Too Cheap to Meter: The History and Economics of a Famous Promise

"Too cheap to meter" is a phrase for a good or service that costs so little to provide that it is easier to charge a flat fee for it, or give it away, than to measure how much each customer uses and bill them for that amount. In general use it describes a pricing situation. In the history of energy it means something narrower: it is the best-known line ever spoken about civilian nuclear power, and it has followed the nuclear industry for more than seventy years.

The line comes from a speech that Lewis L. Strauss, chairman of the United States Atomic Energy Commission (AEC), gave to the National Association of Science Writers on 16 September 1954. Strauss said that the children of his generation would "enjoy in their homes electrical energy too cheap to meter." The next day The New York Times built its headline around the phrase. Electricity from nuclear plants never became too cheap to meter, so the line turned into shorthand for official over-promising. Critics of nuclear power quote it as proof that the industry oversold itself from the start. Defenders have spent decades arguing that it was misread, misattributed, or about a different technology altogether.

This article covers what Strauss actually said and in what setting, how people at the time reacted, the long argument over whether he meant fission or fusion, why the economics of electricity made the promise unlikely from the beginning, how the phrase moved into telecommunications and technology, and what it does and does not tell us about energy prices today.

The phrase in one paragraph

A product is "too cheap to meter" when the cost of measuring and billing each unit of use is about as large as, or larger than, the cost of the units themselves. At that point a supplier gains little by tracking consumption. It can charge a fixed monthly price, bundle the product with something else, or give it away. The idea depends on two quantities: the marginal cost of supplying one more unit and the transaction cost of metering it. The phrase does not mean "free." It means that measuring use no longer pays for itself.

The 1954 speech

Setting and speaker

Lewis Strauss chaired the Atomic Energy Commission, the civilian agency that then held both the job of promoting atomic energy and the job of regulating it. In September 1954 American nuclear policy was changing fast. In December 1953 President Dwight Eisenhower had given his "Atoms for Peace" address to the United Nations, saying the United States would turn atomic technology toward peaceful uses. Congress was rewriting the Atomic Energy Act to open the field to private industry. The country's first full-scale nuclear power station, at Shippingport, Pennsylvania, was being planned and built.

Thomas Wellock, historian of the U.S. Nuclear Regulatory Commission (NRC), has gone through the drafts of Strauss's speech. He reports that staff drafts covered the new Atomic Energy Act, Atoms for Peace, the Shippingport plant, declassification, and the medical uses of isotopes made in reactors. The speech that resulted was mostly about fission: the discovery of nuclear fission, uses of fission products, and whether fission power could pay its way.

What Strauss said

Near the end of the speech Strauss looked ahead. He predicted rapid scientific progress, including "transmutation of the elements, unlimited power," and said the changes could come "all in 15 short years." He then described what the next generation might inherit. Their children, he said, would "enjoy in their homes electrical energy too cheap to meter," and would know of "great periodic regional famines" only as matters of history. He also forecast easy, safe, fast travel and a much longer human lifespan. He called this the forecast for "an age of peace."

Read in full, the passage is a broad utopian vision, not a narrow price forecast. The electricity line sits in a list with ending famine and slowing aging. That mattered later: defenders said it was obviously aspirational, while critics said the chairman of a federal agency had given an official stamp to an unrealistic hope.

The press and Meet the Press

The New York Times of 17 September 1954 ran the story on page 5 under a headline about abundant atomic power that would be "too cheap" for "our children" to meter. William L. Laurence, the paper's Pulitzer Prize-winning science reporter, was in the audience. He went on to use the phrase in articles and in a book, and he predicted that "all signs point to the realization within the next decade of a price for nuclear fuels so low" that only hydroelectric power could compete.

A few days later Strauss repeated the idea on the broadcast program Meet the Press. According to Wellock's account, he said his children and grandchildren would have power "too cheap to be metered, just as we have water today," and that the day might be "close at hand." The water comparison is worth noting. In much of the United States at that time, many households did not have water meters. They paid flat charges based on property characteristics. That is exactly the billing arrangement the phrase describes.

How people reacted in 1954

The speech is often remembered as the moment America fell for an atomic dream. The record from the time is more mixed.

Utilities were skeptical

The electric utilities that would actually have to build and run reactors kept their distance. Wellock reports that the president of Cleveland Electric Illuminating called the idea "a myth," because fuel was only a small part of what a customer paid for electricity. The trade journal Electrical World called it a "delusion," and joked that utilities would welcome more customers, each "with a meter."

Experts had already said nuclear power would be expensive

Several studies and statements before 1954 had concluded that nuclear electricity would not be cheap, at least at first. In his book Atomic Myths, Radioactive Realities (2004), the physicist and nuclear critic Arjun Makhijani quotes C. G. Suits, director of research at General Electric, who said in 1951 that atomic power would be "expensive power, not cheap power as the public has been led to believe." Makhijani also notes that in the 1950s uranium was thought to be scarce, so fuel was expected to be a real cost rather than a trivial one, and that reactors were already known to cost much more to build than coal boilers.

The AEC itself had told Congress that it expected the cost of reactor power to fall only to roughly the level of conventional sources, not far below it. A 2016 survey by M. J. Brown for the Canadian Nuclear Society gathered dozens of statements from the period. Most of them expected nuclear power to cost more than coal-fired power in the near term.

"Nobody took Strauss' statement very seriously"

James T. Ramey, who later served as an AEC commissioner, said, "Nobody took Strauss' statement very seriously." Wellock's conclusion is that the phrase predicted a fission-powered future within the foreseeable future, but that Strauss was speaking for himself, not for the AEC or the industry. Strauss had said elsewhere that a "serious governmental body" should avoid making predictions. As an individual, he seems to have felt free to make them.

The fusion-or-fission debate

The most lasting argument about the phrase is over which kind of nuclear power Strauss meant.

The fusion defense

In 1980 the Atomic Industrial Forum (AIF), a nuclear industry trade group, published a report on public understanding of nuclear energy. It argued that Strauss had been thinking of controlled thermonuclear fusion, not the fission reactors then being designed. The AIF pointed out that the AEC's fusion research program, Project Sherwood, was still classified in 1954. On this reading Strauss could not name the technology he had in mind, so he left the energy source vague. The AIF cited Strauss's son, Lewis H. Strauss, as well as former AEC staff and a Strauss biographer. The report also suggested that Strauss knew nuclear power would not pay for itself for many years, so his vision might only come true for later generations.

The argument came back in 2003. Donald Hintz, chairman of the Nuclear Energy Institute, told a conference that the industry had been "plagued since the early days by the unfortunate quote: 'Too cheap to meter'." He said the phrase had been taken out of context and that Strauss had been describing fusion.

Some supporters of the fusion reading also cited a line that industry "would have electrical power from atomic furnaces in five to fifteen years," which appears close to the "too cheap" remark in some quotation collections.

The case against

Wellock looked for evidence of the fusion reading in Strauss's papers at the Herbert Hoover Presidential Library and found none. He also points out several problems:

The speech as drafted and delivered was about fission. Its themes were fission discoveries, fission products, and the economic prospects of fission power.
The "five to fifteen years" line about atomic furnaces came from Dr. Lawrence Hafstad, head of the AEC's Reactor Development Division, who was speaking in Brussels. It concerned fission reactors.
On Meet the Press a few days later, Strauss answered questions about fission power without saying he had meant something else.
When Strauss did publicly reveal the fusion program in August 1955, he was much less optimistic. He said there was "nothing in the nature of breakthroughs" to suggest fusion was anything other than a "very long range" prospect.

None of this means Strauss had no interest in fusion. Historian Joan Lisa Bromberg, in Fusion: Science, Politics, and the Invention of a New Energy Source (1982), records that he saw fusion as the ultimate energy source and reportedly suggested a million-dollar prize for the first team to succeed. Still, the documents show that the public, the press, and the utilities all understood the speech to be about fission, and Strauss did not correct them at the time.

Why the debate matters

If the fusion reading were right, the phrase would say nothing about the economics of the fission plants that were actually built. If the fission reading is right, as the documents suggest, the phrase is a real if personal forecast about those plants, and it can fairly be compared with what happened. Most historians, including the NRC's own, now take the second view while also noting that Strauss's industry colleagues did not share his optimism.

A note on false attribution

The phrase is sometimes credited to Walter Marshall, a leading figure in British nuclear power. There is no documentary evidence that Marshall coined it or used it. The 1954 Strauss speech is the documented origin.

Why electricity was never going to be too cheap to meter

The utility executive who called the idea a "myth" had a simple point. Even free fuel would not make electricity free, because fuel is only one part of what a customer pays.

What is in an electricity bill

What a customer pays for electricity covers several different costs:

Cost component	What it pays for	Does cheap fuel remove it?
Fuel	Coal, gas, uranium, or nothing for wind and sunlight	Yes, this is the only part it removes
Generation capital	Building the power plant and paying back the money borrowed to build it	No
Generation operation	Staff, maintenance, security, insurance	No
Transmission	High-voltage lines and substations from plants to load centers	No
Distribution	Local poles, wires, transformers, and connections to each building	No
Customer service and billing	Meters, meter reading, billing systems, call centers	No, though it is the cost that "too cheap to meter" would eliminate

Makhijani's estimate makes the point in numbers. Even if nuclear fuel cost almost nothing, he wrote, about 85 percent of the cost of electricity for residential and small-business customers would remain, and about 60 percent for the largest industrial users. Most of what a household pays for goes to equipment and the network that delivers power, not to fuel.

Nuclear power's cost structure makes this worse

Among generating technologies, nuclear plants are an extreme case of high fixed cost and low fuel cost. A nuclear plant is expensive to build, slow to build, and must pay off a large loan before it earns a profit. Its fuel is a small part of its total cost. So the very feature that made nuclear power seem cheap, the enormous energy content of a small amount of uranium, says little about the price of its electricity. That price depends mostly on construction cost, financing cost, and how many years the plant runs.

Hypothetical worked example: when does metering pay?

The following example is invented to show the logic. It does not describe any real utility.

Suppose a utility serves 100,000 homes. Each home uses on average 900 kilowatt-hours a month. Assume:

The cost of generating and delivering one more kilowatt-hour (the marginal cost) is 5 cents.
Owning, reading, and billing a meter costs the utility 2 dollars per home per month.

The energy each home uses then costs 900 × 0.05 = 45 dollars a month. Metering costs 2 dollars, about 4 percent of that. Metering clearly pays: it lets the utility charge heavy users more and light users less, and it gives everyone a reason not to waste power. If the utility dropped meters and charged everyone the average, light users would subsidize heavy users and total use would probably rise.

Now change one number. Suppose the marginal cost fell to 0.1 cents per kilowatt-hour. A month of energy would cost 90 cents per home, less than the 2 dollars spent metering it. In this case the utility could save money by removing meters and charging a flat monthly fee to cover its fixed costs. That is the condition Strauss's phrase describes.

The example also shows the catch. A flat fee still has to cover the utility's fixed costs: the plant, the wires, the staff. Ending the meter does not make electricity free. It only changes how its cost is shared. And if the marginal cost is near zero only at some times of day, as with solar power at noon, meters become more useful, not less, because they let prices change with the hour.

Flat fees, two-part prices, and who pays

Economists who study utilities have a vocabulary for the choices hidden inside the phrase. A flat fee charges every customer the same amount regardless of use. A volumetric price charges per unit used, such as per kilowatt-hour or per gallon. A two-part tariff combines the two: a fixed monthly charge plus a price per unit. Most electricity bills in the United States are some form of two-part tariff, with a fixed customer charge and a per-kilowatt-hour charge.

Each design spreads costs differently.

A pure volumetric price puts most of the cost on heavy users. It encourages conservation, because each extra unit costs the customer money. But if the per-unit price is set above the real marginal cost, in order to recover fixed costs, it also discourages use that would actually be cheap to supply.
A pure flat fee spreads costs evenly. It is simple and cheap to administer, and it suits goods whose marginal cost really is tiny. But it gives no reason to conserve, so light users pay for heavy users.
A two-part tariff tries to get the best of both. The fixed charge covers the cost of the network, and the per-unit price can be set close to the true marginal cost, so customers face the right signal for each extra unit.

Seen this way, "too cheap to meter" describes the limit in which the per-unit part of a two-part tariff shrinks toward zero and the bill becomes almost all fixed charge. The phrase does not say the fixed part disappears. The plant and the wires still have to be paid for.

Hypothetical worked example: how a flat fee changes behavior

This second example is also invented, to show why suppliers are cautious about removing meters even when per-unit costs are low.

Imagine a small town water system that charges 1 cent per gallon. A typical household uses 6,000 gallons a month and pays 60 dollars. The town decides metering is a nuisance and switches to a flat 60 dollars a month for everyone. Nothing has changed except that an extra gallon now costs the household nothing.

Some households start watering lawns more and fixing leaks less. Suppose average use rises by a fifth, to 7,200 gallons. The town must now pump, treat, and deliver 20 percent more water, which over time may mean bigger pumps and a larger treatment plant. Its costs rise, so the flat fee has to rise for everyone, including the careful households whose own use did not change. If the town is in a dry region, it may also run short in summer, with no price signal to tell anyone to cut back.

The same pattern appears in the AOL story described later in this article. It is the main reason most utilities, even those with low marginal costs, keep some form of measurement.

What nuclear power actually cost

Shippingport

The first full-scale nuclear power station in the United States, the Shippingport Atomic Power Station on the Ohio River in Pennsylvania, reached criticality on 2 December 1957 and first fed electricity into the grid on 18 December 1957. Eisenhower formally dedicated it on 26 May 1958. Shippingport was a government-backed demonstration project. Some historians note that its design requirements made it costly, and some sources give the title of first fully commercial pressurized-water reactor to Yankee Rowe in Massachusetts instead.

The long record

Over the following decades nuclear power became a large part of the electricity supply in the United States, France, and elsewhere. But it did not become notably cheaper than its competitors, and in many periods it was more expensive to build. The industry and its critics argue about why. Explanations include rising safety requirements after accidents, the lack of a standard reactor design, the long time it takes to build plants, the high cost of borrowing during construction, and changes in regulation. When the AEC's promotional and regulatory roles were split in the mid-1970s and the Nuclear Regulatory Commission was created, the political setting for nuclear power changed as well. Whether regulation raised costs because it was needed or because it was excessive is still disputed, and good evidence exists on both sides.

Vogtle 3 and 4

The most recent large American nuclear project shows the cost problem clearly. Units 3 and 4 at the Vogtle Electric Generating Plant in Georgia were originally expected to cost about 14 billion dollars and to start operating in 2016 and 2017. The project suffered years of delays, including the 2017 bankruptcy of Westinghouse, the reactor's designer. Unit 3 entered commercial service on 31 July 2023 and Unit 4 on 29 April 2024. Reported total costs vary by source and by which costs are counted, from "more than 30 billion dollars" to around 34 billion dollars or higher. By any of these figures the project cost more than twice the original estimate.

Comparing today's generating costs

Analysts often compare generating technologies with the levelized cost of energy (LCOE). LCOE spreads a plant's lifetime building, financing, fuel, and operating costs over all the electricity it is expected to produce. The investment bank Lazard publishes a widely cited annual LCOE analysis. Its 2024 edition (version 17) gave these unsubsidized ranges in U.S. dollars per megawatt-hour:

Technology	Unsubsidized LCOE range (Lazard 2024)
Onshore wind	about $27 to $73
Utility-scale solar	about $29 to $92
New nuclear	about $142 to $222

LCOE has well-known limits. It does not capture the value of power available at any hour, the cost of backing up variable sources, or the cost of extending existing plants, which is often much lower than building new ones. Still, on this common measure, new nuclear electricity in the United States is far from too cheap to meter. It is one of the more expensive options.

The phrase outside nuclear power

The phrase has had a second life as a general description of business models built on abundance.

Water

When Strauss compared future electricity to water in 1954, he was describing how much of the United States bought water. New York City is a well-documented case. Before 1985 the city did not require water meters on most buildings. Owners paid charges based on things like the size of the building and the number of fixtures. In the mid-1980s the city began requiring meters, and building owners objected, as The New York Times reported in 1985. The city later carried out a Universal Water Metering Program, which a 2006 New York State audit reviewed. A main reason for metering was conservation: if water is not measured, people have little reason to use less.

The water case shows the main problem with flat pricing. A good can be cheap to supply in normal times and still be scarce in a drought. Once a good is not metered, the price no longer tells anyone when it is running short.

Dial-up internet and the AOL rush

Early online services billed by the hour. On 2 December 1996 America Online (AOL) moved most of its customers to a flat 19.95 dollars a month for unlimited use. Microsoft Network had announced a similar flat rate the month before. The result shows what happens when metering stops. AOL customers logged a record 102 million hours online in December 1996, up from 45 million in September. Its network could not keep up. Customers reported constant busy signals and "all circuits busy" messages, and under pressure from state attorneys general AOL agreed to give refunds to subscribers who could not connect.

The lesson is not that flat pricing failed. Flat-rate home internet became the norm. The lesson is that removing the meter changes how people behave, and the supplier must build enough capacity for the new level of use.

Broadband and mobile data

Home broadband in many countries is still sold at a flat monthly price, often with no hard limit on data. In 2016 the market research firm ABI Research published a report titled 5G: Too Cheap to Meter?, asking whether the next generation of mobile networks would make mobile data cheap enough to sell the same way. Mobile carriers have since offered many "unlimited" plans, though these often include speed reductions after a certain amount of use, which is a form of metering under another name.

"Tech is too cheap to meter"

In June 2009 Chris Anderson, then editor of Wired, published an essay with the title "Tech Is Too Cheap to Meter," drawn from his book Free: The Future of a Radical Price. His argument was that the costs of computing power, storage, and bandwidth had fallen so far that companies should design products around abundance rather than scarcity. Many online services are free to the user and paid for by advertising or by selling premium features to a minority of customers.

Anderson's argument works where the marginal cost really is close to zero, such as copying a digital file. It works less well where large fixed costs, such as data centers and network cables, still have to be paid. Those businesses still need revenue. They just collect it in ways other than charging per unit.

Electricity markets now: brief moments of near-zero prices

There is an odd modern echo of Strauss's promise. In electricity grids with a lot of solar and wind power, wholesale prices sometimes fall to zero or below during sunny or windy hours, when supply is more than demand and some generators would rather pay to keep running than shut down. The analytics firm Modo Energy reported that the California grid operator CAISO recorded 712 hours of negative wholesale prices in 2024, more than three times as many as in 2023, mostly because of growing solar output and flat demand.

This is not electricity too cheap to meter in Strauss's sense, for three reasons:

Wholesale is not retail. The retail price a household pays includes transmission, distribution, and other charges that do not fall when wholesale energy is cheap.
Cheap hours alternate with expensive hours. The same grid that has negative prices at noon may have high prices in the evening, when the sun sets and demand peaks.
Low prices are a reason to meter more, not less. When the value of electricity changes by the hour, meters that record use by time are what let customers benefit from cheap hours, for example by charging an electric car at midday.

Strauss imagined a future where power would be so plentiful that nobody would need to keep track. The grid that is actually developing is one where energy is sometimes abundant and sometimes scarce, and measuring it by time is more valuable than ever.

Why the phrase has lasted

It is easy to quote

The phrase is short, striking, and easy to repeat. It sums up a whole argument in four words. Opponents of nuclear power use it to show early promoters' overconfidence, and the economist Steve Cohn used it as the title of his 1997 book, Too Cheap to Meter: An Economic and Philosophical Analysis of the Nuclear Dream.

It stands for a larger pattern

The phrase also stands for a recurring pattern in technology forecasting: an early technology is judged by its most exciting physical property, here the energy in a small amount of fuel, rather than by its full cost to build and run at scale. Similar patterns can be seen in early predictions about many technologies, from supersonic air travel to fusion. A useful question to ask of any forecast like this is: "Which costs does the new technology remove, and which does it leave in place?"

Applied to Strauss's speech, that question gives a clear answer. Nuclear fission did promise to shrink one cost, fuel, to a small fraction of its former size. It did nothing to reduce the cost of building plants, and it added new costs of its own: radiation shielding, safety systems, waste handling, security, and decommissioning. It also did nothing to reduce the cost of transmission lines, local distribution, or customer service. A forecast built on the first point and silent on the others was bound to be too optimistic.

Two more questions help when reading any bold price forecast:

What scale is the forecast assuming? Early technologies are often judged by what they might cost once built in large numbers with a mature supply chain. That may take decades, and it may never happen if demand or politics change.
Who is making the forecast, and in what role? Strauss was a promoter of atomic energy as well as its regulator. Promotional settings, such as a speech to science writers, reward vivid language. Technical reports written for budget decisions tend to be more cautious, and the cautious reports of the early 1950s turned out to be closer to the truth.

It is fair and unfair at the same time

Wellock's view is that both critics and defenders are partly right. Critics are right that the head of the government's atomic agency made a public forecast about fission power that did not come true. Defenders are right that he was speaking for himself, that many experts and utility executives disagreed at the time, and that the industry never formally adopted the claim. Using the phrase to suggest that everyone in the 1950s believed nuclear power would be nearly free misreads the history. Using it to show how a confident official forecast can shape public expectations reads it correctly.

Common misconceptions

"Too cheap to meter" was the official position of the nuclear industry. It was not. It came from one speech by the AEC chairman. Utility leaders called it a myth and a delusion at the time, and a later AEC commissioner said nobody took it seriously.

Strauss was definitely talking about fusion. No contemporary document supports this. The speech was about fission, the press reported it that way, and the NRC's historian found no evidence in Strauss's papers that fusion was his hidden subject. The fusion reading was first put forward publicly in 1980, more than two decades later.

The phrase meant electricity would be free. Strictly, it meant that electricity would be cheap enough that measuring each customer's use would no longer be worth the cost. A flat charge would still be needed to pay for the plants and wires.

Free fuel would mean free electricity. Most of the cost of delivered electricity is equipment and networks, not fuel. Makhijani estimated that even free fuel would leave about 85 percent of the cost for small customers.

Walter Marshall coined the phrase. There is no documentary evidence of this. The documented source is Strauss in 1954.

Negative wholesale prices mean the promise has finally come true. Brief negative wholesale prices are a sign of mismatched supply and demand, not of permanently free power. Retail bills still include network costs, and the same grids have expensive hours.

How good is the evidence?

The basic facts about the speech are well documented. The NRC holds a copy of Strauss's prepared remarks dated 16 September 1954, and the New York Times article of 17 September 1954 is in the paper's archive. The Meet the Press remark and the reactions of utility leaders are recorded in Wellock's history for the NRC, which draws on archival research.

The interpretation of Strauss's intentions is less certain, because no one can read a speaker's mind. The fusion claim rests on later statements by family members and former colleagues, collected by an industry group in 1980. The fission reading rests on the content of the speech, its drafts, how listeners understood it at the time, and the lack of supporting evidence in Strauss's papers. The documentary evidence clearly favors the fission reading, but historians usually describe it as the best-supported reading rather than as settled fact.

Cost figures for modern projects such as Vogtle vary with the source and with which costs are counted, such as financing and the shares of different owners. Lazard's LCOE estimates are a respected industry benchmark, but they rest on assumptions about interest rates, plant lifetimes, and how often plants run, so they should be taken as ranges, not exact prices.

Frequently asked questions

Who said "too cheap to meter"? Lewis L. Strauss, chairman of the U.S. Atomic Energy Commission, in a speech to the National Association of Science Writers on 16 September 1954.

What exactly did he say? He said the children of his generation would "enjoy in their homes electrical energy too cheap to meter," as part of a broader vision that also included the end of great regional famines and longer lives.

Did he say it more than once? Yes. A few days later on Meet the Press he said future generations would have power "too cheap to be metered, just as we have water today."

Was he talking about fusion power? That claim was made by an industry group in 1980 and repeated by the Nuclear Energy Institute's chairman in 2003. The NRC's historian found no evidence for it in Strauss's papers, and the content of the speech was about fission.

Did anyone believe it at the time? Some journalists repeated it enthusiastically, including the New York Times science reporter William Laurence. Utility executives and many technical experts did not. Earlier studies had concluded that nuclear power would at first be more expensive than coal.

Is nuclear power cheap today? Existing nuclear plants that have paid off their construction costs can produce power at competitive cost. New plants in the United States have been expensive. Vogtle Units 3 and 4 cost more than twice their original estimate, and Lazard's 2024 analysis put new nuclear's unsubsidized levelized cost at about 142 to 222 dollars per megawatt-hour, well above onshore wind and utility-scale solar.

What does "too cheap to meter" mean in business? It describes a product whose cost per unit is so low that measuring and charging for each unit is not worth it. Examples include flat-rate home internet and many free online services.

Is there anything today that is truly too cheap to meter? Some digital goods come close, because copying a file costs almost nothing. Even then, the businesses behind them have large fixed costs, which they cover with subscriptions, advertising, or bundling rather than per-unit charges.

Why did water used to go unmetered? In places like New York City, water was billed by building characteristics rather than measured use. Cities later introduced meters mainly to encourage conservation, because unmetered water gives people no reason to use less.

Key takeaways

The phrase comes from a 16 September 1954 speech by AEC chairman Lewis Strauss, who predicted that the next generation would enjoy electricity "too cheap to meter."
Utility leaders and many experts were skeptical at the time, and Strauss was speaking for himself rather than for the AEC.
The later claim that he meant fusion has little documentary support.
Even free fuel would not have made electricity free, because most of its cost is equipment and networks.
New nuclear power in the United States has turned out to be expensive, as the Vogtle project shows.
The phrase now describes any good whose marginal cost is below the cost of measuring it, from flat-rate internet to free software, and the AOL example shows how removing the meter changes behavior.
Modern grids with lots of solar sometimes see negative wholesale prices, but this makes time-based metering more valuable, not less.