The Fiber-Optic Network Bubble: Back to the Future

Nearly two years after the great bull market of the 1990s sank, the debris continues to wash ashore in the form of layoffs, bankruptcies and shuttered Web sites. Among the largest casualties were once-high-flying companies that aimed to construct and manage the information superhighway. The Web-hosting giant Exodus Communications, the DSL-service provider NorthPoint Communications and the ambitious Internet network company PSINet, whose sponsorship of Baltimore's gleaming football stadium has lasted far longer than the company itself, have all sought shelter in Chapter 11.

Companies that built and assembled global networks of fiber-optic cables have been hit particularly hard. Global Crossing, 360Networks and Level 3 Communications had business plans that were as ambitious as their names -- and loaded on the debt to match. They aimed at nothing less than girdling the globe with ultrathin fiber-optic cable, which possesses thousands of times the transmission capacity of copper wire.

The fiber-optic cable market, which was jolted to life in 1982 when MCI placed an order for 100,000 kilometers of Corning's "waveguides," achieved liftoff in the late 1990s. Spurred by the putative need for infinite bandwidth, the promise of interactive television and computing, and the exponential growth of the Internet, companies rushed to finance and build. With few limits set by regulators, American companies were free to expand as rapidly as they wished. In all, network builders in the United States plowed some $30 billion into 90 million miles of fiber-optic cable. The spending spree, in turn, lit a fire under the stocks of suppliers, including JDS Uniphase, Corning, Lucent and Nortel.

These companies succeeded in creating a national broadband infrastructure almost overnight -- one that is now used by nearly all public schools, universities and institutions, as well as millions of homes and businesses. But despite the enormous growth in Internet and data traffic, there simply weren't enough bits and bytes moving to keep all that newly laid fiber lit. In fact, on most days, just 5 percent of the nation's fiber-optic capacity is being used. Accordingly, wholesale prices for fiber-optic capacity have fallen at an annual rate of 60 percent -- far, far beyond the price cuts one would have expected to be driven by technological change and economies of scale. Not surprisingly, the stocks of Internet infrastructure companies that were largely financed with debt have been massacred.

This high-tech glass meltdown has inspired a great deal of finger pointing. And there's obviously plenty of blame to go around: credulous Wall Street analysts and bankers touted the securities; incompetent and/or arrogant entrepreneurs cast caution to the wind; optimistic technology gurus came to believe their own sound bites. Meanwhile, cable and telephone companies turned the process of linking residences and small businesses to the broadband networks -- what has come to be known as "the last mile" -- into a logistic nightmare. And federal regulators either intervened too much or too little -- depending on whom you ask.

For years now, the denizens of Silicon Valley's office parks, Wall Street's towers and the Federal Reserve's marble halls have been declaiming loudly about the way in which information technology, including these humongous fiber-optic networks, was forging a New Economy -- an economic nirvana that constituted a clean break from history. But when it comes to the construction of nationwide networks for commerce and information, the cycle we have just seen in fiber optics -- overbuilding, excess capacity, ruinous competition, falling prices, bankruptcy and consolidation -- is nothing new. For nearly two centuries, entrepreneurs and established players have set out to build networks that link portions of the country. And while the networks have survived and prospered, their creators have rarely fared well.

Indeed, similar infrastructure manias surrounded the telegraph in the 1850s, the railroads in the 1880s and 1890s, and telephony in the 1890s and 1900s. To be sure, the circumstances surrounding these bubbles differed. But their stories bear striking similarities.

• In each, entrepreneurs aimed first to link the business centers of the Northeast and then to stretch long lines across sparsely populated portions of the country.
• Each took place in a climate in which it was possible to raise large sums at low cost.
• Each benefited from the uniquely American approach of relying on private enterprise to manage infrastructure build-outs.
• In each, network builders were hampered by the lack of a common standard that would facilitate seamless connections among rival networks.
• Each created platforms for American businesses to grow in size and scope and was an important midwife to globalization.
• Finally, at the end of each cycle, a few well-endowed players reaped immense profits.

The Telegraph Boom And Bust 1847-1865

In 1842 Samuel Morse, a 51-year-old professor at New York University, strung wires between two Congressional committee rooms and sent messages back and forth. Impressed, the lawmakers granted him $30,000 to construct a telegraph line from Washington to Baltimore. Supplementing the appropriation with private funds, he completed the line. On May 24, 1844, the first telegraph message was sent from the Supreme Court chamber to Baltimore: "What hath God wrought."

Morse helped to form the Magnetic Telegraph Company in 1845, which intended to wire the New York-Washington corridor. But returns were slow in coming; in its first six months of operation, the company notched revenues of $413 and expenses of $3,284. Nor did Morse have the field to himself. Other inventors and entrepreneurs, using rival patents (or appropriating those belonging to Morse), began to build small networks to deliver "lightning," as the telegraph was dubbed.

The number of telegraph wire miles rose from 40 in 1846 to 2,000 in 1848 to 12,000 in 1850. But the early efforts to lash together Northeast business centers foundered on technological and commercial obstacles. Frequently, the "lightning" was reduced to the pace of snail mail, since many incipient networks refused to carry or transmit messages that originated on other lines. By mid-1846, one could send a message on Magnetic Telegraph's line from Washington to Jersey City. But then, as now, the last mile proved problematic. Telegraph messages transmitted to northern New Jersey were ferried by boat across the Hudson.

Despite such problems, hundreds of would-be communications titans invested in start-ups. Barriers to entry were comparatively low; stringing a mile of wire cost just $150. And by 1852, 23,000 miles of line were in use, with another 10,000 under construction. But much of the capacity was duplicative or built far ahead of demand.

In this era of what the historian Robert Luther Thompson called "methodless enthusiasm," few proved more methodless or enthusiastic than the Irish immigrant Henry O'Rielly. In 1845 O'Rielly, who used technology developed by the Scottish inventor Alexander Bain, strung a line from Lancaster to Harrisburg -- the start of the grandly named Atlantic & Ohio Telegraph Company. But revenues were appallingly low; in the week of Feb. 6, 1846, the line reaped $4.50. After three months, the copper wire was taken down and sold to pay some of O'Rielly's debts.

Undeterred, O'Rielly next set out to wire the sparsely populated territory along Lake Erie, a 400-mile stretch running from Buffalo to Detroit. O'Rielly's Lake Erie Telegraph Company immediately found itself in competition with the Erie & Michigan Telegraph Company, a concern aligned with the Morse interests. There was insufficient business to support the debt service of one such line, let alone two, and by 1854 both had failed.

Despite such stumbles, the telegraph soon came to be recognized as a crucial tool for American businesses. Newspapers, bankers, brokers, wholesalers and retailers became heavy users. And, to their delight, prices fell sharply on routes where there was more than one service provider. Before 1849, the pioneering New York & Boston line charged 50 cents to transmit messages of 10 words or less, and 3 cents for each additional word. But that year, two other lines were completed: the Boston & New York Printing Telegraph and the New York New England Telegraph, which was backed by the irrepressible O'Rielly. The price for Boston-New York traffic immediately fell to 2 cents a word, and then to a penny per word. Not surprisingly, the Boston & New York Printing Telegraph went bankrupt and was liquidated by 1852.

These and other bankruptcies provided opportunities for the relatively established players. In 1852, the telegraph pioneer Ezra Cornell snapped up the New York & Erie, which had been capitalized at $110,000 in 1849, for $7,000. With the disappearance of competition on key lines, the survivors were able to hold the line on prices. Magnetic Telegraph, which managed to pay dividends of just 2 percent in 1851, increased its dividend payouts to 9 percent in 1852, and to a hefty 13 percent in 1853 and 1854.

As the systems matured, the ability to transmit messages to more cities through a single network became a huge competitive advantage. Consequently, systems that used Morse patents began to consolidate. One such consolidator, the Rochester-based New York & Mississippi Valley Printing Telegraph Company, was formed in 1851 and quickly snapped up 11 lines in five states in the Midwest. In 1856, it changed its name to Western Union. The next year, Western Union banded together with five other large networks, agreeing to interchange traffic and to connect only with one another. Where the cartel members competed, they successfully divided business to avoid price wars.

In 1862, five million messages were transmitted over 32,000 miles of line. And now, those who aimed to get into the business did so in a big way. In 1864, promoters started United States Telegraph, which rapidly built and bought a 10,000-mile network in the Northeast and Midwest. In a process that is likely to resonate with contemporary investors, bullish Wall Street brokers claimed the stock would rise to $150 by July 1865 -- even as it was hemorrhaging cash. Instead, the share price fell from $98 in March to $20 in October. In March 1866, Western Union acquired United States Telegraph and dismissed two-thirds of its employees.

A few months later, Western Union took over the America Telegraph Company, which had unfortunately consolidated north-south connections on the eve of the Civil War. The domestic game was essentially over; Western Union commanded 37,380 miles of line and 2,250 telegraph offices.

Eventually, both technology and the business model caught up to the dream. Innovations of the 1860s and 1870s -- transatlantic cables, automatic telegraphy and capacity-expanding duplexing -- made the medium far more efficient. Western Union continued to dominate the telegraph, controlling 90 percent of the United States telegraph market in 1915. And while investors who had initially snapped up shares of telegraph lines like the Indiana, Ohio & Illinois were left with worthless scrip, Western Union shareholders collected dividends for generations.

The Railroad Boom And Bust, 1873-1894

On July 4, 1828, 90-year-old Charles Carroll, the last surviving signer of the Declaration of Independence, broke ground for America's first railroad, the Baltimore & Ohio. Between 1830 and 1860, a network slowly took shape between the Atlantic and the Mississippi, consisting mostly of small short lines like the B&O. In 1850, however, the federal government helped kick-start large-scale growth by making its first railroad land grant. The subsidy allowed the Illinois Central to complete a 705-mile north-south route parallel to the Mississippi River in 1856. And between 1850 and 1871 generous legislators made more than 170 million acres available to 80-odd railroads. (Nearly half of the projected lines were never built or completed, resulting in a forfeit of 35 million acres -- an area roughly the size of Florida.) By 1860, the United States was linked by some 30,600 miles of track.

As with the telegraph, early railroads were plagued by the lack of a universal standard, which prevented transfers of rolling stock. In 1861, seven different gauges were used in United States and Canada, and only about half of the 351 railroad companies in North America ran on what eventually came to be the standard gauge (56.5 inches). The three great rail systems coming into New York were largely blocked from through-connections with lines that terminated in rival cities. Shipping goods through New York to points beyond would have been like flying into Newark Airport, then being required to drive to La Guardia to make a connection.

In the three decades after the Civil War, a host of developments paved the way for a massive boom in coast-to-coast rail transit. Congress continued to extend both credit and free land to railroads that aspired to connect the coasts. Railroads began to conform to a standard gauge, and new switching technologies permitted faster and more efficient operations. Wall Street grew into a money center largely to finance railroad dreams. And the iron horse was endowed with the sort of transformative power that would later be ascribed to the Internet.

"With perhaps two exceptions", the North America Review noted in 1867, the railroad was "the most tremendous and far-reaching engine of social revolution which has ever either blessed or cursed the earth." The growth in railroads, it continued, would make the 19th century "different from all others -- a century of greater growth, of more rapid development."

Indeed, the newly reunited nation came to be gripped by railroad fever. Between 1870 and 1880, the investment in rails more than doubled to $5.4 billion. And between 1880 and 1890, it would double again.

As with the telegraph, excess capacity was the order of the day. Within a single generation after the Civil War, not one but four railroads crossed the continent. Not all of them made it to the Pacific intact. The Northern Pacific, capitalized at $100 million, completed just 500 miles before going bankrupt in 1873. It went into receivership, and Henry Villard took control for $10 million.

Burdened with debt, Gilded Age railroads faced a constant mandate to run at maximum capacity. And in an era in which multiple lines connected key city-pairs like New York and Chicago, there was constant pressure to reduce rates to build business. Many lines offered secret rebates to large customers -- a practice that infuriated less favored shippers. Still others formed regional cartels to allocate traffic and pool profits. And somewhat perversely, in the 1880s rail managers decided that the only way to ensure a continuing flow of traffic was "to construct new tracks or buy existing roads in order to form giant