Pith. sign in

REVIEW 3 major objections 5 minor 53 references

Rebate versus Matching, Again: How Opt-in Reshapes the Effectiveness of Price-Equivalent Subsidies

T0 review · 3 major / 5 minor · reviewed 2026-07-12 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read When donors must opt in, matching stays effective while rebate loses half its power—even though both give the same price under forced assignment.

desk verdict Clean 2×2 experiment shows voluntary take-up restores a large matching advantage via opposite selection patterns; the internal result is solid, the external-validity claim is the soft spot. read the letter →

arxiv 2607.03113 v1 pith:ZWNHZ6J3 submitted 2026-07-03 econ.GN q-fin.EC

Shusaku Sasaki , Takunori Ishihara , Hiroki Kato This is my paper · ORCID
classification econ.GNq-fin.EC
keywords charitablesubsidytaxincentivematchingrebatepolicyimplementationself-selectionscalingLATE
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Theory says a 1:1 match and a 50% rebate should produce the same charitable giving because they impose the same out-of-pocket price. Earlier lab results favored matching, but later refinements that equalize budgets and fix confusion largely erase that gap when people are assigned to a scheme. This paper asks what happens when take-up is voluntary, the usual real-world case. In a nationwide incentivized experiment with 2,400 Japanese adults, the authors cross subsidy type with compulsory versus opt-in assignment. Once budgets are equalized and only high-comprehension participants are kept, total amounts received by the charity are statistically indistinguishable under compulsory assignment. Under opt-in the matching advantage reappears and nearly quadruples, not because matching improves, but because rebate loses roughly half its effectiveness. Local average treatment effect estimates show advantageous selection into matching and disadvantageous selection into rebate. The policy ranking of two price-equivalent instruments can therefore be reversed by who chooses to use them.

What carries the argument

A 2×2 design that crosses subsidy type (1:1 matching vs 50% rebate) with assignment rule (compulsory vs opt-in), combined with budget-equalized feasible sets and a LATE decomposition into treatment-on-the-treated and treatment-on-the-untreated, which isolates how self-selection changes realized effectiveness.

What would settle it

A field comparison of matching and rebate take-up and giving in which real administrative costs and timing are allowed to differ by instrument (for example, post-donation tax filing versus designated matching channels) and the ranking of total amounts received is measured under those natural frictions.

Watch

Extended reading notes

Core claim

After equalizing budget constraints and restricting to high-comprehension participants, total giving is statistically indistinguishable under compulsory 1:1 matching versus 50% rebate. Under voluntary opt-in the matching advantage re-emerges and nearly quadruples because the rebate intent-to-treat effect falls by about half while matching stays near its compulsory average treatment effect, with LATE patterns of advantageous selection into matching and disadvantageous selection into rebate.

Load-bearing premise

The experiment’s symmetric, low-cost default-off click is assumed to isolate the same selection-on-gains margin that governs real-world tax rebates and matching campaigns, where procedural burdens and timing differ sharply by instrument.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper tests whether the refined near-equivalence of 1:1 matching and 50% rebate subsidies under compulsory assignment survives voluntary take-up. In a pre-registered, incentivized online experiment with a quota-balanced sample of 2,400 Japanese adults, the authors cross subsidy type with assignment rule (compulsory vs. opt-in), equalize budget constraints via dual upper limits, and restrict attention to high-comprehension participants. Under compulsory assignment the gap in total amount received by the charity shrinks to a statistically insignificant 26 JPY; under opt-in it re-emerges at 97.7 JPY (nearly four times larger) because rebate ITT falls by roughly half while matching ITT remains near its ATE. LATE decompositions following Fowlie et al. indicate advantageous selection into matching (TOT > ATE) and disadvantageous selection into rebate (TOT ≤ ATE). Exploratory evidence links rebate take-up more to consumption motives than to altruism.

Significance. If the result holds, it identifies a distinct implementation margin—voluntary take-up—that can reverse the policy ranking of price-equivalent instruments even after budget constraints and comprehension are controlled. The design is carefully pre-registered, uses a nationwide quota-balanced sample, equalizes feasible sets, and reports ITT/LATE/TOT/TOU estimates with multiple robustness checks (covariate OLS, Tobit, extensive margin, 10 imes hypothetical endowment, Romano–Wolf and List–Shaikh–Xu corrections). These features make the paper a useful contribution both to the rebate-versus-matching literature and to the broader literature on selection into social programs and the scalability of experimental results. The finding that matching preserves effectiveness under opt-in while rebate does not is policy-relevant for tax-based charitable subsidies.

major comments (3)
  1. Section 6 and the Eckel–Grossman (2017) naturalistic acceptance gap (rebate 38% vs. matching 73%) already flag the external-validity limit of the symmetric low-cost default-off click. The central claim that self-selection reshapes the ranking is therefore best read as an experimental demonstration of an implementation margin, not as a direct prediction for real-world tax rebates (post-donation claims, filing costs, timing after the gift is chosen). The manuscript should state more sharply in the abstract and introduction that the ranking result is conditional on this standardized take-up technology, and should treat differential procedural burden as a complementary rather than competing force.
  2. Section 5.1 and Figure 2: the TOU for rebate is recovered by dividing (ATE − ITT) by the non-taker share. This is valid under the maintained assumption that the compulsory ATE is the population average of TOT and TOU, but it is sensitive to any difference in the composition of the compulsory and opt-in samples beyond take-up. The paper should report a formal test of equality of baseline covariates between the compulsory and opt-in arms within each scheme (beyond the overall balance in Table 1) and discuss whether any residual imbalance could reverse the TOT ≤ ATE ordering for rebate.
  3. Appendix D.4 and Section 4.1: the high-comprehension restriction (six or more of eight correct) is not pre-specified, yet it is the focal sample for the main ranking claim (Table 2 cols 5–6; Table 3). The stepwise presentation is transparent, but the paper should also report the opt-in ITT gap and LATE ordering on the full pre-registered sample as a primary robustness check in the main text, not only the low-comprehension columns, so that readers can judge how much the ranking depends on the post-hoc cut.
minor comments (5)
  1. Table 3 and Figure 2: report the exact p-value for the TOT-versus-ATE contrast under rebate (currently only the matching contrast is given as p=0.003) so that the strength of the disadvantageous-selection claim is transparent.
  2. Section 4.2: the Poisson specification of Chen and Roth (2024) is a useful complement, but the mapping from estimated elasticities to γ should note that the 95% CI for γ includes both zero and values near the literature average; the text currently emphasizes consistency with γ=0 more strongly than the width of the interval warrants.
  3. Appendix Table C.1 and Section 3.4: the dual-upper-limit design is clear, but a short sentence in the main text reminding readers that non-takers in the opt-in arms are analyzed under the control upper limit would reduce the chance of misreading the budget-equalization procedure.
  4. Figure 1 is referenced but the caption is minimal; a one-sentence description of the 2 imes2 plus control structure would help readers who encounter the figure before the text.
  5. Minor typographical issues: occasional missing spaces after commas in the abstract and introduction, and inconsistent hyphenation of “opt-in” versus “opt in” as a verb.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: main claims are reduced-form experimental contrasts under random assignment, not forced by fitted parameters or self-citation chains.

full rationale

The paper's load-bearing results are ITT/ATE contrasts from a 2x2 design (matching vs rebate crossed with compulsory vs opt-in) plus LATE decompositions following Fowlie et al. (2021). Budget equalization is a design choice (Section 3.4, Table C.1) that equalizes the maximum total amount received; it does not define the treatment gap by construction. Structural estimates of e and gamma (Section 4.2, Poisson regression of Chen and Roth 2024) are post-hoc interpretive complements that map elasticities to the Hungerman-Ottoni-Wilhelm impure-impact model; they are not used to force the ranking of schemes. Hypotheses 1-4 are derived from that external model and then tested against data; several predictions (equal take-up under gamma approx 0; TOT>ATE for both schemes) are only partially confirmed, and the rebate disadvantageous-selection pattern is reported as a failure of the model rather than a tautology. Self-citations (Sasaki et al. 2022, 2023 pre-registration) document prior design refinements and the PAP; they are not uniqueness theorems that forbid alternatives. No equation equates the target matching-rebate gap to a fitted constant. The derivation chain is therefore self-contained against external experimental benchmarks.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The paper is primarily reduced-form experimental. Load-bearing structure comes from standard RCT identification, the impure-impact utility model used only for hypothesis generation, and the Fowlie-style LATE decomposition. Two structural parameters are fitted but are not required for the central ranking claim. No new physical or economic entities are postulated.

free parameters (2)
  • price elasticity of giving e = -0.632 (S.E. 0.084)
    Estimated via Poisson regression on checkbook giving; used to map to structural parameters and interpret Hypothesis 2, but not required for the reduced-form ITT/LATE ranking.
  • warm-glow parameter γ = 0.168 (S.E. 0.170)
    Backed out from matching vs rebate elasticities; point estimate near zero supports the compulsory near-equivalence narrative, but the main opt-in result does not depend on the point value.
assumptions (5)
  • standard math Random assignment to compulsory or opt-in arms identifies ATE and ITT for each subsidy scheme relative to control.
    Standard experimental identification; balance table supports it (Table 1).
  • domain assumption Impure-impact utility of Hungerman and Ottoni-Wilhelm (2021) generates Hypotheses 1–4 on checkbook giving, total receipts, take-up, and TOT>ATE.
    Adopted in Section 2 and Appendix A; used for prediction, not for estimating the main treatment effects.
  • domain assumption TOT identified by instrumenting take-up with random assignment to the opt-in offer; TOU recovered as (ATE−ITT)/non-taker share (Fowlie et al. 2021).
    Section 5.1; standard LATE assumptions (exclusion, monotonicity) are implicit.
  • ad hoc to paper Using the 500 JPY upper-limit responses for matching and 1,000 JPY for rebate/control equalizes the maximum total amount received by the charity.
    Section 3.4 and Appendix Table C.1; design choice following Blumenthal et al. to remove a known confound.
  • ad hoc to paper High-comprehension subsample (six or more of eight correct) is the focal sample for the main ranking claim.
    Not fully pre-specified (Appendix D.4); stepwise full-sample results are reported but main narrative uses the restricted sample.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Rebate versus Matching, Again: How Opt-in Reshapes the Effectiveness of Price-Equivalent Subsidies." pith.science (2026). https://pith.science/paper/ZWNHZ6J3

@misc{pith2026260703113,
  author       = {Pith},
  title        = {Pith review of: Rebate versus Matching, Again: How Opt-in Reshapes the Effectiveness of Price-Equivalent Subsidies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZWNHZ6J3}},
  note         = {Machine review of arXiv:2607.03113}
}
read the original abstract

Traditional theory predicts equivalent effects of matching and rebate subsidies at equal prices, yet experiments favor matching. Refinements narrow this gap but retain compulsory assignment; take-up is voluntary in practice. We test this implementation margin in a nationwide, incentivized donation experiment with 2,400 Japanese adults, crossing subsidy type and assignment rule. After equalizing budget constraints and accounting for comprehension, total giving is indistinguishable under compulsory assignment. Under opt-in, the matching advantage re-emerges and nearly quadruples because rebate loses effectiveness. LATE estimates suggest advantageous selection into matching but disadvantageous selection into rebate. Self-selection can reshape the policy ranking of price-equivalent instruments.

Figures

Figures reproduced from arXiv: 2607.03113 by the authors.

Figure 1
Figure 1. Research Design. 25 [PITH_FULL_IMAGE:figures/full_fig_p025_1.png] view at source ↗
Figure 2
Figure 2. Treatment Effects on Total Amount Received by Charity. [PITH_FULL_IMAGE:figures/full_fig_p026_2.png] view at source ↗
Figure 3
Figure 3. Treatment Effects on Consumption (Reward to Self). [PITH_FULL_IMAGE:figures/full_fig_p026_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 1 canonical work pages

  1. [1]

    The ideal place for me to spend a vacation is a remote area surrounded by abundant nature

  2. [2]

    I am someone who is always thinking about how my actions affect nature

  3. [3]

    Interacting with nature serves as my spiritual support (mental anchor)

  4. [4]

    I am someone who always pays attention to wildlife

  5. [5]

    Interacting with nature is important for being myself

  6. [6]

    To what extent do each of the following statements apply to you? Please select on a 5-point scale, where 5 = ‘Applies very much’ and 1 = ‘Does not apply at all

    I feel a strong connection to the earth and all living things on it. 40 Appendix B.1.2. Selected Screenshots from Screening Survey Figure B.1: S4: Hypothetical donation amount question. 41 Figure B.2: S6: Comprehension check on the matching scheme (400 JPY donation). Figure B.3: S6 feedback: Correct-answer display. 42 Figure B.4: S8: Comprehension check o...

    2023

  7. [7]

    I feel happy when I see others happy

  8. [8]

    I am very concerned about how others evaluate me

Show all 53 references
  1. [9]

    I am someone who takes the initiative on tasks others would rather not do

  2. [10]

    If I had enough money to spare, I would consider donating

  3. [11]

    I would want to donate even if it meant cutting back on money I could spend on myself

  4. [12]

    When someone does me a favor, I want to repay them

  5. [13]

    I would want to retaliate against someone who treats me unfairly, even if it is costly

  6. [14]

    I feel reassured when I behave like those around me

  7. [15]

    I do not cut in line. 44

  8. [16]

    I willingly cooperate when doing something with others

  9. [17]

    I enjoy accomplishing something through cooperation with others

  10. [18]

    I commit to team play rather than individual play

  11. [19]

    I believe solving societal problems is the government’s responsibility

  12. [20]

    I pay a large amount of tax

  13. [21]

    Note: An on-screen warning was displayed on the next page to respondents who failed the attention check item (15), asking them to read questions more carefully

    [Attention check] For this question, please select the leftmost option. Note: An on-screen warning was displayed on the next page to respondents who failed the attention check item (15), asking them to read questions more carefully. Q2. Patience (time preference). Question:“Ge...

  14. [22]

    50% chance to double, 50% chance to decrease by 60%

  15. [23]

    50% chance to double, 50% chance to decrease by 50%

  16. [24]

    50% chance to double, 50% chance to decrease by 45%

  17. [25]

    50% chance to double, 50% chance to decrease by 30%

  18. [26]

    50% chance to double, 50% chance to decrease by 10%

  19. [27]

    50% chance to double, 50% chance to decrease by 5%

  20. [28]

    For each of the following characteristics, please indicate how much it applies to you. Please answer based on how you see yourself, not how you think others see you

    50% chance to double, 50% chance to decrease by 1% Q5. Rosenberg Self-Esteem Scale (10 items). Question:“For each of the following characteristics, please indicate how much it applies to you. Please answer based on how you see yourself, not how you think others see you.” Respo...

  21. [29]

    I feel that I am a person of worth, at least on an equal plane with others

  22. [30]

    I feel that I have a number of good qualities

  23. [31]

    All in all, I am inclined to feel that I am a failure

  24. [32]

    I am able to do things as well as most other people

  25. [33]

    I feel I do not have much to be proud of

  26. [34]

    I take a positive attitude toward myself

  27. [35]

    On the whole, I am satisfied with myself

  28. [36]

    I wish I could have more respect for myself

  29. [37]

    I certainly feel useless at times

  30. [38]

    Approximately how much did you donate in monetary donations during the year 2022 (January–December)?

    At times I think I am no good at all. Prior donation experience (Q6–Q8) 46 Q6. Annual donation amount in 2022. Question:“Approximately how much did you donate in monetary donations during the year 2022 (January–December)?” Notes provided to respondents:–“Monetary donation” ref...

    2022

  31. [39]

    There is a real 1-in-10 chance of winning the additional 1,000 JPY in reward points

  32. [40]

    If you win, one of your two answers is randomly selected and the donation will be executed as you answered (the remaining amount is yours to keep as additional reward)

  33. [41]

    You cannot change your answers after winning, so please consider them carefully

  34. [42]

    If you do not win, you receive nothing additional and no donation is executed

  35. [43]

    A matching gift scheme has been introduced. Our research team will add an amount equal to your donation and deliver it to the charity

    Winners will be notified by email at a later date. I8S. Charity introduction (with images). A two-image introduction page showed the recipient organization and its mission. Images depicted (1) the Karibayama–¯Ohirayama forest ecosystem protection area in Hokkaido and (2) the C...

  36. [44]

    Nature Conservation Society of Japan

    I think the recipient “Nature Conservation Society of Japan” is trustworthy

  37. [45]

    reward to self

    I carefully confirmed the automatically calculated “reward to self” and “total amount received by the charity” and am satisfied with my decision

  38. [46]

    I believe I will definitely receive the 1,000 JPY

  39. [47]

    Under the following scheme, please choose the option that applies. Tax deduction (rebate): ‘A third party covers 50% of your donation amount and refunds it to you

    If I win the 1,000 JPY, I believe the responsible researcher will deliver the amount I selected to the recipient. Hypothetical donation experiment (Q11–Q12 and variants; presentation order randomized) Q11 / Q12. Hypothetical donation amount, no scheme (Control). Question:“Toda...

    2022

  40. [48]

    Rebate versus matching, again: Does opt-in matter?

    with 10,000 pairs-bootstrap replications, two-sided, with cluster-robust standard errors. List– Shaikh–Xup-values are from Theorem 3.1 of List et al. (2019), computed viamhtexpwith 10,000 bootstrapreplications, one-sided; thepackagedoesnotaccommodatecluster-robuststandarderror...

    2019 doi

  41. [49]

    Theoretical Background

    The main experiment was scheduled to run from February 17 to 24, 2023, and was completed on February 21, 2023. -Country: Japan. -Ethics approval: This study received ex-ante approval from the IRB at The University of Osaka (Approval Number: 2022CRER0120-2, dated January 20, 20...

    2023

  42. [50]

    Onthetotalamountreceivedbythecharity: ifwarm-glowpreferencesweakentheconventional substitution effect,e m > e t; if they strengthen it,e t > e m; and if there are no warm-glow preferences,e t =e m

  43. [51]

    That is, the matching incentive is less effective than the rebate incentive in increasing the donor’s checkbook giving

    On checkbook giving:e m,o > e t for any preference. That is, the matching incentive is less effective than the rebate incentive in increasing the donor’s checkbook giving

  44. [52]

    isolation effect

    On the opt-in decision: (a) when matching and rebate yield identical donation prices, the rebate is weakly preferred to matching for any donor preference (with equality for pure- impact donors, and no opt-in benefit from matching for pure warm-glow donors); (b) donors with lar...

    2021

  45. [53]

    initially selected amount

    documents that misunderstanding of the mechanisms can contribute to the matching- versus-rebate gap. Although the pre-analysis plan did not pre-specify this restriction, we adopt it to align our analysis with this literature and to isolate responses that are not confounded by ...

Pith tools

Reviewed July 12, 2026 · model on record in the stance chip above.