Vision Exchang Hub #1: Selection Bias and the Influence of Education Resources on Class Stratification

1. Selection Bias and Definition

Imagine yourself bursting into the lecture hall in the economics department of UChicago, arms flailing and feet stomping, eyes wide with joy. “Economists, students—hear me!” You shout, practically vibrating with excitement. “Today, I received my first research fund! My very first!” After a dramatic pause with a deep breath, you gesture towards the astonished, maybe a little freaked-out students. “The goal of the research is to determine whether students who, like you all, are pursuing advanced education will be more likely than those who attend community colleges to land a job with a higher salary!” You lean forward with your eyes gleaming, “Will your exhausting studies in UChicago be a sanctuary from econometrics or just another burden?” You raise your arms even higher as if you are almost showing a surrendering gesture. “I will find out with the full rigor of economic research!” But how will you define the parameters of a good educational system? Have you considered that students differ even before they step foot on campus? Did you truly conduct rigorous research if you overlooked the fact that most students at UChicago come from privileged backgrounds? They are probably more motivated, engaged, and enthusiastic about studying than the average American college student. These factors that you neglected are being defined as “selection bias.”

The key characteristics of the occurrence of selection bias can be concluded into the following points:

1) Non-random sampling: Whether intentional or not, researchers sometimes select experimental subjects that share an irrelevant, identical factor influencing the outcome of interest, which can lead to skewed or misleading conclusions.

2) Failure to represent the entire population: It is challenging to generalize the results when selection bias occurs because the findings reflect the unique characteristics of the sample rather than the intended population.

3) Exaggerated/ weakened effect of certain factors: A study’s selection bias may lead to an overestimation or underestimation of the importance of particular factors. Unaccounted-for variables have the potential to magnify the apparent impact of the main factor under investigation when comparing results. This may produce results that incorrectly credit variations to the variable under study when, in fact, the results are being influenced by other underlying factors.

Selection bias can also be expressed in terms of math notations:

SB=E[yi0|D=1]-E[yi0|D=0]

Where:

• E represents the expected value of the factors within the brackets.
• 𝑦𝑖 is the observed outcome for individual 𝑖 
• 𝑦𝑖1 is the outcome they would have had if they had received the treatment
• 𝑦𝑖0 is the outcome they would have had if they had received the control
• D=1 refers to individuals who are part of the treatment group (those who received the treatment).
• D=0 refers to individuals in the control group (those who did not receive the treatment).

Selection bias happens when the difference E[yi0|D=1]-E[yi0|D=0 is a non-zero value, and then people who receive the treatment differ systematically from those who don’t. In this case, the difference in the expected outcomes arises solely from the fact that the individuals belong to different selected groups, even though they receive the same treatment. The variation in outcomes stems from the underlying characteristics of the groups or other external factors rather than the treatment itself. This suggests that the observed difference is driven by inherent differences between the groups, such as socioeconomic background, prior conditions, or other unobserved variables, rather than the effect of the treatment. Thus, the selection bias reflects the influence of these pre-existing group characteristics on the outcomes, independent of the treatment received.

2. Hypothesis #1: Selection Bias Overestimates Educational Resources

The immobility of a college education, according to a large number of people, is a major factor in class stratification, which has become a hot topic in political and economic discourse. Wealthy children are more likely to attend prestigious universities, while less fortunate children are more likely to receive an average education. This can be proved upon data provided by the NPR news: “Among a number of other discoveries, the economists find that kids from the richest 1% of American families are more than twice as likely to attend the nation’s most elite private colleges as kids from middle-class families with similar SAT scores” [1]. Now, the question is this: Do students who attend college well actually have a higher chance of landing a high-paying job? 

If A study were to have the purpose of defining objectively the extent of the ability a good education would have on the opportunity of receiving a high-wage job, it would first of two groups: one treatment group with students that received a good education from top universities, and a control group of those who went to community colleges. Within mathematical terms, we would define the treatment group as D=1 and the control group as D=0. To examine whether or not selection bias occurs in this process, we would first gather the data of E[yi0|D=0], meaning the expected wage rate if the ones that received underprivileged college education actually received underprivileged college education, which would be the average wage rate of the control group. E[yi0|D=1], or the expected wage rate of those who currently have a good education if they were to receive an underprivileged education, is the second piece of data that we would require. This is where the contentious issue arises: in order to obtain official data from this group, we would need to coerce those who have accepted an offer to attend UChicago to voluntarily forfeit it and enroll in a community college, which sounds impossible for operation for no one would be willing to sacrifice an offer from a prestigious school in real life. E[yi0|D=1] would then need to be derived from an estimated value, but what standards and variables need to be taken into account? This value’s ambiguity turns the study into a question with a questionable answer rather than one with a conclusive one.

On the one hand, some researchers think that the challenges associated with seeking a high-paying job have led to an overestimation of education’s effectiveness. Before entering college, a student from a reputable college may already have certain advantageous qualities: they may be more resilient, intelligent, and committed to their studies. A higher likelihood of obtaining these job opportunities may also stem from their comparatively affluent family background, which may have given them access to better networks, extracurricular activities, and enrichment, as well as stricter family rules resulting in higher moral standards. As a result, in this circumstance, it is believed that merely the discrepancy in different qualities of college education does not contribute fully to the difference in salaries in jobs.

2. Hypothesis #2: Education Resources Impact Largely on Class Stratification

On the other hand, people believe educational resources are more than just the accumulation of knowledge; they serve as gateways to a wealth of experiences that create more advantages for people in terms of job applications that extend beyond the classroom. Particularly, prestigious colleges and universities offer access to a wide array of social networking, career development programs, extracurricular activities, and leadership opportunities. These resources help students cultivate not only intellectual growth but also essential life skills such as communication, collaboration, and problem-solving. On the other hand, average schools might not be able to offer such opportunities.

Networking with peers, faculty, and professionals through university events, internships, and clubs can open doors to career opportunities and mentorships that are crucial for personal and professional development. These kinds of encounters frequently result in lifelong connections that influence a student’s prospects long after they graduate. Fraternity membership in college lowers GPA by 0.25 points but increases future income by 36%, according to research data [2]. These findings clearly imply that significant increases in social capital are causally produced by fraternity membership. It’s also true that fraternities at elite universities provide even greater benefits than those at regular universities.

More than just networking, a well-rounded college experience fosters students’ emotional and social development by fostering their sense of autonomy, self-awareness, and resilience. In order to give students a more well-rounded education, universities frequently provide a wide range of support services, including career coaching, counseling, and community engagement initiatives. These services assist students in overcoming their emotional and personal obstacles. Through group projects, research opportunities, and branching out into new areas, the college also fosters critical thinking and creativity. Students gain the ability to challenge presumptions, exercise critical thought, and come up with creative solutions to challenging issues in this setting. These skills are invaluable, not just in the workforce but in everyday life.

Last but not least, the signaling effect is a key reason why some prefer prestigious colleges over average ones. The concept of the “signaling effect” is used in economics and other fields to describe how one party, known as the sender, can signal their underlying qualities to another party, known as the receiver, by acting in a certain way or exhibiting a particular attribute. This is frequently discussed in relation to education or the job market, where people use their academic accomplishments to convey to potential employers their aptitude, competency, or output. When looking for employees, employers would find it exceedingly challenging and inefficient to select their staff by trying to comprehend each candidate’s unique set of skills. Employers have a tendency to believe university judgments for enrollments because they think superior schools select more capable students and mediocre schools select less capable students. Some may argue that it is overly generalizing to say that prosperous entrepreneurs like Bill Gates did not complete their college educations. In Bill Gates’ case, he is so exceptional that he doesn’t need documentation of his college degree. All the same, these are very rare occurrences because not everyone is as creative and capable of leading as Bill Gates is, even without a college education. Thus, students who attend a prestigious college and are, for instance, branded as intelligent “Harvard Graduates” are more likely to have their job offers accepted by the general public in the majority of situations, according to the “signaling effect.”

Ultimately, how do the aforementioned points connect to the idea that educational resources significantly influence class stratification? We have previously demonstrated that children from wealthy families are more likely than those from average families to attend a prestigious university. Prestigious universities offer resources and education of a higher caliber, so their students have a greater chance than others of landing well-paying jobs. As a result, wealthier families continue to benefit financially, which perpetuates socioeconomic inequality between generations.

4. Conclusions

Both theories are supported by the observed phenomenon: admission to a prestigious university may be largely dependent on inherited privilege; the question is whether the resources these institutions provide are sufficient to maintain and amplify these advantages. Were these advantages that resulted in students receiving better job offers established prior to attending college? Or does a college education significantly amplify the advantages wealthy students are born with? There is still much debate on this subject because there is insufficient data to provide a definitive response. It depends on how people define and interpret the term “educational resource,” and perhaps political stances play a small role in the perspective.

5. References

[1] “Affirmative Action for Rich Kids: It’s More than Just Legacy Admissions.” NPR, 24 July 2023, www.npr.org/sections/money/2023/07/24/1189443223/affirmative-action-for- rich-kids-its-more-than-just-legacy-admissions. Accessed 6 Oct. 2024.

[2] Blumberg, Yoni. “Here’s How Much More Money You Could Make Just from Joining a Frat.” CNBC, 9 Oct. 2017, www.cnbc.com/2017/10/09/joining-a-fraternity-could-boost-your-income-significantly.html#:~:text=In%20a%20paper%20titled%20%E2%80%9CSocial,traditional%204%2Dpoint%2Dscale. Accessed 6 Oct. 2024.

‌