The American progressive project has a delivery problem. Healthcare, housing, climate transition, tax administration, benefits enrollment, public transit, child care — all involve promises at scale that the state now struggles to keep. The policy side of the project is lively; the delivery side is a graveyard of failed rollouts, unusable websites, accumulated procurement rules, consultancy capture, and civil-service hiring constraints that together guarantee technology projects in the public sector will be slow, expensive, and bad. This is not an indictment of the public servants doing the work — it is a structural feature of the way the American state has been built, unbuilt, and rebuilt over the past forty years.
The question this topic tracks is how the state might leverage contemporary technology — not least artificial intelligence — for the common good, and what institutional reforms that would require. The answer is not reflexive techno-optimism. The early Internet produced Wikipedia and open science; it also produced surveillance capitalism, platform monopolies, and concentrated fortunes. Which trajectory won was not inevitable; it turned on specific institutional choices — DARPA and NSF funding, open protocols, net neutrality, academic culture, antitrust enforcement — and when those choices weakened, enshittification followed. AI is at a comparable inflection point. Whether it replaces labor or augments it, whether it concentrates wealth or distributes it, whether it equips the state to serve citizens better or equips extractive private interests to capture more of public life, depends on policy and institutional design.
The literature below centers on what might be called the state-capacity left — the argument, developed most clearly by Mariana Mazzucato and Jennifer Pahlka, that progressive ends require a reinvigorated public sector capable of setting missions, managing complex projects, and delivering services at scale. This is distinct from the abundance liberalism associated with Ezra Klein and Derek Thompson, which overlaps on reform targets (housing, clean energy, permitting) but centers regulatory streamlining — giving the left legitimate reasons to be wary, especially in a political moment when "deregulation" is the active project of a hostile administration. Both strands appear below; the distinction matters, and this topic is centered on the first.
Annotated bibliography
The state as active shaper of technology
Mariana Mazzucato, The Entrepreneurial State (2013) and Mission Economy (2021) — the foundational case that the state is not merely a corrector of market failures but an active shaper of markets and a primary driver of transformative innovation. The iPhone, the Internet, GPS, the core of the pharmaceutical pipeline: all built on decades of public investment. Mission Economy pushes the argument forward: what would an ambitious, mission-oriented, climate-serious, technology-competent public sector actually do? Close to the intellectual core of what "state capacity" means in this tradition.
Daron Acemoglu and Simon Johnson, Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity (2023) — the deep historical argument that technology's gains do not automatically flow to workers or to the broader society, and that the trajectory is determined by political and institutional choices. The conclusion is directional rather than pessimistic: AI can augment rather than replace, can distribute rather than concentrate — but not by default. Without deliberate redirection, the current trajectory is labor-replacing and wealth-concentrating. The case for why state-capacity reform is urgent.
The delivery problem
Jennifer Pahlka, Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better (2023) — the most important recent book on the American delivery gap. Why government technology projects fail, what the hidden constraints actually are (procurement rules, civil-service hiring, waterfall mandates, consultancy capture), and what an executable reform program looks like. Pahlka founded Code for America and helped establish the US Digital Service; the analysis is drawn from inside the machine.
Abundance liberalism and the left critique
Ezra Klein and Derek Thompson, Abundance (2025) — a self-described progressive case for a "liberalism that builds": housing, clean energy, transit, digital public services. The book argues that accumulated "process liberalism" — layered environmental review, procurement rules, permitting requirements, community-input regimes — has made the American state unable to deliver material progress, and that progressives should lead this reform rather than concede it to the right. The book has been sharply contested from the left, and the critique has teeth: the framing of regulation as the primary bottleneck overlaps with libertarian talking points on permitting and environmental review; it underweights the distributive and labor-power questions that Piketty, Hacker, and Anderson treat as central; and the "process liberalism" / "everything-bagel liberalism" language reads as anti-left when deployed against coalition members rather than against incumbents and NIMBYs. The timing compounds the problem. A book calling for progressive deregulation lands very differently in a moment when "deregulation" is the Trump administration's active project of hollowing out environmental, labor, consumer, and civil-rights enforcement. Abundance is included here because it is part of the live conversation about state capacity, not because its framework is endorsed — and it is best read against Pahlka and Mazzucato, who reach overlapping conclusions about the need for competent government without treating regulation as the primary obstacle.
Preventing the worst outcomes
Shoshana Zuboff, The Age of Surveillance Capitalism (2019) — the argument that the dominant form of the Internet economy, absent countervailing institutions, is the extraction and commodification of human behavioral data. Zuboff's book is the case for why active institutional design matters; without it, technology tends to concentrate wealth and information in unaccountable private hands. A necessary pessimism to set against the more optimistic builders.
Cory Doctorow, The Internet Con: How to Seize the Means of Computation (2023) and the running project across his essays and novels — the case for interoperability, data portability, and adversarial interoperability as structural anti-monopoly tools. Doctorow's point is that platform monopolies are not inevitable features of digital technology; they are the product of specific legal and regulatory choices (anti-circumvention laws, lax antitrust enforcement) that can be reversed. The most concrete practical-technical side of the state-capacity agenda.
Historical context and political-economy background
See The Welfare State Under Pressure for the longer-run political-economy story of state capacity built (postwar) and then hollowed out (1980s onward). The current delivery problem is partly the accumulated consequence of forty years of deliberately weakening the administrative state.
See Justice as Fairness for the philosophical framework. Rawls's property-owning democracy is, among other things, a regime in which the state has enough capacity to sustain broadly distributed economic power rather than simply redistribute income after the fact. State capacity is the institutional precondition for justice-as-fairness at scale.