All Commentary
Saturday, December 1, 2001

Wire and Rails: Comparing the Web and Railroads

The Internet Has Little in Common with the Transcontinental Railroads

Larry Schweikart teaches history at the University of Dayton.

Not long ago the television show Silicon Spin glumly reviewed the latest news of the cellular phone industry. The guests concluded that even if tech stocks, especially telecoms, had hit bottom, it would be 2003 before the experts thought the majority of them could again struggle back to profitability. Virtually all the problems, they concluded, stemmed from the industry’s rapid advance into broadband at a time that the market did not yet exist.

On the heels of the Silicon Spin analysis, the Wall Street Journal, focusing on a pair of telecoms called Qwest and Level 3 Communications, came to a similar judgment. The Web was “overbuilt,” resulting in a “fiber glut” that “underlies much of the uncertainty plaguing the telecom sector.”1 Claiming that only 2.6 percent of the capacity was actually in use, the Journal bleakly concluded that “much of [the fiber] may remain dark forever.”

What was interesting in both the Journal and the Silicon Spin analyses was the comparison with the railroads in the nineteenth century and the argument that both the Web and the railroads were “built ahead of demand.” It behooves investors—let alone anyone concerned with the future of high-tech America—to learn if telecoms indeed do have a similar historical pattern to that of the railroads, and if so, how they also differ.

America’s first railroad construction came through private financing, although it didn’t take long before the state governments got involved. Economic historians have argued that the capital demands of the railroads dwarfed those of any other industry. While that might have been true in the 1850s, it was not the case in the 1830s or 1840s. Founders of many early roads had simple objectives, often merely linking a single city to a river, or even building a shortcut between a long and difficult bend in a single river, as was the case in Alabama.

After the canal craze of the 1830s, in which many state governments subsidized construction by guaranteeing the canal companies’ bonds, the precedent was set for the states to support railroads. A number did so enthusiastically, especially in the south. Still, private capital dominated the construction of the best—and most viciously fought-over—roads. Cornelius Vanderbilt’s New York and Harlem Railroad, then later the Erie, attracted the attention of journalists and the general public, because of the titanic struggles between Vanderbilt and a bevy of opponents, including Jim Fisk, Jay Gould, and Daniel Drew.

Until that point, the extent of most governments’ involvement with the railroads took the form of legislators’ and aldermen’s taking bribes in return for granting, or withholding, various rights to cross certain territory. This, in turn, allowed those with substantial railroad stock to manipulate the market in railroad securities. Of course, for every buyer there must be a seller, and to the dismay of Fisk, Gould, and Drew, the Commodore often would not be (pardon the pun) railroaded: quite the contrary, Vanderbilt sent his own agents into the securities market to sell when his opponents tried to drive prices up, or to buy when they tried to drive them down. His massive transactions often disrupted the schemes of stock manipulators, and in the process he taught the silent partners in Albany lessons they never forgot.

By the Panic of 1857, most railroads relied on private investment. The depression of that year put many of them into bankruptcy, however. Ironically, while economists have for years thought that the origins of that panic lay in disruptions of the wheat market or foreign instability, it turns out that the Dred Scott decision—which overthrew the Missouri Compromise and opened up all territories to slavery—so terrified investors that the bonds of east-west roads collapsed. (Significantly, none of the roads running predominantly north and south collapsed, because their future business and traffic would have been relatively unaffected by the decision.)2

What would have become of the rail networks had the Civil War not intervened, of course, is difficult to determine. It is likely that, had slavery been prohibited from the territories per the Missouri Compromise post-1860, a normal construction program by private investors, largely absent state government intrusions, would have occurred. It is worth noting that the single most important business transformation in American history—the rise of the so-called “managerial hierarchies”—had already taken place in the private sector to address the capital needs of the railroads—a full decade before the first government-backed transcontinental railroad was launched.

Move Along, No Consumers Here

For many years, the nation had sought to support construction of a railroad to the Pacific for military reasons—largely to supply the forts on the frontier, but also to ensure quicker and more reliable support to California and Oregon. If there was agreement over the necessity of a transcontinental, there was disagreement over the route: southerners wanted a route running from Nashville or New Orleans, while northerners wanted a Chicago locus. It was Senator Stephen Douglas’s introduction of legislation to build a railroad through Nebraska that had touched off “Bloody Kansas,” and, eventually, the war itself. When the Civil War broke out, the Union government needed California’s gold and silver as well as the endless supply of horses and cattle provided by the frontier west. Supply lines were secured by the army, and Congress, without southern opposition, immediately passed the Pacific Railroad Act of 1862, which became a badly flawed blueprint for most of the transcontinental railroads. Under the Act, the government gave the railroad a substantial land grant (which was within its constitutional authority to do, and fit the Articles of Confederation’s provisions of the Land Ordinance of 1784 that reserved four sections of every township to the federal government).

Having the authority and using it wisely were two different things, however, and the free land proved as much a curse to the railroads as a blessing. But the other part of the Pacific Railroad Act involved a provision to give a subsidy to the construction of railroads in the form of United States bonds that the companies could sell on the market, then apply the proceeds to construction costs.

Under any circumstances, this was bad economics. The structure of the subsidies, though, proved even more short-sighted. For each mile of track constructed, the railroad received $16,000 worth of government bonds, rewarding the railroad for miles of track laid instead of actual services provided.

Despite what appear as lucrative inducements to modern Americans, few investors jumped at the opportunity to invest in a transcontinental railroad. People could do the math, and they concluded that it would be decades before such a project turned a genuine profit. There simply were not enough customers on the Great Plains to support the railroads. Congress thus sweetened the pot, doubling the land grant and allowing the railroads to sell their own bonds in addition to the government securities for construction. To make a long story short, this resulted in the infamous Credit Mobilier scandal.

After the government had sufficiently jump-started the Union Pacific and Central Pacific, another competitor subsidized by government land grants, the Northern Pacific, also entered the building frenzy. Whatever other feelings this must have evoked from Native Americans, the sight of these endless and obviously expensive tracks must have struck them as incredibly silly. The Indians, because they negotiated the Plains regularly, saw that even after the sodbusters arrived there were no people out there to speak of.

One railroad builder did not rely on government land grants or bond subsidies for his transcontinental railroad. James J. Hill, a Canadian who was blind in one eye, had started his own transcontinental by steadily marching across Minnesota. Hill recognized the simple economic fact that the other roads missed—largely because they were on the government dole—that there were no customers for the railroads to serve. To that end, Hill decided to create his own customer base, a strategy that would have remarkable implications for the telecom industry in the 21st century. He enticed future customers—in this case, farmers—to places where his railroad ran by offering them land. In a direct inversion of the practice of the other roads, namely getting land from the government, Hill bought land and gave it away! Hill also experimented with a variety of new wheat strains, cattle breeds, and agricultural advances that would keep his future customers profitable, in the process putting his railroad in the black. Burton Folsom’s The Myth of the Robber Barons has thoroughly dealt with the stark differences between the government-funded transcontinentals and the Great Northern. For our purposes, the key factor is that Hill did not “build ahead of demand,” but rather ensured through his own efforts that the demand would be there when he needed it. It was the ultimate essence of Say’s Law.

Back to the Future

Jump ahead now more than a century to the late 1990s and the exploding telecom, wireless, fiber-optic, and high-tech boom. Once again, there are concerns that businesses are building ahead of demand. But the comparison to railroads doesn’t hold water. The railroads—at least, all but the Great Northern—were subsidized by the government and had no incentive to develop a customer base.

Indeed, the government was the customer until the legislators finally took away the golden goose. By the 1880s enough people had moved into the Great Plains and onto the frontier to obscure what had happened: the government had constructed several roads that would not have been profitable if left to their own resources, or at least not using the strategies that operators employed with Uncle Sam.

Economists have beaten the railroad data to death, and while there is near unanimity on the fact that the railroads provided “social savings” in the neighborhood of 5 to 15 percent of GNP by 1890, these conclusions universally ignore the dynamics of a free market. “Social savings” is a term that postulates general benefits to everyone (lower travel costs and shipping prices) that might not have existed in the absence of the railroads. The best that economists have come up with to actually test these theories is an imaginary—but potentially real—system of interstate canals that could have been constructed in place of the railroads. Again, however, proving that no private investors would have built the railroads in the absence of subsidies is impossible. Moreover, it is unlikely none would have done so, given the history of not only Hill and Vanderbilt, but also of the intrastate railroad construction prior to 1860. In other words, absent the government, the railroad entrepreneurs in the 1800s knew their market and had a good grip on the size of the customer base.

For the telecom industry in the modern period, the pundits are solemnly nodding their heads in unison that the “fiber barons” do not know their industry or their customer base—that they “built ahead of demand.” But is that the case? The Wall Street Journal has even argued that, to paraphrase Senator Fritz Hollings, “There’s too much competin’ goin’ on out there.” Further, competitors came in, the Journal implied, because of the “easy availability of funding.” Aha! All those stupid capitalists were so anxious to lose their money that they marched into fiberworld like lemmings, dumping millions into dark cable, all because of the “easy availability” of capital.

One must pause, then, to consider the warnings of tech guru and prophet George Gilder, who has for a decade warned that the high-tech industry (and bandwidth/cable in particular) is underfunded! The key to Internet growth has been the venture capital firms, the top 20 of which, from 1974 to 1995, beat even Warren Buffett’s well-known Berkshire Hathaway returns of 32 percent per year. Following four years of George H. W. Bush and two years of Bill Clinton, the investment in technology firms began to wane.

Three things heated up the Internet economy after 1994, however. First, energy prices continued to remain low and as silicon chip companies gobbled up electricity at rapidly accelerating levels, this component of their price fell steadily. Second, Web browsers appeared in 1994, making the Internet genuinely commercial. Third, the GOP Congress lowered capital gains taxes to 20 percent, while the Fed kept the lid on inflation, with prices rising at a tolerable 2 percent in the 1990s. Low interest rates, low energy costs, and low taxes—you can’t beat that combination for economic growth, and the sector that is on the cutting edge will grow the fastest.

Magnify all of this by the “Law of the Telecosm,” which states that the value of a network grows by a square of the processing power of all the terminals attached to it, and “Gilder’s Law,” which postulates that bandwidth grows at least three times as fast as computing power, and the high-tech economy of the 1990s became a skyrocket. The United States rode this skyrocket, claiming some 80 percent of the Web domain names and dominating Internet traffic. At the same time, the cost of manufacturing the “nuts and bolts” of computer—chips—shrank to almost immeasurable proportions, falling at rates of nearly 70 percent a year and forcing the price of a bit down to a millionth of a cent. Although the Journal may question the claim that Internet bandwidth and traffic are doubling every few months, even its own position—that traffic is doubling more like once a year (based on a single AT&T researcher’s paper)—if true, is still an astonishing growth rate. Given this incredible level of expansion, is there anyone out there—aside from the Journal staff—who seriously wants to argue that the availability of venture capital has grown at exponential levels to match the rest of the Internet?

Low interest rates have not sparked a torrent of new venture capital. Nor has anticipation of the desperately needed, but modest, Bush tax cuts. Part of the bottleneck in venture capital has rested in the Byzantine FCC laws governing bandwidth and the monopoly positions of the Bells. Supposedly the 1996 Telecommunications Act required the Bells to unload their networks and free up local areas to competitors, and in return the Bells would receive the authority to re-enter long-distance data transmission. A burst of new “competitive local exchange carriers” (CLECs) suddenly appeared, only to find that the Act did not provide a timetable for or fines for not allowing the competitors into the loops. Then U.S. Representatives Billy Tauzin and John Dingell introduced legislation to allow the Bells into long-distance data transmission regardless.

Complaining that the “companies focused on the easy part of building a network [while] too little money went into [getting] into homes and offices,” or a phenomenon called the “last mile” connections, the WSJ nevertheless admitted that the culprit might be the Baby Bells, which owned most of the “last mile” connections, not venture capital. Where the WSJ sees plummeting prices for dark fiber and the tech stock leaders, however, analysts such as Gilder see nothing but opportunities and market-driven growth. Falling prices are to be expected and celebrated, Gilder would argue.

Internecine Debate

The proper course of action on the Baby Bells, for the time being, is the fulcrum of analysis, pitting free marketeers with different solutions and approaches to de-monopolizing against each other. On the one hand, the “pure” marketeers such as Gilder brushed aside the impact of anti-Bell legislation, arguing that it was irrelevant in the long run what the Bells did. The key, he maintained, was fiber, which “trumps both copper and the airwaves,”3 and cable companies, not the Bells, command the fiber. And since the important regulations were those governing cable, the Bell monopoly-related legislation was a sideshow: the real action was under the ground.

Other bona fide free marketeers saw the Tauzin/Dingell legislation as a crippling blow to the telecoms. William Lehr, an MIT economist, and James Glassman of the American Enterprise Institute constructed a stock index of the CLECs and found that their market capitalization plunged in direct proportion to the Tauzin/Dingell legislation’s progress.4 However, Gilder responded in an “open letter” to the new FCC chairman that these CLECs were merely “litigation shops” intent on forcing the Bells to share their copper.5 They were not, he maintained, interested in dynamic change.

Gilder’s solution is perhaps politically unfeasible, but eminently sensible: let the cable companies make hay out of their temporary monopoly status. Forget the Bells, and give them the profits from copper. As he correctly notes, no Internet advantage will last more than a fleeting moment. Quit trying to level the playing field. The market is the ultimate groundskeeper. Competition from broadband, DSL, and other technologies will quickly eliminate any advantages the Bells ever held.

One thing is certain. The Internet, and the glass fibers on which it should travel, has little in common with the transcontinental railroads. One was built by entrepreneurs, one by the government. One caused prices to plummet and services to spread; the other allowed companies to pool and price fix with Washington’s blessing. One still suffers from not enough venture capital, while the other nearly collapsed because of too much free money from the taxpayers. The most interesting result of the comparison of wires and rails is that to succeed, one must waste its resource, bandwidth, because bandwidth is potentially the cheapest commodity in the world, driving costs down. The other wasted material resources, because it was subsidized by the government, driving costs up. James J. Hill would have appreciated the differences.


  1. Rebecca Blumenstein, “Web Overbuilt: How the Fiber Barons Plunged the U.S. Into a Telecom Glut,” Wall Street Journal, June 18, 2001.
  2. Charles Calomiris and Larry Schweikart, “The Panic of 1857: Causes, Transmission, and Containment,” Journal of Economic History, December 1990, pp. 807–34.
  3. George Gilder, Telecosm: How Infinite Bandwidth Will Revolutionize Our World (New York: Free Press, 2000), p. 266.
  4. See James K. Glassman, “Look to Politics to Find Broadband’s Market Cap Shortfall,”, June 22, 2001.
  5. George Gilder and Brett Swanson, “The Broadband Economy Needs a Hero,” Wall Street Journal, February 23, 2001.

  • Larry Earl Schweikart is an American historian and professor of history at the University of Dayton. He is the author of more than a dozen books; his best known popular book is A Patriot's History of the United States.