Abstract: The United States has experienced a significant decline in firm entry rates and entrepreneurship since the 1980s. I document that this decline is more pronounced among married households and men, coinciding with changes in demographic composition (including the share of married households, skilled individuals, and marital sorting) and the rise in female labor force participation. To explore the relationship between demographic shifts and entrepreneurship, I develop a model of occupational choice that incorporates marital status, education, and gender. My findings suggest that changes in demographic composition account for 76% of the decline in entrepreneurship in the U.S.

Abstract: Since the 1980s, the U.S. has undergone a marked shift in business organization: the output share of pass-through entities (S-corporations, LLCs, partnerships, and sole proprietorships) has nearly doubled, while the share of C-corporations has declined. A natural candidate explanation for this reallocation is changes in tax policy. Indeed, tax policy, opportunities for tax avoidance, and the borrowing capacity of private businesses have all changed substantially over this period. To quantify the contribution of these factors to the observed reallocation of output, we develop a dynamic model of entrepreneurship featuring borrowing constraints, taxation with endogenous tax avoidance, and overhead costs. We find that changes in taxation, borrowing capacity, and tax avoidance play only a secondary role, together accounting for less than 20% of the observed reallocation. However, when we consider changes in overhead costs for pass-through entities consistent with the evidence, the model can account for up to two-thirds of the observed reallocation. Our results suggest that additional mechanisms are needed to fully account for the patterns in the data.

Abstract: We study how joint household decisions shape gender gaps in hours and wages. Using U.S. data from the American Community Survey and the American Time Use Survey, we document that gender gaps vary with both spouses’ occupation types, hourly wages are nonlinear in hours, and increases in non-market time associated with family size operate primarily through childcare and reductions in female market hours. To interpret these facts, we develop a static household model in which couples jointly choose market hours, childcare time, and expenditures under occupation-specific nonlinear earnings schedules and progressive taxation. In counterfactual experiments, margins that change the returns to hours worked, such as child-related labor-income penalties, nonlinear earnings schedules, and the secondary-earner tax wedge, generate large increases in mothers’ labor supply and reduce the aggregate gender gap from about 0.31 log points to between 0.21 and 0.09, whereas subsidizing monetary child investment substantially increases child spending but leaves the aggregate gender gap essentially unchanged.

Abstract: We study whether tax rules that increase the cost of executive incentive pay affect corporate innovation. The 2017 Tax Cuts and Jobs Act (TCJA) eliminated the Section 162(m) deduction for qualified performance-based compensation, differentially exposing firms that relied more on such pay before the reform. Using pre-reform reliance as a continuous treatment in difference-in-differences and event-study designs, we find that more exposed firms reduce R&D, patent applications, and intangible capital, while capital expenditures do not respond significantly. These firms also reduce performance-based compensation and executive vega, consistent with weaker incentives for managerial risk-taking. The effects are larger among firms with weaker growth opportunities and lower financial resilience, and are robust to controls for other TCJA provisions and alternative exposure measures. The results show that compensation-tax policy can affect innovation through managerial incentive contracts.

Abstract: This paper studies whether dispersion in borrowing costs across a firm's lenders affects firm financing and activity beyond the average cost of credit. Using matched administrative firm-bank data from Turkiye for 2010-2024, we construct loan-share-weighted measures of the average borrowing rate and cross-lender rate dispersion. To isolate lender-driven variation, we instrument both measures with the corresponding mean and dispersion of bank credit-supply shocks across each firm's pre-existing lenders. We find that higher borrowing-cost dispersion reduces credit growth and utilization, shortens loan maturity, and shrinks firms' lender networks. Firms do redirect borrowing toward their cheaper banks, but those banks do not expand enough to replace what is lost elsewhere, and the resulting contraction reaches sales, employment, value added, capital, and productivity. Moreover, the effects are substantially larger for highly leveraged firms and when disagreement about future inflation is high. In particular, a one-standard-deviation increase in inflation forecast disagreement makes the sales elasticity with respect to dispersion about 0.10 more negative while the effect of the average rate weakens. Using lender-specific forecasts, we further show that financing and real activity contract more when the rate  differences come disproportionately from banks whose inflation forecasts were below those of their peers, and most of all when those banks were already charging the firm relatively high rates. Our results suggest that average lending rates can understate financial tightening when borrowing costs diverge across lenders that firms cannot readily substitute.

Abstract: A firm that goes public acquires two things at once: capital, which is its own, and a public record of its condition, which is available to everyone who deals with it. Using linked administrative data covering the universe of formal firms in Türkiye, we follow newly listed firms, their banks, and the suppliers that sell to them, and establish three sets of results. (i) Listing transforms the issuer. We find that sales rise by about 45 percent within four years, assets and capital expand by roughly a third, and exports respond only after a two-year delay. In addition, measured productivity rises by roughly a quarter within two years and stays there, profitability improves, and R&D spending increases. Firms also reorganize as they grow: managers and high-skill workers account for most of the employment response, low-skill employment does not move, and wages rise only modestly and evenly, leaving the distribution of pay inside the firm unchanged. (ii) Banks reprice the newly public firm. Borrowing spreads fall by about 70 basis points, credit capacity expands, and the number of lending relationships rises, while the debt the firm actually draws changes less. (iii) That repricing does not stop at the issuer. Suppliers exposed to a newly listed customer sell more to that customer but show little change in other customer relationships, employment, or measured productivity. Yet their borrowing spreads fall by roughly 40 basis points, their credit limits increase, and they provide more trade credit to their own customers. We also find that these effects are larger when the IPO customer accounts for a greater share of pre-listing sales and are negligible when the relationship is economically small. Our results suggest that going public changes not only the scale, efficiency, and organization of the issuer, but also the financing conditions of private firms connected to it through production relationships.

Abstract: This paper studies how exchange-rate volatility shapes firms' expectations and real activity, and how its effects propagate through domestic production networks. Using matched survey and administrative data for Turkish manufacturing firms from 2009 to 2025, we find that higher FX volatility generates a stagflationary response: firms raise their inflation, own-price, and unit-cost expectations while lowering capacity utilization and becoming more pessimistic about sales, employment, and the macroeconomic outlook. These revisions are accompanied by lower realized activity, with sales falling by 1.2 percent and purchases by 1.5 percent, as well as a decrease in the number of active supplier and customer relationships. Using exact survey response dates and hourly movements in the one-month USD/TRY forward rate, we show that firms update their expectations within days of major intraday volatility spikes and that these revisions remain substantial after conditioning on the signed event-day exchange-rate change. We also find that production networks shape these effects, with firms connected to import-intensive suppliers reporting larger increases in expected costs and prices and sharper reductions in purchases and supplier relationships, while those selling to import-intensive customers report weaker demand and larger declines in sales and customer relationships. Our results suggest that exchange-rate volatility reaches firms not only through what they import themselves, but also through the suppliers they depend on and the customers they sell to.