All podcasts / People I (Mostly) Admire / Summary

167. The Secret of Humanity? It's Common Knowledge.

2025-09-27 - source - Read full transcript

Steve Levitt (host)Seth BerkleyMorgan Levey (host)

Key insights

Gavi delivers the WHO's full 11-vaccine childhood schedule for about $24 per child, versus $1,300 for the same vaccines in the US, a 98 percent price reduction achieved through bulk buying and expanding the manufacturer base.

Berkley says the cost-benefit works out to roughly $21 returned per dollar invested in narrow terms, rising to about $54 per dollar when broader health and productivity gains are counted, a return he says has no parallel elsewhere in healthcare.

vaccine-economics

Vaccines for diseases confined to the developing world are systematically underfunded because pharma companies have no market incentive to invest in them.

Berkley says the incentive problem is real, but Gavi's fix was showing manufacturers that serving low- and middle-income markets adds volume and economies of scale that lower costs and raise profit in their primary high-income markets too, turning a charity ask into a business case.

vaccine-economics

Only about 7 percent of vaccine candidates that enter trials are ultimately approved, and even late-stage failures are common and expensive.

Berkley attributes this to the field's ceiling on non-human testing: animal models are unreliable ('mice lie, monkeys don't always tell the truth'), so efficacy and rare safety issues can only be confirmed in large, costly human trials, which is why pre-Covid vaccine development typically took four to ten years.

vaccine-economics

The Covid vaccine's 327-day development time was possible only because of prior, unfinished research on related coronaviruses (SARS, MERS) combined with the speed of the new mRNA platform.

Berkley notes the first vaccine sequence was assembled in days and the first dose was ready in 63 days, an unprecedented pace built on Operation Warp Speed's funding and on scientific groundwork that had already been done for other coronavirus outbreaks.

pandemic-preparedness

Preventive vaccines undermine their own popular support because the deaths and suffering they prevent are invisible and impossible to imagine.

Levitt illustrates this with his own son's death from pneumococcal meningitis in October 1999, four months before the vaccine (Prevnar) that would likely have saved him was approved; because there is no counterfactual disease experience to point to, people underweight vaccines' benefit while overweighting rare visible side effects.

vaccine-hesitancy

Vaccine hesitancy around conditions like autism is fueled by a causality trap created by the timing of the childhood vaccine schedule.

Because most vaccines are given during the exact developmental window (roughly six months to two years) when conditions like autism first become visible, parents wrongly infer that the vaccine caused a developmental change that was actually coincidental or genetic in origin.

vaccine-hesitancy

Modeling suggests roughly a 50 percent chance of another Covid-like pandemic within 25 years, yet current US policy is cutting the exact preparedness capacity (mRNA research funding, institutional expertise) that made the Covid vaccine response possible.

Berkley calls the NIH's cancellation of mRNA vaccine research 'the stupidest science decision' a colleague has seen in his career, since the same platform speed that matters for pandemics also matters for individualized cancer immunotherapy.

pandemic-preparedness

Vaccine stockpiles for rare, unpredictable outbreaks (like Ebola) require accepting that most of the stock will expire unused, which is a good trade given the cost.

Berkley describes the old mindset of avoiding 'waste' by storing vaccine candidates in bulk rather than in ready-to-use vials, which cost precious time once an outbreak began; spending a few million dollars on a stockpile that may never be used is, in his framing, an extremely good buy against catastrophic global risk.

pandemic-preparedness

Politicizing evolving public health guidance during Covid (masks, six-foot distancing) damaged the credibility of the underlying scientific process.

Berkley argues public health recommendations are supposed to update as evidence comes in, an iterative process, but once Covid became a partisan issue, changing guidance was read as incompetence rather than normal scientific learning, which he calls a genuinely destructive effect on trust.

public-health-policy

The all-pay auction creates an escalation trap where every bidder's rational move is to keep bidding, because stopping locks in the full loss while one more bid only risks losing slightly more.

Levitt explains that once a bidder passes the item's face value, the marginal decision at every subsequent turn favors bidding again rather than eating a larger loss, which is why the auction has no natural stopping point in a classroom setting.

auction-theory

The same auction mechanism that reliably works in a classroom failed when a listener tried it with his own family, because family relationships override the 'consumer' behavioral rules people default to with strangers.

Levitt attributes the listener's failed experiment to two factors: people act by different social rules within families than in an anonymous bidding pool, and running an effective auction requires deliberate facilitation skill (keeping momentum, cutting off side conversation) that the listener lacked.

auction-theory

Books referenced

Fair Doses: An Insider's Story of the Pandemic and the Global Fight for Vaccine Equity - Seth Berkley - Berkley's own book, publishing October 28, 2025, covering his years running Gavi and the fight for global vaccine equity; referenced repeatedly for anecdotes about Ebola stockpiling and vaccine manufacturing.

Media referenced

World Development Report - paper - World Bank report Berkley co-authored making the cost-effectiveness case for vaccines; Bill Gates read it twice and it helped spur his initial $750 million investment that helped launch Gavi.
Conversation with Sendhil Mullainathan (2021 episode) - podcast - Earlier People I (Mostly) Admire episode where Levitt and Sendhil Mullainathan discussed the all-pay ($20) auction, which a listener later tried on his own family.

Companies

Gavi - Public-private vaccine alliance Berkley ran for over a decade; has vaccinated 1.1 billion additional children and is credited with saving over 20 million lives.
Pfizer - Cited for selling its pneumococcal vaccine to developing countries at a 95 percent discount while still profiting through economies of scale.
Novavax - Maker of an adjuvanted protein Covid vaccine; struggled to compete against the faster-to-market mRNA vaccines despite technical advantages.
Coalition for Epidemic Preparedness Innovations (CEPI) - Organization set up after Ebola to fund novel vaccines for epidemic-potential diseases; its head, Richard Hatchett, worked with Berkley at Davos to kick off what became Covax.
Rockefeller Foundation - Berkley previously ran its health programs; notes its entire global health budget was about $15 million, dwarfed by what Gavi later mobilized.

Techniques and frameworks

All-pay (ascending) auction - Classroom demonstration where every bidder pays their highest bid, not just the winner; Levitt explains the escalation trap it creates and why a listener's attempt to replicate it with his own family failed.
Technology transfer financing - Covax's strategy of paying to move vaccine manufacturing know-how to additional producers so global supply volume (and price) could scale during Covid.
Cost-per-life-saved analysis - Levitt and Berkley walk through the math showing Gavi's routine immunizations cost roughly $1,000 per life saved, with an estimated $21 to $54 return per dollar invested.

Summary

Steve Levitt talks with Seth Berkley, the epidemiologist who ran Gavi, the Vaccine Alliance, for more than a decade and later co-led the international Covax effort to distribute Covid vaccines globally. The core of the conversation is an economics lesson dressed as a public health one: Gavi delivers the WHO's full 11-vaccine schedule for about $24 per child, a 98 percent discount off US pricing, achieved by using its pooled market power to give manufacturers volume and price certainty in exchange for access to low- and middle-income markets. Berkley frames this as a genuine win-win rather than pure charity, since expanding a company's global volume lowers its cost of goods everywhere, including in its most profitable markets. The math is striking: Levitt calculates roughly $1,000 spent per child's life saved through routine immunization, and Berkley cites a $21 to $54 return per dollar invested depending on how broadly "return" is defined.

A recurring thread is why some vaccines never get made at all. Diseases confined to poor countries, malaria and Ebola among them, attract little pharmaceutical R&D because there is no paying market, and Berkley traces how Gavi and CEPI engineered artificial incentives (stockpile guarantees, purchase commitments) to get manufacturers to invest anyway. The conversation also digs into why vaccines are so hard to build and manufacture: unlike a chemically pure drug, a vaccine is a living biological product where every step of the process has to be revalidated if anything changes, and only about 7 percent of candidates that reach trials are ultimately approved. The Covid vaccine's 327-day turnaround, the fastest in history, only worked because of unfinished prior research on SARS and MERS combined with the new mRNA platform's speed.

The back half of the interview turns personal and political. Levitt discloses that his own son died of pneumococcal meningitis in 1999, four months before the vaccine that would likely have prevented it was approved, and uses the story to explain why preventive medicine fights against human psychology: prevented deaths are invisible, so people discount vaccines' enormous benefit while overweighting the rare, visible side effects. Berkley extends this into a broader explanation of vaccine hesitancy around autism, noting the childhood vaccine schedule coincides with the exact developmental window when autism symptoms first become visible, creating a false but psychologically compelling causal link. Both are alarmed by the Trump administration's cuts to mRNA research and public health infrastructure, which Berkley calls short-sighted given models suggesting a 50 percent chance of another Covid-like pandemic within 25 years, and given that the same mRNA speed advantage that helped with Covid also matters for individualized cancer treatment.

The episode closes with a lighter listener-question segment. Producer Morgan Levey relays a story from a listener who tried to replicate the "all-pay auction," a demonstration Levitt uses in class where every bidder pays their final bid regardless of whether they win, at a family dinner and lost 38 pounds without teaching anyone the intended lesson. Levitt walks through why the auction reliably traps strangers in an escalating bidding war in a classroom (once you have bid past the item's value, every marginal bid looks better than stopping) but collapsed with family members, who default to different social rules than an anonymous bidding pool, and required facilitation skill the listener didn't have.

Notable Quotes

"For every dollar invested you get a $21 return. But if you take the broader sense of what people want in terms of having good health, having a life without morbidity and other conditions, it goes up to, for every dollar invested, you get a $54 return. And there's nothing like that in healthcare." - Seth Berkley

"Mice lie, monkeys don't always tell the truth, and therefore, you have to ultimately do it in humans to really know whether it's going to work." - Seth Berkley, quoting a vaccinologist colleague

"The vaccine works against its own acceptance by effectively getting rid of the horror." - Steve Levitt

"Every single time as you think on the margin, Should I increase my bid or not? You are caught in a trap." - Steve Levitt