How Much Would You Pay for a Smarter Baby (Economist)

Summary

Business feature on the Silicon Valley embryo-screening startups (Nucleus, Orchid, Herasight) selling polygenic scores (PGT-P) for disease risk, height, IQ, longevity and even hair and eye colour, at up to $50,000. Lays out the market maths, the science dispute, and the ethics fight. Pairs with the Leader In Praise of Designer-ish Babies (Economist). Concept page: Polygenic Embryo Screening.

Market

• US IVF births 2024: 100,000, 2.8% of all births. Global fertility market ~$25–30bn/yr (David Sable, investor and former reproductive endocrinologist); could reach $400–500bn/yr if the full IVF process fell to ~$15,000 (one US cycle today costs $15,000–20,000; only 15 states mandate insurer coverage).
• US VC into fertility nearly doubled $254m (2019) → $496m (2022) (PitchBook), since slowed. Peter Thiel and Elon Musk framed low birth rates as a national threat and seeded the sector; Founders Fund (Thiel) is invested.
• Gaia (fertility insurance/payment plans) claims to save patients ~$15,000 across the process by pooling risk.
• Survey (Science, 2023): 43% of IVF participants would use polygenic screening to raise a child's odds of an elite university if safe and free; a third said the same for genetic editing. Nearly three-quarters of Americans say they would use polygenic screening if already doing IVF.

The three screens

1. PGT-A — chromosome count; some studies claim lower miscarriage risk, many experts sceptical (the IVF add-on review found no effect: Do High-Tech IVF Add-ons Work (Economist)).
2. PGT-M — single-gene disorders (cystic fibrosis, Huntington's); well established.
3. PGT-P — polygenic scores from genome-wide association studies (GWAS), sometimes adjusted for ancestry and family history. The new, contested layer.

The science dispute

• Within-family validity. Selection on polygenic scores works across many offspring, as in animal breeding, but is "less sure" for a single child (Kevin Mitchell, Trinity College Dublin). Herasight validates its scores on sibling data from GWAS datasets to show within-family predictive value.
• Herasight's headline claim: testing ten embryos for a couple where both parents have type 2 diabetes cuts the child's absolute risk by 12–20 percentage points from a 40–60% baseline (company claim, not independently verified in the piece). See Type 2 Diabetes Remission for what the condition looks like once it arrives.
• Effect sizes (Cell, 2019): choosing among ten viable embryos, the median gain is 2.9 cm height or 3 IQ points; gains "drop precipitously" with fewer than five embryos. Larger training databases could roughly double the IQ gain. Dagan Wells (Oxford): for many conditions environment matters more than genes.
• Pleiotropy. Selecting against one disease may raise risk of another the provider doesn't score; some studies find most genetic correlations between polygenic diseases are positive or insignificant, which would limit the downside. Mitchell's summary: the tech is "a promissory note".
• Ancestry gap. Training data under-represents non-European populations, so scores are less accurate for them; expected to improve.

Ethics and regulation

• Screening for sex, polygenic conditions or physical traits is illegal in Britain and many jurisdictions; America has had a de-facto ban on embryo editing since 2016.
• Industry view (Delian Asparouhov, Founders Fund): risk scores are "best in class"; genetic counsellors' insistence on clinical-grade interpretation is gatekeeping; "If this is eugenics, then so is Hinge." All three startups do employ genetic counsellors.
• Case study: Arthur Zey and Chase Popp chose one of six embryos on Herasight scores, aiming at lifespan rather than height or IQ.

What the article flags as unproven or contested

• Whether PGT-P predicts outcomes for an individual embryo rather than a population; the only quantified gains are modelled, modest, and shrink fast with fewer embryos.
• Company absolute-risk claims (Herasight's 12–20 points) are the company's own.
• Whether disclosure suffices or whether offering IQ and aesthetic scores shows the firms are selling "super-babies" regardless of how they describe their customers.

Wikilinks

• In Praise of Designer-ish Babies (Economist) — the Leader: regulate by disclosure, don't ban.
• Polygenic Embryo Screening — concept page.
• Do High-Tech IVF Add-ons Work (Economist) — PGT-A's evidence record.
• Newborn Genome Sequencing — genetic prediction after birth; same penetrance caveat.
• Elon Musk — early fertility-sector investor.
• The Economist — Business.

Captured-by trigger

Source file: raw/processed/Economist 2026-08-15 Business - How much would you pay for a smarter baby.md.