Do Americans live today in a new Gilded Age, dominated by robber barons? It depends on what you mean by “gilded age” and “robber barons.” The real parallel to draw is between the generation that followed the Civil War and the present one. In both eras, the greatest problem hasn’t been big business as such, but rather a particular kind of big business—private infrastructure corporations that control chokepoints in the U.S. economy, like railroads in the first Gilded Age and online giants like Google, Amazon, Facebook, and Twitter in the second. The solution to the ills of our second Gilded Age is the same as the one for the first—not antitrust campaigns indiscriminately targeting all large and successful firms, but public regulation of a few essential commercial and social infrastructure firms.In popular memory, shaped by progressive historians in the first half of the 20th century, industrial capitalists wearing top hats and gold watch chains came to power during the Civil War, defeated the agrarian Confederates and tyrannized factory workers and small farmers until the New Deal of the 1930s. But this cartoon version of history confuses two different eras that followed the Civil War.In the first period, from 1865 until the 1890s, there were few big businesses in the U.S. other than the railroads. When Mark Twain and Charles Dudley Warner wrote their satire The Gilded Age: A Tale of Today (1873), the richest Americans were railroad infrastructure executives like Cornelius Vanderbilt, Leland Stanford, Collis P. Huntington, and Jay Gould, and financiers who invested in railroads like J.P. Morgan.In the late 1880s the only big manufacturing firm listed on the New York Stock Exchange was one related to the railroads, Pullman’s Palace Car Company. As late as 1881, the value of Carnegie Steel was only $5 million, at a time when 41 American railroads were valued at $15 million or more. In 1886, when John D. Rockefeller’s Standard Oil Company controlled 90% of the U.S. oil industry, then used mostly for illumination by kerosene lamps, Standard Oil’s value of $33 million was dwarfed by that of numerous individual railroads, including the New York Central Railroad. Only with the Great Merger wave of the 1890s and early 1900s, and subsequent waves of business growth and mergers, did giant manufacturing firms come to overshadow railroads, including many giants that are familiar today, from the Ford Motor Co. to General Electric.Firms in both manufacturing and infrastructure sectors benefit from scale. Manufacturers often need long production runs and high sale numbers to recoup the costs of initial investment in plants and benefit from increasing returns to scale and diminishing marginal costs of production. A car factory is expensive to build, but once built, if it can crank out one car it can crank out 10,000. By contrast, infrastructure grids benefit from network effects—a railroad that connects a hundred towns is more efficient and useful than a single railroad line between two towns.The phenomenon of increasing returns to scale in manufacturing makes it hard, but not impossible, for new entrants to break into manufacturing sectors dominated by existing big firms or their modern equivalents: firmlike supply chains with many suppliers orchestrated by “original equipment manufacturers” (OEMs) like Apple and Nike. Even so, outright monopolies are rare in manufacturing, where there tend to be a few large, rival, oligopolistic firms in sectors like automobiles, planes, and telephones.Automobile manufacturers are big companies, to be sure, but none has more than a small share of its particular market. In 2019, before the COVID-19 pandemic, Toyota had the highest global market share of all carmakers—a measly 10.2%, followed by Volkswagen (7.59%), Ford (5.59%) and Honda (5.46%).The global market for actual physical phones, as opposed to service networks like Apple’s, is also competitive. In the first quarter of 2021, Samsung led with 22% market share, followed by Apple with 17%, China’s Xiaomi with 14%, Oppo with 11%, and Vivo with 10%.Software firms, as opposed to online infrastructure firms, can be competitive as well. For example, Google’s office suite software is gaining on Microsoft’s still-dominant version.Monopoly is a much greater danger in infrastructure than in manufacturing or services because the benefits of scale for infrastructure firms come from network effects. Bigger networks are more efficient, and the most efficient would be a single network. Even libertarians like Milton Friedman recognized that in networks like electricity and water, free market competition would tend to produce a single “technical monopoly” (to use Friedman’s phrase).Like the phrase “gilded age,” the term “robber baron” is often misunderstood. Today “robber baron” is used loosely as a pejorative term for “capitalist” or “rich person.” But the metaphor of the robber baron itself alludes to economic predation by owners of chokepoints in transportation. The original robber barons were medieval German aristocrats with estates along the Rhine River who forced river-borne merchants and travelers to pay tolls in order to proceed.Like the railroad barons of the first Gilded Age a century and a half ago, most of today’s superrich tend to have made their money not by inventing flying cars or robots but by controlling important commercial networks and information: Jeff Bezos in retail, Mark Zuckerberg in social media, Michael Bloomberg in terminals disseminating stock market information. The few American celebrity billionaires who actually make physical things in the United States tend to be de facto government contractors, dependent on government tax subsidies or contracts for what they make: electric cars and batteries in the case of Elon Musk, and rockets in the case of Musk and Bezos. Both Musk and Bezos originally made their fortunes in the commercial infrastructure sectors of finance and retail, respectively.Pity the country whose richest and most influential people are infrastructure tycoons like Cornelius Vanderbilt in the 19th century, or Mexico’s Carlos Slim today. The largest part of Slim’s fortune is based on the privatization of Telmex, the Mexican telephone monopoly, during the 1990s. In 2019, Slim’s business empire accounted for 70% of the mobile phone line market in Mexico and 80% of the landline market. Slim is not an entrepreneur who invented a new product, like Thomas Edison, Henry Ford, or Steve Jobs. He was already rich, bought a privatized government utility, and raked in the fees from a captive consumer market.Largely as a result of the windfall from the privatization of the Telmex public utility, in April 2021 Slim was the 24th richest person in the world, according to the Bloomberg Billionaires Index. For a time starting in 2015, he was the largest shareholder of The New York Times.Many of today’s household-name American billionaires, including Zuckerberg and Bezos, are more like Carlos Slim than like Edison, Ford or Jobs. Instead of making profits by selling innovative goods or services, they rake in toll fees from the need of businesses and individuals to use the quasi-monopolistic platforms they control.Warren Buffett has explained his ideal business: “High pricing power, a monopoly.” In 2009, his firm Berkshire Hathaway purchased 77.4% of the shares of BNSF Railway, the largest purchase by Berkshire Hathaway to date.Essential infrastructure firms in transportation and telecommunications tend to have power over other businesses and consumers in ways that manufacturers can only envy. In the 1920s, Americans who did not like Ford Model Ts could buy Packards, Chryslers, Studebakers or, if they were adventurous, Stutz Bearcats. But in the 1880s, if you were a farmer or small manufacturer in the United States and the local railroad monopoly refused to ship your products or charged crippling fees, you were out of luck.Even if they are not operating in physical space in the way a railroad company does, firms that control chokepoints in commerce can act as de facto infrastructure monopolies or oligopolies. The search market, for example, may seem competitive. In 2019, according to Marketing Land, 44% of people went directly to Amazon to begin searching for products, while only 34% used Google, Bing, or Yahoo.But here’s the problem. Unlike in manufacturing or other markets, in practice even a moderately high market share for an infrastructure grid or online platform can function like an absolute monopoly. How many shoppers looking for coffee makers, after all, check Amazon—then check Google, then Bing, then Verizon Media, then write down and compare lists of the various results to see if any coffee makers were left out? Similarly, a coffee maker company that Amazon arbitrarily chooses not to show suffers through no fault of its own, even if it were listed on lesser-used sites.Today’s robber barons and their tollbooth enterprises wield vast and unchecked social and intellectual power, as well as market power. Amazon is far from monopolizing groceries, but it is a de facto infrastructure monopoly when it comes to book sales, and it uses that raw power to censor books that question the progressive party line for ideological rather than commercial reasons. For example, Amazon banned Ryan T. Anderson’s 2018 bestseller When Harry Became Sally: Responding to the Transgender Moment, and explained in a letter to Republican members of Congress: “As to your specific question about When Harry Became Sally, we have chosen not to sell books that frame LGBTQ+ identity as a mental illness.” Although there is a debate among psychiatrists about how to classify gender dysphoria, Amazon has decided to censor scientific and theological debate on issues where it threatens the party line of progressive Democratic orthodoxy. When you search Amazon today, When Harry Became Sally is nowhere to be found—but in truly Orwellian fashion, Amazon does sell Kelly Novak’s rebuttal to Anderson’s now-disappeared book, Let Harry Become Sally: Responding to the Anti-Transgender Moment. Amazon also sells Mein Kampf.There is a point at which a retailer ceases to be a firm in a marketplace and becomes, in effect, a marketplace itself, in which other vendors compete to sell their goods and services. A de facto public marketplace cannot complain when the government regulates it. Capitalist markets were, in their origins, public markets in medieval Western Europe created by kings or towns, and subject to public regulations and laws. For the public to regulate a firm that was once a retailer but has since metamorphosed into a de facto public marketplace is merely to reassert the authority of the sovereign state—like subjecting a private security company to stricter government regulation once it has expanded into a private army, even if it started out as three guards at a shopping mall.The crisis of the new Gilded Age, then, is not caused by big business as such, or even by competitive businesses owned by tech billionaires, as the Microsoft and Google software rivalry suggests. The threat in today’s second Gilded Age comes from the lack of public regulation of a small number of commercial and social infrastructure platforms and grids that vast numbers of other businesses and consumers must use, with no realistic alternatives.Fortunately, we know what to do. In the early 20th century, it looked as though the private electric infrastructure empires of tycoons like Samuel P. Insull would rival those of the railroad tycoons of an earlier generation. But in the modern United States there are no electricity kings or water barons, thanks to public utility regulations from the Progressive Era and the New Deal. Electricity and water are provided at regulated rates by private firms supervised by government commissions or, in some cases, direct provision by municipal governments. (I address my checks for my electric utility bills to the City of Austin, Texas). Similarly, apart from a few toll roads and highways, streets in the United States are free and paid for out of taxation, to the perpetual distress of libertarians who want to make you pay a toll to the shareholders and managers of a national or global corporation every time you get on a highway.Search engines like Google, online sales platforms like Amazon, and even Facebook, to the extent that they are essential in connecting consumers with advertised products, are in different ways successors to the Yellow Pages phone directory in the days of Ma Bell’s government-regulated monopoly. There is no need to recreate something like the old AT&T regulated telephone monopoly of the 20th century. Instead, the federal government should license any firm that functions in whole or in part as a search engine or general sales platform—but only in return for the firm’s agreement to obey a single set of industrywide rules and standards set by Congress or an executive agency under congressional, presidential, and judicial oversight.What those government-imposed rules on all government-licensed online infrastructure platforms should be can be debated. That’s what democracy is for. The government might make it illegal for platforms to promote their own product lines (the Yellow Pages did not advertise its own Yellow Pages lawn mowers and Yellow Pages bicycles). Political or ideological discrimination against users of a platform should be illegal, with narrowly defined exceptions for pornographers and the like. (Note to censorious, authoritarian progressives: The new, uniform government rules for online commerce and communications should not define all conservatives or libertarians as fascists for the purpose of banning them from public infrastructure utilities.)To name one example, the federal government—not individual firms—should write the standardized public terms of service for all internet firms that are allowed to access customers and do business in the United States. No longer would you click on a website and see a pop-up asking you to “Agree” or “Disagree” to giving up your right not to be tracked and not to have your browsing habits sold to other companies. A single standard contract specifying user rights and dispute resolution methods should be set by federal law, not imposed on users by vendors in the form of different but equally unintelligible pseudo-contracts.In short, essential online platforms that function as de facto monopolies should be highly regulated, low cost, low profit, or no profit public utilities like water and electricity. States and cities should have the authority to create their own low-cost municipal search engines and online retail platforms, paid for out of taxes, if they choose, like existing public water and electric utilities.Today, Consolidated Edison (Con Ed) is one of the largest private but publicly regulated electrical utility companies in the country. Yet you’re unlikely ever to have heard of its chairman, president, and CEO, Timothy P. Cawley. In 2020 Cawley’s total compensation, in cash and equity, was $2.8 million.In a well-governed American economy, the celebrity tech tycoons whose banal pronouncements, marital lives, and vanity trips into outer space are covered by the media, and whose firms increasingly determine what Americans can and cannot read, view, or listen to, would be CEOs of regulated utilities like Con Ed. They would be as relatively obscure, modestly paid (in comparison), and lacking in discretionary power over other companies and individuals as Timothy Cawley.As the example of Con Ed suggests, we Americans can find models for regulating today’s essential intangible infrastructure utilities in the way that we regulate railroads, electric, and water utilities. Far from harming American capitalism, turning today’s unregulated online infrastructure companies into privately owned or publicly owned regulated utilities would not harm firms in other sectors, and might even free most companies and consumers alike from excessive charges and arbitrary power.A mixed economy with a dynamic market sector does not necessarily produce infrastructure robber barons. In fact, industrial capitalism works best without them.