Cambridge Personal Statement Example: Applicant deciding MSc education or MSc education policy (Score 93)

Research Problem and Aims China's higher education system enrolled fewer than four million students annually in 1998; by 2023 that figure exceeded ten million. This expansion was designed, in part, to widen access and reduce socioeconomic stratification in labour market entry. Whether it has achieved that purpose remains empirically contested. When degree attainment becomes widespread, employers may respond by placing greater weight on institutional prestige, field of study, or social networks rather than credentials alone — a process often described as credential inflation or positional competition. If sorting along these secondary axes correlates with parental education or household income, expansion may reproduce inequality through a different mechanism rather than dissolve it. This proposal investigates that possibility. The central research question is: To what extent does the expansion of higher education in China between 2000 and 2020 moderate or amplify the relationship between family socioeconomic background and graduates' early labour market outcomes, and through which observable sorting channels does any residual stratification operate? Two subsidiary questions follow. First, does the stratifying effect of family background operate primarily through institutional tier attended, field of study chosen, or post-graduation job-search resources — or through some combination? Second, are these effects stable across the expansion period, or do they intensify as degree density increases in the labour market? Literature Positioning Scholarship on credential inflation draws on two partially competing frameworks. The positional competition literature, associated with work in the sociology of education on effectively maintained inequality, argues that advantaged families shift investment to qualitative distinctions — elite institutions, high-return fields — once a credential becomes near-universal. A separate strand, rooted in human capital theory, predicts that expansion raises average productivity and compresses wage gaps by raising the floor rather than lowering the ceiling. Empirical work on China has produced findings consistent with both: some studies using household survey data report that returns to a degree have declined at lower-tier institutions while remaining stable at selective ones; others find that field of study mediates much of the family background effect. A gap in this literature is that most existing quantitative studies treat the expansion as a single event rather than a process unfolding across two decades, and few decompose sorting channels simultaneously within a single analytical framework. This proposal addresses that gap by modelling institutional tier, field, and job-search capital as parallel mediating pathways across a longitudinal window. I should be clear that this literature positioning is provisional. A full review would be conducted in the first months of the MPhil, and the framing may shift depending on what that review reveals. Methodology The proposed design is quantitative and uses secondary data. The primary data source would be repeated cross-sectional graduate employment surveys conducted by Chinese universities and aggregated at the national level, supplemented where possible by publicly released waves of a relevant household panel. The unit of analysis is the individual graduate; the outcome variables are employment status, occupational prestige score, and wage at six and twelve months after graduation. The key independent variable is a composite socioeconomic background index constructed from parental education and reported household income quintile. The analytical strategy has three stages. The first uses descriptive decomposition to document how the distribution of graduates across institutional tiers and fields of study has changed across the expansion period, stratified by family background. The second estimates a series of regression models — ordinary least squares for continuous outcomes, logistic for binary employment — with interaction terms between cohort year and family background to test whether the association between background and outcomes strengthens over time. The third applies a causal mediation framework to decompose the total background effect into pathways operating through institutional tier, field of study, and a proxy for job-search capital such as internship participation or urban hukou status. I have prior training in regression analysis and mediation methods from undergraduate coursework and from a quantitative project completed in early 2025, in which I synthesised evidence on graduate employment patterns and produced a structured analytical memo. That project gave me working familiarity with the logic of mediation decomposition and with the limitations of cross-sectional inference. I am aware that the mediation estimates will be sensitive to unmeasured confounding, and I plan to address this through sensitivity analyses and by reporting bounds rather than point estimates where identification is uncertain. Feasibility, Ethics, and Timeline The proposed data sources are administrative or survey records that are either publicly available or accessible through institutional data-sharing agreements that Chinese universities routinely extend to researchers. No primary data collection involving human participants is planned, which reduces ethical complexity, though secondary data use will still require confirmation that the original collection met ethical standards and that any re-use is consistent with the terms under which the data were released. I will seek guidance from the Faculty Research Ethics Committee at the outset. The main feasibility risk is data access: some survey waves may be incomplete or inconsistently coded across years. A contingency plan is to restrict the analysis to the subset of years and variables that are consistently available, accepting a narrower longitudinal window if necessary. A secondary risk is that institutional tier classifications change across the period; I would address this by constructing a time-consistent tier variable using published admission score thresholds. A provisional timeline for a one-year MPhil would allocate the first term to literature review and data acquisition, the second term to descriptive analysis and model specification, and the third term to mediation analysis, write-up, and revision. This is tight but achievable for a proposal of this scope, which is deliberately bounded to a single country context and a defined outcome window. Cambridge Fit and Expected Contribution The Faculty of Education at Cambridge has active research on education systems, inequality, and the relationship between schooling and labour market outcomes. The quantitative and policy-analytic orientation of the MPhil in Education Policy aligns with the analytical approach proposed here. I am particularly interested in working within research groups that examine comparative education systems and the political economy of education reform, where the China expansion case can be read alongside parallel expansions in other middle-income contexts. The expected contribution of this project is modest and appropriate for an MPhil. It would provide a methodologically explicit decomposition of sorting channels across a two-decade expansion window — something the existing literature has not done in a single unified framework for the Chinese case. The findings would be directly relevant to ongoing policy debates about whether selective university admissions reform or field-of-study guidance can reduce socioeconomic stratification in graduate employment, and they would generate a clearly specified set of questions for subsequent doctoral or comparative work.