Continuous Process Effort

An Operating System for Federal Policy Iteration

Part of the We The People Platform · v1.0 · May 13, 2026

Acronyms referenced in this document: Bureau of Labor Statistics (BLS).

Better results come from better decisions. Better decisions come from better knowledge. Better knowledge comes from a continuous, iterative discipline of understanding what is true, testing what could be done, measuring what actually happens, and refining what to do next based on what was learned. This document describes that discipline and proposes its adoption as the default operating system across federal policy work.

The methodology has many names — Plan-Do-Study-Act, the Deming cycle, rapid-cycle evaluation, evidence-based policymaking, kaizen — and is well-established across industrial quality management, healthcare delivery, and a growing number of federal agencies. What is novel in this platform's framing is the proposal that the methodology should be the default operating system across every policy domain, not an opt-in available to specific programs that happen to fund evaluation.

1. What Continuous Process Effort Is

LINEAGE

The methodology has industrial-quality-management origins. Walter Shewhart, a statistician at Bell Telephone Laboratories in the 1920s, developed the Plan-Do-Check-Act cycle as a framework for systematic quality improvement. W. Edwards Deming, working with Japanese industrial recovery efforts in the 1950s, refined the cycle (later renamed Plan-Do-Study-Act to emphasize learning over verification) and turned it into a foundation of the Toyota Production System and, through that lineage, into the Lean and Six Sigma movements that now permeate manufacturing, healthcare, and software engineering. The methodology is empirically successful across more than a century of application; the question is how to translate it into federal policy practice.

THE CORE IDEA

Every problem worth addressing in policy is one where the current state of knowledge is incomplete. The best available evidence today suggests one course of action; that action, taken at scale, produces real-world data about what works and what does not; that data updates the evidence base; the updated evidence base supports a refined course of action. The methodology is the discipline of running this loop deliberately rather than letting it happen by accident, with documentation at each step so that what is learned does not get lost.

Each pass through the loop produces three outputs: a refined intervention (what to do next), a more complete picture of the problem (what is known now that was not known before), and an explicit record of the reasoning that connects the two. The third output is what distinguishes continuous process effort from ad-hoc trial and error. Without the documentation, the knowledge gets lost when personnel change, when administrations change, when funding cycles end. With it, the knowledge compounds across successive teams.

2. The Six Steps

The methodology has six steps that repeat indefinitely. They are described below as discrete; in practice they overlap and inform each other continuously.

Step 1 — Understand the issue.

Begin by mapping the problem against current available knowledge. What is known with confidence? What is contested? What is unknown? Which data sources exist; which are reliable; which are stale; which are missing entirely? The output is a clear, written statement of the problem and the evidentiary state, with explicit acknowledgment of gaps. This step is rarely given enough time in conventional policy work; the temptation is to jump to proposed solutions before the problem is well-characterized.

Step 2 — Identify data gaps.

Determine what data would need to exist to make a higher-quality decision than is currently possible. Some gaps can be filled by new data collection; some require new research; some are gaps in the population reached by existing data (the survey reaches one demographic well and another poorly). The output is a prioritized list of data needs with cost estimates and responsible parties. Federal statistical agencies (BLS, Census, BEA, NCHS) are existing infrastructure for much of this; the gap is often coordination across agencies, not creation of new capacity.

Step 3 — Design a researched solution.

Use the available evidence to propose an intervention. The intervention should be specific enough to test, ethical and fair by the standards of the affected population, and proportional to the problem. Include the expected outcomes (what success would look like) and the metrics that would measure them. The output is an intervention design with explicit hypotheses about what will happen and why.

Step 4 — Test.

Run the intervention at a scope where its effects can be observed and measured. The scope can be a pilot region, a subset of beneficiaries, a phased rollout. The key requirement is that the test be designed to produce interpretable data — not so small that random variation overwhelms the signal, not so large that course corrections become politically impossible. Federal agencies have well-developed authorities for pilot programs and demonstration projects; the methodology requires using those authorities deliberately for learning, not just for compliance.

Step 5 — Analyze the results.

Compare actual outcomes to expected outcomes. Was the hypothesis confirmed, partially confirmed, or refuted? What was unanticipated — both desirable outcomes that exceeded expectations and undesirable outcomes that should have been foreseen? What does the data say about the underlying mechanism? The output is an analytic report that connects the test's data to the intervention's design, identifying both what worked and what was learned.

Step 6 — Refine and repeat.

Use what was learned to design the next iteration. The result of the previous test becomes the data input to the next iteration's understand-the-issue step. The methodology is recursive: the documentation produced in this iteration is the starting evidence for the next. Over many iterations, the cumulative knowledge becomes substantial in a way that single-shot evaluations never produce.

3. Existing Federal Infrastructure

The proposal in this document is not that the methodology should be invented from scratch in U.S. government. Substantial infrastructure already exists. The proposal is that the methodology should become the default rather than the exception.

Foundations for Evidence-Based Policymaking Act of 2018 (P.L. 115-435)

The Evidence Act is the legal foundation. It requires each federal agency to designate a Chief Evaluation Officer, establish a multi-year learning agenda, and develop an annual evaluation plan. It establishes the Standards and Practices for Evaluation in Federal Agencies. It creates the Advisory Committee on Data for Evidence Building. The Act provides the statutory authority for continuous process effort; what is missing is the operational adoption.

Office of Evaluation Sciences (OES)

Created within the General Services Administration in 2015 (then known as the Social and Behavioral Sciences Team), the OES partners with federal agencies on rapid-cycle evaluation. The OES has run more than two hundred evaluations across thirty agencies, demonstrating that the methodology can be applied at federal scale. Its scaling constraint is staff capacity, not methodological maturity.

What Works Clearinghouse

Operated by the Department of Education's Institute of Education Sciences, the What Works Clearinghouse is the largest sustained federal example of evidence-based policymaking applied to a single domain. It reviews education research for methodological quality, summarizes findings, and publishes intervention reports that practitioners use to select evidence-supported approaches. The model is replicable; analogous clearinghouses exist in narrower form for criminal justice (CrimeSolutions.gov), child welfare (Title IV-E Prevention Services Clearinghouse), and labor (Pathways to Work).

OMB Evidence Agenda and Performance Improvement

The Office of Management and Budget has issued guidance under Circular A-11 and the related Evidence Build-Out memoranda directing agencies on evidence and evaluation requirements. The President's Management Agenda includes evidence-based decision making as a cross-agency priority. The OMB is the natural coordinator if the methodology becomes the default operating system.

SYNTHESIS

The statutory authority exists. The institutional capacity exists in early form. What remains is to make the methodology the default rather than the exception. Every program subject to iteration, every iteration documented, every decision traceable to the evidence that informed it. This is a scaling and culture change, not a from-scratch invention.

4. How This Platform Already Uses the Methodology

The methodology proposed for the country is the methodology this platform was built under. Documenting that worked record is part of the case for the proposal.

Iterative Hardening Process

The platform's Iterative Hardening Process documentation captures the iteration discipline applied to this content-production effort. Each version is numbered (currently 3.7.68); each version has a changelog describing the user request, the diagnosis, the fix, the verification approach, and the audit result; each version is associated with one or more sections of the Open Issues Registry that document the architectural decisions made during that iteration. The result is a worked record of more than 130 consecutive iterations applied to a single multi-year project.

Open Issues Registry (OIR)

The OIR is a 174-section document tracking every architectural decision the platform has made, cross-referenced to the analysis that produced the decision. Each section identifies the issue addressed, the diagnostic work, the resolution, and the iteration in which the resolution shipped. The v3.1.2 Mitigated/Status binary discipline (introduced after a methodology review) separates author-responsibility (Mitigated: Y/N) from content-completeness (Status: OPEN/CLOSED), producing transparent tracking of where work remains and what has been resolved.

Audit script as automated check

Every iteration runs through an audit script that performs more than a dozen automated checks: language consistency, CSS insertion verification, catalog field completeness (added as the simulate-renderCard check in v3.7.61 after a runtime regression), per-page metadata consistency, findability of every shipped document. The audit script is itself iterated: each new class of issue gets a new check after the issue is diagnosed, so the same class of issue cannot recur silently in a future iteration. This is the methodology applied to the methodology — continuous process effort applied to the quality-control system itself.

Diagnose-then-verify discipline

The current iteration narrative requires that every fix follow a strict discipline: identify the issue, diagnose the specific cause (not just the symptom), propose a fix based on the diagnosis, apply the fix, verify the fix actually resolved the underlying issue before claiming success. The discipline was established in response to a user complaint that earlier iterations had claimed fixes that did not fully work. The verification approach has been extended over time from text-grep-presence checks (does the patch appear in the file) to runtime simulation (does the patch actually produce the intended behavior). Each extension came from a specific failure mode the previous level of verification had missed.

5. Application Across the Twelve Pillars

The methodology applies to every policy domain. A sketch of what continuous process effort looks like for each pillar:

• Pillar 1 (Community Contribution Plan): pilot contribution-rate variants and progressivity-band placements; measure participation rates, contributory adequacy, and labor-market response; refine the rates and bands based on what produces the best participation-adequacy-employment combination.

• Pillar 2 (Empirical Wage Floors): pilot regional-variation wage floors derived from BLS Occupational Employment and Wage Statistics; measure employment effects, household-income outcomes, and small-business impact; refine the methodology based on the employment-income trade-off observed.

• Pillar 3 (Sovereign Education Fund): pilot variations in retraining-eligibility windows, field-coverage definitions, and living-stipend amounts; measure enrollment, completion, post-completion employment, and field-of-study distribution; refine eligibility based on what produces the highest enrollment-to-employment ratio in target fields. The fund's quarterly asset allocation review is itself the methodology applied to capital deployment.

• Pillar 4 (Universal Healthcare Access): pilot risk-adjusted reimbursement models in selected regions; measure access, cost, and outcomes; refine the reimbursement formula based on what produces the best access-cost-outcome ratio.

• Pillar 5 (Universal Childcare): pilot quality-rating systems and provider-payment levels in selected metros; measure enrollment, parent satisfaction, child developmental outcomes; refine the payment levels based on quality-outcome correlation.

• Pillar 6 (Universal Mental Health Access): pilot integrated primary-care mental-health screening in selected federal facilities; measure detection rates, follow-through rates, outcomes; refine the screening protocol based on detection-to-outcome conversion.

• Pillar 7 (Civic Infrastructure): pilot service-delivery improvements in selected federal agencies; measure citizen satisfaction, transaction completion rates, error rates; refine the service designs based on satisfaction-completion-error patterns.

• Pillar 8 (Universal Paid Family Time): pilot leave-policy variants in selected federal workforces; measure usage rates, employer continuity, family outcomes; refine the policy based on what produces the highest combined usage and continuity.

• Pillar 9 (Universal Long-Term Care): pilot home-care versus facility-care reimbursement balances in selected states; measure care quality, family burden, cost; refine the balance based on quality-cost-burden trade-offs.

• Pillar 10 (Federal Housing Investment): pilot housing-finance and supply-side incentive variations in selected metropolitan areas; measure affordability outcomes, supply-elasticity response, demographic equity; refine the program design based on the affordability-supply-equity profile.

• Pillar 11 (Climate Architecture): pilot carbon-pricing-revenue allocation variants and adaptation-investment portfolios in selected ecosystems and economic sectors; measure emission outcomes, distributional impact, technology-deployment rates; refine the allocation based on outcome-cost ratios. Each iteration pays down the inherited environmental damage debt incrementally.

• Pillar 12 (Immigration Architecture): pilot pathway variants and integration-support programs in selected entry corridors; measure integration outcomes, labor-market matching, family-stability indicators; refine pathway designs based on outcome distributions across cohorts.

Each pillar produces, every iteration, a documented evidence base that successor administrations can build on rather than relitigating. The cumulative effect is institutional learning at federal scale.

6. What Goes Wrong Without It

When the methodology is absent, policy work degrades in predictable ways. Five failure modes recur across domains:

Institutional amnesia.

Programs persist after their initial assumptions have been disproven, because the disproof never gets written down where the next administration can find it. Successful pilots fail to scale because the conditions that made them work are not documented. Failed interventions are not formally retired; they continue absorbing budget and political capital. The same approach gets reinvented every five years because there is no shared record of what has already been tried.

Disconnect between data and decisions.

Federal agencies collect enormous quantities of administrative data, but the data often does not flow back into the decisions the data is supposed to inform. Analysis is performed but the analytic reports do not reach the decision-makers. Decision-makers act on intuition or political considerations because the data they would otherwise rely on is six months stale or filed in a system they cannot access. The methodology requires the data-decision loop to be deliberate.

Optimization for measurable proxies.

Without explicit hypothesis-and-test discipline, programs drift toward optimizing the metrics that are easiest to measure, regardless of whether those metrics reflect the program's actual goals. The methodology requires the hypotheses to be stated in advance — including the relationship between the measurable proxies and the actual outcomes of interest — so that goal drift becomes visible.

Premature scaling.

Politically successful pilots get scaled before the evidence justifies it, because the political reward for scaling exceeds the analytical reward for patience. The methodology pushes back by requiring that scaling decisions reference the evidence that has accumulated from the pilot, not just the political enthusiasm.

Failure to formally retire failed interventions.

Programs that have been tested and found to underperform their stated goals often continue indefinitely, because no formal retirement step exists. The methodology requires it: an explicit decision, supported by the analytic record, to discontinue an intervention and free its resources for something better-supported.

7. Implementation Path

Making continuous process effort the default federal operating system is a multi-year change-management effort. A practical sequence:

Year one — codify and scale existing capacity.

Expand OES staff and per-agency Chief Evaluation Officer roles to cover every federal agency at meaningful scale. Standardize the evaluation documentation format across agencies so that evaluations conducted by one agency are legible to others. Establish a centralized federal evaluations repository, modeled on the What Works Clearinghouse but cross-cutting.

Year two — link methodology to budget cycle.

Require each agency budget submission to include an evidence summary for each major program: what is being tested, what has been learned in prior cycles, what the planned iteration is for the coming year. Require any request for funding above a threshold to reference the evidence base. Refusal to evaluate becomes a budget red flag.

Year three — extend across statutory programs.

Amend authorizing statutes for major federal programs to include explicit iteration provisions: pilot scopes, evaluation requirements, sunset clauses that trigger formal review at scheduled intervals. The statutory infrastructure currently varies by program; the goal is convergence on a common pattern.

Year four and beyond — operate steady-state.

The methodology is the operating system. Every program subject to iteration, every iteration documented, every decision traceable. The cost of doing this work is real; the cost of not doing it accumulates as institutional amnesia and wasted resources, and exceeds it within a few budget cycles.

8. Closing — Why It's Worth the Effort

The methodology is not glamorous. It does not produce dramatic announcements; it produces incremental improvements that compound over decades. The political return on continuous process effort is modest in any single year and substantial in any twenty-year stretch. The economic return is the difference between a country that learns and a country that does not.

Better results come from better decisions. Better decisions come from better knowledge. Better knowledge comes from the discipline of actually testing what is believed to be true and refining it based on what is learned. The discipline does not need to be invented; it needs to be adopted as the default.

Cross-References

• An Educated Country Pays the Debt (companion vision document, 02_Vision_and_Communication) — articulates the cross-pillar synthesis this methodology supports

• Iterative Hardening Process Documentation — worked example of the methodology applied to multi-year platform development

• Open Issues Registry — operational implementation of methodology applied to architectural decision tracking

• Foundations for Evidence-Based Policymaking Act of 2018 (P.L. 115-435) — statutory foundation for federal-scale application

This document is part of the We The People Platform's 05_Analytical_Framing suite. It is intended for policy practitioners and advocacy organizations interested in the methodology behind the platform's substantive proposals.

Robertson, J. (2026). Continuous Process Effort — An Operating System for Federal Policy Iteration. We The People Platform (Version 3.8.46). https://wethepeopleplatform.com/_web_html/05_Analytical_Framing/05_Continuous_Process_Effort.html

Robertson, Jason. 2026. "Continuous Process Effort — An Operating System for Federal Policy Iteration." We The People Platform v3.8.46. https://wethepeopleplatform.com/_web_html/05_Analytical_Framing/05_Continuous_Process_Effort.html.

@misc{wtpp_2026_121_continuous_process_effort_an_operating_s,
  author    = {Robertson, Jason},
  title     = {Continuous Process Effort — An Operating System for Federal Policy Iteration},
  year      = {2026},
  publisher = {We The People Platform},
  version   = {3.8.46},
  url       = {https://wethepeopleplatform.com/_web_html/05_Analytical_Framing/05_Continuous_Process_Effort.html},
  note      = {Document 121 of 142}
}