







































 

 

 

 

 

Transportation Infrastructure and Family 

Educational Expectations: An Empirical 

Analysis Based on the China  

Family Panel Survey

Qingfeng Cai,
1
 Guanchen Wu,1 Wuyuan Chen

2
 

 
1. Xiamen University, School of Economics, Xiamen 361005, Fujian, China. 

2. Xiamen University, Institute of Education, Xiamen 361005, Fujian, 

China 

 

Abstract: Based on data from the China Statistical Yearbook and China 
Family Panel Survey (CFPS), this article investigated the influence of 

regional transportation infrastructure development on family education 
expectations. Research results show that there is a significant positive 

correlation between transportation infrastructure and family educational 

expectations and that the improvement of transportation infrastructure 
can enhance the family’s willingness to engage their children in higher 

education. According to the results of influencing mechanism analysis, 

transportation infrastructure development affects family education 
expectations by increasing family income and changing family 

reproduction concepts and educational ideas. In addition,  the impact of 
transportation infrastructure construction on family education 

expectations is heterogeneous: the positive correlation between them is 

particularly significant in rural areas with less developed transportation 
infrastructure, in areas with low average education level of parents and 

slow information communication, and among families of lower social 
classes; however, the relationship between the two variables is not 

prominent in urban areas, in areas where information communication is 

fast and parents have a higher average education level, and among 
families with higher social classes. These research findings offer the 

nation and government new implications for better promoting poverty 

alleviation through education and rural revitalization in underdeveloped 
areas. 

Best Evidence in Chinese Education 2022; 12(2):1643-1651. 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No.2, 2022 1644 

Doi: 10.15354/bece.22.ab017 

How to Cite: Cai, Q., Wu, G., & Chen, W. (2022). Transportation infrastructure and 

family educational expectations: An empirical analysis based on the China family 

panel survey. Best Evidence in Chinese Education, 12(2):1643-1651. 

Keywords: Transportation Infrastructure, Family Educational Expectation, Heterogeneity 

Analysis 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

Correspondence to: Wuyuan Chen, Xiamen University, Institute of Education, Xiamen 361005, Fujian, China. E-

mail: skc@xmu.edu.cn  

Conflict of Interests: None. 
 

© 2022 Insights Publisher. All rights reserved. 

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the 

Creative Commons Attribution-Non Commercial 4.0 License 

(http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of 

the work without further permission provided the original work is attributed by the Insights Publisher. 

mailto:skc@xmu.edu.cn
http://www.creativecommons.org/licenses/by-nc/4.0/


Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1645 

 

Introduction 

As a prerequisite for the social and economic development of a region or nation, the 

development of transportation infrastructure lays a crucial foundation for the regional 

resource exchange, information communication, opportunity sharing and advantages 

complementation, which also profoundly affect regional education resource availability 

and family educational ideas. Improved transportation infrastructure help boost the in-

come of local families, as a result, raising family educational expectations; at the mean-

time, it bolsters the inter-regional flow of people and cultural exchange, popularizes 

urban culture, transforms the backward educational ideas in underdeveloped areas, and 

thus heightens local families’ awareness of the importance of education,  

Family education expectations refer to parents’ expectations of their children’s 

future education level and academic attainment, which are a key representation of the 

family’s educational outlook. Generally speaking, the impact of family educational ex-

pectations on child educational achievement is mainly reflected in two aspects: positive 

family educational expectations increase parental investment in child education, includ-

ing increased money input in as well as time and energy devotion to children’s learning; 

at the same time, there exists Rosenthal effect in family educational expectations 

(Rosenthal effect means that expectations of others will produce an “aspiration effect” 

on individuals, thus promoting the actualization of individual goals). In this context, this 

article utilized data from the China Family Panel Survey (CFPS) and China Statistical 

Yearbook to establish panel data which contain 17073 subjects to examine the influence 

of transportation infrastructure development on family educational expectations as well 

as the influencing mechanisms.  

Research Design 

Sources of Data 

The study sourced data from the China Statistical Yearbook 2014-2018 issued by Chi-

na’s National Bureau of Statistics and the CFPS 2014, 2016 and 2018 undertaken by 

Institute of Social Science Survey of Peking University. The China Statistical Yearbook 

contains basic data of China’s economy, transportation, and population. The CFPS 

samples come from 25 provincial administrative regions, with a target sample size of 

16,000 households. The survey subjects include all family members of the sampled 

households, and data on family educational expectations, family economic circumstanc-

es, parents’ education levels and other related aspects are collected. 

The Determination of Variables 

The Predicted Variable: Family Educational Expecta-

tions 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1646 

 

Family educational expectations are the dependent variable in this study. The corre-

sponding question in the CFPS questionnaire is “what is the minimum level of educa-

tion you expect from your kids?” The value of this variable ranges from 0 to 7. The 

bigger the value, the higher the family educational expectations it represents, with 0 = 

“no education expectation for kids,” 1= “the primary school,” 2 = “the junior secondary 

school,” 3 = “the senior secondary school/ secondary technical school/ secondary voca-

tional school,” 4 = “the junior college,” 5= “the bachelor’s degree,” 6 = “the master’s 

degree,” and 7 = “the doctor’s degree.” 

The Core Explanatory Variable: The Index of Transpor-

tation Infrastructure  

This study chose railway transport as the index to measure the development level of 

transportation infrastructure in an area, because compared with other transport modes, 

railway transport has its unique advantages such as the large delivery capacity, cost-

effectiveness, long travel distance, wide coverage, independence from weather condi-

tions, etc., which can promote the inter-regional exchange of labor and information and 

bolster local social and economic development. 

Control Variables 

Based on the previous research literature and the available data from the CFPS, this 

article used the education levels of the father and mother and family book collection to 

represent the family cultural capital; the family’s monthly post and telecommunication 

expenses as the proxy variable to represent the home social capital; per capita house-

hold income and household assets to represent home economic capital. Child character-

istics include child gender, age, and being the only child of the family or not. Regional 

characteristics include the logarithm of the per capita GDP of each province every year 

along with the core explanatory variable of the regional transportation infrastructure 

development level. 

The Research Model 

The following model was constructed to explore the causal relationship between trans-

portation infrastructure and family educational expectations: 

 

Edui,t = α+βTrai,t+δXi,t+εi,t 

 

In this equation, subscript i denotes the sampled individual, and Edui,t represents the 

parental educational expectation for i in the year of t; Trai,t represents the development 

level of transportation infrastructure in province i in the year of t; Xi denotes the family 

background, child characteristic, regional characteristic, and other control variables; εi,t 

means the residual. The directivity and significance of coefficient β are the main focus 

in the model. As the predicted variable in this model is an ordinal categorical variable 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1647 

 

with a value range of 0-7, this paper chose to use an ordinal logistic model to perform 

regression after controlling for the fixed effects. 

Analyses of Empirical Research Findings 

Findings from Main Empirical Models  

Estimate results of the variable of family educational expectations demonstrate that the 

higher the level of transportation infrastructure and the GDP of a province, the higher 

the family educational expectations there. However, when the variable of household 

economic capital is added into the model, the variable of the GDP of each province los-

es its statistical significance, which indicates that the GDP of each province affects fam-

ily educational expectations mainly through its effect on the economic capital of local 

families.   

The regression results show that: (1) The coefficient of the core explanatory 

variable - the development level of transportation infrastructure - in all models is signif-

icantly positive, indicating that the development of transportation infrastructure plays a 

substantial role in raising the educational expectations of local families, as it has been a 

driving force to the social, economic, and cultural advance in the region. It prompts 

families to change their educational ideas and improves outcomes of family educational 

input, and as a result, parents have higher requirements and expectations for children’s 

education attainment. (2) The richer the family capital, the higher the family education-

al expectations for its kids. (3) The coefficient of the variable of child gender is signifi-

cantly positive. Boys are likely to receive higher family educational expectations, indi-

cating that gender discrimination still exists in children’s education. (4) The coefficient 

of the variable of being the only child of the family or not is significantly positive, sig-

naling that parents have higher educational expectations for the only child than for chil-

dren with siblings.  

Whichever model it is, the advance in transportation infrastructure increases 

the number of parents who expect their children to obtain doctor’s, master's, or bache-

lor's degrees, and reduces the number of parents who limit their educational expecta-

tions for children to the junior college or below. In other words, the development of 

transportation infrastructure can significantly boost parents’ willingness to engage their 

children in higher education. 

The Robustness Check 

Altering the Index of Transportation Infrastructure 

To test the robustness of the above results, the development level of highways was used 

as the index to measure the level of regional transportation infrastructure and performed 

the regression of the main empirical models again; an additional indicator of transporta-

tion infrastructure was produced by combining the standardized railway density and per 

capita railway kilometers and the coefficient of the explanatory variable was estimated 

again. The regression results show that the symbol and significance of the explanatory 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1648 

 

variable remain unchanged, whether the development level of highways or the recon-

structed railway-based index is used to replace the original index of transportation in-

frastructure, which indicates that improved transportation infrastructure can significant-

ly raise educational expectations of local families. 

Using the Ordinal Probit Model  

The ordinal probit model was used to conduct regression in order to further verify the 

stability of the regression results of the ordinal logit model. The regression results show 

that the symbols and significance of the coefficient of the variable of transportation in-

frastructure have not changed 

Propensity Score Matching 

This study used the propensity score matching (PSM) method to address the 

endogeneity issue that may be caused by the sample self-selection bias: (1) The sample 

was divided into areas with developed transportation infrastructure (N=8964) and areas 

with ordinary transportation infrastructure (N=8109). The former was set as the experi-

mental group and the latter as the control group. (2) The control variables in the main 

regression model 5 were used as the predictor variables to predict the probability of the 

family residing in an area with developed transportation infrastructure and the propensi-

ty scores were calculated. (3) One to one matching, one to four matching and one to 

four radius matching was performed on sample respectively. The results of propensity 

score matching show that the standardized deviation of variables is substantially re-

duced, the absolute deviation of all variables is less than 5%, and the t-test results of 

most variables do not contradict the assumption that there are no systematic differences 

between the experimental group and the control group, indicating that the matching re-

sult is good. The sample regression results based on PSM show that the symbols and 

significance of explanatory variables are consistent with those in previous analysis, that 

is, the improvement of transportation infrastructure can significantly heighten the edu-

cational expectations of local families. Therefore, the conclusion of this article is robust 

and reliable. 

Analyses of Influencing Mechanisms 

The Effect of Family Income 

A three-step mediation analysis was undertaken: (i) to regress the effect of transporta-

tion infrastructure on family educational expectations with the results showing that the 

influence of transportation infrastructure on family educational expectations remains 

prominently positive when the control variable of family income is not included in the 

model; (ii) to regress the effect of transportation infrastructure on per capita household 

income with the results displaying that the improvement of transportation infrastructure 

can significantly bolster local family income; (iii) to refer to the regression results of the 

main empirical models which have proved that both household income and transporta-

tion infrastructure are positively correlated with family education expectations. The 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1649 

 

above analysis indicates that household income can mediate the effect of transportation 

infrastructure on family educational expectations. In other words, the improvement of 

transportation infrastructure can heighten educational expectations of local families by 

promoting the increase of family income. 

The Effect of Cultural Concepts 

The advance in transportation infrastructure not only has a positive impact on the re-

gional economy, but also speeds up inter-regional information communication, expand 

people’s travelling spheres, and thus boost inter-regional cultural exchange. As a result, 

some new concepts and thoughts are introduced into the lives of residents, gradually 

transforming the local culture and residents’ ideas. Further, the change in regional cul-

tural background and family cultural capital significantly modifies family educational 

expectations. 

Further Heterogeneity Analyses 

The foregoing analyses focus on the influence of transportation infrastructure on family 

educational expectations and its mechanism --- transportation infrastructure enhances 

the educational expectations of local families by promoting the economic development 

of areas within its coverage and changing the family educational concepts. To go fur-

ther, is there any heterogeneity in the effect of transportation infrastructure on family 

education expectations? To address this question, this study performed a heterogeneity 

analysis in the four dimensions: urban and rural division, internet penetration, parental 

education level, and family social capital, with reference to previous research on this 

topic. 

The Urban vs. Rural Heterogeneity  

Regression analyses were carried out on rural and urban households separately and 

demonstrated that the improvement of transportation infrastructure has a significant 

positive effect on educational expectations of rural families but has no significant im-

pact on educational expectations of their urban counterparts.  

Disparities in Education Level among Parents 

To investigate the differences in the impact of the advance in transportation infrastruc-

ture on parents with varying education levels, this paper divides the whole sample into 

two groups: families with high education levels (N = 8,566) and with low education 

levels (N = 8,507), according to the median of the average education level of parents in 

a family and conducted regression on them separately. The regression results reveal that 

improved transportation infrastructure is positively related to educational expectations 

of parents with lower education levels but has no significant influence on educational 

expectations of those with higher education levels. 

The Divide in the Internet Use 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1650 

 

This article chose the Internet penetration rate of each region in every year as the proxy 

variable of information communication speed and divided the sample into the fast in-

formation communication group (N = 9,006) and the slow information communication 

group (N = 8,067) in accordance with the median of the Internet penetration rate. Ac-

cording to the results of group regression, the coefficients of the variable of transport 

infrastructure in both the two groups are positive and significant at the 1% level. By 

comparing the coefficients of the two groups, it is easy to find that the positive impact 

of transportation infrastructure advancement on family educational expectations is 

stronger in the slow information communication group. 

Disparities in Social Capital 

To investigate the heterogeneous effects of the improvement of transportation infra-

structure on educational expectations of families with differential social capital, this 

study divided the sample into two subsamples: higher social class families and lower 

social class families and conducted group regression. The results show that the devel-

opment level of transportation infrastructure is positively correlated with educational 

expectations of lower social class families but has no significant impact on educational 

expectations of higher social class families. 

Conclusions 

Through the ordinal logit model analysis of data from the CFPS, the article drew the 

following conclusions: (i) The more developed the regional transportation infrastructure, 

the higher the family educational expectations. The improved transportation infrastruc-

ture results in stronger intention of families to engage children in higher education. (ii) 

The advanced transportation infrastructure positively affects family educational expec-

tations by increasing local household income and transforming parents’ family repro-

duction concepts and educational ideas. (iii) The improvement of transportation infra-

structure in underdeveloped rural areas can significantly heighten educational expecta-

tions of local families but that in cities has no prominent influence on educational ex-

pectations of urban families. (iv) The advance in transportation infrastructure imposes a 

significant positive impact on family educational expectations in both fast and slow 

information communication regions, with a stronger impact on the latter. (v) Upgrading 

transportation infrastructure leads to higher educational expectations in families with 

lower parental education levels and less social capital but has no remarkable influence 

on educational expectations of families with higher parental education levels and rich 

social capital. 
 

 

The Chinese version of this article has been published in Journal of East China Normal Universi-

ty, 2022, 40(8):96-114. The English version has been authorized for being publication in BECE by 

the author(s) and the Chinese journal. 

 

蔡庆丰, 吴冠琛 & 陈武元. 交通基础设施改善会提升家庭教育期望吗？基于“中国家庭追踪

调查”的实证分析. 华东师范大学学报, 2022, 40(8):96-114. 



Cai et al. (Fujian). Transportation Infrastructure and Family Educational Expectations. 

BECE, Vol.12, No. 2, 2022 1651 

 

 

 

Received: 30 September 2022 

Revised: 11 October 2022 

Accepted: 05 November 2022 

 

 

 

 

 


	3Absbract-WuyuanChen-BECE_TitleNov_2022
	3Absbract-WuyuanChen-BECE_MaintextNov2022

