Publications, Press & Outside Writing

The City Reporter · September 2026
Kaitlyn Schwanemann's voter guide to the five Charter amendments on the November ballot. On Proposal 2, which targets City contracting delays, it cites my finding that the median City RFP takes 387 days end to end, and quotes me in support of the proposal: shaving 30 days off a process that long shows how much work remains. See the data ↗ · Read at The City Reporter ↗

Vital City · August 2026
Vital City's reform agenda for the Mamdani administration, beyond what a Charter revision can do. Its case for reviving the dormant commission that retires needless reports rests on my count of 2,231 reporting requirements across 183 agencies, nearly half with no record of ever being filed, and on my tracker's count of more than 5,000 standing obligations imposed on agencies since 2014. Read at Vital City ↗

The City Reporter · August 2026
Katie Honan's report on whether the Mamdani administration's efficiency push should take on the City Council's reporting requirements, sized with data from my Legislation Implementation Tracker and quoting me on the disconnect between legislators announcing what they passed and the agencies tasked with carrying it out. Read at The City Reporter ↗

EV Grieve · August 2026
EV Grieve's report on a semi that rode over a bollard turning from Clinton Street onto Houston, blocking buses and pushing drivers into the new bike lane. It uses the crash to revisit my Streetsblog op-ed: Clinton between Delancey and Houston carries roughly 16,000 vehicles a day, 4 to 6 times its neighbors, and my Low Traffic Neighborhood plan would remove the Williamsburg Bridge off-ramp onto it and send uptown drivers to Essex and Allen. Read at EV Grieve ↗

EV Grieve · January 2026
EV Grieve's report on the campaign Zachary Thomas and I ran, as members of Community Board 3's Parks, Recreation, Waterfront & Resiliency Committee, to put a metal tree guard around a Ginkgo on East 7th Street near Avenue B. The morning it ran, the campaign sat a few dollars short of its $2,000 goal. Read at EV Grieve ↗

Searchlight (Citizens Union) · August 2026
Argues that when the Council passes a law it reliably tells an agency what to do and unreliably supplies the money and people to do it, drawing on the 8,185 obligations catalogued by my Legislation Implementation Tracker. Proposes four fixes that preserve oversight: a labor price tag on every mandate, on-the-record funding decisions, default sunsets for new reporting requirements, and a standing docket for the Charter commission that retires dead reports. Read at Searchlight ↗

Streetsblog NYC · July 2026
Clinton Street between Delancey and Houston carries roughly 16,000 vehicles a day, 4 to 6 times its neighbors, because the Williamsburg Bridge off-ramp funnels traffic onto a narrow residential block that has seen over 100 crashes and multiple fatalities in a decade. Argues for a Low Traffic Neighborhood redesign that removes the Manhattan-bound off-ramp and redirects through-traffic to Essex and Allen while preserving local and emergency access, on a block where over 80% of residents don't own a car. Read at Streetsblog ↗

Vital City · June 2026
Examines how nearly all of NYC's post-pandemic job growth has concentrated in a single lower-paying sector, leaving the local economy fragile and the city budget squeezed from both ends. Lays out strategies the Mamdani administration can pursue to diversify the jobs base and revive high-paying employment. Read at Vital City ↗

With Sarah Sachs · Reboot Democracy · June 2026
How Block Party built a searchable archive of Manhattan Community Board 3's 20 years of public meeting records, using semantic search and summarization to make institutional knowledge accessible to residents and journalists. Argues that understanding the civic problem, more than the AI tooling, is what created the value. Read at Reboot Democracy ↗

Manhattan Institute · May 2026
Introduces my NYC Council Fiscal Impacts Tracker, built from the 1,563 fiscal impact statements filed with Council legislation since 2014. Council laws added more than $7.5 billion in net costs, with CityFHEPS, the Streets Plan and school paraprofessional pay mandates alone adding over $6 billion in five years, and 97% of the local laws in the tracker name only the General Fund as their funding source. Argues that the 2024 Charter changes to fiscal impact statements did not go far enough, and that Council members should say where the money for a new mandate will come from. Read at the Manhattan Institute ↗

With Amy X. Chen · Responsive Cities Network Sprint, Harvard Kennedy School · Cambridge, MA · September 2026
Lunch session for the city teams in the Responsive Cities Network, run by Data-Smart City Solutions at the Bloomberg Center for Cities and the Knight Foundation: why community board records are officially transparent but practically opaque, how Block Party turned them into a searchable archive, a demo of mapping resolutions by the places they name, and how New York's 59 community boards compare with the closest local bodies in 10 other cities. View the slides ↗

CivStart State of GovTech 2026 · Washington, DC · September 2026
Main-stage session and live demo at CivStart's State of GovTech: why community board decisions are practically opaque, and how Block Party's transcript archive and resolutions database, now covering 26 boards across Manhattan and Brooklyn, open them up. View the slides ↗

With Sarah Sachs · NYC OMB Institute · July 2026
Walks the NYC Office of Management and Budget Institute through Block Party's community board archive, and the story of scaling the resolution engine from one board to eleven: source adapters, parser iterations, and the corpus as it stands today. View the slides ↗

With Sarah Sachs · NYC School of Data 2026 · March 2026
Explains what community boards are and why their resolutions matter, then shows how Block Party turned two decades of Manhattan Community Board 3 meeting records into a searchable public archive with semantic search and AI summarization. View the slides ↗

With Shigeru Fujita & Valerie Ramey · NBER Working Paper No. 33041 · October 2024
Examines why U.S. unemployment rose only modestly despite a dramatic postwar contraction in federal spending, challenging standard models of aggregate demand. Applies panel data and decomposition methods to historical employment records. Read at NBER ↗

With Enghin Atalay, Shigeru Fujita, Sreyas Mahadevan & Ryan Michaels · Federal Reserve Bank of Philadelphia Research Brief · July 2020
Builds a dataset that scores industries on COVID-19 contact risk (O*NET teamwork, customer-contact and physical-proximity indices, plus the share of workers able to work from home) and matches it to the essential-industry lists in the stay-at-home orders of all 50 states and D.C., then tracks reopening by state and industry from late April to early July 2020. States closed high-contact industries such as leisure and hospitality far more often, but did not hold them back during reopening, so those industries reopened faster than low-contact ones such as manufacturing. Read at the Philadelphia Fed ↗

With Wenli Li · Rutgers Center for Real Estate · September 2018
Uses CoreLogic deed records from 2004 to 2015 to show New Jersey home prices recovering after the Great Recession while homeownership fell from 70.1% in 2005 to 64.0% in 2015. Institutions nearly doubled their share of single-family purchases, from 3% to just under 6%, bought about 20% of foreclosed homes, and grew fastest in high-foreclosure counties such as Salem and Cumberland. Read at Rutgers ↗

Journal of Politics & International Affairs, Ohio State · March 2019
Constructed a cross-national database linking educational attainment to income levels across developing economies; identified non-linear gradients in returns to schooling.

With Peter Slattery · MIT FutureTech · April 2025
Reports from MIT FutureTech's Workshop on the Role of AI in Science: foundation models trained on scientific data that help researchers generate hypotheses and design experiments, Ross King's robot scientist that designs and runs its own experiments, Varun Chandola on the NSF's National AI Research Resource pilot to open AI infrastructure to academics, and FutureTech research on the gap between industry and academic access to large models. Read at MIT FutureTech ↗

With Peter Slattery, Emanuele Del Sozzo & Haoran Lyu · MIT FutureTech · January 2025
Third in a series on the inputs behind AI progress: how compute is measured (FLOPS, MIPS, memory bandwidth), how GPUs, TPUs and parallel computing kept performance rising as Moore's Law slowed, and research finding that compute scaling contributed roughly twice as much as algorithmic progress to effective compute. Read at MIT FutureTech ↗

MIT FutureTech · June 2024
Sets published forecasts of generative AI's economic impact, from under $1 trillion a year in added global GDP (Accenture, Goldman Sachs) to $2.6–4.4 trillion a year (McKinsey), against three benchmarks: 2010s forecasts for AI, current forecasts for technologies such as advanced robotics and genomics, and the measured productivity effects of past technologies from the steam engine to IT. GenAI's forecasts land mid-range against current technologies and past ones, and far below the 2010s AI forecasts on productivity. Read at MIT FutureTech ↗

With Peter Slattery · MIT FutureTech · March 2024
Shows AI models' demand for training data growing faster than the supply of suitable data, with Epoch projecting high-quality language data to run out within a few years, then explains why data is unlikely to stop AI progress: few-shot and transfer learning, data augmentation, and smaller, better-labelled training sets. Read at MIT FutureTech ↗