A new National Association for Regulatory Administration white paper advocates smart child care regulation over cycles of deregulation and stricter rules. According to NARA, “smart regulations preserve essential health and safety protections while directing oversight toward the greatest risks.” Anchoring Human Care Oversight in Smart (Balanced) Regulation, released in August, offers a roadmap to implement a risk-based model in a phased approach.

GL Solutions interviewed Jim Murphy, Executive Director and CEO of NARA about the white paper. On behalf of NARA, he helps to explain how to apply the report’s recommendations at your licensing agency through differential monitoring, simpler compliance processes, and accountable use of technology.

Key Points

  • Focus oversight on the greatest risks while preserving essential protections.
  • Combine compliance indicators, risk assessments, quality measures and ongoing evaluation.
  • Pilot changes, establish baselines, train inspectors and scale proven approaches.
  • Simplify compliance processes without weakening responses to serious hazards.
  • Validate technology and AI, protect privacy and keep people accountable for decisions.

An example of smart regulation

Can you provide an example of smart regulation in the human care sector?

NARA: A practical example is differential monitoring in child care licensing. Every provider remains responsible for meeting the full set of applicable licensing requirements, including essential protections such as supervision, staff-to-child ratios, background checks, safe sleep, medication administration, emergency preparedness, and transportation safety. What changes is the scope and intensity of monitoring, based on reliable evidence.

Consider an illustrative comparison. A center with a sustained record of compliance may qualify for an abbreviated inspection using validated compliance indicators and checks of critical safety requirements. A center with repeated supervision or safe-sleep violations would receive a more comprehensive inspection, prompt follow-up, targeted technical assistance, and enforcement proportionate to the danger. An abbreviated visit should expand when findings reveal concerns, and complaints or serious incidents should trigger reassessment regardless of the provider’s previous record.

The distinction matters: compliance indicators identify rules that empirically predict broader compliance, while weighted risk assessments identify violations with the greatest potential for harm. They serve different purposes and should work together. As the paper explains, an effective differential monitoring system includes compliance indicators, risk assessment, quality indicators, and ongoing validation. Any change in inspection frequency must remain within applicable requirements; a strong compliance history does not justify eliminating oversight.

First practical steps

For a childcare licensing agency that wants to move toward smart regulation, what are the first practical steps you recommend?

NARA recommends starting with a defined improvement project rather than a statewide redesign. The agency should first identify where oversight is missing serious risks and where providers or inspectors are spending time on processes that add little protective value. Providers, families, licensing staff, and partner agencies should help identify those problems.

Next, review the rules and workflows. Use Caring for Our Children Basics as a reference for the evidence-based safety floor, reconcile it with applicable requirements, and identify duplicative submissions, unclear language, and conflicting instructions across programs. Establish baseline measures for serious and repeat violations, complaint response, provider administrative time, and inspector workload.

Before developing monitoring tiers, assess the completeness and consistency of inspection, complaint, incident, and enforcement data. Define how indicators will be validated, how violations will be weighted for risk, and what triggers a comprehensive review. Document the criteria and train inspectors to apply them consistently, including through shared case reviews and competency calibration.

Finally, pilot the approach with a representative group of providers across settings and risk levels. Explain the process publicly, gather feedback, and evaluate both safety and burden before expanding. Scale only after the agency can show that the pilot maintains protective oversight and that its monitoring tools remain valid. This follows the paper’s phased roadmap of assessment, system development, pilot and rollout, and ongoing evaluation.

Reducing burden while preserving safety

The paper argues that states can preserve an evidence-based safety floor while reducing unnecessary burden. How can agencies reduce provider burden without compromising safety?

Regulatory agencies should first simplify how providers demonstrate compliance. For example, a provider should be able to submit a current training record once and have authorized programs reuse it, rather than upload the same record separately for licensing, subsidy participation, and quality improvement. Shared records need clear responsibilities for verification, updates, and access.

Other practical changes include plain-language applications, consistent interpretive guidance, prefilled renewal information that providers verify, coordinated monitoring where appropriate, and electronic corrective-action submissions. Agencies can align overlapping requirements across licensing, the Child Care and Development Fund, quality rating and improvement systems, the Child and Adult Care Food Program, state prekindergarten, and local authorities while retaining each program’s necessary safeguards.

Differential monitoring can reduce unnecessary inspection burden when eligibility and escalation criteria are validated and transparent. Minor administrative deficiencies may warrant guidance and a reasonable correction period. Serious hazards require an immediate protective response; agencies should not require an education-first sequence when children face imminent danger.

The regulatory agency should measure whether these changes reduce time and cost while maintaining detection and correction of serious hazards. Proposals affecting ratios or qualifications deserve separate, carefully evaluated pilots with protective safeguards before broad adoption. Administrative simplification alone should not be presented as proof that relaxing a safety standard is justified.

Capabilities of modern licensing systems

What capabilities should modern licensing and inspection systems provide to support differential monitoring, risk assessments, public transparency, and continuous improvement?

Modern systems should make a provider’s history usable at the point of decision. They need a reliable provider identifier, consistent rule and violation definitions, and linked inspection, complaint, incident, corrective action, and enforcement records. Authorized connections to background-check systems, subsidy administration, workforce registries, and quality improvement programs can reduce duplicate entry and missing information.

For differential monitoring, the system should distinguish validated compliance indicators from weighted safety risks and quality measures. It should support configurable monitoring tiers, inspection scopes, scheduling rules, and triggers for escalation. Staff should be able to see the evidence behind a classification, record a justified override, and trace which rule or model version produced the recommendation.

Inspectors need mobile tools, including offline access where necessary, consistent checklists, secure evidence capture, and corrective-action tracking. Alerts should flag serious incidents, repeat high-risk violations, overdue inspections, and missed follow-up. Leaders need dashboards that show unresolved hazards, response times, workload, and variation in findings across inspectors and regions.

Public-facing information should explain license status, inspection findings, seriousness, correction status, and enforcement in plain language, with dates and context. Families should be able to distinguish an allegation from a substantiated finding and a resolved violation from an outstanding hazard. Confidential information must remain protected.

Continuous improvement requires data-quality checks, audit trails, exportable records, and tools to compare safety outcomes and administrative burden over time. Licensing agencies should be able to test whether abbreviated inspections miss important problems and whether monitoring criteria perform consistently across provider types. Procurement should require interoperability, accessibility, security, and agency access to its own data.

Opportunities and guardrails for AI

As states increasingly explore AI, predictive analytics, and integrated regulatory platforms, what opportunities and guardrails should agencies consider when applying these tools to childcare licensing and oversight?

The most useful starting points are tasks that help staff organize and review information: flagging incomplete applications, identifying duplicate records, summarizing a provider’s documented history, and drafting plain language explanations for staff review. Integrated platforms can connect information that otherwise sits in separate systems. Predictive tools may help prioritize review or technical assistance, but their value must be demonstrated in the agency’s operating environment.

Licensing agencies should distinguish rule-based automation, predictive risk models, and generative AI because they have different failure modes. A predictive model can reproduce bias in historical inspection or complaint patterns. Generative AI can produce plausible but inaccurate statements. Neither should be treated as established evidence of a violation or used as the sole basis for a licensing denial, sanction, or reduction in oversight.

Before deployment, define the intended use, validate data quality, and test performance against current practice. Examine false negatives that could miss hazards and false positives that could subject providers to unnecessary scrutiny. Test results across provider types and communities, accounting for differences in inspection exposure and complaint reporting. A low risk score must not override a serious incident, substantiated complaint, or required inspection.

Keep trained staff accountable for consequential decisions. Record the evidence and rationale, provide a meaningful way to correct inaccurate records and seek review, and explain how automated recommendations affect oversight. Protect child, family, staff, and provider information through limited access, secure handling, and clear retention rules. Regulatory software vendor agreements should control secondary use of agency data, require notice of model changes, and permit audits and data export.

Start with a limited pilot, monitor errors and changing performance, and retain the ability to suspend the tool. These guardrails extend the paper’s emphasis on validation, transparency, and professional competency to emerging technology. Agencies should adopt tools because they improve protective oversight and efficiency, with evidence to support that conclusion.

Additional comments on the report

Anything else you would like to comment on about this report?

The report’s central contribution is a practical framework for improving protection while addressing the complexity that providers and regulators experience. Licensing establishes minimum protections; quality 3 improvement supports better practice beyond that floor. Aligning these functions can make the system easier to navigate without treating a quality rating as a substitute for safety compliance.

Implementation requires investment in people as well as systems. Clear interpretive guidance, inspector training, consistent decision-making, and provider technical assistance are essential. NARA’s work with Dr. Rick Fiene and states and Canadian provinces, described in the paper, offers a methodological foundation for differential monitoring. Each jurisdiction still needs to validate its tools and evaluate results in its own context.

NARA encourages regulatory agencies to judge success by fewer serious and repeat violations, faster correction of hazards, more consistent oversight, lower unnecessary administrative burden, and clearer information for families. Changes in access, affordability, and workforce stability should also be tracked, without assuming that regulatory redesign alone will resolve funding or staffing shortages. The framework applies across child care, child welfare, and adult care, with safeguards tailored to each population and setting.

Source documents: NARA, Anchoring Human Care Oversight in Smart (Balanced) Regulation.

Time to Modernize

GL Solutions helps your regulatory agency run, grow and adapt with modern software and automation designed to address your agency’s biggest challenges. GL Solutions embeds AI into regulatory agency workflows in GL Suite, giving staff practical, policy-driven support that boosts efficiency, improves consistency and advances modernization—while keeping agency control and accountability at the forefront. Contact us to learn more.