This paper studies a multi-product monopolist facing a privately informed consumer with additive and type-increasing valuations. Separate monopoly pricing is robustly optimal with respect to the type distribution if and only if the marginal rate of substitution between each pair of goods is monotone. This result is obtained by extending the conventional Myersonian approach.
We introduce a survey instrument to measure earnings risk allowing for the possibility of quitting or being fired from the current job. We find these transitions to be the key drivers of subjective risk. A link with administrative data provides multiple credibility checks and reveals that subjective earnings risk varies systematically across the population. It is also many times smaller than traditional estimates even when conditioning richly on demographics and job history. We show that subjective earnings risk can help explain why many hold limited liquid assets.
We investigate how technologies that reduce the costs of monitoring by central authorities have shaped the historical transition from small patrimonial states to large bureaucratic organizations. Our analysis is based on a novel dataset that traces changes in the organizational structure and geographic presence of the U.S. federal government over the 19th century. To identify causal effects, we develop a new identification strategy that exploits the expansion of the railroad network as a source of variation in the travel time-and thus monitoring costs-between Washington D.C. and other locations. We present three main findings. First, reductions in travel time to Washington D.C. significantly increased the likelihood of federal government presence in a location. Second, this effect is stronger for occupations and tasks characterized by more severe agency problems. Third, decreases in travel time to Washington D.C. are associated with a decline in patrimonial features of the federal government in the location, in line with enhanced monitoring capacity reducing dependence on personal trust and connections.
Abstract The advent of large language models (LLMs) creates new opportunities to conduct qualitative interviews at scale, with thousands of respondents, thereby bridging qualitative and quantitative methods. We develop a simple, versatile approach for researchers to run AI-led qualitative interviews, including voice interviews. We assess its robustness by drawing comparisons to human experts and with several respondents-based quality metrics. The versatility of the approach is illustrated through four broad classes of applications: eliciting key factors in decision making, political views, subjective mental states, and mental models of the effects of public policies. High performance ratings are obtained in all of these domains. Our applications highlight the potential of AI-led interviews as a tool for measurement, hypothesis generation, and discovering mechanisms.
Racial covenants in housing deeds were widely used in the United States during the early 20th century to segregate neighbourhoods. In 1948, the Supreme Court decision in Shelley v. Kraemer made racial covenants unenforceable in court, stopping short of prohibiting them entirely. We surmise that even after losing their judicial enforceability, such covenants could have shaped segregation if they served as a focal point for coordination on an initial equilibrium in neighbourhood formation, a context characterised by investment durability and path dependency. To test this hypothesis, we assemble novel parcel-level data on racial covenants and develop a quasi-experimental design that exploits delays between covenant execution and housing construction to isolate exposure to enforceable covenants at the time neighbourhoods were built. Covenant enforceability indeed affected early-stage neighbourhood formation in terms of housing characteristics, public infrastructure placement, and the composition of initial residents. These early differences proved durable: A one standard deviation increase in neighborhood exposure to enforceable covenants accounts for 6%-24% of the observed neighbourhood racial segregation from 1980 to 2020 and 3.1%-4.4% of housing price differentials in the 21st century. We conclude that by coordinating beliefs and decisions during neighbourhood formation, the law generated spatial inequality that has long outlived racial covenants themselves.
