To what extent are Contextualised Admission policies in China just and fair? And how could they be improved?

Abstract

Contextualised admission (CA) to higher education (HE) is a policy attempt to reduce the stratification of student intakes, in widespread use internationally. This study examined whether the use of residence in impoverished provinces, a rural hukou or a minority ethnicity in China could reliably, steadily and accurately identify disadvantaged students for contextualised admission purposes. And it examined where there are better indicators that could be used instead, or as well.
This study involved a structured review of existing evidence on the disparities in HE participation in terms of popularly discussed indicators in China. It then employed secondary data analysis including administrative data, universities’ data, and two nationally representative cohort datasets with more than 10,000 cases (Chinese General Social Survey (CGSS) and China Family Panel Study (CFPS)) to answer research questions. Finally, in order to bring this data up to date and obtain some further information, a cross-sectional survey of 800 middle school students was conducted asking about their plans for higher education.
The findings of the structured review and descriptive analysis of CGSS and CFPS confirmed that there are important disparities in university participation between different provinces in China, especially in terms of hukou (residential registration) status, and family socio-economic background. According to CGSS and CFPS, the currently used indicators also include living in an under-developed province, a rural hukou and having a minority ethnicity. Not all of these indicators are necessarily appropriate to be used for contextualised admissions.
For example, group-level or area-based indicators such as ethnicity, or province or urban/rural residence will exclude some truly disadvantaged students, and falsely include some clearly advantaged students. The analysis of CGSS/CFPS showed that both false identification of disadvantage, and non-identification of disadvantage, were relatively common.
The results of CGSS and CFPS analyses also showed that other potential indicators could be problematic. They provided no evidence that the sex of the student and their month of birth are clearly associated with disadvantage. Other indicators such as social class and parental occupation are hard to define and hard for administrators to verify. Furthermore, some indicators such as having non-party member parents cover a very large sub-set of the population, and school-level indicators have the same deficits as area ones – they mistake the individual for their peers.
The most promising single indicator, according to the findings, might be parental educational credentials. Students with parents who have only completed compulsory education have more difficulty in accessing HE, and they are more likely to be part of a less advantaged family. Parental education information is usually officially verifiable and this information is accessible to higher education institutions (HEIs), making it safe and relatively easy to use.
The survey results illustrate the key point that current contextualised admission policies in China only focus on National College Entrance Examination (NCEE) candidates, and so ignore around half of students in each age cohort. This means that the policies are only concerned with students who are already on track to university. There is little or no widening participation on offer for the large group who do not take NCEE and are largely invisible. This group is stratified by sex, age-in-year, and other contextual variables, as well as by parental education, socio-economic status (SES), and province.
The study has several suggested implications. For non-NCEE students, alternative policies are needed leading to a possible HE future for them, or else they need to have at least the chance to be part of contextualised admissions. For all, removal of artificial barriers such as the restrictions associated with hukou status would be a major step towards a fairer system, less predicated on accidents of birth. For contextualised admissions, authorities need to move away from reliance on group or area-based indicators, and towards use of accessible, verifiable indicators of genuine disadvantage.