New Haven, 1981. A cassette recorder sits under a crib in the home of two Yale economists. The child in the crib is not yet two. After her parents leave the room, she talks. She reviews the day. She rehearses tomorrow. She works out who went where and why, in long monologues delivered to no one.<\/p>\n
Her parents noticed the talking and mentioned it to the developmental psychologist Katherine Nelson<\/a> (1930-2018). Nelson asked them to record it. Over fifteen months they taped 122 monologues, and Nelson assembled a team of scholars, including Jerome Bruner<\/a>, to analyze them. The result, Narratives from the Crib (1989), became a classic case study in how children build language, memory, and a self. The child was identified only as Emily.<\/p>\n The child was Emily Oster<\/a> (b. February 14, 1980). Harvard University Press<\/a> reissued the book in 2006 with a foreword by the subject, who reflected on having been a research subject before she could consent or even understand. She noted that the toddler on the tapes already did what the adult does: took the raw confusion of experience and ordered it into an account of what happened, what would happen next, and what it meant. She has spent her career doing that for other people, in domains where the confusion is greater and the stakes are higher: sex ratios in Asia, HIV in Africa, pregnancy, parenthood, and a pandemic.<\/p>\n Oster is the JJE Goldman Sachs University Professor of Economics at Brown University<\/a>, a research associate at the National Bureau of Economic Research<\/a>, and the founder and chief executive of ParentData, a subscription media company. Her four books, Expecting Better (2013), Cribsheet (2019), The Family Firm (2021), and The Unexpected (2024, with maternal-fetal medicine specialist Nathan Fox), have sold more than a million copies. She is often described as a data scientist or a medical researcher. She is neither. She is an applied economist trained in econometrics and causal inference, and her career is a long test of what that training can and cannot do when pointed at questions doctors, parents, and governors have to answer.<\/p>\n The household explains a lot. Her mother, Sharon Oster<\/a> (1948-2022), taught competitive strategy at the Yale School of Management<\/a> and became its first female dean. Her father, Ray Fair<\/a> (b. 1942), built one of the best-known macroeconometric models of the American economy and a famous equation predicting presidential elections from economic conditions. Emily was one of three children in a home where incentives, scarcity, and opportunity cost were table talk. The family story she tells most often concerns groceries. She asked her mother why their family had food delivered instead of shopping like other families. Her mother explained that her time had a high opportunity cost. Emily found the answer satisfying. In that house, a choice was never just a taste. It was a price paid in forgone alternatives.<\/p>\n She went to Choate Rosemary Hall<\/a>, graduated in 1998, then to Harvard College<\/a>, graduating magna cum laude in economics in 2002. She stayed for the doctorate, finishing in 2006. At Harvard she met Jesse Shapiro<\/a>, an economist who would become known for work on media markets and political polarization. They married in 2006, later joined the Brown faculty together, and have two children, who appear throughout her books as illustrations of her decision process, never as templates for anyone else's.<\/p>\n Her early research was bold to the point of recklessness, and the pattern of that boldness set the terms of everything after.<\/p>\n The first target was the HIV epidemic in sub-Saharan Africa. Her 2005 paper in The Quarterly Journal of Economics<\/a>, “Sexually Transmitted Infections, Sexual Behavior, and the HIV\/AIDS Epidemic,” argued against the dominant explanation for Africa's catastrophic prevalence rates. The conventional account emphasized behavior. Oster argued that untreated bacterial sexually transmitted infections raised the biological probability of HIV transmission per sexual encounter, so that similar behavior produced far higher prevalence in Africa than in the United States. The policy implication followed: treating other infections might do more than campaigns for restraint. In later work she asked why sexual behavior in Africa responded so weakly to HIV risk, and her answer was an economist's answer. Prevention is an investment in future life. People facing poverty, violence, and high mortality from other causes have less future life to invest in, and her data showed the behavioral response concentrated among the richer and the safer. A third paper, “Routes of Infection,” connected export booms to spikes in new infections along trucking and trade routes, and suggested that Uganda's celebrated decline in HIV owed something to declining exports, complicating the story that education campaigns had done it alone.<\/p>\n Then came the missing women. Amartya Sen<\/a> (b. 1933) had estimated that roughly a hundred million women were missing from the populations of Asia, victims of sex-selective abortion, infanticide, and unequal food and medicine. In 2005, in the Journal of Political Economy<\/a>, Oster proposed that a large share of them had never been discriminated against at all. Medical literature suggested that mothers carrying hepatitis B bore unusually high proportions of sons. Hepatitis B was endemic in the countries with skewed sex ratios. Run the numbers and perhaps half the missing women dissolved into virology. The paper grew out of work she had begun as an undergraduate, and it made her famous in the profession before she had defended her dissertation. It also carried an enormous moral payload. If a virus explained the imbalance, the indictment of Asian son-preference shrank.<\/p>\n The claim was wrong. When Oster and collaborators obtained individual-level data from a hepatitis vaccination program in China, covering thousands of mothers whose infection status was known, carrier mothers did not bear more sons. In 2008 she published the finding under a title that left no exit: “Hepatitis B Does Not Explain Male-Biased Sex Ratios in China.” She has said since that the honest move was cheap for her and would be cheap for science if it were normal, and that the interesting question is why it is rare. The episode became as central to her reputation as the original paper. Admirers cite it as a model of how a scholar should behave when the evidence turns. It is that. It is also a record of how far an elegant argument can travel on aggregate correlations and a shaky medical premise before anyone checks, and of how much attention a young researcher can earn by overturning a morally charged consensus. The attention came first. The check came later. Both facts belong in the account.<\/p>\n Her development work continued in the same contrarian register with better luck. “The Power of TV” (2009), in The Quarterly Journal of Economics<\/a>, tracked the arrival of cable television in rural Indian villages and found it associated with reduced acceptance of wife-beating, reduced son preference, lower fertility, and greater female autonomy. Soap operas set in cities showed village women employed heroines, small families, and marriages run on different terms, and the villages moved toward what they watched. The paper became a staple of development economics, and later replication work questioned how much of the effect survives scrutiny, which is the normal afterlife of a famous natural experiment. A randomized trial in Nepal pointed the other way: distributing menstrual cups to schoolgirls did not raise attendance, because menstruation caused far less absence than the aid narrative assumed. She has been as willing to publish the null as the sensation, and the Nepal paper is the proof.<\/p>\n A second strand of her research asks how people handle threatening medical information, and it may be the deepest thing she has done. With E. Ray Dorsey<\/a> and Ira Shoulson<\/a> she studied Americans at risk for Huntington disease<\/a>, a fatal inherited neurological disorder for which a definitive genetic test exists. Most at-risk people refuse the test. The untested plan their lives, their retirements, their childbearing, as if they do not carry the mutation, while those who test positive change course: they retire earlier, complete less education, invest less in futures they will not have. Oster's interpretation ran through a model of optimal expectations. Information has value, and so does hope, and a person may rationally decline knowledge whose emotional price exceeds its planning benefit. The finding cuts against the assumption beneath her entire public career, that people want the data and are better off with it, and she published it anyway. Related work found that households cut calories only modestly after a diabetes diagnosis, and a paper with Heidi Williams<\/a> traced America's excess infant mortality to the post-neonatal months, concentrated among poorer families, pointing at inequality and conditions after the hospital rather than at American neonatal medicine.<\/p>\n Inside the profession, her most cited contribution is a method. Applied economists had long argued for their results by showing that adding control variables did not move the coefficient. Oster formalized when that argument works, incorporating movement in both the coefficient and the regression's explanatory power, and produced a sensitivity statistic now called Oster's delta: how strong would selection on unobservables have to be, relative to selection on observables, to erase the result? The method spread through empirical economics, acquired its own critical literature identifying where it misleads, and made her name a common noun in seminar rooms. A companion paper, “Health Recommendations and Selection in Health Behaviors” (2020), described a trap that explains much of nutritional science. Once a behavior is recommended, health-conscious people adopt it, and later observational studies find the behavior associated with good outcomes because of who adopted it. The recommendation manufactures its own confirmation. Anyone who has watched eggs, coffee, or red wine cycle from poison to elixir and back has seen the machine she described.<\/p>\n The University of Chicago<\/a>, where she taught from 2006, did not give her tenure. She has speculated that the time she spent writing a trade book weakened her file, an interpretation the university has never confirmed. Brown hired her with tenure in 2015, promoted her to full professor in 2016, and has since given her a university professorship, a term chairing its tenure and promotions committee, and associate editorships at The Quarterly Journal of Economics<\/a> and the Journal of Health Economics. The institution that penalized the hybrid career lost her to one that rewarded it.<\/p>\n The trade book that may have cost her Chicago began in a Chicago obstetrician's office. Pregnant with her first child around 2010, Oster met the standard American protocol: lists of forbidden foods, caffeine limits, weight targets, a testing schedule, all delivered as rules without denominators. She responded the way her mother's daughter would. She went home and pulled the studies. What she found was that the rules compressed a landscape of weak evidence, strong evidence, and value judgments into a single tone of command, and that a trained reader could take the landscape apart. Expecting Better (2013) did the taking apart in public: caffeine, sushi, deli meat, bed rest, prenatal testing, epidurals, induction. Its argument was procedural. Evidence estimates consequences. It cannot weigh them, because the weights are the woman's own, and two women reading the same numbers can decide differently without either being wrong.<\/p>\n One chapter made the book. Reviewing the observational literature on light drinking, Oster concluded that it showed no demonstrated harm from an occasional glass of wine, particularly in later trimesters, once studies accounted for the differences between light drinkers and everyone else. The medical establishment's position, held by the CDC<\/a> and the American College of Obstetricians and Gynecologists<\/a>, is that no amount of alcohol in pregnancy is known to be safe. The two statements are about different things. Oster's is a claim about what studies show. The CDC's is a population rule built for a world of varying susceptibility, imperfect self-knowledge, and the impossibility of a randomized trial. Pediatricians who treat fetal alcohol spectrum disorders denounced the chapter, and hundreds of thousands of women bought the book, many of them saying some version of the same sentence: someone finally treated me like an adult. The fight has never resolved because it cannot. It is a fight about who bears the risk of nuance, and the answer depends on where you sit.<\/p>\n Cribsheet extended the method through infancy: breastfeeding, sleep training, co-sleeping, screens, daycare. Its recurring finding was deflationary. Choices marketed as destiny mostly wash out. The celebrated associations between breastfeeding and IQ shrink toward zero in sibling comparisons, where the same mother fed one child differently than another. Sleep training leaves no measurable scar. Vaccination stood out as a case where she found the evidence one-sided and said so. The book's real product was absolution with footnotes. Against a parenting culture that treats every decision as a referendum on the parent, Oster argued that outcomes ride on genes, money, schools, and accumulated years, and that most individual choices operate inside a wide zone where several answers are fine. This made her a guru whose doctrine is the overthrow of gurus, a position with its own ironies, since a million buyers now take her word on when to disregard expert words.<\/p>\n The Family Firm carried the project into the school years and hit the limit of the data. There are no randomized trials of whether your ten-year-old should play travel soccer. So the book shifted from evidence to procedure, proposing that families make big decisions the way well-run organizations do: state the priorities, gather the facts, meet, decide, revisit. Reviewers mocked the mission statements, and the mockery missed the target. Her point was that families drift into their largest conflicts through a thousand rushed weekday negotiations, assuming love guarantees agreement, and that an hour of explicit planning surfaces the disagreement before it hardens into resentment.<\/p>\n Then the pandemic made her a political figure.<\/p>\n In the spring of 2020 American schools closed, and no federal agency built a system to track what happened when they reopened. Oster did, assembling the COVID-19 School Response Dashboard with Qualtrics<\/a> and school associations, collecting infection data from districts that volunteered it. In October 2020 she published “Schools Aren't Super-Spreaders” in The Atlantic<\/a>, reporting low case rates among roughly 200,000 students and staff and arguing that the data, imperfect as it was, should be weighed against the mounting educational damage of closure. The title outran the evidence, as she partly conceded. Volunteering districts were not a random sample. They were disproportionately the schools with the resources and the confidence to open and count. Detection depended on testing. The data could not carry the headline's certainty, and she knew the alternative to imperfect data was not perfect data. It was decisions made on fear and anecdote.<\/p>\n Watch the reception from three seats. A mother in Brooklyn, a lawyer supervising a second-grader's Zoom kindergarten from her kitchen table while her own billable hours evaporate, reads the article and feels that someone with credentials has finally counted her child's losses as losses. A veteran teacher in Providence, fifty-eight years old, prediabetic, reads the same article and notes that the economist telling her the building is safe will not be standing in it, and that her union's dead are not in the dashboard. And in Tallahassee, Governor Ron DeSantis<\/a> cites Oster's research while banning school mask mandates, at roughly the same time CDC Director Rochelle Walensky<\/a> cites it to justify relaxing classroom distancing rules. Both parties in an otherwise total war reached for the same economist's numbers. Few other figures of the pandemic can say that, and the price was that both sides also learned to distrust her.<\/p>\n The later evidence delivered a split verdict. Her own co-authored national study found that reopening schools did not raise community case rates on average, with real variation: some regions and modes saw increases, especially where community transmission already ran high. Her test-score work across eleven states found that districts staying remote lost far more ground, with math worse than reading and the largest losses in districts serving more Black students, so that the closures adopted in the name of protecting the vulnerable widened the gaps the same institutions spend their days deploring. On the learning question she has been vindicated. On the transmission question the honest summary is that she was more right than her angriest critics and more confident than her data.<\/p>\n In October 2022 she wrote the essay that made her radioactive to both camps at once. “Let's Declare a Pandemic Amnesty,” in The Atlantic<\/a>, opened with her own family hiking in cloth masks in April 2020, her masked children shrieking “SOCIAL DISTANCING!” at other kids who came too close, and argued that most people, herself included, had made their pandemic mistakes in genuine darkness, that Los Angeles closing its beaches made no more sense in hindsight than her masked hikes, and that the country should forgive the honest errors and get on with fixing the damage. The essay was ratioed into legend. The left read it as the school-reopener demanding her critics concede. The right read it as the laptop class requesting absolution before any accounting of who closed the schools, who lost the businesses, who was fired, and who was censored, and noting that amnesty is a thing the powerful grant, not a thing they request. Her stated motive was narrower: she had watched the vindicated turn vindictive on test scores and wanted off the carousel of recrimination. The word “amnesty” sank the argument. Amnesty follows a verdict. She was proposing to skip the trial, and a country that had buried a million people and locked its children out of school for a year was not ready to skip it.<\/p>\n Her fourth book stepped back from the wars. The Unexpected (2024), written with Nathan Fox, an obstetrician specializing in high-risk pregnancy, addresses the third or so of pregnancies that go wrong: miscarriage, preeclampsia, preterm birth, gestational diabetes, hyperemesis, postpartum depression. The question organizing the book is recurrence. A woman who has been through one of these wants to know the odds of going through it again and what can change them, and her twenty-minute appointment rarely tells her. The collaboration also answered, structurally, the oldest criticism of her franchise. The economist reads the studies; the physician who has managed thousands of these pregnancies says what the studies miss. It is the book of hers that medicine has received most warmly, and the least like economics: recurrence tables interleaved with scripts for asking your doctor why it happened and grief that no denominator absorbs.<\/p>\n Around the books she has built a company. ParentData, founded as a newsletter in 2020, now reaches hundreds of thousands of readers weekly with more than 1,100 articles, personalized editions keyed to a due date or a child's age, and an AI search tool, Petey, that answers questions from her archive, an AI librarian whose answers are as good as the shelf behind it. The masthead has grown past her to include obstetricians, endocrinologists, and psychologists, diluting the founding tension of an economist interpreting clinical literature for the masses without dissolving it. In February 2026 she extended the franchise past parenting entirely, launching Wellness, Actually, a weekly iHeart podcast with F. Perry Wilson, a Yale physician, working through the claims of the wellness economy: GLP-1s, colostrum, red-light therapy, microplastics, continuous glucose monitors, full-body scans. The premise of the show is the premise of her career, that the gap between what a study shows and what an influencer says it shows is wide enough to build a life in.<\/p>\n The commercial turn changes the epistemics, and she knows it. A referee lets a paper end in “we cannot distinguish these hypotheses.” A subscription business cannot publish that sentence three times a week. Her academic work taught that people sometimes rationally refuse information and that recommendations corrupt the later evidence for themselves. Her company sells information and recommendations on a weekly schedule. Managing that contradiction, keeping uncertainty visible in a product people pay to have uncertainty removed, is the tightrope of her second act.<\/p>\n The criticisms of her worth taking seriously are four. Her framework prices choices for people who have them; a mother without paid leave does not need a better decision procedure about breastfeeding, and the books can make a structural constraint look like a personal optimization problem. Economists reading medical literature miss things clinicians catch, a gap she has narrowed by hiring the clinicians. The confidence that makes her readable erases, at the margin, the line between “the studies do not establish harm” and “this is safe,” and that margin is where fetal alcohol syndrome lives. And her politics of individual informed choice belongs to no coalition, which is why the left got to call her a DeSantis asset and the right got to call her a lockdown Karen in the same news cycle, sometimes about the same article.<\/p>\n One circulating credential should be retired. Online biographies state that Oster won the Elaine Bennett Research Prize in 2013. The American Economic Association<\/a>'s records do not list her among its recipients; Anna Mikusheva<\/a> won in 2012 and Emi Nakamura<\/a> won in 2014. Her real honors, the Time 100 listing in 2022 with a citation written by Steven Pinker<\/a> (b. 1954), the named chairs, the publication record across the American Economic Review<\/a>, the QJE, and the JPE, need no padding, and a writer whose brand is checking claims deserves to have her own claims checked.<\/p>\n What she has built, across the journal articles and the bestsellers and the dashboard and the podcast, is a single sustained argument about authority. Official guidance, she keeps showing, is rarely a direct translation of decisive science. It is evidence blended with precaution, liability, institutional habit, and someone else's values, presented as fact. Exposing the blend improves informed consent, and it corrodes trust, and the corrosion does not stop where she would like it to stop. A reader taught that the alcohol rule overstates the evidence may conclude the vaccine schedule does too, and she then spends a chapter arguing it does not, from an authority she has taught the reader to interrogate. She has no solution to this beyond doing the work in the open and correcting herself in public, which is more than most of her critics or her imitators offer.<\/p>\n The tapes under the crib caught a two-year-old ordering the day into a story so she could sleep. The adult does the same for an audience of millions who cannot sleep either: the pregnant woman with the glass of wine she cannot mention to her doctor, the parents at the kitchen table with the school reopening survey, the man holding a genetic test kit he may never open. She tells them what the numbers can say, where the numbers stop, and that past that line the decision is theirs. It is a modest doctrine that took an immodest career to establish, and its central risk has never changed since the hepatitis paper: the story is always cleaner than the data. She has been better than almost anyone at admitting it afterward. The harder discipline, the one her whole enterprise now depends on, is saying it at the time.<\/p>\n Notes<\/p>\n Scenes and their sourcing. The crib opening rests on the Harvard UP<\/a> reissue of Narratives from the Crib<\/i><\/a>: 122 monologues taped over fifteen months, Oster<\/a>‘s 2006 foreword on having been a research subject without knowing it; retail listings confirm the details at AbeBooks<\/a>. Jerome Bruner<\/a> contributed to the volume; verify before publishing if you keep his name. The amnesty hiking scene comes from her own essay’s opening; a critical account describes her cloth-masked kids shouting \u201cSOCIAL DISTANCING!\u201d at other children on family hikes, The American Spectator<\/i><\/a>; original at The Atlantic<\/i><\/a>. The Brooklyn mother and Providence teacher are the kind of extrapolation I license: composite viewpoints, no named individuals, drawn from the documented class split in the reopening fight.<\/p>\n The both-parties detail. One critical profile notes that DeSantis<\/a> cited her study when banning school mask mandates while CDC Director Walensky<\/a> referenced her research before relaxing classroom distancing guidance, Substack<\/a>, quoting a critical article about her influence.<\/p>\n Wellness, Actually<\/i><\/a>. Verified: weekly iHeart<\/a> podcast with Emily Oster and physician Perry Wilson<\/a>, covering health news, listener questions, and wellness deep dives, launched February 2026; Apple Podcasts<\/a>. Wilson is an associate professor of medicine and public health at Yale<\/a>.<\/p>\n The Economist at the Border: Emily Oster Through Andrew Abbott<\/strong><\/p>\n Chicago, around 2010. A pregnant economics professor sits in an obstetrician’s office and receives a list. No sushi, no deli meat, two hundred milligrams of caffeine, a weight target, a testing schedule. The list comes with the full weight of the white coat and none of the reasoning behind it. She goes home, opens the journals, and discovers that the reasoning is a mixed bag: some rules rest on strong evidence, some on weak studies of heavy exposure extrapolated downward, some on liability and habit. Three years later she publishes the reasoning, and a licensed profession finds a competitor standing on ground it had assumed was fenced.<\/p>\n Andrew Abbott<\/a>'s The System of Professions<\/a> (1988) exists to explain that scene. Abbott argued that professions should be studied through their work, and their work through jurisdiction: the link a profession claims between itself and a set of tasks, defended against rivals in three arenas, the law, the workplace, and public opinion. Professions rise and fall through interprofessional competition over tasks, and the weapon of that competition is abstract knowledge. A profession holds its territory by showing that its abstractions diagnose, reason about, and treat a problem better than the abstractions next door. Jurisdictions are never safe. They are attacked when they weaken, and they weaken from within, when a profession reduces its work to rules a layman can inspect, and from without, when technology or organizations restructure the tasks.<\/p>\n Emily Oster<\/a> is a jurisdictional dispute with a face. Read through Abbott, her career stops looking like a personal brand built on contrarian nerve and starts looking like a predictable border war between economics and medicine, fought over a strip of territory medicine held but had stopped defending.<\/p>\n Start with the task. Abbott decomposes professional work into diagnosis, inference, and treatment. Diagnosis takes information into the professional knowledge system; treatment brings instructions back out; inference connects them, and inference is where professions live or die, because pure inference is what cannot be routinized. Medicine’s jurisdiction over pregnancy is anchored at the treatment end. It holds delivery, surgery, prescription, and emergency by license, and no economist ever threatened any of it. The advisory task was different. Prenatal counsel had been compressed into rules: the list in the office, the pamphlet, the ban on wine and tuna, “What to Expect When You’re Expecting” as the delegated civilian arm. Abbott holds that a profession which reduces a task to codified rules invites invasion, because rules can be inspected, and inspected rules can be found wanting by anyone whose own abstractions handle the underlying problem. The underlying problem of prenatal advice is inference from observational data under confounding: what does a correlation between light drinking and child outcomes mean when light drinkers differ from abstainers in twenty other ways? That is an econometrics problem. Medicine had the treatment and had let the inference go slack. Oster attacked the slack. She never claimed the fortress. She claimed the flank the garrison had left to pamphlets.<\/p>\n Abbott would insist on the external conditions that opened the border, because in his system jurisdictions shift when the larger environment restructures tasks and audiences. Three changes mattered. PubMed<\/a> and open repositories put the primary literature in front of laymen for the first time; the doctor’s monopoly on the studies dissolved a decade before Oster wrote. The audience changed with it: her readers are college-educated professional women trained to read evidence in their own fields and offended by instruction without justification, an audience that did not exist at scale in 1970. And the demand side changed, because obstetric practice had compressed counsel into the twelve-minute visit, leaving the advisory task effectively vacant during the six months of terror between appointments. Abbott’s rule is that vacant tasks get occupied. The question was never whether someone would occupy prenatal advice. The question was which knowledge system would get there first, and the candidates were wellness culture, religious tradition, the mommy internet, and applied economics. Medicine’s rivals for the territory were already inside it. Oster merely arrived with better artillery.<\/p>\n The artillery is abstraction, and Abbott has a doctrine about its caliber. Knowledge too concrete degrades into clerical work anyone can take; knowledge too abstract loses connection to the task and persuades no one. Jurisdiction goes to optimal abstraction, general enough to claim new problems, concrete enough to solve the one in front of the client. Oster’s package sits near the optimum. Confounding, selection, absolute versus relative risk, the difference between “no demonstrated harm” and “demonstrated no harm”: these are portable abstractions, which is why the same apparatus moved from sex ratios to pregnancy to sleep training to school closures to creatine without modification. And they cash out in a concrete instruction, the coffee is fine, the sushi from a decent restaurant is fine, the sleep training leaves no scar. Medicine’s competing product on this task was a rule without a denominator. Between an abstraction that explains and a rule that commands, the educated client chose the explanation, and Abbott would say the outcome was overdetermined.<\/p>\n Now the arenas, because Abbott’s sharpest tool is the observation that jurisdiction is claimed in three places at once and won in them separately. In the legal arena Oster claimed nothing and can claim nothing. She holds no license, prescribes nothing, and every book carries the disclaimer conceding medicine’s legal monopoly intact. In the workplace arena, the clinic, her presence is real but unofficial: she enters as the patient’s preparation, the printout, the question the obstetrician now has to answer with reasons instead of authority. Abbott notes that workplace jurisdiction always runs ahead of or behind formal jurisdiction, and the exam room where a patient cites Expecting Better to a physician is his example updated, an unlicensed competitor exercising influence the law does not recognize. Her real conquest came in the third arena, public opinion, where jurisdiction means the public’s belief about who legitimately speaks to a problem. A million books, the Time 100<\/a> listing, the newsletter audience in the hundreds of thousands: by the measure of public opinion, an economist now shares jurisdiction over the American pregnancy, and the sharing was won there and nowhere else, which is why it can be contested there and nowhere else.<\/p>\n The counterattack followed Abbott’s script for jurisdictional defense. A profession defending territory does not usually argue the technical point; it argues the client. The response to the alcohol chapter from pediatricians who treat fetal alcohol spectrum disorders, and the standing restatements from the CDC<\/a> and the American College of Obstetricians and Gynecologists<\/a> that no amount of alcohol is known to be safe, all took the same form: the population contains vulnerable members, nuance will be misread as permission, and the layman cannot handle the inference and must be given the rule. That argument is the oldest move in professional defense, the claim that the client’s incompetence requires the profession’s simplification, and Abbott treats it as the standard rhetoric by which professions justify holding inference away from the laity. What medicine could not do was answer her on the evidence, because on the evidence the observational literature is what she said it is. So the defense shifted from the studies to the audience, from what is true to what the public can safely be told, which conceded the inference and fought for the treatment relationship. The fight has never resolved because the two sides are defending different links in Abbott’s chain. She holds the inference. They hold the treatment. The client stands between.<\/p>\n Abbott catalogues the settlements available when neither side can expel the other: full jurisdiction, subordination, division of labor by task, advisory jurisdiction, division by clientele. Oster’s career has run through the catalogue in order. Expecting Better was a raid, a claim of advisory jurisdiction pressed unilaterally from outside. The pandemic dashboard was occupation of an abandoned task. And The Unexpected (2024), written with the maternal-fetal specialist Nathan Fox, is a signed treaty, a division of labor on the page itself: the economist reads the studies and prices the recurrence risks, the physician supplies what the studies miss and manages the patient, and the book’s structure concedes what twelve years of solo authorship had contested, that the advisory task splits into an inference component and a clinical component and the invader holds only one. ParentData’s masthead is the same treaty at industrial scale. The platform that began as one economist interpreting clinical literature now employs obstetricians, endocrinologists, and psychologists, which is to say it hired the profession it raided. Abbott would recognize the arc at once: an interloper who cannot win the legal arena stabilizes her workplace gains by incorporating the incumbents, and the settlement converts a border war into an org chart.<\/p>\n ParentData is also the part of the story Abbott’s 1988 apparatus strains to hold, and the strain is instructive about the present. Abbott’s professions institutionalize jurisdiction through associations, schools, and licenses. Oster institutionalized hers through a subscription company. The jurisdiction is real, a defined task, a knowledge system, a clientele, a payment stream, but it is held by a firm rather than a profession, transmitted by brand rather than credential, and heritable by no colleague group. There is no association of evidence-based parenting advisers, no examination, no ethics board, no successor. This is where Gil Eyal<\/a>'s The Crisis of Expertise (2019) extends the spine. Eyal distinguishes expertise, the task of making knowledge relevant to decisions, from the professions that historically packaged it, and argues that the crisis of our period is the unbundling of the two: expertise now circulates through networks, media, and markets that professions no longer control, generating the twin spectacle of collapsing trust in credentialed authority and insatiable demand for expert guidance. Oster is the unbundling with a newsletter. Her readers arrive because they distrust the official rule and stay because they crave an authoritative answer, and she supplies expertise stripped of profession, judgment without license, jurisdiction as a service. Eyal’s paradox, that attacks on experts coexist with dependence on them, is her business model stated as sociology.<\/p>\n The pandemic ran the whole cycle in fast motion and in public. In 2020 the task of counting infections in reopened schools belonged by rights to public health, which held the jurisdiction and abandoned the work; no federal system existed. Abbott’s vacancy rule operated within months. An economist with survey software occupied the task, and occupation of the task carried, as it always does, a claim on the inference: her data, therefore her reading of what schools did to transmission, therefore her standing in the reopening fight. The breadth of her public-arena jurisdiction at the peak can be measured by a single fact from the bio: a Republican governor cited her research while banning school mask mandates at roughly the same time the CDC director referenced it to relax distancing guidance. Warring authorities do not cite the same outsider unless the outsider holds the territory they both need to cross. And the amnesty essay of 2022 measured the limit of that kind of holding. A credentialed profession absorbs a member’s public misstep through the colleague group, the association, the shared jurisdiction that no single controversy can revoke. Oster’s jurisdiction was personal, held in the public arena alone, and the public arena is the one venue where standing is revocable overnight by the same opinion that granted it. Both coalitions burned her at once, and there was no guild to stand behind. Abbott’s system predicts the exposure: jurisdiction without institution is jurisdiction on a month-to-month lease.<\/p>\n One more of Abbott’s engines is already turning, and it points at her. Codified knowledge invites takeover; that was the vulnerability through which she entered medicine’s advisory territory, rules written down where a rival could read them. Her archive is now codified too, eleven hundred articles and four books compiled into Petey, a chatbot that answers the client’s question directly from the corpus. The tool that scales her jurisdiction also completes its codification, and Abbott’s logic does not exempt the victor: an advisory jurisdiction reduced to a searchable rulebook is open to the next system of abstraction that handles the inference better, whether that is a rival platform, a general-purpose model trained on the same literature, or medicine reclaiming the counsel it once delegated to pamphlets, this time with the denominators included. She took the territory because the incumbents held it as text. She now holds it as text.<\/p>\n What the frame settles, and what it prices, comes down to this. The Abbott reading dissolves the personality story. Oster’s rise required no unusual courage and her opponents no unusual bad faith; a weakly defended advisory task, a newly literate clientele, a portable abstraction, and a vacant inference produced the invasion, and the counterattack, the treaty, and the incorporation followed the standard sequence of jurisdictional conflict as written twenty-five years before she published. The frame also identifies the durable stakes under the alcohol quarrel, which was never about wine. It was about which profession’s chain of inference governs the space between the study and the patient, and that question outlives every particular rule. What the frame cannot price is the client’s own experience, the woman in the office with the list, for whom the border war is not sociology but the question of whether she is to be reasoned with or managed. Abbott’s system explains why she was managed, why an economist was the one who finally reasoned with her, and why the reasoning arrived as a product with a subscription price. It does not say whether she is better governed by the guild that held the treatment or the firm that took the inference. That verdict belongs to her, which is, as it happens, the invader’s doctrine too.<\/p>\n The Anthropologist at the Baby Shower: Emily Oster Through Mary Douglas<\/strong><\/p>\n A restaurant in any American city. A woman in her second trimester orders a half glass of red with dinner. The waiter’s eyes drop to her belly and come back up. At the next table a stranger stops chewing. Someone will say something, or someone will hold herself back from saying something, and the woman knows it, which is why she almost ordered the sparkling water. She drove to the restaurant on the freeway, a risk to her pregnancy measurably larger than the wine, and no one at any table has ever stopped a pregnant woman in a parking garage to ask whether her trip was necessary.<\/p>\n Mary Douglas spent a career explaining that scene. Purity and Danger (1966) argued that pollution beliefs are never really about hygiene. Dirt is matter out of place, and a community’s rules about what defiles are a map of its categories, policing the points where the categories might blur. The abominations of Leviticus were her showpiece: the forbidden animals are the ones that straddle classes, water creatures without fins, land creatures that swarm, and the dietary law is a meditation on order, holiness as keeping kinds distinct. Among the Lele of the Kasai she found the inversion that proved the rule, the pangolin, a scaly, tree-climbing, fish-like mammal that bears its young singly as humans do, so anomalous that the Lele made it the object of a fertility cult. Anomaly draws power. What a classification system cannot place, a community must either abominate or venerate, and it cannot leave it alone.<\/p>\n Sixteen years later, in Risk and Culture (1982), Douglas and the political scientist Aaron Wildavsky carried the argument into the modern world that believed itself past all that. Every society faces more dangers than it can attend to, so every society selects, and the selection is social. Communities police the risks that reinforce their moral order and ignore the ones that don’t, then experience the selection as objective danger. The book scandalized readers by applying this to American environmentalism, arguing that the choice of industrial pollution as the master danger served the boundary needs of sectarian groups defining themselves against the center. The particular target has aged; the machinery has not. Risk, for Douglas, is the modern vocabulary of taboo, pollution belief conducted in the idiom of probability, and its function is what pollution’s function always was: to mark boundaries, bind members, and, when misfortune comes, to allocate blame.<\/p>\n Now set Emily Oster inside the machinery, because her entire public career is a collision with it, and the collision explains what her own vocabulary cannot: why she wins the arguments and loses the fights.<\/p>\n Begin where Douglas begins, with the anomalous body. The pregnant woman is the boundary case of the human classification system, two beings in one skin, a person who is also a vessel, neither one nor two, and Douglas noted in Purity and Danger that societies everywhere surround transitional beings, the initiate, the corpse, the unborn, with special danger and special rules. The pregnant body is a margin walking around in public, and margins are where pollution rules concentrate. Ingestion rules concentrate hardest of all, because eating is the act in which the outside world crosses the body’s boundary, and in pregnancy every crossing is double: what passes the mother’s boundary passes the child’s. Seen from this height, the American prenatal code, no sushi, no deli meat, no soft cheese, no wine, snaps into a familiar shape. It is a dietary law. Its items are not ranked by hazard; listeria from deli turkey, mercury from tuna, alcohol from a Tuesday glass, and caffeine sit side by side in the pamphlet as they never would in an actuarial table, because the list’s coherence is categorical rather than statistical. These are the substances that defile the vessel. Driving, which kills and injures more pregnant women than everything on the list combined, appears nowhere, because a car crash does not symbolize a boundary violation. Nothing crosses into the body. There is no defilement to police, so the community does not police it, and then, exactly as Douglas predicts, the community experiences its selection as science.<\/p>\n The tell is the enforcement. Medical advice is delivered by physicians in offices. Taboo is enforced by everyone, everywhere. The pregnant woman with the wine glass is corrected by waiters, mothers-in-law, and strangers, none of whom could state the relative risk within three orders of magnitude and none of whom needs to, because they are not making a risk claim. They are protecting a category. A visibly pregnant woman drinking in public is matter out of place, the vessel defiling itself, and the flash of anger she draws is the anger Douglas taught us to read: the emotion that guards the line between kinds. No one is angry at a pregnant woman merging onto the 405. Anger is reserved for pollution, and the distribution of anger, not the distribution of harm, maps the actual rule.<\/p>\n Into this walks an economist with a denominator. Expecting Better treated the prenatal code as a set of empirical claims about dose and response, went through the studies, and reported that the claims varied from solid to unsupported, with the occasional-glass-of-wine claim near the unsupported end. As analysis, this was competent and has never been seriously rebutted on its own terms. As anthropology, it was a category error of the purest kind, and Douglas would have seen the explosion coming from a mile off. The rule Oster tested was never a dose-response claim wearing plain clothes. Its form gives it away. “No amount of alcohol is known to be safe at any time in pregnancy” is not the shape of a statistical finding; findings come with magnitudes, intervals, and gradients. It is the shape of a taboo: absolute, unranked, admitting no degree, because taboos guard categories and a category is either intact or breached. You cannot be a little bit defiled. When Oster answered the rule with a coefficient, she was not correcting an error inside the system. She was denying the system’s kind, translating a boundary into a slope, and the fury that came back at her was not the irritation of experts contradicted on evidence. Experts contradicted on evidence write rebuttal letters about confounders. This was the other thing, the heat that greets sacrilege, the accusation not that she was wrong but that she was dangerous, that her book would license defilement, that women would hear permission. Douglas’s terms fit the reaction with almost embarrassing snugness: the priesthood did not defend the coefficient, it defended the boundary, and it defended it with the argument boundary-keepers always use, that the laity cannot be trusted with degrees.<\/p>\n Blame completes the circuit, and Douglas’s later work, Risk and Blame among it, put blame at the center of modern risk talk. Pollution beliefs earn their keep when misfortune strikes, by supplying a culprit and thereby reassuring the community that the moral order still governs events. A miscarriage, a birth defect, a diagnosis in childhood: these are the intolerable events, misfortune without meaning, and the prenatal code converts them into meaning by converting them into verdicts. She drank. She ate the cheese. She didn’t take the folate. The mother is the accused in waiting, and the code she followed or breached is the indictment pre-drafted. This is the cruelty Oster’s readers write to her about, the women who miscarried and spent months auditing their own coffee, and it is also the code’s deep function, which is why the code cannot be reformed by data. A community that gave up the taboo would have to sit with unexplained loss, and Douglas’s whole life’s work says communities would rather have blame than chaos. Oster offers the chaos with error bars. The relief her readers report, and it is the most consistent testimony in her reviews, is the relief of the absolved, women released from a purity regime they could never satisfy, and the fury she draws is its mirror, the reaction of those for whom the regime is load and meaning at once.<\/p>\n The frame keeps paying as the career widens. The pandemic was a pollution event on the Douglas scale, an invisible contaminant crossing every boundary at once, and the response ran on purity machinery from the first week: the six feet, the wiped groceries, the masks worn alone on trails, her own children shrieking “SOCIAL DISTANCING!” at strangers, lay enforcement again, children as taboo police, the surest sign a rule has left the domain of technique and entered the domain of defilement. Some measures worked and some did not, and the point of the Douglas reading is that their persistence tracked their symbolic yield as much as their epidemiological one; the beach closures she mocked in the amnesty essay are inexplicable as disease control and transparent as boundary display. Then the amnesty essay itself, read through Risk and Blame, becomes a study in predictable failure. Oster asked a wounded community to forgo blame for its misfortune, and blame, in Douglas’s account, is what a community’s risk beliefs exist to produce. She was proposing to decommission the machinery at the moment of its maximum use. Both moral communities refused, in identical form and opposite content: each side had its polluters, the reckless unmasked or the cruel closers, and each heard “amnesty” as the demand that defilement go unpurified. Communities purify. They do not amnesty. She keeps making the same offer, evidence in exchange for absolution-rights, and the offer keeps being refused because it asks people to trade the thing pollution beliefs are for.<\/p>\n There is a last turn, and honesty requires it, because Douglas never granted anyone an exemption, least of all the rational. Purity and Danger’s standing provocation was that the moderns who laugh at the Lele have their own pollutions and cannot see them. Oster’s tribe included. The evidence culture she comes from runs on an explicit purity hierarchy, and it talks like one: economists praise “clean identification,” worry about data “contamination,” treat selection as a corrupting influence that must be washed out, and rank the randomized trial as the immaculate design from which all observational work is a fall. The hierarchy of evidence is a hierarchy of ritual purity, and it has its own anomalies it cannot leave alone and its own defilements it polices with heat; sit in any seminar when someone presents a correlation as causal and watch the room enforce. This does not level the difference between the CDC’s rule and Oster’s coefficient; a purity system disciplined by prediction differs from one that answers only to the community, and the difference is most of what we mean by science. It does mean her position is a rival order rather than a view from outside all orders, a community of the clean-identified with its own boundaries, its own membership badges, and its own anger. Her readers convert to it as one converts to a lighter law, trading the taboo of the vessel for the taboo of the confounder, and finding the second easier to live under because it indicts studies instead of mothers.<\/p>\n That may be the fairest summary the frame allows. Oster has not led America’s parents out of pollution thinking, because there is no out; Douglas closed that exit in 1966. She has offered a migration between purity systems, from one that locates danger in the mother’s body and pays out in blame, to one that locates danger in bad inference and pays out in doubt. The first is older, harsher, and self-enforcing at every restaurant table. The second is kinder to the woman with the wine glass and harder on the human need to know why the bad thing happened. The migration is real, a million copies of it, and it remains a minority movement for the reason Douglas would name last and mean most: when the loss comes, doubt is thin, and blame, whatever it costs the accused, tells the community that the world still has rules.<\/p>\n","protected":false},"excerpt":{"rendered":" New Haven, 1981. A cassette recorder sits under a crib in the home of two Yale economists. The child in the crib is not yet two. After her parents leave the room, she talks. She reviews the day. She rehearses … Continue reading