[1] Wealth Taxation and Household Saving: Evidence from Assessment Discontinuities in Norway
[WP] [REStud link] Review of Economic Studies, October 2024 (Forthcoming) Neither theory nor existing empirical evidence support the notion that wealth taxation reduces saving. Theoretically, the effect is ambiguous due to opposing income and substitution effects, and empirically, the effect may be confounded by misreporting responses. Using geographic discontinuities in the Norwegian annual net-wealth tax and third-party reported data on savings, I find that wealth taxation causes households to save more. Each additional NOK of wealth tax increases annual net financial saving by 3.76, implying that households increase saving enough to offset both current and future wealth taxes. The increase in financial saving is primarily financed by extensive-margin labor supply responses. These responses are the combination of small negative effects of increasing the marginal tax rates on wealth and larger positive effects of increasing average rates. These findings imply that income effects may dominate substitution effects in household responses to rate-of-return shocks, which has important implications for both optimal taxation and macroeconomic modeling.
See this Microeconomic Insights article for a more accessible summary of the article.
[2] Entrepreneurial Wealth and Employment: Tracing Out the Effects of a Stock Market Crash
[PDF] [Journal of Finance link] [Internet Appendix]Journal of Finance, 2023 (78:6) I provide evidence that adverse shocks to the wealth of business owners during the Financial Crisis had large effects on their firms' financing, employment, and investment. I use individual-level portfolio data from Norway to exploit the dispersion in stock returns during 2008–09 as a source of exogenous variation in entrepreneurs' wealth. I then trace out the effects of these shocks to the entrepreneurs' privately-held firms. I find that the adverse employment and investment effects are primarily driven by young firms who—relative to mature firms—obtain considerably less bank financing following an owner wealth shock. Firms adjust employment primarily through hiring less, rather than firing, consistent with firms providing extensive-margin insurance for existing workers. These findings provide a causal link between asset price shocks and the real economy; and document that equity-financing frictions and the procyclicality of entrepreneurial wealth are important channels through which economic shocks amplify.
[3] Wealth Taxation and Charitable Giving
[REStat link] [longer WP version w/ EIS and fixed cost estimation], with Thor O. Thoresen Review of Economics and Statistics, January 2025 (Forthcoming) We use third-party reported data and two quasi-experiments from Norway to study how tax incentives affect charitable giving. First, using a shock to wealth tax exposure, we estimate the semi-elasticity of giving with respect to the after-tax return on wealth. Inconsistent with the notion that households accelerate giving to reduce future taxes, we find that a 1% wealth tax reduces giving by 26%. Second, using bunching at an income-tax deduction threshold, we estimate a moderate own-price elasticity of giving of -0.44. This elasticity exhibits little heterogeneity with respect to income and wealth but is considerably larger for religious giving.
[4] A wealth tax at work [published version]
CESifo Economic Studies, 2022, with Thor O. Thoresen, Odd E. Nygård, and Jon Epland We provide descriptive evidence from Norway to address key questions surrounding the current wealth tax debate. In particular, focusing on a subset of ordinary entrepreneurs (who fully own only one firm), we find a likely limited role for the annual 1% net wealth tax in inducing liquidity constraints through including private equity in the tax base. While the wealth tax accrues above a fairly low threshold (about $150,000), the annual marginal wealth tax bill from entrepreneurial assets accounts for less than 1% of sales for 95% of entrepreneurs.
[5] How much and how fast do investors respond to equity premium changes? Evidence from wealth taxation
Revise and Resubmit at Journal of Financial Economics (March 2026), with Luigi Guiso and Andreas Fagereng We use a wealth tax reform that differentially affected the after-tax returns on risky and safe assets to study how households respond to persistent changes to the equity premium. We find that households respond slowly. It takes five years for households to reoptimize as prescribed by canonical portfolio models. Our quasi-experimental findings can be rationalized by a coefficient of relative risk aversion between 2 and 3 in combination with adjustment frictions, moderate one-time entry and small per-period participation costs. Our results provide supportive evidence for the asset pricing literature that builds on portfolio-adjustment frictions to explain asset pricing puzzles, and they have implications for optimal taxation when tax rates can differ across asset classes.
[6] Optimal Delayed Taxation in the Presence of Financial Frictions
[Substantially revised November 2025!] [PDF] with Spencer Bastani [old version titled Financial Frictions and the Non-Distortionary Effects of Delayed Taxation] In the presence of financial frictions, the timing of cash flows matters. We apply this insight to optimal income taxation by proposing a new policy: delayed taxation. Introducing a delay between the accrual and payment of income taxes yields two sources of welfare gains when some agents are borrowing constrained. First, it improves consumption smoothing by enabling constrained agents to borrow at a lower rate. Second, it lowers the present-value tax rate from the perspective of constrained agents, thereby reducing the distortionary effects of income taxation. We study marginally delayed taxation in a dynamic optimal tax model, contrasting these gains with the fiscal costs caused by income effects and defaults. We then characterize optimal delayed tax reform in a simple calibrated model. This quantitative analysis reveals significant welfare gains from delayed taxation, even when the government has no lending advantage over private credit markets. These gains persist on top of other policies, such as age-dependent taxation, nonlinear tax schedules, and subsidized lending. The key source of these gains is delayed taxation’s ability to undo distortions caused by income taxation. The final part of our paper empirically tests this mechanism using a natural experiment from Norway. There, a kinked income-contingent student debt conversion scheme replicates delayed taxation. Bunching analyses reveal elasticities that are considerably lower than those found for a regular income tax threshold. Consistent with our theory, proxies for financial constraints are associated with lower sensitivities to the de facto delayed tax, but not to the regular tax. Taken together, our theoretical, quantitative, and empirical findings underscore the potential for delayed taxation to be a powerful new component of optimal tax policy.
[7] Regulating Entrepreneurship: The Case of Capital Requirements
[PDF] Uploaded February 2024, updated November 2025, with Annika Bacher, Andreas Fagereng, and Ella Getz Wold[Initially titled Financial Constraints and Selection into Entrepreneurship; and Capital Requirements and Entry into Entrepreneurship] Governments have a long history of regulating entrepreneurs' use of limited liability by mandating a minimum amount of paid-in equity. While opponents argue that these requirements pose an unnecessary barrier to entrepreneurship, proponents maintain that they protect stakeholders, in part by screening out low-quality entrepreneurship. We exploit a 2012 Norwegian reform that lowered the capital requirement from $17,000 to $5,000 to study this quantity-quality trade-off. Our setting includes detailed data on both firms and the underlying entrepreneurs, providing comprehensive proxies for entrepreneurial quality. We first show that incorporation rates double after the reform and remain elevated over time. Post-reform entrants are considerably smaller in terms of assets and revenues but do not differ in terms of growth, survival rates, profitability, productivity, or leverage. Furthermore, post-reform entrants have similar past income levels, age, and educational attainment as pre-reform entrants. We further find no difference in educational achievement, captured by proxies for quantitative skills, communication skills, physical ability, and illicit traits. Thus, while the reform led to an influx of smaller firms, there is no indication that new entrepreneurs or their firms were riskier or of lower quality. These findings challenge the existing motivation for capital requirements. In explaining the mechanism behind the increase in entry, we find no evidence that the reform alleviated financial constraints. Rather, our empirical findings and enriched entry model suggest that returns-to-scale heterogeneity plays an important role. Optimally small entrepreneurs have more concave production functions and only enter when capital requirements are low.
[8] Insuring labor income shocks: The role of the dynasty
[NBER WP] Updated September 2025, with Andreas Fagereng, Luigi Guiso, and Luigi Pistaferri We provide empirical evidence on the importance of a relatively understudied channel of insurance against labor income shocks: transfers from (cash-rich) parents to (cash-short) children when the latter experience negative wage shocks. Matching population data for Norway across two generations, we establish several results. First, parents make a transfer, i.e., run down liquid assets, when adult children experience negative labor income shocks. Consistent with dynastic insurance, we observe no transfers when income shocks are positive. Second, parents' responses depend on the nature of the shock. If losses are temporary, parents dissave; if they are persistent, parents save in order to make future transfers. Parental transfers offset 43% of temporary and 27% of persistent losses. Third, insurance is lower when children have other smoothing options, like spousal labor supply, and is greater for shocks to their own child versus a child's spouse. Support also increases if the spouse's parents can contribute, suggesting competition for attention. Lastly, insurance ows are one-way: children do not insure their parents against income losses.
[9] Personal Holding Companies, Tax Regressivity, and Inequality
with David Seim and Gabriel Zucman [10] Taxing Entrepreneurial Wealth: Evidence from Norway, 2021–2025
