Eroom’s Law
Eroom’s law describes how, between 1950 and 2010, the number of new drug approvals per billion dollars of inflation-adjusted R&D fell by about half every nine years.
Eroom’s law, the name Jack Scannell and his colleagues gave in 2012 to a sobering trend, is the observation that the number of new drugs approved for each billion dollars spent on research and development, adjusted for inflation, fell by half roughly every nine years between 1950 and 2010. The coinage spells Moore’s law backward, setting the pharmaceutical laboratory against the computer chip, whose capacity doubled on a comparably steady schedule. In those decades drug discovery acquired combinatorial chemistry, high-throughput screening, and the sequenced human genome, instruments whose power improved by orders of magnitude, while its yield per dollar declined. Many who watch computing expect every field to improve at its pace, so that cures for disease and aging arrive on schedule. Eroom’s law records a field in which more money and better tools produced fewer results.
Many Medicines
The prophet Jeremiah, pronouncing judgment on Egypt after its defeat at Carchemish, addressed the nation in the language of the sickroom: “Go up into Gilead, and take balm, O virgin, the daughter of Egypt: in vain shalt thou use many medicines; for thou shalt not be cured” (Jeremiah 46:11). The taunt carried particular force against a land renowned for its physicians. Homer had praised Egypt as the country whose fertile soil bore the greatest store of drugs, where every man was a healer skilled beyond all others. Jeremiah’s word for “medicines,” rephu’ot, is the plural of rephu’ah, “remedy,” a noun from the same root as rapha, “to heal,” and the oracle sets the multiplication of remedies against the failure of the cure, the very shape of the curve Scannell would plot.
The Gospel of Mark records the same shape in the life of a single patient. A woman with an issue of blood “had suffered many things of many physicians, and had spent all that she had, and was nothing bettered, but rather grew worse” (Mark 5:26). Her expenditure rose as her health declined, precisely the relation Eroom’s law describes across an industry. Scannell and his colleagues proposed four causes for the decline. Each new drug must improve on a growing library of effective older ones, a difficulty they called the “better than the Beatles” problem; regulators grew more cautious after past disasters; firms tended to throw money at problems; and researchers came to place excessive faith in the brute-force screening of molecules against isolated biological targets.
Remedies That Reach the Cause
Alma 46:40 links healing to removing the cause of disease through “many plants and roots.”
The Book of Mormon reports that fevers were less deadly among the Nephites “because of the excellent qualities of the many plants and roots which God had prepared to remove the cause of diseases” (Alma 46:40). Alma’s phrase states the standard by which scripture measures medicine. Ezekiel’s vision of the restored temple pictures a river whose banks bear trees of which “the fruit thereof shall be for meat, and the leaf thereof for medicine” (Ezekiel 47:12), and the Revelation of John recapitulates the image in the tree of life, whose leaves “were for the healing of the nations” (Revelation 22:2). Ezekiel’s word for medicine, teruphah, belongs to the same family of healing words and occurs nowhere else in the Hebrew Bible, so that its single appearance falls within a vision of the house of the Lord.
The usual explanation of Eroom’s law holds that the easiest targets were found first. In 2016 Scannell and Jim Bosley advanced a less obvious diagnosis, locating the larger loss in the predictive validity of the models researchers use to decide which compounds to pursue. Mid-century pharmacologists tested fewer compounds, often in whole animals or in patients, and their tests forecast human benefit well. Industrial screening tests millions of molecules against isolated proteins, and in their model a small decline in how faithfully a test predicts outcomes in patients can cancel the gain from a hundredfold increase in throughput. A revelation teaches that every blessing is “predicated” upon a law and obtained “by obedience to that law” (D&C 130:20-21). The workings of the human body may be counted among such laws, discerned only by tests that reach the cause.
Wisdom and Order
King Benjamin counseled his people to “see that all these things are done in wisdom and order; for it is not requisite that a man should run faster than he has strength” (Mosiah 4:27). His counsel concerned gifts to the poor, yet it applies with some precision to an industry that sought to purchase speed. Since about 2010 the decline has halted and partly reversed, and in 2020 Scannell and his colleagues credited the change chiefly to higher success rates in clinical trials, which they traced to better evidence for choosing targets, including human genetic evidence that ties a target to a disease. Scannell has since cautioned that the rise in approvals coincides with fewer eligible patients for each new drug, owing in part to a turn toward rare diseases.
Mormon transhumanists anticipate that machine learning will hasten this recovery and turn it toward the diseases of aging, and the first results lend that hope some footing. In 2025 rentosertib, a drug whose target and molecule were both proposed by generative software, reported encouraging results in a small mid-stage trial for idiopathic pulmonary fibrosis, a scarring disease of the lungs. The danger lies in the pattern Scannell diagnosed. Machine learning applied to the same poorly predictive screens would only accelerate the production of failures, running faster than the evidence has strength. A second danger concerns access, since drugs for small populations often reach approval at prices that place them beyond many who need them. Chris Bradford has asked that life extension, when it arrives, be extended equitably.
Eroom’s law counts new drugs per billion dollars, and set beside Jeremiah’s oracle and Mark’s account the count measures how far the multiplication of remedies has drifted from the work of healing. Egypt’s many medicines and the woman’s many physicians failed to reach the cause of the malady, and the industry’s costly screens, by Scannell’s account, predicted little about patients. The Restoration expects the leaves of the tree of life to heal the nations, and Mormon transhumanists hope that genetic evidence and learning machines will bend the curve toward treatments for aging. Bending it requires tests faithful to the laws of the body and a distribution that reaches the poor, so that the count of remedies and the number of persons healed rise together.