Théo Valentin

Abstract

This paper investigates the extent to which corporate taxpayers exhibit inattention to tax policy changes and quantifies how such inattention affects behavioral responses to reform. I exploit a French corporate income tax reform that replaced a flat-rate system with a progressive two-tier structure featuring a discrete threshold. I leverage a distinctive institutional feature to identify inattentive behavior: the self-reported tax liability. First, I find that up to 19% of firms display inattention to the reform, continuing to apply the pre-reform tax schedule rather than the prevailing code. Second, I estimate the taxable income elasticity by exploiting excess bunching at the threshold, demonstrating that aggregate frictions, among which inattention, attenuate the estimated elasticity by 39% in the short run. Combining these estimates, I calculate that inattention imposes an average cost of 903€ per firm-year, representing 2.36% of taxable income. Overall, firm inattention reduces the short-run welfare gains from the reform by 19%.

Presented at: UC Berkeley Public Finance Seminar (Berkeley, 2025), Center for Business Taxation Doctoral Conference (Oxford, 2025), UC Santa Barbara Brown Bag Seminar (Santa Barbara, 2025), University of Utah Lunch Seminar (Salt Lake City, 2025), EEA (Bordeaux, 2025), PSE Applied Economics Seminar (Paris, 2024), CESifo Public Economics Area Conference (Munich, 2024), ZEW Public Finance Conference (Mannheim, 2024), The 80th Annual Congress of the International Institute of Public Finance (Prague, 2024)

Abstract

I study how corporate tax cuts reshape firm entry, using a 2001 French reform that introduced a reduced corporate rate on profits below a threshold. Comparing corporate (CIT) and personal (PIT) business creation within département x sector cells in exhaustive registry data, I estimate a 38% increase in corporate relative to personal entry, identified as the net reallocation between organizational forms. I then show that a part of this response is not genuine entry but entity splitting: personal-form incumbents carve out new corporate entities to shift income into the lower-rate bracket. Using firms’ GPS coordinates, I document that new corporate entrants co-locate at the exact addresses of personal-form incumbents at a rising rate after the reform, compared to new personal-form entrants. I show that the personal-form incumbents experiencing the arrival of a CIT entrant experience a larger decrease in sales and revenues than when a PIT entrant sets up closeby, consistent with entity splitting. The split is corroborated by a second, confound-robust strategy matching entrants to incumbents on shared address and surname, which the general rise in corporate entry cannot manufacture, and by a sharp contemporaneous decline in the matched incumbents’ reported income onto successive personal-income-tax brackets. Together these establish the existence and direction of the splitting channel, though its magnitude is identified only suggestively given the chosen timing of splits. The findings imply that corporate entry responses to threshold-based tax cuts overstate genuine business creation, since avoidance-driven entity creation contaminates the entry margin which has direct consequences for the welfare evaluation of such reforms.

Abstract

A substantial fraction of social media consumption relies on users' behavioral biases. Yet as opposed to standard sin goods, e.g. tobacco and sodas, they are freed from corrective taxation. This paper develops a model of optimal taxation of a monopoly platform running a social media that jointly invests in an addictive technology and sets its ad-load. While generating consumer surplus, the social media harms users through overconsumption from addiction and harms non-users through negative consumption externalities. Because addiction investment is observable only by the platform, the social planner relies on a tax on total engagement as an indirect corrective instrument. A key friction is that the platform can offset the tax by adjusting its ad-load, limiting its effectiveness. I characterize the optimal tax through a sufficient statistics formula: the tax equals a welfare-weighted sum of the elasticities of addiction, ad-load and engagement to the tax. When the welfare gains of reducing addiction are large, the tax rate reaches its upper bound through the platform's participation constraint and the social planner optimally extracts all platform surplus.

Presented at: CREST Internal Seminar (2026), Doctoral Workshop on the Economics of Digitization (CESifo, Telecom Paris, TSE, 2026)

Abstract

Charitable giving is widely subsidized through the tax system, and the design of these subsidies relies heavily on estimates of the tax-price elasticity of giving. Yet charitable tax incentives differ along dimensions beyond their generosity, including which tax liabilities they offset and which organizations they make eligible. In this paper, we study wealthy taxpayers, who can simultaneously reduce their tax liabilities through charitable giving via both an income tax credit and a wealth tax credit in France. We investigate the existence of resistance to wealth taxation, which can generate responses to changes in the wealth tax credit that are disproportionate relative to the statutory price change: a household that donates instrumentally to reduce its wealth tax may drastically reduce its giving when the wealth tax credit disappears, rather than substituting toward the income tax credit. We provide empirical evidence of this large sensitivity to the wealth tax deduction exploiting two complementary reforms to the French tax system. The 2011 and 2018 reforms changed wealthy taxpayers' access to a 75 percent wealth tax credit for charitable donations while leaving unchanged the existing 66 percent income tax credit. Combining these reforms with exhaustive French administrative tax records, we show that losing access to the wealth tax credit leads to a large decline in total charitable giving and little substitution toward the income tax credit. The implied price elasticity of donations ranges from −1.6 to −2.0, consistent across both reforms and different segments of the wealth distribution. This large response is difficult to reconcile with existing estimates of the income-tax-price elasticity of giving, and is consistent with the existence of wealth tax resistance among wealthy donors. Our findings suggest that the effectiveness of charitable tax incentives depends not only on their generosity but also on the tax they offset and the organizations they subsidize.

Presented at: ERC PARTICIPATE Closing Conference (SciencesPo, 2026)