Behavioral science offers systematic frameworks for designing verification systems that accommodate rather than fight human cognitive architecture. Work requirements beginning December 2026 can be implemented through systems that help people comply or systems that catch people failing. The choice reflects design philosophy, not technical constraint. Current approaches assume beneficiaries should adapt to bureaucratic requirements. Behaviorally-informed approaches assume bureaucratic requirements should adapt to how humans actually behave.
The distinction matters because behavioral design can shift compliance outcomes by 15 to 30 percentage points without changing underlying requirements. Text message reminders increase enrollment by 10 to 19 percentage points. Form redesign raises completion rates from 73 to 96 percent while reducing errors by 60 percent. Default enrollment with opt-out reverses participation patterns, with automatic enrollment producing 50 percentage point increases over systems requiring affirmative action. These are not marginal improvements. They represent fundamental differences in who maintains coverage under identical eligibility rules.
Behavioral economics reveals that the intention-action gap is universal, not exceptional. People genuinely intend to comply with verification requirements. They fail to convert intention into action due to predictable cognitive mechanisms including present bias, decision fatigue, planning fallacy, and prospective memory limitations examined in Article 15B. Traditional program design treats intention-action gaps as moral failures justifying coverage termination. Behavioral design treats them as engineering problems requiring system redesign.
The default principle demonstrates how system architecture shapes outcomes independently of user preferences. Brigitte Madrian and Dennis Shea’s 401(k) research found automatic enrollment increased retirement savings participation by approximately 50 percentage points. Eric Johnson and Daniel Goldstein’s organ donation analysis documented participation differences exceeding 60 percentage points between opt-in and opt-out countries. Same populations, same underlying preferences, dramatically different outcomes based solely on whether participation required affirmative action or passive acceptance.
Current work requirement systems invert this principle. They require affirmative action at every step: portal login, document location, file upload, submission confirmation. Each step creates dropout opportunity. Each step loses people who intended to comply but encountered friction they could not overcome. The default outcome is coverage loss. Behaviorally-informed design would presume compliance unless evidence suggests otherwise, requiring member action only when automated verification fails. This flips the default from termination to continuation.
Automated data matching demonstrates recognition architecture in practice. States already receive quarterly wage data from unemployment insurance systems, SNAP ABAWD verification, and federal data hubs. These systems capture approximately 70 to 80 percent of expansion adult employment. Automatic verification through existing data requires zero member action, eliminates documentation burden, and produces verification accuracy exceeding manual self-reporting. The technology exists. The barrier is philosophical commitment to catching non-compliance over enabling compliance.
Friction mapping identifies where system design creates unnecessary barriers to legitimate compliance. Password requirements demanding complex credentials changed periodically create persistent friction for people accessing portals infrequently. Document specifications requiring specific formats and file sizes produce rejection errors when members photograph pay stubs on phones. Unclear instructions using bureaucratic language generate confusion wasting time and causing errors. Each friction point represents a design choice, not a necessity.
The behavioral design response systematically documents every step users must complete, identifies dropout points, measures time required at each stage, tests whether typical users can complete processes without assistance, and eliminates friction serving no legitimate program integrity purpose. This produces systems where enrollment is easy and disenrollment is hard rather than the reverse. Some friction protects against inappropriate termination. Most friction simply reflects systems designed without understanding user constraints.
Implementation intentions leverage planning specificity to double completion rates. Research by Peter Gollwitzer demonstrates that when people specify when, where, and how they will perform intended actions, follow-through approximately doubles compared to general intentions. A verification notice asking “When will you submit your verification? Write down the day and time here” costs nothing to implement and produces substantial compliance improvements. The technology is a sentence. The barrier is recognizing that design matters.
Timing interventions optimize when communications arrive relative to deadlines, cognitive states, and action opportunities. Fresh start moments following natural temporal landmarks like month beginnings produce higher engagement than arbitrary mid-period communications. Reminder timing matters: messages sent seven days before deadlines outperform those sent 30 days before or one day before. Too early and people forget. Too late and people lack time to respond. Optimal timing accommodates prospective memory constraints documented in executive function research.
Loss aversion framing recognizes that people weight potential losses roughly twice as heavily as equivalent gains. “Your coverage will end unless you submit verification” produces higher response than “To maintain eligibility, submit verification.” Same deadline, different framing, measurably different outcomes. The underlying information is identical. The psychological impact differs substantially. This is not manipulation. It is recognition that how information is presented affects whether people act on it.
Social proof leverages human tendency to look to others’ behavior as guidance for our own. Notices stating “73% of people in your county have already submitted verification” increase submission rates compared to identical notices without social comparison. The mechanism operates through normative influence and implied capability. If most people have submitted, non-submission becomes psychologically salient. If most people can submit, the task seems achievable rather than impossible.
The behavioral toolkit examined in Article 15D extends these principles through specific interventions including text message reminders, calendar integration, navigator-facilitated implementation intentions, commitment devices, and accountability partnerships. These interventions work because they accommodate cognitive limitations rather than pretending they do not exist. They shift burden from people with impaired executive function to systems with unlimited administrative capacity.
But behavioral interventions have limits. They reduce friction. They do not eliminate structural barriers. Text messages help people who have phones. They do not help people without digital access. Form simplification helps people who can read. It does not help people with limited English proficiency. Pre-population helps people whose data systems capture. It does not help people working informally for cash. The nudge toolkit assists some people at the margin. It does not address underlying inequalities making millions vulnerable to administrative exclusion.
The assumption-reality gap centers on what systems are for. If work requirements exist to identify people unwilling to work, system design should minimize false positives even if this means some people who could work maintain coverage. If work requirements exist to catch people not working, system design should maximize detection even if this means people who are working lose coverage due to documentation failures. These purposes are not equally legitimate. Medicaid is a healthcare program, not a compliance testing program.
Design philosophy beneath the tools reflects fundamental choice about whether compliance failures represent moral failures or design failures. Arkansas data showing 95% of coverage losses among people working or exempt suggests design failure, not moral failure. Georgia Pathways enrollment below 6% of projections despite minimal requirements suggests design failure, not work resistance. When populations demonstrably willing to comply cannot navigate systems, the system is broken, not the people.
For states implementing work requirements, behavioral design offers frameworks for distinguishing people unwilling to work from people unable to navigate bureaucracy. The former represents legitimate policy concern. The latter represents administrative failure. Currently, systems cannot distinguish between these populations. Both lose coverage. Behavioral design enables separation by building systems that support rather than impede legitimate compliance.
For MCOs managing affected populations, behavioral design principles suggest specific operational interventions. Text reminder sequences cost pennies per member but prevent coverage loss costing thousands in risk adjustment degradation. Form redesign requires one-time investment but produces ongoing compliance improvements. Navigator training in implementation intention techniques enhances effectiveness without requiring additional staffing. The return on investment from behavioral design typically exceeds 10:1 when measured against coverage retention value.
The recognition versus compliance distinction examined in Series 19 becomes operationally concrete through behavioral design lens. Recognition systems that automatically identify compliance through existing data represent ultimate behavioral intervention: zero member burden, maximum accuracy, elimination of intention-action gap. Compliance systems requiring monthly self-reporting represent behavioral design failure: maximum burden, minimum accommodation of cognitive constraints, systematic exclusion of populations whose barriers are administrative rather than motivational.
Work requirements policy emerged from assumptions about dependency and labor force attachment. Behavioral science reveals that how policy gets implemented determines who succeeds more powerfully than policy content itself. Systems designed around false assumptions about human capacity produce inhuman results. Systems designed around actual cognitive architecture produce outcomes aligned with stated intentions. The evidence base exists. The technology is available. The choice remains.