Public Policy Bulletin (7th Issue - December 2023)

2 A 2008 survey conducted in Hong Kong estimated that 280,500 community-dwelling elderly individuals required assistance in daily living, but only 47.6% were served by caregivers, of whom 65.6% received care from family members and 25.6% were supported by domestic helpers. (Census and Statistics Department, HKSAR 2009). The current situation, in which the lion’s share of total LTC costs are paid out of pocket, represents a catastrophic financial risk for older adults. Conventional economic theories suggest that its high cost and low probability of being needed make LTC dependency an ideally insurable risk, but it faces a paradox that reflects a remarkably underdeveloped LTC insurance market, even in fully mature market economies. On the other hand, an individual's decision to purchase insurance is determined not only by self-assessment of risk and anticipated needs but also by the structural features— coverage, premium, and term length of insurance plans. Previous studies have found that income, financial status, and bequest motives are closely related to individuals’ decisions to purchase LTCI. The value individuals place on LTCI depends on their beliefs about their future care needs. Some factors have been shown to crowd out the wish to purchase LTCI, such as the presence of children who can provide informal care. Hong Kong is an ideal place in which to study and experiment with private LTCI to offer useful guidance to other aging societies; apart from its superaging characteristic, the city is an international finance hub with a highly developed insurance industry where the extensive penetration of private insurance reflects wide public acceptance. Study Methodology This study used actuarial projection to compute the attributes that indicate willingness to purchase and pay the premiums charged for LTCI plans; the hypothetical plans were built on Anderson’s behavioral model of health services. A discrete choice experiment was then conducted to determine which of the hypothetical plans was most appealing to middle-aged respondents. Because the COVID-19 pandemic was in progress during this investigation, a telephone survey was adopted instead of a standard household survey. A tobit model was used for multivariate analysis and a robustness check was performed to corroborate the statistical results using a fractional probit model. The three groups of factors illustrated in Anderson’s model have been applied widely in LTCI studies; these are predisposing, enabling, and needs factors. Predisposing factors, such as age, gender, and belief, exist before care uptake but may influence behaviors related to purchase decisions. Enabling factors facilitate or inhibit the purchase of LTCI, and these include income, assets, financial literacy, and cognitive capacity. Needs factors represent perceived or real needs for LTC that directly influence purchase decisions. Plan Parameters Plan 1 Plan 2 Monthly benefit HK$20,000 HK$20,000 Inflation protection 3% Annual growth Nil Years of premium contribution 15 Years (age 50-65) 35 Years (age 30-65) Monthly premium HK$1,237 (men) HK$2,150 (women) HK$180 (men) HK$298 (women) Table 1. Hypothetical LTCI plans designed for the study Employing a two-stage stratified sampling method for this study, invitations were first sent to a random sample of household addresses drawn from registered quarters and segment samples were obtained from the Census and Statistics Department. A household member aged 40-59 was then chosen using the “next birthday” method to ensure unbiased random selection. The sample in this study closely resembles the overall demographic characteristics of the Hong Kong population, as those aged 50-59 years in Hong Kong occupy 51% of the middle-aged group (40-59 years), and women account for 57% of the entire middle-aged group. The key features of both plans were presented to the respondents, who were instructed to indicate their willingness to subscribe on a 0%-100% scale. All those who did not indicate full willingness (did not select 100% on the scale) were asked to choose from a list of reasons, including dissatisfaction with the terms, lack of available alternative funding, distrust of commercial insurers, the belief that it is government’s responsibility to finance LTC, the desire to reserve funding for bequests, lack of familiarity with private LTCI, and others. The predisposing factors were presented using binary variables: sex (1=female), age (1=50-59 years old), marital status (1=married), parenthood (1=having at least one child), education (1= some tertiary education), living arrangement (1= living alone), availability of family care (1=available). Four attitudinal statements about LTCI were included to indicate bequest motives, formal care preferences, the principle of familial obligation, and concerns about insurers. Findings and Analysis Table 2 summarizes the basic statistics extracted from the survey. The preference for Plan 1, as shown in the table, is higher than that for Plan 2, indicating that middle-aged adults find the primary insurance product generated for the discrete-choice experiment more appealing than other cohorts do. Around one-fifth of the respondents attributed their lack of interest to their Private Long-Term Care Insurance in a Super-Aging Society: A Purchase Motivation Study in Hong Kong Public Policy BULLETIN