Theme

Register of 2024 public money · GF1005

Unemployment

In short. Studies find longer or easier benefits slowed the return to work and the 2019-2021 cuts sped it; the 2021 cut lowered the benefits affected claimants received by about a third.

2024

€48.5bn, of public money, line 11 of 155

Strength of evidence

Strong level 4 of 4, on 1 of its 2 goals: Access and return to employment (P102-1) · on rules or parts of the line, which have no separate cost

Studies

24 causal studies tied; 16 findings on its goals

Summary

Version 1 · written 19 October 2026 · history

What it is and who gets it

Cash benefits for people out of work: mostly the unemployment insurance that Unédic funds for job seekers who have contributed, and the State's solidarity allowance (ASS) once insurance rights run out.

Goals and the evidence on each

  • Replacement income for job seekers: Very limited, 1 finding
  • Access and return to employment (P102-1): Strong, 15 findings

What the studies find

Of the 16 findings on its goals: 3 in the goal's direction, 7 mixed, 0 with no significant effect, 6 against it; 7 of them study a cut in the measure.

What is missing

How this summary was made

Literature searched on 19 October 2026: causal evaluations of unemployment benefits published since the search of 4 October 2026, and those the search notes list as unread: the 2019-2021 reform (eligibility, degressivity, daily reference wage, bonus-malus), the 2023 cut in benefit duration, the solidarity allowance (ASS), working while claiming, older job seekers, benefits for founders. Searched: the HAL API (11 queries, 2025-2026), the Crossref API (8 queries, from 2025), the OpenAlex API (1 query, then refused for the day), Unédic's list of publications (unedic.org, 4 pages), the IPP's site (ipp.eu), the Internet Archive's copies of the Dares publications; the Dares site answered with a captcha, HAL's full texts with a bot check, SSRN with 403. Strength of evidence rated by the rule on the Method page.

  • Version 1, 19 October 2026, written · by writer: First version. Latest updates

Under the register's two definitions

Broad definition: a causal study in the line

evaluated

At least one causal study is tied to this line, on all of it or on one of its rules; the line then counts in full, so this is an upper bound.

Strict definition: a single measure with its own cost and study

not yet evaluated

A spending category: the register holds no separate cost for the measures inside it, so none of its money counts under this definition.

24 causal studies are tied to this line, the latest from 2026. The share of public money under each definition is on the evaluations page.

Against the Haut-commissariat's perimeters of aid to companies:

  • not placed: the report lists no measure of this line

The perimeters, and the share of each with a causal study

The literature

The studies that touch this line

  • Laura Khoury et al. (2019). Entitled to Leave: the Impact of Unemployment Insurance Eligibility on Employment Duration and Job Quality. Working paper (HAL halshs-02393383; SSRN 2020; updated as Brébion, Briole, Khoury 2025 'Unemployment Insurance Eligibility and Employment Duration', SSRN 10.2139/ssrn.5362825).

    • Design, as recorded: Separation response at the eligibility threshold around the 2009 reform; regression discontinuity at the eligibility criterion for the effect of benefit receipt
    • Goal measured: Access and return to employment (P102-1)
    • Impact: adverse « a large negative impact of UI benefits receipt on employment probability up to 21 months after meeting the eligibility criterion »
    • Leakage: not informed
    • Permanence: not applicable
  • Laurent Brembilla et al. (2025). Réforme 2019-2021 : premiers résultats d'évaluation du bonus-malus. Unédic, Analyses (March 2025).

    • Design, as recorded: Preliminary comparison of separation-rate trends, firms subject to bonus-malus vs firms in the same sectors not subject (2,916-firm control group); short, Covid-affected pre-period so parallel trends not fully testable
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: effect « Les premières évaluations tendent néanmoins à montrer que le dispositif a joué un rôle, limité, dans cette baisse »
    • Leakage: not informed
    • Permanence: not applicable
    • another outcome, not counted
  • Laurent Brembilla and Marie-Hélène Nguyen (2025). Réforme 2019-2021 : Évaluation de la dégressivité des allocations les plus élevées (Document d'étude). Unédic, Analyses - Document d'étude (March 2025).

    • Design, as recorded: Difference-in-differences: claimants above the allowance threshold for degressivity vs claimants slightly below it, over a period spanning implementation; plus survey of affected and similar non-affected claimants
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « Nous ne trouvons pas d’effet statistiquement significatif de la dégressivité sur la probabilité de création d’entreprise »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • Brice Fabre and Justine Obser (2026). Réforme 2019-2021 : Évaluation de la nouvelle condition d'ouverture de droit : quels effets sur la durée des contrats ? (Document d'étude). Unédic, Analyses - Document d'étude (September 2026).

    • Design, as recorded: Difference-in-differences comparing hires of fixed-term contracts of different durations (under 7 months vs 7 months) before and after the reform; 2009 symmetric reform also examined but inconclusive
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: mixed « L’effet net sur le nombre de jours de contrat prévus dans les embauches en CDD de moins de 7 mois dans les entreprises de plus de 50 salariés est négatif »
    • Leakage: not informed
    • Permanence: not applicable
    • another outcome, not counted
    • a study of a cut
  • Maxime Le Bihan and Brice Fabre (2025). Réforme 2019-2021 : Évaluation du nouveau mode de calcul de l'allocation chômage (Document d'étude). Unédic, Analyses - Document d'étude (March 2025).

    • Design, as recorded: Regression discontinuity on contract-end date around 1 October 2021, differenced with the same window in 2018 (placebos 2017 and 2022); complementary difference-in-differences using rights opened from October 2021 with pre-reform contract end as controls
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « cet effet est porté par une hausse des contrats temporaires de moins de six mois »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • Florent Fremigacci and Antoine Terracol (2014). L'activité réduite : effet d'enfermement et effet tremplin (Subsidized temporary jobs in France: lock-in and stepping stone effects). Travail et Emploi (also Applied Economics 2013, doi 10.1080/00036846.2013.797644).

    • Design, as recorded: Duration model separating in-treatment (lock-in) and post-treatment effects of reduced activity on the hazard to regular employment (timing-of-events type), with heterogeneity by implicit tax rate and simulations
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « Simulations suggest that the lock-in effect first dominates, but that the overall effect eventually becomes positive. »
    • Leakage: not informed
    • Permanence: not applicable
  • Pierre Granier and Xavier Joutard (1999). L'activité réduite favorise-t-elle la sortie du chômage ?. Economie et Statistique, vol. 321(1), pp. 133-148.

    • Design, as recorded: Duration model of transitions to employment with immediate and persistent effects of past reduced activity, by gender; no quasi-experimental variation stated in retrieved text (weak identification)
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « Although the immediate effect is sometimes negative, working a limited number of hours would appear to make future integration into the labour market less difficult. »
    • Leakage: not informed
    • Permanence: not applicable
  • Maëlle Fontaine and Julie Rochut (2014). L'activité réduite des demandeurs d'emploi. Quel impact sur la qualité du retour à l'emploi ?. Revue économique, vol. 65(4), pp. 621-643.

    • Design, as recorded: Dynamic matching on timing of entry into reduced activity
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « Quality of employment, measured by wages and number of hours worked, is not significantly better after reduced activity. »
    • Leakage: not informed
    • Permanence: not applicable
  • Hélène Benghalem et al. (2023). The Lock-in Effects of Information on Part-Time Unemployment Benefits. Journal of Human Resources (working-paper versions 2021-2022: 'The Lock-In Effects of Part-Time Unemployment Benefits').

    • Design, as recorded: Randomized controlled trial: emails presenting the activité réduite rules sent to a random subset of about 150,000 UI recipients
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « The information provision had a significant positive impact on the propensity to work while on claim, but reduced the unemployment exit rate »
    • Leakage: leakage found « The extension of the duration of compensated unemployment counterbalanced the increase in the number of days worked while on claim »
    • Permanence: not applicable
  • Brigitte Dormont et al. (2001). L'effet de l'allocation unique dégressive sur la reprise d'emploi. Economie et Statistique, n°343, pp. 3-28.

    • Design, as recorded: Before/after comparison of two successive benefit regimes (1986-1992 single-drop schedule vs 1992 AUD) with duration models of exit to employment; macro-conditions partly controlled
    • Goal measured: Access and return to employment (P102-1)
    • Impact: adverse « the type of degression implemented from 1992 to 1996 probably slowed down the return to work »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • Ioana Marinescu and Daphné Skandalis (2020). Unemployment Insurance and Job Search Behavior. Quarterly Journal of Economics (online 2020; SSRN WP 2018).

    • Design, as recorded: Event study around benefit exhaustion with longitudinal correction for composition; workers eligible for 12-24 months of UI compared with UI-ineligible workers
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: effect « search effort (the number of job applications) increases by at least 50% during the year preceding benefits exhaustion »
    • Leakage: not informed
    • Permanence: not applicable
    • another outcome, not counted
  • Laura Khoury (2021). Choosing Unemployment Benefits: the Role of Adverse Selection and Moral Hazard. Working paper (HAL halshs-02057137); chapter of PhD thesis (tel-03168298, 2019).

    • Design, as recorded: Fuzzy regression discontinuity using an eligibility criterion for the choice of schedule; correlation tests for adverse selection
    • Goal measured: Access and return to employment (P102-1)
    • Impact: adverse « choosing a short schedule with higher average benefits increasesunemployment duration by six months »
    • Leakage: not informed
    • Permanence: not applicable
  • Luc Behaghel et al. (2008). The perverse effects of partial employment protection reform: The case of French older workers. Journal of Public Economics (quote from 2005 SSRN working-paper abstract).

    • Design, as recorded: Difference-in-differences around legislative changes since 1987, in particular the 1992 change (comparison of unemployed just above vs just below age 50)
    • Goal measured: Access and return to employment (P102-1)
    • Impact: adverse « an adverse effect of the tax on the firms' propensity to hire older workers »
    • Leakage: not informed
    • Permanence: not applicable
  • Béatrice Boutchenik and Raphaël Lardeux (2020). The Take-Up of Unemployment Benefit Extensions (Le recours aux extensions d'indemnisation à l'assurance chômage). Insee working paper (HAL insee hal-05387278).

    • Design, as recorded: Regression discontinuity on benefit-exhaustion date around 1 October 2014 (letter sent to claimants exhausting after that date)
    • Goal measured: Replacement income for job seekers
    • Impact: touched « the letter reduced by 14 points the share of claimants who do not certify their full work history »
    • Leakage: not informed
    • Permanence: not applicable
    • a goal only touched, not counted
  • Laura Khoury (2023). Unemployment Benefits and Redundancies: Incidence and Timing Effects. Journal of Public Economics.

    • Design, as recorded: Bunching at a discontinuity in UI benefits at a job-tenure threshold
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: not informed
    • Leakage: not informed
    • Permanence: not applicable
    • another outcome, not counted
  • Olivier Baguelin and Delphine Remillon (2014). Unemployment insurance and management of the older workforce in a dual labor market: Evidence from France. Labour Economics, vol. 30, pp. 245-264.

    • Design, as recorded: Natural experiment: 1 January 2003 reduction of potential benefit duration; analysis of age patterns of UI inflow before/after
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: effect « the reform increased the mean age at job termination of workers dismissed close to retirement by 4months »
    • Leakage: leakage found « dismissals of insiders close to retirement are often scheduled so that they can receive benefits until retirement »
    • Permanence: not applicable
    • another outcome, not counted
    • a study of a cut
  • Thomas Le Barbanchon et al. (2019). Unemployment insurance and reservation wages: Evidence from administrative data. Journal of Public Economics (NBER WP 23406, 2017).

    • Design, as recorded: Difference-in-differences around a reform of UI rules changing potential benefit duration; regression discontinuity as alternative strategy
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: no significant effect « We cannot reject that the elasticity of the reservation wage with respect to PBD is zero »
    • Leakage: not informed
    • Permanence: not applicable
    • another outcome, not counted
  • Thomas Le Barbanchon (2016). The effect of the potential duration of unemployment benefits on unemployment exits to work and match quality in France. Labour Economics, vol. 42, pp. 16-29.

    • Design, as recorded: Regression discontinuity design at the contribution-record threshold giving 15 instead of 7 months of potential benefit duration
    • Goal measured: Access and return to employment (P102-1)
    • Impact: adverse « the rate of return to work decreased by 28 per cent »
    • Leakage: not informed
    • Permanence: not applicable
  • Pierre Cahuc et al. (2020). Taxation of Temporary Jobs: Good Intentions with Bad Outcomes?. The Economic Journal, vol. 130(626), pp. 422-445.

    • Design, as recorded: Structural job search and matching model estimated on French data; policy simulation
    • Goal measured: Access and return to employment (P102-1)
    • Impact: adverse « the taxation of temporary jobs does not reach its objectives: it reduces the mean duration of jobs and decreases job creation »
    • Leakage: not informed
    • Permanence: not applicable
    • a simulation, not counted
  • Johan Hombert et al. (2020). Can Unemployment Insurance Spur Entrepreneurial Activity? Evidence from France. Journal of Finance (SSRN 2013, NBER w20717 2014).

    • Design, as recorded: Comparison of start-ups created before vs after a large-scale reform (design details not given in retrieved abstract)
    • Goal measured: Access and return to employment (P102-1)
    • Impact: mixed « because new firms are more productive, the reform has the impact of raising aggregate productivity »
    • Leakage: leakage found « the effect of the reform on employment is largely offset by large crowd-out effects »
    • Permanence: not applicable
  • Laurent Brembilla et al. (2024). Impact de la réforme de la condition d'affiliation minimale à l'Assurance chômage sur les comportements de retour à l'emploi et d'inscription à France Travail (ex-Pôle emploi). Dares, Document d'études n° 273 (February 2024).

    • Design, as recorded: Difference in differences: contract ends of workers with 4 to 6 months worked in 24 (losing eligibility) against those with 6 to 8 months, before and after 1 November 2019; common trends tested on the pre-reform period (p. 15); main specification on contracts over 3 months, to rule out anticipation
    • Goal measured: Access and return to employment (P102-1)
    • Impact: effect « la réforme a un impact positif et significatif sur la probabilité de retrouver un emploi dans les deux mois suivant leur fin de contrat »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • Léonie Fauvre and Raphaël Lardeux (2025). Quels effets de la réforme de la dégressivité sur la durée indemnisée et le retour à l'emploi ?. Dares, Document d'études n° 281 (April 2025).

    • Design, as recorded: Three strategies: difference in bunching of exits at month 7, difference in differences, and a regression kink at the exposure threshold; the authors call parallel trends a strong assumption (p. 21), reject them for one outcome (p. 42) and read their estimates as lower bounds with strong uncertainty (p. 49)
    • Goal measured: Access and return to employment (P102-1)
    • Impact: effect « les allocataires concernés par la dégressivité au taux de 30 % ont réduit leur durée indemnisée de près de deux mois et leur durée de non-emploi salarié d’environ un mois et demi »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • François Fontaine et al. (2025). Réforme du salaire journalier de référence et trajectoires professionnelles. Rapport IPP n° 55 (April 2025), for the Dares (also Dares, Valorisation de la recherche n° 16; HAL halshs-05042640).

    • Design, as recorded: Difference in differences with exposure graded by the share of time between contracts in the reference period; pre-reform trends parallel, with a triple difference against differential seasonality (report p. 106); mainly the effect of the lower daily benefit for careers built before the reform (p. 17)
    • Goal measured: Access and return to employment (P102-1)
    • Impact: effect « la réforme entraîne une hausse du nombre de jours travaillés chez les individus impactés durant les six mois qui suivent la fin de leur contrat »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • Sabina Issehnane et al. (2025). Discontinuité de l'emploi et réforme de l'assurance chômage : quels effets sur le revenu et le retour à l'emploi des demandeurs d'emploi ?. Dares, Valorisation de la recherche n° 17 (April 2025); HAL hal-05349013.

    • Design, as recorded: Difference in differences between claimants with non-contiguous and contiguous contracts in their reference period, for two inflow windows (May-June and October-November 2021), with coarsened exact matching; OLS, ordered probit and Cox models; one pre-period, so no pre-trend test
    • Goal measured: Replacement income for job seekers
    • Impact: adverse « se traduisant par une diminution d’environ 35% de leur allocation journalière effective »
    • Leakage: not informed
    • Permanence: not applicable
    • a study of a cut
  • Louis de Lachapelle (2025). Unemployment insurance experience rating: How do firms react to the implementation of employer-specific contribution rates?. Dares, Document d'études n° 280 (April 2025).

    • Design, as recorded: Difference in differences and event study: firms of the covered sectors against firms of the same sectors excluded by their collective-agreement code; no pre-trends in the main event study (p. 23), large ones without quintile-by-time fixed effects (p. 27); first year only
    • Goal measured: an outcome that is not one of the line's goals; its reading has no colour and is not counted
    • Impact: effect « they decrease their separation rate by 2.4 points and their hiring rate by 2.5 points »
    • Leakage: not informed
    • Permanence: not applicable
    • another outcome, not counted

Each study is coded by this site from the phrase quoted beside each verdict: impact on the goal it measures, leakage to people the law does not target, and permanence where the aim is a lasting change. The rows are in criteria.csv and studies.csv.

Searches

When the literature was searched

  • 2026-10-04: every line, for causal evaluations of its measures and rules. Searched: the HAL, Crossref, OpenAlex, PubMed and IDEAS/RePEc catalogues of research and publishers' pages, line by line, and a general web search engine until its allowance ran out; some research departments' sites refused automated access, so a study they publish may have been missed (data/literature/gaps). Found: literature/studies.csv.
  • 2026-10-19: causal evaluations of unemployment benefits published since the search of 4 October 2026, and those the search notes list as unread: the 2019-2021 reform (eligibility, degressivity, daily reference wage, bonus-malus), the 2023 cut in benefit duration, the solidarity allowance (ASS), working while claiming, older job seekers, benefits for founders. Searched: the HAL API (11 queries, 2025-2026), the Crossref API (8 queries, from 2025), the OpenAlex API (1 query, then refused for the day), Unédic's list of publications (unedic.org, 4 pages), the IPP's site (ipp.eu), the Internet Archive's copies of the Dares publications; the Dares site answered with a captcha, HAL's full texts with a bot check, SSRN with 403. Found: five evaluations of the 2019-2021 reform, coded as rows 371-375 (Dares, Documents d'études n° 273, 280 and 281; Rapport IPP n° 55; Dares, Valorisation de la recherche n° 17); not coded: the evaluation committee's final report (Dares 2025), a review of them; the journal version of row 66 (Economie et Statistique n° 550, 2026); descriptive and qualitative Dares and Unédic studies; nothing on the ASS or the 2023 cut.

Back to every line of the register

Sources: Eurostat, Ministère de l'Économie et des Finances, Direction de la Sécurité sociale, and the publishers of the studies, each named in its panel.

The page is static HTML and reads in full without JavaScript. The script adds the tooltips, traces a flow through a diagram, opens in place the evidence kept on separate pages, shows and hides the diagrams on a phone, recomputes the wealth calculator, places you on the page that asks where you stand, and keeps the theme you pick: light, dark or high contrast.

Questions, ideas or corrections: open an issue at github.com/samoht/budget/issues. Who makes this site, and the corrections made to it, are on the Method page.

Every data file, with its columns, licence and checksum: the Data page. Every chart, by question: the index of the charts. Every line of the register: the register's index.

The evidence summaries of the register as they are written: Latest updates (Atom feed).

Text and data: Licence Ouverte / Open Licence 2.0, except third-party data, which keep their publishers' terms. Code: ISC. See LICENSE.md.

Version unreleased-ef7a7ba2, built on 2026-11-12 from ef7a7ba2.