When Incentives and Nudges Meet: Promoting Budget Allocations for Undervalued Policies

Makoto Kuroki; Shusaku Sasaki

REVIEW 3 major objections 3 minor 10 references

When Incentives and Nudges Meet: Promoting Budget Allocations for Undervalued Policies

T0 review · 3 major / 3 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A 50% national subsidy and a loss-framed peer nudge each raise municipal budget assessments by about ¥650,000, and combining them adds roughly ¥1.1 million; only the nudge lifts the locally funded share.

desk verdict Solid pre-registered experiment showing only the nudge moves self-financed budgets; the incentive's headline effect is the subsidy itself, and the authors should reframe the story. read the letter →

arxiv 2505.08323 v1 pith:F72Y5G2H submitted 2025-05-13 econ.GN q-fin.EC

Makoto Kuroki , Shusaku Sasaki This is my paper · ORCID

classification econ.GNq-fin.EC

keywords budgetallocationsnudgesfinancialincentivessurveyexperimentlocalgovernmentHPVvaccinationlossframingpeerinformation

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

Budget officers in local governments tend to underfund programs whose benefits arrive in the future, such as vaccination awareness campaigns. In a nationwide mail experiment with 490 Japanese municipal budget officers, the paper shows that a 50% central-government subsidy and a non-financial nudge—telling officers that neighboring municipalities have already launched similar campaigns and that failure to act will keep vaccination rates low—each raise the assessed budget by roughly ¥650,000 to ¥670,000 relative to no intervention. Combining the two yields the largest increase, about ¥1.1 million. The nudge is the only treatment that raises the self-financed portion of the budget, suggesting it works through psychological mechanisms rather than simply providing money. These findings imply that low-cost behavioral messages can be as effective as fiscal transfers in encouraging forward-looking budget decisions, at least in a hypothetical scenario.

What carries the argument

The engine of the experiment is a 2×2 between-subjects design in which two kinds of information are appended to an otherwise identical hypothetical budget-request scenario: a financial incentive (a 50% central-government subsidy) and a non-financial nudge (peer information that neighboring municipalities have already launched similar campaigns, combined with a loss-framed statement about the consequences of inaction). The outcome is the assessed budget amount, recorded both as the officer's individual assessment and as the amount after consultation with a manager, and separately as the total amount and the amount excluding the national grant. The loss-framed message targets time discounting and psychological hesitation; a time-discounting question (choosing ¥100,000 today versus ¥110,000 in one year) supports a heterogeneity analysis of whether the treatments correct myopic undervaluation.

What would settle it

Run a field version of the intervention in an actual budget-request cycle, sending the nudge and subsidy information to a randomly assigned set of municipalities and comparing final approved budgets to a control group; if the real increases fall far below the 650,000 to 1,100,000 JPY effects found in the survey, or do not appear at all, the claim that these interventions improve real budget assessments lacks support.

Watch

Extended reading notes

Core claim

The paper reports a 2×2 between-subjects field-based survey experiment embedded in a realistic hypothetical budget request for an HPV vaccination awareness campaign. Across 490 valid responses from budget officers in Japanese municipalities, the total assessed budget (including any national grant) rose by 678,350 JPY for the financial-incentive group, 661,960 JPY for the non-financial-nudge group, and 1,066,810 JPY for the combination group, each relative to the control and significant at the 1% level. When the national-grant component is removed, only the nudge group shows a significant increase in the self-financed amount (661,960 JPY); the incentive and combination groups show no significant change in what the municipality itself pays. The authors interpret this as evidence that the nudge changes the intrinsic willingness to allocate local resources, whereas the subsidy expands the total budget without affecting the local contribution. The paper also finds that the treatment effects concentrate among budget officers classified as myopic, moving their assessments close to the level of far-sighted officers.

Load-bearing premise

Budget officers' written assessments in a hypothetical mail survey reflect the amounts they would actually allocate in real budget decisions.

Editorial extensions

If this is right

If stated assessments translate into real budgets, a peer-information and loss-framing nudge can raise total funding for an undervalued campaign by roughly the same amount as a 50% national subsidy, at near-zero fiscal cost to the central government.
Combining the subsidy with the nudge yields a larger total increase than either alone (about ¥1.1 million over the control group), suggesting the two interventions operate through partly separate channels.
Only the nudge increases the self-financed local share; the subsidy alone or combined with the nudge leaves the municipality's own contribution unchanged, so policymakers who value local fiscal autonomy should prefer the nudge.
The treatment effects appear mainly among budget officers who discount future outcomes, moving their assessments up to the level of more far-sighted officers, which indicates the interventions correct undervaluation rather than simply inflating budgets.
Managerial review adds little except in the incentive-only group, since 0.867 JPY of every 1 JPY of the initial individual assessment survives the review process.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural extension not tested here is cost-effectiveness: because the nudge is nearly costless to administer, its effect per yen spent likely exceeds that of a 50% subsidy, and a follow-up could compare the marginal cost of each intervention.
The combination result—that the nudge's self-financed effect disappears when a subsidy is present—is consistent with motivation crowding-out, and it suggests that financial incentives can blunt the benefit of social-comparison framings in other intrinsic-motivation settings.
The HPV case is one instance of a broader class of delayed-benefit policies; the same 2×2 design could be applied to climate adaptation or preventive health, though the peer-information and loss-framing messages would need cultural adaptation.
The external-validity caveat points to a sharper test: linking survey responses to actual budget execution records for the same municipalities in subsequent fiscal years would reveal whether stated assessments predict real allocations.

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.

Desk Editor's Note

read the letter

Colleague, here's my read of arXiv:2505.08323. The headline result to know before you get to the paper: the financial-incentive arm's increase in total assessed budget is real but mechanical—the outcome includes the national grant, and the self-financed portion is unchanged (Table 3, Cols. 3–4). The nudge is the only treatment that raises self-financed spending. The paper reports this split clearly in Section 5.2, but the abstract and policy conclusions lead with the total-budget framing, which can easily be misread as the incentive making officers value the campaign more. That's the main weakness and it's addressable through reframing, not new data. What's genuinely new is the 2x2 design itself, applied to real budget officers in 490 Japanese municipalities: a 50% central-government matching grant, a peer-information/loss-framing nudge, and the combination. The prior literature, including the authors' own nudge-only field experiment, hadn't compared these head-to-head. The execution is solid. Stratification by municipality type, balance checks, pre-registration (AEA RCT Registry), cluster-robust standard errors at the prefecture level, and covariate-adjusted estimates are all there. The effect sizes are meaningful: roughly 1.1M JPY total for the combination and 0.66M for each alone. The internal validity holds up. The central claim in the abstract—'all three treatments significantly increased assessed budget amounts'—is true under the definition used, and the paper goes one step further and tells you the self-financed numbers, which is where the policy story actually lives. That split isn't a statistical error; it is the honest finding that only the nudge moves the locals' own money. The 2x2 also gives a sensible heterogeneous-types story for why the combination adds up. Smaller soft spots: the myopic/far-sighted classification rests on a single hypothetical time-discounting question, and the external validity of hypothetical budget assessments is an acknowledged limitation. No replication package is posted, which is a miss for a paper with pre-registration. Who should read this: people working on behavioral public budgeting, fiscal federalism, or nudge-vs-incentive comparisons in government. It's a legitimate extension and the data-collection effort is above the norm for this literature. My recommendation: send it to peer review. It deserves a serious referee. The right referee will push the authors to de-emphasize the total-budget framing for the incentive arms and make the self-financed result the headline, and to clarify that the incentive adds a grant without changing local contributions.

Referee Report

3 major / 3 minor

Summary. This paper reports a mail-based survey experiment with 490 Japanese municipal budget officers, using a 2x2 factorial design to test the effects of a 50% national subsidy (financial incentive) and a loss-framing/peer-information nudge on assessed budget amounts for an HPV vaccination awareness campaign. The authors find that all three treatments raise the total assessed budget amount, with the combination having the largest effect, and that only the nudge raises the self-financed portion. The paper interprets this as evidence that low-cost nudges can match financial incentives in changing budget assessments.

Significance. The experimental design is a clear strength: the study is pre-registered, uses stratified randomization, reports balance tests, employs cluster-robust standard errors, and samples real budget officers from all Japanese municipalities. If the main finding were correctly stated, the paper would make a useful contribution to behavioral public administration and public budgeting. However, the central interpretation is undermined by the non-comparability of the total-budget outcome across treatment arms, as detailed in the major comments. The nudge finding on the self-financed margin is novel and policy-relevant, but the framing of the incentive effects needs substantial revision.

major comments (3)

[Appendix 1, Table 3 (Columns 3-4), Abstract] The total assessed budget amount is defined differently across treatment arms: in the control and nudge arms, total equals the self-financed amount, while in the incentive and combination arms, total includes the central government grant (see the instruction '国からの補助額を含めた総額を記入してください' in Appendix 1). Table 3, Columns 3-4 show that relative to control, self-financed amounts change by -34,350 JPY (Incentive) and +140,470 JPY (Combination), both statistically insignificant, while the nudge increases self-financed by 66,538 JPY (p<.01). The incentive arms' total increases of 678,350 and 1,066,810 JPY are therefore largely or entirely mechanical, reflecting the inclusion of the grant rather than an increase in the local budget decision. The abstract's claim that 'Both the financial incentive and the nudge independently increased assessments' is thus misleading: the incentive did not increase the amount the officer chose to allocate from local resources. The authors should either make the self-financed amount the primary outcome or clearly frame the total as total project cost, and adjust the abstract and policy conclusions accordingly.
[Section 5.2, Abstract] The statement that the financial incentive and the nudge had 'comparable' effects holds only for the total-budget outcome. For the self-financed outcome (Columns 3-4 of Table 3), the nudge is significant while the incentive is not. Presenting the two interventions as equivalent on a decision-relevant margin is inaccurate; the only intervention that increased the local government's own contribution is the nudge. This is not a cosmetic wording issue, because it reverses the policy conclusion about the incentive's effectiveness relative to the nudge.
[Section 6.3] Using the far-sighted group's control-group mean as a normative benchmark to claim that the treatments 'corrected' undervaluation is a post-hoc interpretation that was not pre-registered. The argument that the combination treatment may have led to overestimation is speculative, and the subsequent claim that the 'appropriate level' might exceed the far-sighted group's mean is ad hoc. This section should be labeled as exploratory, and the subgroup analysis should not be given a causal interpretation because time-discounting tendency was not randomly assigned.

minor comments (3)

[Table 2 and Table 3] The units of the budget amounts are not clearly stated in the main text; values such as 187.246 in Table 2 and 111.522 in Table 3 appear to be in units of 10,000 JPY, but this should be stated explicitly for readers.
[Figure 1] The figure caption should define 'National grant' and 'Self-financed' and explain how the self-financed amount is computed for the incentive arms (total minus the 50% grant), so the reader can follow the decomposition without consulting the appendix.
[Section 6.1] The manipulation check reports sensitivity to 'projected funding for FY2024' and 'performance targets for FY2024,' but it is not immediately clear which items in the survey (E1-3) correspond to these constructs; please map the figures to the specific survey questions.

Circularity Check

1 steps flagged · score 6.0 of 10

The incentive arms' headline total-budget effect is largely an accounting identity: the outcome is defined to include the 50% subsidy, while the self-financed portion is unchanged; the nudge arm's effect is genuine.

self definitional [Section 4.2 / Appendix 1 response-field note; Section 5.2 and Table 3, Cols. 1-4]
"※国からの補助額を含めた総額を記入してください。該当する介入群にのみ記載 ... These results suggest that while the combination and financial incentive groups did not alter the self-financed amount compared to the control group, the total assessed amount was higher due to the inclusion of the national grant."

The headline outcome is the total assessed budget amount, and in exactly the arms that receive the treatment (50% national subsidy) the questionnaire instructs respondents to record the total including that subsidy; the control and nudge arms have no such component. With a 50% matching grant, Total = Self-financed + Grant = 2 × Self-financed. Table 3 Cols. 3-4 show the self-financed coefficients for the incentive arm (-34,350 JPY) and the combination arm (+140,470 JPY) are statistically insignificant, while the total-budget effects are 678,350 and 1,066,810 JPY (p<.01).

full rationale

The paper is a randomized survey experiment, not a derivation from assumptions, and most of its machinery is self-contained. The nudge result is independent: for that arm total and self-financed amounts coincide, and the 661,960 JPY increase is a real behavioral effect. The self-citations (Kuroki and Sasaki 2023a, 2023b) are used for motivation and effect-size comparison, not as load-bearing evidence, so they do not raise the score. The circular component is the financial-incentive headline: because the outcome is defined as the total including the 50% subsidy, and because Table 3 shows the self-financed portion did not change, the large significant effect on total assessed budget in the incentive and combination arms is substantially an accounting identity rather than evidence that budget officers increased local allocations. The paper is transparent about this in Section 5.2 and Section 7, but the abstract's statement that 'the financial incentive ... increased assessments' is nonetheless partly definitional for those arms. I therefore score this as partial circularity, not full circularity, since the nudge effect and the self-financed null results are genuine empirical findings.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper does not fit constants or construct a model; treatment effects are estimated from a randomized comparison, so no free parameters are introduced. The causal interpretation rests on standard experimental assumptions (random assignment, no interference, outcome measurement as instructed) plus the domain assumptions listed above, which are the load-bearing premises for the paper's policy conclusions.

assumptions (4)

domain assumption Hypothetical scenario responses reflect real budget allocation behavior
Section 7 acknowledges that the experiment relied on hypothetical scenarios and that external validity is the main limitation; the policy relevance of the estimated effects depends on this premise.
domain assumption The far-sighted group's mean assessed amount is an appropriate benchmark
Section 6.3 interprets the far-sighted group's constant term (1,265,674 JPY) as the appropriate assessment level and uses it to argue that treatments correct myopic underassessment; this normative equivalence is assumed, not derived.
domain assumption A single time-preference question classifies budget officers into myopic and far-sighted types
Section 6.3 uses the 100,000 yen now versus 110,000 yen in one year choice to split the sample; classification noise would weaken the subgroup interpretation.
domain assumption Peer information and loss framing act as a single coherent nudge intervention
Sections 2 and 4.2 bundle neighboring-municipality information with a loss-framed outcome message into one treatment arm, so the effect of either component cannot be identified separately.

how reviews work

0 comments

Cite this review

Pith. "Pith review of When Incentives and Nudges Meet: Promoting Budget Allocations for Undervalued Policies." pith.science (2026). https://pith.science/paper/F72Y5G2H

@misc{pith2026250508323,
  author       = {Pith},
  title        = {Pith review of: When Incentives and Nudges Meet: Promoting Budget Allocations for Undervalued Policies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F72Y5G2H}},
  note         = {Machine review of arXiv:2505.08323}
}

read the original abstract

Budget officers often assess public project proposals based on available financial support and expected outcomes. However, behavioral factors such as time discounting and psychological hesitation may lead to underinvestment in programs with delayed but significant benefits. This study investigates whether financial incentives and non-financial nudges can influence budgetary decisions in local governments. We conducted a nationwide mail-based survey experiment targeting budget officers in Japanese municipalities and received 490 valid responses. Using a 2*2 randomized design, we tested the independent and combined effects of a financial incentive (a 50% national subsidy) and a non-financial nudge (loss framing and peer information). All three treatments significantly increased assessed budget amounts compared to the control group. The largest effect appeared in the combination group (approximately 1.1 million JPY higher, p < .01). Both the financial incentive and the nudge independently increased assessments by approximately 650,000-670,000 JPY (p < .01). Notably, only the nudge raised the self-financed portion of the budget. These findings demonstrate how low-cost behavioral interventions can improve budget assessments for undervalued projects. As an application, we embedded the experiment in a case involving HPV vaccine promotion in Japan, but the approach can be broadly relevant to other public policy domains requiring forward-looking budget decisions.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

10 extracted references · 7 canonical work pages

[3]

We used cluster-robust standard errors at the prefecture level

Individual and Team’s Decision-Makings (Total Budget Amount) Notes: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. We used cluster-robust standard errors at the prefecture level. Incentive×Nudge Incentive Nudge Individual assessed budget amount Team’s assessed budget amount Covariates 108.092*** 68.757*** 66....

work page 2023
[19]

and Willoughby, K

https://doi.org/10.1111/pbaf.12289 Melkers, J. and Willoughby, K. (2001), Budgeters ’ views of state performance -budgeting systems: distinctions across branches, Public Administration Review, V ol. 61 No. 1, pp. 54-64. Melkers, J. and Willoughby, K. (2005), Models of performance -measurement use in local governments: understanding budgeting, communicatio...

work page doi:10.1111/pbaf.12289 2001
[245]

Rosen, and James R

https://doi.org/10.1080/10967494.2015.1022676 Case, Anne C., Harvey S. Rosen, and James R. Hines

work page arXiv 2015
[1993]

Journal of Public Economics 52(3): 285–307

Budget Spillovers and Fiscal Policy Interdependence. Journal of Public Economics 52(3): 285–307. https://doi.org/10.1016/0047- 2727(93)90036-S DellaVigna, S., & Linos, E. (2022). RCTs to scale: Comprehensive evidence from two nudge units. Econometrica, 90(1), 81-116. Demaj, L

doi:10.1016/0047- 2022
[2006]

Public Administration Review

Does Performance Budgeting Work? An Examination of the Office of Management and Budget’s PART Scores. Public Administration Review. 66: 742-752. https://doi.org/10.1111/j.1540-6210.2006.00639.x Heinrich, C. J

arXiv 2006
[2012]

Public Administration Review

How Credible Is the Evidence, and Does It Matter? An Analysis of the Program Assessment Rating Tool. Public Administration Review. 72(1): 123 -34. https://doi.org/10.1111/j.1540-6210.2011.02490.x Ikeda, S., Y . Ueda, A. Yagi, S. Matsuzaki, and E. Kobayashi,

arXiv 2011
[2014]

Experimentation in Political Science

“Experimentation in Political Science.” In Cambridge Handbook of Experimental Political Science, edited by J. N. Druckman, D. P. Green, J. H. Kuklinski, and A. Lupia, 3-11. Cambridge University Press. Epperson, R., & Reif, C. (2019). Matching subsidies and voluntary contributions: A review. Journal of Economic Surveys, 33(5), 1578-1601. Frey, B. S., & Jeg...

work page 2019
[2016]

Annual Review of Political Science

Policy Making for the Long Term in Advanced Democracies. Annual Review of Political Science. 19:1, 433-454. Karlan, D., & List, J. A. (2007). Does price matter in charitable giving? Evidence from a large -scale natural field experiment. American Economic Review, 97(5), 1774-1793. Kuroki, M. (2022). Impact of Depreciation Information on Capital Budgeting a...

work page doi:10.1111/auar.12355 2007

Show all 10 references

[2019]

Expert Review of Vaccines , 18(4): 323 -325, DOI: 10.1080/14760584.2019.1584040 Jacobs A

HPV vaccination in Japan: what is 18 happening in Japan? . Expert Review of Vaccines , 18(4): 323 -325, DOI: 10.1080/14760584.2019.1584040 Jacobs A. M

2019
[2024]

Mohr, Z., and L. Kearney. (2021). Behavioral-Experimental Public Budgeting and Financial Management: A Review of Experimental Studies in the Field. Public Finance and Management. 20(1). Namba, M., Y . Kaneda, C. Kawasaki, R. Shrestha, and T. Tanimoto. (2023). Underlying backgr...

2021

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [3]

We used cluster-robust standard errors at the prefecture level

Individual and Team’s Decision-Makings (Total Budget Amount) Notes: *, **, and *** indicate statistical significance at the 10%, 5%, and 1% levels, respectively. We used cluster-robust standard errors at the prefecture level. Incentive×Nudge Incentive Nudge Individual assessed budget amount Team’s assessed budget amount Covariates 108.092*** 68.757*** 66....

work page 2023

[2] [19]

and Willoughby, K

https://doi.org/10.1111/pbaf.12289 Melkers, J. and Willoughby, K. (2001), Budgeters ’ views of state performance -budgeting systems: distinctions across branches, Public Administration Review, V ol. 61 No. 1, pp. 54-64. Melkers, J. and Willoughby, K. (2005), Models of performance -measurement use in local governments: understanding budgeting, communicatio...

work page doi:10.1111/pbaf.12289 2001

[3] [245]

Rosen, and James R

https://doi.org/10.1080/10967494.2015.1022676 Case, Anne C., Harvey S. Rosen, and James R. Hines

work page arXiv 2015

[4] [1993]

Journal of Public Economics 52(3): 285–307

Budget Spillovers and Fiscal Policy Interdependence. Journal of Public Economics 52(3): 285–307. https://doi.org/10.1016/0047- 2727(93)90036-S DellaVigna, S., & Linos, E. (2022). RCTs to scale: Comprehensive evidence from two nudge units. Econometrica, 90(1), 81-116. Demaj, L

doi:10.1016/0047- 2022

[5] [2006]

Public Administration Review

Does Performance Budgeting Work? An Examination of the Office of Management and Budget’s PART Scores. Public Administration Review. 66: 742-752. https://doi.org/10.1111/j.1540-6210.2006.00639.x Heinrich, C. J

arXiv 2006

[6] [2012]

Public Administration Review

How Credible Is the Evidence, and Does It Matter? An Analysis of the Program Assessment Rating Tool. Public Administration Review. 72(1): 123 -34. https://doi.org/10.1111/j.1540-6210.2011.02490.x Ikeda, S., Y . Ueda, A. Yagi, S. Matsuzaki, and E. Kobayashi,

arXiv 2011

[7] [2014]

Experimentation in Political Science

“Experimentation in Political Science.” In Cambridge Handbook of Experimental Political Science, edited by J. N. Druckman, D. P. Green, J. H. Kuklinski, and A. Lupia, 3-11. Cambridge University Press. Epperson, R., & Reif, C. (2019). Matching subsidies and voluntary contributions: A review. Journal of Economic Surveys, 33(5), 1578-1601. Frey, B. S., & Jeg...

work page 2019

[8] [2016]

Annual Review of Political Science

Policy Making for the Long Term in Advanced Democracies. Annual Review of Political Science. 19:1, 433-454. Karlan, D., & List, J. A. (2007). Does price matter in charitable giving? Evidence from a large -scale natural field experiment. American Economic Review, 97(5), 1774-1793. Kuroki, M. (2022). Impact of Depreciation Information on Capital Budgeting a...

work page doi:10.1111/auar.12355 2007