The Chinese population is ageing and will continue to age dramatically. The United Nations projects that the percentage of Chinese people aged 60 years or above was 12.4% (168 million people) in 2010 and will increase to 28% (402 million) by 2040 [1
]. The pace of population aging in China is much faster than that in other developed and developing countries [2
]. Together with this demographic change, China is simultaneously witnessing great socioeconomic transition. Accelerated by the one-child policy, the so-called 4-2-1 family structure (i.e., a family constituted by four grandparents, two parents and one child) has become the main stream family structure in urban China [3
]. This socioeconomic transition weakens the traditional familial duties of caring and supporting the elderly [4
]. As a result, a huge number of elderly Chinese are now choosing to live alone [5
]. Given that long-term care is quite costly in developed countries [6
], the above demographic and socioeconomic shift makes the financing of long-term care a significant concern for policymakers in China.
Long-term care insurance (LTCI) is acknowledged as the most desirable policy choice among the existing public financing models of long-term care in China [7
], which is also highly recommended for a middle-income country familiar with the public insurance system for the financing of long-term care [8
]. In June 2016, the Ministry of Human Resources and Social Security in China issued a document “Guidance on Pilot Cities to Launch Long-Term Care Insurance”, which signified the official initiation of LTCI in China. A total of 15 cities were designated as pilot cities. Based on the experience in the pilot cities, China aims to formally design the policy framework of LTCI by 2020. The policy recommendations made for pilot cities to design LTCI’s target participants, eligible criteria, financing mechanism, and benefit package include the following focal points: (1) The participants of LTCI are in principle those covered by Urban Employees Basic Medical Insurance (UEBMI), a public health insurance covering the employed urban population; (2) It is the severely disabled to whom LTCI mainly provides financial protection; (3) In the pilot stage, it suggested that LTCI raise funds by optimizing the structure of UEBMI funds, transferring UEBMI pooled funds, and adjusting the contribution rate of UEBMI and so on; (4) LTCI is advised to pay 70% of the costs that meet the requirements of reimbursement; (5) Pilot cities are encouraged to gradually expand participation and relax the restrictions on eligibility based on their own circumstances, to explore the multi-channel financing mechanism step by step, and to enhance the benefit package on a gradual basis according to the development of the economy [9
One of the greatest challenges in the development of China’s LTCI is the mobilization of sufficient funds [10
]. In the trial run, the financing channel of UEBMI funds, which is heavily emphasized in the guidance, is convenient for China to gain experience in LTCI. However, it is not a sustainable channel [10
]. Therefore, the information on how much the participants are willing to pay for LTCI is crucial for Chinese policy makers to design a sustainable financing policy for LTCI.
Two methods are usually used to estimate willingness to pay (WTP): a revealed preference approach and a stated preference approach. The former is based on actual choices. The latter is based on hypothetical choices in surveys and is widely used to estimate the monetary value of a non-marketed commodity, such as health care [12
]. The contingent evaluation method (CVM) is one of the most frequently used stated preference approaches [13
]. With regard to health insurance, revealed preferences data can only be obtained from post-scheme design studies. Thus such studies are rarely used to give policy implications in the design of the schemes [14
]. In low- and middle-income countries, a large number of papers have been published relying on CVM to elicit WTP of the proposed health insurance, aiming at informing policy makers about the financing design of the schemes [15
], among which several were based in China [22
However, we could identify only a small number of papers focusing on WTP for LTCI globally. Some researchers relied on CVM to estimate WTP for LTCI in Japan [24
] and Spain [25
]. Other used a discrete choice experiment, another stated preference approach, to explore WTP for LTCI in Italy [26
] and the U.S. [27
]. However, until now, no papers have been published on WTP for LTCI in China. Only two papers were identified that were related to LTCI in China. One paper qualitatively evaluated the emerging models to finance long-term care for the aging population [7
]. The other was concerned with the factors associated with the preference for public or private LTCI plans and the level of the appropriate premium of each plan [28
]. The latter paper did not focus on how much respondents were willing to pay for LTCI, but on the premium that they consider appropriate, which is different from their WTP. In addition, the latter paper did not describe the benefit package of LTCI plans. This limitation affects the results from that paper since respondents were asked to express their views on the appropriate premium of LTCI plans without knowing what these plans offered them.
This study aims at filling this gap by estimating WTP for LTCI in China using CVM and exploring the determinants of demand for LTCI in China.
This study makes an important contribution to the available literature as it is one of the very few studies exploring WTP for LTCI in China. The main results in this paper were that more than 90% of the respondents expressed their willingness to buy LTCI and the median WTP accounted for 2.29% of average annual per capita disposable income. Our finding is best to be interpreted with the current individual contribution rates for LTCI in other developed countries. For example, in Germany the contribution rate for LTCI (split by employers and employees) is 2.55% for parents with children and 2.8% for childless citizens— if older than 23 years—of gross income [8
]; the individual contribution rate is 0.9% of gross income for Japanese adults aged 40–64 [36
]. We also found that the vast majority of pilot cities mainly rely on UEBMI funds as the financing source for LTCI. China’s LTCI started with a limited benefit package and strict eligibility rules, which is the recommended way for middle-income countries to establish LTCI [8
]. As the guidance from the central government foresees, China will gradually expand LTCI [9
]. Our findings are very enlightening, especially since China’s LTCI is in the initial period of development and is facing the challenge of financing sustainability [10
]. Based on the WTP estimated in this study, we suggest that it is acceptable, from the participants’ perspective, that individual contribution is considered an important source of mobilizing funds for LTCI in China. We also demonstrated that among different age groups, the older population (65 or above) revealed less demand for LTCI than the younger population (35 or below), but no significant difference from the population in the middle age group (35 to 64). These results are similar to one systematic review on the demand for health insurance indicating that age was negatively correlated with WTP for health insurance [37
] and similar to the previous study showing that older individuals in the U.S were no more likely to reveal a higher possibility of buying LTCI [27
]. However, these results are contradictive to the previous study based in Spain revealing that age was positively associated with the demand for LTCI [26
]. Generally, the older an individual is, the higher the likelihood that one could benefit from LTCI at average premiums. However, since the younger population was born when the one-child policy was strictly implemented in China, they may already realize the importance of LTCI for older and disabled people and thus display a high demand for LTCI. At the time being, there is a heated debated concerning the starting age of collecting individual contributions for LTCI in China [10
]. Considering that the demand for LTCI is relatively high among the younger population, we suggest that it is feasible and reasonable to collect the individual contribution for LTCI at a young age.
In line with previous studies on demand for LTCI [24
] and on health insurance [15
], we found that the demand for LTCI increased as income, a proxy of ability to pay, increased. The findings that demand for LTCI among the lowest income quartile was quite low were worrisome since this population segment is already the most vulnerable group. Additionally, deprivation of the social protection gained from LTCI will worsen their social-economic status and may let them fall into deeper poverty. These findings suggest that the Chinese government needs to provide proper subsidies to the low-income population who cannot afford the insurance premium. In addition, the positive correlation detected between education and the demand for LTCI aligns with the previous evidence on the demand for LTCI [26
] and on health insurance [15
]. The findings could be attributed to the fact that individuals with higher education usually have a stronger awareness of the actual risk of the costs induced by future long-term care.
Another important finding in this study was that the demand for LTCI in Qinghai (located in western China) showed no significant difference from that in Zhejiang (located in eastern China). China has started to establish LTCI even though China as a whole lacks long-term care providers [7
]. LTCI can serve as a stimulus for the development of long-term care delivery systems in middle-income countries [8
]. However, a long-term care delivery system will be more difficult to develop in western provinces than in eastern provinces in China, since the western provinces are less developed. To meet the high demand for LTCI in Qinghai, the government has already made and will in the future make more efforts to strengthen affordable high quality long-term care in such provinces [38
]. In contrast with prior studies on the demand for health insurance [18
], suffering from chronic conditions, a representative of poor health status, was not a significant factor associated with the demand for LTCI. These results indicate that when controlling other variables, such as age, income, education, price and so on, health status is not significantly associated with the demand for LTCI in the settings of this study. However, further studies are needed in order to more deeply explore the relationship between health status and the demand for LTCI in other settings in China.
A few limitations of this study need to be acknowledged. First, the starting bid bias, i.e., the fact that the first bid biases respondents’ answers to the subsequent bids, may incur when using the bidding game method. Considering that the median WTP for LTCI estimated in this study was above the starting bid, respondents’ WTP is possibly biased downwards [22
]. The starting bid in this study was not varied due to practical constrains. Further study is needed to account for the effect of the starting bid in the estimation of WTP for LTCI. Second, though this paper was a part of the study on the public health insurance system, the LTCI designed in this study did not clearly point out whether it was public or private. The public LTCI plan was found to be more popular than the private in terms of participation and contribution in China [28
]. Thus, it can be inferred that if we had clearly pointed out the public nature of LTCI, the actual WTP would have been even higher than the estimates we obtained from this study. Third, the coefficients from our model can be used to estimate the potential enrollment size of LTCI under different prices for sub-groups defined by the independent variables (e.g., age, gender, income). However, the gap always exists between respondents’ behavior in a hypothetical market and in real life. Thus, more research is needed in order to set an appropriate premium for LTCI in China. Fourth, respondents may underreport the annual household income due to recall bias. So the ratio of an individual’s WTP for LTCI to the annual per capita income may be biased upwards. Fifth, we used literature review and interviews with key policy makers to develop our WTP questionnaire. Thus, the hypothetical LTCI designed in this study did not reflect the expectations of the potential participants of LTCI. Further study is needed in order to incorporate community preferences for LTCI into related studies. Meanwhile, the design of the hypothetical LTCI in this study did not account for the government subsidy. Currently, half of LTCI schemes do not include government subsidy as the financing source of LTCI. However, further study is still needed to understand how respondents’ attitude towards government subsidies in LTCI. In addition, comparing the LTCI designed in this study with the current LTCI schemes in pilot cities, our LTCI had a larger benefit package and broader eligibility since it featured similar copayment as those in the current schemes, but without a payment ceiling and without eligibility in terms of the severity of disability. Though our data was collected in 2010, we still believe that our findings have great policy implications for the development of LTCI in China; China’s LTCI will gradually expand its benefit package and relax its restrictions on eligibility, but the fact remains that it is experimenting with a suitable sustainable financing mechanism [10
]. However, one needs to be cautious in applying the median WTP for LTCI estimated in this study into reality, since the average per capita income, consumer price index and so on are rising year by year. Last, this study addressed the question of whether respondents were willing to buy a designed LTCI under different bids. However, in reality the decision on whether to invest in LTCI is related to how households set priorities on household consumption. More study is needed to explore the trade-off between the investment in LTCI and other aspects of household consumption, such as childcare, purchasing property, etc.