What we’re reading, May 8, 2026
Are we kind of being pricks?
Happy Friday! Here’s what caught our attention this week:
“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
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
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
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
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
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
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
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
The Inclusive Abundance Institute 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 launched a newsletter from Rep. Josh Harder, covering the caucus’s bipartisan work on housing, infrastructure, and permitting, and featuring a couple of our posts.