wei’s (1996) work with data from US cities, the median half-life for perishables and non perishables are, respectively, 3 and 5 quarters. In this study, the half lives are shorter for nonperishable commodities than perishables; they are shortest for those goods exported and imported. Without non-zero city means, the average and median half-life are 7.5 and 3.5 quarters for perishables and 4.1 and 3.5 for non perishables. With city-specific effects, the average and median are comparable across commodities and range between 0.4 and 0.6 quarters. In other words, if we allow markets to exhibit separate long-run equilibrium price differentials, deviations from these equilibria shrink by 50% within 2 months on average.

Finally, we examine the link between transport costs and the rates of convergence. To the extent that geography matters a great deal for market integration and conditions the rates of convergence, it could then be concluded that transportation costs explain the empirical failure of LOP. To address this issue, we re-estimated our equations after introducing distance separately and interacted with intercity price differentials. We pooled the data together for all commodities but allowed the two groups — perishables and nonperishables - to have separate slopes, i.e., different rates of convergence. We find that transport costs conditioned the rates of convergence; convergence to LOP is slower in cities that are farther apart. The half-life of deviations range between 0.5-2.5 quarters for perishables and 0.49-2.01 quarters for non-perishables. Could the high convergence rates in Egypt be justified by appeal to the transportation revolution alone? Our measure of transport costs, distance between cities in rail kilometers, overlooks the speed and flexibility gained with the introduction of railways (Landes 1969). Imagine the following counterfactual illustration: if we assume using our simple measure of transport costs that, in the absence of the transportation revolution, transport costs would have doubled - a reversal of what in fact took place between 1890 and 1910 — then the implied half-life of deviations from LOP would increase to the rates observed in cross-country empirical studies.

VI Conclusions

Although the empirical results in the present investigation of price integration in Egypt are still preliminary, a number of conclusions are evident. First, previous notions of segmented markets in local trade must be revised in light of the overwhelming evidence of market integration. Second, Egyptian commodity markets exhibit mean-reversion tendencies that compare very favorably with those obtained for the US in recent years and in cross-country studies of market integration. Third, the revolution in transportation and communications has contributed to the creation of a nationally integrated economy.

Endnotes

1 For example, the Egyptian Market Company was formed in 1898, a British firm given a concession to build new weekly markets and enclose old ones within walls. By 1938, there were 500 weekly markets in Egypt, all of which were privately owned.

2 Beginning in 1921, the Ministry of Finance sent out a monthly questionnaire to 355 civil servants and laborers, requesting information on the cost of living in their respective localities. The returns, which also included the costs of housing, utilities and transportation contained the prices of the 26 essential commodities including 24 food items. The ministry used the data to construct a monthly cost of living index for various urban regions in Egypt. The commodity list was assigned a 52% weight in the index and, hence, it provides a reliable indicator not only of inflationary trends, but also on the extent of price dispersion across markets. See the November 1921 issue of the Monthly Bulletin of Agricultural and Economic Statistics for details on the design and use of the monthly survey.

3 The regression results establish a strong link between transportation costs and intercity price differentials. Not surprisingly, price variability is related more strongly to distance in the case of non-perishable commodities, which were stored, transported and traded domestically and internationally.

4 The half-life is the length of time in quarters that it takes for deviations from LOP to shrink by 50%.

Select Bibliography

 

Abdel-Rahim Mustafa, Ahmed, "The Breakdown of the Monopoly System on Egypt after 1840," in Political and Social Change in Modern Egypt, P. Holt, ed. (London: Oxford University Press, 1968).

Baer, Gabriel, "Social Change in Egypt: 1800-1914," in Political and Social Change in Modern Egypt, P. Holt, ed. (London: Oxford University Press, 1968).

El-Sherbini A. and A. El-Sherif, "Marketing Problems in an Underdeveloped Economy," L’Egypte Contemporaine, July 1956, 42-46.

Engel, C. and J. Rogers, "How Wide Is the Border?" American Economic Review, December 1996.

Frankel, J. and A. Rose, "A Panel Project on Purchasing Power Parity: Mean Reversion Within and Between Countries," Journal of International Economics, 1996, 209-224.

Hansen, Bent, The Political Economy of Poverty, Equity, and Growth: Egypt and Turkey (Washington: Oxford University Press, 1991).

Hurd, J., "Railways and the Expansion of Markets in India," Explorations in Economic History, 1975, 263-88.

Issawi, Charles, An Economic History of the Middle East and North Africa (New York: Columbia University Press 1982).

Issawi, Charles, Egypt: An Economic and Social Analysis (London: Oxford University Press, 1947).

Kindleberger, Charles, Economic Laws and Economic History (New York: Cambridge University Press, 1989).

Landes, David, The Unbound Prometheus (Cambridge: Cambridge University Press, 1969).

Larson, B., "The Rural Marketing System of Egypt Over the Last Three Hundred Years," Comparative Studies in Society and History, 1985, 494-530.

Latham, A. and L. Neal, "The International Market in Rice and Wheat," Economic History Review, 1983, 260-79.

North, Douglass, Institutions, Institutional Change and Economic Performance (Cambridge,: Cambridge University Press, 1990).

Martin, G. and I. Levi, "Le marche egyptien et l’ utilite’ de la publication des mercuriales," L’Egypte Contemporaine, 1910.

Metzer, Jacob, "Railroad Development and Market Integration: The Case of Tsarist Russia," Journal of Economic History, 1974, 529-550.

Ministry of Finance, Monthly Bulletin of Agricultural and Economic Statistics, 1919-39. Cairo.

Parsley, D. and S. Wei, "Convergence to The Law of One Price Without Trade Barriers or Currency Fluctuations," Quarterly Journal of Economics, 1996, 1213-1236.

Obstfeld, M. and A. Taylor, "Nonlinear Aspects of Goods Market Arbitrage and Adjustment," Journal of Japanese and International Economics, 1997, 441-79.

Owen, E.R.J. (1969), Cotton and the Egyptian Economy, 1820-1914. Oxford: Oxford University Press.Rivlin, H., The Agricultural Policies of Mohamed Ali in Egypt (Cambridge, Harvard University Press, 1961).