Daria Matviienko

A long-established concern about corporate lobbying is regulatory capture: incumbent firms shaping regulation in their favor at the expense of the public interest. We study which regulations firms lobby for and how this affects industry concentration, innovation, and welfare. To measure what firms lobby for, we use LLMs to classify the federal lobbying reports of public firms and trade associations, assigning each request to a specific policy tool from a new, comprehensive taxonomy, along with the direction of the requested change and the firms it covers. We find that lobbying for firm-specific benefits such as tax decreases and government contracts is one-sided: nearly every firm asks for higher benefits. Lobbying over regulatory costs is instead contested: the largest firms are the most likely to ask for higher costs, while trade associations mostly ask for lower ones. To quantify the consequences of both kinds of lobbying, we build a dynamic general equilibrium model of oligopolistic competition and innovation in which firms lobby for firm-specific transfers and over industry-wide regulatory costs, which are set in a contest between strategic incumbents and an association of the competitive fringe. The largest firms favor higher regulatory costs mostly because their profits rise as fringe firms exit, while smaller incumbents favor them mainly as protection against displacement by innovating entrants. In the model estimated on US data, lobbying raises regulatory costs and lowers consumption-equivalent welfare through industry composition: although each incumbent innovates more, industries become more concentrated, lowering output and slowing growth. Almost all of the welfare loss comes from lobbying over regulation rather than for firm-specific benefits.

Borrowers who face similar shocks default at very different rates. We explore how much of that dispersion reflects ex ante heterogeneity in preferences of households rather than the realization of income and house-price risk. To answer this question, we estimate a life-cycle model of housing, mortgage choice, saving, and default in which households differ ex ante in patience, elasticity of substitution, and permanent income. We use PSID microdata and the simulated method of moments to estimate the model over the realized housing boom and bust. Preference heterogeneity is essential to matching the data: a single-type version of the model cannot reproduce the hand-to-mouth gradient in default, the consumption response to hand-to-mouth status, and the timing of the aggregate foreclosure spike during the financial crisis. We find that most defaults come from impatient households, who are also persistently hand-to-mouth. We also confirm the finding from the literature that defaults follow income declines and mostly occur with positive home equity: foreclosure is driven by a liquidity shock; it’s not a strategic choice driven by negative equity. In addition to decomposing the contribution of preferences in the aggregate foreclosure rate, we find that targeted liquidity provision to households based on a history of hand-to-mouth status can increase the effectiveness of policies similar to the Homeowner Assistance Fund.

Macroeconomic research on the green transition and finance research on sustainable investing—by funds, individual investors, and central banks through green-asset purchases—have developed largely in isolation. This paper connects them by asking how a higher cost of capital affects firms that produce and innovate with both clean and dirty technologies. Evidence shows that higher capital costs for brown firms can increase pollution, and that dirty firms produce many high-quality green innovations. I develop a parsimonious perfect-foresight general-equilibrium model with directed technical change and multiproduct firms. Dirty-input taxes and clean research subsidies reduce emissions, but a higher cost of capital amplifies the existing direction of technical change: when dirty innovation is relatively more productive, firms shift further toward it, which can slow the green transition. ESG strategies and central-bank green-asset purchases should therefore account not only for emissions intensity, but also for the substitutability of clean and dirty production and the structure of firms' innovation.

This paper studies the optimal level of US government debt in a continuous-time heterogeneous-agent model with incomplete markets, portfolio choice, and aggregate disaster risk, extending Aiyagari and McGrattan (1998). The income process uses autoregressive dynamics with normal-mixture innovations to match the observed income distribution. Disaster risk calls for a fiscal buffer, so the maximum welfare-improving debt-to-GDP ratio is lower than in models without it. The optimal US debt-to-GDP ratio is between 120% and 140%.