So will elite universities really create elite lives? This time, the administrative data of Chile, the most unique country in South America, is selected to identify the causal effect of elite university admission on graduates' future career development, and the breakpoint regression design is used to analyze whether the admission of elite (business) projects affects the opportunities of students from different backgrounds to enter top positions and top income classes. Previous conclusions:
First of all, a few business graduates from elite universities with extremely low admission rate occupy a large proportion of corporate leadership positions and the highest income level. Only 65,438+0.8% students were admitted to the business degrees of three elite universities, but these students occupied 465,438+0% senior management positions, 27% top 65,438+0% income, 39% top 0. 1% income and 45% top 0.0.
Second, the probability of business degree students in elite universities entering the top 0. 1% income is 16.6 times that of medical degree students with medium admission rate. Therefore, the admission of elite universities intensifies the class inequality caused by gender and socio-economic status.
Third, gender and family income are important determinants of elite business degree admission. Entering an elite university degree increases the average income of boys from wealthy families and the probability of entering senior positions. Boys from wealthy families are 3.4 times more likely to enter leadership positions than boys from other family backgrounds, and 4.2 times more likely to enter the top 0. 1% income class.
Fourthly, the research of breakpoint regression design shows that the admission of elite university degree increases the probability of students obtaining leadership positions by 45%, and the probability of students entering the top of income by 0. 1% and by 5 1%. However, this kind of reward only works for boys with high family status, and the chances of such students getting high positions and high incomes are 54% and 69% higher than other students.
Fifth, it is difficult to explain the high returns of men from high socio-economic classes by job skills, high-income industries and college graduation rate. However, high flyers's classmate effect may be an important mechanism for elite universities to achieve high returns. The results show that when male graduates from elite universities are also born in wealthy families, these graduates will be more likely to enter the management of their classmates' companies.
Among universities in Chile, University of Chile (UC) and Catholic University of Chile (PUC) are the two most famous elite universities. High school students in Chile passed the standardized college entrance examination last year. After getting the test results, students can fill in many colleges and universities and professional volunteers. Relevant schools and their majors grade students according to the test results. If students are not admitted to the first choice, they will be postponed to the second choice. If the conditions of the second choice are met, the third choice will automatically become invalid. The list of students who meet or exceed the admission line of a university and the list of candidates who are close to the admission line will be published in the newspaper. This breakpoint regression design is to compare the students at the bottom of the admission list with the students at the top of the candidate list.
Chile's college entrance examination system is quite procedural, and students can't change their admission results after the college entrance examination unless they wait for another year to retake the exam. The admission line of a school changes every year, depending on the number of applicants and the number of places allocated by the university, so it is difficult for students to accurately predict the admission score. Therefore, it can be said that the configuration variable (admission score) of breakpoint regression in this design has not been obviously manipulated, which meets the first prerequisite of breakpoint regression. In addition, the admission of elite universities may be influenced by the family background of high school students. Because there is no data on college students' family income and parents' education level, we use the type of high school students' study to replace their family background. Private high schools are expensive, and families that can afford them should be high-income families.
The university application and admission data we selected this time include three data. The first data is all the data from 1980 to 200 1 who took the college entrance examination. The second data is the data of admission results of all university majors in Chile from 198 1 to 200 1. The third data are 1974 to 200 1 the admission results of the law, engineering and business majors in elite universities in Chile. Using the high school information in university application records, we matched the information of 79% elite university students from 1980 to 200 1, and 60% elite university students from 1974 to 19 1. We also used the web crawler to capture the data of senior executives and board members of listed companies in Chile from 1975 to 20 13, and finally got 10220 company leadership positions, of which 2522 were applicants from elite universities after 1974. We also used Chilean tax data to match the income data of 94% university applicants during the period of 1974-200 1, including labor income, pension, rent, capital income, dividends and other income. Based on this income data, the main dependent variable is constructed, that is, whether the income of college applicants is in the top 0. 1% of the income in 20 13 years (that is, the annual income reaches or exceeds 330,000 US dollars).
This breakpoint regression design is to compare the differences in future positions and incomes between the students at the bottom of the admission list and the students at the top of the candidate list in two elite universities in Chile. The model structure of breakpoint regression design is as follows:
When studying the influence of admission to elite universities on students' positions, the dependent variable Yipc indicates the position status of student I applying for degree program P in application group C. There are two methods to measure the dependent variable: one is the number of times the student has served as a board member or an executive position, and the other is whether he has served as a board member or an executive position. When studying the influence of admission on students' income, Yipc= 1, that is, students' income is 0. 1% before income. Dipc represents the difference between the score of student I's degree application and the admission score. Aipc is the driving variable of breakpoint regression, where dipc >: 0, Aipc= 1, that is, students are admitted to the degree program, otherwise Aipc=0. Because it is not excluded that some students appear in the data many times, for example, the student is admitted to UC after being rejected by PUC, the breakpoint regression uses the cluster standard deviation at the student level. The discussion of optimal bandwidth will not be repeated here. In the choice of breakpoint bandwidth, the bandwidth of position model is "bandwidth = 10 point" and the bandwidth of return model is 20 points.
A basic assumption of breakpoint regression is that the driving variables are not manipulated by individuals, that is, the driving variables are evenly distributed on both sides of the breakpoint. Draw and observe the dipc density distribution on both sides of the breakpoint. The histogram of the score in Figure 3 shows that there is no obvious jump to the right of the breakpoint. The sub-sample test also found that the scores of students studying in private high schools and students studying in non-private high schools did not jump significantly on both sides of the breakpoint.
From Figure 6, we can clearly find the phenomenon of jumping near the elite score line. Being admitted to elite colleges has a significant positive impact on students' entry into leadership positions and top income. PanelA's mixed sample of men and women. As shown in Figure 6, entering elite colleges and universities increased the proportion of students who entered the top 0. 1% by 50%, and the number of students who obtained leadership positions increased by 44%. Comparing PanelA and PanelB, we can find that the positive influence of elite colleges on income and position is mainly determined by the background factors of male students in private high schools. After being admitted to elite universities, the proportion of male students' income of 0. 1% increased by 50%, and the number of leadership positions increased by 54%. On the contrary, entering an elite university has little effect on girls' entering the high-income class, but it has a slight negative effect on women's chances of entering leadership positions. In addition, the spillover effect of elite university admission is only applicable to boys who graduated from private high schools, which increases the proportion of such students entering the high-income class by 69% and the number of leadership positions obtained by such students by 54%, but elite university admission has no obvious effect on students who graduated from non-private high schools.
We also tested the influence of elite university admission on students' chances of entering the former 1% and 10% classes and their actual income (log). The results are very similar to the trend in Figure 6. Being admitted to elite universities only increases the income of boys who graduated from private high schools and the number of leadership positions. Among them, the proportion of boys who graduated from private high schools entered the pre-1% income level reached 33%, and the proportion of boys who entered the pre-10% income level reached 13%, so their actual income increased by 20%. For girls and boys who graduated from non-private high schools, the influence of elite universities on their top incomes of 1% and 10% is close to zero.
We analyze whether the influence of elite university admission changes with different universities, majors, students' other family backgrounds and other income measurement dimensions. The research finds that: firstly, the influence of PUC and UC admission on leadership positions is not much different from the actual income, but entering the high-income class is mainly produced by PUC admission. Secondly, there are significant professional differences in the spillover effects of positions and incomes, and the spillover effects of law and business degrees are stronger than those of engineering degrees. Thirdly, we find that the job spillover benefits of elite universities do not seem to exist for those departments with lower incomes, such as government departments and education departments. Fourthly, the spillover effect of elite universities mainly affects the graduates of the seven best private high schools, which are characterized by high tuition fees, high scores and strict admission. But for public high schools, even if the admission is strict and the score is high, the result seems to be useless.
Why only boys who have graduated from private high schools will have higher income and job prospects after being admitted to elite universities? The possible explanations here are: first, private high schools are better prepared for their studies; Second, geographical preferences and constraints; Third, the choice of high-income occupations; Fourth, private high school graduates do well at the university level; Fifth, private high school graduates have established valuable classmates' relationship in university study.
Our test shows that there is little difference between boys in private middle schools and boys in non-private middle schools in math and reading. Private middle schools are only 7.8 points and 4.8 points higher than non-private high school students in two grades, and the gap between them is even smaller in the academic achievements of the reported degree programs. Therefore, it can be said that the learning ability of boys with different backgrounds in high school is very similar, which will not affect their learning ability in college and their future career. In terms of preferences and constraints, 82% of private high school students come from Santiago, Chile, and 70% of non-private students come from Santiago. Then we test whether these factors directly affect the admission effect of elite universities. By constructing the interaction between elite university admission and the above factors (scores and geographical location), it is found that except for private high schools, high school mathematics and reading scores and geographical location factors have no significant influence on income and leadership positions.
The higher career and income prospects of students admitted to elite universities may be because they choose higher-paying careers. The test results show that there is not much difference in career choice between private high school graduates and non-private high school graduates. 47% of private high school graduates and 465,438+0% of non-private high school graduates choose three major business sectors for employment (real estate/leasing/business services, wholesale/retail, finance), but a slightly larger proportion of non-private high school graduates work in the government and education departments. The test results of breakpoint regression show that the admission of elite universities has little influence on students' engagement in the above three major occupations. Except for the government and the education industry, the admission of elite universities has little influence on the employment choice of other industries.
The sample is limited to the sample of students who apply for medical degrees from elite universities. Breakpoint regression analysis shows that, unlike the business degree from elite universities only affects boys who graduated from private high schools, the admission of medical degrees has a significant positive impact on the income prospects of boys and girls and private and non-private high school students (before entering the 10% income class, before entering the top 5% and actual income), but it has a significant positive impact on the income of students before 0.65433. Therefore, although the admission to the non-business degree of elite universities can improve the future income of students of different genders and different family incomes, it has no obvious help for them to enter the top income.
The analysis shows that, on the whole, the graduation rate of students is very high. In business and medical degrees of elite universities, the graduation rate of women is higher than that of men. The graduation rate of boys who graduated from private high schools with the same education is higher than that of boys who graduated from non-private high schools. Women with business degrees are more likely than men to graduate smoothly. The above findings show that the academic performance of elite university degrees is not the main reason for the high return of boys who graduated from private high schools.
The last mechanism we need to test is the classmate relationship of elite university students. This influence channel is not put forward out of thin air. The existing research and some authoritative media reports show that classmate relationship has an important influence on individuals entering the top management of enterprises. If your classmate holds an important position in the company, then you may enter the upper level of the company with the help of your classmate and get a high income. If you and your classmates work in the same company, because of the previous classmates, your work is more harmonious and your work performance is correspondingly higher. At this time, you have a greater chance of getting a high income when you enter the top. The most important thing is that you and your classmates are both rich men, and this kind of classmate relationship may be more valuable.
In order to identify the influence of classmate relationship on career return more cleanly. We compare the proportion of a pair of college students holding the same leadership position in the same company with the proportion of a pair of non-classmates (studying the same degree at different times and studying different degrees at the same time) holding the same leadership position in a company.
The result is shown in Figure 9. In Figure A of Figure 9, the proportion of male students with the same education and different education in the same company who hold leadership positions is compared. The left picture shows all private high school graduates, and the right picture shows 7 elite high school graduates. The results of the two pictures show that the proportion of a pair of students studying the same degree together in the same company is significantly higher than that of other students. With the increase of age difference, the proportion of reading classmates will decrease linearly in the future, but it is still higher than that of not reading classmates. Obviously, if two students have different learning levels and lack this kind of classmate bond, it is unlikely that they will have the opportunity to lead a company together.
Figure 9 PanelB uses the breakpoint regression design to show the influence of elite university admission on the chances of students who apply for degrees at different times but have no degrees to become the same leaders in the future. The results show that there is no significant difference in the chances of students who are not admitted to elite university degree programs, whether they are classmates or not, to lead a company. However, when they are admitted to elite universities, whether they are classmates has a great influence on their grades. Students with clear peer relationship are more likely to become peer leaders of the company in the future, which is significantly higher than those with the same education (the same prog, dif year) and different education (the same year, dif prog). If there is no classmate relationship, even if students are admitted to elite universities, the chances of becoming company leaders in the future are no higher than those who are not admitted to elite universities.
The above picture only illustrates the difference between private high school graduates and future leaders. Then, does the classmate relationship between students who graduated from non-private high schools also help them to become company leaders in the future? We have established single-difference and double-difference models to test the causal effect of the classmate relationship of non-private high school graduates on students' position. The forms of single-difference model and double-difference model are as follows:
Yiji = 1 indicates that a pair of students I and J hold leadership positions in the same company, ti and tj indicate the years when I and J applied for degrees, and pi and pj indicate the elite university degrees that I and J applied for respectively. If |titj？ | = g, then Cg (ti,? tj？ )= 1。 If pi? = pj？ , then S( pi, pj? )= 1。 The coefficient of difference is πg, that is, the influence of classmate relationship on the probability of a pair of classmates holding leadership positions together. The test results are shown in figure 10. Only the classmate relationship (the same queue) improves the probability that students * * * share leadership positions. Other relationships, whether students are in different years, have different academic qualifications in the same year, or have different academic qualifications in different years, have no influence on the probability of students sharing leadership positions. The results of single difference show that the relationship between classmates increases the probability of two people leading the same company by 1.26 times. If both students come from elite private high schools, the probability of them leading the same company will increase by 1.37 times. In addition, the results in the last two columns of table 10 show that classmate relationship seems to only affect students in private high schools. If one of a pair of classmates comes from a non-private high school, the classmate relationship seems unreliable, which has no obvious influence on the future company leadership position and the coefficient is very small.
In short, this part of the study shows a cold social reality: the relationship between classmates is particularly important, and the origin is more important. After a private high school graduate is admitted to an elite university, it is easier for him to lead a company with his classmates from the same wealthy family. A student who comes from an ordinary family and graduated from a non-private high school, even if he is admitted to an elite university degree, even if his classmates come from a rich family, can't lead a company with his classmates. It is very likely that you are the abandoned party in the classmate relationship. ?
We use data from Chile to tell a sad story: you have been admitted to a good school through hard work, but your gender and family background may hinder your further development, but you are helpless. The professional income of elite universities is likely to be monopolized by men and students from wealthy families. Children from poor or ordinary families who are admitted to elite universities may change their fate, and their income may be close to the top income, but it is still impossible to become the top elite in society and enter the highest income class. Especially in the commercial and financial industries, the highest-paid class and company leadership positions are occupied by men who graduated from elite universities and came from wealthy classes. In the top class and professional circles, personal relationships, such as classmates' relationships, have become particularly important. Your background doomed you to be unable to integrate into the circle of classmates from your wealthy family. As a friend once said, people who say that big cities will succeed by their ability may not know that the connections of the top people are more important, but you just don't know. As written in the lyrics of "Waiting for you to finish class": "Maybe I shouldn't be in your world. When you receive the love letter, it means that I have gone far. "
The data interest theory, interpreted with scientific data, tells you the principles behind some social phenomena.