Health Poverty Alleviation Project in Rural China: Impact on Poverty Vulnerability, Health Status, Healthcare Utilization, Health Expenditures

Figures & data

Table 1 Definition of Variables

Variable	Definition
Dependent variables
Vulnerabilityit	The probability that per capita household income at time t will be below the poverty line at time t+1. 1 for the probability of poverty vulnerability exceeds 0.5, 0 for otherwise.
Health_declineit	1 for the individual i of self-assessed health level at time t+1 less than self-assessed health level at time t, 0 for otherwise.
Independent variables
DHPAPit	1 for individual i in treat group has been affected HPAP in time t, 0 for otherwise.
Age	The age of respondents in the survey year.
Gender	1 for female, 0 for male.
Marital status	1 for married or cohabiting, 0 for otherwise.
Education level	The years of education.
Party membership	1 for party membership, 0 for otherwise.
Ethnic minority	1 for ethnic minority, 0 for otherwise.
Family size	The family number of people living together in a household.
ln(household income)	Total income earned by all family members (in logarithmic form).
Asphalt roads	1 for asphalt roads having been built in a community or village, 0 for otherwise.
Sewer system	1 for sewer system having been built in a community or village, 0 for otherwise.
ln(per capita GDP)	Per capita gross domestic product (GDP, unit: 100 million yuan) in city (in logarithmic form).
Medical beds	The number of medical beds per 10,000 people in province.

Notes: The variable of marital status is set to 1 for married or cohabiting. This emphasize married life as an economic community that lives together and plans families’ financial expenditures together. We set years of schooling for the variable of education level as follows: (1) No formal education (illiterate) for 1.5. (2) Did not finish primary school but capable of reading and/or writing for 3. (3) Elementary school (Sishu/home school) for 6. (4) Middle school for 9. (5) High school (Vocational school) for 12. (6) Two/Three Year College/Associate degree for 15. (7) Four-Year College/Bachelor’s degree for 16. (8) Master’s degree for 18.5; and (9) Doctoral degree/Ph.D. for 22. Both household income and GDP per capita take logarithmic form. The definition of the dependent variables for the impact of HPAP on healthcare utilization and healthcare cost are reported in Table A2. All currency units are in RMB Yuan.

Table 2 The Characteristics of the Interviewees Included from Four Waves CHARLS

Variable	Total (N=40384)	Access to HPAP (N=5946)	Not Eligible for HPAP (N=34438)
Dependent variables
Vulnerabilityit, n(%)	12,625 (31.26%)	2499 (42.03%)	10,126 (29.4%)
Health_declineit, n(%)	33,879 (97.62%, N=34704)	5289 (96.97%, N= 5454)	28,590 (97.74, N=29250)
Independent variables
Age	60.47 (9.80)	63.61 (10.07)	59.92 (9.65)
Female, n(%)	21,019 (52.05%)	2,999 (50.44%)	18,020 (52.33%)
Married, n(%)	35,377 (87.6%)	4,666 (78.47%)	30,711 (89.18%)
Education level	4.94 (3.28)	4.14 (2.95)	5.08 (3.32)
Party membership, n(%)	2,864 (7.09%)	481 (8.09%)	2,383 (6.92%)
Ethnic minority, n(%)	2,648 (6.56%)	598 (10.06%)	2,050 (5.95%)
Family size	2.96 (1.60)	2.73 (1.55)	2.99 (1.61)
ln(household income),	9.21 (1.63)	8.91 (1.32)	9.27 (1.67)
Asphalt roads, n(%)	24,267 (60.09%)	2856 (48.03)	21,411 (62.17%)
Sewer system, n(%)	7160 (17.73%)	618 (10.39%)	6542 (19%)
ln(per capita GDP)	10.49 (0.59)	10.36 (0.55)	10.52 (0.60)
Medical beds	48.94 (10.59)	50.29 (10.76)	48.71 (10.54)

Notes: For the dummy variables, Table 2 reports the number of samples and the percentage of samples that take the value 1. For continuous variables, Table 2 reports the mean and standard deviation. The summary statistics reported in Table 2 is based on the baseline regression sample. The sample size for the variable of health decline is smaller than for the other variables. The reason is that the variable of health decline has been defined since 2013. The Summary statistic of the outcomes variables for the impact of HPAP on healthcare utilization and healthcare cost are reported in Table A3.

Figure 1 Physical examination and inpatient care as the mediator. Path c’ represents the channel of the direct effect on poverty vulnerability. Path a and path b represent the channel of the indirect effect on poverty vulnerability by physical examination and inpatient care.

Figure 2 Flowchart of population selection.

Table 3 The Impact of HPAP on Poverty Vulnerability and Health Decline

Variable	Vulnerability	Health Decline
	(1)	(2)
DHPAPit	−0.0330***	−0.0184**
	(0.0127)	(0.0077)
Age	−0.0052***	−0.0011***
	(0.0003)	(0.0002)
Gender	−0.0909***	−0.0008
	(0.0052)	(0.0028)
Marital status	−0.2559***	0.0196***
	(0.0087)	(0.0045)
Education level	−0.0440***	0.0009**
	(0.0008)	(0.0004)
Party membership	−0.0291***	−0.0008
	(0.0098)	(0.0045)
Ethnic minority	−0.0943***	−0.0011
	(0.0109)	(0.0049)
Family size	0.0096***	−0.0005
	(0.0019)	(0.0009)
ln(household income)	−0.0328***	0.0005
	(0.0018)	(0.0004)
Asphalt roads	−0.0861***	0.0019
	(0.0060)	(0.0030)
Sewer system	−0.2033***	−0.0030
	(0.0067)	(0.0034)
Ln(per capita GDP)	−0.1015***	−0.0096***
	(0.0050)	(0.0025)
Medical beds	−0.0050***	−0.0001
	(0.0005)	(0.0002)
Observations	40,384	34,704
Adj. R^2	0.3987	0.0734

Notes: Standard errors are in parentheses. The meaning of statistical significance symbol represented by ***, ** is as follows: *** denotes p<0.01; ** denotes p<0.05, respectively. We control for individual fixed effects and time of year fixed effects in all estimation results.

Table 4 The Impact of HPAP on Healthcare Utilization

Variable	Outpatient Care	Inpatient Care	Self-Treatment
	(1)	(2)	(3)
DHPAPit	0.0036	0.0934***	0.0410***
	(0.0129)	(0.0125)	(0.0157)
Age	0.0006**	0.0032***	0.0030***
	(0.0003)	(0.0002)	(0.0004)
Gender	0.0240***	0.0040	0.0326***
	(0.0051)	(0.0042)	(0.0065)
Marital status	−0.0133	−0.0106	0.0307***
	(0.0082)	(0.0071)	(0.0102)
Education level	−0.0020**	−0.0007	0.0037***
	(0.0008)	(0.0007)	(0.0011)
Party membership	0.0120	0.0152*	0.0002
	(0.0103)	(0.0089)	(0.0128)
Ethnic minority	−0.0327***	0.0027	0.0263**
	(0.0101)	(0.0084)	(0.0129)
Family size	0.0053***	−0.0014	0.0017
	(0.0018)	(0.0015)	(0.0022)
ln(household income)	0.0014	−0.0014*	0.0033***
	(0.0010)	(0.0008)	(0.0012)
Asphalt roads	0.0009	0.0177***	−0.0253***
	(0.0057)	(0.0047)	(0.0072)
Sewer system	−0.0096	0.0040	−0.0192**
	(0.0071)	(0.0058)	(0.0088)
ln(per capita GDP)	−0.0152***	−0.0172***	−0.0307***
	(0.0049)	(0.0041)	(0.0062)
Medical beds	−0.0016***	0.0015***	0.0063***
	(0.0005)	(0.0004)	(0.0006)
Observations	45,818	45,932	45,932
Adj. R^2	0.0096	0.0259	0.0201

Notes: Standard errors are in parentheses. The meaning of statistical significance symbol represented by ***, **, * is as follows: *** denotes p<0.01; ** denotes p<0.05; * denotes p<0.1, respectively. We control for individual fixed effects and time of year fixed effects in all estimation results.

Table 5 The Impact of HPAP on Healthcare Costs

Variable	Loutpacare_total	Loutpacare_oop	Loutpacare_total1
	Last Month	Last Month	Last Time
	(1)	(2)	(3)
DHPAPit	−0.4989*	−0.7399***	−0.5893**
	(0.2885)	(0.2838)	(0.2936)
Control	Y	Y	Y
Observations	8712	8609	8625
Adj. R^2	0.0441	0.0345	0.0474
Variable	Loutpacare_oop1	Linpacare_total	Linpacare_oop
	Last time	Last year	Last year
	(4)	(5)	(6)
DHPAPit	−0.7280**	0.0192	−0.6631
	(0.2838)	(0.3039)	(0.4575)
Control	Y	Y	Y
Observations	8513	5555	5447
Adj. R^2	0.0397	0.0563	0.0643
Variable	Linpacare_total1	Linpacare_oop1	Lselfmedicare_total1
	Last time	Last time	Last month
	(7)	(8)	(9)
DHPAPit	−0.0447	−0.8939**	0.1090
	(0.2983)	(0.4348)	(0.0734)
Control	Y	Y	Y
Observations	5486	5345	23,093
Adj. R^2	0.0560	0.0992	0.0477
Variable	Lselfmedicare_oop1	Lmedi_expen	Lfctp_nofood
	Last month	Last year	Last year
	(10)	(11)	(12)
DHPAPit	0.1098	0.2136***	−0.1687***
	(0.0760)	(0.0664)	(0.0487)
Control	Y	Y	Y
Observations	22,645	31,625	37,241
Adj. R^2	0.0451	0.0556	0.1854

Notes: Standard errors are in parentheses. The meaning of statistical significance symbol represented by ***, **, * is as follows: *** denotes p<0.01; ** denotes p<0.05; * denotes p<0.1, respectively. We control for individual fixed effects and time of year fixed effects in all estimation results.

Table 6 The Impact of HPAP on Ex Ante Moral Hazard

Variable	Physical_examination	Smoke	Drink	CHE
	(1)	(2)	(3)	(4)
DHPAPit	0.0645***	0.0097	−0.0276*	0.1259***
	(0.0150)	(0.0149)	(0.0147)	(0.0184)
Control	Y	Y	Y	Y
Observations	45,932	32,828	45,869	35,709
Adj. R^2	0.3198	0.2621	0.0827	0.0658

Notes: Standard errors are in parentheses. The meaning of statistical significance symbol represented by ***, * is as follows: *** denotes p<0.01, * denotes p<0.1, respectively. We control for individual fixed effects and time of year fixed effects in all estimation results. CHE stands for Catastrophic Health Expenditure.

Table 7 The Mediation Effect Test of Physical Examination and Inpatient Care

Variable	Physical_examination	Vulnerability	Inpatient_care	Vulnerability
	(1)	(2)	(3)	(4)
DHPAPit	0.0645***	−0.0308**	0.0934***	−0.0290**
	(0.0150)	(0.0127)	(0.0125)	(0.0127)
Physical examination(Inpatient care)		−0.0204***		−0.0331***
		(0.0059)		(0.0078)
Age	0.0069***	−0.0050***	0.0032***	−0.0050***
	(0.0003)	(0.0003)	(0.0002)	(0.0003)
Gender	0.0230***	−0.0905***	0.0040	−0.0910***
	(0.0055)	(0.0052)	(0.0042)	(0.0052)
Marital status	0.0131	−0.2556***	−0.0106	−0.2562***
	(0.0089)	(0.0087)	(0.0071)	(0.0087)
Education level	0.0048***	−0.0439***	−0.0007	−0.0441***
	(0.0009)	(0.0008)	(0.0007)	(0.0008)
Party membership	0.0643***	−0.0277***	0.0152*	−0.0286***
	(0.0113)	(0.0098)	(0.0089)	(0.0098)
Ethnic minority	−0.0279**	−0.0953***	0.0027	−0.0947***
	(0.0109)	(0.0109)	(0.0084)	(0.0109)
Family size	−0.0067***	0.0095***	−0.0014	0.0096***
	(0.0019)	(0.0019)	(0.0015)	(0.0019)
ln(household income)	0.0093***	−0.0326***	−0.0014*	−0.0330***
	(0.0010)	(0.0018)	(0.0008)	(0.0018)
Asphalt roads	0.0079	−0.0859***	0.0177***	−0.0855***
	(0.0061)	(0.0060)	(0.0047)	(0.0060)
Sewer system	0.0529***	−0.2021***	0.0040	−0.2030***
	(0.0077)	(0.0067)	(0.0058)	(0.0067)
ln(per capita GDP)	0.0485***	−0.1005***	−0.0172***	−0.1019***
	(0.0052)	(0.0050)	(0.0041)	(0.0050)
Medical beds	0.0021***	−0.0050***	0.0015***	−0.0050***
	(0.0005)	(0.0005)	(0.0004)	(0.0005)
Observations	45,932	40,375	45,932	40,375
Adj. R^2	0.3198	0.3991	0.0259	0.3993

Notes: Standard errors are in parentheses. The meaning of statistical significance symbol represented by ***, **, * is as follows: *** denotes p<0.01; ** denotes p<0.05; * denotes p<0.1, respectively. We control for individual fixed effects and time of year fixed effects in all estimation results.

Table 8 The Estimation of Dynamic DID and Robustness Check

Variable	Vulnerability	Health_decline	Vulnerability	Vulnerability
	(1)	(2)	(3)	(4)
DHPAPi,t-5	0.0116
	(0.0202)
DHPAPi,t-3	0.0260	0.0007
	(0.0189)	(0.0067)
DHPAPi,t+1	−0.0594***	−0.0182**
	(0.0145)	(0.0079)
DHPAPit(placebo)			−0.0379***	−0.0019
			(0.0127)	(0.0106)
Control	Y	Y	Y	Y
Observations	40,384	34,704	40,384	40,384
Adj. R^2	0.4032	0.0737	0.4346	0.3985

Notes: Standard errors are in parentheses. The meaning of statistical significance symbol represented by ***, ** is as follows: *** denotes p<0.01; ** denotes p<0.05, respectively. We control for individual fixed effects and time of year fixed effects in all estimation results.