What makes an American drug have to work? Not the market, which rewards what sells, and not a physician’s impression, which can mistake habit for evidence. What makes it have to work is a federal statute that bars any new medicine from sale until its maker proves, through adequate and well-controlled investigations, that the drug does what its labeling claims. That statute is the Drug Amendments of 1962, known ever since as the Kefauver-Harris Amendments, and it reorganized the relationship between drug makers, the drug agency, and prescribing physicians more thoroughly than any other single enactment in the history of American medicine. Its demand is simple to state and exacting to meet: safety is not enough, because a safe drug that does nothing still costs the patient money, time, and the chance of real treatment.

The strange fact about the statute is that nobody wrote it to do this. It began as an antitrust bill about drug prices, the product of years of Senate hearings on pharmaceutical patents and promotion, and by the spring of 1962 it was dying in committee, stripped of the compulsory licensing provisions its sponsor cared about most. Then a birth defect catastrophe in Europe, a sedative marketed as exceptionally safe, and a stubborn agency medical officer who had refused to approve the American application collided with the stalled bill in a single summer. Newspaper coverage made the reviewer a national figure, the President honored her in August, and within weeks the dying price bill had been rewritten around proof of effectiveness and passed by Congress.

This article traces the statute from its antitrust origins through the thalidomide summer to the effectiveness standard that became the operating system of American drug regulation. It walks through the substantial evidence requirement, the shift from default approval to affirmative approval, the birth of the investigational drug process, the quieter provisions on consent, manufacturing, and advertising, the retroactive review of thousands of older drugs, the economic debate over whether the law slowed innovation, and the later statutes that renegotiated its trade-offs. The story begins where the bill did, in a Senate subcommittee investigating why Americans could not afford their medicine.

The Antitrust Origins of the Kefauver-Harris Amendments

The statute that requires American drugs to prove they work began as an antitrust bill about prices. Before the Kefauver-Harris amendments became the Drug Amendments of 1962, Public Law 87-781, the measure that President John F. Kennedy signed on October 10, 1962 was a Senate proposal aimed at the cost of prescription medicine, the reach of pharmaceutical patents, and the marketing tactics of the industry that sold them. Congress.gov records the original introduction of S. 1552 on April 12, 1961, more than a year before the thalidomide crisis would give the bill its second life. The safety and effectiveness machinery that came to define American drug regulation was not in the bill’s original design, and the price provisions that were in it did not survive. Understanding that original design, and why it was dying in committee by the spring of 1962, is the only way to see the enacted statute clearly: it is the residue of a failed antitrust effort, repurposed in a matter of weeks.

This section traces that first life. It follows Senator Estes Kefauver’s Antitrust and Monopoly Subcommittee through its hearings on pharmaceutical pricing and patent practices, lays out the principal provisions of the 1961 bill as introduced, records the industry’s organized resistance, and reconstructs the bill’s slow stall and the stripping of its most contested provisions through the spring of 1962.

Why did a Senate antitrust subcommittee start investigating drug prices?

A Tennessee Democrat famous for televised hearings on organized crime turned his subcommittee’s antitrust machinery toward prescription medicine after years of complaints that Americans could not afford the drugs their doctors prescribed, and the hearings that followed became one of the most publicized congressional inquiries of the era.

Estes Kefauver of Tennessee chaired the Senate Subcommittee on Antitrust and Monopoly, a unit of the Judiciary Committee, and by the late 1950s he had made it the Senate’s most visible platform for challenging corporate concentration. Kefauver had first become a national figure through his televised Senate hearings on organized crime in 1950 and 1951, and he carried the same investigative instincts into economic questions in the years that followed. The antitrust panel operated with a small but aggressive staff, and its chief economist directed the industry-by-industry price studies that supplied the hearings with their exhibits. Where other Senate investigations chased headlines for a season, Kefauver’s subcommittee worked as a standing research operation, publishing interim reports that built the case for legislation piece by piece. The subcommittee’s guiding doctrine was the theory of administered prices, associated with the economists Gardiner C. Means and John K. Galbraith, which held that large firms in concentrated industries set prices by administrative decision rather than by the play of supply and demand. Kefauver had already aimed the theory at steel, automobiles, and other heavy industries, and his staff economists spent years compiling price data industry by industry. Medicine entered the portfolio because the letters did. Kefauver told the subcommittee at the outset of the drug inquiry that the country had the best drugs in the world while a great number of citizens were experiencing difficulty being able to purchase them, and a Kaiser Health News retrospective published in February 2019 records that framing as the stated rationale for the inquiry.

The drug hearings opened in late 1959 and ran intermittently until September 1960, a ten-month span in which the subcommittee heard roughly 150 witnesses, according to the historian Daniel Scroop’s account of Kefauver’s investigations. The sessions concentrated on the wonder drugs of the period: antibiotics, corticosteroids, and tranquilizers, the high-volume products whose prices most affected household budgets and public programs. The inquiry had two tracks that ran together throughout. One track was price: how much a brand-name prescription cost the patient against what the same compound cost when sold under its generic designation, and how wide that gap could run when patent protection and brand promotion did their work. The other track was patent practice: how firms obtained fresh patents on modest molecular modifications of existing compounds, extending market exclusivity without producing a meaningfully different therapy. A pharmacology professor from what later became the University of California, San Francisco, Dr. Frederick Meyers, told the subcommittee that companies often took an existing product and modified the original drug just enough to secure a patentable alternative, a characterization that became one of the most quoted moments of the hearings. The subcommittee’s staff economists framed the whole inquiry in the language of administered prices, treating the pharmaceutical business as another concentrated industry where large firms set prices by decision rather than by competition. That framing mattered because it determined the remedy. If high drug prices were the product of monopoly structure, then the answer was structural: weaken the patent exclusivity, force generic naming, and police the promotion that sustained brand loyalty. President Kennedy would later describe the effort as twenty-eight months of intensive investigation and hearings, a figure that counts the staff work before and after the public sessions and signals how deeply the subcommittee had invested in the diagnosis before a bill was ever drafted.

What did the drug hearings reveal about prices and patents?

The hearings produced a public record of wide brand-to-generic price gaps, heavy spending on physician-directed promotion, and patent strategies built on small modifications, alongside an industry defense that attributed the price level to research costs and the affordability problem to low incomes rather than to excessive charges.

The price evidence came first and drew the headlines. Subcommittee staff laid out comparisons between brand-name prices and the prices of chemically equivalent products sold under generic designations, and the spreads were large enough to carry the inquiry for months. On promotion, a committee staffer displayed the pile of samples, advertisements, and flyers that a single Minnesota physician had received in just one month, a theatrical exhibit meant to suggest that marketing budgets, not manufacturing costs, explained much of the retail price. Kefauver himself suggested at one point that if companies spent less on advertising, armies of salesmen, and executive stock options, they could lower the price of drugs too. Prescription advertising in 1960 was aimed at doctors and hospitals through medical journals rather than at consumers, yet the volume struck the subcommittee as extravagant.

The promotion evidence cut close to the industry’s daily practice. Detail men, the traveling sales representatives who visited physicians’ offices, were shown to be a major channel of influence over prescribing, and the subcommittee heard that promotional spending shaped which products doctors reached for as much as clinical judgment did. One expert witness proposed that the government publish an official list of patented drugs with their generic equivalents and prices, so that the cost differences the hearings had exposed would be visible to every prescriber. The proposal went nowhere in 1960, but it showed the direction of the subcommittee’s thinking: transparency and generic substitution as weapons against brand-driven pricing.

The industry’s witnesses answered with a consistent counter-narrative. John Connor, then president of Merck, told the senators he had deep sympathy for people unable to afford medicine. Francis Brown, president of Schering, said that some people undoubtedly found it difficult to pay for needed medication but would also have difficulty meeting rent and food bills, framing the issue as inadequate income rather than excessive prices. The most pointed industry testimony came from Austin Smith, president of the Pharmaceutical Manufacturers Association, who told the subcommittee there was no need to limit drug prices, that attempts to prove excessive industry profits were doomed to failure, and that drug prices were rising more slowly than consumer prices generally. The executives urged Congress to create government programs to help people pay for health care, a position the Kaiser Health News retrospective noted was vindicated a few years later when public coverage expanded. None of the testimony moved Kefauver. The subcommittee’s investigators left the hearings convinced that monopoly pricing and patent manipulation, not research expense, explained the price level, and Kefauver began drafting legislation to match that diagnosis.

What was in the Kefauver drug bill as introduced?

S. 1552, introduced on April 12, 1961 and recorded by Congress.gov as the bill that became Public Law 87-781, was principally a price and competition measure whose headline provisions were compulsory licensing of drug patents, tougher patent standards for modified compounds, generic naming requirements, and regulation of drug advertising.

The compulsory licensing provision was the bill’s most aggressive element and the one the industry feared most. Under the draft, any firm could license and sell a patented drug in return for royalties paid to the patent holder, once three years had passed, a mechanism designed to break the link between patent exclusivity and high prices by inviting generic competition while the patent was still in force. Alongside it, the bill would have withheld patents for modified drugs unless the new molecule produced a therapeutic response significantly greater than the original, directly answering the molecular-tweaking practice that Dr. Meyers had described to the subcommittee. A related provision required consideration of therapeutic effectiveness in the granting of patents for drugs that were modifications of other drugs, so that a minor chemical change could not by itself earn a fresh period of exclusivity.

The remaining provisions aimed at the information environment around prescribing. The bill provided for standardization of drug names, so that physicians would encounter and prescribe by nonproprietary designations rather than by brand names alone, which Kefauver believed would steer prescribing toward lower-priced equivalents. It required disclosure of adverse as well as beneficial effects in drug promotion, a direct response to the advertising exhibits the subcommittee had displayed during the hearings. And it added safeguards meant to assure the reliability and effectiveness of drugs placed on the market, language that already hinted at quality control but did not yet carry the proof-of-effectiveness requirement that would define the enacted statute. The advertising provisions deserve emphasis because they addressed the exact practices the hearings had exhibited. Requiring that promotion disclose adverse effects alongside beneficial ones struck at the one-sided medical-journal advertising the subcommittee had condemned, and standardizing drug names struck at the brand-name system that let a firm sell a long-available compound as a premium product. Together these provisions treated the information given to physicians as a competition problem: if doctors prescribed by generic designation and saw the full picture of risks, the subcommittee reasoned, brand premiums would erode. President Kennedy’s letter of April 10, 1962 to Senate Judiciary chairman James Eastland catalogued these same provisions, compulsory licensing, reliability and effectiveness safeguards, standardized names, adverse-effect disclosure, and the patentability test for modified drugs, which confirms that the bill’s center of gravity in the spring of 1962 was still competition and price.

How did the drug industry fight the bill?

The pharmaceutical industry met S. 1552 with a coordinated campaign led by the Pharmaceutical Manufacturers Association, arguing that the bill’s price and patent provisions would undermine research incentives, that compulsory licensing would punish successful innovators, and that the industry’s own pricing record needed no correction.

The association’s public case rested on three claims, each advanced repeatedly by Austin Smith and by company executives who appeared before the subcommittee. First, the industry said its profits were not excessive and could not be shown to be, a position Smith stated flatly in testimony. Second, it said drug prices were rising more slowly than consumer prices generally, which made medicine a bargain in relative terms rather than a problem case. Third, it warned that compulsory licensing and tightened patent standards would discourage the research spending that produced new therapies, a claim aimed at senators who might otherwise sympathize with the price argument. Behind the testimony ran a lobbying effort directed at the Judiciary Committee and its Patents subcommittee, where the technical patent provisions would live or die. The industry found a receptive audience for delay. The bill’s patent and licensing machinery was complex, its opponents were well funded and well organized, and the Senate’s calendar in an election year offered many quieter ways to kill a measure than voting it down. The association’s campaign worked the committee structure rather than the floor. Compulsory licensing was a patent question, and patent questions belonged to the Patents subcommittee, whose members were readier to treat the provision as a technical problem for study than as a moral question for action. Kefauver’s own subcommittee had produced the investigation, but it did not control the bill’s fate; jurisdiction over S. 1552 sat with the full Judiciary Committee, where the chairman, James Eastland of Mississippi, had no appetite for a fight with a major industry. That mismatch between the investigators and the gatekeepers shaped everything that followed.

The opposition’s most important victory came without a floor fight. The Judiciary Committee’s Subcommittee on Patents took up the compulsory licensing feature and, as Kennedy reported to Eastland in the April 1962 letter, decided that the provision required further study and consideration. In Senate practice, a referral for further study of a bill’s central provision is a polite form of removal, and the industry understood it as such. By the spring of 1962 the measure that had been introduced as a frontal assault on drug prices and patent practices was losing the provisions its sponsor cared about most, and no replacement momentum had yet appeared.

Why did the bill stall in committee?

The bill was stalling because its price and patent core was being dismantled in committee, with compulsory licensing parked for further study and the remaining provisions lacking the political force to carry the measure forward on their own, a retreat the White House acknowledged when it asked that the surviving provisions be salvaged.

The procedural picture in the spring of 1962 was bleak for Kefauver’s original design. S. 1552 sat before the Judiciary Committee, where Eastland’s chairmanship gave industry allies every procedural tool for delay. The Patents subcommittee’s decision to set compulsory licensing aside for further study removed the provision that had given the bill its bite, and the patentability test for modified drugs faced the same skeptical reception from senators attuned to the industry’s research-incentive arguments. Kennedy had endorsed the bill’s aims in a consumer message to Congress on March 14, 1962, calling for new legislative authority to protect consumers in the marketing of drugs, and his April 10 letter to Eastland urged that the loss of compulsory licensing not delay enactment of the other provisions. The letter is the clearest contemporary evidence of the stripping process: a president asking a committee chairman to salvage the bill’s secondary provisions because its primary one was gone.

What remained was a weakened measure whose original purpose was slipping away. The antitrust investigation had run its course, the hearings had produced their headlines, and the legislation meant to convert those findings into law was being reduced provision by provision. Kennedy’s engagement showed how far the bill had fallen and how hard its friends were working to keep it alive. His March 14, 1962 consumer message had asked Congress for new authority over drug marketing, and the April 10 letter to Eastland pressed the case again, arguing that S. 1552 incorporated the major recommendations of that message and would give consumers better, safer, and less expensive medicines. The letter’s concession on compulsory licensing was the tell: the White House was asking the chairman to move the rest of the bill precisely because the lead provision was being set aside. By the time the spring session wore on, the bill that Kefauver had introduced in April 1961 looked unlikely to reach the floor in anything like its original form. The transformation that rescued it, and turned a dying price bill into the statute that requires American drugs to prove they work, arrived in the summer of 1962 from outside the antitrust story altogether.

The catalyst was a sedative most Americans had never heard of, and a medical reviewer who refused, month after month, to let it onto the American market.

The Catalyst: Thalidomide and Frances Kelsey

How did thalidomide reach millions of patients across Europe?

A German firm launched the sedative in 1957 under the brand Contergan, promoted it as exceptionally safe including for pregnant women, and licensed it to manufacturers worldwide, so that within a few years it was sold in 46 countries under dozens of brand names, including Distaval in Britain, as a standard remedy for insomnia and morning sickness.

The sedative that remade American drug law began as a German pharmaceutical product introduced to the German market on October 1, 1957, by Chemie Grunenthal under the brand name Contergan. Philip J. Hilts, in Protecting America’s Health, a history of the Food and Drug Administration published by Alfred A. Knopf in 2003, records both the launch date and the way the company framed the product, as a sedative for insomnia that was also promoted against the nausea of pregnancy. The drug’s active ingredient, thalidomide, had been developed in the early 1950s and was licensed rather than invented by the German firm. Its appeal to physicians rested on one striking claim, that it was so nearly nontoxic that no overdose could be lethal, and advertising materials emphasized safety for pregnant women without evidence from tests on pregnant women. Through licensing agreements with manufacturers around the world, thalidomide reached pharmacies in 46 countries under dozens of brand names, including Distaval in the United Kingdom, where it was distributed by the Distillers Company, and it became a standard prescription for morning sickness. In the United Kingdom, advertisements stated that Distaval could be given with complete safety to pregnant women and nursing mothers without adverse effect on mother or child.

The German company sought an American partner before the drug’s launch in Germany. Representatives approached Smith, Kline and French with a request to market and distribute the drug in North America, and a memorandum later found in the agency’s archives showed that the American firm ran animal tests and a clinical trial involving 875 people in 1956 and 1957 before declining to commercialize the product, for reasons that stayed hidden for decades. In 1958, Grunenthal reached a separate agreement with the William S. Merrell Company of Cincinnati, Ohio, a prescription drug maker that later became Richardson-Merrell, to market and distribute thalidomide throughout the United States. The company filed a new drug application with the federal food and drug agency in September 1960 seeking permission to sell the drug under the trade name Kevadon. Thalidomide thus arrived in the United States on paper, through a four volume dossier, at a moment when the governing 1938 statute let a new drug reach the market automatically if the agency raised no objection within sixty days of filing.

Why did one medical reviewer refuse to approve the American application?

Dr. Frances Oldham Kelsey, the agency’s newly hired medical officer, found the American application incomplete: translated documents contained errors, the animal studies were a poor guide to human toxicity, and the clinical reports resembled testimonials rather than controlled studies, so she repeatedly demanded more data and restarted the sixty day clock each time.

The dossier landed in September 1960 on the desk of Dr. Frances Oldham Kelsey, a physician and pharmacologist who had joined the agency the previous month. Kelsey was born in Cobble Hill on Vancouver Island, trained in pharmacology at the University of Chicago and earned a medical degree there as well, and the thalidomide application was the first drug assignment she received at the agency, according to a Food and Drug Administration press release issued in October 2000. Under the procedures of the day, a three person team reviewed each new drug application: a chemist, a pharmacologist, and a medical officer. Each member of the team assigned to thalidomide found problems. The chemist found errors in the translation of the German source documents. The pharmacologist judged the animal studies a poor guide to possible human toxicity because animals absorb thalidomide poorly. Kelsey determined that the clinical studies were incomplete and that the claims in the file were not adequately supported. Many of the physician submissions read more like advertising testimonials than well designed scientific studies, and some carried the names of authors she recognized from questionable journal papers, as recounted in a Discover magazine profile of her work.

Rather than clear the application within the sixty day window, Kelsey asked the company for more data, which restarted the clock. The company resubmitted in January 1961, and Kelsey refused again. In early 1961 she came across a letter in the British Medical Journal in which the British distributor described cases of peripheral neuritis, a painful numbing and tingling in the hands and feet, among patients taking thalidomide. The American application had said nothing about that side effect, and the company later said it had not known of it. Kelsey wrote to Merrell asking for additional information and further proof of safety, and she pressed for data on the nerve damage reports that were accumulating in Europe. The company kept pushing for approval and contacted the agency’s Bureau of Medicine roughly fifty times over the following year, as one contemporary account tallied it. Every sixty days, the cycle repeated: the company submitted more material, Kelsey found it inadequate, and the drug stayed off the American market. The repeated refusals continued from September 1960 through 1961 and into 1962. Because the application was never cleared, thalidomide was never approved for sale in the United States, a point that the record makes explicit and that later retellings sometimes blur.

One feature of the old law made her stand narrower than it looked. The 1938 statute governed the marketing of a drug and said almost nothing about its investigation. While the Kevadon application sat pending, the company could legally distribute the unapproved drug to physicians for what was called investigational use, with no required protocol, no required informed consent, and no required records. More than one thousand two hundred American doctors received tablets, and roughly twenty thousand patients took them, including several hundred pregnant women, according to an interview with agency historian John Swann recorded for the fiftieth anniversary of the 1962 amendments. The agency later recognized seventeen American children born with thalidomide related deformities, of whom about ten were traced to the investigational distribution and the rest to drugs brought in from abroad, Swann said in the same interview. The distinction mattered: Kelsey’s refusals kept the drug out of American pharmacies, but they could not stop the tablets already circulating hand to hand.

What evidence connected the drug to the birth defects?

Adult nerve injury reports came first, then the German pediatric geneticist Widukind Lenz traced a surge of phocomelia to the drug in November 1961, and the Australian obstetrician William McBride reported severe congenital abnormalities in The Lancet that December, completing the causal picture before the year ended.

The first clinical warnings concerned the nervous system rather than pregnancy. A letter in the British Medical Journal in December 1960, from a physician named Leslie Florence, described four patients who had developed severe paraesthesia, a numb or burning sensation in the hands and feet, while taking thalidomide, and noted that the symptoms persisted months after the drug was stopped. Distillers’ own letter in early 1961 supplied the British confirmation. These reports were the red flags that made Kelsey demand more data, and they established a pattern of nerve injury in adults that the American application had omitted.

The birth defect link was established in Germany in the autumn of 1961. Widukind Lenz, a pediatric geneticist at the University of Hamburg, had been tracking an alarming rise in a rare pattern of limb malformation known as phocomelia, in which the hands and feet attach almost directly to the trunk because the long bones of the arms and legs fail to develop. By August 1961, Lenz estimated that the incidence of gross long bone defects had risen tenfold in Hamburg and by a far greater multiple in some towns of Rhineland and Westphalia, as documented in a history of thalidomide published by Open Book Publishers. Working through case records in the first weeks of November 1961, Lenz found repeated references to Contergan in interviews with the parents of affected infants. He telephoned Grunenthal with his concern in mid November, and on November 16, 1961, he sent the company a registered express letter setting out his reasons for connecting the drug to the deformities. On November 18, 1961, he laid his findings before a meeting of pediatricians in Dusseldorf, though one detailed account holds that he did not name the medication in his remarks that day. The full results of his study were published in June 1962, but the November presentation was the decisive public step: the drug was pulled from circulation in Germany within days, in the last week of November 1961. Other countries followed. The British distributor, Distillers, withdrew the drug on December 2, 1961, according to the Thalidomide Trust timeline.

In December 1961, a letter from William McBride, an obstetrician in Sydney, Australia, appeared in The Lancet under the heading “Thalidomide and Congenital Abnormalities.” McBride reported that in recent months he had observed an incidence of multiple severe abnormalities in babies delivered of women who had been given thalidomide, sold locally as Distaval, as an antiemetic or sedative during pregnancy, and he put the incidence at nearly twenty percent of such births, against a background rate he estimated far lower. McBride had himself helped popularize the drug for morning sickness in Australia, which gave his warning added weight. With Lenz’s findings in Germany and McBride’s letter reaching the medical world, the picture was complete before the end of 1961: a sedative advertised as exceptionally safe had produced an epidemic of limb deformities across the countries where it had been sold. Grunenthal’s withdrawal in Germany was accompanied by the company’s public position that the action was a response to press sensationalism rather than an admission of cause, a framing recorded in the press coverage of the period.

Merrell withdrew its Kevadon application in the United States on March 8, 1962, conceding a fight it had pressed for eighteen months. The drug never received American approval, and the withdrawal closed the formal application. But the political story was only beginning, because most Americans had still never heard the name thalidomide.

How did newspaper reporting move a stalled bill?

A July 15, 1962 Washington Post front page story by Morton Mintz told readers that Kelsey’s refusal had kept a drug off the American market that had injured thousands of European babies, and the nationwide coverage that followed gave the stalled bill the public urgency it had lacked.

Through the first half of 1962, the birth defect disaster was a European story and the American public knew almost nothing of it. Physicians who tried to warn Congress fared no better than the early European case reports. In May 1962, at the request of Senator Estes Kefauver, the cardiologist Helen Taussig testified before Representative Emanuel Celler’s Antitrust Subcommittee about the drug’s effects on the unborn, and the American press passed over her testimony entirely, according to an account of the episode published in a Newcomb College Institute journal. Kefauver’s bill, which by then had been stripped of its most contested provisions and sat without momentum, gained nothing from the testimony.

The summer changed that. In mid July 1962, Kefauver and his staff decided to bring the thalidomide story to the press as a way to revive the legislation, and Morton Mintz, a reporter at the Washington Post, received information that the drug would almost certainly have reached the American public if not for Kelsey’s persistence. The tip reached Mintz through the Post’s own chain of desks, from a congressional staffer to reporter Bernard Nossiter to assistant city editor Sy Fishbein, who held the assignment until the paper’s medical writer went on vacation and then gave it to Mintz, as Mintz later recalled in an interview with the International Consortium of Investigative Journalists. Mintz interviewed Kelsey and wrote the story that the Post put on the front page on July 15, 1962, above a photograph of Kelsey, under the headline “‘Heroine’ of FDA Keeps Bad Drug Off Market.”

The article told readers that a medical reviewer had stood alone against a pharmaceutical company and that the drug she blocked had injured thousands of European babies. It was picked up by newspapers across the country within days and produced a wave of editorials and follow up stories about drugs and drug control. One later retrospective in Scientific American, tied to the Lost Women of Science podcast series, described the July 15 story as the pivot from which everything else spun, noting that before that date most Americans knew nothing of the drug and that thousands of them were still taking it unknowingly. The reporting also forced into public view the two features of the old law that the thalidomide file had exposed: that effectiveness had never been required for approval, and that an unapproved drug could be distributed to doctors as samples or for investigation without agency oversight. In August 1962, President Kennedy used a press conference to warn the country that thalidomide might still sit in American medicine cabinets, a remark that Swann cited in his fiftieth anniversary interview as a measure of how the story had seized public attention.

Kelsey herself entered the public record directly. On August 1, 1962, she testified before the Senate about the Bureau of Medicine’s role in keeping the sedative off the American market, and her testimony criticized the lobbying tactics of the drug industry’s trade associations and of Richardson-Merrell in opposing the drug bill, an account preserved in the Newcomb College Institute journal. The transformation of the bill’s prospects dated from those weeks. A measure that had spent the spring being stalled and stripped carried the most potent political fuel available in Washington: a named catastrophe abroad, a named heroine at home, and a press corps that had made both familiar to the electorate.

What honor did the President bestow on the reviewer?

On August 7, 1962, President John F. Kennedy presented Frances Oldham Kelsey with the President’s Award for Distinguished Federal Civilian Service, the highest civilian honor for government employees, crediting her judgment with preventing a major tragedy of birth deformities in the United States.

On August 7, 1962, in a White House ceremony, President John F. Kennedy presented Frances Oldham Kelsey with the President’s Award for Distinguished Federal Civilian Service, the highest civilian honor established for government employees. The records of the John F. Kennedy Presidential Library document the ceremony, which honored six recipients that day on the South Lawn, and list Kelsey by her title as medical officer at the Food and Drug Administration. She became the second woman ever to receive the award. An agency press release issued in October 2000 reproduced the engraving on the award, which credited her exceptional judgment in evaluating a new drug for safety and stated that it had prevented a major tragedy of birth deformities in the United States through her high ability and steadfast confidence in her professional decision.

The honor fixed Kelsey’s refusal in the national memory as the act that had spared the country, and it bound the thalidomide story to the fate of the drug bill in Congress. Within days of the ceremony, the political calculation around the legislation had inverted. Opposition that had been organized and effective in the spring dissolved in the glare of the summer’s coverage, and the bill that Kefauver had introduced as an antitrust measure about drug prices was reworked around the very subject Kelsey had administered: proof that a drug works, and control over how drugs are tested and advertised.

How did the bill reach the President’s desk?

Kefauver reworked the bill around agency-drafted safety and effectiveness provisions, dropped the contested price and compulsory licensing measures, and the final conference report passed both chambers in the first days of October 1962 without recorded dissent; Kennedy signed it on October 10 as Public Law 87-781.

Kefauver reworked his legislation to carry provisions drafted by the agency itself, aimed at preventing another thalidomide. An agency consumer publication describing the history states that Kefauver reintroduced his legislation to include provisions drafted by the Food and Drug Administration specifically designed to prevent a disaster like thalidomide, with Oren Harris as the chief cosponsor in the House. The price and compulsory licensing provisions that had made the original bill so bitterly contested largely fell away, and safety and effectiveness provisions took their place. The Senate had passed an earlier version in August 1962 with recorded dissent, but the final conference report moved through both chambers in the first days of October without recorded dissent, and an agency account of the history describes the legislation as passed unanimously by both houses of Congress. The House agreed to the conference report on October 4, 1962, and the Senate concurred. President Kennedy signed the bill into law on October 10, 1962, as Public Law 87-781, the Drug Amendments of 1962.

The speed of the ending remains the measure of the catalyst’s force. A bill that had been introduced in 1961, investigated for years through an antitrust subcommittee’s hearings on pharmaceutical pricing, stalled and stripped through the spring of 1962, completed its passage within weeks of the July newspaper stories and the August award. The statute that required American drugs to work was therefore the residue of two histories fused by accident: a long antitrust campaign against drug prices that could not pass on its own terms, and a birth defect catastrophe that supplied, within a single summer, the public horror and the public heroine that the antitrust campaign had never possessed. Kelsey’s refusals, which had begun as a routine exercise of a sixty day review clock in September 1960, ended as the premise of a law she had not set out to create, passed by the Congress that had spent the previous year refusing to pass the bill its sponsor had actually written.

The repurposed bill’s answer to the next question, what a drug maker must actually prove, is the single most consequential sentence in American drug law.

The Effectiveness Requirement at the Heart of the Amendments

The Drug Amendments of 1962 are remembered as the law that made American medicines prove they work, and the instrument of that change was a single definitional turn. Before the amendments, the Federal Food, Drug, and Cosmetic Act of 1938 asked one question of a new compound: is it safe. The 1962 statute added a second question and gave it teeth: does it do what its labeling says it does. That second question, framed as the substantial evidence standard and answered only through adequate and well-controlled investigations, rearranged the relationship between the drug maker, the regulator, and the physician, and it created the clinical trial system that still organizes human drug testing. This section traces the statutory language, the procedural reversal from default approval to affirmative approval, the birth of the Investigational New Drug process, and the way the requirement operates once a compound moves from the laboratory toward the market.

What does the substantial evidence standard require?

Substantial evidence is evidence consisting of adequate and well-controlled investigations, including clinical investigations, conducted by experts qualified by scientific training and experience, on the basis of which those experts could fairly and reasonably conclude that the drug has the effect its labeling claims under the stated conditions of use.

Section 505(d) of the act, codified at 21 U.S.C. section 355(d), defines the term in language Congress wrote with visible care, and every clause of the definition does work. The evidence must consist of investigations, plural, and it must include clinical investigations, which foreclosed reliance on animal data alone or on the accumulated impressions of physicians who had prescribed the compound. The investigations must be adequate and well-controlled, a phrase the agency was left to interpret but which the statute plainly aimed at studies designed to distinguish the effect of the compound from the effects of chance, of the underlying disease, and of the expectations of patients and investigators. They must be conducted by experts qualified by scientific training and experience to evaluate effectiveness, which moved judgment about what counts as proof away from the prescribing physician and toward people trained to design and read experiments. And the standard of judgment is set inside the definition: the experts must be able to fairly and reasonably conclude that the compound will have the effect it purports or is represented to have under the conditions of use prescribed, recommended, or suggested in its labeling. The statute thus tied effectiveness to the labeled claims rather than to any free-floating notion of therapeutic value, and it made the quality of the proof, not the sincerity of the belief, the test.

The contrast with the 1938 regime was deliberate and total. The 1938 act defined a new drug as one not generally recognized by experts as safe for use under the conditions stated in its labeling, and section 505(b) required the application to carry full reports of investigations made to show whether the compound was safe. Effectiveness entered the picture only at the margins. The 1938 act reached false therapeutic claims through its misbranding provisions, which meant that the question whether a compound actually worked was fought out after marketing, in court, through prosecutions and seizures, with judges weighing testimony about cures. The 1962 amendments rewrote the statutory definition of a new drug itself so that general recognition now had to run to safety and effectiveness together, and they required the application’s full reports to show both. As the agency’s own clinical investigator training materials later put it, until 1962 medicines had to be shown safe to be marketed, but there was no requirement to show effectiveness, and the amendments changed that in one stroke. A former agency official and longtime student of drug regulation, Robert Temple, observed in a 1994 chapter on the development of American drug law that the radical element was not the effectiveness requirement as such, which a creative reading of safety might have implied, but the demand for adequate and well-controlled studies as the only acceptable basis for approval. That observation captures why the standard mattered: Congress did not ask the agency to form an opinion about efficacy in the abstract. It prescribed the kind of evidence that would count.

Congress also gave the plural its teeth. The statute says investigations, and section 505(b) likewise speaks of full reports of investigations. The agency read that plural, supported by language in the Senate report suggesting the phrase was meant to address the quantum as well as the quality of the proof, as a general expectation of more than one adequate and well-controlled study, each convincing on its own. That reading held as the working norm for decades. The Food and Drug Administration Modernization Act of 1997 later softened it, permitting approval on the basis of one adequate and well-controlled clinical investigation supported by confirmatory evidence, but the default picture the 1962 statute painted was replication: one trial could be a fluke, a second independent trial answering the same question made the finding sturdy. Nothing in the standard required the agency to compare the compound against competing treatments. The question was always whether the medicine did what its own labeling claimed, not whether it did it better than something else, and the statute left that comparative judgment to the physician at the bedside.

How did the sixty-day default approval work before the amendments?

Under the 1938 statute, a new drug application automatically became effective on the sixtieth day after filing unless the Secretary issued an order stopping it. The amendments replaced that default with a requirement that the Secretary affirmatively approve the application, within one hundred and eighty days of filing, before the drug could be marketed.

The mechanics of the old regime explain why the change mattered as much as the new standard itself. Section 505(c) as written in 1938 provided that an application became effective on the sixtieth day after filing unless, before that day, the Secretary issued an order refusing to permit it to take effect. Inaction was consent. The agency had two months to find a reason to object, and if its reviewers were still working through the file, still waiting on data from the sponsor, or still debating among themselves, the clock resolved the question in the sponsor’s favor. The provision reflected the 1938 act’s limited ambition: the agency was a safety checkpoint with a short fuse, and the default posture of the law was that commerce in a new compound should proceed. A reviewer who needed more time, or who harbored doubts the file did not yet answer, had no statutory shelter; delay past the sixtieth day meant the compound reached the market without the agency ever having said yes.

The 1962 amendments inverted the default. The revised section 505(c) required the Secretary, within one hundred and eighty days after the filing of an application, either to approve it upon finding that none of the grounds for denial applied, or to give the applicant notice of an opportunity for a hearing on whether the application was approvable. The one hundred and eighty day period could be extended by agreement between the agency and the applicant, which in practice turned the deadline into a framework for negotiation over what further evidence was needed rather than a hard guillotine. Section 505(d) then listed the findings that compelled refusal, and lack of substantial evidence of effectiveness stood among them alongside inadequate safety testing, deficient manufacturing, and false or misleading labeling. Taken together, the two provisions shifted the burden of inertia. Before 1962, a sponsor who filed a thin application could win by running out the clock. After 1962, a sponsor who filed a thin application lost by standing still, because approval required an affirmative order grounded in findings the file had to support. The agency’s history of clinical trial regulation describes the change exactly this way: the sixty-day approval default under the 1938 act was removed, and new medicines had to have positive and specific approval before reaching the market.

The reversal also changed what it meant to refuse. Under the old default, stopping an application meant issuing an order against the current of the statute; the agency had to act to prevent marketing. Under the new scheme, approval itself was the act, and refusal was the resting state. That reallocation of the burden of action gave the agency’s scientific reviewers something they had never possessed: the ability to say not yet, and to have the law stand behind the pause while the evidence was gathered. The reviewers no longer needed to prove a compound ineffective or unsafe within sixty days to keep it off the market. They needed only to find the evidence insufficient, and the application waited.

How did the Investigational New Drug process begin?

Section 505(i) gave the Secretary authority to exempt drugs shipped for investigational use from the premarketing approval requirements. The agency’s 1963 investigational drug regulations built that authority into a filing called the Notice of Claimed Investigational Exemption for a New Drug, shortened to IND to parallel NDA, which sponsors submitted before testing a compound in people.

The investigational exemption solved a problem the effectiveness standard created. If no compound could be marketed without proof of effectiveness from human studies, then the law had to create a lawful way to conduct those human studies before approval existed. Section 505(i) supplied it: the Secretary could issue regulations exempting drugs intended solely for investigational use from the approval requirements of section 505, subject to conditions the regulations imposed. The agency built on that grant with the investigational drug regulations it issued in 1963. Under the new system, a company sponsoring a compound had to file its notice of claimed investigational exemption before human trials began, and the notice was in substance a package of materials: preclinical pharmacology and toxicology data sufficient to show the compound could be given to people with reasonable safety, a general plan of investigation, the qualifications of the investigators, and the manufacturing information bearing on the identity and quality of the material to be tested. The acronym IND was coined deliberately to parallel NDA, and technically the IND is what its original name said it was: an exemption from the normal premarketing requirements, namely the submission and approval of a new drug application, granted so that the studies supporting a future application can lawfully be done.

The conditions attached to the exemption reached deep into how the studies were run. The regulations required that a qualified expert personally supervise each clinical investigation and that the investigator agree not to hand the compound to other investigators, a response to the loose distribution networks through which experimental compounds had circulated. Investigators had to keep records and report adverse events to the sponsor, and sponsors had to report them onward to the agency, planting the seed of the adverse event reporting system the amendments also created. And on the insistence of Senator Jacob Javits of New York, troubled by the number of people who had taken an experimental sedative in the thalidomide episode without understanding its status, the statute required that investigators obtain the informed consent of human subjects, so that no patient would again receive an experimental compound without being told it was experimental. A Sixth Circuit decision in 1964, in a case brought by a sponsor whose investigational exemption the Commissioner had terminated, shows the system already operating within a year of the regulations: the court recounted a notice filed in June 1963, a September letter from the Commissioner listing inadequacies including missing animal data and no real plan, and a November termination of the exemption, with the court sustaining the agency’s authority to police the exemption’s conditions.

Out of this investigational machinery grew the phased structure of human testing that later became the default method for studying medicines in people. The familiar sequence runs from first introduction of the compound into humans, typically in small groups of twenty to eighty subjects, to determine metabolism, pharmacologic action, and the side effects associated with increasing doses; through controlled studies in patients with the disease or condition under study, usually involving no more than several hundred subjects, to evaluate effectiveness and determine the common short-term side effects; to expanded controlled and uncontrolled trials in thousands of patients, performed after preliminary evidence of effectiveness has been obtained, to gather the definitive evidence on which approval rests. A 2016 white paper from the Friends of Cancer Research traces this three-phase concept to the wake of the new law, when the agency’s interpretation of what adequate and well-controlled investigation required gave drug makers their working template, and notes that the agency adopted it as the standard method for studying medicine in humans from that point on. One point of precision is worth recording for the article’s accuracy: the phase labels themselves were not part of the 1963 investigational regulations. The three phases were codified as 21 CFR section 312.21 in the IND rewrite the agency completed in 1987, published at 52 Federal Register 8831, which also standardized the contents of the IND submission and formalized the agency’s interaction with sponsors during development. The concept grew out of the 1962 framework; the numbering entered the regulations a quarter century later.

How does the effectiveness requirement operate in practice?

A sponsor files an IND, conducts clinical studies under agency oversight, and submits a new drug application with full reports of the investigations showing the compound is both safe and effective. Approval is refused unless those reports supply substantial evidence of effectiveness, and the Supreme Court sustained that framework in Weinberger v. Hynson, Westcott and Dunning in 1973.

In operation, the standard works as a chain of gates, each one carrying the burden forward. The sponsor’s obligation begins with the content of the application itself. Section 505(b) requires the applicant to submit full reports of the investigations made to show whether the compound is safe for use and whether it is effective in use, along with the compound’s components, its composition, the manufacturing methods and controls, samples, and the proposed labeling. The phrase full reports matters: the agency was entitled to the underlying data, not summaries or conclusions, so that its own reviewers could reanalyze what the sponsor’s investigators had found. When the agency’s clinical investigator training describes what makes an investigation adequate and well-controlled, it names three aims that have organized review practice ever since: a valid control group similar enough to the test group that the difference can be attributed to the compound, minimization of bias in how subjects are selected, treated, observed, assessed, and analyzed, and documentation detailed enough to permit critical evaluation of whether the study’s design actually supports its conclusions. Randomization, blinding, and prespecified endpoints became the working vocabulary of this requirement, not because the statute named them but because they were the techniques by which the statutory aims could be met and demonstrated.

The agency’s refusal authority gave the standard its bite. When the reports failed to supply substantial evidence, the Secretary was required to issue an order refusing approval, with findings of fact, after giving the applicant notice and an opportunity for a hearing. That structure made the scientific dispute reviewable but placed it inside the agency first: the sponsor argued to the agency’s scientists that its trials were adequate and well-controlled, and only then to a court. The Supreme Court’s 1973 decision in Weinberger v. Hynson, Westcott and Dunning sustained that architecture, upholding the agency’s power to withdraw approvals for lack of substantial evidence; the decision, and the retroactive review of pre-1962 compounds out of which it arose, are taken up in this article’s section on the retroactive review.

The requirement also reached backward by design. Congress did not limit the new effectiveness standard to compounds first submitted after 1962. The amendments directed the agency to review the effectiveness of the thousands of compounds approved during the safety-only era between 1938 and 1962, working with the National Academy of Sciences, and to remove from the market those lacking substantial evidence. That retrospective program, the Drug Efficacy Study Implementation, is the subject of this article’s section on retroactive review, but its existence belongs in the account of how the standard operated: effectiveness was not merely a hurdle for future compounds. It was a condition that past approvals had to survive.

The stringency of that regime has been contested since its first years, and the argument over whether proof of effectiveness slowed innovation or bought safety belongs to this article’s section on the innovation debate. What the record establishes as the starting point is the mechanism itself: since 1962, no new compound has reached the American market without an affirmative agency finding, grounded in adequate and well-controlled investigations, that it does what its labeling says it does.

The comparison below sets the 1938 regime against the 1962 amendments on the four dimensions that mattered most to a manufacturer: what had to be shown, who bore the burden, what the agency could do, and how long it had.

What changed	Under the 1938 act	Under the 1962 amendments
What a manufacturer had to show	Safety of the new drug under the conditions stated in its labeling	Safety and effectiveness, with effectiveness shown by substantial evidence from adequate and well-controlled investigations
Who bore the burden	The agency, which had to issue an order stopping a drug within sixty days or it reached the market by default	The manufacturer, which had to win an affirmative approval order before marketing a single tablet
What the agency could do	Judge safety on the submitted record, then largely close the file; fight false therapeutic claims after marketing through misbranding prosecutions	Require proof of effectiveness, control human testing through the investigational exemption, demand adverse event reports, inspect factories and records, and police prescription drug advertising
How long the agency had	Sixty days from filing before the application became effective automatically	One hundred eighty days to approve or notice a hearing, extendable by agreement between the agency and the applicant

Effectiveness was the headline requirement, but the amendments carried four quieter ones that changed the daily operation of drug regulation, and a backward look that gave the new standard market-wide force.

The Other Requirements Beyond Effectiveness

The effectiveness standard drew the headlines in the autumn of 1962, and it has kept them ever since, but four quieter provisions of the amendments rewrote the daily operating law of American drug regulation. One changed the relationship between the investigator and the human subject. One imposed a standing duty on manufacturers to tell the agency when drugs hurt people after approval. One put federal inspectors inside the factory as a permanent presence. And one moved the policing of prescription drug promotion from the trade regulator to the drug regulator. Together they explain why the statute governed far more than the approval decision.

What did the informed consent requirement actually require?

Congress required investigators to obtain informed consent from the human subjects of drug investigations, except where the investigator judged consent not feasible or contrary to the subject’s best interests. The provision answered the discovery that women had taken thalidomide without being told the drug was experimental. Regulations that followed added protections for prisoners and institutionalized persons.

The thalidomide episode supplied the fact pattern that made the provision unanswerable. Women had been given the drug without being informed that it was experimental, without knowing that they were research subjects, and without being told that its safety had not been established. The agency’s own historical account of clinical drug trials records that the legislative history discussion, published at 44 Federal Register 47714 in August of 1979, traced the consent provision directly to that failure. Senator Jacob Javits of New York sponsored what became one of the most consequential provisions of the law: the requirement that investigators maintain personal supervision over clinical investigations, agree not to hand the drug to other investigators, and obtain consent from every human subject so that patients would have to be told specifically when a drug they received was experimental.

The consent rule that Congress enacted was not the strongest version that had been proposed. The House bill would have required informed consent in all clinical trials of drugs. The version reported out of conference allowed the exceptions that became law, recorded at House Report 2526 of the 87th Congress, second session, dated October 3, 1962, at pages four and five. Investigators could dispense with consent where they deemed it not feasible or, in their professional judgment, contrary to the best interests of the subjects. The Congressional Record for October 3, 1962, at pages 22038 and 22042 to 22043, accepted professional responsibility, grounded in what the record called the greatest exercise of conscience, as the basis for permitting administration of investigational drugs without consent. The statute thereby embedded informed consent as a condition of lawful investigation while leaving a narrow door for the investigator’s judgment, and the exceptions reflected the conference committee’s reluctance to forbid placebo-controlled trials outright.

The framework did not stay where Congress left it. The 1978 round of rulemaking developed the protections for the subjects least able to refuse. On August 8, 1978, the agency proposed regulations on institutional review boards, published at 43 Federal Register 35189, that would make board review a precondition for clinical investigations and that refused to waive the requirement in three situations, one of them investigations involving institutionalized human subjects. Later that year, on November 16, 1978, at 43 Federal Register 53655, the Department of Health, Education, and Welfare finalized additional protections for prisoners as research subjects, codified at 45 Code of Federal Regulations part 46, subpart C. Those rules limited the kinds of research in which prisoners could be enrolled and required special review, on the reasoning that incarceration could constrain the ability to make a truly voluntary and uncoerced decision about participation. The consent provision that began as a response to unwitting thalidomide exposure thus matured, through the regulations, into a system that singled out confined and institutionalized persons for the strongest procedural safeguards.

What changed in adverse event reporting and manufacturing standards?

The amendments required manufacturers to report adverse drug reactions to the agency and authorized it to set standards for how drugs were made. The first manufacturing regulations followed in 1963, giving inspectors authority over factory conditions and access to company production and control records. The two provisions created a federal presence in the factory, not a single judgment at approval.

The reporting duty grew out of the recognition that safety evidence could not be gathered entirely before marketing. Under the 1938 statute the agency judged safety on the record the manufacturer submitted, and then the file largely closed. The amendments imposed the opposite discipline: adverse drug reactions were to be reported to the agency as they occurred in the course of marketing, so that the record of a drug’s risks accumulated across its commercial life. That change turned post-approval experience into a regulatory input rather than a private fact, and it gave the agency the information base on which later withdrawal decisions would rest.

The manufacturing provisions went further than any earlier federal drug law in regulating the factory itself. On June 20, 1963, the agency announced three sets of regulations governing the manufacture, effectiveness, and promotion of drugs, the first comprehensive federal standards for how drugs had to be made. The hearings that produced the amendments had prepared this ground. Louis Lasagna, then a prominent clinical pharmacologist at the Johns Hopkins School of Medicine, advised Congress during the 1962 hearings on the proper conduct of clinical research, and those hearings prompted the recognition that good manufacturing practices had to cover the quality of the process as well as the quality of the finished product. The regulations required establishments to abide by current good manufacturing practices designed to ensure that drugs met standards of identity, strength, quality, and purity.

Congress paired the manufacturing standards with inspection powers that made them enforceable. The agency gained enhanced authority to examine company production and control records, and the historical account on the agency’s own site describes the grant plainly: greater powers to access company production and control records in order to verify the practices the regulations required. In an interview recorded for the fiftieth anniversary of the amendments, the agency historian John Swann summarized the package as enhanced inspectional authority, more access to drug records and quality control records, and good manufacturing practices made systematically the rule of law. The effect was to convert the factory from a private site visited when trouble appeared into a regulated environment whose records the agency could inspect as a matter of routine.

Why did Congress move prescription drug advertising from the FTC to the FDA?

Congress moved jurisdiction over prescription drug advertising from the Federal Trade Commission to the Food and Drug Administration. The Kefauver hearings had spotlighted advertising practices, and the shift put promotional claims under the same agency that evaluated a drug’s safety and effectiveness. Over-the-counter advertising remained with the trade commission.

The advertising question had surfaced early in the legislative process. Senator Kefauver’s antitrust and monopoly subcommittee hearings spent years on pharmaceutical pricing and patent practices, and advertising practices came into that record alongside them. In the fiftieth-anniversary interview, the agency historian John Swann identified advertising as one of the issues that came out of Kefauver’s hearings and noted the statutory result: the agency would be assigned control over prescription drug advertising, which had been the responsibility of the trade commission. The transfer was deliberately bounded. Only prescription drugs moved; the trade commission kept the advertising of products sold directly to consumers.

The logic of the transfer was institutional rather than punitive. The agency that reviewed the clinical investigations and decided what a drug could lawfully claim in its labeling would now also judge what the manufacturer could say in promoting it to physicians. That closed a gap in which the trade commission, whose expertise ran to commercial deception generally, had policed a technical domain whose claims turned on clinical evidence. The result has been discussed in the series’ companion treatment of the trade commission’s consumer protection work (/2019/01/15/ftc-consumer-protection-rulemaking/), where the advertising jurisdiction of the two agencies is traced through the twentieth century.

The Retroactive Review and Its Defense in Court

The amendments did not stop at the new-drug door. Congress required the agency to turn around and judge the drugs it had already let through. Between 1938 and 1962 the agency had permitted more than 9,000 new drug applications to become effective, of which some 4,000 were still on the market when the amendments took effect, and thousands of additional “me-too” products had been marketed without any clearance at all, in reliance on approvals held by other manufacturers. Every one of those products had been judged on safety alone. The statute gave manufacturers two years to develop substantial evidence of effectiveness for their previously approved products, during which the agency could not withdraw those approvals for lack of effectiveness, and then it demanded the reckoning. That backward look, carried out through the Drug Efficacy Study Implementation and defended in the Supreme Court, is what gave the effectiveness standard its market-wide force.

How did the DESI review work?

In 1966 the agency contracted with the National Academy of Sciences and National Research Council to evaluate the effectiveness of drugs approved for safety alone between 1938 and 1962. Panels reviewed over 3,000 products and 16,000 claims, sorting each as effective, ineffective, or needing further study. The agency implemented the findings through the Drug Efficacy Study Implementation program.

The contract was the largest independent scientific review of marketed drugs the country had attempted. Thirty panels of six professionals each, organized by drug category, conducted the review. The panels’ reports arrived in the late 1960s and early 1970s, the last of the National Academy and National Research Council reports was submitted in 1969, and the contract was extended through 1973 to cover continuing questions. The agency’s own task force completed its initial review of the reports in November 1970. Administrative implementation of the findings took the name Drug Efficacy Study Implementation, and the agency formed the DESI program in 1968 to carry the panels’ recommendations into regulatory action.

The scale of the undertaking is best stated in the agency’s own figures. The study evaluated over 3,000 separate products and over 16,000 therapeutic claims. By September of 1981 the agency had taken final regulatory action on 90 percent of the DESI products. By 1984 final action had been completed on 3,443 products, of which 2,225 were found effective, 1,051 were found not effective, and 167 were pending. The agency’s historical account of clinical drug trials states that the program removed over 1,000 ineffective drugs and drug combinations from the marketplace, and thousands more products were relabeled to reflect their demonstrated therapeutic qualities more accurately.

One early effect of the study reached beyond the drugs it judged. Products that the review found effective but whose labeling required changes needed a regulatory path that fit their situation, and the development of the Abbreviated New Drug Application grew out of that need. The DESI program thus did double duty: it cleared the market of products that could not meet the new standard and it supplied part of the machinery for the generic drug system that would follow.

The reviews also exposed the thinness of the evidence on which pre-1962 approvals had rested. Robert Temple, who began work at the agency in 1972, recalled that most of the old studies were inadequate beyond belief, that there was often no protocol at all, that there was almost never a statistical plan, and that sequential analyses were unheard of. The recollection, preserved in the agency’s history of clinical drug trials, captures what the panels kept finding: products judged on safety alone had often never been subjected to anything resembling a controlled investigation. In 1970 the Court of Appeals reinforced the new evidentiary discipline in Upjohn v. Finch, holding that commercial success alone did not constitute substantial evidence of a drug’s safety and efficacy. The market’s verdict on a product, in other words, could not substitute for the investigations the statute required.

What did the Supreme Court decide in Weinberger v. Hynson?

In 1973 the Supreme Court upheld the agency’s power to withdraw approval from drugs lacking substantial evidence of effectiveness in Weinberger v. Hynson, Westcott and Dunning, Inc., 412 U.S. 609. The Court held the agency could deny a formal hearing when a manufacturer submitted no adequate and well-controlled investigations. The decision secured the retroactive review against delaying litigation.

The dispute began with a product called Lutrexin. Hynson, Westcott and Dunning, Inc. held a new drug application for the drug that had become effective on safety grounds under the 1938 statute. After the 1962 amendments added the effectiveness requirement, the company submitted evidence of the drug’s efficacy, the National Academy and National Research Council review found that evidence inadequate, and the Commissioner moved to withdraw the approval. The company demanded a formal hearing. The question before the Court was whether the agency had to grant one when the manufacturer’s submission, on the agency’s reading, contained no evidence meeting the statutory definition of substantial evidence.

The Court answered no. Argued on April 17, 1973, and decided on June 18, 1973, the opinion held that the agency could deny a hearing where the submitted evidence did not present substantial evidence as the statute defined it, meaning evidence from adequate and well-controlled investigations by qualified experts on the basis of which such experts could fairly and responsibly conclude that the drug would have its claimed effect. Manufacturers had full and precise notice of the evidence they had to present, the Court reasoned, and no purpose would be served by a hearing whose only outcome could be the rejection of evidence that fell short of that standard. The opinion added that Congress had great leeway in setting standards for releasing drugs to the public, drugs that might be miracles or, in the Court’s phrase, merely money-making schemes dressed in mysterious scientific clothing.

The decision settled more than the hearing question. It construed the statute’s grandfather clause, which exempted from the effectiveness requirements any drug that on the day before enactment was commercially used or sold in the United States, was not a new drug as defined in the 1938 statute, and was not covered by an effective application. And it confirmed the agency’s threshold power to decide whether products were new drugs subject to its jurisdiction at all, a power the Court treated as necessarily included in the comprehensive authority Congress had conferred. Companion cases decided alongside Hynson, involving CIBA, Bentex Pharmaceuticals, and USV Pharmaceutical Corporation, applied the same principles to other withdrawal disputes. The practical consequence was that the retroactive review could proceed through administrative findings rather than through a hearing in every case, and the DESI program’s withdrawal orders gained the judicial backing that manufacturers had been seeking to deny them.

The retroactive review invites the same two readings that the amendments as a whole invite, and both are stated with their named sources in this article’s section on the innovation debate. The tension between them is the price of a statute that chose to judge every drug by evidence rather than by the market.

That tension, between the safety bought and the innovation forgone, is where the statute’s critics have pressed hardest, and the argument deserves its full hearing.

The debate: did the effectiveness requirement slow innovation?

The effectiveness requirement drew its most sustained intellectual attack about a decade after the Drug Amendments of 1962 took effect, and the attack came from economists rather than from physicians or regulators. Sam Peltzman, an economist at the University of Chicago, published “An Evaluation of Consumer Protection Legislation: The 1962 Drug Amendments” in the Journal of Political Economy in 1973, and expanded the argument in an American Enterprise Institute book, Regulation of Pharmaceutical Innovation: The 1962 Amendments, in 1974. Peltzman argued that the amendments had reduced the flow of new drugs to patients so sharply that the losses to sick consumers exceeded whatever was saved by keeping ineffective or dangerous drugs off the market. The claim supplied numbers to an intuition that already circulated in industry and medical circles: requiring proof of effectiveness carried a hidden price, measured in therapies that arrived late or never arrived at all.

Peltzman’s work did not simply assert a decline in drug introductions. It constructed a formal comparison between the drug market that the amendments appeared to have created and the drug market that would have existed without them. That counterfactual construction is the source of both the study’s influence and its vulnerability, and every serious response in the decades that followed has been an argument about how much of the observed decline in new drugs belonged to the statute and how much belonged to everything else that was changing in pharmaceutical research at the same time.

What did Peltzman measure about the flow of new drugs?

Peltzman measured the annual flow of new chemical entities reaching the market before and after the amendments took effect. Using a model fitted to pre-1962 data, he estimated that the law reduced introductions from about 41 to about 16 new chemical entities per year, a decline of roughly 60 percent, and added at least two years to development time.

The method deserves attention because the conclusions depend on it. Peltzman built an economic model of drug development and estimated its parameters from data gathered before the amendments. The model reproduced the major changes in drug innovation during the pre-amendment period, which gave him reason to trust its structure. He then ran the same model forward into the post-amendment years to generate a prediction of what innovation would have looked like had the law not changed. The gap between that prediction and the observed flow of new drugs was persistent, and Peltzman attributed the difference to the regulation.

He was careful about what counted as a new drug. A new chemical entity, meaning a genuinely new molecule put to therapeutic use, was distinguished from recombinations of existing drugs and from new dosage forms of drugs already on the market. Following earlier work by the economist Joseph Schnee, Peltzman converted the mixed flow of introductions into a single unit, the new chemical entity equivalent, defined as one new chemical entity plus three tenths of a new combination product plus sixteen hundredths of a new dosage form, with the weights drawn from Schnee’s estimates of the relative research cost of each type. The decline he reported was a decline in this weighted measure, not in a raw count of approvals.

The point estimate was striking. Annual introductions of new chemical entities fell, in his estimate, from 41 per year to 16 per year, a reduction of about 60 percent, with a standard error of under two and a half entities. Summarized differently, he found about 25 fewer new drugs per year as a result of the amendments. He also noted that American introductions of new chemical entities fell 54 percent relative to British introductions over the same period, which made it harder to explain the whole decline as a natural exhaustion of research opportunities, since British firms worked from the same body of science.

Peltzman also measured the cost side in research terms. Before 1960, real research and development spending per new chemical entity equivalent had been rising at about 14.8 percent per year, a trend he extrapolated forward. The extrapolated values consistently underpredicted what was actually spent after 1962. For the years from 1965 to 1969, the extrapolated spending averaged about half the actual spending, which implied that the amendments had roughly doubled the research cost of producing one new drug. The extra cost came to between five and ten million dollars per new chemical entity equivalent, which Peltzman noted was about a year’s sales for a fairly successful new drug. On development timing, he put the added delay at a minimum of two years, treating that figure as conservative.

How did he weigh the benefits against the costs?

Peltzman set the savings from keeping ineffective and dangerous drugs off the market against the losses from forgone and delayed therapies. He concluded the gains were small relative to the costs, and put the net burden at the equivalent of a tax of at least 300 million dollars per year on the sick.

The benefit side of his ledger was real but, in his accounting, modest. The amendments were intended to reduce the number of new chemical entities by screening out ineffective ones, and Peltzman examined drug evaluations by the American Medical Association to test whether that screening produced measurable gains. Those evaluations showed that drugs designated ineffective had been more common before 1963, and that such drugs lost market share after introduction. So the mechanism Congress had in mind appeared to operate: some drugs that would previously have been sold were not sold, and some drugs that were sold found thinner demand.

The cost side was larger and more varied. Peltzman listed the elimination of some effective drugs, which never reached the market because higher development and approval costs made them uneconomic to pursue; delay in the introduction of major drug innovations, which postponed benefits for patients who needed them; higher drug prices, because fewer treatment alternatives meant weaker competition facing sellers of existing drugs; and more costly, and therefore less abundant, drug information for prescribers and patients. The central measure of cost was forgone consumer surplus, the value patients would have captured from the drugs that were never developed.

To weigh the benefit, Peltzman compared the growth of market shares of drugs launched before 1962 with the growth of market shares of drugs launched after 1962. His assumption was that if the amendments raised the average efficacy of new drugs relative to drugs already on the market, the newer products should capture greater initial market share. He found that the gains from stiffer regulation arose mainly from reducing the number of ineffective drugs and eliminating some drugs that would otherwise have caused major harm, but that against these gains stood the larger losses from the missing and delayed drugs. Even under assumptions he described as favorable to the amendments, he could not find that they had been beneficial on net. In the book version of the study, he stated the net cost as the equivalent of a tax of at least 300 million dollars per year imposed on the sick.

Who contested the estimates, and on what grounds?

Henry Grabowski, John Vernon, and Lacy Glenn Thomas, in a 1978 study in the Journal of Law and Economics, accepted that drug introductions fell but argued the decline could not be attributed entirely to the amendments. They pointed to shrinking research opportunities, heightened risk aversion after thalidomide, and rising research costs that operated independently of regulation.

Their study, titled “Estimating the Effects of Regulation on Innovation: An International Comparative Analysis of the Pharmaceutical Industry,” appeared in volume 21, number 1 of the journal, on pages 133 to 163. It remains the most prominent early attempt to test Peltzman’s attribution rather than merely his arithmetic. Grabowski, Vernon, and Thomas confirmed the raw decline. They reported that introductions of new chemical entities fell from 233 in the five year period from 1957 to 1961 to 93 in the period from 1962 to 1966 and to 76 in the period from 1967 to 1971. They added a detail that cut against the simplest defense of the statute. The share of total ethical drug sales accounted for by new chemical entities declined roughly in proportion to the number of drugs, from 20.0 percent in the 1957 to 1961 period to 5.5 percent in the 1967 to 1971 period. If the amendments had merely eliminated the most trivial and insignificant drugs, as a defender might claim, the sales share should have fallen by less than the count did. The parallel fall suggested that the missing drugs were not all inconsequential.

But the authors refused to treat the entire decline as the statute’s doing. They attempted to measure the marginal reduction plausibly attributable to the American amendments after controlling for other forces that could have reduced drug introductions at the same time. They listed three in particular. First, the depletion of new product opportunities, meaning that the easiest therapeutic targets had been exploited in the fertile postwar years and the remaining ones were harder. Second, the thalidomide tragedy itself, which they argued may have made manufacturers and physicians more risk averse, reducing demand for new drugs and raising internal testing standards even without any legal change. Third, pharmacological advances that may have raised research and development costs independently of regulation, as the science of drug development grew more complex.

Their most important device was a comparison with Britain. The United Kingdom had a strong and successful research based pharmaceutical industry, which made it a credible comparator, and it had not adopted efficacy requirements until 1971. Its 1963 safety requirements were statistically unrelated to the flow of new discoveries. That structure offered something close to a natural experiment: two countries working from similar science, one with the American style efficacy requirement and one without it during the crucial years. By comparing the trends in new chemical entity discoveries across the two countries, Grabowski, Vernon, and Thomas sought to isolate how much of the American decline survived when the common forces were differenced out.

Other economists pressed the critique on different fronts. Peter Temin, in a 1979 study, questioned Peltzman’s methods for estimating costs and benefits, arguing that the benefits may have been understated and the costs overstated by ascribing the decline in new chemical entities solely to regulation. A 2005 review by the economist Patricia Danzon, surveying the evidence on safety and efficacy regulation for the National Bureau of Economic Research, summarized the state of the argument this way. Peltzman’s 1973 study remained the only significant attempt to weigh both the benefits and the costs of the amendments, and his conclusion that the costs dominated depended critically on his methods, which had been questioned. Danzon noted that the study stayed important because it offered a theoretical and empirical framework for evaluating the efficacy requirements, even as its estimates remained contested.

What is the drug lag, and how did America compare with Britain?

The drug lag named the pattern in which medicines reached patients in Britain before reaching patients in the United States. William Wardell documented the gap in a 1973 study comparing the two countries, and a 1978 analysis found that France, Britain, and Germany were introducing almost three times as many new drugs per year as the United States.

Wardell, a physician and pharmacologist, published “Introduction of New Therapeutic Drugs in the United States and Great Britain: 1962 to 1971” in Clinical Pharmacology and Therapeutics in 1973. Working with Louis Lasagna at the University of Rochester, and later as adjunct scholars at the American Enterprise Institute, Wardell and Lasagna documented what they called the drug lag: the difference between drugs available earlier in the United Kingdom than in the United States. The term caught on because it gave a name to an experience physicians already reported, that of reading about a useful new therapy in the European literature and being unable to prescribe it at home.

Wardell returned to the comparison in a 1978 article in the Journal of the American Medical Association, and the findings sharpened the picture. France, Britain, and Germany were on average introducing almost three times as many new drugs each year as the United States. The lag was not only about first introductions. Approval for additional uses of an already marketed drug could also trail by years. Wardell’s work cited propranolol, a beta blocking drug, for which the American approval for treatment of angina as well as for hypertension came about seven years after the corresponding British action.

The drug lag literature was not a simple indictment. Wardell himself noted an important limit in the comparison. For the period from 1972 to 1978, mutually available drugs, meaning drugs that by the end of the observation period were available in both the United States and Britain, accounted for less than half of the total number of new chemical entities introduced into either country. The two markets were choosing different drugs, not merely sequencing the same drugs differently, which complicated any claim that every missing American drug was a British drug delayed.

Other evidence complicated the attribution story further. Research costs and development times were rising in Britain too. A survey of products licensed in the United Kingdom from 1971 to 1981, published in the British Journal of Clinical Pharmacology, found that development times for new drugs originated by British owned firms had increased from about four and a half years in the 1960s to nine years in the 1970s, with a corresponding reduction in effective patent life. If longer development was partly a global phenomenon, driven by advancing science and heightened caution after thalidomide, then not all of the American slowdown could be laid at the feet of the American statute.

How did later reviews assess the argument?

Later reviews treated Peltzman’s 1973 study as the founding attempt to weigh the costs and benefits of the effectiveness requirement while questioning its methods. A 2005 survey by Patricia Danzon for the National Bureau of Economic Research noted that the benefits may have been understated and the costs overstated by ascribing the entire decline in new drugs to regulation.

The broader economic literature absorbed the debate without settling it. A 1993 review in the Journal of Economic Perspectives, surveying the rising cost of drug development, reported that the economist Joseph DiMasi and his colleagues had estimated that bringing a successful drug of 1980s vintage to market entailed research, development, and testing outlays averaging 53 million dollars in the clinical stage plus 73 million dollars at the preclinical stage, in 1990 dollars and counting the cost of failures, with 23 percent of new chemical entities put into human testing surviving all hurdles to approval and the average time from the start of clinical trials to approval exceeding eight years. The review judged that some of the observed cost increase would have occurred even without legal and regulatory changes, as drugmakers tightened their own testing to reduce liability exposure, but attributed roughly half to tougher regulation, citing Grabowski and his colleagues and a 1990 study by Lacy Glenn Thomas.

In 2022, the economist Casey Mulligan returned to Peltzman’s original question in the Journal of Law and Economics, in an article titled “Peltzman Revisited: Quantifying 21st-Century Opportunity Costs of Food and Drug Administration Regulation.” Mulligan applied an excess burden framework to measure the opportunity cost of regulatory delays, including substitution toward remedies outside the agency’s jurisdiction that need not demonstrate safety or effectiveness. His restatement of Peltzman’s point estimate, a reduction from 41 to 16 new chemical entities per year with a small standard error, and of the 54 percent fall of American introductions relative to British ones, showed how durable the original numbers remained as reference points even as the methods around them evolved.

The honest treatment of the debate requires presenting both of the moral claims that sit beneath the numbers, because they are genuinely in tension. The safety case holds that the effectiveness requirement, together with the retroactive review of drugs approved between 1938 and 1962, removed from the market drugs that had no demonstrated value and spared patients the cost and risk of taking them, and that the systematic clinical evidence the law demanded made it harder for another thalidomide pattern to pass undetected. The access case holds that the same requirement delayed therapies for patients who could have used them and deterred the development of drugs that would have helped, and that those losses, precisely because the missing drugs cannot be examined, are systematically undervalued in public argument. Both claims can be true at once. The literature has not produced a consensus measure of the trade, because measuring the trade requires observing the drugs that were never developed, and that counterfactual remains, as Mulligan’s retrospective acknowledged, the permanent difficulty at the center of the argument.

Congress did not leave the trade where the 1962 statute set it. Three later interventions renegotiated it, and the approval system they inherited is the amendments’ most durable legacy.

The approval system the amendments built

The price provisions that Senator Kefauver cared about most died on the way to enactment, and the antitrust machinery he had spent years assembling was dismantled in the compromises. What lived was the thing his hearings had never been about: a regime that made proof of effectiveness a condition of market access in the United States. That regime became the load-bearing wall of every later argument about how drugs reach patients, because every subsequent statute in this field had to work through it, against it, or around it. The companion article on FDA approval law mechanics traces the modern pathway in depth; this section records how the 1962 amendments built the foundation it stands on, and how three later interventions answered the pressures that foundation created.

Before the Drug Amendments of 1962, the Federal Food, Drug, and Cosmetic Act of 1938 demanded proof of safety but nothing more, and its procedure barely slowed a determined sponsor. A new drug application became effective automatically after sixty days unless the agency intervened, a structure that put the burden of stopping a product on the government rather than the burden of winning approval on the manufacturer. The amendments reversed that default. After October 10, 1962, no drug reached the market without an affirmative agency decision, rendered within one hundred eighty days or as the review required, and no drug won that decision without substantial evidence of effectiveness drawn from adequate and well-controlled investigations. The FDA history office describes the change as the moment the agency began requiring manufacturers to prove that new drugs worked, not only that they would not poison anyone.

How did the amendments change the way drugs were approved?

The 1962 amendments required every new drug to clear two gates before reaching the market. A sponsor first obtained agency permission to test the drug in people under an Investigational New Drug filing, then submitted a New Drug Application proving safety and effectiveness before selling a single tablet.

The investigational and application machinery behind those two gates is examined in this article’s section on the effectiveness standard, as is the Drug Efficacy Study Implementation that turned the new standard backward on the thousands of products already on pharmacy shelves, and the consent, adverse event reporting, manufacturing, inspection, and expert review requirements are taken up in this article’s section on the other requirements. What matters here is what the reorganization meant: the statute replaced a notification regime, in which the government had to catch a dangerous product in time, with a permission regime, in which the developer had to earn entry, and it built an approval system dense enough that later lawmakers took its shape for granted as the natural form of drug regulation, which is how a revolution hides inside an institution.

Did the effectiveness standard slow drug development?

The record gives ammunition to both sides. Review times lengthened after 1962 as the agency applied the new effectiveness standard, and development costs rose. Yet the safety argument points to withdrawn and blocked drugs, so the question remains contested rather than settled.

The argument that stringency cost patients access to medicines rests on timing and on counts, and its strongest evidence comes from the decades when review times were at their longest. The Tufts Center for the Study of Drug Development compared the late 1980s with the late 1990s and found that approval times for priority-reviewed new chemical entities, drugs promising significant improvement for serious conditions, had fallen to about seven months from about twenty eight months, while standard-reviewed new chemical entities fell to about seventeen months from thirty five to thirty seven months. The decline was sharp, and it dated from the Prescription Drug User Fee Act of 1992, which let the agency charge sponsors fees and hire reviewers. A 2002 Government Accountability Office report on the user fee program measured the same drop from the agency’s side: median approval time for standard new drug applications fell from twenty seven months in 1993 to fourteen months in 2001, while priority applications held steady at six months from 1997 onward. An analysis by Ernst Berndt, Adrian Gottschalk, and Tomas Philipson, published as a National Bureau of Economic Research working paper in 2004, estimated that approval times fell forty two percent between 1991 and 2002 and attributed roughly two thirds of that decline to the user fee program, while noting that times had already been drifting downward before the program began. The access reading of this evidence holds that the 1962 standard, applied with limited staff, created a backlog that kept useful medicines from patients, and that money for reviewers, not a lower bar, fixed it.

The safety reading answers with a different set of counts. Sidney Wolfe, director of the Health Research Group at Public Citizen, counted nine drugs approved between 1993 and 2000 that were later withdrawn for safety reasons, among them the cholesterol drug Baycol, the heartburn drug Propulsid, and the diabetes drug Rezulin, against five withdrawals among drugs approved in the eight years before the user fee law took effect. Wolfe argued that industry funding of the review process had damaged its rigor. The safety position further holds that the drugs the effectiveness standard blocked or delayed represent harms prevented rather than benefits lost, and that the thalidomide episode which supplied the amendments’ momentum is the standing example of what a permissive regime lets through. A 2024 study of FDA approvals from 1980 to 2022, published in Scientific Reports, found that median review time had fallen from 26.6 months before the user fee era to 9.9 months after the 2012 Food and Drug Administration Safety and Innovation Act, while approvals grew fastest in cancer, immunology, biologics, and orphan drugs. That study also found that more than half of recent approvals traveled expedited pathways that did not require demonstrating therapeutic benefit over available alternatives. Presented with equal care, the two readings do not reconcile. One counts the cost of delay in medicines patients never received; the other counts the cost of speed in medicines patients had to stop taking. The 1962 amendments made that trade-off the permanent structure of American drug politics, which is why nearly every later statute in this field is best read as an attempt to renegotiate it.

What the later statutes answered

How did Congress answer the cost and delay complaint?

Congress split the difference in the Drug Price Competition and Patent Term Restoration Act. Generics won an abbreviated application proving bioequivalence instead of repeating full trials, while brand sponsors won restored patent time lost to regulatory review and a formal process for litigating patent disputes before approval.

The 1984 act, Public Law 98-417, sponsored by Senator Orrin Hatch and Representative Henry Waxman, is the statute that most directly answered the charge that the 1962 standard had made drug development too slow and too expensive. A Congressional Research Service analysis described it as a complex and multifaceted compromise between innovative and generic pharmaceutical companies, and the structure of the law earns that description. Before 1984, a company wishing to sell a generic copy of an approved drug had to run the full approval gauntlet nearly from scratch, repeating clinical trials whose outcome was already known, which meant the effectiveness standard protected brand-name products long after their scientific novelty had worn off. The act created an Abbreviated New Drug Application under section 505(j) of the Food, Drug, and Cosmetic Act, codified at 21 U.S.C. 355(j), through which a generic applicant won approval by demonstrating that its product was bioequivalent to the approved brand drug, without supplying its own clinical evidence of safety and effectiveness. The same title of the law added a second shortcut, the section 505(b)(2) application, sometimes called the paper new drug application, which contained full reports of safety and effectiveness but could rely in part on published literature rather than new trials. Together the two pathways let competitors stand on the pioneer’s data, which collapsed the cost of entry that the 1962 standard had raised.

What the brand sponsors received in return was time and procedure. The Congressional Research Service reports detail the patent term restoration mechanism: a patent covering a drug that lost years to clinical testing and agency review could be extended by a period reflecting half the time between the Investigational New Drug filing and the New Drug Application submission plus all of the agency’s review time, capped at five years, with the restored patent’s remaining term after approval never exceeding fourteen years. The extension was not automatic; the patent owner had to apply to the patent office within sixty days of marketing approval and had to show diligence in pursuing it. The statute also created the machinery that governs brand-generic patent conflict to this day. Brand sponsors listed their patents in the FDA’s Orange Book, the register of Approved Drug Products with Therapeutic Equivalence Evaluations, and each generic applicant had to certify against those patents under one of four paragraphs: paragraph one when no patent was listed, paragraph two when all listed patents had expired, paragraph three when the applicant would wait out the patents, and paragraph four when the applicant claimed the patents were invalid or would not be infringed. A paragraph four certification let the first applicant to file win one hundred eighty days of exclusivity against other generics, and it gave the brand sponsor forty five days to sue for infringement, with the agency’s approval of the generic stayed for thirty months while the dispute was litigated. The act also added the research safe harbor that answered the Federal Circuit’s 1984 decision in Roche Products v. Bolar Pharmaceutical, so generic firms could conduct the testing needed for approval before the relevant patents expired. The guide article on the Hatch-Waxman Act of 1984 treats this bargain at full length. Its shape matters here because it shows what the 1962 amendments had made necessary: a separate statute to create competition that the effectiveness standard had incidentally suppressed, and a separate set of patent rules to compensate the innovators for the delay that standard imposed. A Health Affairs analysis of the patent system, drawing on a Federal Trade Commission working paper, reported that once multiple generics entered a market, prices fell to as much as ninety percent below the brand price. That price cliff is the measure of what the 1984 compromise bought, and it became the economic argument for every later expansion of the abbreviated pathway.

Why did rare diseases need their own statute?

A drug that works would still lose money if only a few thousand people need it. The Orphan Drug Act answered that arithmetic directly, offering seven years of market exclusivity, tax credits, and research grants for drugs treating conditions that affected fewer than two hundred thousand Americans.

The Orphan Drug Act of 1983 addressed the class of patients the 1962 standard had most visibly left behind. A rare disease, however devastating, offered too few customers to justify the cost of the clinical trials the amendments demanded, so sponsors simply did not develop drugs for it. Public Law 97-414, chiefly sponsored in the House by Henry Waxman as H.R. 5238, passed the House on December 14, 1982 and the Senate by voice vote on December 17, and President Ronald Reagan signed it on January 4, 1983. Its definition drew a bright line: a drug qualified as an orphan if it treated a disease or condition affecting fewer than two hundred thousand people in the United States. For such drugs the act assembled the most generous incentive package in the drug code. The first sponsor to win approval for a designated orphan use received seven years of market exclusivity, during which the agency could approve no competing version of the same drug for the same condition. The sponsor could claim a tax credit for fifty percent of qualified clinical testing expenses, a credit Congress later reduced by half in 2017. The act funded research grants for orphan product development, offered protocol assistance to small sponsors navigating the agency, expanded access to investigational drugs for rare disease patients, and was later amended to waive the user fees otherwise charged under the prescription drug user fee program. The guide article on the Orphan Drug Act of 1983 covers these mechanics in full.

The incentive logic met the arithmetic head on, and the counts record the result. Before the act, only thirty eight drugs had ever been approved in the United States for rare diseases, and fewer than ten reached the market in the decade before 1983, according to the National Organization for Rare Disorders. By the act’s thirtieth anniversary, roughly four hundred twenty orphan drugs had been approved across twenty seven hundred thirty designations, and the organization’s federal policy director later counted more than six hundred orphan drugs developed since the law’s passage. About nine percent of those drugs went on to reach blockbuster sales, including products for conditions that proved larger than their original designations. That the same Representative Waxman who sponsored the orphan incentives in 1983 co-sponsored the generic pathways a year later is more than coincidence; it marks the two statutes as companion answers to the same problem, the one subsidizing drugs the market would not produce under the 1962 standard, the other cheapening drugs the 1962 standard had protected from competition.

The sequence and the advertising settlement

The 1962 amendments belong to the earliest layer of modern American health legislation. They amended the 1938 food and drug statute and preceded the 1965 creation of Medicare and Medicaid by three years, which places them at the head of the post-1950 sequence surveyed in the article on health legislation since 1950, ahead of the managed care statutes of the 1970s, the 1996 Health Insurance Portability and Accountability Act, the 2003 Medicare drug benefit, and the 2010 Affordable Care Act. Because they came first, later health laws inherited their definitions, their agency, and their standard of proof without having to legislate any of it again. That inheritance is the quiet part of the legacy: the amendments set the evidentiary vocabulary, substantial evidence, adequate and well-controlled, that every later drug statute speaks.

The final piece of the settlement concerned how medicines could be sold in words, and the transfer of prescription drug advertising oversight from the trade commission to the drug agency is examined in this article’s section on the other requirements. The new authority carried substantive rules: prescription drug advertisements had to include a brief summary of side effects, contraindications, warnings, and directions for use, and had to maintain what the law called fair balance between claims of effectiveness and disclosures of risk. That balance, administered through agency guidance, governed prescription drug promotion for decades and supplied the framework within which direct-to-consumer advertising later grew, after 1997 guidance from the agency permitted broadcast advertisements that did not devote equal time to promotion and risk. The article on FTC consumer protection rulemaking covers the commission’s side of this history.

The statute that Congress passed as an antitrust bill about drug prices ended its legislative life as the effectiveness law, and its sponsor’s signature provisions did not survive. That is the caution the namable claim requires. The amendments are remembered as the charter of modern drug safety, but they were not written to be one, and the parts of them that were written on purpose, the pricing and patent provisions Kefauver fought for, either died in committee or were stripped on the floor. A reader who starts from the statute’s reputation and reasons backward to its intent will get the story exactly wrong. What endures from 1962 is not a purpose but a procedure: the two-gate approval system, the effectiveness standard, and the trade-off between safety and access that every later statute, the abbreviated pathways of 1984, the orphan incentives of 1983, the user fee programs, and the advertising rules, has spent the decades since trying to price correctly.

The statute that requires American drugs to prove they work was never meant to be that statute, and the assessment has to start there. A Senate antitrust subcommittee spent years building a case that drug prices were monopoly prices, a sponsor wrote a bill to break the patent and promotion machinery behind them, and the industry killed the machinery in committee. What survived was the residue: a repurposed bill whose effectiveness standard, affirmative approval, investigational controls, consent rules, manufacturing standards, and advertising jurisdiction became the operating system of American drug regulation. The provisions Kefauver cared about most did not survive; the provisions nobody had campaigned for became the law’s entire reputation. That is the caution the record forces on any reader tempted to read legislative purpose backward from a statute’s fame.

The honest measure of the amendments is the trade they installed. The effectiveness standard filtered out drugs of no proven value, built the controlled trial into the shared language of medical evidence, and gave reviewers the power to say not yet with the law behind the pause. It also raised the cost of development, lengthened the path to market, and left the hardest question unanswerable, because the drugs never developed cannot be examined. The economic literature has argued about that trade for fifty years without settling it, and the absence of a settlement is itself a finding: no statute that judges every product by evidence rather than by the market can avoid choosing between the patients it protects and the patients it delays.

Readers who want to follow the threads outward have a mapped cluster. The form the two-gate system took in the decades that followed, the Investigational New Drug filing and the New Drug Application as they operate in later practice, is traced in the companion guide to FDA approval law mechanics. The 1984 bargain that answered the cost and delay complaint, abbreviated generic approval in exchange for restored patent time, is the subject of the guide to the Hatch-Waxman Act. The incentive answer for the patients the effectiveness standard priced out, seven years of exclusivity and the tax credit for rare conditions, is covered in the guide to the Orphan Drug Act. And the place of the 1962 amendments at the head of the post-1950 health legislation sequence, ahead of Medicare and everything after, is set out in the survey of American health legislation since 1950.

The final assessment is simpler than the literature. Before 1962, an American drug had to be safe; after 1962, it had to work, and the maker had to prove it before selling a tablet. Every later argument about drug regulation, speed against rigor, access against evidence, price against incentive, has been an argument about where to set the dials on machinery this statute built. Readers keeping keep your drug-law study notes together on VaultBook will find the comparison table and the passage timeline above the natural anchors for those notes, because the statute’s two histories, the antitrust campaign that failed and the catastrophe that redirected it, are the two things any serious account of American drug law has to hold together.

Frequently Asked Questions

Q: What did the Kefauver-Harris amendments require?

The amendments required every new drug to be proven both safe and effective before marketing, replacing the 1938 regime that had demanded only safety. Effectiveness had to be shown by substantial evidence from adequate and well-controlled investigations, which created the modern clinical trial system. The statute also replaced the sixty day default approval with a requirement of affirmative agency approval, brought human testing under agency oversight through the investigational drug exemption, required informed consent from research subjects, imposed adverse event reporting on manufacturers, authorized manufacturing standards and factory inspections, and transferred jurisdiction over prescription drug advertising from the trade commission to the drug agency. It further ordered a retroactive review of the thousands of drugs approved between 1938 and 1962, removing those that could not meet the new effectiveness standard.

Q: How did thalidomide lead to the Kefauver-Harris amendments?

Senator Estes Kefauver’s drug bill had begun as an antitrust measure about prices and patent practices, and by the spring of 1962 it was stalled in committee with its most contested provisions stripped away. The thalidomide catastrophe broke into American public view that summer, when a Washington Post front page story reported that an agency medical officer, Dr. Frances Oldham Kelsey, had repeatedly refused to approve the sedative for the American market while it was injuring thousands of babies in Europe. The story made Kelsey a national figure, President Kennedy honored her in August, and the political situation inverted within weeks. Kefauver reworked the bill around safety and effectiveness provisions drafted by the agency itself, the price provisions largely fell away, and the amended measure passed both chambers in early October and was signed on October 10. The statute that requires American drugs to work is therefore the residue of a failed price bill, rescued by a catastrophe its sponsor had never legislated about.

Q: Who was Frances Kelsey in the Kefauver-Harris story?

Dr. Frances Oldham Kelsey was the agency medical officer whose refusal to approve thalidomide for the American market became the moral center of the 1962 amendments. A physician and pharmacologist trained at the University of Chicago, she had joined the agency in 1960 and received the American thalidomide application as her first drug assignment. Finding the clinical studies incomplete and the file unsupported, she repeatedly demanded more data, restarting the sixty day review clock with each request, and she pressed for information on the nerve injury reports accumulating in Europe. Her refusals kept the drug out of American pharmacies from September 1960 until the manufacturer withdrew its application in March 1962. When the European birth defects became public that summer, newspaper coverage made her a national figure, and President Kennedy presented her with the President’s Award for Distinguished Federal Civilian Service in August 1962.

Q: Was the Kefauver-Harris bill originally about drug prices?

Yes. The bill Senator Estes Kefauver introduced in 1961 as S. 1552 was principally an antitrust and competition measure, the legislative product of years of hearings by his Antitrust and Monopoly Subcommittee on pharmaceutical pricing and patent practices. Its headline provisions were compulsory licensing of drug patents after three years, tougher patent standards for modified compounds, generic naming requirements, and regulation of drug advertising. Through the spring of 1962 the industry fought the bill in the Judiciary Committee, compulsory licensing was set aside for further study, and the measure was losing its most contested provisions. The safety and effectiveness machinery that defines the enacted statute was not in the original design; it was added in the summer of 1962 after the thalidomide catastrophe transformed the bill’s politics, and the price provisions largely fell away in the substitution.

Q: What is substantial evidence of effectiveness under Kefauver-Harris?

Substantial evidence is evidence consisting of adequate and well-controlled investigations, including clinical investigations, conducted by experts qualified by scientific training and experience, on the basis of which those experts could fairly and reasonably conclude that the drug has the effect its labeling claims under the stated conditions of use. The definition, written into section 505(d) of the food and drug statute, ties effectiveness to the labeled claims rather than to any free-floating notion of therapeutic value, and it makes the quality of the proof, not the sincerity of anyone’s belief, the test. The agency read the plural investigations as a general expectation of more than one adequate and well-controlled study, each convincing on its own, a reading that held for decades. The standard never required comparing the drug against competing treatments; the question was whether the medicine did what its own labeling said, leaving comparative judgment to the prescribing physician.

Q: What happened to drugs approved before Kefauver-Harris?

Congress did not limit the new effectiveness standard to future drugs. The amendments directed the agency to review the effectiveness of the thousands of drugs approved on safety grounds alone between 1938 and 1962, working with the National Academy of Sciences, and to remove those lacking substantial evidence. Manufacturers received a two-year grace period to develop the evidence before the agency could withdraw approvals for lack of effectiveness. The agency contracted with the academy in 1966, panels evaluated more than 3,000 products and 16,000 therapeutic claims, and the findings were implemented through the Drug Efficacy Study Implementation program. By 1984 the agency had taken final action on 3,443 products, finding 2,225 effective and 1,051 not effective, and its historical account records that more than 1,000 ineffective drugs and combinations left the marketplace, with thousands more relabeled to match their demonstrated qualities.

Q: Did Kefauver-Harris slow down new drug approvals?

The question has been contested since the first years of the statute’s operation. The economist Sam Peltzman argued in a 1973 study that the amendments reduced new drug introductions from about 41 to about 16 new chemical entities per year, roughly doubled the research cost per new drug, and added at least two years to development time, with the forgone benefits exceeding the safety gains. The pharmacologist William Wardell documented the drug lag, the pattern of medicines reaching Britain before the United States. Other economists contested the attribution: Henry Grabowski and his colleagues pointed to shrinking research opportunities, post-thalidomide risk aversion, and rising costs independent of regulation, and Peter Temin questioned Peltzman’s methods. The literature never produced a consensus measure, because measuring the trade requires observing the drugs that were never developed, a counterfactual that remains the permanent difficulty at the center of the argument.

Q: How did Kefauver-Harris pass Congress?

Kefauver’s bill had been stalled and stripped through the spring of 1962 when the thalidomide catastrophe broke into public view that summer. He reworked the legislation around safety and effectiveness provisions drafted by the drug agency itself, with Representative Oren Harris as the chief House cosponsor, and the contested price and compulsory licensing provisions largely fell away. The Senate passed an amended version in August 1962 with recorded dissent. The final conference report then moved through both chambers in the first days of October without recorded dissent, the House agreeing on October 4 and the Senate concurring, and President John F. Kennedy signed the bill on October 10, 1962, as Public Law 87-781. The speed of the ending remains the measure of the catalyst’s force: a dying price bill completed its passage within weeks of the summer’s newspaper stories and the August award to Frances Kelsey.

Q: What did the informed consent requirement change for clinical trials?

Before the amendments, the 1938 statute governed the marketing of a drug and said almost nothing about its investigation, so a manufacturer could distribute an unapproved drug to physicians for investigational use with no required protocol, no required informed consent, and no required records; more than a thousand American doctors received thalidomide tablets this way. The amendments required investigators to obtain informed consent from human subjects, except where the investigator judged consent not feasible or contrary to the subject’s best interests, so that no patient would again receive an experimental compound without being told it was experimental. Investigators also had to maintain personal supervision of clinical investigations and keep records. Later rulemaking built on the provision, adding institutional review board review and special protections for prisoners and institutionalized persons.

Q: Why did the amendments move prescription drug advertising to the drug agency?

Before 1962, prescription drug advertising belonged to the Federal Trade Commission as a consumer fraud matter, an arrangement that had survived the drafting of the 1938 act because industry insisted the commission, not the drug agency, should police marketing. The Kefauver hearings had spotlighted aggressive physician-directed promotion, including the one-sided medical journal advertising the subcommittee exhibited. The amendments transferred jurisdiction over prescription drug advertising to the Food and Drug Administration, the agency that reviewed the clinical investigations and decided what a drug could lawfully claim in its labeling, so that promotional claims would be judged by the same body that evaluated the evidence behind them. Over-the-counter advertising stayed with the trade commission. The new authority carried substantive rules requiring a brief summary of risks and what the law called fair balance between effectiveness claims and risk disclosures.

Q: What are good manufacturing practices under the amendments?

The amendments authorized the drug agency to set standards for how drugs were made, going further than any earlier federal drug law in regulating the factory itself. The first manufacturing regulations followed in 1963, requiring establishments to abide by current good manufacturing practices designed to ensure that drugs met standards of identity, strength, quality, and purity. The hearings that produced the amendments had prepared this ground: clinical pharmacologist Louis Lasagna advised Congress during the 1962 hearings on the proper conduct of clinical research, and the proceedings prompted the recognition that good manufacturing practices had to cover the quality of the process as well as the quality of the finished product. Congress paired the standards with inspection powers, giving the agency enhanced authority to examine company production and control records to verify compliance as a matter of routine.

Q: What was the Drug Efficacy Study Implementation?

The Drug Efficacy Study Implementation, known as DESI, was the agency’s program for carrying out the amendments’ retroactive review of drugs approved on safety grounds alone between 1938 and 1962. In 1966 the agency contracted with the National Academy of Sciences and National Research Council to evaluate effectiveness; thirty panels of six professionals each, organized by drug category, reviewed more than 3,000 products and 16,000 therapeutic claims, sorting each as effective, ineffective, or needing further study. The agency formed the DESI program in 1968 to turn the panels’ recommendations into regulatory action. By 1984 final action had been completed on 3,443 products, of which 2,225 were found effective and 1,051 not effective. The program removed more than 1,000 ineffective drugs and combinations from the marketplace, and its machinery helped produce the Abbreviated New Drug Application for generics.

Q: What did Sam Peltzman argue about the amendments?

The economist Sam Peltzman, in a 1973 Journal of Political Economy article expanded into a 1974 book, gave the access critique of the amendments its most cited early form. He estimated that the law reduced new drug introductions from about 41 to about 16 new chemical entities per year, a decline of roughly 60 percent, added at least two years to development time, and roughly doubled the research cost of producing one new drug. Weighing benefits against costs, he concluded that the gains from screening out ineffective drugs were small relative to the losses from forgone and delayed therapies, and he put the net burden at the equivalent of a tax of at least 300 million dollars per year on the sick. His study remains the only significant attempt to weigh both sides of the trade, though later economists contested his methods and the counterfactual on which the estimates rested.

Q: What did Weinberger v. Hynson decide?

In Weinberger v. Hynson, Westcott and Dunning, decided in 1973, the Supreme Court upheld the drug agency’s power to withdraw approval from drugs lacking substantial evidence of effectiveness. The dispute began with a product called Lutrexin, whose manufacturer demanded a formal hearing after the agency moved to withdraw its pre-1962 approval. The Court held that the agency could deny a hearing when the manufacturer’s submission contained no adequate and well-controlled investigations, reasoning that no purpose would be served by a hearing whose only outcome could be rejection of evidence falling short of the statutory standard. The decision secured the retroactive review against delaying litigation, confirmed that the effectiveness requirement reached drugs already on the market, and established that the agency, not the prescribing profession at large, decided whether the evidence met the standard.

Q: How did the Hatch-Waxman Act change the system the amendments created?

The Drug Price Competition and Patent Term Restoration Act of 1984, sponsored by Senator Orrin Hatch and Representative Henry Waxman, was the most direct legislative answer to the charge that the 1962 standard had made drug development too slow and too expensive. It created the Abbreviated New Drug Application, through which a generic maker won approval by proving bioequivalence to an approved brand drug without repeating full clinical trials, and the section 505(b)(2) application, which could rely in part on published literature. In return, brand sponsors won restored patent time lost to regulatory review, capped at five years, and a formal process for litigating patent disputes before generic approval, including the Orange Book listings and the paragraph four certification machinery. The act thus preserved the effectiveness standard for new drugs while building the abbreviated pathways that let competition reach the market faster.

Q: How did the Orphan Drug Act respond to the effectiveness standard?

The Orphan Drug Act of 1983 addressed the patients the 1962 standard had most visibly left behind: a rare disease offered too few customers to justify the cost of the clinical trials the amendments demanded, so sponsors did not develop drugs for it. The act defined an orphan drug as one treating a condition affecting fewer than two hundred thousand Americans and assembled an incentive package for such drugs, including seven years of market exclusivity from approval, a tax credit for fifty percent of qualified clinical testing expenses, research grants, protocol assistance, and expanded access to investigational drugs. The results were large: before the act, only thirty eight drugs had ever been approved for rare diseases, while by the act’s thirtieth anniversary roughly four hundred twenty orphan drugs had been approved across nearly three thousand designations.

Q: What is an Investigational New Drug filing?

An Investigational New Drug filing, known as an IND, is the submission a sponsor makes before testing an experimental compound in people. Section 505(i) of the food and drug statute authorized the Secretary to exempt drugs shipped for investigational use from the premarketing approval requirements, and the agency’s 1963 regulations built that authority into a filing, originally called the Notice of Claimed Investigational Exemption for a New Drug, containing preclinical pharmacology and toxicology data, a general plan of investigation, the qualifications of the investigators, and manufacturing information. The exemption solved a problem the effectiveness standard created: if no drug could be marketed without proof from human studies, the law had to create a lawful way to conduct those studies first. Investigators had to keep records, report adverse events, and obtain informed consent from subjects.

Q: How long did makers of older drugs have to prove effectiveness?

Manufacturers of drugs approved between 1938 and 1962 received a two-year grace period to develop substantial evidence of effectiveness for their previously approved products. During that window, the agency could not withdraw those approvals for lack of effectiveness; after it closed, the reckoning began. The period recognized that the thousands of products on pharmacy shelves had been judged on safety alone and that their makers needed time to run the controlled investigations the new standard demanded. The agency used the interval to set up the retroactive review machinery, contracting with the National Academy of Sciences in 1966, and the Drug Efficacy Study Implementation program then carried the panels’ recommendations into regulatory action over the following years.

Q: Why did President Kennedy honor Frances Kelsey?

President John F. Kennedy presented Frances Oldham Kelsey with the President’s Award for Distinguished Federal Civilian Service on August 7, 1962, the highest civilian honor established for government employees, and she became the second woman ever to receive it. The award credited her exceptional judgment in evaluating a new drug for safety, stating that it had prevented a major tragedy of birth deformities in the United States through her high ability and steadfast confidence in her professional decision. The honor fixed her refusal to approve thalidomide in the national memory as the act that had spared the country, and it bound the thalidomide story to the fate of the drug bill in Congress. Within weeks of the ceremony, the political calculation around the legislation inverted, and the reworked bill passed both chambers.

Q: Did the amendments control drug prices?

No. The provisions Senator Kefauver cared about most, compulsory licensing of drug patents, tougher patent standards for modified compounds, and generic naming requirements, were the price and competition core of his original bill, and they did not survive to enactment. Compulsory licensing was set aside for further study in the spring of 1962, and the safety and effectiveness provisions added that summer displaced the price agenda entirely. The statute that emerged required drugs to prove they worked, controlled how they were tested and advertised, and set manufacturing standards, but it gave the government no power over what they cost. Later statutes addressed the cost question through different machinery: the 1984 Hatch-Waxman Act built the abbreviated generic pathway that collapsed prices once competition entered, and the 1983 Orphan Drug Act subsidized development where the market alone would not pay.