A 2025 series of medically framed controlled experiments showed that adding a pay-for-performance (P4P) bonus — a discrete, 20%-level bonus granted when a physician's treatment meets a quality threshold that is adjusted to each patient's severity of illness — to either fee-for-service (FFS) or capitation (CAP) significantly reduces non-optimal medical service provision and raises quality of care, with effect sizes depending on patient severity and the base payment scheme. Yet behavioral responses are heterogeneous across individual physicians in ways the experiment does not explain. The source paper therefore asks what drives this heterogeneity — specifically the role of individuals' social preferences, attitudes, and personality traits — and whether such characteristics predict self-selection (sorting) into payment systems, which matters for whether schemes should be uniform, tailored, or self-selected. A 2026 follow-up by the same research group, pooling 659 subjects from this experimental program with Big Five personality measures, shows that traits explain part of the heterogeneity (smaller blended-payment gains for highly conscientious and agreeable subjects under CAP, mainly ceiling effects), but it explicitly reports power limitations, leaves the roles of elicited social preferences and attitudes untested, and does not address sorting. The open problem is an individual-level account of heterogeneous responses to performance pay and of selection into payment systems.

Background

How to pay physicians so that they deliver the right amount of care is a classic problem of health economics. Under fee-for-service (FFS) — a payment per service rendered — physicians have an incentive to overprovide; under capitation (CAP) — a fixed lump sum per patient — they have an incentive to underprovide (the physician-agency framework of Ellis and McGuire). Pay-for-performance (P4P) schemes try to correct these distortions by granting a bonus when measured performance crosses a threshold; the United Kingdom's Quality and Outcomes Framework is a large-scale example. Whether P4P actually improves quality of care is contested: field evidence is moderate and heterogeneous at best, while controlled experiments consistently find positive average effects. Brosig-Koch, Groß, Hennig-Schmidt, Kairies-Schwarz and Wiesen (2025) report a systematic experimental comparison within this literature. In a series of medically framed experiments (physicians, medical students, and non-medical students in the role of physicians, across laboratory and artefactual field settings), each subject decides the quantity of medical services for exogenously given patients whose severity of illness and marginal health benefit vary. A patient's health benefit is monetized, and accumulated benefits are donated to a charity funding cataract surgeries for real patients, which makes patient-regarding behavior incentive-compatible. Quality of care is measured as the deviation of the chosen quantity from the patient-optimal quantity. P4P is introduced at a within-subject level, i.e. each subject is observed under the base scheme (FFS or CAP) and under the same base blended with a threshold-based P4P bonus whose quality threshold is adjusted to the patient's severity of illness (analogous in spirit to diagnosis-related groups in hospital payment); the bonus level of 20% mirrors the UK Quality and Outcomes Framework. Main findings: P4P significantly reduces both overprovision (under FFS) and underprovision (under CAP) and improves quality; the size of the P4P effect varies systematically with the patient's severity of illness and with whether the base is FFS or CAP; a cost-benefit analysis restricted to incentive costs shows bonus expenditures rising disproportionately relative to the induced health-benefit gains. Beyond these average effects, the individual-level responses to P4P are heterogeneous, and the paper closes by asking what explains this heterogeneity: what role do individuals' underlying social preferences (such as altruism toward patients), attitudes, and personality traits play, and might the same characteristics not only explain responses to performance pay but also self-selection — who sorts into which payment system? Prior and adjacent knowledge frames this question. Personality traits, measured with the Big Five Inventory (the five broad traits Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism), had barely been linked to provider payment responses: Donato, Miller, Mohanan, Truskinovsky and Vera-Hernández (2017), in a field experiment among maternity care providers in India, found negative correlations of Conscientiousness and Neuroticism with performance-contract effects; Brosig-Koch, Hennig-Schmidt, Kairies-Schwarz, Kokot and Wiesen (2024), studying a representative primary-care physician sample in an analogous threshold-based P4P system, elicited self-reported motives (almost all physicians reported being motivated by patient benefit or both benefit and profit) and documented crowd-out-like quality reductions for some providers; Oxholm, Andersen, Waldorff, Wehberg and Pedersen (2025) show that survey-measured intrinsic motivation of Danish general practitioners relates to performance and moderates their response to a non-financial incentive (accreditation). On the selection side, experiments exist in which subjects may sort into payment schemes (Brosig-Koch, Kairies-Schwarz and Kokot 2017; Bejarano, Green and Rassenti 2017 for credence goods), and prosocial motivation is known to predict sorting into health-sector jobs (Lagarde and Blaauw 2014 for South African nurses; Ashraf, Bandiera, Davenport and Lee 2020 for public-service delivery) and into health professions (Fedele, Tonin and Wiesen 2025), while Opitz, Sliwka, Vogelsang and Zimmermann (2025) demonstrate in a large non-medical experiment that algorithmic, trait-based targeted assignment of incentive schemes can outperform any single uniform scheme. But no study connects directly measured social preferences or attitudes both to individual P4P response heterogeneity and to the choice of payment system with its consequences for realized quality of care.

Problem Statement

Provide an individual-level account of the heterogeneous physician responses to performance-based pay documented by Brosig-Koch, Groß, Hennig-Schmidt, Kairies-Schwarz and Wiesen (2025) and the broader experimental program on physician payment systems, in two linked parts. (i) Explain what drives the observed heterogeneity: identify which directly measurable individual characteristics — social preferences (e.g., incentivized measures of altruism toward patients), professional and general attitudes, and personality traits — predict how a provider's quantity and quality of care change when a severity-adjusted, threshold-based pay-for-performance (P4P) bonus is added to a fee-for-service or capitation base, and determine whether individual differences in response reflect genuine moderation of incentive effects, ceiling effects, or crowding-out of patient-regarding behavior. (ii) Establish whether these same characteristics predict self-selection into payment systems: when providers can choose among payment schemes (e.g., fee-for-service, capitation, or blended P4P variants), do measured preferences and attitudes systematically predict who chooses which scheme, and what does the resulting sorting imply for realized quality of care and for the design of uniform versus tailored or choice-based payment systems?

References

  1. Brosig-Koch, Jeannette; Groß, Mona; Hennig-Schmidt, Heike; Kairies-Schwarz, Nadja; Wiesen, Daniel (2025). 'Physicians' incentives, patients' characteristics, and quality of care: a systematic experimental comparison of performance-pay systems.' International Journal of Health Economics and Management 25(2): 217-243. DOI 10.1007/s10754-025-09390-x. https://doi.org/10.1007/s10754-025-09390-x
  2. Brosig-Koch, Jeannette; Hennig-Schmidt, Heike; Kairies-Schwarz, Nadja; Kokot, Johanna; Wiesen, Daniel (2026). 'On the role of personality traits in shaping responses to physician payment systems.' Theory and Decision 100(3): 857-906. DOI 10.1007/s11238-025-10107-2. https://doi.org/10.1007/s11238-025-10107-2
  3. Brosig-Koch, Jeannette; Hennig-Schmidt, Heike; Kairies-Schwarz, Nadja; Kokot, Johanna; Wiesen, Daniel (2024). 'A new look at physicians' responses to financial incentives: Quality of care, practice characteristics, and motivations.' Journal of Health Economics 94: 102862. DOI 10.1016/j.jhealeco.2024.102862. https://doi.org/10.1016/j.jhealeco.2024.102862
  4. Brosig-Koch, Jeannette; Kairies-Schwarz, Nadja; Kokot, Johanna (2017). 'Sorting into payment schemes and medical treatment: A laboratory experiment.' Health Economics 26(S3): 52-65. DOI 10.1002/hec.3616. https://doi.org/10.1002/hec.3616
  5. Bejarano, Hernán; Green, Ellen P.; Rassenti, Stephen (2017). 'Payment scheme self-selection in the credence goods market: An experimental study.' Journal of Economic Behavior & Organization 142: 396-403. DOI 10.1016/j.jebo.2017.08.004. https://doi.org/10.1016/j.jebo.2017.08.004
  6. Donato, Katherine; Miller, Grant; Mohanan, Manoj; Truskinovsky, Yulya; Vera-Hernández, Marcos (2017). 'Personality Traits and Performance Contracts: Evidence from a Field Experiment among Maternity Care Providers in India.' American Economic Review 107(5): 506-510. DOI 10.1257/aer.p20171105. https://doi.org/10.1257/aer.p20171105
  7. Oxholm, Anne Sophie; Andersen, Merethe Kirstine; Waldorff, Frans Boch; Wehberg, Sonja; Pedersen, Line Bjørnskov (2025). 'The role of providers' intrinsic motivation for quality of care and responses to a non-financial incentive.' Social Science & Medicine 382: 118378. DOI 10.1016/j.socscimed.2025.118378. https://doi.org/10.1016/j.socscimed.2025.118378
  8. Fedele, Alessandro; Tonin, Mirco; Wiesen, Daniel (2025). 'Self-Selection into Health Professions.' IZA Discussion Paper No. 17941, June 2025. https://docs.iza.org/dp17941.pdf
  9. Ashraf, Nava; Bandiera, Oriana; Davenport, Edward; Lee, Scott S. (2020). 'Losing Prosociality in the Quest for Talent? Sorting, Selection, and Productivity in the Delivery of Public Services.' American Economic Review 110(5): 1355-1394. DOI 10.1257/aer.20180326. https://doi.org/10.1257/aer.20180326
  10. Lagarde, Mylene; Blaauw, Duane (2014). 'Pro-social preferences and self-selection into jobs: Evidence from South African nurses.' Journal of Economic Behavior & Organization 107: 136-152. DOI 10.1016/j.jebo.2014.09.004. https://doi.org/10.1016/j.jebo.2014.09.004
  11. Opitz, Saskia; Sliwka, Dirk; Vogelsang, Timo; Zimmermann, Tom (2025). 'The Algorithmic Assignment of Incentive Schemes.' Management Science 71(2): 1546-1563. DOI 10.1287/mnsc.2022.03362. https://doi.org/10.1287/mnsc.2022.03362
  12. Ogundeji, Yewande Kofoworola; Sheldon, Trevor A.; Maynard, Alan (2018). 'A reporting framework for describing and a typology for categorizing and analyzing the designs of health care pay for performance schemes.' BMC Health Services Research 18(1): 686. DOI 10.1186/s12913-018-3479-x. https://doi.org/10.1186/s12913-018-3479-x
  13. Meacock, Rachel; Kristensen, Søren Rud; Sutton, Matt (2014). 'The cost-effectiveness of using financial incentives to improve provider quality: A framework and application.' Health Economics 23(1): 1-13. DOI 10.1002/hec.2978. https://doi.org/10.1002/hec.2978
  14. Cadena, Brian C.; Smith, Austin C. (2022). 'Performance pay, productivity, and strategic opt-out: Evidence from a community health center.' Journal of Public Economics 206: 104580. DOI 10.1016/j.jpubeco.2021.104580. https://doi.org/10.1016/j.jpubeco.2021.104580