Publications

Using four months of high-frequency round-trip-time measurements from 99 residential probes in Chicago, this paper detects latency anomalies shared across devices without relying on network topology, and develops a sampling algorithm that covers 95% of aggregate anomaly impact using less than half the probes.

The push to demonstrate effectiveness has made randomized controlled trials the presumed “gold standard” for nonprofit evaluation. Drawing on extensive interviews with nonprofit managers, evaluators, and foundation officers, this book explains why RCTs are so often the wrong tool for the sector — what happens inside organizations that adopt them, the equity problems they create, and what to do instead.

This paper develops a regionalization method for crowdsourced latency data, grouping geographic areas by similarity in internet performance. Applied to a major U.S. city, it reveals coherent regional patterns that cut across administrative boundaries and correlate with network infrastructure — a new lens for understanding internet inequity.

The first systematic analysis of the FCC Broadband Data Collection challenge process — who disputes reported availability, which areas they target, and how often they succeed — revealing gaps that may limit its ability to correct inaccuracies in federal broadband maps.

A practical guide to the structure, limitations, and analytical possibilities of the FCC Broadband Data Collection — the most comprehensive federal effort to map U.S. internet availability — giving researchers and policymakers the tools to use it responsibly.

Examines whether the use of randomized controlled trials in philanthropic evaluation can address concerns about democratic accountability, finding that while RCTs may improve some accountability dimensions, they introduce new equity concerns that limit their democratic potential.

Internet performance can vary sharply over short distances and times, yet most research aggregates at coarse scales. This paper examines variability at hyperlocal levels — census blocks and below — revealing inequality that larger analyses miss.

The first rigorous side-by-side comparison of the two most common speed-test platforms — Ookla Speedtest and M-Lab’s NDT7 — examining how they differ in methodology, user base, and measured outcomes, and what that means for research and policy.

Argues that benchmark-based measures of internet performance (fixed minimum speeds) miss the equity dimensions of service, and proposes an equity-centered framework that asks who gets what quality of service relative to others in their community.

Develops and evaluates best practices for speed-test data collection — sampling, instrumentation, timing, and reporting — that materially affect the validity of broadband findings.

Documents internet inequality in Chicago across three domains — adoption, affordability, and availability — and shows they do not always align, so closing the divide requires policies targeted at each dimension separately.

Uses a unique dataset of New York City government contracts with nonprofits to study whether social-services funding tracks neighborhood need, distinguishing local organizations serving their immediate area from distributive organizations serving many.

Analyzes the micro-relations of governance across politics and nonprofits, identifying the district-based politician as the key actor linking neighborhood and citywide social organization. Contract-allocation networks reveal two dynamics: exclusive, durable patronage and citywide, short-lived partnership.

Argues that the dominant social-disorganization lens obscures the many ways formal organizations produce and manage urban poverty, and proposes an approach combining urban governance and strategic-action-fields frameworks to open new research directions.

Proposes a governance framework — drawing on political science and organizational sociology — as an alternative to economic models of government-nonprofit relations, situating those relations within the broader structures of urban governance.

Argues that the “what works” movement, as implemented, imposes heavy evaluation burdens on human-service nonprofits without generating knowledge that improves practice, and proposes a social-work-centered alternative prioritizing organizational learning and equity.

Reframes government funding of nonprofits: rather than asking how public funds constrain private action, it asks how government deploys nonprofits to secure a public good. A New York State child-welfare case reveals a deficit model in which nonprofits are deputized to secure rights but under-resourced to do so.

Offers a framework for how organizations matter to the production and amelioration of urban poverty, using the classical concept of integration to take seriously both organizations’ territorial embeddedness and their field-level dynamics.

Using 1997–2001 New York City contract data, finds that social-services contract dollars are positively associated with neighborhood disadvantage overall, though distributive organizations are less likely to be located in the neediest neighborhoods.

Traces how the privatization of welfare-state functions produced an extensive nonprofit service-delivery infrastructure in New York City, and how contracting-out from the 1970s–1990s institutionalized a dense system of community organizations intertwined with local government.

Argues that urban sociology — both Marxian political economy and Chicago-school work — treats organizations as derivative rather than productive of urban life, and proposes taking organizations seriously as active producers of urban outcomes.

Through ethnographic fieldwork at eight community-based organizations in Williamsburg and Bushwick, Marwell shows how the relationships these groups form with larger political and economic institutions outside the neighborhood shape the lives of the poor. The book widens the lens of urban poverty research from individuals and families to the organizations that collectively drive urban life.

Theorizes mixed-form markets in which nonprofits and for-profits compete and collaborate, drawing on organization theory and economic sociology to explain why such markets emerge and what they mean for service quality, access, and equity.

Shows how publicly funded nonprofit community-based organizations generate greater contract revenue by adding electoral politics to their service and community-building roles — producing a new organizational form, the machine-politics CBO.