> For the complete documentation index, see [llms.txt](https://cultural-physics.gitbook.io/n/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cultural-physics.gitbook.io/n/applications-per-discipline/public-policy-governance.md). # Public Policy / Governance ### Overview Public policy is the institutional practice of designing the membranes, nodes, and gates that shape the perceptual field of an entire society. Unlike HR (organization‑scale) or L\&D (workforce‑scale), public policy operates at the scale of polities—cities, regions, nations, and increasingly the planetary membrane (p. 328–329). A tax code, a public health regulation, a zoning ordinance, a privacy law—each is an amplitude field intervention that shapes millions of collapses across generations. In Cultural Physics terms, public policy is the engineering of **societal amplitude fields**—structured distributions of permissible, encouraged, and prohibited meaning that guide how citizens perceive, behave, and co‑exist. The policy designer is a **Gatekeeper** (controlling access to resources, rights, and protections), a **Stabilizer** (embedding governance patterns through laws and institutions), and, when legitimacy fractures, a **Repairer** (re‑entraining public trust after rupture). This research brief integrates behavioural public policy, implementation science, algorithmic governance, administrative law, public value management, and contemporary governance practice into the Cultural Physics framework. *** ### Part 1: Core Concepts – What Public Policy Actually Does #### 1.1 Policy as Field Engineering, Not Rule‑Setting Standard policy frameworks define public policy as authoritative decisions by governments that allocate values, constrain behaviour, or pursue collective goals. Cultural Physics reveals a deeper truth: **policy shapes the perceptual field before it shapes behaviour**. A speed limit is not primarily a rule about driving; it is an amplitude peak in the civic field—a signal that “here, safety is prioritized over speed.” An income tax is not primarily a revenue instrument; it is a membrane that redistributes the felt weight of collective obligation. **Cultural Physics translation:** Policy is not merely a set of rules; it is a **perceptual infrastructure**. It defines what collapses are possible (what is legal), what collapses are encouraged (what is subsidized, promoted, or default‑designed), and what collapses are foreclosed (what is banned, taxed out of reach, or socially stigmatized through law). #### 1.2 The Policy Value Chain Through a Cultural Physics Lens | Policy Function | Traditional Framing | Cultural Physics Framing | | ---------------------------- | ---------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **Agenda Setting** | Identifying problems requiring state action | Field reading: detecting decoherence, identifying nodes of unaddressed social charge | | **Policy Formulation** | Designing interventions to address problems | Amplitude encoding: embedding predictive templates (mandates, incentives, defaults) into legal and programmatic form | | **Adoption/Legitimation** | Gaining authorization through legislative or executive processes | Membrane ratification: the governing body formally approves the field intervention | | **Implementation** | Executing the policy through public administration | Field activation: translating legal text into operational amplitude fields that citizens and administrators actually collapse | | **Evaluation & Learning** | Assessing effectiveness, efficiency, equity | Field diagnostics: measuring coherence, resonance, gravity, and decoherence patterns; detecting where policy amplitude peaks are not being collapsed as intended | | **Adaptation & Termination** | Modifying or ending policies based on performance | Hatcher Mechanic (p. 127–130): bending the field without breaking; removing nodes that have lost charge | #### 1.3 The Distinctive Crisis of Public Policy Governance today is marked by three intersecting crises that are, at root, crises of field coherence: * **Legitimation crisis:** Trust in public institutions has collapsed. Across OECD countries, the share of people who trust their national government has fallen to historic lows. As the Brookings Institution notes, mistrust in government data is growing, even as governments collect more of it. This is **membrane failure**: citizens no longer believe the gate is legitimate. * **Implementation crisis:** Policies are passed, but their intended collapses do not occur. The “commitment–implementation gap” has been documented across sectors. The EPA‘s Acid Rain Program case shows that even well‑designed policies can be undermined by administrative overload and “policy triage”—the deliberate prioritization of some implementation tasks over others when capacity is limited. This is **field drift**: the policy amplitude peak is encoded in law but never materializes in practice. * **Algorithmic governance crisis:** AI systems are rapidly being integrated into public administration without adequate transparency, accountability, or public consent. As Dedyaev‘s study documents, by 2025 all U.S. states had adopted some form of AI governance measures, yet a consistent framework for accountability remains absent. This is **basis hijack at population scale**: citizens are subject to algorithmic decisions whose measurement frames are opaque. *** ### Part 2: Behavioural Public Policy – Nudging as Amplitude Shaping #### 2.1 The Rise of the “Nudge Unit” The most influential example of Cultural Physics–style field engineering in governance is the Behavioural Insights Team (BIT) in the United Kingdom, known colloquially as the “Nudge Unit.” Created in 2010 within the UK Cabinet Office, BIT was the world‘s first government team dedicated to applying behavioural science to public policy. Its approach: use subtle, low‑cost, non‑coercive interventions—nudges—to help citizens make choices aligned with their long‑term interests and the public good. The classic example: a letter to late tax‑payers was rephrased to say, “Most people in your town have already paid.” Compliance soared without new laws or fines. By 2014, BIT had become a not‑for‑profit consulting company, and similar teams have since emerged across the globe—in the US (Office of Evaluation Sciences), Europe, Japan, Singapore, Saudi Arabia, Peru, Australia, New Zealand, and many other countries. **Cultural Physics translation:** The nudge unit is an **institutionalized field engineering practice**. It treats policies not as commands but as amplitude fields; citizens are not passive subjects but active Observers whose collapses can be gently shaped through choice architecture. Nudges are low‑mass, high‑precision amplitude injections—small peaks that bias collapse without brute force. #### 2.2 Effectiveness and Limits A meta‑analysis by DellaVigna and Linos (2020) examined 165 evaluations affecting over 24 million people across two leading behavioural economics units, including BIT and the US Office of Evaluation Sciences. The analysis found a clear positive impact of nudges, estimating an average behavioural improvement of more than 8%. Such low‑cost interventions can yield high benefits for both individuals and society. But critiques are substantial. Critics charge that nudges can undermine individual autonomy, involve manipulation that takes advantage of non‑conscious cognitive processes, and erode the accountability that should govern public authority. Defenders respond that many public policies are already paternalistic; nudges are simply “soft” instruments with less intrusive intervention than traditional regulation. A more fundamental critique: nudges often produce non‑persistent effects, compensatory negative spillovers, or psychological reactance. Banerjee, Grüne‑Yanoff, John, and Moseley argue that behavioural public policy must promote **agency**—people‘s ability to form intentions and act on them freely. They advocate for agency‑enhancing approaches: *boosts* (competence‑building), *debiasing* (reducing automatic responses), and *nudge+* (enabling citizens to think alongside nudges and evaluate them transparently). These approaches aim to create longer‑lasting, more meaningful behaviour change by leaving citizens “in the light”. **Cultural Physics translation:** The critique of nudges as paternalistic manipulation is a critique of **collapse without consent**. An ethical nudge discloses its basis; it respects the citizen‘s role as an Observer with the right to know why the field is shaped as it is. Nudge+ is an **amplitude transparency protocol**—a design that makes the shaping visible, allowing the citizen to choose whether to align or resist. #### 2.3 Scalability and Replicability The effectiveness of behavioural interventions can differ dramatically across contexts. Interventions that succeed in one setting or for one group may not apply to other situations. The NASEM (2023) report concludes that “the field of behavioural economics has not yet produced generalizable and implementable practice guidance”. This is a classic **field transfer problem**: the amplitude peaks that collapse effectively in one field may be inert or disruptive in another. *** ### Part 3: Citizen Participation – When the Membrane Opens #### 3.1 The Crisis of Representative Democracy Traditional representative governance operates as a **gatekeeping membrane**: citizens collapse their preferences into votes at periodic intervals; representatives then collapse those votes into policy. But this membrane is leaking. As the New America report documents, in the face of deepening political dysfunction and disaffection, fewer than half of U.S. states provide access to usable ballot initiative processes, leaving millions without a direct channel to shape public policy. Voters‘ distrust and cynicism deepen when there is no other path to change that aligns with public preferences. **Cultural Physics translation:** The representative membrane has lost permeability without losing brittleness. Citizens cannot easily influence the field between elections, yet the field is constantly shifting. The result is **field capture by extraction**—policy amplitude peaks that serve incumbent interests, not citizen collapses. #### 3.2 Democratic Innovations as Field Re‑permeabilization Democratic innovations—citizens‘ assemblies, participatory budgeting, civic assemblies, public deliberation—are **field repair mechanisms**. They create temporary apertures through which citizens can directly influence policy collapses. Research has found that civic assemblies have positive impacts on participants: they are more likely to change policy attitudes, and participation positively affects political knowledge, internal efficacy, and reasoning skills. Around the world, approaches such as citizens‘ assemblies, participatory budgeting, and public deliberation provide opportunities for citizens to help shape laws and policies, offering a promising response to the disconnect between public will and political action. The OECD notes that an open and protected civic space is a prerequisite to enabling meaningful participation. **Cultural Physics translation:** Democratic innovations are **gates without gatekeepers**—thresholds where citizens, not officials, control the measurement basis. They open the membrane from the inside, allowing the field to be re‑collapsed by those who inhabit it. #### 3.3 Data‑Driven Participatory Governance The Brookings Institution has documented how cities can rebuild public trust by creating data systems that are transparent, participatory, and grounded in community priorities. Effective dashboards combine user‑friendly tools, public training, and staff who can teach, ensuring residents can access and use city data meaningfully. Engaging elected officials early, establishing advisory boards, and evaluating community outcomes are key to building a culture of data‑driven policymaking. **Cultural Physics translation:** Participatory data governance is **membrane transparency**. Citizens are not merely the subjects of data collapse; they are co‑collapsers, able to see the data basis on which policy decisions rest. *** ### Part 4: Algorithmic Governance – AI as Automated Gatekeeping #### 4.1 The Scale of Algorithmic Integration The public sector is undergoing rapid algorithmization. Public‑sector investment in AI is projected to grow 19% annually between 2022 and 2027—faster than any other sector. According to Deloitte, task automation at the federal level alone could eliminate between 96.7 million and 1.2 billion working hours annually, corresponding to estimated savings of USD 3.3 to 41.1 billion. Within a single year, the number of officially registered AI use cases across major U.S. federal agencies more than doubled. **Cultural Physics translation:** AI is being deployed as a **high‑speed, high‑volume field actuator**—an automated Gatekeeper that can manage millions of individual collapses (benefits claims, tax filings, regulatory compliance checks) in real time, without human intervention. #### 4.2 Two Modes of Algorithmic Governance Dedyaev‘s comparative analysis of thirty AI implementation cases across federal, state, and municipal levels in the United States identifies two broad modes of algorithmic governance: | Mode | Definition | Cultural Physics Translation | | ---------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------- | | **Control‑oriented systems** | AI used for surveillance, enforcement, regulatory oversight, high‑stakes gatekeeping (e.g., welfare administration, public health eligibility) | *Membrane enforcement*—AI decides who is inside (eligible) and who is outside (ineligible), often without transparency | | **Support‑oriented systems** | AI used to streamline operations, improve direct interactions, provide decision support | *Field smoothing*—AI assists human collapsers, reducing friction but not replacing judgment | The findings reveal a clear functional differentiation across levels of government. Federal AI is most often institutionalized for high‑stakes control. State governments occupy a middle ground where AI frequently combines support with algorithmic gatekeeping. Municipal governments tend to deploy AI in more pragmatic, service‑oriented ways. The character, function, and risks of AI in the public sector are fundamentally shaped by the level of governance at which these systems are deployed. **Cultural Physics translation:** The same AI technology operates as a **different field actuator** depending on its governance context. At the federal level, AI is a hard gate; at the municipal level, it is a soft path. This suggests that the *institutional field* (the rules, norms, and capacities of the administering agency) is as important as the algorithm itself. #### 4.3 The Accountability Gap The rapid deployment of AI in public administration has outstripped legal and governance frameworks. Scholars Mark and Morison (2025) examine how key jurisdictions—the US, EU, and China—frame the risks of algorithmic decision‑making. They find that despite divergent framings (US: “innovation gap,” EU: “trust deficit,” China: “stability risk”), there has been a post‑2024 trend toward regulatory softening that privileges innovation over precautionary safeguards. The accountability gap is especially acute in benefits adjudication. Stanford‘s “Evaluation as Due Process” project documents how merit staffing requirements, designed to ensure procedural fairness, have also impeded responsible innovation in automated decision systems. **Cultural Physics translation:** The accountability gap is a **measurement basis conflict**. The algorithm collapses on one basis (efficiency, cost reduction, statistical consistency). The citizen collapses on another basis (fairness, dignity, individual circumstance). Neither collapse is illegitimate, but the field has no mechanism to reconcile them. The result is not coherence but **field fragmentation**—two incommensurable realities. #### 4.4 Algorithmic Governance Legislation In response, legislative action is accelerating. By 2025, all 50 U.S. states had introduced various AI‑related legislation. The **Sectoral AI Governance Act of 2026**, introduced by Rep. Sara Jacobs, would give federal agencies a consistent framework for writing and issuing rules when AI is likely to materially contribute to violations of existing federal laws. As Jacobs stated: “Federal laws shouldn‘t become optional just because technology is new. We can’t let the American people‘s rights and protections become meaningless the moment a company outsources a decision to an algorithm”. **Cultural Physics translation:** The Sectoral AI Governance Act is an attempt to **re‑assert the measurement basis**. It re‑states that public values (fair housing, equal credit, due process) are the correct basis for collapse—even when those collapses are mediated by AI. *** ### Part 5: The Policy Implementation Gap – When Amplitude Peaks Attenuate #### 5.1 The Commitment–Implementation Gap Policies are often designed sensibly but fail to produce intended outcomes because of ineffective implementation. As a 2024 Oxford study of education policy found, “most of the policy‑practice gap can be explained by ineffective service delivery”—not by naive design. The Brookings Institution provides a systematic account of the causes of policy failures and implementation gaps. Key causes include: limited administrative capacity, fragmented responsibilities, competing priorities, and the simple fact that more policies are layered onto overstretched administrations than are repealed. **Cultural Physics translation:** The implementation gap is **gravity attenuation without active maintenance**. The policy amplitude peak is encoded in law (the policy is “passed”), but the **ritual density** required to maintain that peak—inspections, training, compliance monitoring, public communication—is absent or under‑resourced. The policy‘s amplitude decays back toward the baseline before it ever produces a collapse in the field. #### 5.2 Policy Triage and Administrative Overload The EPA‘s Acid Rain Program case study is a sobering lesson. The 1990s cap‑and‑trade program was a landmark environmental initiative, widely praised. But as the EPA took on this major new responsibility, it quietly scaled back inspections and enforcement in unrelated areas. This was not mismanagement; it was **policy triage**—the deliberate prioritization of some implementation tasks over others when administrative capacity is limited. **Cultural Physics translation:** Policy triage is **field decoherence through resource starvation**. The field cannot maintain all its amplitude peaks simultaneously. Administrators must choose which peaks to reinforce and which to allow to decay. This is not a failure of design; it is the tragic necessity of limited capacity. The solution is not to design better policies but to **increase the field’s maintenance resources**. *** ### Part 6: Public Value Management – The Normative Basis of Collapse #### 6.1 Public Value Theory Mark H. Moore‘s **Public Value Theory** (1995) defines government success by outcomes the public values as legitimate, equitable, and effective. Public value governance extends this view to cross‑sector collaboration. Public managers, on behalf of the public, should actively strive to generate greater public value, as managers in the private sector seek greater private business value. **Cultural Physics translation:** Public value is the **normative amplitude field**—the distribution of social meaning weighted by what the public deems legitimate. A policy that produces efficient outcomes but is perceived as illegitimate has low public value; its amplitude peak is high in the administrative field but low in the civic field. #### 6.2 Public Value Misalignment in Algorithmic Tools Ingrams and Giest (2025) examine how management inputs produce—or fail to produce—public value in algorithmic decision‑making. Using the case of the U.S. public comments platform Regulations.gov, they show how operational capacity and external support affect strategic considerations of managers. **Public value misalignment** occurs when the values embedded in the algorithm (efficiency, consistency) conflict with the values of the public (fairness, voice, accountability). **Cultural Physics translation:** Public value misalignment is **basis conflict between different levels of the field**. The algorithm‘s collapse basis is technical (statistical, cost‑effective). The citizen‘s collapse basis is moral (just, respectful). The field has no mechanism to adjudicate between them—so it fragments. *** ### Part 7: The Industry in Transformation – Policy in the Age of AI #### 7.1 The New Policy Disciplines Policy design is developing new disciplines that mirror Cultural Physics concerns: | Discipline | Focus | Cultural Physics Translation | | ----------------------------------- | ----------------------------------------------------------------------------- | ----------------------------------------------- | | **Behavioural Public Policy (BPP)** | Applying behavioural science to policy design and implementation | *Amplitude shaping through choice architecture* | | **Algorithmic Governance** | Designing governance frameworks for AI and automated decision systems | *Membrane enforcement by code* | | **Implementation Science** | Systematic study of how to close the gap between policy intent and outcomes | *Active maintenance of policy amplitude peaks* | | **Public Value Management (PVM)** | Creating outcomes that citizens value as legitimate, equitable, and effective | *Normative basis setting for the policy field* | #### 7.2 The Role of Policy Labs Policy innovation labs have emerged as spaces for co‑designing policy with citizens, using behavioural insights, rapid prototyping, and iterative testing. The OECD Observatory of Public Sector Innovation (OPSI) documents numerous behavioural insights projects across local governments—from increasing tax compliance to reducing littering to improving public participation. **Cultural Physics translation:** Policy labs are **temporary field apertures**—safe spaces where policy amplitude peaks can be prototyped, tested, and refined before being deployed at scale. They reduce the risk of large‑scale field decoherence by allowing small‑scale calibration. #### 7.3 The Policy–Administration Distinction in an Algorithmic Age The traditional distinction between policy (setting the ends) and administration (choosing the means) is eroding as AI systems embed policy choices into code. Stanford‘s “Evaluation as Due Process” project traces how merit staffing requirements, designed to ensure procedural fairness, have inadvertently impeded responsible innovation in automated decision systems. Bureaucratic rules that were designed for a human‑mediated world are poorly adapted to an algorithmic one. **Cultural Physics translation:** The collapse of the policy‑administration distinction means that **code is policy**. The algorithm‘s collapse basis *is* the operationalized policy. If the basis is not debated and consented to by the public, then policy is being made without the membrane‘s consent. *** ### Part 8: Ethical Dimensions – The Responsibility to Govern Collapse #### 8.1 Nudge and the Consent Problem The most persistent ethical criticism of behavioural public policy is that nudges operate without transparency and may undermine individual autonomy. As the CNMC report notes, because nudges “subtly influence behaviour and often take advantage of cognitive processes that are not necessarily conscious,” individuals are less likely to recognize them than conventional instruments, eroding the principle of accountability that should govern public authorities. **Cultural Physics translation:** The consent problem is **collapse without awareness**. The citizen is being guided toward a collapse they cannot see happening. Ethical behavioural policy must disclose its basis—making the nudge visible and giving citizens the option to opt out (the “nudge+” approach). #### 8.2 Algorithmic Gatekeeping and Due Process Algorithmic gatekeeping systems—used in welfare administration, public health eligibility, benefits adjudication—raise profound due process concerns. When an algorithm determines that a citizen is ineligible for benefits, by what right does it do so? What recourse does the citizen have? As the Sectoral AI Governance Act recognizes, existing federal laws (the Fair Housing Act, the Equal Credit Opportunity Act, the Administrative Procedure Act) already provide the basis for accountability—but agencies lack clear authority to apply them to algorithmic decisions. **Cultural Physics translation:** Algorithmic gatekeeping is **gate without review**. The measurement basis is hidden; the gatekeeper cannot be questioned; the field has no appeal mechanism. Re‑establishing due process in the algorithmic age requires making the algorithm‘s basis transparent and subject to challenge. #### 8.3 Participatory and Deliberative Governance The most robust ethical response to the crises of legitimation, implementation, and algorithmic governance is **participatory and deliberative democracy**. As the New America report documents, direct democracy mechanisms (ballot initiatives, citizens‘ assemblies) can inoculate democratic systems against capture by elites. Citizens‘ assemblies provide the opportunity to deliberate and discuss, providing clear and balanced input that makes policies more representative and legitimate. **Cultural Physics translation:** Participatory governance is **field re‑permeabilization**. It opens the membrane from the inside, allowing citizens to shape the basis on which policies collapse. It is the antidote to gatekeeping capture. #### 8.4 Algorithmic Transparency as Basis Disclosure The core ethical demand of algorithmic governance is **transparency**: citizens have a right to know the basis on which algorithms make decisions about their lives. This is not merely a technical requirement; it is a **normative requirement** of democratic legitimacy. As the Brookings research on data dashboards shows, transparency alone is insufficient; citizens also need training, access, and meaningful opportunities to shape how data is collected and used. **Cultural Physics translation:** Algorithmic transparency is **membrane visibility**. It makes the gate‘s basis visible to those subject to it. Without visibility, the gate is not a legitimate threshold; it is a trap. *** ### Part 9: Research Agenda for Cultural Physics – Public Policy / Governance | Research Area | Questions | Methods | | ---------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------------------------- | | **Nudge as amplitude field** | How do different nudge types (defaults, social norms, framing) affect collapse speed, depth, and persistence? What is the dose–response relationship between nudge intensity and behavioural change? | Randomised controlled trials with varying nudge intensities; longitudinal tracking of collapse persistence | | **Policy implementation as field maintenance** | What administrative rituals (inspections, training, reporting) are necessary to maintain policy amplitude peaks? What is the minimum ritual density for policy gravity? | Comparative case studies of successful vs. failed implementation; administrative load analysis | | **Algorithmic basis disclosure** | What level of transparency is sufficient for citizens to contest algorithmic decisions? Can basis disclosure be automated? | User testing with varying transparency conditions; legal analysis of due process requirements | | **Participatory governance as field repair** | Under what conditions do democratic innovations (citizens‘ assemblies, participatory budgeting) restore field coherence? Do they produce measurable increases in public trust? | Pre‑/post‑intervention trust surveys; comparative analysis of participating vs. non‑participating citizens | | **Algorithmic governance at different governance levels** | How does the function and risk of AI in public policy differ across federal, state, and municipal levels? What governance structures are appropriate for each? | Dedyaev‑style comparative case analysis; institutional ethnography | | **Public value misalignment diagnostics** | Can we measure misalignment between algorithmic basis and public value basis? What are the early warning signs of basis conflict? | Q‑methodology; public value surveys; algorithmic audit | | **Regulatory softening and the innovation–precaution trade‑off** | Has post‑2024 AI governance converged toward regulatory softening? What are the field consequences of prioritising innovation over precaution? | Cross‑jurisdictional regulatory analysis; longitudinal tracking of algorithmic harms | *** ### Summary: Public Policy / Governance in One Page \| **Core Mechanic** | Policy is the institutional engineering of societal amplitude fields; it shapes which collapses are possible, encouraged, or prohibited | | **Behavioural Public Policy** | Nudge units (UK BIT, US OES) use low‑cost, non‑coercive interventions to shape collapse; nudges produce 8%+ average behavioural improvement but raise autonomy concerns; agency‑enhancing approaches (nudge+, boosts) offer ethical alternatives | | **Citizen Participation** | Democratic innovations (ballot initiatives, citizens‘ assemblies, participatory budgeting) re‑permeabilize the membrane; participants show increased policy knowledge, efficacy, and reasoning skills | | **Algorithmic Governance** | AI is being deployed at scale (19% annual growth), with control‑oriented systems at federal level and support‑oriented at municipal level; accountability gap is acute; Sectoral AI Governance Act (2026) seeks to re‑assert existing legal bases | | **Implementation Gap** | Policies fail not because of design but because of attenuated amplitude peaks; administrative overload forces policy triage; closure requires adequate field maintenance resources | | **Public Value Management** | Public value (Moore) is the normative amplitude field; misalignment occurs when algorithmic bases (efficiency) conflict with public bases (fairness, voice) | | **Ethical Demands** | Transparency as basis disclosure; algorithmic due process; participatory field re‑permeabilization; nudge+ as consent‑respecting amplitude shaping | | **Key Scholars/Practitioners** | Thaler & Sunstein (nudge), Halpern (BIT), DellaVigna & Linos (nudge meta‑analysis), Banerjee et al. (agency‑enhancing BPP), Moore (public value), Dedyaev (algorithmic governance levels), Mark & Morison (AI regulation), Jacobs (Sectoral AI Governance Act) | *** ### Plain Text Source List (Public Policy / Governance) Banerjee, S., Grüne-Yanoff, T., John, P., & Moseley, A. (2024). It‘s time we put agency into Behavioural Public Policy. Behavioural Public Policy, 8(4), 789–806. Brookings Institution. (2026). How citizens and local governments can advance data-driven policymaking together. Brookings. Dedyaev, M. (2026). Algorithmic Governance in the United States: A Multi-Level Case Analysis of AI Deployment Across Federal, State, and Municipal Authorities. arXiv:2602.08728. DellaVigna, S., & Linos, E. (2020). RCTs to Scale: Comprehensive Evidence from Two Nudge Units. NBER Working Paper. European Union Policy Lab. (2026). Turning policy into real‑world action: exploring the implementation gap. Ingrams, A., & Giest, S. (2025). Public value misalignment and adoption of algorithmic tools: the influence of operational capacity and external support. Public Management Review. Kaplaner, C., Knill, C., & Steinebach, Y. (2025). The EPA‘s Acid Rain Program shows how more policy can often mean less enforcement. LSE US Politics and Policy. Mark, D., & Morison, J. (2025). Governing AI Decision-Making: Balancing Innovation and Accountability. Politics and Governance, 13, Article 10245. Moore, M. H. (1995). Creating Public Value: Strategic Management in Government. Harvard University Press. National Academies of Sciences, Engineering, and Medicine (NASEM). (2023). Behavioral Economics: Policy Impact and Future Directions. New America. (2025). Expanding Citizen-Led Policymaking in the Twenty-First Century. OECD. (n.d.). Adopting a systems approach to citizen participation. OECD Observatory of Public Sector Innovation (OPSI). (2026). Behavioral Interventions in Local Government: Increasing the Efficiency of Local Public Policies. Sectoral AI Governance Act of 2026. (2026). U.S. Congress. Stanford Regulatory Lab. (n.d.). Evaluation as Due Process: Civil Service in an Automated Age. The Decision Lab. (n.d.). Nudge Team. UK Behavioural Insights Team. (2010–present).