Trains and railroads have long captured the imagination of small boys. But the fantasies that locomotives engender do not necessarily end with childhood. Until the mid-1960s, an entire generation of economic historians, including Joseph Schumpeter, Leland Jenks, Chester Wright, Gilbert Fite, Jim Reese, Herman Krooss, August Bolbino, and others saw railroads as the engine of economic growth in America. For them the Iron Horse symbolized the dawn of America’s industrial age. Their belief was born of the notion that railroads reduced transportation costs, increased the demand for products produced by other industries, opened up western development, and encouraged eastern industrialization.

Railroads And Economic Take-off

The American economic historian, Professor W. W. Rostow of MIT, identified the Iron Horse as the “lead industry” that precipitated the “take-off” of American industrialization in the period 1843-1860. The Rostovian take-off thesis was so influential it formed the foundation of US foreign aid and development policies in the 1960s and 1970s under the Kennedy and Johnson administrations.

In the popular mind, and among policy makers, faith in engines of economic growth continues to hold sway. The Iron Horse may be replaced over time by other engines of growth, as technology and circumstances evolve, but the fundamental idea that economic development is spearheaded by one, or a few, pillar industries has maintained its political popularity. For example, in Hong Kong policy makers have continued to cherish these ideas.

Two ground-breaking studies of the railroad industry in 19th century America – one by Albert Fishlow (1965), the other by Robert Fogel (1964) – succeeded in overturning Rostow’s take-off thesis. Both studies were based on doctoral dissertations. Fishlow conducted his research at Harvard and subsequently served as Deputy Assistant Secretary of State for Inter-American Affairs from 1975 to 1976. Robert Fogel undertook his work at Johns Hopkins and later was awarded the Nobel Prize in Economics, in 1993, for his contributions to new quantitative economic history.

Fogel’s research represented a major milestone, not only for understanding the role of railroads in 19th century American economic growth, but much more importantly, for challenging the conventional wisdom that economic development depended upon engines of growth or pillar industries.

Fishlow’s work shows that most locomotives in the antebellum period burned wood, which meant that railroads used surprisingly little coal. As for iron, Fishlow demonstrates that railroads accounted for only 20% of net consumption in the 1850s. Twenty percent was significant, but hardly revolutionary. Nor did railroads single-handedly create the machinery industry. Fishlow argues that the production of locomotives created no strategic breakthroughs in steam engine design and production. Steamboats, in fact, demanded far more in the way of large, sophisticated engines.

Fogel examined Rostow’s thesis of railroads as an indispensable “leading industry” essential for American economic “take-off”. He addressed the question of what impact the railroads had on the change in the share distribution of output among the various industries. Fogel’s results indicate that railroads did not dominate the development of the iron industry in the two decades before the Civil War. Indeed their impact on that percentage distribution was minimal. In the case of iron railroads, it accounted for only 17% of the output change. In fact, his conclusions strongly support Douglass North’s assertion that, from the point of view of backward linkages, it would be as sensible to talk about an iron stove theory of the development of the iron industry as a railroad theory. For coal, it was less than 5%; for lumber, barely 5%; in the case of transport equipment only 25% (only half of the change accounted for by vehicles drawn by animals); and for machinery it was less than 1%. Thus, for all manufacturing, the railroads accounted for less than 3% of the change in share distribution.

Railroads simply did not generate backward linkages of great enough importance to have made a major contribution to the growth of supplying industries between 1843 and 1860. Rostow’s claim that antebellum railroads constituted a “leading sector” that induced widespread industrialization via backward linkages to coal, iron, and machinery was simply not substantiated by the data.

Social Savings From Railroads

Fogel further finds that the more general notion of the discontinuity of the growth process, inherent in the Rostovian take-off thesis, was absent from the American experience. He observes little evidence of any unique discontinuity during Rostow’s indicated period of 1843-60. In fact, equally dramatic structural changes in the period after the Civil War (1860-1865), and also in the 1820′s and 1830′s could be found. Rostow’s discontinuities in the period 1843-60 are an illusion. He was simply influenced by the impressions of the economic historians of his generation.

Another crucial issue addressed by both Fishlow and Fogel was an estimate of the social savings from railroads. In order to evaluate the economic contributions it was necessary to perform a cost-benefit analysis. This entailed estimating the prospective level of America’s GNP in the complete absence of railroads, that is, if they had never been built in the first place. Such a measurement lies at the heart of the economic concept of “opportunity costs.” The additional value of any innovation is the difference between the benefit achieved by adopting the innovation, versus that of using the next best alternative. For example, the benefit of travelling by car should be compared to the benefit of travelling by horse, bicycle, walking or whatever the next best alternative happens to be. To measure the additional benefit correctly one must take into account all the benefits and costs. So Fishlow and Fogel had also to estimate the social savings from railroads.

Fogel compares the 1890 cost of the shipment of agricultural products by some combination of rail, wagon, and water haulage with the estimated costs of shipping the same goods over the same trade pattern in the absence of railroads. This cost differential, which Fogel calls the social saving, is surprisingly small. Even when these estimates are expanded to include all commodities traded within the United States, the maximum value of total social saving is 7.1% of GNP, while the most “reasonable” estimates lie below 5%. Fogel concluded that with a less than 5% social savings estimate the level of per capita income achieved by 1890 would have been delayed by only three months, if railroads had never been invented. Railroads were therefore hardly “indispensible” to American progress because rather good alternative modes of transportation already were available. Fogel himself was surprised by such a low estimate.

Fishlow, on the other hand, estimated the social savings of railroads in 1859 to be 4% of GNP. Extrapolating to 1890, he produced social savings of at least 15% of GNP, far higher than Fogel’s estimated 5%. The key difference rested on the way in which each defined social savings. Fishlow estimated these by comparing railroads to the next best alternative actually available in the antebellum period. Fogel, on the other hand, calculated the social savings of railroads compared to a vast system of improved roads and canals that 19th century Americans might have built in the absence of railroads. Fogel, in essence, compared railroads to a “hypothetical counterfactual” economy; one that did not exist in 1890, but could reasonably have been expected to have occurred. Fishlow’s “counterfactual”, in effect, assumed that the American economy would not have made any improvements to its internal transportation network in the period 1860-1890 if railroads had not been built.

Agriculture the Biggest Winner

The Fishlow “counterfactual” is obviously false unless the economy stagnates. It could therefore be treated as providing an absolute upper-bound estimate of social savings. The Fogel “counterfactual” is obviously hypothetical and contains inevitable errors and biases. In constructing and estimating his hypothetical “counterfactual” Fogel took great care to bias it in favor of higher estimated social savings. His 7.1% of GNP is therefore an upward-biased high estimate and a more reasonable estimate is less than 5%. This latter number has survived numerous challenges in the past half century.

Fogel’s study, like Fishlow’s, concluded that, although cheap transportation made an important contribution to the development of agriculture in the North Central states, the combination of wagon and water transportation could have equally provided a relatively good substitute for the fabled Iron Horse.

In Fishlow’s account, Midwestern farmers and agricultural processing industries (such as flour milling) benefited the most from railroads. Railroads led to the creation of new farms and the growth of towns and cities that could market and process the growing surplus of grains, hogs, and cattle. Fishlow persuasively argued that these railroads were not built ahead of demand. Midwestern railroads, in fact, ran through densely populated areas. Profits that intensified development in locales best suited for commercial agriculture. Almost from the very beginning, these railroads made substantial profits, which one would not expect from developmental enterprises built ahead of demand. Private capital markets (with occasional help from local governments) financed most Midwestern railroads, thus confirming that investors expected these companies to make money sooner rather than later.

Fishlow’s argument has important implications for understanding the relationship between government policy and economic development. Since antebellum railroads generally made money, investment from the national or state governments was not important. To the extent that it occurred at all, government investment led to excess and wasteful construction. The American case showed that investment in railroads — an example of infrastructure capital — produced high social rates of return, but only in the context of a vibrant market economy.

Fishlow warned that underdeveloped nations — especially those wracked with large and unproductive agricultural sectors, illiteracy, concentrations of wealth, frequently wasteful government intervention — should avoid mechanistically investing in infrastructure capital to magically replicate the American experience. The generally poor record of large-scale infrastructure projects in many parts of Africa, Asia, and Latin America underscores the salience of Fishlow’s point.

China’s Rail Grows to 100,000 Km

At 91,000km, China has the third longest rail network in the world after those of the US and Russia, but rail traffic density is four times higher than the US and one-third higher than Russia. China handles one-quarter of the world’s total rail traffic on just 6% of the world’s total track length. Although freight trains remain the chief means of moving bulk commodities around the country, overcrowding means China’s railways meet only an estimated 35% of freight demand. Poor interconnections and crowded freight wagons mean it can take 10 days or more to transport freight the 2,000km between Chongqing and Shanghai.

To unclog this economic bottleneck the central government has grand plans to invest Rmb2,000bn (US$250bn) over a 15-year period, expanding the railway network to100,000km. The eternal problem with such grand plans is where the money will come from. Until the global financial tsunami of 2008, annual railway investment had averaged around Rmb50bn; fulfilling the planned targets will require, on average, more than double that – Rmb120bn – each year until around 2020.

Issuing debt is one solution, but a partial one at best. Investment from state insurance companies is another avenue. The National Council of Social Security Funds responsible for China’s national pension savings is also a prospective source of funds. To bridge the funding gap the government hopes for huge private investment. But so far efforts to attract these investors have failed dismally. Since 2002 the State Council has opened 72 rail projects to private investors, but take-up has been poor. The main obstacle is the prospect of dealing with Ministry of Railway (MOR), an ossified monopoly.

MOR, which controls 90 per cent of the rail network, is divided into 18 regional railway bureaus, which oversee day-to-day operations in different parts of the country and have sole responsibility within their individual jurisdictions. Fragmentation of authority among bureaus creates staggering inefficiencies: a train running from one end of the country to another must pass through numerous different rail administrations; and bureaucracy often insists that it must change locomotives when entering a new region. This contrasts greatly with railway operations of 19th century America that operated in a market economy.

MOR Keeps Most Lucrative Projects For Itself

In China, private investors can own pieces of the track but traffic management and scheduling remain in the hands of MOR’s railway bureaus. Tariffs are set by the National Development and Reform Commission (NDRC), which keeps passenger prices low for social reasons and freight tariffs down to prevent commodity price inflation.

Moreover, for all the talk of opening up to private investment, MOR keeps the most lucrative projects to itself. The Beijing-Shanghai high-speed line was not open to outside investors, while the Wuhan-Guangzhou route was taken off the table once MOR and the central government realized how profitable it could be. The route is now reserved for domestic institutional investors, including social security funds. So far the only real private investment in the railways has come from shareholders in listed railway lines.

Until China is willing to shake up the MOR monopoly and open its rail expansion to significant and bona fide private investors, investment in railroads is likely to fall well short of targets. By contrast American railways in the 19th century attracted global investors from Britain, the Netherlands, France and Germany. For them America then was the newest and largest “emerging” economy. Fishlow noted that these investors were able to make money sooner, rather than later, at that time.

Railroad’s Contribution Exaggerated

It is important to note that Fishlow’s and Fogel’s conclusion was not that a well-developed internal transportation network was not valuable to American development. Rather, the significance of the railroad invention has been greatly over exaggerated. Fogel’s work also indicated that there was no other industry that was likely to have been more important than the railroads; and thus, if not railroads, no other industry could have played the role attributed to the rails. This is an even more significant conclusion for it implies that economic progress entails the contribution of numerous, often insignificant, sectors; and that even the most impressive ones are but one of many rather than any “leading sector”. Fogel’s findings are consistent with the ideas of Jane Jacobs on recognizing diversity as the source of productivity. Robert Lucas reaffirms such a view and states: “But my own sense is that patents and ‘intellectual property’ more generally play a very modest role in the overall growth of production-related knowledge. I have sought a formulation that emphasizes individual contributions of large numbers of people, in which the role of market power is minimized.”


Albert Fishlow, American Railroads and the Transformation of the Ante-Bellum Economy, Harvard University Press, 1965.

Robert W. Fogel, Railroads and American Economic Growth, Johns Hopkins University Press, 1964.

W. W. Rostow, The Stages of Economic Growth: A Non-Communist Manifesto, Cambridge University Press, 1960

Share 分享到: 新浪微博   腾讯微博   人人网   FaceBook   Twitter   Google+  
Print Friendly

此文章还有以下语言版本:Chinese (Simplified), Chinese (Traditional)

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>