The impact of nonboarding on the development of disadvantaged boarding students in western rural China

Abstract

Rural China has seen an increase in its migrant workers returning home. As a result, many of these workers’ children, who had previously boarded at school, needed to return home as well. While the existing research indicates that boarding affects the development of disadvantaged children, the effect of the switch to nonboarding on the growth of vulnerable boarding children remains unknown. Using two-stage data from 20,594 fourth- and fifth-grade students in rural Shaanxi and Gansu provinces as well as the difference-in-differences method, this study estimates the impact of switching to nonboarding on the academic performance and mental health of vulnerable boarding students. The results suggest that the shift toward nonboarding significantly reduces boarding students’ academic performance, and further testing shows that these results are robust. Additionally, the switch to nonboarding insignificantly increased the standardized mental health scores of rural primary school students but significantly increased their standardized impulsive tendency scores. Heterogeneity analysis found that boarding students whose mothers had lower educational achievement or whose families belonged to lower economic levels had poorer academic performance after switching, while boarding students whose parents had higher education achievement or myopia possessed better mental health after switching. This study offers novel, policy-relevant insights into potential strategies that would improve the academic performance and mental health of students who transition to nonboarding, especially those with low-educated parents and those belonging to poor families.

Appendices

Appendix

Appendix 1: The effect of non-boarding status on academic performance of boarding students (N = 3322)

Dependent variable: ${\Delta \mathrm{Score}}_{\mathrm{is}}$=${\mathrm{Score}}_{\mathrm{i}2013}-{\mathrm{Score}}_{\mathrm{i}2012}$		Model Adjusted Only for Baseline Standardized Math Scores	Full model
		(1)	(2)
[1]	Non-boarding (1 = boarding at baseline but non-boarding at end-line; 0 = boarding at baseline and end-line)	− 0.076	− 0.109**
		(0.051)	(0.050)
[2]	Age (years)		− 0.138***
			(0.016)
[3]	Gender (1 = male; 0 = female)		0.018
			(0.029)
[4]	If the student is in grade 4 (1 = yes; 0 = no)		0.037
			(0.050)
[5]	If the student has myopia (1 = yes; 0 = no)		0.067**
			(0.034)
[6]	Father’s education level (1 = complete high school or above; 0 = others)		0.031
			(0.039)
[7]	Mother’s education level (1 = complete high school or above; 0 = others)		0.003
			(0.061)
[8]	Both father and mother migrate to urban areas for work (1 = yes; 0 = no)		− 0.025
			(0.042)
[9]	Family assets		− 0.007
			(0.017)
[10]	Distance from the school to the student’s resident county (km)		− 0.000
			(0.001)
[11]	Baseline Standardized Math Scores	− 0.443***	− 0.467***
		(0.018)	(0.018)
[12]	Observations	3,322	3,322
[13]	R-squared	0.261	0.287

1. Control variables that are missing are imputed for regressions. The strata fixed effect was added in all regressions. Robust standard errors in parentheses, adjusted for clustering at the school level. A negative regression coefficient indicates an association with lower end-line mathematics scores
2. ***p < 0.01, **p < 0.05, *p < 0.1

.

Appendix 2: The effect of non-boarding status on mental health and sub-scales of boarding students (N = 3322)

Dependent variable: ${\Delta \mathrm{Score}}_{\mathrm{is}}$=${\mathrm{Score}}_{\mathrm{i}2013}-{\mathrm{Score}}_{\mathrm{i}2012}$		Mental Health Scores	Mental Health Scores	Study Anxiety Scores	Social Anxiety Scores	Loneliness Scores	Self-punishment Scores	Sensitivity Scores	physical anxiety symptoms Scores	Fear Scores	Impulsiveness Scores
		(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
[1]	Non-boarding (1 = boarding at baseline but non-boarding at end-line; 0 = boarding at baseline and end-line)	0.092	0.096*	0.066	0.075	0.106*	0.049	0.045	0.081*	0.030	0.127**
		(0.057)	(0.056)	(0.048)	(0.054)	(0.056)	(0.052)	(0.051)	(0.046)	(0.052)	(0.054)
[2]	Age (years)		0.061***	0.040**	0.060***	0.108***	0.029*	0.029*	0.055***	0.013	0.042**
			(0.018)	(0.016)	(0.017)	(0.018)	(0.017)	(0.016)	(0.018)	(0.016)	(0.017)
[3]	Gender (1 = male; 0 = female)		− 0.193***	− 0.134***	− 0.129***	− 0.049	− 0.155***	− 0.123***	− 0.111***	− 0.253***	− 0.074**
			(0.031)	(0.027)	(0.031)	(0.036)	(0.033)	(0.031)	(0.030)	(0.026)	(0.033)
[4]	If the student is in grade 4 (1 = yes; 0 = no)		0.001	− 0.042	0.004	0.055	− 0.010	− 0.042	0.048	− 0.018	0.004
			(0.043)	(0.039)	(0.037)	(0.046)	(0.043)	(0.036)	(0.044)	(0.031)	(0.044)
[5]	If the student has myopia (1 = yes; 0 = no)		0.048	0.041	0.071	0.010	0.068	0.054	0.036	− 0.018	0.045
			(0.046)	(0.040)	(0.044)	(0.043)	(0.042)	(0.043)	(0.040)	(0.033)	(0.045)
[6]	Father’s education level (1 = complete high school or above; 0 = others)		0.066	− 0.017	− 0.017	0.015	0.017	0.011	0.099**	0.069	0.175***
			(0.050)	(0.047)	(0.055)	(0.055)	(0.056)	(0.052)	(0.042)	(0.047)	(0.057)
[7]	Mother’s education level (1 = complete high school or above; 0 = others)		0.016	− 0.027	− 0.049	0.000	0.087	− 0.018	0.005	0.021	0.064
			(0.066)	(0.054)	(0.059)	(0.071)	(0.062)	(0.055)	(0.060)	(0.063)	(0.057)
[8]	Both father and mother migrate to urban areas for work (1 = yes; 0 = no)		− 0.014	− 0.027	− 0.102**	0.035	− 0.012	− 0.017	0.067	0.017	0.055
			(0.057)	(0.055)	(0.052)	(0.051)	(0.055)	(0.054)	(0.053)	(0.049)	(0.053)
[9]	Family assets		0.013	0.010	0.034*	− 0.002	0.005	− 0.001	− 0.003	0.020	0.011
			(0.019)	(0.015)	(0.018)	(0.018)	(0.019)	(0.017)	(0.018)	(0.014)	(0.018)
[10]	Distance from the school to the student’s resident county (km)		− 0.002*	− 0.001	− 0.001	− 0.001	− 0.001	− 0.001	− 0.001	0.000	− 0.001*
			(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)	(0.001)
[11]	Baseline Standardized Mental Health or Sub-scales Scores	− 0.383***	− 0.393***	− 0.554***	− 0.612***	− 0.613***	− 0.609***	− 0.613***	− 0.552***	− 0.472***	− 0.527***
		(0.016)	(0.016)	(0.017)	(0.018)	(0.019)	(0.016)	(0.017)	(0.019)	(0.015)	(0.017)
[12]	Observations	3322	3322	3322	3322	3322	3322	3322	3322	3322	3322
[13]	R-squared	0.191	0.204	0.313	0.331	0.302	0.329	0.334	0.303	0.245	0.271

1. Control variables that are missing are imputed for regressions. The strata fixed effect was added in all regressions. Robust standard errors in parentheses, adjusted for clustering at the school level. A negative regression coefficient indicates an association with lower end-line mental health scores
2. ***p < 0.01, **p < 0.05, *p < 0.1