Poverty Does Make Us Sick

Nazim Habibov; Alena Auchynnikava; Rong Luo

doi:10.5334/aogh.2357

Figures & Tables

Table 1

Main results (outcome variable from 5 = very bad health to 1 = excellent health).

Variables	Model 1	Model 2	Model 3	Model 4	Model 5
Variables	2SLS	2SLS	2SLS	2SLS	OLS
Household wealth (1 = the wealthiest quintile and 5 = the poorest quintile)	0.218***	0.258***	0.263***	0.259***	0.020**
	(0.025)	(0.035)	(0.036)	(0.036)	(0.007)
Age	0.018***	0.018***	0.018***	0.018***	0.019***
	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
Female	0.113***	0.078***	0.081***	0.081***	0.086***
	(0.015)	(0.021)	(0.022)	(0.022)	(0.020)
University education	–0.043*	–0.008	–0.018	–0.017	–0.113***
	(0.019)	(0.027)	(0.028)	(0.028)	(0.022)
Married	0.004	–0.043	–0.041	–0.042	–0.093***
	(0.017)	(0.023)	(0.025)	(0.025)	(0.022)
Unemployment	0.195***	0.204***	0.207***	0.209***	0.291***
	(0.019)	(0.026)	(0.028)	(0.028)	(0.023)
Urban area	–0.088***	–0.101***	–0.108***	–0.103***	–0.001
	(0.019)	(0.027)	(0.028)	(0.028)	(0.022)
Frequent unjustifiable absence of healthcare personnel		–0.017	–0.007	0.020	–0.007
		(0.034)	(0.036)	(0.042)	(0.039)
Disrespectful treatment by healthcare personnel		0.062*	0.036	0.035	0.052
		(0.030)	(0.032)	(0.036)	(0.033)
No required drug available		0.110***	0.106***	0.110***	0.124***
		(0.026)	(0.027)	(0.032)	(0.029)
Healthcare facilities not clean		–0.007	–0.012	0.030	0.025
		(0.042)	(0.044)	(0.053)	(0.048)
Payments required for services that should be free-of-charge		0.183***	0.181***	0.151***	0.135***
		(0.028)	(0.030)	(0.034)	(0.032)
Long waiting time		0.016	0.008	0.054	0.045
		(0.024)	(0.026)	(0.030)	(0.028)
Generalized trust in other people			–0.049*	–0.053*	–0.062*
		(0.024)	(0.026)	(0.024)
Trust in government			–0.037*	–0.033*	–0.032*
		(0.015)	(0.016)	(0.015)
Trust in parliament			–0.039*	–0.030	–0.027
		(0.017)	(0.017)	(0.016)
Trust in political parties			–0.006	–0.006	–0.002
			(0.013)	(0.014)	(0.013)
Community generalized trust				0.042	–0.038
				(0.060)	(0.054)
Community trust in government				–0.016	–0.032
				(0.052)	(0.048)
Community trust in parliament				–0.064	–0.033
				(0.060)	(0.055)
Community trust in political parties				0.021	0.028
				(0.035)	(0.032)
Community level frequency of unjustified absence of doctors				–0.046	0.021
				(0.086)	(0.079)
Community level disrespectful treatment by healthcare personnel				0.003	–0.033
				(0.085)	(0.078)
				–0.004	0.021
Community level no required drug available				(0.062)	(0.058)
				–0.139	–0.106
Community level healthcare facilities not clean				(0.100)	(0.092)
				0.103	0.051
Community level free services that are charged for				(0.067)	(0.061)
				–0.157**	–0.179***
Community level long waiting time				(0.059)	(0.054)

Country dummies included	YES	YES	YES	YES	YES
Number of cases	13233	6605	5851	5851	5851
Wald test of equality of all regression coefficients in the main stage χ²	4513.26***	2539.60***	2331.99***	2369.15***
F-statistic of equality of all regression coefficients					71.86***
Testing instrument relevancy
First-stage robust F statistic	299.04***	158.40***	146.50***	147.59***
Minimum eigenvalue statistic	299.04	158.40	146.50	147.59
Stock and Yogo’s statistic	19.93	19.93	19.93	19.93
Testing instrument validity
Sargan χ²	0.01	0.02	0.07	0.16
Basmann χ²	0.01	0.02	0.07	0.16
Testing engdogeneity
Durbin χ²	80.59***	61.09***	55.85***	54.34***
Wu-Hausman F statistic	80.96***	61.41***	56.09***	54.47***

[i] Notes: Data rounded up.

Regression coefficients and robust standard errors are reported for 2SLS and OLS.

Significance level: * p < 0.05; ** p < 0.01; *** p < 0.001.

Country-dummies are not shown to conserve space.

Table 2

Robustness analysis of main results.

Variables	Model 6	Model 7	Model 8
Variables	2SLS	2SLS	ivprobit
Subjective assessment of household wealth (1 = the wealthiest 10% of households in the country and 10 = the poorest 10% of households in the country)	0.317***
	(0.047)
Household wealth (1 = wealthiest quintile and 5 = poorest)		0.280***	0.085***
		(0.038)	(0.013)
GDP per capita		0.000***
		(0.000)
GDP growth rate		–0.008
		(0.007)
Gini		0.013***
		(0.003)
Current health expenditure		0.042***
		(0.010)
Out-of-pocket expenditure		0.003*
	(0.001)
Age	0.016***	0.018***	0.004***
	(0.001)	(0.001)	(0.000)
Female	0.092***	0.072**	–0.002
	(0.024)	(0.022)	(0.009)
University education	–0.012	–0.015	–0.003
	(0.030)	(0.028)	(0.012)
Married	–0.005	–0.021	–0.023*
	(0.029)	(0.025)	(0.010)
Unemployed	0.206***	0.203***	0.083***
	(0.030)	(0.029)	(0.012)
Urban area	–0.024	–0.135***	–0.019
	(0.025)	(0.029)	(0.011)
Frequent unjustifiable absence of healthcare personnel	–0.002	0.023	0.004
	(0.045)	(0.043)	(0.018)
Disrespectful treatment by healthcare personnel	0.027	0.033	0.024
	(0.038)	(0.037)	(0.015)
No required drug available	0.077*	0.110***	0.044***
	(0.035)	(0.033)	(0.013)
Healthcare facilities not clean	0.022	0.032	0.019
	(0.056)	(0.054)	(0.022)
Payments required for services which should be free-of-charge	0.107**	0.157***	0.036*
	(0.037)	(0.035)	(0.014)
Long waiting time	0.033	0.056	0.003
	(0.032)	(0.031)	(0.013)
Generalized trust in other people	–0.014	–0.050	0.010
	(0.029)	(0.027)	(0.011)
Trust in government	–0.005	–0.034*	–0.015*
	(0.018)	(0.017)	(0.007)
Trust in parliament	–0.022	–0.028	–0.007
	(0.018)	(0.018)	(0.007)
Trust in political parties	0.013	–0.006	–0.004
	(0.015)	(0.015)	(0.006)
Community generalized trust	–0.012	0.027	0.023
	(0.063)	(0.060)	(0.026)
Community trust in government	–0.077	–0.060	0.019
	(0.056)	(0.050)	(0.022)
Community trust in parliament	0.031	0.036	–0.030
	(0.064)	(0.060)	(0.026)
Community trust in political parties	0.019	–0.007	0.002
	(0.037)	(0.035)	(0.015)
Community-level frequency of unjustified absence of doctors	0.062	–0.082	–0.029
	(0.091)	(0.087)	(0.037)
Community level disrespectful treatment by healthcare personnel	–0.058	–0.011	0.002
	(0.090)	(0.086)	(0.037)
Community level no required drug available	–0.048	–0.048	0.001
	(0.067)	(0.063)	(0.027)
Community level healthcare facilities not clean	0.086	–0.086	–0.065
	(0.111)	(0.101)	(0.043)
Community level free services that are charged for	0.123	0.017	0.049
	(0.071)	(0.065)	(0.029)
Community level long time waiting	–0.170**	–0.165**	–0.060*
	(0.062)	(0.059)	(0.026)
Country dummies included	YES	NO	YES
Number of cases	5791	5851	5862
Wald test of equality of all marginal effects in the main stage χ²	2072.25***	2218.50***	1501.40***
Testing instrument relevancy
First-stage robust F statistic	63.41***	138.32***	146.09***
Minimum eigenvalue statistic	63.41	138.32	146.09
Stock and Yogo’s statistic	19.93	19.93	19.93
Testing of instrument validity
Sargan χ²	0.75	0.01	2.46
Basmann χ²	0.75	0.01	2.45
Testing engdogeneity
Durbin χ²	44.29***	61.17***
Wu-Hausman F statistic	44.32***	61.42***
Wald test of equality of all marginal effects in the main stage χ²			33.82***

[i] Notes: Data rounded up.

Outcome in Model 6 and 7 is ordinal from 5 if the individual reported very bad health to 1 if excellent health.

Outcome variable in Model 8 is binomial and equal to 1 if the individual reported bad or very bad health.

Regression coefficients and robust standard errors are reported for 2SLS, while marginal effects and robust standard errors are reported for ivprobit.