Nanyang Business School Forum on Risk Management and Insurance
Risk-Revealing Contracts for Government-Sponsored Microinsurance in China
In January of 2016, China’s State Council announced the merger of the nation’s two major health insurance programs for low-income residents – the New Cooperative Medical Scheme (NCMS), which serves rural citizens, and the Urban Residents’ Basic Medical Insurance system (URBMI), which serves city dwellers – into the newly created Urban and Rural Residents’ Basic Medical Insurance (URRBMI) system. Like its two predecessor organizations, the URRBMI is a microinsurance facility organized by China’s central and provincial governments (see Wang et al., 2014, and Zhu et al., 2017).
In recent years, microinsurance products have become an important component of life, medical, property, and agricultural risk transfer in developing economies. By the end of 2013, it was estimated that such products covered approximately 500 million people worldwide. Designed primarily to address the basic needs of low-income individuals and households (see, e.g., Churchill, 2007), microinsurance is characterized by premium volumes and profit margins that are very small compared to those of conventional insurance. Consequently, such programs are frequently offered by non-profit entities (including government agencies and NGOs) and/or socially motivated insurance companies focused more on sustainability than profitability. China’s URRBMI operates as a non-profit government monopoly with voluntary policyholder participation.
One major problem faced by voluntary insurance programs is that of adverse selection; that is, higher-risk individuals and households are more willing to purchase a given contract than their lower-risk counterparts, which makes it difficult for insurance providers – whether competitive or monopolistic – to achieve a sustainable market. In the case of a competitive market, adverse selection imposes obstacles to stable equilibrium (see Rothschild and Stiglitz, 1976). If the insurer is a monopolist, then the major challenge is to offer contracts at a reasonable price for lower-risk buyers (i.e., without excessive cross subsidies of higher-risk buyers by lower-risk buyers).
In conventional insurance markets, adverse selection is usually mitigated by policyholder underwriting, in which insurance companies hire experienced employees and/or intermediaries to assess the loss probabilities of potential buyers. However, this approach generally is not applicable in microinsurance, because underwriting costs cannot be sustained by the small premiums and limited profitability. Such is the case with the URRBMI, which employs no underwriting based upon buyer risk levels.
Previous research on the NCMS and URBMI has demonstrated the presence of adverse selection in both programs (see, e.g., Eggleston, 2012, and Milcent, 2018). Although the problem may appear to have been resolved over time by attracting high policyholder subscription rates (over 95 percent of qualified residents in both the NCMS and URBMI; see, e.g., Yu, 2015), this would ignore the true cost of adverse selection: the government subsidies required to attract lower-risk buyers to the programs by offsetting the crosssubsidies from lower-risk buyers to higher-risk buyers. If the government were not willing to provide these subsidies, then some of the lower-risk buyers simply would decline to participate. Although limited data from the URRBMI are currently available (see http://www.nhfpc.gov.cn), it is clear that, despite high subscription rates, the new program is likely to inherit the adverse-selection issues of the NCMS and URBMI.
Our present work addresses the problem of adverse selection for the URRBMI through a practical, risk-revealing contract mechanism. Specifically, we modify the classic competitive-market framework of Rothschild and Stiglitz (1976; hereafter, “RS”) to apply to a non-profit monopoly, and then show how a variety of buyer types – differentiated by their unknown loss probabilities – can be encouraged to reveal their inherent risk levels by selecting from a menu of actuarially priced contracts. In an unsubsidized program, these contracts would allow the insurer to “break even” in the sense that total premiums cover total expected losses; whereas in a subsidized program, they enable the insurer to apply a constant subsidy factor to all buyer types (rather than having to give additional subsidies to lower-risk policyholders). Finally, we use province-level data from the NCMS to illustrate the proposed approach.
The complete paper is available at: