SOCIAL PROGRAMS · DÉPARTEMENT CASE: RSA PROGRAM OVERHAUL
“How do we overhaul the RSA support journey toward greater conditionality without increasing non-take-up among the most vulnerable publics?”
Completed case for a French département facing a national political mandate for conditionality.
THE CONTEXT
A national political mandate: on the program's most fragile publics.
A French département faced a national political mandate: strengthening the conditionality of the RSA minimum income benefit (revenu de solidarité active) under the French Full Employment Act, with 15 to 20 mandatory weekly activity hours, a formal engagement contract, and graduated or immediate sanctions for non-compliance. The principle was settled by the legislator. What remained was designing a departmental support journey that maximizes effective integration without amplifying non-take-up among the most vulnerable publics: isolated claimants in rural areas, single parents with very young children, publics with documented psychological distress, young care leavers.
The departmental council's project team had 8 conditionality scenarios, defined by crossing four parameters: a flexible or rigid contract (the caseworker's margin of appreciation left broad or strictly framed), graduated or immediate sanctions (a progression of warning / partial suspension / full suspension, or direct passage to suspension), in-person or blended support (100% in-person or an in-person/remote combination by situation), and a single or shared caseworker (one stable caseworker over 12 months or rotation between specialists by need). The executive committee had to decide within 3 months, with presentation to the departmental council for public deliberation.
The project team's known risk was reproducing the errors already observed in earlier territorial experiments. Some départements had seen their non-take-up rate jump from 12% to 34% after rigidifying the program: vulnerable publics actively withdrew, with massive administrative delistings and a lasting degradation of the institutional relationship. Others had kept their take-up rate but struggled to demonstrate improved integration trajectories, with a program perceived as bureaucratic and weakly binding. No configuration seemed to balance the two objectives: maximizing effective integration and preserving the institutional anchoring of fragile publics.
That is when the deputy directorate general for solidarity mobilized our system. The brief was to test each of the 8 scenarios on the département's real claimant population, with a 36-month projection of integration and non-take-up trajectories, a fine analysis of the most vulnerable typologies, and identification of the program parameters with the most decisive impact on fragile publics. The result was expected within six weeks, ahead of the project's presentation to the executive committee and its preparation for the departmental council.
THE INQUIRY
Six insights that reoriented the program's design.
Rigid conditionality amplifies non-take-up by +22 pts among vulnerable publics.
Our system projected the effect of rigidifying the contract on the département's 14 typologies of social situations. The 3 most vulnerable typologies, isolated rural claimants with limited mobility, young single parents with children under 3, publics with documented psychological distress, react to rigidification by actively withdrawing: missed appointments, unopened mail, no response to follow-ups, administrative delisting at term. Non-take-up in these 3 typologies rises from 12% (current program) to 34% (rigid conditionality). The flexible conditional journey + blended support preserves their institutional anchoring, the sine qua non of any later return to work. Without anchoring, there is no integration trajectory: there is an exit from the program without recovery.
The single caseworker beats the shared caseworker by +34 pts of adherence.
Our system tested in isolation the effect of a single versus a shared caseworker on adherence to the program. Claimants do not reject conditionality as a principle: 68% consider it legitimate in principle. What they reject is the perceived inhumanity of the arrangement: moving from one caseworker to another each month, re-explaining their situation at every change, filling in the same forms again, feeling like a file more than a person. A single caseworker over 12 months, even less technically experienced than a rotation of specialists, obtains adherence 34 pts higher. Relational stability outweighs technical expertise. This insight contradicted the specialty-based organizational approach carried by part of the HR team.
Graduated sanctions are perceived, immediate sanctions are denied.
Counter-intuitively, a graduated sanctions regime (a written warning first, then a 30% partial suspension, then full suspension) works as a logic of accountability. The written warning is received, read, and triggers a rebound of engagement in 62% of cases. An immediate sanction regime (full suspension at the first non-compliance) triggers cognitive withdrawal: claimants stop believing their situation can evolve within the program, withdraw psychologically, and the measured return-to-work rate at 24 months drops 8 pts versus the graduated regime. Severity perceived as illegitimate destroys the incentive effect. The graduated regime, perceived as fair, preserves the contract's legitimacy.
Blended in-person and remote support fits opposing vulnerabilities.
Our system tested several support modalities. 100% in-person support maximizes the human relationship but creates a mobility constraint that excludes isolated rural publics and single parents with very young children. 100% remote support removes the mobility constraint but excludes publics far from digital tools or in psychological distress needing a strong relational frame. Blended support, adjustable to each claimant's situation, with a mandatory but frequency-flexible in-person base, completed by an accessible remote layer, covers opposing vulnerabilities within the same program. This modularity by typology is identified as a structural lever of the program's success.
The 14 typologies do not respond to the program at the same speed.
Our system modeled the 36-month integration trajectories for each typology. Trajectories are strongly differentiated. Six typologies, claimants in professional transition, young workers recently out of a job, two-parent families with an integration project, reach a return-to-work rate of 45 to 62% at 24 months under the flexible program. Five typologies, long-term claimants, seniors, workers in retraining, reach 22 to 34% at 36 months, on longer trajectories requiring continuous support. Three typologies, the most vulnerable, reach 12 to 18% at 36 months, on non-linear trajectories marked by relapses and rebounds. A program calibrated on the median typologies fails the most vulnerable. A program calibrated on the most vulnerable slows the median trajectories. Journey modularity is the only structural answer.
The cost of non-take-up is social and budgetary, but above all budgetary downstream.
Our system modeled the budgetary consequences of non-take-up for the département. The direct cost of non-take-up is zero for the département (unclaimed credits are not paid). But publics who withdraw from the RSA program do not disappear: they reappear in downstream services: emergency accommodation, food aid, unscheduled emergency care, reinforced child protection. The consolidated downstream cost for the département averages €4,800 per claimant in non-take-up over 24 months, against €3,200 for a claimant inside the RSA program over the same period. Non-take-up is not an apparent saving, it is a transfer toward more expensive services. This modeling was decisive in convincing the département's finance directorate to support the flexible program.
THE METHOD
How we built the inquiry.
Our system rebuilt a synthetic population of the département's 340,000 RSA claimant households, calibrated on public data (INSEE, the national family benefits fund, the département's open data on beneficiaries) and proprietary data provided by the departmental social services (internal typologies, benefit durations, journey histories, non-take-up rates measured by cohort). The population was structured into 14 typologies of social situations, built by crossing six dimensions: age and family structure, seniority in the program, qualification level, residential and geographic situation, documented physical and psychological health, and the relationship to institutions measured by the frequency of honored appointments. No personal records entered the system.
On this base population, our system individually interviewed 4,200 synthetic claimants distributed proportionally across the 14 typologies. Each claimant was exposed to the 8 tested conditionality scenarios, coupled with 12 support programs in variable combinations. The dynamic agents conducted in-depth qualification interviews, following up with each claimant on the tipping points identified in real time: perception of the contract, feeling of recognition, capacity to respond to administrative demands, degree of institutional anchoring preserved after several months. These interviews qualified the precise thresholds beyond which active withdrawal becomes probable for each typology.
Our system then projected integration and non-take-up trajectories over 36 months for each scenario, with specific modeling of the interactions between the RSA program and downstream social services (accommodation, food aid, child protection, care). This projection produced 112 distinct trajectories per scenario, with identification of monthly tipping points and consolidated costs for the département. The strategy of flexible contract + single caseworker + graduated sanctions + modular blended support + a dedicated arrangement for the 3 most vulnerable typologies emerged as dominant, with a differential of +22 pts of non-take-up avoided and +11 pts of return to work at 24 months versus the tested alternatives.
THE DEPLOYMENT
What was decided, what happened.
The departmental council validated the dominant strategy identified by our system, with two adaptations from the dialogue with the departmental social workers' representatives. A flexible contract with the caseworker's margin of appreciation formalized through a shared criteria grid, a single caseworker over 12 months with technical supervision by a pool of specialists, graduated sanctions with a mandatory written warning before any suspension, and modular blended support calibrated by typology with a twice-monthly in-person base and remote resources accessible at all times. The plan was presented to the departmental council with a reinforced support budget of an additional €4.2 million annually, justified by the modeling of the downstream cost avoided.
Deployment began eight months after the departmental council's decision, once the 340 caseworkers concerned were trained and the technical supervision tools calibrated. The 3 most vulnerable typologies received dedicated arrangements: for single parents with very young children (partner daycare, occasional home-based support), for isolated rural claimants (in-person support in nearby town halls, occasional transport solutions), for publics in psychological distress (partnerships with public mental health centers and occupational health services). The single caseworkers received in-depth training and weekly supervision for complex cases.
At 24 months into full deployment, consolidated results validate the program's robustness. Non-take-up among the 3 most vulnerable typologies held at 12% (instead of the 34% projected under the alternative rigid scenario). The measured return-to-work rate at 24 months is 38% on average (above the 34% projection), with differentiated trajectories consistent with the typologies. Caseworkers' professional satisfaction measured at 18 months is 71% (up 14 pts versus the pre-overhaul period). The département's consolidated downstream cost fell by €8 million over the reference year (better-used credits + savings on emergency accommodation, food aid, unscheduled care). The département has shared its program as a methodological reference with three other départements engaged in the same transformation.
- NON-TAKE-UP, VULNERABLE PUBLICS, AT 24 MONTHS
- 34% → 12%vs the rigid conditionality scenario
- RETURN TO WORK AT 24 MONTHS
- +11 ptsabove projection
- CASEWORKERS' PROFESSIONAL SATISFACTION
- 71%up 14 pts vs pre-overhaul
- CONSOLIDATED DOWNSTREAM COST AVOIDED
- €8M / yearaccommodation + food aid + unscheduled care
- ADMINISTRATIVE DELISTING RATE
- divided by 2.4vs the rigid conditionality scenario
- SIMULATION COST VS DOWNSTREAM SAVINGS
- 1 : 46
THE LESSONS
Three principles transposable to structuring social policies.
Non-take-up is not an apparent saving: it is a transfer toward more expensive services.
This case confirmed a structural dynamic of social programs: publics who withdraw from a program do not disappear, they reappear in other downstream social services, often more expensive and less effective. Modeling in consolidated cost completely changes the budget trade-off for local authorities and the state. It justifies investing in flexible programs that preserve institutional anchoring, against rigid programs that look cheaper in the short term but generate multiplied downstream costs. This principle holds for every structuring social policy: the RSA, the in-work benefit, youth programs, minimum social benefits, elderly care.
Relational stability outweighs technical expertise for vulnerable publics.
This case showed that a less experienced single caseworker obtained higher adherence than a technically more qualified rotation of specialists. This asymmetry contradicts the specialty-based organization that structures part of social and medico-social services. It holds beyond the RSA: for supporting young people in integration, child protection, home-based elderly care, and programs for victims of violence. It implies an organization that preserves relational stability as a structural parameter, with technical supervision to complete the expertise without breaking the relationship.
Journey modularity is the only answer to opposing vulnerabilities.
This case showed that a uniformly calibrated program fails the typologies whose vulnerabilities are opposed: one cannot travel, another cannot use digital tools, a third needs a strong relational frame. The answer cannot be a standard program. It must be owned journey modularity, calibrated by typology, with a common base but differentiated resources. This principle holds beyond social policies: for benefits access programs, public health policies, professional integration programs, and elderly care programs.
GET STARTED
Preparing a social program overhaul?
Structuring social program overhauls, the RSA, the in-work benefit, minimum social benefits, youth support, child protection, elderly care, share common mechanics with this case. Publics differentiated by opposing vulnerabilities, non-take-up risk amplified by rigidification, the decisive weight of relational stability, consolidated downstream costs that invert the apparent budget trade-off. Every overhaul is singular, but the analytical levers are transposable.
The dynamic agents scope with you the parameters of a simulation adapted to your situation, ahead of the decision. From initial brief to first deliverable, allow 20 to 30 minutes, depending on the case's complexity and the breadth of the populations to model.