Introduction

The maximally maintained inequality (MMI) posits that ensuring saturation of a specific education level could reduce educational disparities at the given level of education (Raftery & Hout, 1993). However, numerous studies have consistently found that persistent inequality remains, manifested as qualitative distinctions among educational institutions. In response, the effectively maintained inequality (EMI) theory contends that, even when access to a specific level of education is universal, educational disparities continue in the form of qualitative differences among educational institutions (Lucas, 2001, 2009). Under this notion, students from socioeconomically advantaged backgrounds are more likely to improve their standing within a qualitatively varied education system (Andrew, 2017; Entrich, 2019; Luo et al., 2021; Sweet et al., 2017; Valdés, 2022; Wang & Guo, 2019).

South Korea (henceforth Korea) is one of the countries that show a nearly universal transition to high schools and have diversified the education system in both upper secondary and postsecondary levels. In recent years, the qualitative stratification at the upper secondary level has further intensified in Korea. In 2010, Korea’s Ministry of Education (MoE) introduced the high school diversification reform, which introduced autonomous high schools and expanded special-purpose schools, stipulating in the categorization of high schools into four distinct types: vocational, general, autonomous, and special-purpose high schools. Such change in the high school structure could exacerbate educational inequalities within the education system, as it affects students’ academic performance and postsecondary careers (Jang, 2016).

In 2019, the Korean MoE aimed to eliminate the high school stratification by planning to convert special-purpose high schools into general high schools by 2025, but this decision was reversed in 2023 (MoE, 2019, 2023). The shift in policy was based on the belief that promoting diversity can lead to better quality in the overall public education, thereby reducing educational disparities in high school transitions (MoE, 2023). Such hypothesis, therefore, makes Korea a meaningful case as one can test the EMI theory taking into account Korea’s high enrollment rates at the secondary (99%) and postsecondary (80%) levels, as well as diverse qualitative distinctions at both education levels. In addition, there is a lack of research in assessing trends in educational inequality in alignment with the evolving education system and an exploration of how socioeconomic disparities persist in this context. For instance, previous studies have focused solely on the high school tracks, academic versus vocational, and the types of higher education, 2-year colleges versus 4-year universities, which do not adequately reflect the evolving nature of Korea’s education system (Byun & Park, 2017).

In this regard, the study investigates how educational inequalities persist through education transitions at both secondary and postsecondary levels in Korea. First, the study employs the EMI framework to examine qualitative disparities among diverse types of high schools. It seeks to determine whether students from higher SES maintain their advantages as they ascend the stratified hierarchy of high schools. In addition, this research evaluates how attending different high school types affects students’ academic performance in their final high school year Essentially, it investigates whether the high school type can lead to varied learning outcomes for students of comparable academic abilities. Finally, the study contextualizes the qualitative differences among higher education institutions and investigates the potential institutional advantages of attending high-quality high schools, such as autonomous and special-purpose schools, during the transition to higher education.

Background

Theoretical background

Criticizing the lack of appreciation of MMI for qualitative differentials within a given level of education, EMI argues that educational inequality owing to socioeconomic disparities persists in the form of qualitative distinctions within education, such as the type of education and academic curricula (Lucas, 2001; Lucas & Byrne, 2017). These disparities are predominantly upheld through gatekeeping behaviors, such as imposing standardized tests or advanced coursework heavily influenced by parental SES, resulting in socioeconomically advantaged students being more likely to transition to higher-quality educational institutions (Alon, 2009; Andrew, 2017). Consequently, EMI theory contends that qualitative distinctions within an educational system would maintain socioeconomic disparities through students’ likelihood of pursuing different educational trajectories. The scholars testing the EMI hypothesis have found that tracking decisions from lower to upper secondary education often place low-SES students in vocational and technical tracks, which are viewed as less prestigious and offer fewer opportunities for higher education and lucrative careers, a trend observed in countries such as the USA, Germany, and Italy (Andrew, 2017; Guetto & Vergolini, 2017; Protsch & Solga, 2016).

In addition, recent research has highlighted the importance of a context-specific application of EMI, which identifies qualitative differences in the types of education within a country-specific education system (Andrew, 2017; Fernández, 2022). For instance, research in Spain categorized high school education into natural sciences and technology, social sciences, and vocational high schools. Parental SES significantly influenced upper secondary tracks, with socioeconomically advantaged students preferring natural sciences and technology, and disadvantaged students attending vocational or social sciences, humanities, and arts (Fernández, 2022). Similarly, research in Japan, which classified high schools into four tiers based on school rankings, found that students from higher family socioeconomic backgrounds were more likely to progress to higher-tier academic high schools (Entrich, 2019).

Previous studies validating the EMI theory in Korea have focused on the high school track distinction between vocational and academic high schools, and on the type of higher education, comparing 2-year colleges to 4-year universities, finding that students from higher socioeconomic backgrounds are more likely to attend academic high schools and 4-year universities (Byun & Park, 2017; Choi, 2020). However, these studies do not adequately reflect Korea’s evolving education system, which has diversified and stratified over the years. This study fills the gaps by conceptualizing the institutional hierarchy in the contemporary Korean school system and exploring how socioeconomic status affects students’ advancement within this hierarchy.

Contextual background

Transition to high school in Korea

The education system in Korea is based on a K-12 structure divided into primary schools (grades 1 to 6), middle schools (grades 7 to 9), and high schools (grades 10 to 12) (Byun & Park, 2017). The year 2010 saw significant changes that diversified the types of high schools—which used to be mainly general and vocational high schools—into vocational, general, autonomous, and special-purpose high schools to enable tailored educational service based on student demands. General high schools, accounting for 72% of all high school students in 2020, offer advanced education with elective courses and prepare students for college education. Autonomous high schools (8%) consist of autonomous public and autonomous private high schools that have autonomy in school management, academic curriculum, and teacher recruitment. Special-purpose high schools (5%) include foreign language, science, and art high schools that are designed for gifted students in respective areas.

In the last year of middle school, students can apply to a high school of their choice. If not selected by their chosen school, they are generally assigned to a general high school based on their registered address. Admission criteria for autonomous and special-purpose high schools are based on students’ academic performance in middle school, their academic records, and interviews to assess their suitability for specialized curricula (MoE, 2021). These criteria can influence student selection by acting as gatekeeping mechanisms that potentially favor socioeconomically advantaged students. That is, socioeconomic advantages can enhance academic performance through access to private education support and additional resources, providing further advantages in preparing academic records and admission interviews.

It is noteworthy that the premium associated with attending an academic high school track—general, autonomous, and special-purpose—is evident, as 77% of its students progress to college, compared to 58% from vocational schools. Previous studies have also found qualitative distinctions among academic high schools, with students who move on to special-purpose and autonomous high schools generally demonstrating superior academic performance compared to their counterparts in general high schools (Kim & Namkung, 2014; Kim et al., 2016). Additionally, about 39% of new students at Korea’s three prestigious universities came from special-purpose high schools (20%) and autonomous high schools (19%), despite these schools accounting for only 11% of the total high school graduates in 2020 (MoE, 2021). Such qualitative differences among high school types have stratified the Korean education system, resulting in socioeconomic inequalities during students’ transition to high school. This hierarchical arrangement is depicted in Fig. 1.

Fig. 1
figure 1

Source: created by authors based on Andrew (2017) and Byun and Park (2017)

Stratification of upper secondary and higher education.

Full size image

Transition to college in Korea

Korea’s higher education system largely consists of 2-year colleges and 4-year universities. Similar to the high school system, stratification among different types of schools also exists in higher education. First, 2-year colleges are perceived to be of lower prestige than 4-year universities, one reason being the income disparities after graduation. The OECD (2023) reported that in Korea, workers with a bachelor’s degree from 4-year universities earn 51% more than their counterparts with upper secondary degrees, while those with 2-year college degrees earn 11% more.

In addition, the geographic location of universities is significant as higher-ranking institutions are located in the capital area of Korea. About 95% of the top 20 universities are located in the capital area (Lee et al., 2023, see also Appendix G). Previous studies have demonstrated that 4-year universities located in the capital area have a wage premium over those located in non-capital regions. Cho and Kim (2014) found that graduates from Seoul’s 4-year universities had initial incomes 5 to 12% higher than their counterparts from other areas, and overall, 4-year university graduates earned about 13% more than 2-year college graduates. However, despite these disparities, prior research predominantly concentrates on the distinction between 4-year universities and 2-year colleges, often overlooking the regional disparities among higher education institutions (e.g., Byun & Park, 2017; Choi, 2020).

In this regard, this study contextualizes the hierarchical nature of the Korean education system, as illustrated in Fig. 1, by highlighting the perceived distinctions in quality among different high school types and higher education institutions. This study postulates that high-SES students are able to maintain their advantage in progressing to higher-quality institutions, such as autonomous and special-purpose high schools, within this stratified school system. Additionally, the study examines the benefits of attending these higher-quality high schools on students’ academic performance and their transition to different higher education institutions.

Research questions

Following the arguments outlined above, this study poses three research questions (RQs):

  • RQ 1: Do high-SES students have greater mobility in the high school hierarchy? Specifically, are high-SES students more likely to progress to high-quality schools, such as autonomous and special-purpose high schools?

  • RQ 2: Are having high-SES and attending high-quality high schools associated with achieving higher academic performance during high school?

  • RQ 3: Do high-SES students and those who attend high-quality high schools have a greater likelihood of advancing in a qualitatively stratified postsecondary system? Specifically, are they more likely to attend 4-year universities in capital areas?

Methods

Data

This study uses the data from the Korean Educational Longitudinal Study (KELS) (Korea Educational Development Institute, 2023). The Korean Educational Development Institute (KEDI), a government-funded education research institute, administers the KELS, a nationally representative dataset. The KELS randomly selected a total of 7324 fifth-grade students in 2013, with a plan to track the same group of students while collecting annual data until 2030. Along with the students’ information, data on household members and school administrators are also subjects of the survey.

In particular, this study utilized the surveys conducted from 2017 to 2023. The fifth survey gathered information about the academic performance and family background of students during their final year of middle school. Subsequent surveys monitored data pertaining to students’ transitions into upper secondary education and their decisions regarding higher education. Our analysis focused on respondents with available educational trajectory data (see Appendix A), selecting a sample of 4654 individuals from a nationally representative pool (see Appendix B & C).

Measures

Transition to high school

Given that the incidence of students dropping out during this transition is exceedingly low both in our sample and in Korea as a whole (less than 1%), this study narrows its scope to examine respondents who make the transition to high school. This transition is categorized into four distinct high school types, namely vocational (vocational track),Footnote 1 general (academic track), autonomous (academic track), and special-purpose high schools (academic track). In our sample, the distribution of students transitioning to each type of high school closely mirrors that of the national population (please refer to Appendix D).

Transition to college

To contextualize the differences in the quality of college pathways in Korea, this study categorized the sample into those who pursued 4-year universities in the capital area, non-capital 4-year universities, 2-year colleges or employment, and others. Others, here, refer to individuals who are actively seeking employment as well as those who may be engaged in re-studying for the college entrance exam.

Family socioeconomic status

This study measures SES with three variables: monthly household income, father’s education, and mother’s education (Garrison et al., 2017). Parental education is categorized into four levels: secondary education or less, 2-year college, 4-year university, and master’s degree or higher. This study defines low-SES as parental education and family income at or below the 20th percentile of the sample distribution, and high-SES as parental education and family income at or above the 80th percentile. The aim is to analyze the differential transitions of each socioeconomic group into various types of high schools and higher education.

Academic achievement

First, this study accounts for students’ academic performance in their final year of middle school, as assessed by the KELS’ subject-level achievement test. This study combines the scores from three core subjects—mathematics, English, and Korean—on a percentile scale, where higher values indicate superior academic performance. Students’ academic performance is categorized at the 20th and 80th percentiles for low and high achievement to assess how academically similar students from different SES groups transition to different high school types. In addition, this study measures high school academic performance using the Korean National Academic Ability Test scores, which assess students’ performance in the final year of high school across the three mentioned subjects, graded on a scale from 1 (lowest) to 9 (highest). Multiple imputation was used to handle missing data in these three academic subjects’ performance, generating ten imputed datasets (Kenward & Carpenter, 2007). This study also controls for students’ demographic attributes, encompassing students’ gender and residential location (urban or rural).

Analytical strategy

This study utilizes descriptive statistics and incorporates two regression models—multinomial logistic regression and linear regression analysis. Initially, this study explores descriptive statistics to examine variations in individual and household background characteristics during the transition to upper secondary and higher education. Subsequently, the study uses multinomial logistic regression to assess the probability of students attending one of four high school types based on their SES (Byun & Park, 2017; Guetto & Vergolini, 2017; Lucas & Byrne, 2017). The investigation further estimates the predicted probabilities of enrolling in different types of high schools for students from low, middle, and high-SES groups, as well as from high and low academic achievement groups (Andrew, 2017). This approach enables a detailed examination of how high-SES students benefit in transitioning to higher-quality high school types compared to their peers of similar academic achievements.

In continuation, the study employs linear regression to investigate the association between high school type and academic performance during the final year of high school. Additionally, this analysis targets students who ranked in the top 20th percentile of academic performers during middle school, aiming to investigate how their high school academic performance has changed depending on high school type. The goal is to ascertain whether certain types of high schools provide institutional advantages over others, considering comparable previous academic achievements. Finally, this study employs multinomial logistic regression to examine the relationship between high school type and the likelihood of progressing to various higher education categories, evaluating the benefits of attending high-quality high schools.

Findings

Descriptive findings

The descriptive findings in Table 1. revealed a clear trend, showing that the level of students’ family SES increases in the order of vocational, general, autonomous, and special-purpose high schools. In other words, high-SES students are more likely to attend autonomous and special-purpose high schools. For instance, students from families in the highest income quintile demonstrate a preference for enrolling in autonomous high schools and special-purpose high schools, constituting 29% and 33% of this group. Similarly, students with fathers who have completed 4-year university degrees or higher constitute a significant portion of attendees in autonomous (42%) and special-purpose high schools (66%).

Table 1. Descriptive statistics of upper secondary transition across high school types
Full size table

Conversely, students from lower SES, characterized by lower parental education levels and residing in less affluent income quintiles, show a preference for vocational high schools. For example, those with parents with secondary education or less make up 61% of fathers and 64% of mothers in this category. Similarly, students from the lowest income quintile exhibit a stronger inclination toward vocational high schools, comprising 44% of all attendees in this school type.

Regarding gender differences, the analysis reveals variations in high school transitions. Gender differences in high school transitions are evident. Male students are more likely to attend vocational (55%) and autonomous (60%) high schools. In contrast, female students demonstrate a higher inclination toward general (53%) and special-purpose (69%) high schools. Regional disparities are also observed in high school transitions. Concerning their residential background, students from rural areas have lower representation in autonomous (8%) and special-purpose (12%) high schools, relative to the national average of 17%.

Table 2 provides a breakdown of the transition to higher education based on SES and various demographic factors. Over half of the students pursue enrollment in 4-year universities, with 20% opting for universities in capital areas and 37% entering universities in other regions.

Table 2 Descriptive statistics on post-high school destinations across higher education categories
Full size table

The descriptive statistics show that students from higher-income families are more likely to attend 4-year universities in capital areas. For example, the highest income quintile has a substantial presence in these institutions (29%). Similarly, students with fathers who hold 4-year university degrees or higher are more likely to attend 4-year universities in capital areas (56%). In contrast, students from lower SES are more inclined to attend 2-year colleges and seek employment. For example, in the lowest income quintile, 31% of students choose 2-year colleges and 38% of them secured employment, which is notably higher than the percentages in higher income quintiles. This trend suggests that economic constraints or a desire for more affordable educational options may influence this choice.

Regarding high school type, students from autonomous and special-purpose high schools display a distinct pattern with a focus on 4-year universities. Proportionally, nearly half of the special-purpose high school graduates, 74 out of 159 students or 47%, proceed to 4-year universities in the capital area. Similarly, approximately 26% of autonomous graduates attend these universities (110 out of 422). This suggests that these high schools systemically prepare students academically for prestigious university programs. In contrast, students from vocational high schools primarily favor 2-year colleges or employed, constituting 68% of their transitions in this category.

In summary, these statistics emphasize the significant impact of parental SES and the type of high school attended on students’ higher education transitions. Autonomous and special-purpose high schools appear to concentrate on attending students for 4-year university in capital areas. General high school students exhibit a diverse range of pathways, while students from vocational high schools enroll in a 2-year college education or immediate employment.

Regression results

Socioeconomic differences in the transition to high schools

Table 3 displays the findings of a multinomial regression analysis on the transition to different high school types, detailing marginal effect sizes which indicate the change in percentage points regarding the likelihood of attending each school type.

Table 3 Multinomial logistic models predicting transition to upper secondary education
Full size table

The analysis distinguishes between middle- and high-SES groups in comparison to their low-SES counterparts, revealing notable disparities in the likelihood of attending each high school type after controlling for other covariates, including students’ academic performance, their gender, and school area. High-SES students exhibit a 20-percentage point lower likelihood of attending vocational high schools and a 3 and 5 percentage point higher likelihood of attending special-purpose and autonomous high schools, respectively, after controlling other covariates. When examining SES factors separately, students from the highest income quintile show 4 and 2 percentage points increased likelihood of attending autonomous and special-purpose high schools, respectively, and those with parents holding 4-year university degrees have a 2–6 higher chance of attending special-purpose high schools (see Appendix E). These findings indicate a clear SES gradient in the transition to different high school types, with higher SES correlating with a reduced preference for vocational tracks and a higher preference for attending high-quality high schools, such as autonomous and special-purpose high schools. These findings align with prior research (e.g., Byun & Park, 2017; Kim & Ogawa, 2023) indicating that students from wealthier families are more likely to pursue academic tracks.

Significant differences emerge when considering gender and rural location in high school attendance. Female students exhibit notably stronger preferences, being less likely to attend vocational high schools but more inclined to opt for general- or special-purpose high schools, compared to their male peers. This disparity underscores the influence of gender in shaping high school choices. Regional differences were significant but less pronounced in students’ high school choices. Students residing in rural areas were more inclined to enroll in general high schools rather than autonomous high schools. These results are in line with a new theoretical perspective referred to as segmentally-maintained inequality proposed by Luo et al. (2021) who argued that established inequality can be more pronounced for selected social segments.

As anticipated, a positive correlation was found between higher academic performance and a greater likelihood of attending academic tracks, including general, autonomous, and special-purpose high schools. For instance, high-achieving students were 24 percentage points less inclined to choose vocational high schools. However, the effect of high academic performance somewhat diminished within the selected academic tracks. Notably, the impact of academic performance was most moderate when it comes to attendance at special-purpose high schools although still statistically significant (2 percentage points, p < 0.05).

Previous studies have shown that academic performance is crucial in determining high school type, as autonomous and special-purpose high schools generally select students based on their middle school academic performance (Hur, 2018). This study categorized students by similar academic abilities to isolate the impact of socioeconomic background on educational transitions. It aims to investigate how students from wealthier backgrounds are advantaged in transitioning to high schools, especially autonomous or special-purpose high schools, compared to peers with similar or lower academic achievements. This analysis, encapsulated in Table 4, leverages multinomial logistic regression to predict probabilities of attending different school types across different SES levels for students with similar academic performance (full results with marginal effects are provided in Appendix F). In multinomial regression analysis, predicted probability signifies the likelihood of attending a particular high school type based on diverse socioeconomic groups within each academic performance group. For instance, a 44% predicted probability of attending a vocational high school among individuals from low SES indicates a 44% likelihood of a student from the low SES group attending a vocational high school among the low-performing group (result A in model (1) of Table 4).

Table 4 High school transition probabilities by SES and academic performance
Full size table

For low academic performers (result A of Table 4), low-SES students have a high probability of attending vocational (44%) and general high schools (49%), with a very low probability of attending autonomous (4%) and special-purpose schools (3%). In contrast, among low-performing students from high-SES, there is a marked propensity for autonomous and special-purpose schools, standing at 11% and 8%, respectively. Additionally, low-performing but high-SES students were 34 percentage points less likely to attend vocational high schools than their low-SES counterparts (see also Appendix F).

An interesting observation is the comparison between low-achieving but high-SES students (result A) and high-achieving but low-SES students (result C). Despite their lower academic achievements, high-SES students have a higher probability of attending special-purpose high schools (8%) compared to high-achieving but low-SES students (3%). This suggests that for high-SES students, academic performance is less of a barrier to accessing more prestigious educational institutions, potentially indicative of elevated social capital, enhanced access to information, or heightened aspirations (Kim & Ogawa, 2023).

Among high achievers (result C), who have higher chances of being selected for autonomous or special-purpose high schools, the advantage of high-SES is even more pronounced. High-SES students in this group are more likely to attend autonomous or special-purpose high schools, rather than settling for general high schools. This is evidenced by the 16% probability for autonomous schools and 10% for special-purpose schools. The comparatively higher probabilities of entering these high-quality institutions among high-SES students may reflect not only their financial capacity but also parental and social expectations. On the contrary, there is a clear and overwhelming preference for vocational or general schools among the low-SES but high-achieving group, with a predicted probability of 10% and 79%. Notably, low-SES students, with a 10% probability of entering vocational high schools, are more likely to attend vocational high schools compared to both their middle or high-SES counterparts.

In conclusion, the study demonstrates that parental SES significantly influences high school transitions, with wealthier and lower-performing students progressing to specialized prestigious schools by offsetting their lower academic performance. Conversely, evidence shows that low-SES students tend to end up in vocational or general schools, even with high academic achievements. These results indicate that socioeconomic advantages can outweigh academic constraints, thereby profoundly affecting educational trajectories.

Impact of high school type on academic performance

This section evaluates the influence of parental SES and high school type on academic performance in Korean, Math, and English, utilizing scores from the Korean National Academic Ability Test in the senior year, which are scaled from 1 (lowest) to 9 (highest). The study hypothesizes that the high school type and parental SES play a significant role in shaping academic performance during the high school years.

Table 5 displays the regression results for students’ high school performance, taking into account the type of high school they attended, SES, and other variables. The results show that special-purpose high school students excel in all subjects: Korean (0.63, p < 0.01), Math (0.63, p < 0.05), and English (0.75, p < 0.001), indicating a substantial advantage. Autonomous high school students also show moderate improvements in Korean and English (0.33, p < 0.05 for both) but no significant difference in Math compared to their peers. Considering that the test scores for three subjects are graded on a scale of 1 to 9, the reported effect sizes indicate a significant association that could play a crucial role in shaping their college paths.

Table 5 Impact of high school type on high school academic achievement
Full size table

This analysis also reveals the consistent and significant impact of parental SES on students’ learning during high school. That is, students from middle- and high-SES backgrounds perform significantly better than those from low-SES in all subjects. The effect is more pronounced in high-SES students, with substantial increases in grades: Korean (0.90, p < 0.001), Math (0.93, p < 0.001), and English (1.82, p < 0.001).

Female students show no significant difference in academic performance compared to male students. Rural students, however, are at a distinct disadvantage in all subjects (Korean: − 0.34, Math: − 0.31, English: − 0.45, all p < 0.001), highlighting the challenges faced by students in these areas as explained by segmentally-maintained inequality (SMI) proposed by Luo et al. (2021).

The results show a strong positive correlation between students’ academic performance in middle schools and the equivalent in high schools. The analysis shows that high performers from middle school exhibit the most substantial gains across all subjects, suggesting that previous academic success is a strong predictor of continued achievement in high school.

The analysis delves deeper into students who were high performers in the top 20th percentile during middle school, as they are more likely to be selected for autonomous or special-purpose high schools. The purpose of this analysis is to examine whether autonomous and special-purpose high schools provide advantages over other high schools even when previous academic performance was similar (Table 6).

Table 6 High school type impact on high performers’ academic changes (N = 920)
Full size table

Table 6 shows that attending autonomous high schools is significantly correlated with improved grades in all three core subjects—Korean (0.51, p < 0.05), English (0.53, p < 0.001), and Math (0.51, p > 0.001)—compared to their peers in general or vocational high schools. Furthermore, attending special-purpose high schools offers greater academic benefits. Students at these schools show significant enhancements in Korean (0.56, p < 0.001), English (0.60, p < 0.05), and Math (0.88, p < 0.001) scores. This indicates that a student with similar grades in middle school could expect a 0.5 to 0.9-point increase in their academic grade by attending these high schools.

This evidence implies that transitioning to different school types could be associated with students’ academic performance. Autonomous- and special-purpose high schools emerged as an influential factor, implying substantial advantages in academic performance across all subjects. Additionally, SES revealed itself as a consistent and significant determinant, with students from middle- and high-SES consistently outperforming their peers from low-SES backgrounds. This analysis underscores the institutional advantages of selected high school types and socioeconomic advantages in shaping academic achievement.

Impact of SES and high school type on the transition to higher education

This section examines the likelihood of students making different postsecondary choices, focusing on how their SES and high school types influence their transition to various higher education institutions.

The evidence consistently demonstrates that parental SES confers significant advantages to students in their educational choices as outlined in Table 7. Middle- and high-SES students exhibit a notable advantage, with a 5 and 8 percentage point increase, respectively, in their likelihood to enroll in 4-year universities located in capital areas. These findings support the EMI hypothesis, which posits that educational disparities rooted in socioeconomic distinctions persist in the form of educational institution selection. That is, students from more affluent backgrounds consistently ascend the institutional hierarchy during school transitions.

Table 7 Multinomial logistic models predicting transition to higher education
Full size table

In addition, students from special-purpose high schools were 15 percentage points more likely to enroll in 4-year universities in the capital areas compared to students from vocational or general high schools, even after accounting for family SES and academic performance. These findings underscore the institutional advantage offered by special-purpose high schools toward prestigious urban universities, which also offer the highest income premium at later stages in life (Cho & Kim, 2014). The findings suggest that prestigious universities’ institutional selectivity might lean toward favoring graduates from special-purpose high schools, a trend potentially driven by these schools’ distinguished reputation, leading to preferential admissions for these students.

Overall, these findings demonstrate that the type of high school attended can significantly impact students’ postsecondary educational trajectories. This result is in line with previous studies that concluded that students from special-purpose high schools are more likely to progress to 4-year universities in the capital area compared to graduates from general high schools (Shin & Min, 2023).

Discussion and conclusion

Expanding on the extant empirical studies of the EMI theory, this paper contributes to elucidating the dynamics of socioeconomic inequalities during the transitions from upper secondary to higher educational transition within Korea’s evolving educational system, characterized by institutionally hierarchical structure.

In alignment with prior findings, this research underscores the enduring influence of socioeconomic backgrounds in shaping students’ choice of school types (Entrich, 2019; Wang & Guo, 2019). The study reveals that high-SES students leverage benefits, compensating for their varied academic performance levels, which propel them up the institutional ladder—favoring choices like autonomous and special-purpose high schools—when compared to their low-SES peers. The results are consistent with other findings that high-SES students compensate for academic performance by attending prestigious schools (Andrew, 2017; Kim & Ogawa, 2023).

In addition, the study confirmed that there are institutional advantages of attending autonomous and special-purpose high schools (Jeong et al., 2018; Kim et al., 2016). First, among high-achieving middle school students, those attending autonomous or special-purpose high schools demonstrated significant improvements in major academic subject scores compared to their peers in general or vocational schools (Kim et al., 2016). This may be linked to several institutional advantages in autonomous and special-purpose high schools, including higher participation in after-school academic programs and providing selective courses, such as calculus and chemistry, lower student–teacher ratios, and access to selective English language education programs (Jeong et al., 2018; Kim et al., 2016).

More importantly, students from special-purpose high schools have an advantageous position in securing admissions to 4-year universities in capital areas. These results demonstrate that institutional selectivity favors graduates of special-purpose high schools, allowing high-SES students to ascend the postsecondary institutional hierarchy by attending more selective high schools. The evidence shows that high-SES students select academic paths with long-term benefits, whereas low-SES students often choose vocational programs with immediate but lower returns, despite similar academic abilities (Hu et al., 2020).

This result enriches the discourse on educational disparity, positing that educational inequalities occur by the transmission of socioeconomic disparities from families to their children in the form of school type (Lucas, 2017). This study contributes to previous findings by conceptualizing Korea’s stratified school system, revealing that educational disparities are exacerbated by institutional quality differences at secondary and postsecondary transition points. During this process, cumulative institutional advantages of high-quality secondary and postsecondary education are conferred on high-SES students, leading to persistent educational stratification. Given previous literature revealing the wage premium of graduating from 4-year universities in the capital area, these cumulative institutional advantages enjoyed by high-SES students may lead to more pronounced social stratification. Thus, the findings of this study suggest that the MoE’s decision to preserve the existing school structure poses significant challenges in addressing entrenched educational inequalities and may further exacerbate educational stratification.

Overall, this study reveals persistent socioeconomic disparities across critical educational transitions in Korea. However, the mechanisms driving these differential transitions remain unexplored. Notably, the research does not clarify if high-achieving students from low-SES backgrounds, who arguably stand a good chance of gaining admission into elite upper secondary or higher education institutions, are intentionally opting for less prestigious establishments or if they are being overlooked by these renowned institutions. Additionally, the study’s approach to categorize higher education appears somewhat narrow, failing to capture the full spectrum and nuances of the Korean higher education landscape. To understand these patterns, future research should consider whether these decisions are a result of institutional selectivity, or whether they emerge from a self-limiting goal revision process on the part of students and their parents (Hu et al., 2020).