Estimation results

Table 4 shows the results of entering high quality university. Column (1) is the basic model, which we find that those who are in star high school have the relative odd ratio greater than one, implying they are more likely to enter high quality university than non-star high school students. We also find the students with high academic performance also have the advantage, but private tutoring does not has a significant effect. Being cadres for class or high school activities will have advantage to enter good quality university as the position enable them to increase their working knowledge, responsibility, sense of honor. Reading extra-curricular books also help because it broadens the spectrum of common knowledge.

As for individual specialty, good creativity and good leadership turn out to be a disadvantage. This is because in Taiwan university admission mainly relies on

national-wide written entrance examination. Creativity and leadership are not taken into account and those who spend lots of time in participating student clubs or extra-curriculum activities may diverse their degree of academic concentration.

Good physical strength and stamina does not offer a advantage, while those who are good at adapting to disappointment or frustration will bring a advantage. These may imply that good emotional adjustment to surrounding environment tends to escape easily from external disturbance and concentrate more on school study. Those who are freely to express their own ideas when they disagree to the teachers representing a strong sense of self with the nature of individualism do not give them significant advantage. This may again have to do with the cram education for entrance examination, which emphasizes memorization more than creation and self consciousness. Those who always consult their parents before making a decision representing a passive character have a disadvantage but insignificant.

As for individual ability, artistic ability has no advantage and interpersonal communication skill has a disadvantage. These may imply that social network in high school is useless for entrance exam but diverse time available for school study.

However, good English proficiency and math ability does provide a significant advantage!

In gender, males have advantage over females. Our samples have 51% of

females but only 49.5% in good quality university. In general females have high university enrollment rate but males enter better university more than females. On family background, parents with higher education or students from rich family have a advantage to enter good quality university. These may imply that more educated parents learned more information and care more about their children education as they know that education is an effective means for upper class dominance.⁴ Rich family tends to offer more educational resources to their children, while poor family is usually subject to financial liquidity constraint for their children’s education investment.⁵ This is in accordance with Brewer, Eide, and Ehrenberg’s (1999) finding that students from high income family and have more educated parents has greater opportunity to enter good university.

As different senior high school may apply different standards and criteria for academic performance evaluation, we further add an interaction term between academic performance and school quality, column (2) of Table 3 shows that the coefficient of interaction term is positive and significant. In other words, those students in good high school with better academic performance have a significant advantage to enter good university. This may also imply the synergy effect or peer effect of good high school matches with good students. The current entrance        

4 Chuang and Chen (2011) find that in Taiwan education is an effective way to preserve upper class dominance.

5 We further divide family income into five categories, the results is positive and significant. That is, higher family income will bring a significant advantage to enter a good university.

examination education system plays a very important role to match elite schools with good students. For those students, peer effect is usually strong and positive. Note also that the addition of the interaction term does not change the estimation results of all other explanatory variables. However, the relative odd ratio of school quality drops from 2.194 to 1.275 but remain significant. This implies that excluding positive peer effect of elite school, the brand effect of school quality does exist. Moreover, the school brand effect of 1.275 has roughly the same magnitude as the peer effect of 1.188.

In column (3), we further consider the interaction term between after school supplementary education and students English and math ability, the result show a significant and positive effect. This implies that after school supplementary education helps in particular those students with good English and math ability to enter good university. The coefficient of after school supplementary education becomes negative, implying that the effect of after school supplementary education is not straight forward except for those with certain level of English and math ability. In column (4), we add the total score of university entrance exam and with no surprise we find that higher score in entrance exam will increase the odds to enter good university.

In sum, after controlling for individual specialty and ability, physical and mental

status, and extra curriculum experience in senior high school and further extracting out the peer effect of good school matching good students, we find there exists a brand effect for elite high school to enter elite university. That is, there is a brand

value for the elite senior high school!

Table 4 Estimation results of relative odd ratio for general education track

(1) (2) (3) (4) Elite HS 2.194*** 1.275* 1.281* 1.234

(0.072) (0.176) (0.177) (0.174) Acad score 2.009*** 1.762*** 1.764*** 1.700***

(0.038) (0.066) (0.066) (0.065) Supp edu 1.034 1.032 0.649*** 0.556***

(0.036) (0.036) (0.080) (0.070) Cares HS 1.082* 1.082* 1.081* 1.028

(0.046) (0.046) (0.046) (0.044)

Extra Curr 1.084** 1.084** 1.084** 1.048

(0.034) (0.034) (0.034) (0.034)

Creativity 0.942* 0.943* 0.942* 0.944*

(0.029) (0.029) (0.029) (0.030) Leadership 0.866*** 0.866*** 0.865*** 0.857***

(0.026) (0.026) (0.026) (0.026)

Phy Pwr 1.034 1.034 1.032 1.072

(0.052) (0.052) (0.052) (0.054)

Flex adj 1.071* 1.070* 1.070* 1.071*

(0.043) (0.043) (0.043) (0.044)

Own Opn 0.919 0.923 0.924 0.944

(0.066) (0.066) (0.066) (0.069)

Passive 0.918 0.916 0.916 0.928

(0.050) (0.050) (0.050) (0.051)

Art Ability 0.977 0.978 0.978 0.992 (0.014) (0.014) (0.014) (0.014)

Oral Exp 0.989 0.990 0.990 0.995

(0.018) (0.018) (0.018) (0.018) Interper Com 0.933*** 0.932*** 0.932*** 0.942***

(0.018) (0.018) (0.018) (0.018)

Eng & Math 1.040*** 1.040*** 1.005 0.983

Ability (0.006) (0.006) (0.010) (0.010) Gender 1.203*** 1.206*** 1.204*** 1.198***

(0.034) (0.034) (0.034) (0.035) Fath Edu 1.065*** 1.065*** 1.065*** 1.045***

(0.014) (0.014) (0.014) (0.014) Mom Edu 1.033** 1.033** 1.034** 1.023

(0.015) (0.015) (0.015) (0.015) Low Income 0.807*** 0.808*** 0.807*** 0.827***

(0.023) (0.023) (0.023) (0.023)

Elite HS*Acad 1.188*** 1.186*** 1.158***

Score (0.051) (0.051) (0.050)

Supp Edu*Eng & 1.045*** 1.052***

Math Ability (0.012) (0.012)

Entr Score 1.018***

(0.001)

N 25588 25588 25588 25588

pseudo R² 0.073 0.073 0.074 0.091

Note: Figures in the parentheses are standard deviation; *、**、*** represent 1 statistical significance level at 0%, 5%, 1% respectively.

As there are two tracks of education system, other than general education, we further consider the group of vocational and technical education track. Table 5 shows the estimation results for vocational school to entering technical colleges and universities. The results are very similar to what we have in Table 3 for general education track. Good vocational school or good academic performance has a positive and significant effect on entering good technical university. However, unlike the result in Table 3 after school supplementary education for specialty subjects does provide advantage to enter good technical university.⁶ Being cadres of class and school clubs ,        

6 In vocational track, specialty subjects such as accounting in management field and mechanics in

creativity and leadership, and adjustment to disappointment or frustration have no significant advantage, while free to express own ideas with a sense of individualism becomes a disadvantage. As for individual ability, artistic ability, oral expression, interpersonal communication have a significant disadvantage. This again has to do with the entrance examination system that getting good scores in exams is the most important, time spending in things irrelevant to examination will purely diverse one’s concentration in academic study.

As for family background, in contrast to Table 3 students from rich family has a significant disadvantage to enter good technical university. Under the current two-track education system, for rich family, they tend to send their kids to general university and those who stay in technical university used to be with low ability. On the contrary, for poor family due to financial constraint and hoping to acquire a secure job after graduation more able students tend to choose vocational track and has a higher motivation and interests to enter technical university. Thus, students from poor family have a advantage to enter technical university over those from rich family.

The peer effect of good vocational high school matching good students can also be identified. Controlling for the matching effect, we still find a significant brand effect for good vocational high school. As in Table 3, after controlling for the interaction

engineering field are relatively more important than general subjects of English and math for example.

between after school supplementary education and academic performance, after school supplementary education has a advantage only conditional for those with good

academic performance in specialty subjects.

Table 5 Estimation results of relative odd ratio for vocational education track

(1) (2) (3) (4) (5) Elite HS 4.218*** 1.937*** 1.690*** 1.761*** 1.747***

(0.166) (0.325) (0.288) (0.298) (0.296) Acad score 2.030*** 1.740*** 1.578*** 1.592*** 1.599***

(0.048) (0.069) (0.064) (0.064) (0.065) Supp Edu2 1.334*** 1.333*** 1.127*** 1.186*** 0.801***

(0.050) (0.050) (0.044) (0.046) (0.058) Cares HS 1.047 1.052 0.981 0.997 1.003

(0.055) (0.056) (0.053) (0.053) (0.054)

Extra Curr 1.097** 1.099** 1.079** 1.087** 1.088**

(0.042) (0.042) (0.042) (0.042) (0.042)

Creativity 0.960 0.960 0.963 0.956 0.958 (0.040) (0.040) (0.041) (0.041) (0.041) Leadership 1.019 1.019 1.029 1.022 1.020

(0.041) (0.041) (0.043) (0.042) (0.042) Phy Pwr 0.899 0.894 0.925 0.913 0.908

(0.062) (0.062) (0.065) (0.064) (0.064) Flex adj 1.075 1.079 1.059 1.056 1.058

(0.060) (0.060) (0.060) (0.060) (0.060) Own Opn 0.743*** 0.742*** 0.765*** 0.750*** 0.753***

(0.073) (0.074) (0.077) (0.075) (0.075)

Passive 0.827*** 0.826*** 0.828*** 0.820*** 0.822***

(0.057) (0.057) (0.058) (0.057) (0.057)

Art Ability 0.923*** 0.924*** 0.937*** 0.921*** 0.923***

(0.018) (0.018) (0.019) (0.019) (0.019) Oral Exp 0.924*** 0.926*** 0.943** 0.929*** 0.929***

(0.023) (0.024) (0.025) (0.024) (0.024) Interper Com 0.946** 0.945** 0.944** 0.941** 0.939**

(0.025) (0.025) (0.025) (0.025) (0.025) Eng & Math 1.109*** 1.109*** 1.089*** 1.112*** 1.111***

Ability (0.008) (0.008) (0.008) (0.008) (0.008)

Gender 1.250*** 1.259*** 1.335*** 1.265*** 1.265***

(0.047) (0.047) (0.051) (0.048) (0.048) Fath Edu 1.067*** 1.066*** 1.073*** 1.078*** 1.076***

(0.020) (0.020) (0.020) (0.020) (0.020) Mom Edu 1.021 1.020 1.028 1.027 1.026

(0.020) (0.020) (0.020) (0.020) (0.020) Income 0.951** 0.951** 0.954** 0.953** 0.952**

(0.020) (0.020) (0.020) (0.020) (0.020) Elite HS 1.257*** 1.254*** 1.273*** 1.272***

*Acad Score (0.061) (0.062) (0.062) (0.062) Eng & Math 1.008***

Score (0.000)

Entr exam 1.006*** 1.005***

(0.000) (0.000)

Entr Exam 1.004***

*Supp Edu2 (0.001)

N 21391 21391 21391 21391 21391

pseudo R² 0.134 0.135 0.167 0.156 0.158 Note: Figures in the parentheses are standard deviation; *、**、*** represent 1 statistical

significance level at 0%, 5%, 1% respectively.