This paper studies social welfare in markets for natural disaster insurance. Wagner quantifies frictions in uptake, test for adverse selection, and estimate the welfare effects of proposed policy reforms by developing a model of natural disaster insurance markets and compiling new data. The paper has three main findings. First, willingness to pay for natural disaster insurance is remarkably low. In the high-risk flood zones throughout all U.S. Atlantic and Gulf Coast states, fewer than 60% of homeowners purchase flood insurance even though subsidized premia are only two-thirds of their own expected payouts. Second, homeowners select into insurance based on observable differences in houses' defensive investments against natural disasters (i.e., adaptation), but not on private information about risk. Exploiting house-level variation in flood insurance prices and construction codes reveals that requirements to elevate newly constructed homes reduce insurer costs by 31% and insurance demand by 25%. Asymmetric information between homeowners and insurers, however, does not affect average payouts. Third, ignoring how frictions, such as risk misperception, distort demand understates the welfare cost of currently proposed price increases and changes the sign of the predicted welfare effect. In contrast, enforcing a natural disaster insurance mandate increases social welfare.
Schoenherr, Skrastins, Fazio, and Doornik document that a more generous unemployment insurance (UI) system shifts labor supply from safer to riskier firms and reduces compensating wage differentials risky firms need to pay. Consequently, a more generous UI system increases risky firms' value and fosters entrepreneurship by reducing new firms' labor costs. Exploiting a UI reform in Brazil that affects only part of the workforce allows us to compare labor supply for workers with different degrees of UI protection within the same firm, sharpening identification of the results. Altogether, their results suggest that UI provides a transfer system from safe to risky firms.
Consumer data is increasingly available to firms through private exchanges. Jin and Vasserman study a voluntary monitoring program by a major U.S. auto insurer, in which drivers accept short-term tracking in exchange for potential discounts on future premiums. Using a proprietary dataset matched with competitor price menus, the researchers document that safer drivers self-select into monitoring, and those who opt-in become yet 30% safer while monitored. They then model the forces of supply and demand shaping the amount of information revealed in equilibrium. Jin and Vasserman find large profit and welfare gains from introducing monitoring. Requiring the firm to make monitoring data public would have reduced short-term welfare.
Sen and Sharma show that U.S. life insurers used internal models to over-report the value of a large fraction of corporate bonds during the financial crisis. Reported credit spreads of bonds valued using internal models were substantially lower by 220 bps, as compared to bonds that are otherwise similar but valued using external sources. Misreporting is higher for bonds that are likely to be impaired and negatively affect regulatory ratios, for insurers that have low regulatory capital, and for bonds that are held by few insurers. Using novel data on U.S. state regulators, they document that misreporting is negatively correlated with the degree of supervision at the state level, but only within bonds held by multiple insurers. In contrast, supervision has limited impact on misreporting when a bond has a single owner, as the lack of reference prices increases the search cost for regulators. Consistent with these incentives, Sen and Sharma show that insurers "corner the market" by holding a large fraction of a bond's issue, as this allows them to bypass regulatory scrutiny and limit trading, factors that enable misreporting with internal models. Their findings have implications for the micro-structure of a segment of the corporate bond market and for properly assessing the fragility of financial institutions in bad times.
Government marketing activities are often used to increase enrollment in public programs. They are common in market-based public programs, in which private firms provide benefits in regulated markets and conduct their own marketing activities. This paper studies government and private marketing activities in the context of the Affordable Care Act health insurance marketplace. Using detailed TV advertising data, Aizawa and Kim study a key question for designing market-based public programs: should the government engage in marketing activities, or should private firms exclusively engage in marketing activities? The researchers present evidence that government advertising and private advertising are targeted to different geographical areas and provides different messaging content. Then, by exploiting discontinuities in advertising along the borders of local TV markets, Aizawa and Kim estimate the impact of each of these types of advertising on consumer enrollment. They find that government advertising has a market expansion effect, increasing the total program enrollment. In contrast, private advertising tends to steal consumers from other insurers, with little net effect on enrollment. Finally, by estimating an equilibrium model of the marketplace, the researchers explore the impact of changing government advertisement spending. Aizawa and Kim find that an exogenous increase in government advertising increases total program enrollment while reducing inefficient rent-seeking advertising competition among private insurers: a $1 increase in government advertising decreases private advertising by $0.13. Their finding suggests a welfare-enhancing role of government advertising as a market design tool.
This paper was distributed as Working Paper 27695, where an updated version may be available.
Abrardi, Colombo, and Tedeschi study a competitive insurance market in which insurers have an imperfect informative advantage over policyholders. They show that the presence of insurers privately and heterogeneously informed about risk can explain the persistent profitability, the pooling of risk and the concentration levels observed in some insurance markets. Furthermore, the researchers find that a lower market concentration may entail an increase in insurance premia.
Gennaioli, La Porta, Lopez-de-Silanes, and Shleifer propose a new model of homeowners insurance, in which consumers can cheat and make invalid claims and firms can cheat and deny valid claims. In this environment, trust and honesty are critical factors that shape insurance contracts and the payment of claims, especially when the disputed amounts are too small for courts. They characterize the equilibrium insurance contracts, and show how they depend on the quality of the legal system and the level of trust. The researchers then bring the predictions of the model to a data set of both country business unit data, and individual claims data, obtained from 28 independently operated country subsidiaries of a multinational insurance company. Gennaioli, La Porta, Lopez-de-Silanes, and Shleifer investigate the incidence of claims, the disputes over claims, the rejections of claims, and the payment of claims in this data, as well as the cost and pricing of insurance. Particularly for trust, the evidence is broadly consistent with the predictions of the model. Cultural factors appear to shape insurance markets in economically meaningful ways, just as they shape other spheres of human activity.
This paper was distributed as Working Paper 27189, where an updated version may be available.
Despite sharply rising prices, the number of companies choosing to sell private long-term care insurance (LTCI) has dropped from over 100 to just over 30 today. This paper shows that regulators' political incentives had a significant effect on both prices and insurer participation in the LTCI market. Liu and Liu find that four attributes of the state regulator - election cycle, political capital, political affiliation, and campaign funding - significantly affected price changes. Furthermore, companies operating in states with more political frictions earn only .48 times the profits of their counterparts and are more likely to drop out of the market. To quantify equilibrium effects of the regulator in the LTCI market, the researchers develop and calibrate a dynamic structural model. Using counterfactual simulations, Liu and Liu estimate that if regulators only cared about consumer surplus and faced no political frictions, social welfare would increase by roughly $228 million per year.
This paper documents the long-run effects of an important reform of capital regulation for U.S. insurance companies in 2009. Becker, Opp, and Saidi show that its design effectively eliminates capital requirements for non-agency MBS, implying an aggregate capital relief of over $18bn at the time of the reform. By 2015, 40% of all high-yield assets in the overall fixed-income portfolio are MBS investments. This result is primarily driven by insurers' reduced propensity to sell poorly-rated legacy assets. Using a regression discontinuity framework, they can attribute this behavior to capital requirements. The researchers also provide evidence that the insurance industry, driven by large life insurers, crowds out other investors in the new issuance of (high-yield) MBS post reform.
Climate change is leading to significant increases in destructive weather events, especially wildfires. In this paper, Stanton, Vergara-Alert, Wallace, and Issler study wildfires in California from 2000-2018 using a comprehensive data set of houses and mortgages in California, merging data on fires, mortgages, property characteristics, and weather. Using a difference-in-differences approach, confirmed via panel regression, Stanton, Vergara-Alert, Wallace, and Issler find a significant increase in mortgage delinquency and foreclosure after a fire event. More surprisingly, they find that default and foreclosure decrease in the size of the wildfire. The researchers argue that this second result arises from the coordination externalities afforded by large fires, whereby county requirements to rebuild to current building codes and insurance-covered losses work together to ensure that rebuilt homes will be more valuable than they were pre-fire. This is only true as long as there exists a well-functioning insurance market, and the size of recent losses, combined with regulatory distortions in the market, casts doubt on the continued ability of insurance companies to absorb fire-related losses. The technology in this paper can help by providing the California Department of Insurance (CDI) and other insurance regulators with a framework for building benchmark models to evaluate proposed insurance-company models, much like the bank stress-testing carried out by the Federal Reserve System.
Big data, machine learning and AI inverts adverse selection problems. It allows insurers to infer statistical information and thereby reverses information advantage from the insuree to the insurer. In a setting with two-dimensional type space whose correlation can be inferred with big data Brunnermeier, Lamba, and Segura-Rodriguez derive three results: First, a novel tradeoff between a belief gap and price discrimination emerges. The insurer tries to protect its statistical information by offering only a few screening contracts. Second, Brunnermeier, Lamba, and Segura-Rodriguez show that forcing the insurance company to reveal its statistical information can be welfare improving. Third, they show in a setting with naïve agents that do not perfectly infer statistical information from the price of offered contracts, price discrimination significantly boosts insurer's profits. The researchers also discuss the significance their analysis through three stylized facts: the rise of data brokers, the importance of consumer activism and regulatory forbearance, and merits of a public data repository.