How America turned electricity into a regulated monopoly, then spent thirty years taking it apart. A history of the U.S. power industry from 1879 to 2005, told through one question: at each stage, who owned what, and who got paid?
Nakul Nandhu23 October 202615 min read
On the afternoon of September 4, 1882, Thomas Edison's Pearl Street Station in lower Manhattan began supplying electricity to paying customers, the first commercial central power station in the United States (U.S. Census Bureau, 2026). About 400 lamps came on that day, for fewer than 90 customers, all inside a quarter of a square mile. Within a year the station was lighting about 10,000 lamps for 513 customers (Engineering and Technology History Wiki, n.d.).
Today the same basic idea, a power plant connected by wires to people who pay for what they use, runs the whole country. Between those 400 lamps and today, the industry was built into one shape, broken into another, and is now being rebuilt again. This article covers the first two. Part 2 covers the rebuild.
The short version
Electricity started as a free-for-all of competing companies and settled into a regulated monopoly: one company per territory, with prices set by the state.
That deal worked for about fifty years because bigger power plants kept making electricity cheaper.
When that stopped in the late 1960s and 1970s, costs exploded, nuclear plants ran far over budget, and regulators began making shareholders eat the losses.
Washington and about half the states then broke the monopoly apart, but stopped halfway after California's 2000 to 2001 crisis. That left two business models side by side, which is why a company like Vistra and a company like Southern Company are so different today.
Building the monopoly, 1879 to 1965Breaking it, 1965 to 2005
Figure 1. The turning points. Twelve moments that changed who owned the power system and how they got paid. Each is covered below.
Dates from the sources cited in the text.
01First, how electricity actually works
You do not need physics to follow this story, but four ideas will make everything after it make sense.
Almost every power plant is a machine for spinning a magnet.
Move a magnet past a coil of wire and current flows in the wire. Coal, gas and nuclear plants all boil water into steam that spins a turbine. A hydro plant uses falling water and a wind farm uses wind. The fuel changes; the spinning does not. Solar panels are the main exception.
Voltage is the trick that makes distance possible.
Think of electricity in a wire like water in a hose. Voltage is the pressure and current is how much water flows. The energy a wire wastes as heat grows with the square of the current, so to send a lot of power a long way, you raise the pressure and shrink the flow. A transformer does exactly that, and it only works with alternating current (AC). Generators produce power at roughly 2,300 to 24,000 volts. A transformer steps it up to 100,000 volts or more for long-distance transmission, substations step it back down to between 4,160 and 34,500 volts for local lines, and a final transformer brings it to the 120 or 240 volts in your house (PJM Interconnection, 2026).
Figure 2. From power plant to plug.Voltage ranges from PJM Interconnection (2026).
Supply has to equal demand every second.
For most of history, electricity could not be stored at any useful scale. If people use more power than plants produce, the grid becomes unstable and equipment trips offline. Someone always has to be watching the system and turning plants up and down. Keep this fact in mind. It is the reason the industry needs planners, reserve plants and, later, whole organizations just to run the grid.
Capacity is not the same as energy.
A plant's capacity, measured in kilowatts or megawatts, is how much it can produce at one moment. Energy, measured in kilowatt-hours or megawatt-hours, is how much it actually produces over time. A 1-megawatt plant running for one hour produces 1 megawatt-hour. Because demand rises and falls through the day, the system needs cheap plants that run all the time (baseload) and more expensive plants that switch on only at the peak (peakers). A plant that runs only a few hundred hours a year still has to be paid for, and every business model in this story has to solve that problem.
Figure 3. One day of demand. The cheapest plants fill the bottom of the curve around the clock. More expensive plants stack on top as demand rises toward the evening peak.
Illustrative shape, not measured data.
02The race to power America
1879 to 1900
Edison's system used direct current (DC), and in the 1880s DC had a fatal flaw: there was no practical way to change its voltage. Low-voltage power loses too much energy in the wires, so a DC station could only serve customers close by. Lighting a whole city meant building a power station in every neighborhood.
George Westinghouse bet on alternating current, built on Nikola Tesla's designs. With AC, transformers could raise the voltage for the journey and lower it at the other end, so one big plant could serve customers far away. Edison fought back hard, in what became known as the War of the Currents.
AC won in public. In 1893 Westinghouse powered the World's Columbian Exposition in Chicago with Tesla's AC system. The British physicist Lord Kelvin, who had opposed AC, was converted after seeing it there, and the international commission he led to choose a system for harnessing Niagara Falls gave Westinghouse the contract in October 1893 (PBS, n.d.; Niagara Parks, n.d.). The Niagara plant began operating in August 1895, and on November 16, 1896, its power reached the city of Buffalo (Niagara Parks, n.d.; PBS, n.d.).
The equipment business was consolidating at the same time. Edison's company merged with its rival Thomson-Houston in 1892 to form General Electric (Encyclopedia.com, n.d.). From then on, GE and Westinghouse were the two giants selling generators, turbines and transformers to everyone else, a role their descendants still play.
Why this matters for money: AC meant you could build one huge plant and run wires out from it. And a huge plant, as one man was about to prove, makes electricity far more cheaply than a small one.
03Insull and the monopoly bargain
1892 to 1929
Samuel Insull started as Edison's private secretary and in 1892 moved to Chicago to run Chicago Edison. He understood something about the economics of electricity that most of his peers missed.
A power plant is mostly a fixed cost. Once it is built, producing one more kilowatt-hour costs very little. So the cost of each unit depends mainly on how much of the time the plant is running. A giant plant that runs all day is far cheaper per unit than a small one that sits idle at night.
Insull built his business around that idea. He built huge central stations, cut prices to win customers, and deliberately mixed customers whose demand peaked at different times: factories during the day, homes in the evening, streetcars at rush hour. The more evenly demand was spread, the harder his plants worked and the cheaper each unit became, which let him cut prices again. In his first 42 months in Chicago he grew sales almost five times, buying out competitors as he went, and in 1907 he merged Chicago Edison with Commonwealth Electric to form Commonwealth Edison (Encyclopedia.com, n.d.).
There was a problem. In many cities, several companies strung competing wires down the same streets, and franchises from city councils were often bought with bribes. Insull's answer, first made in an 1898 speech to the industry's trade association, sounded strange coming from a businessman: he asked to be regulated (Public Utilities Fortnightly, n.d.).
The deal he proposed is still the foundation of the industry. A utility gets an exclusive territory, so nobody else can compete for its customers. In return, it must serve everyone in that territory, and a state commission sets its prices. On July 9, 1907, Wisconsin became the first state to give a commission power over utility rates, and Massachusetts and New York soon followed (Public Utilities Fortnightly, n.d.).
The utility gives up
The right to set its own prices
The right to pick its customers: it must serve everyone
Privacy: regulators audit its costs
The utility gets
An exclusive territory, with no competitors
A fair return on the money it invests
Customers who must pay for approved costs
Revenue requirement =Operating expenses + ( Rate base × Allowed rate of return )
Operating expenses include fuel, labor, depreciation and taxes. Rate base is the utility's long-lived investments, net of depreciation.
Figure 4. The regulatory bargain, and the formula behind it. Regulated utilities in the U.S. still earn money this way.
Formula and definitions from the CAREC Program (2021).
The formula is worth understanding exactly. The commission works out the utility's revenue requirement: the money it needs to cover its operating costs plus a fair return on its rate base (CAREC Program, 2021). Prices are then set so that total revenue equals that number.
Take a made-up example. A utility has $1 billion of power plants and wires in its rate base, and the commission allows a 10% return. The utility gets to earn $100 million a year on top of covering its costs. Now look at the incentive. If it builds a new $500 million plant and the commission approves it, the rate base grows to $1.5 billion and the allowed profit grows to $150 million. Under this model, a utility grows its profit by spending money on assets. That one fact explains a lot of the next hundred years.
By the 1920s the industry was also being stacked into holding-company pyramids: companies that owned companies that owned utilities, often with heavy debt at every layer. Insull built one of the biggest. By 1932, eight holding companies controlled almost three-quarters of the investor-owned utility business (Hirsh, as cited in Knowledge Problem, n.d.).
04Crash and the New Deal
1929 to 1945
The pyramids worked while money flowed in and collapsed when it stopped. After the 1929 crash, 53 utility holding companies with $1.7 billion of securities went into receivership or bankruptcy (Securities and Exchange Commission Historical Society, n.d.). Insull's empire fell in 1932, wiping out many small investors, and he became the face of the era's excess. By that year, three groups (J.P. Morgan's United Corporation, Insull's empire and Electric Bond and Share) controlled nearly half of all electricity generated in the U.S. (Securities and Exchange Commission Historical Society, n.d.).
Congress answered with the Public Utility Act of 1935. Its first part, the Public Utility Holding Company Act (PUHCA), gave the SEC power to break up the pyramids and simplify the companies that remained (Securities and Exchange Commission Historical Society, n.d.). The number of utility holding companies fell from 216 to 18 between 1938 and 1958 (Hirsh, as cited in Knowledge Problem, n.d.). Its second part created the Federal Power Act, giving the Federal Power Commission, the predecessor of today's FERC, authority over interstate transmission and wholesale sales of electricity (Congressional Research Service, 2017). That split still holds today: states regulate the price you pay, and the federal government regulates the wholesale market and the high-voltage grid.
The government also became a power producer. The Tennessee Valley Authority was created in 1933, and on May 11, 1935, President Roosevelt established the Rural Electrification Administration (NRECA, n.d.-a). Private utilities had largely ignored farms, because stringing miles of wire to a few customers did not pay. In the mid-1930s, nine out of ten rural homes had no electricity. The REA lent money to farmers' cooperatives so they could build their own lines, and by 1953 more than 90% of U.S. farms had power (NRECA, n.d.-a).
The result was three kinds of owners that still exist today: investor-owned utilities, public power (federal agencies and city-owned utilities) and member-owned cooperatives. Put all of this together and you get the first version of the value chain.
Diagram 1. The integrated monopoly, about 1945. One company does everything from power plant to customer, inside a protected territory, in exchange for regulated prices. The owner can be an investor-owned company, a public agency or a cooperative, but the shape is the same.
Based on the sources cited in Sections 3 and 4.
05The golden age, and the cracks
1945 to 1980
For about twenty years after the war, the bargain worked almost perfectly. Demand grew fast, utilities built bigger plants, and each new plant made power cheaper. Residential electricity sales nearly tripled during the 1950s, growing almost 11% a year (U.S. Energy Information Administration, 2026). Utilities became the classic "widows and orphans" stock: boring, regulated and paying a steady dividend.
Nuclear power looked like the next step. The Atomic Energy Act of 1954 allowed private companies to own reactors for the first time, and on December 18, 1957, the Shippingport plant near Pittsburgh began feeding power into the grid of Duquesne Light (American Heritage, 2007; ASME, n.d.).
Then three things went wrong at once.
First, the free lunch ran out. New plants stopped getting much more efficient than the last ones, so building bigger no longer made power cheaper.
Second, the 1970s arrived. The 1973 oil embargo, high inflation and high interest rates hit an industry that runs on borrowed money spent on giant construction projects. Demand growth also slowed sharply, from about 9% a year in the 1960s to about 4% in the 1970s (U.S. Energy Information Administration, 2026), so many plants under construction were no longer needed.
Figure 5. The end of the growth machine. Average annual growth in U.S. residential electricity sales. The golden age (blue) relied on fast, steady demand growth. After 1970 it never came back. Part 2 extends this chart to today.
Data: U.S. Energy Information Administration (2026), Monthly Energy Review, Table 7.6, residential sales to ultimate customers. Growth rates are the author's calculation (compound annual growth between start and end years).
Third, nuclear became a financial disaster. Reactor orders fell sharply after 1974 (Davis, 2011). The average cost overrun on a reactor approached 300%, and the Washington Public Power Supply System walked away from plants in the middle of construction, triggering the largest municipal bond default in U.S. history at that time (“Three Mile Island at 30,” 2009). On March 28, 1979, a reactor at Three Mile Island in Pennsylvania partially melted down. No one died, but the stricter rules that followed raised costs further, and new orders, which had already almost stopped, did not return for decades (American Heritage, 2004).
This is where the regulatory bargain first failed investors. The formula promises a return on "prudent" investment. When a plant came in years late and billions over budget, or was abandoned half-built, regulators started asking whether the spending had been prudent. Where they decided it had not, they refused to let the utility charge customers for it, and shareholders took the loss.
Meanwhile, Congress opened a crack in the monopoly from a different direction. The Public Utility Regulatory Policies Act of 1978 (PURPA) required utilities to buy power from certain independent producers, called qualifying facilities, at the utility's "avoided cost": what it would have cost the utility to produce that power itself. Before PURPA, only utilities could own and operate power plants (International Energy Agency, 2017). The law was meant to save energy after the oil shocks, but it proved something bigger. Companies that were not utilities could build and run power plants perfectly well.
06Deregulation, and where it stopped
1992 to 2005
By the 1990s, the argument for change went like this. The wires are a natural monopoly; it makes no sense to build two sets of power lines down the same street. But generation is not. If independent companies can build plants, let them compete, and let the market rather than a regulator decide which plants get built and what power costs.
The Energy Policy Act of 1992 created a new class of "exempt wholesale generators" that could sell power at wholesale without being regulated as utilities under PUHCA, and it expanded FERC's power to make utilities carry other companies' power over their lines (Watkiss & Smith, 1993). On April 24, 1996, FERC issued Order No. 888, requiring every utility that owned interstate transmission to offer open, non-discriminatory access to it (Federal Energy Regulatory Commission, n.d.). A utility could no longer keep competitors out by refusing to carry their power.
Competition needed a referee. Independent grid operators, known as ISOs and RTOs, took over running the grid and the wholesale market across large regions. FERC's Order No. 2000 in 1999 encouraged utilities to join one but did not require it, and Texas restructured its own grid, ERCOT, under a 1999 state law (POWER, n.d.). These organizations now serve about two-thirds of the country (Americans for a Clean Energy Grid, 2019).
Some states went further and restructured their utilities completely. Utilities were pushed to sell their power plants and keep only the wires, and customers could choose who supplied their electricity. The plants were bought by a new kind of company: the independent power producer, or merchant generator, which owns plants and sells their output into the market at whatever price it can get.
Then California showed what could go wrong. Its new market, opened in 1998, combined a fragile design with capped retail prices. In 2000 and 2001, drought, a shortage of new plants and deliberate manipulation by traders, including Enron, sent wholesale prices soaring. By December 2000, the price on California's power exchange had reached $377 per megawatt-hour, more than 12 times the $29.71 average a year earlier (NRECA, n.d.-b). Because retail prices were capped, utilities had to buy power at those prices and sell it for far less. PG&E, the state's largest utility, spent $9 billion on wholesale power in ten months and filed for bankruptcy on April 6, 2001 (NRECA, n.d.-b). The state suffered rolling blackouts and suspended retail choice that October (“The March Through the Crisis,” 2002). Enron itself filed for bankruptcy on December 2, 2001 (Calpine Corporation, 2004).
The crisis stopped the movement halfway. Today about thirteen states have fully restructured retail markets and six have suspended restructuring (National Governors Association, n.d.). Most of the Southeast and much of the West still run vertically integrated utilities. Georgia, where I study, is firmly in the traditional camp.
Full retail choiceTraditional, limited or suspendedGeorgia
Figure 6. The patchwork. States (plus D.C.) where households can choose their electricity supplier. Restructuring clustered in the Northeast, the Mid-Atlantic, Illinois, Ohio and Texas. Several other states allow choice only for some customers or suspended it after 2001; they are shown in grey.
Count from the National Governors Association (n.d.); state list from ElectricChoice (n.d.). Each square is one state, not drawn to scale.
The merchant generators had their own reckoning. Many had borrowed heavily to build gas plants during the boom, just as power prices fell. NRG filed for bankruptcy on May 14, 2003, and Mirant followed that July (Calpine Corporation, 2003). Calpine, then America's largest merchant power producer with 27,000 megawatts of capacity, filed on December 20, 2005 (ENR, 2006). Remember this for Part 2: when many companies build power plants at once on optimistic forecasts, the market can punish all of them at the same time.
Here is the value chain at the end of the "breaking" era. The single box from Diagram 1 still exists in half the country. In the other half it has been cut into separate businesses, each paid in a different way.
Power flowsMoney flows
Diagram 2. Same industry, two business models, about 2005. In regulated states one company still does everything for a regulated price. In restructured states, generators sell at a market price, a non-profit grid operator runs the market, a regulated wires company earns its allowed return through a delivery charge, and competitive retailers sell to customers.
Based on the sources cited in Sections 5 and 6.
07Why this still matters
The two models in Diagram 2 are not history. They are the two kinds of power companies you can buy shares in today, and you can trace each one back to a moment in this story.
Company today
Model
Where it comes from
Southern Company
Regulated, integrated
Owns Georgia Power. The Diagram 1 model, still intact across the Southeast.
Exelon
Regulated wires
Owns ComEd, the descendant of Insull's Commonwealth Edison. After it separated its power plants into Constellation, it kept only the wires (Constellation Energy Corporation, 2023).
GE's energy business. GE itself dates to the 1892 merger in Section 2.
Table 1. Today's power companies, traced back.
Vistra's story shows what it means to live on the market side. In 2007, private equity firms led by KKR and TPG, with Goldman Sachs, bought TXU, Texas's biggest power company, for about $45 billion, the largest leveraged buyout in history at the time. The bet was that power prices would keep rising. Instead, falling natural gas prices pulled power prices down, and the company, renamed Energy Future Holdings, filed for bankruptcy on April 29, 2014 (NPR, 2014). Its power plant and retail businesses emerged as a standalone company in October 2016 (Renewable Energy World, 2016) and soon renamed themselves Vistra Energy (Oil & Gas 360, 2016).
I used AI (Claude) as my research and drafting partner. The questions, the structure and the cuts are mine, and every figure is checked against the sources listed below. If something is wrong, that is on me, and I would like to know.
This article is for educational purposes only and is not investment advice. Figures are as reported by the cited sources on the dates shown. All views are my own.
