1. 10,000 pieces of type — Nisha Austin

2. Science’s institutional history at 250 — Jordan Dworkin

3. Anthropic develops drugs — Saloni Dattani

4. The Astera metascience essay contest — Matt Clancy

5. Trump v. Slaughter and the end of independent agencies — Willow Latham-Proenca

6. Modular housing and the HUD code — Alex Armlovich

And Dylan is on holiday this week!

10,000 pieces of type — Nisha Austin

One of my favorite 4th of July traditions is to go to the historic fort in town for a reading of the Declaration and firing of the cannons by a historical reenactment group. This year, with the sweltering temperatures, we decided to go to a reading of the Declaration at the local town hall instead. It unexpectedly included a hand-printed copy of the Declaration that was made by setting roughly 10,000 pieces of type by hand and running each sheet one at a time on cotton linen paper, the same slow way it would have been done in 1776. There’s a fitting Massachusetts thread to this: in 1776, we were the only state to ask every town clerk to copy the text into town records and every minister to read it aloud from the pulpit. Two and a half centuries later, a version of this is still happening in our town halls.

There’s a line in the Declaration which defines our unalienable rights as life, liberty, and the pursuit of happiness, rather than the more familiar “property”. Jerusalem Demsas and Kelsey Piper dwell on that choice: the historian Arthur Schlesinger read “pursuit” as practice, as in a right not merely to chase happiness but to have it. Piper goes on to describe all the ways in which Americans live lives of staggering wealth: hot water on demand, a refrigerator, a cupboard holding more kinds of tea than most people who ever lived drank in a lifetime. In 1900, Americans spent 43 percent of their income on food; the White House had no indoor plumbing until the 1840s. Medieval kings would envy us.

Jason Crawford reads the anniversary in a similar way. American history is not a story of decline but one of long, messy accumulation of things built: the reaper, the light bulb, the airplane, the interstate system, the Internet, eradicating polio and smallpox. His piece lands where I did in our foundational texts post: on Lincoln, and what we owe the institutions we inherit. What we have is precious and rare, and it should make us generous. The work is never finished, and every generation has to show up for it.

Science’s institutional history at 250 — Jordan Dworkin

For America’s 250th anniversary, Brad Wible invited a trio of pieces for Science’s Policy Forum tracing the institutional shifts that built the modern US scientific enterprise.

Ronald Daniels writes about the founding of Johns Hopkins in 1876, made possible by a $7m gift (the largest philanthropic gift in US history at the time), which imported the German research university model and laid the foundation for the postwar funding system. Ashish Arora and Sharon Belenzon cover the rise and fall of the corporate research lab, positioning the pursuit of long-horizon research in commercial contexts as a break from the more distributed systems that dominated before (typically individual inventors licensing patents to firms) and after (typically universities producing novel science, startups translating it, and firms commercializing it). And Daniel Gross and Bhaven Sampat trace the federal funding system back to OSRD, which invented the federal R&D contract (described at the time as one of the “great inventions” of the war) and grappled with many questions that remain top-of-mind today, like indirect costs, patent rights, and geographic distribution.

Anthropic develops drugs — Saloni Dattani

Anthropic made two announcements in the last week about its plans in AI for science. First was the launch of Claude Science, a platform that integrates dozens of scientific databases and bioinformatics models and currently runs on Opus 4.81. Second, that it plans to develop drugs of its own.

I found it interesting for multiple reasons: Anthropic committed to working on pre-clinical research – drug development in the lab and potentially in animals – but didn’t rule out taking its drugs through clinical trials as well. It also plans to focus on “neglected diseases” with weak commercial incentives, which it describes as potentially including both rare genetic disorders and tropical diseases, although it hasn’t disclosed any examples of diseases or drug modalities that it might work on.

This follows Anthropic’s acquisition of Coefficient Bio,2 a stealth startup, earlier this year, with roughly ten people who came mostly from Genentech’s computational drug discovery unit; and Anthropic is also building its own wet lab.

But why work on drug development at all? According to Anthropic, it’s to get hands-on experience in the field it works with companies on and to build a feedback loop to learn how to make better tools and models. As their head of life sciences Eric Kauderer-Abrams described it, there’s no substitute for being “in the trenches trying to develop drugs.”

I’d be more optimistic about Anthropic’s chances for developing successful drugs for rare genetic diseases, which tend to be ‘simpler’ in biological terms – as they’re often caused by a single gene or mutation that can be targeted with drugs and gene therapies – than neglected tropical diseases, for which the biological data to understand them, and how to target them with drugs, is often lacking in the first place. And since the best validation of drugs is testing them in human clinical trials, rather than in the lab or animals, I’m left wondering if they’ve deliberately kept that possibility open. Even aside from that, it seems to me that clinical testing and manufacturing, rather than preclinical drug discovery, are more important bottlenecks to progress on both fronts, although most AI tools so far have been built for preclinical work.

News reporting also suggests that their stated focus, on neglected diseases, helps them avoid stepping on pharma companies’ toes. In a sense, thanks to the Orphan Drug Act, faster approval processes, and rare disease priority vouchers, rare diseases aren’t as neglected as they used to be and pharma companies do have some incentives to develop drugs for them (and they now make up roughly half of annual new drug approvals) – but there are so many different rare diseases that the rationale probably still works.

But building labs and hiring top wet lab biologists seems like a first step for biological validation and early drug development rather than an end goal in itself; and since clinical trials and manufacturing are the bigger bottlenecks, delivering on Anthropic’s public mission would probably mean going much farther than that.

The Astera metascience essay contest — Matt Clancy

The most interesting thing I read recently was the winning entrants in the Astera metascience essay contest, for which I was one of the judges. While not every winning essay was related to AI (see these two, for example), the opportunities and challenges of AI dominated the contest:

• It’s a bit of a truism that AI and data are strongly complementary; but what data should you collect? What if we could use scaling laws as an input into this decision process? Collect data and train models on it, until we can see the trendlines emerge, which will let us estimate how much data we would need to collect to achieve a given performance (more from Peter Koo’s winning essay).

• A major problem in science is the “file drawer problem”, wherein null results don’t get published, in part because writing up any result takes time and energy, and the rewards to doing so are very low for academics (more here). But if AI agents are integrated into the research process, they have all the material they need to do this for us at almost no cost. Could we finally open all (well, more) file drawers? (more on this theme from Niveditha Iyer’s winning essay)

• One of the most important problems in science is knowing what question to study. With AI that can finally parse text as good as most people, we are in a position to map out the enormous network of cause and effect that the collective literature has studied. With a map in hand, could we identify promising gaps we would otherwise miss? Or could we at least make progress on automating the identification of important questions? (more from Prashant Garg’s winning essay)

• The use of benchmarks to assess the quality of different AI models has been a notable hallmark of AI progress, and one that seems likely to spread across science more generally. But how do you design good benchmarks? Shaamil Karim and Jaeeon Lee both proposed ideas related to this.

One more interesting AI-related challenge: when we ran the winning essays through Pangram, some of them were flagged as using AI (though the share that used AI was even higher among the ones that had been triaged out, prior to the AI check).

I think all future essay contests are going to need to think hard about an AI policy. In our case, we cared about the arguments and ideas, not the authenticity of the prose (or even the provenance of the idea?), and so we decided not to factor AI use into our decisions. But questions remain: what’s the right size for a prize, when the cost of writing an essay has declined? To ensure the prize is incentivizing ideas that outcompete what you would get by simply asking an LLM for ideas, maybe contest organizers of the future should use AI to generate essays in response to the contest announcement and covertly submit them; a prize is only awarded if the judges pick essays that were not generated this way.

Trump v. Slaughter and the end of independent agencies — Willow Latham-Proenca

To add even more uncertainty to the fraught landscape of energy planning and regulation, the end of the Supreme Court opinion season last month brought a highly controversial ruling in Trump v. Slaughter, where the court held that the president can fire commissioners at independent agencies without cause. While some argue that this is an appropriate tightening of elected authority over independent bureaucracy, a less-hopeful view (articulated by Robinson Meyer in Heatmap last week) is that the ruling represents a continuing erosion of Congressional authority – in this case, powers intentionally delegated to independent technical experts – relative to the executive branch. While we’re unlikely to see major short-term policy changes on the energy front – the current FERC and NRC are already closely aligned with the goals of the current administration – over the long term, this could reshape policy at FERC and NRC in ways that are hard to predict. Robinson Meyer takes the strong view here, arguing that Slaughter essentially turns the commissions into agencies – implementers of the president’s agenda, rather than independent arbiters (Ben Schifman of IFP has an interesting take in the same thread arguing that ratemaking bodies like FERC will remain exempt - time will tell if SCOTUS will re-litigate that argument).

Modular housing and the HUD code — Alex Armlovich

Kimberly Burnett has a new piece at Niskanen on HUD's new best-practices guidance for states and localities, focusing on one key recommendation: regulate modular housing by harmonized and objective performance standards rather than prescribed construction methods. Modular is still only ~3% of US single-family homes against 28% in Japan, and we’ve lived through many modular “hype cycles” in the US. But factory-built manufactured housing using the national HUD code is still between 5% to 10% of single family housing completions in the US, suggesting a key binding constraint to modular methods is regulatory fragmentation: duplicative factory-and-site inspections, codes written for on-site assembly, and thousands of local building departments with no capacity to evaluate factory-built homes. National model code standards for modular homes, like ICC/MBI 1200 and 1205, are spreading but have not yet succeeded on the scale of Congress’ choice to directly preempt local building codes nationwide with the HUD Code in the 1970s.

And finally, a couple additional highlights worth sharing:

• Institute for Progress published a detailed report on reforming Section 106 of the National Historic Preservation Act, which has expanded well beyond its original “stop, look, and listen” function into a significant source of delay for energy and transmission projects. The report proposes right-sizing the review process by limiting scope, capping visual-impact studies, and setting statutory deadlines, in line with recent NEPA reforms.

• California’s SB 79, the Abundant and Affordable Homes Near Transit Act, took effect July 1, establishing statewide zoning standards that allow increased residential density near major transit stops across California’s urban counties. Cities are taking markedly different approaches to implementation, with some complying, some using the law’s exemption provisions to delay.

Housing

1. The 21st Century ROAD to Housing Act passed the Senate 85-5 and awaits the president’s signature (or, absent a veto, will have automatic enactment by July 10th). 45+ provisions cut red tape, overhaul financing, and provide carrots and sticks for local reform. (June 24)

2. New York took two big permitting steps: Governor Hochul’s SEQRA modernization is reportedly final, and the mayor launched the SPEED Task Force report projecting two years off rezoning timelines. (May 15)

3. “Are we kind of being pricks?“ A Marblehead resident’s viral question captures what happens when towns comply with zoning law by putting housing capacity on a golf course where nothing will get built. (May 8)

4. Jerusalem Demsas also shared her reflections on Marblehead. The town’s story shows what happens when local democracy mediates land-use decisions, and when compliance isn’t the same as housing production. (May 28)

5. Nolan Gray goes deep on California data: ADU permits have produced nearly 150K new units, the density bonus law is facilitating thousands more, and major reforms are less than a year old. (May 8)

6. Michael Andersen of Sightline argues the picture outside California is just as encouraging: states across the country have only recently passed reforms and results are beginning to show. (May 8)

7. A new Urban Institute piece asks what happens after big upzonings in strong markets. Answer: a lot. The evidence that significant reform leads to construction continues to pile up. (April 17)

8. Our Zoned Capacity post got shoutouts from NYC’s Deputy Mayor and Matt Yglesias. In high-rent cities, “Missing Massive” upzoning is essential for unlocking construction that pencils out now. (April 10)

9. Evan Soltas translated his LA permitting paper into a NYC estimate: each year of permitting time saved is equivalent to an 8% drop in construction costs. (April 3)

10. The housing burden is falling disproportionately on new buyers and renters: new homeowners now pay 26% of income, renters who recently moved pay a record 33%. (April 17)

11. Arpit Gupta and Steve Teles diagnose why factory-built housing keeps failing: boom-bust cycles, patchwork regulations, transport costs, and parcel assembly challenges. (April 3)

12. Everyone’s talking about building codes. California YIMBY summarized a UCLA Lewis Center report on how the code-writing process works and how we got where we are. (April 10)

13. Vincent Rollet’s MIT paper on zoning and redevelopment is getting attention from city planners; it’s the kind of structural model that can discipline a rezoning’s projected supply impact. (April 24)

14. Pew’s new address chain video explains how new construction frees up housing across incomes in months, not decades. They also launched a report on pre-approved building plans. (May 28)

15. The NYT editorial board endorsed Massachusetts’ “Legalize Starter Homes” ballot measure, which would require towns to allow homes on small lots in areas served by public water and sewer. (June 3)

16. A VoxEU column argues the relative price of US structures has risen 80% since 1970, making construction productivity a drag on economy-wide growth, not just a sector-specific concern. (June 10)

17. Brian Potter asks where the economies of scale are in construction. The “Idiot Index” (cost of a good vs. raw materials) for housing is roughly 2, not far from the highly automated car industry’s 1.8. (June 3)

18. Florida’s insurance crisis may be easing: after tort reforms, insurers report sharp drops in litigation costs and premium growth. The state had 76% of US insurance lawsuits from just 9% of homeowners. (June 10)

19. Sightline’s new report finds apartments are the climate solution hiding in plain sight: more energy efficient, more likely heated by electricity, and located where people can walk, bike, and take transit. (May 15)

20. Larry Katz’s 1981 article on how growth controls drove San Francisco’s home prices from the national median to the highest in the country in just one decade. A hauntingly contemporary time capsule. (April 17)

21. Oliver Kim reviews a book complicating YIMBYism’s invocation of Singapore: 99-year land leases create a looming political problem, and the city is already so dense that upzoning can’t unlock value the way it does in US cities. (May 8)

22. IFP launched its Transit Abundance Playbook: fifteen memos on why the US pays the world’s highest prices to build transit and how to bring costs down, including on bus procurement. (June 24)

    Energy

23. The House passed a bipartisan geothermal package with NEPA categorical exclusions and faster leasing. But the binding constraints (resource characterization and transmission) remain unaddressed. (June 10)

24. Travis Kavulla explains why more places will look like PJM as data center demand rises: the grid has little excess capacity, and new demand requires costly new investment. (May 28)

25. Matt Yglesias argues “conventional environmentalism” is too focused on efficiency and missing the upside of truly abundant energy. We think cheap energy helps, but some sectors have biotech or materials bottlenecks energy alone won’t solve. (May 15)

26. The President’s FY27 budget proposes 11% cuts to DOE civilian energy programs, and the department has been slow-walking grants it already committed to and slowing new solicitations that keep appropriated funds moving. (April 17)

27. Data centers aren’t the only new large loads: heavy industry is increasingly worried about being outbid for scarce generation resources by data centers willing to pay above-residential rates. (June 3)

28. Zane Kasher’s job market paper finds a 12% drop in home prices within 3 miles of a wind farm. Aesthetic responses vary by country, and blinking lights are a big but cheaply fixable driver of annoyance. (April 24)

29. The energy-growth link is less clear-cut than you’d expect: better energy sector productivity may not affect GDP much, but eliminating power outages could matter enormously and growth effects may be asymmetric. (May 8)

30. LBNL updated its Retail Price Trends and Heatmap/MIT launched an Electricity Price Hub with utility-level data. Price patterns remain fragmented and geography-specific. (April 3)

31. Loudoun County gets 45% of its revenue from data centers, meaning lower property taxes, better schools, and general abundance. But there’s a flip side: residents are fighting the transmission lines needed to power them, and the county chair says people no longer care about the tax benefits. (May 8, June 10)

32. Semianalysis published a comprehensive analysis of space-based data centers. Base case: they don’t reach cost parity with earth until ~2040, but we shouldn’t completely discount the idea. (June 24)

    Innovation Policy

33. OMB published a 412-page proposed revision to federal science funding guidance. Concerns include political pre-award review, mandatory termination clauses, and sweeping foreign collaboration bans. Comments due July 13. (June 3)

34. NSF launched the $1.5B X-Labs initiative: decade-long, milestone-driven funding for independent research teams working outside traditional university structures. A small bet (~1% of NSF budget) but an exciting one. (May 28)

35. A new RFI floats capping PIs at two to four grants. The last attempt (in 2017) was dropped within a month. Supporting younger scientists is the strongest case; responses due August 3. (June 17)

36. The NSF FY27 budget proposes a metascience unit and $50M for Tech Labs (exciting), but also proposes eliminating social sciences and cutting funding 55% (not exciting). Congress largely ignored a similar proposal last year. (April 10)

37. The NIH indirect cost cap fight is over (for now): the administration let the Supreme Court deadline pass. Jeremy Berg’s explainer on how IDC rates actually work is worth your time. (April 10)

38. New medical phrases increasingly come from more specialized scientists and are introduced in less accessible language, both of which slow idea diffusion. Our living literature review program was motivated by exactly this concern. (June 25)

39. Three reads on what happens when AI gets good at producing science’s proxies (theorems, papers) before the processes they stand in for. We highlight Bessis on math, The Paper Factory, and Ted Chiang’s prescient 2000 short story. (May 15)

40. Many funders say they support “high-risk, high-reward” research, yet science keeps getting more conservative. Standard review processes aren’t structured to evaluate risk, and HHMI’s selection model shows promise. (May 8)

41. Ben Reinhardt argues the Vannevar Bush era of “free play of free intellects” is ending. Researchers should think about who their customers are, and independent research labs need to prove they can deliver superior quality. (April 17)

42. Eroom’s Law — drug R&D costs keep going up — may persist not because of science but because only a few large firms can commercialize drugs. Innovation is rate-limited by their pipeline capacity, not discovery. (April 10)

43. Two very different visions for reforming science funding: Gibson in City Journal wants lotteries, scouts, and indirect cost caps; Fineberg in PNAS wants sustained federal investment and talent pipelines. Both want more support for young scientists. (April 3)

44. Grant Witness is now tracking NIH and NSF funding curves in real time which is essential for monitoring whether agencies are spending what Congress approved. (April 3)

45. arXiv’s CS chair announced year-long bans for authors submitting papers with unchecked AI content like hallucinated references. Holding authors accountable for accuracy seems worthwhile, even if the post-ban peer review requirement goes too far. (May 28)

46. Refine, an AI review tool for economics, published a benchmark study, winning ~90% of head-to-head matchups against frontier LLMs. More tool-builders should run and publish studies like this. (June 24)

47. Emergent Ventures announced new grants for metascience policy entrepreneurs and science communicators. (April 17)

    Clinical Trials

48. Dario Amodei’s latest post focuses heavily on clinical trials. AI will flood the regulatory pipeline with drug candidates, so we need major reforms to make sure therapies don’t languish for years. (June 17)

49. ASCO 2026 highlights: Daraxonrasib doubled survival for metastatic pancreatic cancer and lorlatinib showed 55% of ALK-positive lung cancer patients still progression-free at 7 years. On the screening side, the NHS Galleri blood test quadrupled cancer detection but mostly caught stage III rather than truly early. (June 4)

50. FDA Commissioner Marty Makary resigned after friction with RFK Jr. He announced promising ideas but also oversaw departures of career scientists and eroding evidentiary standards. (May 15)

51. Joe Lonsdale’s FDA reform agenda pushes promising ideas: surrogate endpoints, Australian-style Phase 1 trials, and eliminating wasted time between trial phases.(June 3)

52. The FDA announced “Real Time Clinical Trials” but the rollout was muddled. Adam Kroetsch writes it’s genuinely promising but requires the FDA to define how they use live data and build interoperability standards. (May 15)

53. We might finally get a vaccine against Lyme disease but the phase 3 confidence intervals were wide and the trial missed its primary endpoint. We had one in the ‘90s and lost it to unfounded fears. (April 3)

54. Drug development fails 95% of the time. Abhishaike Mahajan explains the financial engineering tricks, from hub-and-spoke companies to synthetic royalties, that keep biotech money flowing despite the brutal odds. (May 8)

55. In Development has a great article on pooling regulatory reviews across African countries so manufacturers don’t have to file separately in each one. The African Medicines Agency launched in 2025 but faces big challenges. (May 28)

56. Niko McCarty ran a bounty to surface ideas for cheaper biology experiments: 430 submissions, 20 prizes awarded. Ideas included a protein printer from DNA origami and gel-filtration protein synthesis. (April 10)

57. Saloni discussed clinical trial reform on the Works in Progress podcast, from why ethics reviews are inefficient to how Australia made earlier-stage trials simpler and faster. (April 24)

58. Niko McCarty and Saloni wrote a biotech/medicine roundup: organ donations are vastly more efficient, semaglutide failed for Alzheimer’s, and a transformative pancreatic cancer drug succeeded in phase 3. (April 24)

    Europe

59. Paul Krugman says Europe is keeping up with the US; Luis and Pieter Garicano say it’s falling behind. The biggest question: does US tech dominance actually matter for European living standards? (May 15)

60. European tourists visiting the US for the World Cup are going viral with their own “driving around tests“, though free hotel rooms and celebrity meet-and-greets don’t come standard with road trips. (June 17)

61. Philippe Aghion identifies Europe’s gaps: no single market, insufficient long-run research funding, no DARPA-like agencies, a culture that doesn’t celebrate risk-taking, too much red tape, and more. (June 17)

62. Samuel Hughes asks whether European housing politics should be Americanized. YIMBYism is effectively nonexistent on the continent, despite housing shortages arguably worse than America’s. (June 17)

63. Kelsey Piper and Alexander Kustov explain why America is better at immigration than Europe: flexible labor markets let immigrants work, integrate, and build public support. (April 3)

64. Europe 2031 paints a scenario where AI access becomes nationalized and Europe’s assumption that it can free-ride on US tech breaks down. Eerie timing, published shortly before the US banned foreign access to Anthropic’s latest model. (June 17)

65. The ERC announced stricter application rules for 2027, likely in response to AI-generated proposals flooding review panels. NIH also capped at six applications per year. The optimal solution is an open metascience question. (April 24)

66. The Centre for British Progress released a paper on street votes: residents collectively hire an architect, agree on upzoning, and vote to allow new housing with value capture to homeowners. (April 24)

    Team Announcements

67. Alex Armlovich published Zoned Capacity Is Like an Artificial Oil Deposit, on why cities can claim they have room for thousands of new homes on paper and still not build them, and

68. Alex Armlovich published Tokyo Land Is Still >$85 Million an Acre, looking at what happens to land values when a megacity achieves housing abundance.

69. Matt Clancy published How do you get Abundance and Growth?, an outline of the Abundance and Growth fund’s plans for what kinds of work to fund in 2026.

70. Matt Clancy published A Little Progress Is Worth a Trillion Dollars, with a web tool and interactive explainer for valuing different growth scenarios.

71. Matt Clancy published An Atlas of Innovation, on how to decide when to use a prize, a research grant, or something else.

72. Matt Clancy published Social Science at the NSF, on how history rhymes with today’s debates about federal science funding.

73. Jordan Dworkin published US Science Agencies Have Money; Can They Spend It?, on why and how to keep track of government spending on R&D.

74. Jordan Dworkin published Our Grandchildren’s AI-Science Bottleneck, on why slow fields may determine the future pace of discovery.

75. Dylan Matthews published How Do States Share Ideas?, a brief introduction to the study of policy diffusion.

76. Saloni Dattani published Why Were Covid Vaccine Trials So Fast?, on how the timeline to develop coronavirus vaccines blew many predictions out of the water.

77. The Abundance and Growth team published What got us into abundance and growth, about the works that were personally foundational to our interest in abundance.

    Writing from our Grantees

78. Witold Więcek co-authored a JAMA perspective on the FDA’s Bayesian guidance.

79. IFP published analysis of Section 901 of the ROAD Act, a prevailing wage report, and the Transit Abundance Playbook.

80. Ruxandra Teslo wrote on how the pancreatic cancer breakthrough came about.

81. Lauren Gilbert published “Immigration and House Prices“ in her Migration Living Literature Review.

82. Ben Holland joined David Roberts on the Volts podcast to discuss why climate funders don’t fund housing.

83. The Foundation for American Innovation hosted the Energy Imperatives Summit in June.

    More Abundance

84. Marc Dunkelman and Alex Mechanick grapple with how democracies make decisions when people disagree, and why our current processes put too little weight on growth. (May 8)

85. The USPS owns 8,500+ properties. Rebuilding at surrounding density levels could unlock 117,000 units and earn nearly $1B/year in leases. State capacity and housing supply in one policy? (May 28)

86. The Jones Act waiver data is in: 45 voyages from 35 ships in just two months, mostly moving oil from the Gulf Coast to the West Coast and Puerto Rico. That’s a lot of pent-up demand. (May 28)

87. Sam Peltzman documents a big negative break in American happiness starting around Covid that has persisted well past the pandemic. Both GSS and Gallup data confirm well-being remains near unusual lows. (April 10)

88. 31% of World Bank policy reports are never downloaded, but the bank still writes them. Todd Moss argues this is a tragic waste of talent that could be solving actual problems. (April 10)

89. The Census Bureau is debating noise infusion for anonymizing data. A Trump Commerce Department proposal to ban it entirely could lead the Bureau to just not release certain data at all. (June 24)

90. Charles Mann and Virginia Postrel launched Everyday Abundance, a new podcast. First episode: the history of brushing your teeth. If anyone can make that interesting, it’s these two. (June 10)

91. In Development launched with a piece on cash transfers. A dollar is worth roughly 250x more to someone in extreme poverty. Harvard’s research ethics committee nearly killed GiveDirectly’s first RCT on grounds that giving people money might harm them. (April 10)

92. A big FRI survey asked AI experts, economists, superforecasters, and the public to predict AI’s economic effects. Surprisingly, the groups weren’t far apart, and all fall well short of “double-digit growth” forecasts. (April 3)

93. Alex Imas on the rise of “relational work”: as AI makes information cheap, irreducibly human elements — trust, presence, judgment — become what people spend more on as they get richer. (April 17)

94. The Labor Automation Forecasting Hub collects predictions on everything from specific jobs to specific states. The forecasts will be wrong a lot, but the clarity is appreciated. (April 24)

Most people didn’t believe it was possible to develop vaccines in the timeframe needed for them to be useful. Some looked at past vaccine timelines, which had averaged roughly 8 to 12 years, and thought this one would be similar. Others thought that, even though it was an emergency situation, it would still take at least a year and a half, or two years, or maybe even four.

I had a different conclusion. In a piece I wrote in the summer of 2020, I explained why I believed that vaccines would most likely arrive within a year of the beginning of the pandemic (placing a 58% probability on enough doses for 25 million Americans being approved and available between October 2020 and March 2021, with my central estimate landing around February 2021).

We now know how the timeline panned out and my forecast was, if anything, slightly too pessimistic, since vaccines first became available three months earlier, in December 2020. In this post, I want to go into more detail and take a look back at what happened. Why were Covid vaccine trials so fast?

Subscribe now

The scientific foundation

One of the most surprising things about Covid vaccines is how many of them there are: mRNA vaccines (Pfizer and Moderna), viral vector vaccines (Oxford-AstraZeneca and Johnson & Johnson), protein subunit vaccines (Novavax), inactivated whole virus vaccines (Sinovac and Sinopharm), and others. Although their protection dropped as the virus evolved into new strains, I take this bounty of options as a result of the coronavirus being relatively easy to develop vaccines for.

Part of the reason is the disease itself. With some infections, like HIV, no one clears the virus naturally or develops lasting immunity to it, so it’s hard to know what a vaccine should mimic. Covid-19 was very different: it was evident early on that most people recovered and developed antibodies that could neutralize the virus. This suggested it was possible to prime the immune system with a vaccine.

Another reason was prior research into coronaviruses. Work on SARS and MERS from earlier coronavirus outbreaks meant scientists already understood some of their features. They had identified the key antigen the immune system reacted against – the spike protein – and had animal models and laboratory assays ready to go. Some vaccines against animal coronaviruses had already been developed. Candidate vaccines for the earlier SARS virus had also been developed for humans (and shelved after the 2003 SARS epidemic was contained). When SARS-CoV-2 arrived, lots of this groundwork could be picked up.

There’s another feature that explains why some vaccines ended up highly effective: a stabilized spike protein. The spike protein gives the coronavirus its crown-like appearance, and is the main ingredient in most Covid vaccines. But interestingly, it changes shape during an infection. Before it fuses with a cell, it sits in a compact “pre-fusion” form (on the left, below); afterwards, it springs into an elongated “post-fusion” form (on the right). The former boosts immunity most because it’s what antibodies generally encounter, before the virus has entered cells. But when the spike protein is isolated for a vaccine, it tends to collapse into the later, post-fusion form.

Over the past decade, advances in structural biology, especially in cryo-electron microscopy, have helped scientists see the pre-fusion shape of coronaviruses directly. With that knowledge, they introduced a few mutations to stabilize the spike protein in its pre-fusion form, better for the immune system to recognize. This is a big reason why vaccines with the stabilized pre-fusion protein (including the mRNA vaccines and Novavax) generated much stronger neutralizing antibody responses and were likely much more effective than vaccines without the stabilized form (including the AstraZeneca and Sputnik V vaccines).

All of these features meant the chances of a vaccine succeeding were somewhat high. They’re also why they could be designed very fast – of course, with a lot of work on the part of scientists. It took two days to design the Moderna vaccine, which then spent 63 days in preclinical testing before entering clinical trials, versus a weekend and about 100 days for the Oxford/AstraZeneca vaccine. But the vaccines would still need to be tested in clinical trials and gain regulatory approval afterwards, and this process usually takes several years.

Parallel trials

The timeline for vaccine development typically goes like this. A candidate vaccine is first designed and tested in the lab and in animals, as ‘preclinical testing’, which takes around 1.5 years on average. Next, in ‘phase 1 trials’, it’s tested in a small number of participants for basic data on safety and the immune response, taking around 2.5 years. Phase 2 trials expand this to hundreds of participants to refine the dose and gather more data on immunogenicity and safety, and take around 3 years. Then phase 3 trials test efficacy and safety in hundreds or thousands of people, and take around 2.5 years. After all these stages, all the data is submitted to the regulator as a package, which reviews it, usually taking another year.

Historically, successful vaccines have taken around 10.7 years to make it through the whole pipeline. Each phase happens sequentially and functions as a ‘go / no-go’ system, where candidates are dropped if they flop at a stage.

Covid vaccine trials took place quite differently. Rather than running sequentially, the phases ran in parallel, as you can see in the diagram below. For example, phase 1 and 2 trials were combined, collecting the relevant data for each at once, or phase 2 and 3 were combined in the same way. Ultimately, the same amount of data was collected in a shorter timeframe, by overlapping the phases rather than doing them one by one.

This was a major reason it was possible to condense the whole timeline so much. But even then, the trials still ran much faster than usual. For example, Pfizer’s combined phase 2/3 trial then took about 4.5 months, but it usually takes an average around 2.5 years for phase 3 trials alone. What made that possible?

Subscribe now

Easier recruitment

One of the biggest delays in clinical trials is the process of recruiting enough participants to run them at all. Only half of trials meet their original target for recruitment, and a fifth revise their target downward. But the Covid vaccine trials recruited huge numbers, fast. The phase 3 trials for Moderna and Pfizer enrolled around 30,000 and 44,000 participants respectively, making them some of the largest clinical trials in history.

I think there were a couple of reasons this was easier than usual. One was that people were simply more willing to volunteer. It was a pandemic after all, and everyone’s attention was on it, plus, many people were stuck at home without much to do anyway. Interest in volunteering for trials was very high. Dr Jim Kublin, who helped run a clinical trial network for Covid, recalled: “My colleagues were enrolling a phase one study for the Moderna vaccine, and they had like 10,000 people sign up on their website. They needed 40 people to enroll.”

The infrastructure to recruit participants was also more streamlined than usual. The National Institute of Allergy and Infectious Diseases (NIAID) created a clinical trial network with a shared volunteer registry called COVID-19 Prevention Network (CoVPN), from four existing NIAID trial networks, to supplement recruitment into trials. Volunteers could sign up with a short screening questionnaire to the network as a whole and be routed to whichever trial fit. The registry helped support recruitment for the late-stage trials of Moderna, AstraZeneca, Janssen, and Novavax, and by autumn 2020, almost half a million Americans had signed up through it to take part in Covid vaccine trials. Meanwhile in the UK, a similar shared volunteer registry recruited participants through the NHS. Over half a million volunteers signed up in total, feeding into trials including those for Novavax and Oxford/AstraZeneca.

Covid was Everywhere All At Once

It also took less time to test new drugs and vaccines because the disease was spreading rapidly. It’s challenging to run trials fast when diseases are rare or slow to develop. Taking a simple example, imagine a trial where 100 people are given a placebo and 100 people given a vaccine. If only 1 or 2% of them (so 1 or 2 people) caught Covid during the trial as the base rate, it would be very hard to spot a reduction from that already-low number. Even if vaccines reduced infections by half, you’d hardly be able to tell in this example, not with much confidence anyway. Without a much larger sample size, a difference between the two groups could simply reflect noise.

This reflects the ‘statistical power’ of the trial, and that it’s easier to figure out whether a vaccine is protective if new cases are arising quickly (i.e. the disease has a high incidence rate) or if the sample size is very large. You can see this in the model I made below. It’s also why it’s much harder to develop new vaccines for diseases we’ve reduced massively: it would be a struggle to test new polio vaccines in rich countries that have eliminated the virus, for this reason.

Fortunately for trial statisticians1 – but unfortunately for everyone else – Covid was spreading quickly. During Pfizer’s trial, the monthly incidence rate (the % of the population infected for the first time, per month) was around 0.5%. By September 2020, using seroprevalence data, the CDC estimated that up to 23% of people had already been infected in the hardest-hit areas, and just below 10% across most of the US.2 In this situation, what matters most is how fast new cases are accumulating while a trial is running: during the autumn and winter of 2020, outbreaks were developing fast, and cases were racking up quickly in the placebo group, which made it easier to see a reduction in the vaccine group. Within a few months, Moderna’s trial recorded 185 Covid cases in the placebo group versus 11 in the vaccinated group, and Pfizer’s trial recorded 162 versus 8.

Subscribe now

Emergency Use Authorization

People often point to the ‘Emergency Use Authorization’ pathway as the main reason Covid vaccines were introduced so fast – that they weren’t approved the regular way, and faced lower standards of evidence than usual. Let’s dig into this a little.

In general, an emergency use authorization (EUA) is a conditional authorization that lets the FDA allow a product to be used during a declared public health emergency before it has gone through full approval, and which can be revised or revoked at any time. And the bar is intentionally low: the FDA only has to find that the product ‘may be effective’ and that its known and potential benefits outweigh its known risks, given that there is no alternative that’s approved.

The standard of a product that ‘may be effective’ is much weaker than the evidence of effectiveness required for full approval. One example was the emergency authorization of hydroxychloroquine as a treatment for Covid in March 2020, which relied on observational data rather than efficacy trials. A few months later, when the RECOVERY trial showed, in a randomized study, that it had no benefit, its authorization was pulled.

Does this mean that the standards were so relaxed for vaccines? Not really. The vaccines faced a stricter EUA, called ‘EUA plus’. In its June 2020 guidance, the FDA demanded large randomized controlled trials demonstrating an efficacy of at least 50% before a vaccine could be licensed. Then in October, it added more conditions an EUA would have to meet: a median of at least two months of safety follow-up after the final dose, and review by an independent advisory committee in a public meeting.

This meant that the standards for Covid vaccines were similar to a full approval, with two notable exceptions. 1) While full approval required a complete manufacturing dossier, with validation and facility inspections, the EUA accepted a less complete package, and deferred the validation and inspection until later. 2) While full approval required a longer safety and durability record, typically at least six months, the EUA accepted a two-month median follow up time after the final dose.

It’s worth noting that the vaccines did fulfill the requirements of a full approval later on, starting with Pfizer’s vaccine in August 2021.

Incidentally, the EUA’s shorter follow-up requirement was the subject of political controversy. When Trump complained that the vaccine was being delayed until after the election, it was because of the follow-up requirement, which wouldn’t be complete until later – Pfizer only reached the median two months of safety follow-up in the third week of November, over two weeks after the election, and filed its EUA request on November 20.

I don’t have strong views on the requirement, but it seems hard to strike a balance between collecting enough safety data to build public trust versus speed up the roll-out, and unlike all the other reasons for acceleration, this was the only way in which the safety threshold was laxer than a full approval, which was achieved later on. It seems like the crux is how important public trust would be in increasing vaccination rates and whether the requirement achieved that.3

Rolling regulatory review

Typically, when researchers finish their trials, they compile a complete dossier with lots of information about the trial, the product, the manufacturing process, and so on, and submit it to the regulator, which then reviews the complete package.

It usually takes a while to review the complete package. In the US, the FDA’s goal is to review the data and make a decision within 10 months of receiving the application, or 6 months if it qualifies for ‘priority review; in practice it takes an average of eleven months. Meanwhile in the EU, the European Medicines Agency’s goal is 210 days (roughly 7 months), but the clock only includes the days the agency is actively reviewing; it pauses whenever the regulator sends questions back to the company and waits for answers, so in reality, a review usually stretches well beyond a year.

The Covid vaccines faced a different process. In Europe, they faced a ‘rolling review’ process instead, where regulators assessed the data – on the manufacturing, preclinical research, trial results, etc. – as it arrived, instead of all together right at the end. So by the time the final data was in, most of the review had already been done, and it took much less time to reach a decision. Pfizer submitted its formal EU application on 30 November 2020 and received conditional authorization about three weeks later, on 21 December, because the rolling review had already evaluated the data as it came in.

In the US, the rolling review process applied to the full approval, rather than the initial authorization, since the vaccines went through the Emergency Use Authorization pathway instead, which began on 20th November for Pfizer and granted approval on 11th December 2020. The complete application, which did involve rolling review, was completed in May 2021 and received full approval on 23 August 2021, roughly three to four months, still much shorter than usual.

Subscribe now

Can we apply some of these processes to speed up other drugs and vaccines?

By this point, you’re probably wondering why the same processes couldn’t apply to many other drugs and vaccines. I’ve wondered the same thing, and I think it’s true that various parts of this process could apply elsewhere too – such as clinical trial networks to recruit participants, more fundamental research into pathogens and candidate vaccines, and better trial design. (The Covid experience is a reason I became interested in clinical trial reform more broadly.)

But some parts would be challenging to scale up more widely. Rolling regulatory review, for example, is more demanding on the part of regulators. In the usual pipeline, most drugs or vaccines fail before they reach the finish line and are submitted for review, which means regulators are saved the time of reviewing lots of applications that would have been withdrawn anyway. (You could, of course, expand rolling reviews by hiring and training more regulatory staff, and reduce the trade-off!)

Parallelized trials are also challenging. When trials are run sequentially – with phase one, phase two, and then phase three – the sequence functions as a ‘go / no-go’ system, where developers can drop candidates they think are unlikely to succeed at the next stage. It saves time to run trials in parallel, but it also raises financial risk: instead of spending money on the next phase only once the previous one has paid off, a vaccine developer would have to commit to spending it up front.4 A similar problem applies to building production capacity during the trials, since that capacity would be wasted if the vaccine doesn’t work.

This would have been especially financially risky during a pandemic, because of the degree of uncertainty at the outset. What if the virus had actually been contained by the time the trials were complete, as was the case during the 2003 SARS outbreak? Or what if the fundamentals of the virus were misunderstood and the vaccines were designed incorrectly? By the time such large trials were complete, a lot of money might have been sunk into trials that would ultimately fail. This is where Operation Warp Speed comes in: in my view, it made it possible for a lot of the processes above to function at a lower financial risk, and in a more coordinated way.

Operation Warp Speed and its analogs

Operation Warp Speed was launched in May 2020 by the Trump administration, as a partnership between the Department of Health and Human Services and the Department of Defense. Its goal was to produce and deliver 300 million doses of safe and effective Covid vaccines, with the first doses becoming available by January 2021.

To make it happen, Operation Warp Speed funded the development of a range of candidates and their clinical trials, including the phase 3 trials of Moderna, Janssen, AstraZeneca, and Novavax.5 The CoVPN trial network was developed as part of Operation Warp Speed. Operation Warp Speed also paid for companies to start manufacturing their vaccines at industrial scale while trials were still running, so that millions of doses would be ready to ship once vaccines were authorized.

Importantly, it also committed in advance to buying hundreds of millions of doses of a vaccine that succeeded, meaning that the companies wouldn’t face a massive loss if the outbreak ended up petering out before vaccines were ready. This idea, called an advance market commitment, was originally developed by economists Michael Kremer and Rachel Glennerster to successfully speed up new pneumococcal vaccines in the 2000s, before it was adapted for the pandemic. As then-secretary of Health and Human Services, Alex Azar, described it, the government would invest more than 10 billion dollars to “de-risk vaccine companies’ development efforts and manufacture product in advance.” By the end of the year, it had committed around 13 billion dollars.

As I described earlier, running trials in parallel and manufacturing before trials are complete makes things faster but is financially risky, because a company commits money up front that would be wasted if the candidate vaccine failed. Operation Warp Speed absorbed much of this risk. (One thing that Operation Warp Speed didn’t do was change the regulatory bar: the EUA and rolling review procedures were set by the FDA separately.)

Other countries also used analogs to Operation Warp Speed to accelerate vaccine development and secure doses for their populations. The UK, for example, ran a Vaccine Taskforce, set up in spring 2020 and led by Kate Bingham, which also backed a spread of candidates across different technologies, made advance purchase agreements, and funded manufacturing. The EU ran a Vaccines Strategy, launched in June 2020, which used advance purchase agreements as well: the EU Commission financed part of companies’ upfront costs in return for the right to buy doses. Since it negotiated for all 27 member states, it had a lot of bargaining power, but it started much later; its first major agreement was signed months after the UK and US. And lastly, COVAX created a pooled procurement mechanism for countries around the world, aimed to secure doses for lower-income countries that couldn’t make their own deals (though they fell below their target of later delivering vaccines, delivering half of their target of 2 billion doses by the end of 2021 as high income countries struck bilateral deals quicker).

Operation Warp Speed-type efforts probably work best for large problems that need coordination across many actors, where the goal is well-defined and commercial markets won’t deliver it at the speed required, because the private returns are too small relative to the social value. But it’s also possible to work faster across the board, by creating clinical trial networks and shared volunteer registries to speed up recruitment, innovative trial designs that collect data more efficiently, investment into early-stage research, more regulatory capacity to review evidence as it arrives, and many other ideas we’ve shared on this blog.

Subscribe now

Faster trials saved millions of lives

The speed of vaccine development during the pandemic was unprecedented, but it wasn’t unforeseeable. It came from a coordinated effort that involved dedicated research funding, clinical trial networks, parallel trial phases, rolling regulatory review, and parts of the review being deferred rather than skipped. What eased the process were decisions to de-risk the process financially, with upfront funding and advance commitments to buy the vaccines if they worked.

The result was extraordinary. Millions of deaths globally were averted by Covid vaccines in just the first year of vaccination alone.6 We take it for granted today, but it was not a given that trials would be completed within a year. Nor was it a given that the vaccines would have around 95% efficacy against the original strain of the coronavirus, which was much higher than most scientists anticipated: the median forecast was 50%. Speeding up the process saved millions of people.

When it comes to clinical trials, it’s easy to assume the process is so bureaucratic that we couldn’t move quickly even if we wanted to. The pandemic shows that is not the case. When the need was obvious, thousands of people around the world – including scientists, operators, policymakers, regulators, economists and government officials – worked differently to make sure vaccines became available faster.

It also reveals how slow the current system is. If clinical trials can be sped up so much without compromising on safety, we should find ways to reform them outside of pandemics as well. Every year of delays in testing is a year patients go without potentially lifesaving treatment, risking their lives. We shouldn’t be dragging our feet for the sake of following the status quo. Rather than treating pandemic trials as exceptional, we should ask whether that speed should have been the norm all along.

It’s been six years now since the publication of Ideas are getting harder to find, a famous paper in the economics of innovation which documented that vast increases in research effort are not resulting in materially faster rates of discovery (since then, a number of other papers point to similar conclusions). Why not? My preferred version is that as sciences mature, a variety of stresses emerge. A partial list of those stressors is that, as a discipline matures…

• Ideas that can be quickly or cheaply discovered are the first to be found

• Further progress requires marshalling more detailed knowledge, which requires more years of training to acquire, and older scientists have a harder time with some kinds of breakthrough innovation

• Marshalling more knowledge also increases reliance on research teams, and teams have a harder time developing breakthroughs than individuals

• The requirement to bring more knowledge to bear on outstanding problems, leads people to specialize more. But this impedes making connections across different fields, which is an important part of innovation (see Jordan’s latest blog post on this, in the context of AI that differentially accelerates different fields!)

The last point - that specialization impedes the diffusion of ideas - has seemed likely true to me, but I haven’t ever seen it measured, until this week. A new paper by Gaetani and Berkes tackles this measurement problem in medicine. To measure the spread of ideas, they compare the abstracts of papers that mention a medical phrase in the first year in which it was coined, and then in the subsequent ten years. They then use natural language processing to measure how different the abstracts are that first mention a concept and the ones that use it in later years, with the idea being that a greater difference between the abstracts implies the idea is being used in fields that are further afield from where an idea was developed.

Gaetani and Berkes’ paper suggest that specialization has made it harder for ideas to spread: they find new medical phrases increasingly come from scientists who have worked in fewer fields, and that phrases originating from more specialized scientists don’t diffuse as far. But they also document a dynamic that compounds this issue: over time, scientists have been introducing new ideas in less and less accessible language.

Our living literature review program was motivated in part by precisely these concerns - accessible literature syntheses could help ideas diffuse more quickly over time.

Here’s what caught our attention over the last week:

1. The ROAD Act and the Transit Abundance Playbook — Alex Armlovich

2. A new threat to economic data — Dylan Matthews

3. Refine sets the benchmark for AI review — Jordan Dworkin

4. Remember data centers in space? — Willow Latham-Proenca

5. Operation Trial Blazer — Saloni Dattani

6. The limit of human felicity — Matt Clancy

We’re also planning to post a separate What we’re reading spotlight by Matt tomorrow. Stay tuned!

The ROAD Act and the Transit Abundance Playbook — Alex Armlovich

The Senate passed the 21st Century ROAD to Housing Act 85–5 on Monday; the House takes up the amended package today, and President Trump is expected to sign it within the week. The bill’s 45-plus provisions cover a wide array of reforms: cutting red tape, overhauling funding and financing, and providing technical assistance plus carrots and sticks for state and local reform. The supply-oriented measures trim federal permitting friction, expand the definition of manufactured housing by relieving the HUD code chassis mandate, raise and index FHA multifamily loan limits, lift the bank public-welfare-investment cap from 15% to 20%, and stand up grant programs to convert vacant buildings and rehab aging homes. Most of the binding constraints on housing still live at the state and local level, but this provides direct cleanup of key paperwork burdens and bad rules that exist at the federal level.

IFP just launched its Transit Abundance Playbook: fifteen memos from transit practitioners and researchers on why the US pays the world’s highest prices to build and run transit (and how to bring those costs down). Rohan Aras and I wrote on bus procurement: US agencies pay anywhere from $500,000 to over $1 million per bus, while peers abroad buy comparable vehicles for under $400,000. Bad incentives in federal-local cost-sharing compound with excessive customization at the agency level…plus burdensome procurement regulations, bans on trade in buses even with US allies, and weak federal state capacity, to produce the world’s least competitive & most expensive domestic bus market.

A New Threat to Economic Data — Dylan Matthews

Economists, sociologists, demographers, and other researchers need accurate, ideally individual-level data to understand what’s going on in the world. At the same time, most people don’t want information like their address or income available for everyone to see. How do we reconcile these two goals? Some countries, like Norway and Sweden, err on the side of openness, making even individual income tax records public. The US typically values privacy highly, which means the Census has begun to use “noise infusion,” a technique for anonymizing individual and business-level data by adding random changes to individual data points. At a high level, the random noise cancels out, in theory letting you still learn accurate facts about, say, a county or industry, without violating specific people or businesses’ privacy.

When the Census announced they would do this to the 2020 Census data, leading to strange conclusions like finding that 48 people live on the same island as the Statue of Liberty (in reality zero people do, at least legally), there was a major backlash and the agency eventually backed down. Now, the Trump Commerce Department is proposing to ban any use of noise infusion. Nathan Goldschlag, a former staff economist at the Census Bureau, has a post at the Economic Innovation Group’s blog arguing this goes much too far and risks leading the Bureau to just not release certain data at all. I am far from an expert on this and honestly think the US tends to overweight privacy in these matters, but it’s the kind of debate that could have far-reaching ramifications for how we understand the economy.

Refine sets the benchmark for AI review — Jordan Dworkin

There are a growing number of AI tools for various components of the scientific process, but there are rarely careful, public evaluations of how they perform and compare. Last week Refine, an AI-powered review tool for economics research, published the kind of benchmark and evaluation study that I would love to see more of in this space. They ran ~1300 head-to-head matches on 150 economics preprints, comparing the quality of Refine’s reviews to both single-shot frontier LLM reviews and scaffolded review systems. Refine won roughly 90% of the match-ups (~95% against single-shot models and ~85% against scaffolds). Worth caveating that the benchmark’s unit of evaluation is a “paper-grounded atomic concern”, a specific verifiable issue tied to a location in the paper; this is a reasonable operationalization of review quality, but is also one that informed Refine’s design. Refine’s win rate is also strongest in matchups where it raises more unique concerns than the competitor (~90-95% when it has more concerns vs ~70% when it has fewer) - that could very well be signal, but it could also be an evaluator bias towards longer/more detailed reviews (a preference that has been observed in human evaluators), which would be both useful and interesting to check. But overall, it seems clear that one major takeaway is “Refine is very good at producing high-quality, technical feedback, and if you’re an economist you should consider giving it a try”, and another is “more tool-builders should run and publish studies like this.”

Remember data centers in space? — Willow Latham-Proenca

(...probably). Semianalysis is out with a comprehensive comparative analysis of the costs and uncertainties around space-based datacenters, arguing that while technical, operational, and cost barriers mean that space-based data centers start out at at least 4x earth-based costs, we shouldn’t completely discount the idea. They’re not exactly bullish on timing, though – in their base case, space doesn't reach cost parity with earth until ~2040, driven by falling launch and hardware costs. Even then, space would be a genuine necessity only if earth-based data centers have exhausted all potential capacity from grid supply, BTM generation, converted bitcoin mining, and industrial capacity/manpower to build additional infrastructure, which their base case doesn’t show happening at a meaningful scale through the middle of the century. However, what’s interesting is how quickly regulatory barriers (and capacity barriers dressed up as regulatory barriers) could potentially start to bite – by the early 2030s, space-based data centers could be running at only a ~30% premium, easily in the range where permitting and regulatory barriers become the deciding factor on whether to launch if earth-based capacity remains artificially constrained. Depending on how fast we solve some of the technical hurdles that currently make space-based centers a financial non-starter – as well as the universal constraint on chip deployment anywhere, semiconductor production – political will could end up being of central importance, making the most important argument for space…well, the space (and even that is limited by narrow orbital regimes!).

Operation Trial Blazer — Saloni Dattani

This week, the US Department of Health and Human Services (HHS) announced a new initiative called ‘Operation Trial1 Blazer’ that aims to bring back early-stage clinical research, which has increasingly been outsourced to countries like China and Australia, by reducing how long it takes to get approval to start phase 1 trials as well as how long it takes the FDA to review data from phase 1 trials.

The initiative includes several parts. One is an FDA pilot program to set up a network of qualified research institutions that help researchers prepare the protocol and supporting components before they submit it, though the FDA still decides whether a trial can proceed. It also provides more guidance on what manufacturing (CMC) data is needed for phase 1 and clarifies this is lower than for later-stage trials – companies have so far tended to submit more data than necessary to secure approval. And it includes guidance on choosing the first doses given to humans, recommending these are based on statistical modelling of the drug’s pharmacology rather than animal toxicology studies, which were often poor predictors of side effects in humans. Finally, there’s a rolling platform for the FDA to review components of the package as they arrive rather than all at the end.

Most of this sounds positive, though I’d add a caveat of the main bottleneck facing the FDA right now: staffing capacity, given the cuts over the last few years (roughly a fifth of the agency’s workforce, around 3,500 FDA staff, were cut last year), and whether these reforms ease that problem or worsen it further. Rolling reviews especially require more staffing time, because reviewing at the end typically means some attrition from researchers who don’t make it to that stage; similarly, if the FDA still has to make the final decision on whether phase 1 trials can proceed, especially given this initiative is new and will take their input. Still, there’s good stuff here: the guidance on manufacturing data and dose selection doesn’t depend on FDA capacity the way the pilot does, so some of these changes could result in improvements soon.

For more thoughts on the new policy, Ruxandra Teslo and Adam Kroetsch also weighed in.

The limit of human felicity — Matt Clancy

I’ve been listening to the wonderful new podcast Everyday Abundance (no affiliation!) by Virginia Postrel and Charles Mann, based on Dylan’s recommendation two weeks ago. Each episode looks at the history of some aspect of modern life, which we tend to take as just part of the background now. The last episode I listened to was about listening to music, and featured this quote from Edward Bellamy, writing in 1888:

[I]f we could have devised an arrangement for providing everybody with music in their homes, perfect in quality, unlimited in quantity, suited to every mood, and beginning and ceasing at will, we should have considered the limit of human felicity already attained.

I estimate I spent about 16% of my waking hours in 2024 listening to music.

We have just one announcement to share this week: our colleagues at Coefficient Giving launched a $10-30 million RFP on global health and wellbeing in an era of transformative AI, looking for proposals on everything from LMIC clinical trial infrastructure to how labor market disruption could reshape health systems. Applications are open through August 21.

1. The driving around test (Euro edition) — Matt Clancy

2. You should read Dario Amodei's latest essay — Dylan Matthews

3. It's 2017 again at the NIH — Jordan Dworkin

4. 40+ ways to plug in a data center — Willow Latham-Proenca

5. Should European housing politics be Americanized? — Alex Armlovich

6. Oops, I ordered the wrong antibody again — Saloni Dattani

7. Updates from Congress — Nisha Austin

The driving around test (Euro edition) — Matt Clancy

As part of the month-long debate about the extent of the European/American gap in living standards, Luis and Pieter Garicano suggested a “driving around test”: “Go to the periphery of any modern American city and see a level of new-built material wealth that is extremely uncommon in Europe.” With the World Cup underway in America, some European tourists are doing their own “driving around” tests and going viral, though these driving around tests are hardly scientific (the US may be wealthy, but free hotel rooms and meet-and-greets with our favorite singers do not come standard with road trips).

Nonetheless, for those worrying about the US-EU living standards gap, there’s plenty to read from the last few weeks. To start, Simon Johnson’s interview with economist Philippe Aghion covered a wide range of plausible factors explaining the gap: a lack of a single market, insufficient funding for long-run research, lack of DARPA-like agencies, a culture that doesn’t celebrate risk-taking, and excessive red tape. For more on this theme, see also Alex’s post below about European YIMBYs (or the lack thereof). While Aghion believes these factors are real, he pushes back when Johnson characterizes Europe as troubled:

I wouldn’t say “troubled.” Europe is not on the verge of a major crisis. Instead, the danger is a long period of stagnation or sluggish growth.

Unfortunately, the authors of Europe 2031 (the latest AI scenario microsite) worry that we are in an era when stagnation will impose heavy costs on Europe in the years ahead. One theme of the US-EU living standards debate has been on the importance of having a thriving domestic tech sector. So far, many of the benefits of the US tech sector have been shared by Europe, which, after all, has as much free access to Google and Facebook as it wants. The Europe 2031 authors paint a scenario where this assumption breaks down, and access to AI becomes nationalized as its security implications grow and compute for inference remains scarce. The microsite’s timing is eerie, coming shortly before the US government banned access to Anthropic’s latest model for foreign nationals (though, at least in this case, the ban meant that US citizens also lost access).

You should read Dario Amodei’s latest essay — Dylan Matthews

If you’re staying on top of AI news, you probably don’t need me to tell you to read Anthropic CEO Dario Amodei’s latest essay. That said … you should read Dario Amodei’s latest essay. It’s the most detailed look at the company’s public policy attitudes to date, and particularly interesting to the AGF team because of how much Amodei focuses on clinical trials. As he writes, AI seems likely to "greatly increase the rate at which new drug candidates enter the regulatory pipeline"; but that pipeline is so long and slow that it will be hard for customers to enjoy those drug candidates in a timely way. That implies that we need major reforms to the way the Food and Drug Administration and European Medicines Agency evaluate drug candidates, to make sure new, promising therapies aren't languishing for years before approval.

It’s 2017 again at the NIH — Jordan Dworkin

It’s 2017 again: the Gorillaz are topping charts, Star Wars is in theaters, and the NIH is discussing a potential cap on the number of grants a single PI can hold at once. Last week the agency published an RFI floating potential caps at two, three, or four simultaneous research project grants, with the stated goals of spreading funding more widely across labs, institutions, and researchers, and supporting more early- and mid-career scientists.

The agency’s last attempt at this – the 2017 Grant Support Index, which would’ve capped support at roughly three R01s’ worth of funding – was dropped within a month following pushback from elite universities. Last week’s RFI cites a handful of metascience papers in support (specifically showing diminishing marginal returns to additional funding, and reduced disruption as a function of team size), though in my view the metascientific jury is still out on the cost-benefit of a hard cap; getting more money to younger scientists seems like the strongest case for the cap, but cases against include diverting funding from meritorious proposals, increasing process burden, and making large-team science more difficult to sustain. Since 2017, the agency has tried a few other approaches for achieving these goals, including a dedicated fund for early- and mid-career scientists (which directly replaced the GSI when it was dropped) and a softer version of a cap at NIGMS, in which explicit justification is required for funding above a certain threshold. STAT has a good overview of the background and the various perspectives. Responses to the RFI are due August 3.

Should European housing politics be Americanized? — Alex Armlovich

"Should European housing politics be Americanized?" asks Samuel Hughes. American policy frames usually map badly onto European debates, but housing is an exception: Europe's housing shortages are arguably worse than America's, with European house prices now above US levels and roughly 80% of the postwar increase attributable to regulatory restrictions on building. Despite that, YIMBYism is effectively nonexistent on the continent, and housing arguments often default to rent control, zero-sum expropriation without new construction, and public housing. Hughes's sharpest move is historical: zoning was arguably invented in Europe, and he walks through Berlin's 1905 villa-colony scheme to show the same map performing the same exclusionary function 120 years later. I went in a little nervous about how a British writer would explain American YIMBYism abroad, but Sam’s argument was novel and deftly executed.

40+ ways to plug in a data center — Willow Latham-Proenca

It may not be the most scintillating reading for the general audience, but the Lawrence Berkeley National Lab is out with an extremely cogent, clear rundown of over 40 proposals to speed up how fast large loads – data centers, of course, but also industrial users or fleet EV charging – can connect to power. These range from the frequently-discussed (BYOG, interruptible service) to the arcane (ERIS-like service for loads, one of several proposals to adapt tools from the generation interconnection toolkit to the load side). As usual, the national lab that brought you quantitative evidence that new load doesn’t (necessarily) increase prices is calibrated on how much to panic – they compare load forecasting errors from the 1970s over-forecast to now, and find we might actually be doing better. Worth noting that proposals to speed up construction of the transmission needed to connect many of these loads to electrons – which can take up to three times as long as the planning and study process, in the report’s estimation – are out of scope here.

Oops, I ordered the wrong antibody again — Saloni Dattani

The science sleuth Sholto David recently uncovered a mix up that’s affected hundreds of papers in biology. Scientists who were intending to work with the p16-INK4a antibody, to flag a tumor suppressor linked to cancer and aging, accidentally ordered another antibody p16-ARC instead, which tags a component of the cell’s internal skeleton.

Some of the consequences were mostly harmless, where scientists ordered the correct antibody but described the wrong one in their papers. Others were unfortunate, as lots of experiments may not have worked because of the mistaken antibodies. But some are potentially fraudulent, where in some experiments, scientists reported results using the wrong antibody that could have only worked if they were using the right ones.

The whole naming confusion reminds me of another related episode in biology, where hundreds of papers were affected by an error in which the gene SEPT1 got automatically converted into the date ‘1st September’ by Microsoft Excel. Eventually, the Gene Nomenclature Committee solved the problem by simply renaming the gene SEPTIN1. Let this be a warning about the dangers of dating.

Updates from Congress — Nisha Austin

The first major federal housing bill in roughly 30 years just took a big step forward. After months of ping-ponging between the House and Senate, all four corners (the chairs and ranking members of the Senate Banking and House Financial Services committees) released unified bill text yesterday for the 21st Century ROAD to Housing Act. Senate leadership is expected to bring it to a floor vote this week. We’ll be watching.

On energy permitting, there are two bipartisan moves worth keeping an eye on. Senators Cotton and Cortez Masto introduced a Senate version of the FREEDOM Act, which would prevent agencies from revoking permits or halting construction on fully permitted energy projects, and is a direct response to the administration’s stop-work orders on offshore wind. The Senate version rolls in much of the House’s geothermal package from earlier this month and is notably cleaner than the House’s: it strips the most stringent agency penalties and moves fees into a freestanding “Permitting Performance Fund,” which should quiet concerns that the original structure would create a cost spiral for agencies. Meanwhile, Senator King added an amendment to the NDAA targeting the DOD’s increasingly creative use of flight path certification to block onshore wind projects. Neither energy bill is a sure thing, but both signal that permitting certainty has bipartisan legs even in this environment.

Here are a few other highlights and announcements from our team and grantees:

• Jordan published Our grandchildren’s AI-science bottleneck, on why the fields AI accelerates least may end up determining the pace of discovery for everything else.

• Matt published a blog post about the Atlas of Innovation, an interactive guide to choosing between prizes, grants, contracts, and other tools for funding innovation, developed with IFP and the Market Shaping Accelerator.

• Saloni wrote up What’s new in biology for Works in Progress, rounding up recent developments in the field, and covering everything from antibody mix-ups to new cancer research.

• Abundance New York launched a voter guide ahead of the city’s elections.

• California YIMBY endorsed Xavier Becerra for governor.

This jagged frontier is familiar from broader conversations about AI’s impacts on the economy and society, and operates at multiple levels of abstraction (i.e. tasks, jobs, sectors). At the task level, uneven automation is commonly discussed as a key bottleneck for large-scale productivity gains: if some essential tasks remain hard to automate or improve, the returns to progress elsewhere are constrained.

But science faces a second, nested version of this problem. Discoveries are not only bundles of tasks, but also inputs to other discoveries. If AI accelerates some fields much more than others, then the cross-pollination and interdependence that drive scientific discovery may turn less-accelerated disciplines into long-run bottlenecks for accelerated ones.

The economic context

Several recent papers lay out the foundational economic argument for how uneven automatability can constrain overall productivity gains. Jones & Tonetti (2026) show that output is constrained by the weakest/scarcest tasks, so overall growth is dependent on progress within the weakest links of the system. Aghion, Jones, and Jones (2019) similarly model growth constraints as a function of tasks that are essential but difficult to improve.

Jones (2025) applies these ideas to R&D tasks and research outcomes, showing that AI’s impact on the rate of scientific progress depends on the share of tasks that AI can do, how good it is at them, and how readily progress on one task can replace progress on another. He finds, similarly to findings in occupational tasks, that the returns to intelligence are bounded by the share of tasks AI covers.

Jones also points to a higher-level bottleneck. Even if AI overcomes the task-level constraints within a given research process, the outputs of different processes still have to combine into economy-wide productivity, and that aggregation averages down again towards the areas advancing slowest. This is the more macro-level version of the jagged frontier. If growth depends on the interaction across industries and supply chains, then variability in how much AI accelerates each one means overall productivity gets bottlenecked by the least accelerated sectors.2

Weak link opacity in science

Within a market, the same complementarity that creates the bottleneck also creates a signal for relieving it. When one input becomes scarce or slow relative to the rest, its marginal value rises, its relative price rises, and its share of spending and (ideally) investment into its growth and efficiency rise in turn.3 Because task and supply chain bottlenecks tend to be relatively legible and sensitive to price signals, reallocation of labor and capital to the least accelerated subdomains can help alleviate the worst effects of weak links in some cases.

Task-level bottlenecks within science, much like their occupational analogues, will be fairly legible and amenable to strategic reallocation of resources within roles and labs. When AI is very good at some parts of a field’s workflow and worse at others, researchers will often be able to see the gap and shift attention and labor towards it.

But problems arise one level up. In the broader economy, higher-level bottlenecks arise due to linked cross-sector contributions to growth; science’s analogy to cross-sector aggregation is, roughly, cross-disciplinary recombination of discovery and invention.

We are all living through one example of this dynamic, with once-niche statistical learning algorithms from computer science starting to facilitate new methods and discoveries across science. Other examples include the mid-20th century development of radiocarbon dating in physical chemistry, which has since transformed archaeology and geology; and the discovery of Thermus aquaticus by microbiologist Thomas Brock in the 1960s, which allowed for the invention of the polymerase chain reaction (PCR) in 1983 by biochemist Kary Mullis.

If some domains of science are more responsive to AI-driven acceleration than others, the fact that discoveries and inventions depend on prerequisite innovations in other fields will result in important discovery pathways becoming constrained by less-accelerated disciplines.

But unlike the broader economy, science lacks clean, contemporaneous price signals for these higher-level bottlenecks. The welfare returns to research are often invisible, especially prospectively; we typically can’t see in real time what fields or discoveries will turn out to be load-bearing for others – the way we might map sector-level dependencies across a supply chain – leaving field-level bottlenecks relatively opaque.4 As a result, the natural response to a jagged frontier within science could deepen the macro-level bottleneck rather than ameliorate it.

Compounding the illegibility of scientific bottlenecks is the nature of scientific incentives. Prestige is based on priority, novelty, access to funding, and peer recognition of being on the frontier. If AI allows progress in certain fields to visibly accelerate, those fields may see heavy investment of scientific resources and labor in an attempt to capitalize on progress and pull forward innovation.

This will likely work in the near term, as we make rapid progress in areas that are newly ripe for discovery, with access to “prerequisite overhang” from earlier yet-untapped progress; but in the long term it could starve unaccelerated fields of necessary resources and worsen a future bottleneck.

Right now, within and across fields, resources and focus are starting to flow toward areas that seem most amenable to AI acceleration. This is an explicit policy goal of both the US and the UK,5 and is an area of large-scale and increasing private investment. While we’re still mapping the opportunity space, this enthusiasm and investment is incredibly valuable; but since basic science lacks the market signals to tell us where the equilibrium is, the possibility that policy and investment could deepen long-run bottlenecks is worth having in the back of our minds.

Modeling the bottleneck

When might a bottleneck arise, and how could it be mitigated with targeted policy and investment?

To make the logic a bit more concrete, we can consider a stylized model of scientific progress as a directed network of discoveries; I’ve implemented this model in an interactive companion site, where you can adjust the parameters to match your own intuition and read more about the underlying methods.

Within the network, each new discovery becomes available when a set of prerequisite discoveries has been made; prerequisites often come from nearby fields, but occasionally span disciplines. The extent of AI-driven acceleration varies across fields, and progress on available discoveries is a function of both acceleration and effort share. Under this framework, a few parameters shape the trajectory of discovery.

The first is prerequisite density. New discoveries and inventions depend on previous ones; how many prerequisites, on average, does the typical discovery depend on? Tech-tree style datasets suggest values in the low to mid single digits.6 Within this model, higher numbers would slow progress and worsen bottlenecks. For baseline illustration, we’ll suppose each discovery depends on two gating prerequisites.7

Most prerequisites will come from a related discipline, but some discoveries depend on ideas from outside of one’s own field. How many prerequisite links are cross-field? At a paper level, somewhere between 20% and 40% of references span disciplines; for inventions, perhaps 15-30%.8 Let’s suppose 15% of prerequisite links are cross-field.9

Finally, how many fields can be accelerated by AI, and by how much? We don’t yet have strong evidence for either of these numbers, but for now let’s assume 60% of fields are accelerated or accelerable; those fields will make progress 5x faster than baseline (i.e. conditional on prerequisites being discovered, discoveries will happen 5x faster with the same amount of effort, or at the same speed with 1/5th as much effort), and the other fields will progress at the same pace as baseline.

With these parameters, we can construct networks of latent discoveries and simulate the propagation of scientific discovery over time.10 By simulating the discovery process under different acceleration schemes, we can compare the trajectory of progress under this jagged ecosystem to one in which no fields were accelerated (gray, dashed), and one in which all fields were accelerated equally at the ecosystem average (gold, in this case 100% of fields accelerated 3.4x).

You can see the cost of uneven acceleration, such that a uniform average speed-up produces much faster progress than a heterogeneous one, since overall progress is bottlenecked on the weakest links. If you zoom into the fast and slow fields, you can see what’s happening. The accelerated fields shoot up early and then plateau, having run out of discoveries they can make without the prerequisites from the slower fields. From that point, they progress only as quickly as the slow fields do.

While we can’t necessarily expect market or social forces to naturally allocate effort in ways that will reclaim that penalty, the way that resources flow into this system can have a significant effect. Resources here mean more than just compute; even in AI-heavy fields, scientific discovery will continue to depend on a broad basket of inputs including labor, infrastructure, land, and money. Funders and policymakers will have to make decisions about how to allocate these resources, and those decisions will shape the pace and direction of progress across domains.

If we assume that the resources in a given field proportionally affect how fast progress is made, we can look at overall progress as a function of what share of resources is put into fields sped-up by AI vs fields that are not. If hype and excitement about short-term discovery prospects crowd in more resources (either/both labor and capital) to sped-up fields than their relative proportion of the ecosystem (say, 75% of the resources into that 60% of fields) you get a pattern in which innovation initially progresses faster than if resources were allocated proportionally, but then hits a wall and ends up slowing down overall progress in the long term.

In this case, fast fields end up speed-running their available discovery space more quickly than in the proportional case, but run into the same bottleneck waiting for slow fields, which in this case are even further behind than in the other case because fewer resources have been available.

If you reverse this, and invest disproportionately in the slower fields, the dynamic changes. Initial progress is slower, but fast fields are able to make steadier progress without hitting a wall, and the long-run discovery rate ends up higher than in either the proportional or AI-heavy investment cases.11

Softening the jagged frontier

This obviously is just a toy model.12 The binary fast/slow distinction doesn’t represent the many shades of acceleration we’re already observing, it’s possible that cross-field linkage estimates overstate the relevant dependence structure, and prerequisites are treated as more binding than is realistic. It’s also hard to know, without a formal social welfare construction, when and whether a faster initial discovery rate would be a worthwhile trade-off for less overall progress in the long-run.13

There are also exogenous reasons why a bottleneck like this might not arise. Heavy investment in helping accelerate slow fields rather than investing in their status quo work could increase the proportion of accelerable fields.14 And scientific labor may not reallocate quickly enough for slow fields to be starved in the first place.

But despite the uncertainties, I think there’s sufficient reason to be attentive to this dynamic. This potential bottleneck is one that public funders are well-positioned to design around, and a relatively clear argument for why a rise of private money flowing into AI-for-science won’t necessarily reduce the value of public funding for the rest of it.15 In fact, it plausibly raises that value. If private capital and scientific prestige chase the fields where AI makes progress most visible, public funding becomes critical for keeping the rest from becoming the weak links that slow down the ecosystem.

Funders and policymakers won’t necessarily be able to see in advance which fields are accelerable, which are not, and which will turn out to be pivotal. Fortunately the case here doesn’t require them to; it instead points towards the importance of investing in acceleration where possible, while retaining broad and deep support across science even when the short term returns seem to pale in comparison to frothy fields and the long term returns are uncertain.

In a jagged scientific ecosystem, slow, stubborn fields may not garner much attention, but that won’t stop them from determining the pace of progress that future generations of scientists can achieve.

• Ask researchers to propose research projects and fund the best ones

• Hold a competition, and award a prize to the best submission

• Hire a contract researcher to deliver the innovation you seek

• Form a new research organization to work on the problem

• Make an advance commitment to subsidize the purchase of the innovation

• Make loans to organizations that work on the problem

How do you decide which way to commit funding?

There’s no simple heuristic for this problem. Economists have written a lot of papers that evaluate the empirical and theoretical case for different policies, but these are usually considered in isolation, as in, “how well does peer review predict the eventual scientific impact of a grant program” or “does an advance commitment to subsidize purchase of a new product lead to faster invention and uptake?” When papers do compare policies, they cover only a subset of the options. Over time, a literature has built up that covers a huge range of considerations and tradeoffs, but there is no overarching map of this territory to help decision-makers decide which policy to use. Yet, decisions need to be made, and in the absence of a decision framework, policymakers will use other considerations: familiarity, novelty, or idiosyncratic taste.

Over the past few years, I have been part of a team of economists and policy wonks developing a guide to this decision problem. This team was mostly drawn from the Institute for Progress (IFP) and the Market Shaping Accelerator. The result, which launched last week, is the Atlas of Innovation, an interactive web tool we made for policymakers to help them make sense of funding mechanisms for innovation.

Our most important goal was to come up with a decision framework that pointed policymakers towards the right policy tool. But given the complexity of the problem and the literature that has grown up around it, that meant the framework would end up convoluted and complicated. For example, the image below is an early iteration of our decision framework. But we wanted something a policymaker could use without deep economic expertise.

IFP co-founder Caleb Watney’s vision for an answer to this dilemma was an online tool, where policymakers could answer a series of questions about their innovation-related challenge and get pointed to a recommended policy. The (potentially) complex decision tree could live behind the scenes where it wasn’t necessary for anyone to engage with it.

For example: suppose you want to spur progress on a pan-coronavirus vaccine, which would protect against Covid-19 variants, but also MERS, SARS, and other coronaviruses. If you take this problem to the Atlas of Innovation, the first question you’ll be asked is “how precisely defined is your goal?”, with two options: “narrowly defined” or “broadly defined”, with some explanatory text to help decide which is a better fit.

“Narrowly defined” is a better answer for a specific innovation target like a vaccine. If you can specify the goal you are trying to achieve, then that opens up a set of financing mechanisms that make payouts conditional on achieving some target, like advance market commitments, procurement contracts, or certain kinds of prizes. If not, you’ll be better served by financing mechanisms that defer goal setting to the research performers. Depending on the answer you give, the Atlas takes you to a new set of questions that help zero in on a good policy option.

After a series of questions, the Atlas recommends a specific policy. For example, for a pan-coronavirus vaccine, a plausible recommendation is an Advance Market Commitment. The landing page for each recommended policy explains the reasons we recommended this policy as a solution to the answers given. These landing pages also play a few other roles for us: they let us describe more of the evidence base related to a policy, to describe subvariants of the policy, point to other related policies (in case our recommendation didn’t quite hit the mark), and give examples and a bibliography.

You may have noticed the screenshots of the Atlas have a lot of imagery. To encourage people to actually use the thing, we wanted the experience of using the Atlas to be better than the experience of reading a White Paper. So, in addition to the metaphor of navigating through an archipelago, each landing page also has an illustration from R. Kikuo Johnson, which tries to capture the spirit of the policy tool in metaphor.

For example, R&D tax credits are relatively untargeted, and can incentivize lots of different kinds of firms to engage in lots of different kinds of R&D. That’s represented here as a wind that speeds lots of different kinds of sailboats. Meanwhile advance market commitment creates a market for a new product to serve, if it can be invented. That’s represented here as building a town, to entice the railroad to extend a line out to that area.

The result lives at AtlasOfInnovation.org. Check it out, or check out the accompanying paper which describes our decision framework (presented at the 2026 NBER Entrepreneurship and Innovation Policy and the Economy Conference and forthcoming in the accompanying volume).

But I hope this isn’t the end of the story.

To start, one thing I came away from this experience with was a sense of just how uneven the scholarship is across these different policy tools, at least, as applied to the problem of fostering innovation. One hope is that the Atlas can help illuminate these gaps, fostering new research. Second, while we put a lot of work into this decision framework, it is still only a first attempt. We hope now that there is one attempt out there, researchers and policymakers can critique it, study its implications, and help come up with something better. Third and finally, for this version of the Atlas, we opted to focus on ways the government (or other funders) can spend money to pursue an innovation target. But there are a range of other policy options governments have to accelerate innovation: immigration, education, intellectual property rights, regulation, standards, and so on. I hope there will be future versions of the Atlas that can draw on richer evidence, reflect improved understanding of the decision framework, and incorporate policies beyond funding mechanisms.

But for now, I’m happy that Version 1.0 is out there.

To close, I want to thank all the people involved in the project, and for letting me contribute to it. Special thanks to IFP’s Matthew Esche and Caleb Watney who led the overall project. Less special thanks to my collaborators Sarrin Chethik, Claire McMahon, Siddhartha Haria, Christopher Snyder, and Heidi Williams. Thanks to Beez Africa for creative direction, Emma Steinhobel for design, joodaloop for web development and design, and Ben Murphy for web technical direction. Thanks also to Eamonn Ives, Hewson Duffy, Joseph Fridman, and Santi Ruiz for all their work editing this into something that non-economists would want to read. And finally, thanks to the large number of policy and economics experts who gave us detailed feedback on this project (see the full list here).

Here’s what caught our attention over the last week:

1. Everyday Abundance — Dylan Matthews

2. Geothermal policy — Willow Latham-Proenca

3. The cost of building and the cost of insuring — Alex Armlovich

4. Clinical trial innovation barriers — Saloni Dattani

5. Data center power lines in Loudoun County — Nisha Austin

Jordan is whale watching in Alaska; he’ll be back next week. And we’re also planning to post a separate post by Matt on the Atlas of Innovation launch later this week.

Everyday Abundance — Dylan Matthews

I just finished reading Charles Mann’s The Wizard and the Prophet, which has to be a top-ten abundance and growth book for its description of the Green Revolution’s effect on crop yields and recounting of the failed apocalyptic predictions of “prophets” like William Vogt and Paul Ehrlich. So it was serendipitous that just as I wrapped up that Mann work, I heard about a new one: Everyday Abundance, a new podcast he and Virginia Postrel (another science journalist I love) are starting. Their first episode is, naturally, about the history of brushing your teeth. If anyone can make that interesting, it’s these two.

Geothermal policy — Willow Latham-Proenca

The House passed a bipartisan geothermal package (plus a freestanding bill) last week – big hits are statutory formalization of a NEPA categorical exclusion, expanding on the April 2026 categorical exclusion to cover development (not just exploration) of geothermal resources, and a faster leasing cadence (while provisions setting timelines for approvals are a move in the right direction, they’re probably less likely to be effective). Whether this passes the Senate, either on its own or riding on a larger permitting bill, remains an open question.

While it’s exciting to see policy movement on America’s most basically likeable energy source, it’s worth remembering that two of the most important binding constraints - better characterization of underground resources to make projects more bankable, and transmission to get electrons to where they’re needed - remain to be addressed. And Fervo’s March ‘26 comment on a California PUC transmission financing concept paper illustrates (part of) why transmission is an unpopular, if critical, cousin - allocating the costs remains fiendishly controversial. Nisha’s dispatch from Loudoun County this week shows what that opposition looks like on the ground.

The cost of building and the cost of insuring — Alex Armlovich

A new VoxEU column argues that declining construction productivity has fueled an 80% increase in the relative price of US structures since 1970, offsetting the benefits of continued declines in the cost of equipment. “Construction productivity is no longer a sector-specific concern”, they say, positing more expensive structures as a key drag on economy-wide growth. The authors don’t cleanly distinguish between land costs and construction cost inputs, so their model includes the combined effects of land use regulations and pure construction costs on the final user cost of real estate. While endorsing land use regulation reform, they point to reforms to building codes and investigating other construction cost inputs as the neglected next frontiers (a view widely shared among professional YIMBYs).

Meanwhile, insurance costs have also been a hot discourse topic, especially in housing and real estate. Some of this is pure money illusion: The average price level of all goods and services increased ~30% since 2020 because of the largest coordinated fiscal and monetary policy stimulus since the New Deal, so it’s easy to find a trend piece about almost any product up by 20% to 40% in nominal dollars. Still, Florida had a real problem with homeowners insurance, reportedly originating 76% of US insurance lawsuits from just 9% of US homeowners. After recent liability and tort reforms, insurance firms and Florida regulators are reporting a sharp drop in litigation costs and premium growth. Every insurance market is different, but there’s never a bad time to identify and reform rent-seeking that threatens housing affordability.

Clinical trial innovation barriers — Saloni Dattani

Have you ever wondered why some innovative approaches only get used once or twice? It’s something I’ve often thought about with clinical trials. Take platform trials, which test multiple treatments within a single trial, or adaptive trials, where new treatments can be added as they become available, or protocols can change in response to incoming data (in a predetermined way).

Designs like these are often piloted or used by academic researchers, and sometimes they even show up in FDA guidance, but they rarely get adopted widely by pharmaceutical companies. A new post by Adam Kroetsch explains why.

The recurring theme is uncertainty on what regulators will accept, and since trials are so expensive and slow, with a lot riding on whether a drug application is accepted or rejected, the stakes of failure are enormous. The first company that tries an innovative design carries huge risks. But even a single pilot trial usually isn’t enough to convince companies that a given protocol will be accepted more widely.

He lays out how to get these designs to scale: the FDA should formally recognize modern trial practices so sponsors know in advance what protocols will be accepted; the government should set interoperability standards so trial software can be harmonized; and the NIH should fund more large, phase 3-scale trials and build reusable infrastructure, de-risking innovative designs by making them ready for regulators, repeatable, and adoptable.

Data center power lines in Loudoun County — Nisha Austin

Last month, Dylan wrote about Loudoun County, Virginia, home of the largest concentration of data centers in the world, and where data centers account for 45 percent of county revenue, lower property tax rates, and well-funded schools. But there’s a flip side playing out over the transmission infrastructure needed to power them. Dominion Energy’s Golden-Mars line, the final piece of a reliability loop, was approved by Virginia’s State Corporation Commission in April, but residents have turned out by the hundreds to fight the route, which runs near schools and homes. The county school board asked for the lines to be buried; the commission said no. County Chair Phyllis Randall captured the shifting politics: residents used to accept data centers for the tax benefits, but “now I say, ‘this is what it does for your tax rate’ and they say, ‘I don’t care.’”

If there’s a local story like this you think we should know about, we’d love to hear from you.

Here are a few other highlights and announcements from our team and grantees:

• We provided prize funding for the Metascience Novelty Indicators Challenge with UKRI, a global competition to develop scalable methods for assessing the novelty of research papers (see coverage in Science here). The winning team from Forschungszentrum Jülich built an AI-based approach that compares the knowledge contribution of a paper to the state of knowledge, as represented by the text of the papers it cites.

• The Zoning Project launched a publicly available national dataset of housing regulations, built using LLMs to read and code zoning documents across U.S. municipalities. It makes it possible to systematically compare how rules like minimum lot sizes and density restrictions vary across the country.

• Lauren Gilbert published “Immigration and House Prices“ in her Migration Living Literature Review, reviewing the evidence on whether immigration increases housing costs.

• California YIMBY published their latest Homework newsletter, rounding up recent housing legislation and policy developments across the state.

There have been lots of big breakthroughs in cancer treatment recently, so I was excited to follow the news from the American Society of Clinical Oncology conference (ASCO 2026), which took place in Chicago this week1, and wanted to share some of my highlights.

The biggest was daraxonrasib, a new drug against pancreatic cancer. Pancreatic cancer is one of the deadliest, and it has been considered essentially untreatable for a long time, with 5-year survival rates of only ~13%. In a phase 3 trial, daraxonrasib roughly doubled median survival among people with metastatic pancreatic cancer (to about 13 months). This is probably only the beginning, since the drug targets a cancer protein (KRAS) that’s also common in cancers of other organs, including the lungs. Ruxandra Teslo wrote a great piece on how the breakthrough came about here.

Then there were 7 year results for lorlatinib, a precision drug for metastatic lung cancers: specifically, a type (ALK-positive) that’s common in non-smokers. You can see the results of the phase 3 trial below: 55% of patients were still progression-free after 7 years, versus just 3% on the older drug crizotinib. I have never seen a survival curve like it.

A new prostate cancer drug, talazoparib, cut the risk of progression by about half when added to hormone therapy, compared to hormone therapy alone, in a phase 3 trial for prostate cancers carrying specific DNA repair gene mutations. Talazoparib belongs to a class of drugs (PARP inhibitors) that have already advanced treatment of breast and ovarian cancers with BRCA mutations. This trial suggests the same strategy also works for prostate cancer.

There was also long-term data on a recent endometrial cancer drug, dostarlimab, an antibody that was approved in 2023. After 4 years on the phase 3 RUBY trial, 58% of patients whose tumours carried a ‘mismatch repair deficient’ signature still hadn’t progressed when the drug was added to chemotherapy, versus just 16% on chemotherapy alone. That’s a huge difference! The drug is a checkpoint inhibitor: T cells have built-in ‘checkpoints’ that hold them back from attacking cells unnecessarily, but tumours exploit these to avoid being attacked, so blocking the checkpoint frees the T cells to go after the tumour.

An early cancer detection blood test2, the NHS Galleri blood test, was tested in a large RCT in England with about 142,000 people. The results are a little confusing, so let me start by describing a ‘good result’ in this scenario: detecting a non-benign cancer that would’ve missed otherwise, or catching one earlier so treatment could start sooner.

On the first, it succeeded: roughly quadrupling the detection rate and cutting cancers caught only at emergency presentation by around a quarter. On the second, it detected some cancers earlier: with a 16% rise at stage I–II, and 14% reduction at stage IV. But the trial missed its primary goal: there was no significant drop in combined stage III and IV cancers, because stage III diagnoses rose more than expected and cancelled out the stage IV reduction. The way I’d interpret it all – since the largest jump in detection was at stage III and the combined late-stage count barely changed – is that the ‘early’ detection test was mostly catching cancers at stage III, rather than genuinely early. That may still be useful, of course.

And finally, an mRNA cancer vaccine that might actually work! In people with high-risk melanoma, who’ve already undergone surgery, it cut the risk of cancer recurrence or death by around half, compared to Keytruda (the blockbuster cancer immunotherapy drug) alone, in a phase 2b trial over 5 years.

As I’ve learnt from Ruxandra, many previous mRNA cancer vaccines have been doomed by how they pick their targets. They sequence a tumour’s mutations and design the vaccine against the resulting ‘neoantigens’ (newly formed mutant proteins in a cancer cell), but in many cancers, those neoantigens are never actually displayed on the cell surface, so immune cells can’t recognize them. Melanoma is different: its neoantigens are readily displayed and recognized, which is why drugs like Keytruda already work well against it, and why a vaccine has a much better chance of working as well.

We’re living through a very special time in cancer research.

Here’s what caught our attention this week:

1. Joe Lonsdale’s FDA reform agenda — Dylan Matthews

2. OMB’s proposed changes to federal science funding — Jordan Dworkin

3. Brian Potter on housing construction productivity — Matt Clancy

4. Energy scarcity and heavy industry — Willow Latham-Proenca

5. Massachusetts’ starter homes ballot measure and the single-stair code win — Alex Armlovich

6. State abundance agendas — Nisha Austin

We’re also planning to post a separate What we’re reading spotlight by Saloni later this week. Stay tuned!

Joe Lonsdale’s FDA reform agenda — Dylan Matthews

Joe Lonsdale’s recent Substack post, “What America Needs from a New FDA Commissioner,” is notable as much for who’s saying it as for what it says. Lonsdale is a Palantir cofounder turned venture capitalist who is closely connected with the Trump administration and tech right. That means his comments on regulatory policy have a decent shot at being heard by people who can act on them. The good news is that Lonsdale is pushing a lot of promising ideas to remove unnecessary red tape holding back clinical trials. That includes investigating and approving new “surrogate endpoints” that drug studies can use as outcomes to measure; adopting an expedited Australian-style approach to Phase 1 trials; and trying to eliminate wasted time between trial phases. These are paired with a focus on beating China and encouraging companies to test in the US. That framing wasn’t enough to get much done on these topics under former Commissioner Marty Makary, but maybe they will be under his successor.

OMB’s proposed changes to federal science funding — Jordan Dworkin

Last week the Office of Management and Budget (OMB) published a proposed revision to the Guidance for Federal Financial Assistance. It runs 412 pages, and carries significant implications for federally funded science in the US.

Many of the proposed changes are concerning: the guidance would be classified as a binding regulation, requiring agencies to implement it directly unless changes are approved by OMB;

• a new pre-award review would require political appointees to assess awards to ensure they “demonstrably advance the President’s policy priorities”;

• discretionary termination clauses would become mandatory in all discretionary awards (unless prohibited by statute), making it easier for agencies to terminate grants on policy or topical grounds;

• and sweeping restrictions would be introduced on foreign collaborations, prohibiting the use of federal funds across all agencies to support collaboration with “covered foreign countries” likely including China, Russia, and Iran (the related Wolf Amendment has been around since 2011 but only applied to NASA and OSTP, and specifically restricted certain collaboration with China).

One positive note is that the document explicitly declines to propose changes to the indirect costs system – a possibility that we previously flagged. And one to keep an eye on is the proposed prohibition on publication costs, like article processing charges (APCs) and open access fees, unless expressly approved by the agency on a case-by-case basis (this could be a big deal, but it’s possible that including an APC line within the itemized budget of an approved grant would be sufficient). Instructions for commenting can be found here, and comments can be submitted at this link until July 13.

Brian Potter on housing construction productivity — Matt Clancy

One factor in the high cost of housing is that we’re not better at building houses; construction productivity is relatively flat (not only in the US, but in many countries). It’s often presumed that this is because housing construction has mostly failed to move inside the factory (this piece by Arpit Gupta and Steve Teles, which we’ve covered before, gives some reasons why that might be). But in a new blog post, Brian Potter is skeptical of this explanation.

Brian reminds us that factory built housing is actually a thing, in the manufactured home industry, and it’s telling that this sector has often not found it worth centralizing production in large automated factories. He also suggests there might be a low ceiling on how much automation can reduce the cost of housing, by looking at the so-called Idiot Index: the cost of a good divided by the cost of its raw materials. In housing, Brian argues this is roughly 2, compared to 1.8 in the highly automated car manufacturing sector. Maybe the reason construction productivity is flat here and around the world, is that it’s already near a local maximum? And yet, and yet - other work finds manufactured homes cost substantially less to build than comparable site-built homes.

Energy scarcity and heavy industry — Willow Latham-Proenca

There’s been so much focus on energy for data centers that it’s possible to forget that they’re not the only new large loads. A short report this week highlights a potential challenge for heavy industry – access to increasingly scarce affordable generation resources.

Industrial users have historically paid below-residential rates as a result of essentially “buying in bulk,” since their high, predictable demand and direct grid connection made them easier to serve. As data centers increasingly offer (at least the spectre of) prices above residential rates, new large loads in metallurgical or manufacturing industries - including sectors targeted for strategic reshoring - are increasingly concerned they’ll be outbid. Aluminum could seem like a straw man - it’s famously energy-intensive enough to earn the nickname of “congealed electricity,” but so are many other upstream commodities in the electro-tech stack, argues Seaver Wang of the Breakthrough Institute. Given the long timelines from investment to production - an aluminum facility takes about 5 years to build and bring online - it’s plausible that uncertainty now could cast a long shadow over what capital investments firms are willing to risk.

Massachusetts’ starter homes ballot measure and the single-stair code win — Alex Armlovich

The NYT editorial board endorsed the Massachusetts “Legalize Starter Homes” ballot measure, which would require localities to allow homes on small lots in areas served by public water and sewer. The endorsement lands alongside a volunteer-driven signature push to secure the November ballot slot. That combination bodes well for MA YIMBYs: contested ballot measures are expensive, since building name and message recognition with paid media costs real money. Ballot measures can also be risky in that a win can convince a legislature that they’ve mentally overindexed on NIMBYs in the electorate, while a loss could make reform radioactive. Fortunately, no organized opposition has yet appeared, and organic volunteer energy and platinum earned media like a Times endorsement are both free in dollars…and probably signal strong underlying voter support.

Stephen Smith just landed a stunning, albeit incremental, single-stair win: his proposal to raise America’s “International” Building Code single-stair apartment height limit from three stories to four passed its final membership vote and will officially make the 2027 IBC. The political dynamics here are understudied: this is "private politics," the work of lobbying and advocating inside the International Code Council, a private, nongovernmental body that drafts the model building codes later adopted into state and local law. The simplest early lesson might be the old Gretzky adage about missing 100% of the shots you don't take: YIMBYs walked into the room at ICC for the first time ever, and good things happened. (The burst of state and local-level single stair code wins across the US and Canada since 2021 were key points of leverage as well).

State abundance agendas — Nisha Austin

One thing we’ve been tracking is the proliferation of state and local abundance agendas. In the last year, we’ve seen Abundance New York publish a detailed policy roadmap across housing, transit, energy, and governance; Governor Newsom sign a package of housing and permitting reforms under the banner of the “California Abundance Agenda”; Abundant Housing Massachusetts put together a legislative agenda for the current session; Utah’s Office of Energy Development launch Operation Gigawatt, a statewide plan to address growing energy demand; and a recent ChinaTalk episode covered what an abundance playbook looks like for semiconductor reshoring. Finally, Inclusive Abundance is accepting proposals for their federal abundance agenda through June 18.

These vary a lot in scope, specificity, and political orientation, and we’re not endorsing every item in any of them, but the pattern itself is notable: “abundance” is becoming a shared vocabulary across very different political contexts, and concrete policy agendas are starting to follow. Are there other state or local abundance agendas in development that we should know about? We’re thinking about keeping a running directory, and we’d love to hear what we’re missing!

A few grantee and team shoutouts and other announcements worth sharing:

• Matt worked with IFP and the Market Shaping Accelerator to launch the Atlas of Innovation this week, an interactive guide to innovation funding mechanisms for policymakers and philanthropies.

• Alex joined the Gotham Gazette podcast to discuss Mayor Mamdani’s new affordable housing plan.

• Dylan published How do states share ideas? this week, a deep dive into the political science of policy diffusion and what it means for spreading growth-promoting reforms across states.

• The FDA recently began releasing its drug rejection letters, but the law firm Covington & Burling has petitioned to stop it, arguing the letters are too useful to competitors to not be considered confidential. Adam Kroetsch explains in a new blogpost that this gets it backwards.

• Alex and Reed Schwartz published a piece at the Niskanen Center on New York State’s newly authorized land value tax for transit investment.

• New York State’s newly enacted FY 2027 budget includes major new mass transit projects for New York City, among them the Interborough Express (IBX), which would be the city’s first new end-to-end rapid transit line in almost a century.

• The near-term federal transit funding outlook is rocky, but the budget includes a powerful funding and financing tool that will allow the state transit authority to “capture” the added value that the IBX and future projects will create.

• The project will need a market-grade underwriting, but we find that the lower-bound estimate of funding available for the IBX through this state tool is in the billions of dollars.

In her 2026 State of the State address in January, New York Gov. Kathy Hochul outlined an ambitious transit infrastructure agenda. Not only would the Metropolitan Transportation Authority (MTA) move ahead with the Interborough Express (IBX), the city’s first new end-to-end rapid transit line since 1937, but, Hochul said, it would also extend the Second Avenue Subway to 125th Street in Upper Manhattan as long envisioned.

But big ambitions come with big price tags. The next phase of the Second Avenue Subway extension is projected to cost over $7 billion; the IBX is projected to cost $5.5 billion. While the MTA is looking into potential cost savings, the question of how New York will foot the bill remains. Federal transit funding is uncertain given the Trump administration’s hostility to transit investments, and even in less politically volatile times, the procedural requirements for federal grants can delay projects and inflate costs.

New York State’s long-overdue FY 2027 state budget, adopted late Tuesday, includes a powerful alternative to fund those projects: land value capture. Widely used in Japan, Hong Kong, and elsewhere and known as the “Rail plus Property model,” land value capture encompasses policy tools that connect the private real estate value created by local public investments with the funding needed for those investments to trigger a virtuous cycle of cost-effective growth.

When the MTA expands the transit network, residents near new stations can access more jobs and services in a fixed amount of time, and businesses can access a larger labor pool and customer base. These benefits substantially increase the value of nearby properties. Land value capture allows the public to recoup a share of that windfall — on the principle that public investments shouldn’t exclusively enrich private landowners — and reinvest it for public benefit.

New York City and the MTA should recognize this principle and use land value capture to fund the IBX, finally resuming the continuous expansion of the subway system that stalled before World War II.

Value capture under Section 119-R

New York City’s ability to recapture the value of its transit investments is embedded in Section 119-R of the New York General Municipal Laws, enacted in 2016 and recently extended to April 1st, 2027.

The provision enables value capture for transit investments, both directly and indirectly, through three mechanisms. The first two are new taxes that a city can impose on property owners in conjunction with transit projects: special transit assessments and land value taxes. A special assessment is a tax on nearby properties, both the parcels of land and their improvements. A land value tax is a more targeted levy on land alone, rather than on the property as a whole.

The third mechanism is known as tax increment financing (TIF), in which public agencies such as the MTA can borrow against the property taxes that their projects will generate. Unlike the other two provisions, TIFs are financing mechanisms and do not entail additional value capture.

The land value taxes that this provision enables may be the most important. The value of underlying land, not physical structures, increases in response to nearby improvements, so this tax would fall more directly on the value generated by the state’s investments — without penalizing construction.

This matters because it’s been more than a decade since New York last used any kind of value capture to fund a subway extension. That project entailed the sale, in 2015, of air rights and a TIF-like redirection of payments in lieu of property taxes to help fund the extension of the MTA’s No. 7 train from Times Square to the redeveloped Hudson Yards on Manhattan’s West Side waterfront. In that alternative model of value capture, a fee is collected if a developer builds, but not if they leave the land undeveloped. In effect, it taxes the decision to build.

But the Hudson Yards approach is no longer viable for the IBX: New York has since layered on new programs that already capture much of the value from upzoning. Between inclusionary zoning, prevailing wage rules under Section 485-x of New York State Real Property Tax Law, and other exactions, there is little left for the city’s Planning Department to claim through District Improvement Bonuses (the fees developers pay for additional floor area) near proposed IBX stations. Besides, the incremental property taxes from upzoning are already exempted to offset the costs of inclusionary zoning and prevailing wage mandates: property taxes that have already been forgiven can’t fund transit.

A land value tax avoids this problem by capturing the added value — the value uplift — in both the build and no-build scenarios. It will capture the land value created by transit access without relying on the meager leftover value unlocked by residential upzoning alone.

The Second Avenue Subway precedent

At a cost of $4.5 billion, the 2017 Second Avenue Subway extension was the most expensive two miles of transit ever built. The project involved an extraordinary amount of waste and delay that must be shaved off future projects. Yet even with its bloat, it was well worth the cost. In their 2022 paper “Take the Q Train,” economists Arpit Gupta, Stijn Van Nieuwerburgh, and Constantine Kontokosta found that the extension boosted property values in the Second Avenue corridor by 8 percent, generating $5.5 billion in uplift — enough to pay for the entire line with over $1 billion left over. Despite the excess expense, a centrally located subway line in the densest neighborhood in North America still generated enough social value to pay for itself.

But existing property taxes only capture some of that windfall — the authors estimate that only ~30 percent of the private uplift value will ultimately flow back to the city’s general fund. The remaining uplift generated by public investment will stay in private hands. The more substantial levy that Section 119-R authorizes would ensure that the landowners who stand to benefit the most from public investments pay a commensurate share.

Value capture along the IBX

The value generated by the first phase of the Second Avenue Subway is not obviously replicable with other projects, given the exceptional land values of Manhattan’s Upper East Side. But if the IBX extension generates even a fraction of the $5.5 billion value that the Second Avenue Subway created, the case for value capture is strong.

To estimate the IBX’s potential, we took the value uplift percentages that Gupta, Nieuwerburgh, and Kontokosta found, and then applied them to notably undervalued city property assessments near the planned IBX route. Using a 0.3-mile (as the crow flies) buffer around proposed stations and the 5.6 percent uplift rate Gupta et al. found for properties nearest the new Second Avenue Subway stations yields approximately $1.5 billion in added value. Expanding to the properties accessible on a .5-mile (approximately 15-minute) walk raises that to $2.1 billion. At the higher 8 percent rate that the authors found along the Second Avenue corridor itself, these projections rise to $2.3 billion and $3.3 billion, respectively.

And there are sound reasons to believe that, on a percentage basis, IBX could exceed the returns to the Second Avenue project. This potential depends on two competing considerations.

On one hand, the Second Avenue Subway serves some of the densest, most valuable real estate on the planet. Paring even small amounts of time off commutes in the area could create outsized gains for the millions of residents who live in and commute to the area. On the other hand, the Upper East Side was generally well served before the expansion; though the expansion was highly productive, it was a marginal improvement for an area that already had high transit access. In contrast, many of the IBX stations are sited in transit deserts. Providing residents of the outer reaches of Brooklyn and Queens with access to jobs in Manhattan could be even more transformative, even if the expansion is located on the periphery rather than in the city’s center.

In addition, zoning restrictions constrain land values in Brooklyn and Queens. If the IBX expansion is paired with transit-oriented upzoning to lift those constraints, new transit access would unlock more housing capacity, higher land values, and greater fiscal returns. Even though the maximum allowed density in the IBX corridor is far below that on the Upper East Side of Manhattan, the corridor still has more unused existing “zoned capacity” than the Manhattan areas that the Second Avenue Subway extension served at the time it was completed. Even if much of the incremental upzoning value is captured by inclusionary zoning and 485-x wage rules, the IBX’s potential to catalyze development could result in both short-term value uplift and long-term revenue expansion.

As calculated, the IBX’s uplift may not offset its full $5.5 billion price tag. But it could substantially offset capital costs and, if paired with upzoning, generate a steady, growing return on public investment, funding further infrastructure expansion at a pace New York hasn’t seen in decades.

Putting transit funding on solid ground

Beyond the IBX and Second Avenue extensions, establishing value capture as a permanent funding feature would significantly improve current mechanisms for transit planning and development.

First, the MTA’s use of broader-based taxes does not distinguish between those who benefit most from capital improvements and those who do not; value capture would allocate the funding burden directly onto the greatest beneficiaries.

Second, it enables local autonomy over capital decisions. The total federal funding pool for transit is limited; even when New York wins grants, the delays caused by application and compliance work can add years (and huge costs) to projects. Even in more stable environments, autonomy for local decision-makers is extremely valuable. In a previous value capture success, the Hudson Yards tax increment financing enabled then-Mayor Michael Bloomberg to extend the No. 7 train without federal help.

Third, using 119-R value capture to support new projects would create a virtuous cycle of investment and recaptured returns. By building institutional knowledge in the relevant agencies, a development pipeline supported by value capture might even mitigate many of the cost issues that plagued inexperienced planners during the Second Avenue Subway extension.

Conclusion

By enacting the 119-R extension, the governor and the legislature have equipped New York City with the ability to capture the value that investment in public transportation creates. As a project both sufficiently transformative to deliver the required value and large enough to demonstrate how value capture can work at scale, the IBX is the right place to start.

I live in Washington DC, and here there’s a natural tendency to obsess over the federal government. But for a lot of issues that the Abundance and Growth Fund cares about — housing supply, renewable energy buildout, transportation, state capacity, commercial deployment of scientific innovation — state-level policies can be as, if not more, central. In housing, states can overrule intransigent NIMBY localities to allow more building. On energy, they can roll back regulations that delay or doom things like solar farms and electrical transmission lines.

Moreover, many policies seem to “bubble up” from the states until they affect a huge swathe of the country, rather than being imposed top down by the federal government. Take housing, for instance. As our CEO Alexander Berger has noted, the YIMBY movement was trivially small at its onset around 2016, when we began supporting state-level YIMBY work in California. But the success of the movement there has led to growing movements in other states, and now a dozen-odd states have passed policies to encourage more housing production:

In political science, this is known as “policy diffusion,” and between-state diffusion of policy has been a major topic of study since a landmark article by Jack Walker in 1969. I’ve been starting to dig into this literature a bit in search of ideas for how to spark diffusion of good, growth-promoting policies. Penn State professor Daniel Mallinson’s excellent meta-analysis of the area has been a great help in getting my bearings, and the SPID (State Policy Innovation and Diffusion) database1, covering hundreds of specific state-level policies, has enabled a spree of really excellent work just in the past decade.

I can’t do the whole field justice in a short blog post, but I’ll highlight a few big takeaways I’ve gotten so far.

Partisanship is starting to beat proximity

Much of the diffusion literature explores to what extent policy decisions at the state level are driven by internal factors (how rich the state is, how liberal it is, etc.) versus influence from neighboring states. Obviously both factors matter, and at least since a 1990 paper from Frances Stokes Berry and William Berry, political scientists have been building models where both play a role. Those papers have tended to only focus on one particular policy area, though (in the Berrys’ case, state lotteries). SPID has changed that by allowing researchers to track diffusion patterns across many different policy domains over time.

Economists Stefano DellaVigna at Berkeley and Woojin Kim at Stanford recently published a very helpful paper tracking how the roles of partisan and geographic proximity have evolved from the 1950s to present. They conclude that from the ‘50s through the ‘90s, the most important factor was geographic proximity. If you want to know what will happen next in Iowa, check out what’s happening in Minnesota or Missouri. But by the 2000s and 2010s, the more important factor was state partisanship. Being nearby didn’t lose any of its importance; its coefficient in the main regression is basically unchanged. Partisanship just became much, much more important at the same time.

Other researchers have found the same thing. Also using SPID, Penn State’s Mallinson has found that partisanship became a more important factor from 1960 to 2014. Unlike DellaVigna and Kim, he finds that the role of geography has actually decreased, rather than staying constant over the decades. But the results are aligned in finding partisanship’s importance growing.

This pair of maps from DellaVigna and Kim illustrates the overall dynamic well:

Here, DellaVigna and Kim are asking us to suppose that California introduces a new policy that no other state has (they find that California was the most “innovative” state in recent decades, in terms of introducing novel policies). The first map shows how likely different states would be to adopt the policy in turn according to the dynamics that prevailed in the 1990s: darker means more likely to copy California, and lighter means less likely. The second map does the same simulation but using dynamics from the 2010s.

The differences are subtle but noticeable. Blue states like Illinois and New York become substantially more likely to copy California by the 2010s. While Utah, Idaho, and even Texas are geographically close enough to California that in the 1990s you would’ve expected them to learn from it, by the 2010s the very different political stances of the states starts to dominate. One thing to note here is that in addition to the policy diffusion dynamics changing, California itself became much more left-leaning during this period; it had Republican governors from 1982 to 1998 and even voted for George H.W. Bush in 1988. So it makes sense that in the 1990s, Republican governors in places like Utah might have learned from California.

This paper’s point is descriptive, not causal; the authors are measuring the extent to which partisanship and proximity predict policy change, but their methodology doesn’t allow us to say something like “Iowa would not have passed this law if Missouri hadn’t passed it first.” Moreover, this research question is complicated by the fact that states might have the ideologies/partisan leanings they have because they’re close to states with those leanings. Being close to Massachusetts might make New Hampshire more Democratic-leaning, which makes separating out the influence of geography versus partisanship that much trickier. That said, being able to predict which policies spread from place to place is useful even if we don’t get good causal metrics.

Policies are more likely to diffuse these days

DellaVigna and Kim found that the importance of geographic proximity in spreading policies between states remained basically constant from the 1950s onward, but the importance of partisanship rose. Those two changes combined imply that there’s just more policy change happening, period, today than there used to be.

Sure enough, the baseline probability that a state adopts a given law in the dataset in a given year rose from 0.03 early in the sample to 0.05 later on, implying that passage odds nearly doubled. States are doing more learning from each other than they used to.

Some issues are still nonpartisan

The SPID dataset includes incredibly polarized issues like abortion, Medicaid expansion under the Affordable Care Act, and voting rights. But there are still areas of policies that are more technical and where divisions between the parties and ideological camps are not as sharp, and the dynamics in those areas seem more driven by geography than partisanship.

Case in point: Louisiana State’s Ryan Yang Wang and Penn State’s Krishna Jayakar have a nice paper from 2024 looking at state-level broadband internet policies. The policies included everything from laws governing right-of-way for fiber optic and other cabling; to financing and tax incentives for broadband rollout; to municipally run broadband services; to policies around mapping and planning future broadband rollout.

They find surprisingly little role for partisan control or affiliation in predicting if a state will adopt another state’s broadband policy. States with Democratic governors aren’t likelier to borrow from other states with Democratic governors; legislatures with unified Republican control aren’t likelier to borrow from other states with unified Republican control.

However, geography remains very important. They find that if two states are adjacent, they’re three times more likely (odds ratio of 3.028) to have one borrow a broadband policy from the other than if they weren’t bordering. Wang and Jayakar’s interpretation, which strikes me as correct, is that this is an area that’s both relatively unpolarized and where the specific geographies of states (eg how population-dense they are, how hard it is to dig underground wiring, etc.) are very important. That suggests a bigger role for geography than ideology.

Perhaps more surprisingly, Missouri’s Scott Lacombe, University of Iowa’s Caroline Tolbert, and the University of Tampa’s Samuel Harper find that diffusion of election laws is more strongly determined by a nonpartisan, objective indicator of how well states run elections (the Election Performance Index) than any partisan leaning. This surprised me, given that the list of election laws they considered included some that are highly politically contentious, like voter ID requirements, term limits, and moving to the popular vote for presidential elections. But it also includes unobjectionable good-government stuff like letting people register to vote at the DMV, The authors note that “the median policy was adopted by 31 states, and fewer than 20% of the policies were overwhelmingly adopted by states controlled by a single party.”

What does this mean for abundance and growth?

None of the studies here have focused specifically on the issues we at the Abundance and Growth Fund work on: housing supply, energy abundance, cheaper and faster clinical trials, more and better scientific funding, etc. Indeed, Mallinson’s meta-analysis calls out housing, transportation, and environmental policies as particularly lacking in policy diffusion studies.

But I still think there are some useful takeaways for our team:

• Policy diffusion is real. Getting a policy passed in one state can meaningfully increase the odds that another state adopts it. That matters for our estimates of the impacts of state legislation. If funding an effort to have, say, Colorado simplify wind power permitting leads to neighboring states adopting similar reforms, that means our grant there is higher-impact than it looks like at first glance.

• On certain technical issues, proximity is probably the best predictor of where that kind of diffusion will happen. But otherwise, we should expect diffusion mostly to happen between politically aligned states: between blue states, and between red states, but not crossing party lines.

• It might be worthwhile to support parallel efforts in one blue and one red state to make progress on a certain topic. That way, you can set up parallel diffusion patterns, with Republican states learning from each other and Democratic states learning from each other, but the two groups arriving at a similar place in terms of actual policy.

We think this research suggests pro-abundance policies have a decent chance of diffusing at the state level, and we're going to keep investigating how to make that happen. If you have more papers on this or related topics, please link them below in the comments! We’re still very much in a learning process here, and welcome any and all suggestions.

We’re glad to be back after a short break! Here’s what caught our attention over the last couple weeks:

1. The Jones Act, the American law that bans ships manufactured outside the US from traveling between US ports, has been an economic burden on the country for decades. The best study on its impact estimates the cost at 0.1 to 0.4 percent of GDP — which may sound small, but given the size of the US economy that means $31.8 billion to $127.4 billion in economic damage every single year. It’s particularly bad for clean energy, because offshore wind turbines need specialized ships to install them, and the US has built a grand total of one of those ships. Luckily, the Trump administration has waived the Jones Act for oil, fertilizer, and related products since March as part of the ongoing Iranian oil crisis. Colin Grabow at the Cato Institute pulled the data and found that in just two months, 45 voyages from 35 ships have taken place under the waiver, mostly moving various oil products from the Gulf Coast of Texas and Louisiana to the West Coast and Puerto Rico. That suggests meaningful pent-up demand for transit using foreign-built ships that the Jones Act is suppressing. — Dylan Matthews

2. Earlier this month NSF announced its new X-Labs program, a $1.5 billion, decade-long initiative to fund independent, milestone-driven research teams working outside of traditional university and corporate structures. The program aims to address a long-held concern that existing grant mechanisms and institutional models aren’t well suited for team-oriented, engineering heavy, long-term research efforts. This is a relatively small bet, all things considered (it’s ~1% of NSF’s annual budget); but it’s an exciting one that will test a new model for doing science, and support a new ecosystem of research labs taking big swings. — Jordan Dworkin

3. In other news, arXiv’s CS section chair Thomas Dietterich announced that authors who submit papers with “incontrovertible evidence” of unchecked AI content (e.g. hallucinated references or leftover chat content like “here is that manuscript you requested”) will be banned for a year. After that, future submissions will be required to clear peer review before being posted on arXiv. This caused a bit of a stir online, with some researchers supportive and many strongly opposed. One opposition camp was frustrated that this move shifts arXiv even further away from the openness and lack of gatekeeping that made it valuable in the first place and is needed going forward. I’m sympathetic to that concern, especially with regard to the post-ban requirement that submissions pass peer review. But another camp seems to find it unreasonable to expect authors to verify their claims and citations, which I am less sympathetic to. It is certainly true that references have included mistakes or misattributions since long before AI, and these are usually relatively harmless (though they don’t suggest great citation practices on the part of authors). But it’s never been good when the knowledge ecosystem is flooded with unsourced claims, and AI allows that to happen at speed and scale previously unthinkable; if researchers can work with AI to write a paper that is accurate and rigorous, fantastic! But having AI write an inaccurate paper that is never carefully reviewed is not fantastic. Holding authors accountable for the accuracy of the work they’ve put their name on seems worthwhile. — Jordan Dworkin

4. It seems obvious that increased demand for electricity from data centers must raise prices. But electricity prices can move in counter-intuitive ways. We’ve talked before about how electricity price increases largely hadn’t yet been empirically linked to new demand (read: data centers) in most geographies outside the PJM Interconnect (the electricity market that houses data center alley). And in places like North Dakota, increased demand for electricity has been associated with lower electricity prices, since fixed costs of infrastructure could be spread over a larger ratepayer base. Travis Kavulla (incidentally, potentially the next head of the Bonneville Power Administration) explains in American Affairs this week why more places will probably look like PJM - where the grid has little excess juice to squeeze, and new demand mostly does require costly new investment - than like North Dakota as data center demand continues to rise. Additionally, while grid capacity can at least in theory cut both ways, equipment price inflation and higher capital costs really only drive up the cost to serve demand, relative to existing customers. Kavulla explains why bringing on new demand at the same rates as existing customers - whether through typical ratemaking or longer-term take-or-pay agreements - means socializing those higher costs across all utility customers, and argues that treating large loads like other customers doesn’t make sense when they’re potentially driving multiples of existing demand. To avoid the malignancies of the famously strangled generation queue on the load interconnection side, he argues, among other propositions, for an “open season” for large load interconnection to manage grid access, together with strict BYOG (Bring Your Own Generation) requirements that actually match the cost and risk of new supply to new demand. — Willow Latham-Proenca

5. Pew’s Housing Policy Initiative is making a splash with its new video explaining “address chain” filtering: How new market rate construction frees up housing opportunities across incomes and neighborhoods as occupants of newly built homes free up their existing, cheaper homes (not over decades but in just months). Pew also launched a new report on pre-approved building plans — reusable, pre-cleared design blueprints that let builders bypass discretionary design review. The report documents results from ~40 US jurisdictions running some version of the program, with the strongest outcomes in small and mid-size cities like South Bend, Indiana (223 new homes built, ~7% of permits over three years) and Seattle (ADU approval times cut from 160 days to 54). Federal interest is growing: the Accelerating Home Building Act, folded into the 21st Century ROAD to Housing Act, would launch a HUD pilot grant program to scale the model. — Alex Armlovich

6. “The USPS owns over 8,500 properties and leases over 25,000 more. It also needs money.” In a new Brooking Report, Aaron Shroyer, Ben McAdams, and Glen Nuckolls argue we can kill two birds with one stone: the Postal Service should let developers build housing above post offices (or otherwise use their real estate to support housing). They estimate that rebuilding Postal Service real estate with buildings that merely have the same level of density as nearby census block groups could unlock an additional 117,000 housing units in markets with medium-to-high housing demand, and earn nearly a billion dollars a year from the leases. They also point to precedents for this kind of thing, both in the US and abroad. What could be more in tune with abundance than a proposal that simultaneously eases housing shortages and enhances state capacity? — Matt Clancy

7. What is it like to know that a new drug exists – and has been proven safe and effective – but that it simply isn’t available where you live, and you might die without it? This was the reality for millions of Africans during the AIDS crisis in the 2000s. Even when tenofovir was approved in the US in 2001, most Africans couldn’t access it for nearly a decade, instead having to rely on a cheaper but toxic alternative that roughly a third of patients would abandon due to its side effects. These delays have shortened but remain prevalent across much of the world today.
   The new magazine In Development has a great article on one solution: pooling regulatory reviews across countries, so that manufacturers don’t have to file separately in each one and overstretched agencies don’t duplicate each other’s work. The European Medicines Agency effectively does this across Europe, and the African Medicines Agency, launched in 2025 to do something similar across Africa. I learnt a lot from the piece about how it came about and the problems it faces today. Some of the continent’s biggest pharmaceutical markets haven’t signed on, some countries refuse to accept each other’s assessments, and the whole project runs on donor funding without a clear path to self-sufficiency. — Saloni Dattani

8. By now you’ve probably seen the viral clip of Marblehead resident David Modica asking his fellow townspeople the question heard ‘round the internet. Jerusalem Demsas wrote about it in The Argument, making the case that Marblehead’s decision to zone a golf course for housing (knowing nothing would ever get built) is what happens when local democracy mediates land-use decisions. She’s right.

   But I sit on Marblehead’s Housing Committee, and I’ve been watching the aftermath up close. At our May 12 meeting, the mood, in part, was relief: the years-long 3A fight was finally over, compliance was in sight, and the town could stop bleeding resources on it. I get it. The town is in a budget crisis, and the committee chair’s own job has been on the line through all of this. Still, my reaction was that compliance isn’t the right metric for the Housing Committee, housing production is.

   What strikes me, though, is that most of the people involved aren’t acting in bad faith, they’re operating within constraints they didn’t design. Voters first accepted (through record-breaking attendance at Town Meeting) and then rejected (through a dormant referendum process) the more meaningful plan. The state signed off on the weaker one. Everyone did their job, and the system still produced the wrong outcome.

   We just started a team book club reading Alain Bertaud’s Order Without Design, and our first discussion kept circling this tension. Planners often want to do the right thing but are hemmed in by systems that make good outcomes structurally hard. Marblehead is a small-scale version of that story. — Nisha Austin

And a few grantee and team shoutouts and other announcements worth sharing:

• Matt published “Social science at the NSF” earlier this week, looking at the history of fights over social science funding at the NSF and how the current proposed elimination of the SBE Directorate rhymes with the Reagan-era cuts of the early 1980s.

• Progress Ireland published a series on housing in Europe, making the case that the EU has a significant opportunity to reshape the built environment in 2026 through permitting reform, the Construction Services Act, and Brussels-level policy change.

• Mayor Mamdani released “Block by Block: The Housing Plan for a New Era,” New York City’s new 10-year housing plan, which aims to build 200,000 new affordable homes and preserve 200,000 more, backed by $22 billion in city housing funding over five years.

The 2027 National Science Foundation Budget Request proposes to eliminate the Social, Behavioral, and Economic (SBE) Directorate, which funds social science research.1 So far, all signs suggest it means to follow through on this proposal (the NSF’s SBE Directorate has made essentially no research awards so far this year, which is highly atypical). We care a lot about R&D at the Abundance and Growth Fund, but let’s set aside for a minute the question of whether the NSF should be supporting social science research.2 What I find interesting about this whole episode is that we’ve been here before, a little over 40 years ago.

Reagan Revolution

Let’s go back to 1981, shortly after Ronald Reagan took office. The following account of the social science fights of the first Reagan administration are based on a few histories of the period that I read earlier this year.3 The Reagan administration ultimately ended up cutting inflation-adjusted non-defense R&D by a third by 1983, but it was especially hostile to the social sciences. In the FY1982 President’s Budget, the administration called for a sharp reduction in spending on social science research, more severe than proposed for the overall NSF.

The Executive Branch wasn’t an outlier though; Congress was also skeptical. In the same year, it passed a rescission package reducing the appropriated funds for social science at the NSF by $10mn, out of a total allocation of $33mn (Congress eventually restored $1mn). The next year, the Reagan administration proposed cutting the NSF social science budget from the post-rescission level of $24mn by an additional $14mn, but Congress split the difference and appropriated funds equivalent to another $6.4mn cut. Funding bottomed out in 1982 and then began to slowly recover. But the social sciences would not attain the inflation-adjusted level of funding they had enjoyed in 1980 until 1996.

While the scale of these cuts caught the social sciences off guard, they didn’t come out of nowhere. Support for social science research at the NSF had always been more tenuous than other fields: some draft legislation founding the NSF had even banned the social sciences from receiving funding, though an outright ban wasn’t in the legislation that ultimately established the NSF. Decades later, in the lead up to the 1981 cuts, multiple bills had been proposed that cut social science funding at the NSF (one passed the House, but not the Senate, in 1979). Social science research was also the frequent subject of ridicule, for example via Senator Proxmire’s “Golden Fleece” awards, which kicked off in 1975 with a criticism of an NSF-funded social science study on why people fall in love. Perhaps as a consequence of all this doubt, funding for the social sciences had been growing at a significantly slower rate than other sciences at the NSF leading up to the 1980s.

Make America Great Again

Fast forward 44 years to 2025, and the second Trump administration had just come into office. It also arrived highly skeptical of federal support for R&D; President Trump’s FY2026 budget proposed a 20% cut to overall R&D, with much larger cuts for non-defense R&D (57% at NSF, 40% at NIH4). This time, however, Congress was not on board with the proposed cuts. The normal appropriations process for discretionary spending (like R&D) is subject to the filibuster, which means any large cuts to NSF and NIH would have had to garner supermajorities to pass. This meant a large cut through the normal appropriations process was pretty unlikely.

That said, it was possible Congress might pass large cuts at the NSF and NIH with a rescission package instead. Rescissions are a mechanism for the executive branch to seek approval not to spend appropriated funds, and as noted above, Congress passed one in 1981. Through much of 2025, the administration’s spending at NIH and NSF was below trend (see figure below); a large pot of unspent funds near the end of the year could potentially be used as a pretext for rescissions, which unlike normal appropriations, can be passed by a simple majority.5

However, that isn’t what ultimately happened. In the second half of the summer of 2025, spending at NSF and NIH accelerated and ended the year on trend or above (see figure). No rescission packages targeting R&D were proposed. And the eventual appropriations from Congress cut the budget of the NSF by 3.4% (roughly $300mn), rather than 57%. The budget appropriated for NIH actually increased.

The social sciences, however, fared worse. As under the Reagan administration, one can see some indications that support for the social sciences has been less robust than for the rest of the sciences for many years. SBE as we know it today was formed in 1991, following years of advocacy from social science associations. But since then, support has gradually weakened, relative to the overall NSF Research and Research Activities Budget. While it made up 5% of research spending in 1995, that has drifted down to 4% at the time the Trump administration took office.

In 2025 though, support plunged. At SBE, spending in the first half of the year was on trend. But spending in the second half of the year failed to keep up with normal trends and the budget for the social sciences ended the year significantly below trend.6 And five months into the new year, new spending is close to zero.

What happens next?

Perhaps the past can give some clues as to how things might unfold in the future this time around. In response to the 1981 cuts, various disciplinary associations of social sciences founded the Consortium of Social Science Associations, an organization that had been previously proposed but never formally established. This new organization, COSSA, followed a pretty standard playbook to influence the US Federal budget: it hired staff to lobby Congress; mobilized social scientists to visit, telephone, and send letters to their representatives; focused on winning over members of Congress with large universities in their district; and targeted both Democrats and Republicans as part of a message that funding for the social sciences was not a partisan issue. The size of the proposed cuts (the 1982 proposal was a 75% cut relative to what the previous administration had proposed) also galvanized significant media coverage.

The timeline is consistent with this response working, at least a bit. COSSA would have had more time to develop and execute its strategy ahead of the 1982 budget battles than the 1981 rescission fight, and the budget cuts in 1982 were indeed smaller in absolute terms and as a share of the administration’s proposal, than the 1981 cuts. That said, if the campaign moderated the proposed cuts, it could not stop them. According to a contemporaneous article by James Zuiches, the partial reversal of the cuts that began after 1983 owed less to anything specific the social sciences did, and more to a renewed interest by the Reagan administration in research in general. Social science benefitted from the halo around all research (though even so, its growth rate remained below the growth rate of other sciences at NSF).

COSSA is still active today, but it’s less clear what Congress can do in 2026. The explanatory text for the latest NSF appropriations package already specified “No directorate shall receive more than a 5 percent reduction relative to the fiscal year 2024 enacted level.” However, while the overall budget of the NSF is set by Congress, the executive branch has considerably more latitude in how that funding is allocated (see my colleague Jordan’s earlier blog post about this). The preceding explanatory text, for example, does not have statutory force (though such statements usually guide agency actions).7

Zooming out, is there some future where the social sciences can enjoy the levels of political support more common in the life sciences and hard sciences? That’s an interesting question, and one I’ll come back to in a follow-up to this post.

Note: there will be no “What we’re reading” blog post from the Abundance and Growth Fund this week, as a large share of the team is traveling.

Hope you’ve had a great week! Here’s what we’ve been reading:

1. Our scientific ecosystem often prizes outputs that, themselves, don’t necessarily drive scientific progress, but instead are useful proxies for the more intangible processes that do. Three reads this week have me thinking about what happens if AI gets good at creating these outputs before (or instead of) getting good at the things they’ve been proxying for. First is David Bessis’s fascinating essay on AI in math. He discusses, among many other things, the idea that theorem-proving has historically served as a legible demonstration of underlying conceptual innovation, because solving major open problems typically required first building a new framework that made the solution tractable and expressible; AI may break that coupling, producing correct but unintelligible proofs and capturing social rewards without the accompanying distillation and canonization that allow the field to build on the new knowledge. Second is Engzell and Wilmers’ recent preprint, “The Paper Factory,” which, though different in goal and tone, describes a related dynamic in social science. The authors develop a multi-agent LLM workflow that automates most of the steps of creating a publishable empirical paper, but struggles with the intangibles that form the foundation of meaningful outputs, like problem selection and judgment. For a more speculative take on the topic, read Ted Chiang’s prescient short story “Catching crumbs from the table,” published in 2000 in Nature, which imagines human scientists reduced to interpreting the incomprehensible discoveries of superintelligent successors. — Jordan Dworkin

2. The US is richer than Europe. This is not especially controversial: residents of just about every European country, save for tax havens like Luxembourg and the petro-state of Norway, lag the US in median income. But is Europe falling further behind the US, or keeping up? Paul Krugman says “keeping up”; Luis and Pieter Garicano say “falling behind.” Some of their debate is technical and about what the proper purchasing power parity (PPP) metric to use when comparing living standards between the US and Europe. But the biggest question in the debate is how to think about US dominance in tech. The five most valuable companies on Earth (NVIDIA, Google, Apple, Microsoft, and Amazon) are all American tech firms, and the smallest of them is nearly five times more valuable than the most valuable European tech firm (ASML). To Krugman, this is basically irrelevant to living standards; Europeans can still use those American firms’ products, often (as with Google) for free, so the physical location of the firms isn’t so important. To the Garicanos, this pattern is evidence of a fundamental lack of dynamism in Europe that is already hurting residents and will hurt them more as AI scales up. — Dylan Matthews

3. Alan Durning at Sightline has a new report arguing that the climate movement is underrating one strategy for reducing climate change: apartment buildings. Apartment buildings have lower carbon emissions than single-family homes for several reasons. They are more energy efficient; more likely to be heated by electricity; and when sited near offices, shops, and public transit, let people get around more on foot, by bike, and by public transit instead of driving. But perhaps just as important is the fact that housing policy is tractable. Building more apartments doesn’t require new government spending, nor asking people to make hard tradeoffs. Private developers want to build apartment buildings, and people want to live in them, if only they are allowed. We’ve made a lot of progress in the last year to enable building more apartment buildings, but there is a long way to go. — Matt Clancy

4. Matt Yglesias dropped an avowedly contentious take on energy abundance, arguing that “conventional environmentalism” is overly focused on efficiency and missing the upside of truly abundant energy generation. Yglesias argues that “too cheap to meter” energy would not only make efficiency arguments irrelevant, but unlock vast new frontiers in energy-constrained areas like synthetic fuels, vertical farming, and lab-grown animal proteins. Leaving aside the political positioning of energy efficiency for the moment, I found this a helpful provocation on what we should and shouldn’t expect from more generation capacity. While more cheap energy is clearly a good idea, and we're sympathetic to many of the conclusions here (permitting reform and interregional transmission would be important wins), my read is that we still risk missing a step by focusing too closely on energy prices alone as a driver of innovation. For some sectors, the cost of energy probably is a binding constraint (synfuels, desalination, potentially some green industrial materials), so it follows that cheaper energy would lead to growth there. For others — including cultured meat and potentially vertical farms, but also next-gen aviation — there are still major materials science/biotech bottlenecks (or physical constraints) to be addressed. It’s not clear to what degree even too-cheap-to-meter energy would actually induce the innovations to make those technologies commercially viable; at least some of the literature on directed technical change implies that energy prices can shift R&D budgets within the energy sector, but we’re (to my knowledge!) missing an empirical through-line from energy prices to R&D in complementary fields. If we don’t know how strong the link between energy price and complementary innovations is, it’s worth asking whether driving energy prices as close to zero as possible is the most cost-effective way to induce those innovations. That said, there are plenty of good reasons to build more cheap energy — not least that growing demand has made ever tighter supply-side constraints an increasingly real possibility even for “conventional” growth. — Willow Latham-Proenca

5. New York took two big steps on housing permitting this week. After weeks of late-budget deadlock, Governor Hochul’s SEQRA modernization is reportedly final with bill text coming next week. The deal reportedly exempts NYC housing projects up to 500 units in medium/high-density districts and 250 in low-density areas, with smaller carveouts upstate. On the city side, the mayor launched the SPEED Task Force report. Originally scheduled to arrive after a final Albany budget, the report is explicit that its headline pre-certification cut — slashing ULURP rezoning pre-cert from two years to six months — is contingent on the state SEQRA carveout actually landing. Beyond SEQRA-enabled reform, SPEED promises: five months trimmed from new-construction permitting, five months from office-to-residential conversion permitting and lease-up, and a ground-up rewrite of the affordable housing lottery that takes median time-to-move-in from 210 days to under 100. All told, the report projects two years off the timeline for projects requiring rezoning and eight months off all affordable housing — backed by $14M in the executive budget and ~96 new agency hires where capacity actually binds the expected flood of streamlined housing permit applications. With Evan Soltas’s LA permitting work implying each year saved is likely worth ~8% in construction costs, that’s the stacked feasibility shift needed to help turn theoretical zoned capacity into actual homes. — Alex Armlovich

6. FDA Commissioner Marty Makary resigned this week, just over a year in the role. It seemed sudden from the outside, but was reportedly a long time coming internally, with months of friction with RFK Jr. Apparently, the final straw was his reluctance to approve flavored vapes. I find the chaos worrying, and although Makary announced some promising ideas, like streamlining clinical trials, Bayesian analysis, and reducing animal testing, as STAT news describes, he also oversaw the departure of experienced career scientists, more politicized approval decisions, and eroding evidentiary standards. — Saloni Dattani

7. One reason that clinical trials are slow, expensive, and wasteful is that the data they generate are locked in siloed systems and manually transcribed. Recently, the FDA announced a pilot project to fix this, which would create “Real Time Clinical Trials”. The idea is to let regulators see trial data in real time, as it’s collected, potentially compressing the dead time between study phases. But as Adam Kroetsch writes in a new post, the announcement was muddled; speakers discussed topics like rural access and site burden, which made it unclear what was actually being proposed. His take is that the idea is genuinely promising, but will be difficult to deliver at scale, and requires the FDA to go much further. He recommends they define how they actually use live data, build interoperability standards, and give sponsors explicit regulatory cover to abandon unnecessary verification practices. — Saloni Dattani

Here are a few other highlights and announcements from our team and grantees:

• Alex Armlovich published “Tokyo land is still >$85 million an acre” looking at what happens to land values when a megacity actually achieves housing abundance. The short answer: rents fall, but land values don’t, which matters for the political economy of getting zoning reform passed.

• NSF officially launched the $1.5 billion X-Labs initiative, funding independent research teams with large, flexible block grants to tackle scientific challenges outside traditional university structures. The initiative draws on ideas from across the science policy community, including the Institute for Progress, whose X-Labs proposal last August laid out a framework for this kind of funding model.

• The Foundation for American Innovation is hosting the Energy Imperatives Summit on June 9-10 in Washington, DC, a two-day forum on energy policy featuring speakers from government, industry, and finance.

Megacity YIMBYism lowers rents but not land values—plus who wins, who loses, and why most mortgage-holders shouldn’t worry.

When you zoom out on American politics, one pattern that snaps into focus is how much local elected officials focus on preserving, even boosting, home values. It’s a reasonable political instinct. Roughly two-thirds of American adults live in owner-occupied housing, and for most of them the home is by far their largest asset. The threat that finally building enough housing to satisfy pent-up demand would crash widely-held asset values is taken seriously across the spectrum—by populists who promise rising prices to homeowners, by mortgage-market watchers who remember 2008, and by political-economy theorists in the William Fischel tradition who treat homeowner profit as the prime mover of American zoning.

That fear is mostly wrong, but in an interesting way. Good public policy should indeed raise total land values: insofar as land markets are efficient at capturing all the “goods” and “bads” about living in a particular location, policymakers and society at large should want high land prices because they proxy for good wages and high quality of life.

Thus the right question isn’t whether YIMBYism would lower the user cost of housing—that’s the whole point of the YIMBY movement and the unanimous upshot of the academic urban economics literature—but whether it would lower the value of land overall. The answer will almost certainly turn out to be no, because higher land prices are substantially severable from, and can coincide with, lower structure prices. And Tokyo, the world’s largest city and the only one that features anything like a realistic best-case version of housing abundance at megacity scale, is the cleanest place to see it.

The world’s largest YIMBY metropolis still has more than $85M-an-acre dirt

Tokyo is the world’s largest single metropolitan labor market. Greater Tokyo holds about 37M people, and it’s still growing as Japan’s countryside empties out and reforests. As a comparison, Tokyo first surpassed NYC in population in the early 1960s; after decades of steady growth, its nearly twice the size of metropolitan NYC’s 19.3M residents.

By the standards of Anglosphere megacities—New York, London, Toronto, LA—Tokyo housing is famously cheap. Builders can put up small-lot single-family homes, midrises, microapartments, and single room occupancy-style shared housing units with ease and in large volumes. (High-rises are not allowed by-right everywhere, so this isn’t a pure laissez-faire experiment, just the closest real-world example.) If “Tokyo regulation” is what real-world YIMBY victory looks like in a global megacity, it’s the best dataset we have.

So what are Tokyo’s land prices?

Prices for land range from well over $100M per acre in the urban core (in 2024 USD, PPP-adjusted) to just under $20M per acre on the periphery of Tokyo proper’s 23 wards. Whether one uses PPP (~$154M/acre) or market conversion rates (~$90M/acre) is not particularly important to the thesis: The point is that land prices are extremely high, and demonstrate a typical metropolitan bid-rent gradient from the core to the periphery.1

The famous (relative) cheapness of Tokyo housing is not a story about cheap land. A one-acre detached house in central Tokyo would cost more than $100 million in dirt before you broke ground. The land is pricey but the structures are cheap. Admittedly, Tokyo’s rents and prices are not as cheap per square foot as buildings in the US Sunbelt’s midsize cities, but cheap by the standards of any 10-million-plus Anglosphere metro area. Tokyo built its way to relative affordability without ending up with low land values, and the values themselves look reasonable for a productive, agglomerated megacity that simply didn’t artificially restrict its own supply.

Land prices and structure prices are related but different things

Standard urban economics tells us that wages and amenities accessible from a location capitalize into the price of land at that location. The harder practical upshot, which Ed Glaeser and his coauthors have spent their careers teaching, is that whether high land prices translate into high structure prices is a function of land-use regulation.

High land prices are a price signal. They tell builders to economize on land per unit by stacking up more structure—more floors, smaller footprints, higher floor area ratios per unit of land. In a liberally regulated environment, the bid-rent gradient for land can be extremely steep, while the bid-rent gradient for structures is much shallower.2 Structure costs do rise with height—the U.S. model building code introduces real cost steps as a structure grows in density. A build starts needing tuned mass dampers and bespoke engineering somewhere past 50 stories—but those increases should be shallower than the underlying land-price gradient if regulation allows.3

As a useful sanity check on what land prices are really capitalizing: Consider tightly zoned waterfront cities, where the going price (and waitlist) for houseboat moorings—physically not land at all—behaves like part of the residential property market. The product being priced is not actually dirt! It’s locational access to the bundle of public and private goods available at that point on the map. Upzoning lets builders amortize that location access cost over more buildable square feet. It does not, and could not, reduce the value of the access itself as long as the net external congestion costs of growth don’t exceed the net external amenity and wage-agglomeration benefits of growth.

When policymakers upzone widely with by-right permitting across a high-demand metro, two things should happen at once. Measured per-acre (like farmland), land value rises, because the parcel has been granted a valuable option to host more buildable area. Land value measured like NYC-area developable land per buildable square foot falls, because that higher per-acre value is being amortized over much more floor area.

These move in opposite directions, and neither one alone is “the land price” in the sense political conversation usually means. This distinction dissolves a lot of the political-economy panic around housing abundance. Again, land prices quoted the way NYC land brokers quote them, on a “Zoning Square Feet” or “Buildable Square Feet” basis, will fall even as the total value of land is at least stable or rises in the metro area being upzoned.

The Homevoter Hypothesis Can’t Explain Everything

The leading academic framework for explaining American NIMBYism is Dartmouth economist Bill Fischel’s pairing of fiscal zoning and the homevoter hypothesis.

The fiscal-zoning argument is that single-family-only zoning and large minimum lot sizes are profit maximizing tools for managing the local tax base: By excluding apartments and small houses, exclusive suburbs effectively “faregate” the municipal border, restricting admission to households whose property taxes will exceed their consumption of local public services like schools. Local public goods become club goods, and the club charges admission at the municipal border rather than the schoolhouse door. Instead of hosting private for-profit schools, these suburbs act as private for-profit neighborhoods with an ostensibly “public” school inside.4

The homevoter hypothesis is the political mechanism: Homeowners turn out in local elections to defend the asset values that this fiscal arrangement protects.

Fischel’s homevoter hypothesis rests on two claims: that single-family zoning maximizes the asset value of incumbent owners, and that those owners (the electorally pivotal median voter in most US jurisdictions) turn out to defend it.5 The second half holds up in the suburban cases Fischel built the theory around. The first half breaks in the cases that matter most for national productivity: the high-demand cores of America’s largest cities.

Tokyo’s land and property market conditions show why: Tokyo combines metropolitan housing abundance with land prices that remain high, not low. The two facts coexist because the relevant scarcity isn’t housing supply; it’s centrally-located land. A homeowner sitting on a single-family parcel in a high-demand Tokyo ward holds a portfolio that is overwhelmingly land. Upzoning her parcel raises the value of that land, sometimes dramatically, because the parcel can now carry dozens of units rather than one. The option to redevelop is worth more than the right to preserve neighborhood character. Restrictive zoning, in this regime, is asset-destructive for a land-rich homeowner in a high-demand area.6

If American single-family homeowners in the cores and first-ring suburbs of New York, San Francisco, Boston, or LA were maximizing the dollar-denominated asset values Fischel says they care about, they would be voting to become like Tokyo landowners — to unlock the redevelopment option on their parcels. They are not. Whatever they’re maximizing, it isn’t profit.7 (See Agenda for Abundant Housing for a full-length argument.)

That’s not a fatal blow to the homevoter framework if we’re willing to relax the “objective function” of voters from strictly defined profit-maximization. As any good undergraduate economics professor will remind a student who discovers people valuing non-pecuniary interests: “Firms maximize profit. People maximize utility.” This is of course the beginning, not the end, of the question: It is the task of the other social sciences and humanities to help economists figure out what it is that people see as utility-maximizing or otherwise in their best all-things-considered interests, when consumers are not behaving in a way that maximizes apparent pocketbook dollars and cents.

Amenity preferences, socioeconomic and positional status concerns, and outright economic confusion all seem relevant — Clayton Nall and co-authors find that a slight majority of Americans believe new supply raises prices.8 But it does mean the rational-asset-defense story isn’t sufficient for the high-demand core cases…which is to say, for the cases where land-use reform would do the most national good. Maximizing “Baby Boomer suburb aesthetics” in neighborhood character is, in the cores of the most important metropolitan labor markets, a goal in direct tension with homeowner profit maximization–and all Americans, homeowner and renter alike, are collectively paying a tremendous price.

The bottom line

Tokyo is the closest thing we have to a real-world YIMBY megacity, and Tokyo land is still very expensive. Tokyo builds at a Sunbelt city pace–around 2% per year–far higher than other global megacities. Cheap housing and expensive land are the very likely fruits of successful housing abundance in America’s superstar cities, and also in most of the desirable areas of our other cities.

Aggregate land values should rise with wage and amenity access under abundance, even as the geographic distribution of land values will change. Per-unit prices will fall. Land-rich owners will profit. Structure-rich owners will likely take a haircut. All renters win. And, importantly, the mortgage market will not break provided we don’t time a supply rollout to coincide with a recession.

Finally, the political-economy theory that homeowner asset defense explains American zoning has to reckon with the fact that, as Tokyo’s high post-Abundance land values indicate, the assets in question would be uplifted, not bankrupted. The narrow profit-maximizing excuse for NIMBYism does not apply to the homeowners with strong redevelopment options in the metro areas where YIMBYism can do the most good.

Happy Friday! Here’s what caught our attention this week:

1. “Are we kind of being pricks?,” asked a resident at Marblehead’s annual Town Meeting this week in a video that went viral (full disclosure: I live here and serve on the Housing Committee). After four votes over three years, the town approved a zoning overlay to comply with the state’s MBTA Communities Act, this time by concentrating the required multifamily capacity on a private golf course, where it’s all but certain no new housing will be built. Marblehead isn’t alone. A Boston Foundation report from earlier this year sorts municipal responses into three categories: towns that exceeded requirements, towns that achieved incremental reform, and towns whose zoning will yield little or no new housing. The law has put nearly 7,000 units into the pipeline across 34 communities and is the most effective state zoning policy Massachusetts has passed in decades. But its flexibility was a deliberate tradeoff: ask too much and the whole thing collapses politically, ask too little and you get compliance plans built around golf courses. The law lets towns put their zoning capacity on land where nothing is likely to get built, which is why it matters when some guy stands up at a microphone and says what’s actually happening. — Nisha Austin

2. Our existing processes for how we decide what is allowed when there is disagreement are a key reason its hard to build things in the USA today. Over the last week, Marc Dunkelman and Alex Mechanick wrote an interesting pair of substacks grappling with this question. Dunkelman is concerned that abundance will require empowering someone to make decisions that not everyone will agree with (a decision he thinks not everyone will agree with!). He argues housing reforms have been successful in large part because they have implicitly selected someone to make a decision, in this case the property owner who decides whether or not to build more densely on land they own. But he points out the same trick won’t work with infrastructure that spans many parcels, like transmission or rail lines. In those cases, a centralized authority will need to make decisions about whose interests win out, and empowering a central authority to make those decisions runs counter to some liberal sensibilities. Meanwhile, Alex Mechanick asks how it can be simultaneously true that the government faces so many veto points that it cannot enact its decisions, but also is so unconstrained by process that it risks autocracy. The resolution for Mechanick is that many procedural checks are badly designed; they don’t constrain in cases where they ought to, and in other cases create delays, uncertainties, and costs, without offering much in the way of benefits. I’m inclined to think the kinds of procedural checks on government are the way we, in fact, make decisions about how to balance competing interests. As I see it, a key focus of the abundance agenda is that (1) these processes don’t work that well and (2) they implicitly put too little weight on the virtues of long-run growth, by making it easy for incumbents to veto projects whose net benefits are large and positive. — Matt Clancy

3. Two recent pro-housing champion advocates make the case that zoning reform is working. First, Nolan Gray of California YIMBY goes deep on the California data in “Where Are All the Cranes?,” documenting how ADU permits have produced nearly 150,000 new units, the state density bonus law is facilitating thousands more, and the signature CEQA and transit-oriented zoning reforms are less than a year old. Then this week Michael Andersen of Sightline Institute argues the picture outside California is just as encouraging: states across the country have only recently passed meaningful reforms and the results are beginning to show. Both pieces spiritually or explicitly respond to Derek Thompson, Marc Dunkelman, and Ezra Klein’s recent podcast episode reflecting on how the political landscape has changed in the year after Abundance was released. In that podcast, they worry that the vibes around abundance are stronger than the outcomes. For our part, we’re far more sanguine. While zoning isn’t everything, its hard to build anything if its not allowed and the evidence that significant and clean reform leads to construction continues to pile up. — Alex Armlovich

4. Oliver Kim has a detailed review of a new book on Singapore’s public housing system that complicates the YIMBY movement’s frequent invocation of Singapore as a model. The country houses 76 percent of the population in high-rise public units called HDBs with 99-year land leases, creating a version of the Georgist’s political economy problem: the state is the residual claimant of those structures when the land leases expire. Eventually the government must retake possession and let the terminal value of leasehold HDB homes hit zero. That’s hard to reconcile with residents’ intuitive sense of property ownership. The other challenge is more structural: because Singapore is already mostly high-rise and homeownership-focused, upzoning can’t unlock land value the way it does in American superstar cities dominated by single-family homes. In contexts with large potential density changes, like the US, allowing a developer to build an apartment building where a single-family house stood creates meaningful land value uplift that helps pay for the transition. In Singapore, unless the government increases allowable lot coverage ratios, there’s little additional value to generate from redeveloping existing HDB blocks. — Alex Armlovich

5. I’ve been thinking a lot about the relationship between energy and growth lately, which is less clear-cut than you might expect. I’ll be writing more about this soon, but this set of papers on energy and development does a nice job illustrating part of why it’s so complicated. Colmer, Lagakos and Shu, in a paper revised and published last month, find that improving productivity in the energy sector might not actually affect GDP growth very much, casting doubt on the idea that energy is a critical “weak link” driving development outcomes. However, a 2020 paper by Fried and Lagakos (yes, same coauthor) shows why the energy/growth link might still be an important driver of productivity, even if it’s more complicated than more generation > more growth; they model long-run general equilibrium effects of eliminating power outages at about 20% of GDP for a sample of 5 African countries. Together, these imply that growth effects might be asymmetrical – resources wasted on expensive self-generation, and the high cost of entry this creates for new firms, might be more important as a drag on growth than new generation capacity is as an accelerant. — Willow Latham-Proenca

6. Loudoun County, Virginia is notable for a few things. It’s the richest county in America by median household income, significantly richer than runners-up in Silicon Valley; it houses Dulles Airport and is thus the exurban area DC residents find ourselves schlepping to most often. But its most important role is as the capital of the internet. Due to proximity to the Defense Department, northern Virginia has long been home to a disproportionate share of the ARPAnet and then internet’s core infrastructure, and it still houses 13 percent of the world’s data centers today, with Loudoun and the town of Ashburn specifically at the epicenter. Judge Glock of the Manhattan Institute has an excellent piece explaining what this has meant for Loudoun as a county. In the 2027 budget, the county projects that it will get a whopping 45 percent of its revenue from data centers. For residents, that has meant lower property tax rates, lots of investment in schools and roads, and general, well, abundance. Data centers are not generally popular with voters, but they’re definitely good for Loudoun County. — Dylan Matthews

7. Drug development fails 95% of the time, making biotech a brutal investment environment. So how does the industry keep the money flowing? Abhishaike Mahajan, also known as Owl Posting, has a great post on the financial engineering tricks that have emerged to make this survivable. The examples range from hub-and-spoke holding companies, which pool uncorrelated drug programs so that one big win can offset many failures, to synthetic royalties — manufactured financial claims on future drug revenue that let biotechs raise cash without diluting shareholders or taking on debt. Priority Review Vouchers also get a mention — tradable tickets that speed up FDA review, originally designed to incentivize neglected disease drugs. As I’ve often thought, he mentions that while they reward getting a drug approved, they don’t reward whether it actually gets manufactured or reaches patients affordably. — Saloni Dattani

8. Many leading science funders have a stated goal of supporting ambitious, “high-risk, high-reward” research, or at least specific programs dedicated to that cause — to give a few examples: “ARIA was built to… go after ideas that may seem far-fetched, but could unlock world-changing capabilities,” “[ARPA-H] provides leadership for high-risk, high-reward biomedical and health research,” “[VolkswagenStiftung] supports groundbreaking and risky research ideas,” and “NSF’s commitment to fund high-risk, high-reward ideas strengthens the U.S. economy...” At the same time, there is relatively widespread concern in the metascience community that science is becoming too risk-averse, conservative, and incremental, and there is growing research evidence to back up the case. What gives? Part of the answer is institutional: Franzoni and Stephan argue that standard review processes, regardless of stated goals, are not structured to properly evaluate risk; and Azoulay and Greenblatt find that risky NIH grants are indeed renewed at lower rates, especially for novel research areas and new investigators. But incentives for/against risk are also social and individual: risk aversion can be rational at the level of the individual scientist, even if it’s collectively suboptimal. So is the scientific enterprise too risk-averse? This week Johns Hopkins hosted a debate on exactly this question, with Tyler Cowen and Brandon Ogbunu arguing for, and Kate Biberdorf and Sethuraman Panchanathan arguing against. In the end, the “too risk-averse” side won more converts among audience members, but both sides agreed that expanding support for science is one of the most important levers the enterprise could pull to increase innovation. — Jordan Dworkin

Shoutouts and a few other things worth mentioning:

• Alex Armlovich joined Jamie Rubin on Vital City’s After Hours podcast to discuss the downsides of a rent freeze and what happens to buildings when operating costs outpace revenue.

• Matt Clancy published an outline of the Abundance and Growth fund’s plans for what kinds of work to fund in 2026, titled How do you get Abundance and Growth?.

• Saloni Dattani joined Ben Southwood on the Works in Progress podcast to discuss how to speed up clinical trials, covering everything from Eroom’s law to why pharma companies are moving early trials to Australia, with Ruxandra Teslo as a guest.

• Ruxandra also got a shoutout on the Ezra Klein show about her work on Clinical Trial Abundance.

• Chris Snyder was interviewed on Freakonomics, where he discussed the Market Shaping Accelerator, a program he co-directs that designs incentives to spur private-sector innovation in areas where commercial returns alone fall short of social need.

• Progress Ireland played a role in Ireland’s new planning exemption for backyard modular homes.

• The Centre for British Progress is looking for someone to work on clinical trials policy

• Inclusive Abundance put out a request for policy proposals as part of their Abundance Agenda, looking for bold federal policy ideas across housing, energy, health, and governance.

• The Build America Caucus’s April newsletter from Rep. Josh Harder covered the caucus’s bipartisan work on housing, infrastructure, and permitting, and featured a couple of our posts.

What happened to growth? Part of the answer is policy. When there is insufficient weight put on the long-run growth effects of a policy, we make bad tradeoffs and pursue better policies with insufficient urgency. We think better policies are possible and there is an opportunity for philanthropy to help get us there.

Our primary goal at the Abundance and Growth Fund is promoting broadly shared economic growth—reducing the cost of living by increasing the supply of goods and services—via better policy. We’re a three-year fund with a $120 million commitment. In this post I want to give a bird’s eye view of the kind of work we’re supporting in 2026.

Policy areas

There are a lot of different policy areas that bear on growth and the cost of living, but we’re starting with five areas that we think are important and tractable: innovation, energy, clinical trials, housing, and state capacity.

Our first area is innovation. Technological progress, drawing on scientific research, is the foundation of economic growth. But government funding for science has not significantly evolved for decades. Spending has been fixed at 0.4% of GDP since the 1970s, despite a large academic literature documenting that federal funding for research has a large impact on productivity growth (it probably pays for itself via increased tax revenue, over a long enough time horizon). Meanwhile, even though the returns to science are high, we think they can be even higher: the science funding ecosystem has not evolved significantly for decades and is starting to show its age. Across a variety of different indicators (Nobel prizes, citations, patents), the impact per dollar of recent science seems to have fallen. In our grantmaking to support increasing innovation, we break it down into three pieces: talent, resources, and productivity. Talent refers to the people doing the innovation, resources to the funding and other support they have, and productivity is how well our talent is able to use resources (including AI) to drive innovation.

Our second area is energy. For the average person to benefit from innovation, new technologies need to actually be deployed at scale. But building new energy infrastructure - whether solar, wind, geothermal, or nuclear power - requires navigating a government approval process that has become dramatically more onerous over time. For example, in the 1970s, environmental reviews were a few dozen pages on average; the average is now above 1700 pages, and they take years to complete (more here). This constraint on the supply of energy is likely to worsen over time as demand for electricity from electric vehicles and data centers grows. Our energy program focuses on policy reforms to make it easier to get new energy technologies built in the real world. Specifically, we’re supporting efforts to bring about federal policy reforms to streamline energy permitting, both in the near term and over the longer run, and to support similar reforms in the states.

Next comes clinical trials. New drugs face a similar dynamic as energy; breakthroughs in biomedical science can lead to better health, but only if people can access them. The cost of getting a drug to market has been doubling roughly every nine years. Much of this cost is associated with the costs of running clinical trials, which have slowed over time from roughly six years in the 1970s to 8-9 years in the 2010s. As with energy, we’re again interested in supporting national policies that reform the clinical trials system to make it faster and less expensive, without compromising safety. But we’re also interested in increasing transparency around the data used in clinical trials, as a way to speed up research on drug efficacy and new applications (especially as AI potentially lowers the cost of analyzing that data).

Our fourth area is housing. Most people spend roughly a third of their income on housing - the single largest expenditure category - so if we want to create broadly shared economic growth, creating abundant housing is a prerequisite. Alas, since the 1960s, many local governments have increasingly restricted the ability to build more housing in the places people most want to live, whether via zoning, excessively restrictive building codes, parking requirements, or any number of other policies developed by the 20th century “growth control” movement. The result is a growing gap between construction costs and home prices in high-income metros, with real consequences for broadly shared growth: migration to high-wage cities has reversed since the 1980s, and lower-wage workers are now worse off if they move to expensive metros once housing costs are factored in. We support work to change these policies to increase the supply of housing. While we have historically focused on the most supply-constrained big cities, this year we want to look at state and regional reforms in jurisdictions that might be more tractable to work in before the wedge between prices and costs gets too large.

Finally, cutting across all of the above is the capacity of the government itself to perform its functions well, which we call state capacity. Just as government policies can impede the supply of energy, medicine, and housing, so too can it impede its own function. For example, policies designed to prevent biases in hiring decisions can mean subject matter experts who best understand what is needed for a role are excluded from key parts of the hiring process. In other cases, policies designed to reduce paperwork increase it. More broadly, Americans are not happy with how well the government works: for decades, between half and two-thirds of Americans have believed the government is generally ‘wasteful and inefficient.’ This matters for growth in two ways: government policy affects the economy and the government is itself a major economic player (US federal government spending is 14% of GDP, even after setting aside social security and paying interest on the debt). To increase the ability of the government to execute, we support reform policies that cut across all of government. Two areas we are especially interested in are rules around civil service hiring and government procurement, which jointly determine whether government objectives are carried out by government employees or contractors who are positioned to do the work well.

Most of our focus is on the United States, but we are also interested in these policy areas in other high income countries,1 and have made some grants to support work abroad (mostly in Europe).

Strategy areas

Within those domains, we support a lot of different kinds of work - anything we think will improve the prospects of a sufficiently good outcome. It can be a bit hard to describe all the kinds of work we do concisely, but one way to organize it is on a continuum from research to fieldbuilding to practice.

On one end of the spectrum is research. Sometimes, it’s not clear what the best policy even is, or the evidence base for a policy is too weak for us (or policymakers) to be sure it’s a good idea. In that case, we may fund academic research to try and build a better evidence base about the tradeoffs of different policies. We’ve made several grants like this related to improving our understanding of what kinds of policy changes can improve science (for example, we are co-funding a series of grants with the UK Department for Science, Innovation and Technology on this topic), and are looking at opportunities to support research on what kinds of permitting reform would have the biggest impact on building energy infrastructure. But as a group focused mostly on policy change, we usually think about supporting research with the objective of shedding light on a specific policy question, rather than open-ended exploratory research.2

Once we have some clarity around what policies will be most effective, we often support fieldbuilding; work that spreads ideas and grows the network of people who care about them. For example, we were founding investors in The Argument, in part because we didn’t see a publication looking at contemporary policies and politics with an abundance frame. We also have a program to support experts part-time to write accessible living literature reviews. Other work focuses on the people who are excited by ideas. For example, we support some of the major conferences in these policy areas (YIMBYTown, Metascience, the Roots of Progress, the Abundance conference) as well as some membership organizations that help people interested in these topics learn more and actively participate. This process can take time; for example, while our housing policy work has recently racked up major wins, we have been supporting YIMBYTown since 2016.

Finally, when there is an ecosystem of people and a solid evidence base on policy design, it’s time to support practice. That can take a lot of different shapes as well. In some cases, we help launch organizations that directly do important work, such as the International Student Resource Center, which provides authoritative immigration information to international students in the USA; Pilot City, which organizes workshops where city agencies pitch research projects that will help them achieve local government goals to academics, and vice versa; and research accelerators like Brains and BiTS. In other cases, our grantees generate and provide information that is directly useful to policymakers. For example, we have supported modeling work on the budget impacts of high skilled immigration reform, or the climate impacts of permitting reform. And, of course, other organizations we have supported directly advocate for specific policies. These might include DC-based think tanks (such as the Institute for Progress), or regional groups that focus on their state or local government (such as California YIMBY, Open New York, or the Sightline Institute).

What’s next

We think innovation, energy, clinical trials, housing, and state capacity cover some of the most important and tractable policy areas for promoting broadly shared growth, but there is a lot more we could do. Part of that will come down to how our budget changes over time; it isn’t fixed and we are seeking additional funding.3 But part will also be the outcome of research into new potential policy areas.

Three new areas we’re particularly interested in looking at are transportation, healthcare, and restrictions on labor choices. Transportation infrastructure facilitates access to the productivity benefits of major cities, but it is also bedeviled by many of the same permitting and approval processes that restrict the supply of energy and housing. Health is about delivery of medical care even more so than the supply of new medical treatments, but the supply of doctors and other health services is also constrained by specific policy choices. Meanwhile, productivity growth rises when people can more flexibly change careers to pursue what they think suits their skills, but occupational licensing regimes and non-competes restrict labor choices.

Meanwhile, even within our initial set of five areas, I expect our plans will change; research turns into clear policy goals, fieldbuilding leads to mature ecosystems, and policies eventually get passed. And we’ll also learn from what’s working and what isn’t, and like everyone else, adapt to a changing world. But for 2026, that’s what the Abundance and Growth Fund will be up to.

Matt Clancy: The British Industrial Revolution in Global Perspective. Why did the Industrial Revolution happen in the UK and not somewhere else in Europe? Allen argues it’s down to economics. It’s only worth doing all the work to develop a steam engine if you expect people will want to buy it, and that will only happen (initially) in places where the cost of labor is expensive and the cost of coal is cheap. The UK qualifies, but most of the rest of Europe didn’t. Allen’s book was the first serious work I read about innovation and it changed my whole career. Not because of Allen’s specific argument, which I’ve subsequently been convinced wasn’t a big part of the story. Instead, it was the notion that technological progress - which I already believed was the big lever of rising material prosperity in human history - had causes which could be both understood and influenced. Eighteen years later, I’m still working on that.

Dylan Matthews: The Land of Too Much: American Abundance and the Paradox of Poverty. My first obsession in public policy was the problem of poverty in the rich world: why, in a nation as wealthy as the United States, do people still have to endure homelessness and hunger? It’s a particularly big question for the US specifically, where poverty rates (especially if defined in “relative” terms) are significantly higher than in peer countries. The most provocative and mind-expanding explanation comes from this 2012 book by sociologist Monica Prasad. America's unusual level of poverty is ultimately the result, she claims, of our outrageous levels of agricultural productivity in the late 19th century. That surge in production led to an agrarian populist movement which demanded not a strong safety net but extensive access to credit. Loans and government redistribution are both ways that people are able to spend money they don't immediately have access to, and the US chose to embrace debt over social insurance, with far-reaching consequences to this day. The value to me, though, is less in the argument’s specifics than in its ability to demonstrate the importance of seemingly ancient policy decisions in shaping the structure of American society today. That suggested that setting up better policy basics now, the way AGF’s grantees seek to, could prove massively high-value.

Jordan Dworkin: As Science Evolves, How Can Science Policy? Coming up in science, my peers and I enjoyed critiquing the institutions in which we worked, and occasionally used our computational skills to probe and quantify the scientific ecosystem. But that exercise often stopped short of attempting to fully understand the structures and incentives that produced the inefficiencies we critiqued, or identifying the paths that might be charted out of them. In the mid-2010s, the work of a growing economics-of-science community helped me start to bridge that gap. This 2011 NBER Innovation Policy and the Economy chapter by Ben Jones was particularly influential. Building on his then-recent “burden of knowledge” theory, Ben laid out the evidence that knowledge accumulation was leading to increased specialization, more extended training, and an emerging dominance of team science. But he also took the important step of grappling with how science policy would need to rethink grant mechanisms, evaluation systems, and incentive designs to adapt to the changing enterprise. There was a lot of that grappling in the roughly three-year period surrounding his piece; in hindsight, 2011-2013 saw the publication of many now-classic ideas that shape metascience today.

Saloni Dattani: I don’t have a foundational text that drew me to clinical trial reform. I learnt about the topic during the Covid-19 pandemic. It was surprising how efficient some pandemic clinical trials were and, despite it being an emergency situation, it seemed to me that a lot of these processes could have been applied more widely. The RECOVERY trial, for example, tested around a dozen drugs within two years and quickly identified treatments estimated to have saved over a million lives. The COVID-19 vaccine trials ran much faster than usual, thanks to parallelized phases, rolling regulatory reviews, Operation Warp Speed, and other reasons I outlined in a piece forecasting when the vaccines would arrive. Some pieces I found memorable from the time were a book chapter on the history of clinical trials in the US and how AIDS activists reshaped the FDA, and an episode with Martin Landray, who co-led the RECOVERY trial. I’d also read Stuart Ritchie’s book Science Fictions, on bias and poor research practices in science, and Ben Goldacre’s Bad Pharma. Both made me wonder about ways to improve scientific rigor while limiting bureaucracy, which I wrote about in a piece later on.

Willow Latham-Proenca: Political Order and Political Decay: From the Industrial Revolution to the Globalization of Democracy. I frequently argue that Francis Fukuyama is an under-acknowledged intellectual godfather to the abundance movement. He coined the term “vetocracy” to describe America’s dysfunctional approach to building things back in 2013, and has written some of the most cogent descriptions of American institutional decay (as well as some of the most elegant defenses of modern liberal democracy) in the decade since. His classic Political Order duo is not just a fascinating tour of societies throughout history but a convincing argument on a universal framework of good governance. This was the first thing I read that really brought the tension between state capacity and the rule of law to life for me – the idea that the norms and institutions that allow the state to function effectively (think the Chinese bureaucracy throughout history, or the Ottoman devshirme) and those that prevent state overreach (the medieval English court system…or the US regulatory system) aren’t just items on a “good government” checklist, but self-sustaining leviathans that require active competition to avoid overgrowth.

Alex Armlovich: How Skyscrapers Can Save The City (2011). Ed Glaeser introduced almost everything YIMBYism is today in this Atlantic feature published years before the first modern YIMBY grassroots groups were even founded. This single paragraph’s math, zoning, and even building code comments are still the core of the YIMBY analysis:

Building up is more costly, especially when elevators start getting involved. And erecting a skyscraper in New York City involves additional costs (site preparation, legal fees, a fancy architect) that can push the price even higher. But many of these are fixed costs that don’t increase with the height of the building. In fact, once you’ve reached the seventh floor or so, building up has its own economic logic, since those fixed costs can be spread over more apartments. Just as the cost of a big factory can be covered by a sufficiently large production run, the cost of site preparation and a hotshot architect can be covered by building up. The actual marginal cost of adding an extra square foot of living space at the top of a skyscraper in New York is typically less than $400. Prices do rise substantially in ultra-tall buildings—say, over 50 stories—but for ordinary skyscrapers, it doesn’t cost more than $500,000 to put up a nice 1,200-square-foot apartment. The land costs something, but in a 40-story building with one 1,200-square-foot unit per floor, each unit is using only 30 square feet of Manhattan—less than a thousandth of an acre. At those heights, the land costs become pretty small. If there were no restrictions on new construction, then prices would eventually come down to somewhere near construction costs, about $500,000 for a new apartment. That’s a lot more than the $210,000 that it costs to put up a 2,500-square-foot house in Houston—but a lot less than the $1 million or more that such an apartment often costs in Manhattan.

Nisha Austin: Around the time when I started getting into ‘Abundance’ (in the way we all think about it here), my mother-in-law gave me Robin Wall Kimmerer’s The Serviceberry. I think she was trying to piece together exactly what I was so excited about and landed on a bestseller about abundance in nature, reciprocity, and community (things I am admittedly also excited about!). Not quite a book about permitting reform or building state capacity, but it turned out to be more relevant than I expected. It named something I’d been circling: that abundance is relational, that all flourishing is mutual, and that wealth means having enough to share. The other book I keep coming back to is Doris Kearns Goodwin’s Team of Rivals and the issues Lincoln grappled with: What do we owe the extraordinary institutions we inherited? How do we hold them together through profound disagreement? How do we keep building when the whole project feels fragile? All of these questions are still relevant. Ultimately, I came to this work from a lot of different directions, but the thread connecting all of it is stewardship of the places, institutions, and systems that let people flourish. Kimmerer reminds me that abundance only matters if it’s shared, while Lincoln reminds me that the answer is never settled. Every generation has to show up and do the work.