The impact of tax-subsidized health insurance on health and out-of-pocket burden in China

Abstract

To achieve a tiered healthcare system in China, the government has beefed up efforts to encourage people to embrace private health insurance. One of the steps is to introduce tax-subsidized health insurance (TSHI) in 2015. This paper assesses whether TSHI is successful in improving health and easing financial burden. Using unique data from the Domestic TSHI Consumers Evaluation Survey in 2019 and Propensity Score Matching method, we find that TSHI has been effective in improving participants’ health status, but not in lowering out-of-pocket (OOP) burden. Besides, TSHI can increase the utilization of both inpatient and outpatient care, which provides a possible explanation for beneficial health effect. From an internal perspective, TSHI has the potential to reduce disparities in health and access to healthcare services. Our results provide implications on modifying the benefit design of TSHI to better complement basic medical insurance system and on reforming the current tax incentive policy to reach out a large number of low-income people.

Appendices

Appendix 1

Studies that estimate the impact of medical insurance on health outcomes, OOP burden, and utilization of medical care

Paper	Nation and program	Health outcomes	OOP burden	Utilization
Feldstein and Gruber (1995)	USA, “major risk policy” approach		A health insurance plan that has a 50% co-insurance but limits OOP spending to 10% of income can substantially reduce total medical spending
Keeler et al. (1988)	USA, RAND HIE			Cost sharing, including co-insurance and deductibles reduces use of medical services
Newhouse (1993)	USA, RAND HIE	No significant effects on a wide range of measures of health status for the average patient, except for persons with poor vision or elevated blood pressure	Health insurance without co-insurance leads to more people using services and to more services per user, referring to both outpatient and inpatient service
Fihn and Wicher (1988)	USA, outpatient services at the Seattle VA Medical Center	Cancelation of VA health benefits associated with increase in blood pressure
Perry and Rosen (2004)	USA, employer-based health insurance	No differences in self-reported health status or in the probability of a number of conditions between the self-employed and wage earners
Doyle (2005)	USA, private health insurance	The uninsured have a substantially higher mortality rate		The uninsured receive 20% less care
Lichtenberg (2002)	USA, Medicare	Increases the survival rate of the elderly		Increases visits to physician’s offices and utilization of ambulatory and inpatient care
Card et al. (2009)	USA, Medicare	Decreases mortality of ED admission		Increases number of procedures and total list charges
Finkelstein and McKnight (2008)	USA, Medicare	Medicare has no impact on overall elderly mortality in its first 10 years	Substantially reduces the elderly’s OOP risk
Engelhardt and Gruber (2011)	USA, Medicare		The extension of Medicare Part D is associated with sizeable reductions in OOP spending
Currie and Gruber (1996a)	USA, Medicaid	Medicaid eligibility significantly reduces children’s mortality		Significantly increases the utilization of medical care, particularly care delivered in physicians' offices
Currie and Gruber (1996b)	USA, Medicaid	Expanded eligibility for pregnant women lowers the incidence of infant mortality and low birth weight
Goldman et al. (2001)	USA, Medicaid	Health insurance can have a dramatic effect on mortality among patients with HIV
Currie et al (2008)	USA, Medicaid	Medicaid eligibility has little effect on children’s current health status, but eligibility in early childhood has positive effects on future health		Eligibility for older children improves current utilization of preventive care
Finkelstein et al. (2012)	USA, Medicaid	Extending access to Medicaid can improve self-reported physical and mental health	Medicaid lowers OOP medical expenditures and medical debt	Medicaid can induce higher healthcare utilization (including primary and preventive care as well as hospitalizations)
Baicker et al. (2013)	USA, Medicaid	The 2008 Medicaid expansion in Oregon generates no significant improvements in measured physical health outcomes in the first 2 years, but raises rates of diabetes detection and management, and lowers rates of depression	Reduces financial strain	Increases use of healthcare services
Taubman et al. (2014)	USA, Medicaid			The Medicaid expansion in Oregon increases overall emergency use by 0.41 visits per person, or 40% relative to an average of 1.02 visits per person in the control group
Goodman-Bacon (2018)	USA, Medicaid	Medicaid’s introduction reduces mortality rates among nonwhite infants and children in the 1960s and 1970s
Hanratty (1996)	Canada, National Health Insurance	Reduces infant mortality by 4% and low birth weight by 1.3%
Lei and Lin (2009)	China, New Cooperative Medical Scheme	No improvement in health status, measured by self-reported health status and by sickness or injury in the past 4 weeks	No reduction in OOP expenditure	Increases the utilization of preventive care, particularly general physical examinations
Chen and Jin (2012)	China, New Cooperative Medical Scheme	NCMS does not affect child morality and maternal mortality, but help improve the school enrollment of 6-year-olds
Wagstaff et al. (2009)	China, New Cooperative Medical Scheme		No reduction in overall OOP expenses, OOP expenses per outpatient visit or inpatient spell	Increases outpatient and inpatient utilization
Yip and Hsiao (2009)	China, New Cooperative Medical Scheme		NCMS is not effective at addressing medical impoverishment, because it overlooks spending on chronic diseases
Hou et al. (2014)	China, New Cooperative Medical Scheme		NCMS does not increase financial protection	NCMS is effective in increasing access to care
Li et al. (2014)	China, New Cooperative Medical Scheme		NCMS fails to prevent catastrophic health expenditure and medical impoverishment
Cheng et al. (2015)	China, New Cooperative Medical Scheme	NCMS improves the elderly enrollees’ activities of daily living and cognitive function, but does not lead to better self-assessed health status	There is no evidence that the NCMS has reduced their out-of-pocket spending	The elderly participants are more likely to get adequate medical services when sick
Cheung et al. (2016)	China, New Cooperative Medical Scheme		The NCMS induces an increase in OOP health expense for mild illness and a decrease in health payments for more serious illnesses; The NCMS also leads to a higher incidence of catastrophic healthcare spending
Zhou et al. (2017)	China, New Cooperative Medical Scheme	There is little evidence that NCMS expansion contributes to mortality decline
Pan et al. (2013)	China, Urban Resident Basic Medical Insurance	URBMI has a positive effect on health	The insured do not pay more for the services	The insured receive more and better inpatient care
Liu and Zhao (2014)	China, Urban Resident Basic Medical Insurance		URBMI does not reduce total OOP health expenses	URBMI increases the utilization of formal medical services, especially for children, low-income families and residents in the poor region
Zhang et al. (2017a, b)	China, three types of Basic Medical Insurance Schemes			NCMS membership is not significantly correlated with outpatient service utilization; access to three types of Basic Medical Insurance Schemes is positively related to inpatient service utilization
Zhang et al. (2017a, b)	China, three types of Basic Medical Insurance Schemes and other insurances		Health insurance reduces inpatient OOP expenditure; OOP spending on pharmaceuticals is reduced for inpatient care but not for outpatient care	Having health insurance increases the likelihood of utilizing healthcare
Wagstaff and Lindelow (2008)	China, three types of health insurance schemes		China's government and labor insurance schemes increases financial risk associated with household healthcare spending, but rural cooperative medical scheme significantly reduces financial risk in some areas but increased it in others; China's new health insurance schemes have also increased the risk of high levels of OOP spending
Guan and Tena (2018)	China, NCMS and URBMI	Social medical insurance schemes have no effect or even a marginally negative effect on children's health status in some cases		Participating in social medical insurance schemes significantly increases children's yearly hospital use, especially for children who come from rural China
Shigeoka (2014)	Japan, Elderly Health Insurance	Lower cost sharing does not improve health measures such as mortality and self-reported physical and mental health	Lower cost sharing yields reduction in OOP expenditure	Reduction in cost sharing increases healthcare utilization for both outpatient visits and inpatient admissions
Grignon et al. (2008)	France, free complementary health insurance plan			The plan is lack of an overall effect on utilization
Hullegie and Klein (2010)	Germany, private health insurance	Positive effects of private health insurance on health		Negative effects of private health insurance on the number of doctor visits, no effects on the number of nights spent in a hospital
Montoya Diaz et al. (2020)	Brazil/Egypt/India, private health insurance		PHI fails to address the problem of high OOP payments on health care in the three countries, because it is beyond the financial reach of those most in need of access to health care and financial protection
Hurley and Emmanuel Guindon (2020)	Canada, private health insurance		Except for drug sector, where it covers a large number of people not covered by public insurance programs, PHI contributes in only a minor way to financial protection	People with private drug insurance tend to use more publicly financed physician services

Appendix 2

Domestic TSHI Consumers Evaluation Survey Questionnaire

Part A Demographic Backgrounds

1. What is your age? ______ years old.

2. What is your gender?

A. Male B. Female

3. What is your marital status?

A.Unmarried B. Married C. Divorced D. Widowed

4. What is the highest degree or level of education you have completed?

A.Junior high school B. High school C. Bachelor's degree D. Master's degree or higher

5. Your after-tax monthly income is about ______ yuan.

6. What is your current employment status?

    A. Work in state-owned enterprises

    B. Work in public institutions

    C. Work in private enterprises

    D. Freelancers

    E. Rural migrant workers

    F. Agricultural workers

   G. Unemployed

   H. Retired

   I. Other

7. Which province do you currently live in?

    (1) Beijing (2) Tianjin (3) Hebei (4) Shanxi (5) Inner Mongolia (6) Liaoning

    (7) Jilin (8) Heilongjiang (9) Shanghai (10) Jiangsu (11) Zhejiang (12) Anhui

    (13) Fujian (14) Shandong (15) Henan (16) Hubei (17) Hunan (18) Jiangxi.

    (19) Guangdong (20) Guangxi (21) Hainan (22) Chongqing (23) Sichuan (24) Guizhou

    (25) Yunnan (26) Tibet (27) Shaanxi (28) Gansu (29) Qinghai (30) Ningxia

    (31) Xinjiang

8. What is your area of residence?

    A. Urban area B. Rural area

Part B Health Status and Medical Insurance

9. In general, how would you rate your current health status?

    A. Poor B. Average C. Good D. Very Good

10. During the last 12 months, how many times did you visit a hospital, health center or clinic for outpatient care?

      A. 0–5 times B. 6–10 times C. 11–20 times D. 21 times or more

11. What was the total cost of outpatient care for the last 12 months? (Includes out-of-pocket part and reimbursement part)

      A. 0–500 yuan B. 501–1000 yuan C. 1001–2000 yuan D. 2001–5000 yuan

      E. 5001 yuan or more

12. What was the reimbursement rate of basic medical insurance for the outpatient care?

     A. 0–10% B. 11–20% C. 21–30% D. 31–40% E. 41–50% F. 51–60% G. 61–100%

     H. 71–80% I. 81–90% J. 91–100%

13. During the last 12 months, how many days did you stay in a hospital to receive inpatient care?

     A. 0–5 days B. 6–10 days C. 11–20 days D. 21–40 days E. 41 days or more

14. What was the total medical cost for all the inpatient care you received during the last 12 months? (Includes out-of-pocket part and reimbursement part)

     A. 0–2 thousands B. 2–5 thousands C. 5–10 thousands D. 10–20 thousands

     E. 20–50 thousands F. 50–100 thousands G. 100–200 thousands H. more than 200 thousands

15. What was the reimbursement rate of basic medical insurance for the inpatient care?

     A. 0–10% B. 11–20% C. 21–30% D. 31–40% E. 41–50% F. 51–60% G. 61–100%

     H. 71–80% I. 81–90% J. 91–100%

16. What was the percentage of your out-of-pocket medical expenses (including inpatient and outpatient care) during the last 12 months to the total household expenditure?

     A. 0–20% B. 21–40% C. 41–60% D. 61–80% E. 81–100%

17. How often do you take a physical examination?

     A. Once every 6 months B. Once a year

     C. Once every 2 or 3 years D. Never took a physical examination in last 5 years

18. What is your basic medical insurance type?

      A. Free medical insurance

      B. Urban Employee Basic Medical Insurance

      C. Urban and Rural Residents Basic Medical Insurance

      D. No basic medical insurance

19. Do you think the basic medical insurance can meet your healthcare needs?

     A. Yes (Skip to q. 21) B. No (Skip to q. 20) C. Do not know (Skip to q. 21)

20. What is the main reason you think basic medical insurance cannot meet your needs?

     A. Low reimbursement rate B. Restricted benefit list C. Hard to deal with critical illness

     D. Other, please specify ______

21 . By which one of the following methods would you prefer to finance out-of-pocket medical expenses?

      A. Engage in precautionary savings B. Borrow from friends or relatives

      C. Purchase private health insurance coverage D. Other, please specify______

Part C Tax-subsidized Health Insurance

22. Do you know of Tax-subsidized health insurance, whose premium can be deducted before personal income tax, with a limit of 2400 yuan per year?

     A. Yes (Skip to q. 23) B. No (Skip to q. 24)

23. Where did you learn about TSHI?

     A. Promotional materials from insurance companies

     B. Advices from insurance brokers

     C. On TV, radio, newspapers and magazines

     D. On the Internet, social media apps like WeChat or Weibo

     E. At the workplace or community

     F. Learn from families or friends

     G. Other

24. Which of the following features of TSHI appeals to you? (Choose all that apply)

     A. Premium is tax deductible, with a limit of 2400 yuan per year.

     B. TSHI covers pre-existing diseases.

     C. Besides medical payments coverage, TSHI provides individual accounts for investment, with the settlement interest rate to be at least 2.5%.

     D. TSHI is a guaranteed renewal policy until the age of retirement.

     E. TSHI has no deductibles or waiting period.

     F. TSHI covers specified drugs, items, or medical materials not covered by basic medical insurance.

     G. None of the above.

25. Do you have TSHI?

     A. Yes (Skip to q. 26) B. No (Skip to q. 31)

26. Where did you buy TSHI?

     A. Through employers

     B. Through direct writers

     C. Through brokers or agents

     D. On the Internet, or mobile apps

     E. Through telemarketing channels

     F. Other

27. How much did you pay for TSHI?

     A. 1–200 yuan B. 200–400 yuan C. 400–700 yuan D. 700–1000 yuan E. 1000–1500 yuan

     F. 1500–2000 yuan G. 2000–3000 yuan H. 3000–4000 yuan I. 4000–5500 yuan

    J. more than 5500 yuan

28. How would you like to rate the premium of TSHI?

     A. The premium was higher than expected.

     B. The premium was appropriate.

     C. The premium was lower than expected.

     D. Do not know.

29. What was the coverage limit of your TSHI?

     A. 200 thousand yuan per year

     B. 250 thousand yuan per year

     C. 300 thousand yuan per year

     D. Other, please specify ______

30. Would you like to renew your TSHI policy?

     A. Yes B. No

31. Do you have any suggestions for the benefit design of TSHI? (Choose all that apply)

     A. Raise the tax deduction limit of 2400 yuan per year, or set no limit.

     B. Raise the coverage limit of 200–300 thousand yuan.

     C. Enlarge the list of covered drugs, items, and medical materials that are not paid by basic medical insurance.

     D. Streamline the tax deduction procedure of TSHI premium.

     E. Other, please specify ______.

Appendix 3

Change of Sample Sizes

The questionnaires were delivered across the nation with the assistance of the Insurance Association of China in August 2019, and after 2 months, 15,885 questionnaires were returned by respondents in total.

Then data cleaning was performed in the following steps. First, the observations that took less than 60 s (5% quantile) to answer the questions were dropped from the sample. Then, our dataset was restricted to respondents from age 18 to 65, as TSHI is not available for children or the elderly. We further discarded observations with missing information on key variables such as after-tax monthly income and proportion of OOP medical expenses in total household expenditure. We then excluded respondents whose after-tax monthly income was beyond the quantile (1%, 99%). Our final sample consisted of 11,939 observations.

In PSM, we employed one-to-one nearest neighbor matching within a caliper distance of 0.01 to match each participant with a non-participant. Further, observations not in the common support area were dropped from the sample in order to satisfy the common support condition. Therefore, after matching only 2744 observations were kept in the sample.

Appendix 4

See Tables 11, 12, 13, 14, 15, 16, 17 and 18.

Table 11 Interval regression results for OOP burden and medical care utilization

Table 12 Impact of TSHI on health and OOP burden (using heteroskedasticity-based instruments)

Table 13 Regressions on subsamples

Table 14 Impact of TSHI on health using different matching methods

Table 15 Impact of TSHI on OOP burden using different matching methods

Table 16 Participants' willingness to renew the policy

Table 17 Health and OOP burden by renewal decision

Table 18 Impact of TSHI on health and OOP burden by renewal decision