Do engineering majors earn higher wages than business, social science, or humanities?
INTRODUCTION
All over the world, millions of people engage in different works ranging from casual to formal employment to earn a living. As employees exchange pay with their skill to the employers, they directly or indirectly drive the world’s economy. Different forms of payment depend on the type, amount, or quality of work done. Some employees get paid through allowances, wages, bonuses, or even monthly salaries. The money paid to workers enables them to plan their financial endeavors for the present and future times. Besides, different employees have different ambitions in life, which include career progression or turnover that is profoundly affected by the amount of pay they receive. Therefore, it is vital to understand the determinants of a higher wage hence being able to decide on the future and self-development.
Education plays a more significant role in deciding the type of job one may land. Most formal employment opportunities require individuals with trained skills and knowledge in a particular field. The collective consciousness is that people who study STEM (Science, Technology, Engineering, and Mathematics) are more likely to earn more than their peers who undertake other courses such as in the field of arts. However, all academic majors are essential in driving the economy since every expertise is essential in one or another. According to Robst (2007), individuals holding STEM degrees are likely to receive less pay than those from liberal majors as technology progresses.
According to several scholars, nations that provide competitive salaries tend to attract the best minds in a globalized economy hence improving their economic growth. Therefore, it is essential for policymakers, financial stakeholders, and the government to understand the factors that affect the salaries offered. Similarly, students should be well versed with the knowledge on the salaries different academic majors attract before joining to study and train in them.
Purpose
This study is focused on understanding the effect of an academic major on the wages offered to citizens on employment.
Research Question
Do engineering majors earn higher wages than business, social science, or humanities?
Research Significance
This research’s outcome complements the existing literature on the earnings to post-secondary education majors. Similarly, evidence from this study presents practical knowledge to policymakers on the distribution of income based on different fields of study by examining the potential causes of heterogeneity in the rates of payoffs by education majors. This study can also inform debates over the interventions of governments in addressing market frictions and visible mismatches in the college majors. For example, most governments all over the world are calling for educational reforms in a bid to increase the number of students graduating in STEM fields. This study also guides students who often make their educational choices based on limited information on the earnings in the labor market.
LITERATURE REVIEW
This section provides a critical review of previous studies on the relationship between academic majors and the number of returns received. The implications in the choices of both college and high school majors have attracted several academic literature and modeling regarding wages. The context of the review of the past literary works is set by analyzing scholarly research on the sociology of the labor market. A keen look at studies dwelling on the relationship between college majors and careers is paramount. Similarly, a review of literature on classical concepts such as human capital and exploitation gives a clear understanding of the college majors’ implications on the wage distribution scale.
Altonji et al. (2012), in their research paper, provide an in-depth analysis of theoretical literature as well as empirical modeling of the heterogeneous nature of education and how it connects with specific occupation paths. The scholars’ study develops a model that draws a distinguishes varying education choices and their causal effects on wages. Decisions made under uncertainties about ability, knowledge, and preferences lead to different possibilities on the wage effects of various education outcomes (Altonji et al. 2012). An increase in the payoff in one field others held constant, say engineering, increase the number of individuals registering for that particular course. Conversely, the move reduces the number of individuals interested in other majors, for example, humanities (Altonji et al. 2012). Altonji et al. (2012 report that there are significant gaps in wage returns across different fields that attract students with similar grades to higher learning institutions.
A considerable variation exists in the earnings of college graduates regarding their field of study. (Arcidiacono, 2004). An agreement to the study findings by Altonji et al. 2012 and Ransom (2014) is echoed by Freeman and Hirsch (2008). In their study, the two scholars prove that the number of graduates in a particular college major concerns the idea of the occupation and the knowledge of the market payoff in the specific field. The scholars add that occupational choice is highly influenced by the skills and knowledge acquired during the study. College graduates have more experience in suitable careers, including the number of market payoffs than individuals who do not attend college (Freeman and Hirsch, 2008). Ransom (2014) reported that occupations of persons holding degrees in specific fields explain the high pays in those majors. However, high-paying in the areas of engineering and physics show varying patterns hence cannot be defined by the occupations held by people in those majors (Ransom, 2014).
Kirkeboen et al. (2014) study why people choose different majors and the resulting payoffs to the choices made. The scholars’ research assesses whether people tend to select fields in which they seem to possess a comparative or absolute advantage. Different areas have significantly varying payoff institutional differences and peer group quality notwithstanding (Kirkeboen et al. 2016). Furthermore, Kirkeboen et al. (2016) provide evidence that individuals have tended to choose fields in which they have a competitive advantage. Their findings agree with Altonji et al. (2012) by suggesting substantial heterogeneity in the returns to studying different college degrees depending on selection. By examining the heterogeneity in levels of earnings according to educational choices, Kirkeboen et al. (2016) determine the magnitude of the payoffs and, at the same time, find that people sort into fields in which they possess an absolute advantage.
According to Yao (2019), the gap present across college majors in terms of earnings is comprehensive and substantial. Yao’s research focused on the mode of generating field premiums and the wage gap across college majors. Heterogeneity in cognitive skills influences the college field choices made by individuals. The increase in literacy and numeracy tends to sway people away from less-skilled courses such as arts, humanities, and education, towards STEM majors (Yao, 2019). According to Yao, there are gender differences when choosing a major in terms of skill, and men have stronger preferences for high skilled majors in STEM than women. The heterogeneous skills of people in the different graduate majors are also reflected in the earnings gap between the majors (Van de Werfhost, 2002). Werfhost provides evidence that numeracy skills are essential for the labor market outcomes than literacy skills. Further, Werfhost suggests that the heterogeneity in skills, together with the variations in course contents, leads to a difference in the professional career choices across majors, subsequently influencing earnings. The field premium is, therefore, generated from the process.
According to Yao (2019), there are still significant field premiums, even after correcting skill heterogeneity. The phenomenal may arise due to a likelihood of the mismatch between college major studied and employment. A similar argument is by Robst (2007), who suggests that mismatched occupational employees receive less pay than well-matched workers having the same degree field. Therefore, premium differentiation can result from the level to which employees in an area are mismatched. According to Robst, individuals working in fields such as engineering, health professions, and law have a high likelihood of being in a related field to their college majors hence low mismatch levels and, therefore, explaining high premiums. On the other hand, Yao (2019) argues that the skills from such types of majors could be highly valued in the market hence the emergence of high premiums. Such premiums could be regarded as occupational labor market rewards available for their knowledge gained in college together with practical skills (Zafar, 2013).
According to several studies, a variation in wages in the labor market could also be a result of other factors such as race and experience. According to Altonji and Pierret (2001), individuals who have been in a particular field for more years could earn more payoffs than new entrants. This could also lead to variations of earnings between different areas of study, taking into account that people with much experience from a major traditionally considered to have low payoffs could be earning more than new entrants in fields considered to have significant returns (Altonji and Pierret, 2001). Factors such as gender and race form part of discrimination in the labor market capital inclination (Blau and Kahn 2017).
DATA AND METHODS
This section is concerned with the research design and the methodology employed. The data collection procedure and the variables of the dataset used in the study are explained. Lastly, the data analysis technique is outlined.
Design
The research design used for the present study was a longitudinal design technique involving different college majors and their variation in payoffs across different times. Differences between the wages in various fields of study are measured through a passive approach. The research uses data at specific times to estimate the heterogeneity in the earnings from the college fields. Therefore, the study’s findings entail being able to predict the amount of salaries a particular college major can influence in a specific time.
Population and sample
The target population consists of employees in the American labor market, possessing degree-level education. The study uses a dataset from a general population of American citizens. A sample of 1067 individuals was randomly chosen from publically available data.
Variables
The primary variables of the research are the undergraduate degree major, the level of education, and the amount of wages earned. Other variables include sex, unique respondent identifier, taken English as a second language, and currently active certification.
Unique respondent identifier: The variable represents each respondent of the survey in the sampled dataset.
Sex: The variable represents the gender of the respondent. It is an independent variable that should predict the wages that respondents receive.
Level of school completed: The variable represents the highest level of education the respondent has completed. The variable was categorical consisting of No high school diploma category, High school diploma, Alternative high school credential, Less than one year credit, One or more years of college, Associate’s degree, Bachelor’s degree, Master’s degree, Professional degree beyond a bachelor’s, and Doctorate degree category. It is an independent variable that is used to estimate the amount of earnings one might receive. The levels were coded from 1 to 10.
Taken English as a second Language: The variable seeks to understand the communication medium of respondents and whether the language used has an effect on the subsequent salaries earned. The variable was binary coded as 1 and 2.
Field of study for the highest level of school completed: The variable represents the college majors of respondents in the sample. The variable has 24 categories including fields in business, education, humanities, engineering, arts, communication, health, among others. It is an independent variable that could determine the total amount of wages earned. The categories were coded 1 to 24.
Currently active certification: The variable represents whether the respondent has an active professional license. However, the licenses do not include business permits offered by governments. The variable is an independent variable that could determine the salaries earned by respondents. The variable was binary coded as 1 and 2.
Earnings in the past 12 months: The variable represents the earnings received by the respondent including salary, commissions, bonuses, and allowances in the past 12 months before deductions. The variable is broken down to nine categories coded 1 to 9. The lowest category is between 0$ to 10,000$, whereas the highest category is in the range of 150,000$ or more. This study uses the variable as a dependent variable.
Data Collection
The data employed in this study was secondary data sourced from a public data repository named kaggle.com. The website contains numerous datasets that are publically accessible to students and scholars. The dataset does not contain raw data values. Instead, it has data that has been pre-processed. The dataset includes data representing the college majors of American citizens and how earning is distributed among the fields of study. Secondary data was preferred to save on time and extensive costs that would have been incurred when collecting primary data. Similarly, secondary data has been pre-cleaned and stored in a format that is easy to use for academic work. Further, the data could not have been collected easily, considering that it consists of information on finances, which can prove to be a sensitive matter among different people.
Data Analysis
Data analysis was performed in the STATA software statistical package. The software is widely used to produce quality and integrated statistical outputs, including easy to understand graphics. The primary data analysis procedure is the use of descriptive statistics to understand the nature of the data and the relationship between the variables. Correlation and multinomial logit regression analysis are carried out.
A regression model is developed to predict the mid-year career salaries. Below is the theoretical model to predict the category of earnings under which an individual falls. The independent variables in the model include sex, highest level of education attained, educational field of study, and the professional certification in the field of employment.
FINDINGS
The table below represents the distribution of the starting median salaries across undergraduate degrees.
Descriptive Statistics
| Variable | Obsv | Mean | Std. Dev. | Min | Max |
| Sex | 1067 | 1.684161 | .465067 | 1 | 2 |
| eduattn | 1067 | 5.137769 | 2.392677 | 1 | 10 |
| Edufos | 1067 | 7.717901 | 7.996369 | -1 | 24 |
| Enroll | 1067 | 1.165886 | .5266253 | 1 | 3 |
| Eslcla | 1067 | 1.962512 | .1900441 | 1 | 2 |
| Readcla | 1067 | 1.982193 | .1323112 | 1 | 2 |
| Cnmain | 1067 | 1.776007 | .4171126 | 1 | 2 |
| Eearn | 1067 | 3.415183 | 3.320169 | -1 | 9 |
Source: STATA
The table above represents some summary statistics about the dataset. The total number of respondents in the dataset was 1067. The table shows the mean and standard deviations of the data in the data set. Similarly, the minimum and maximum response values are illustrated in the table. A table showing the gender distribution in the sampled dataset is given below.
| Sex | Frequency | Percent |
| 1 | 337 | 31.58 |
| 2 | 730 | 68.42 |
| Total | 1067 | 100 |
From the table, there were 730 males representing 68.42 percent of all the respondents in the study. On the other hand, a total of 337 females representing 31.58 percent of all the study participants were involved in the survey. From the numbers, the study was biased towards the male sex than the female. A distribution of the levels of education completed is shown in the table below.
| Education Attained | Frequency | Percent |
| No high school diploma | 72 | 6.75 |
| High school diploma | 186 | 17.43 |
| Alternative high school | 40 | 3.75 |
| Less than one year college | 79 | 7.40 |
| More than one year college | 175 | 16.40 |
| Associate’s degree | 105 | 9.94 |
| Bachelor’s degree | 251 | 23.52 |
| Master’s degree | 112 | 10.50 |
| Professional degree | 26 | 2.44 |
| Doctorate degree | 21 | 1.97 |
| Total | 1067 | 100 |
Source: STATA
From the above table, the highest number of respondents have a bachelor’s degree representing 23.52 percent of the total number of respondents in the study. Those having a high school diploma were second highest representing 17.43 of the sampled participants. Those with more than one year college came in third, while individuals holding a master’s degree were fourth. Individuals holding a doctorate degree were the fewest in the study representing 1.97 percent of the total number of respondents in the sample. The table below shows the amount of earnings received by the respondents for a period of one year.
| Earnings | Frequency | Percent |
| Skip | 247 | 23.15 |
| 0 to $10,000 | 124 | 11.62 |
| $10,001 to $20,000 | 84 | 7.37 |
| $20,001 to $30,000 | 101 | 9.47 |
| $30,001 to $40,000 | 86 | 8.06 |
| $40,001 to $50,000 | 90 | 8.43 |
| $50,001 to $60,000 | 60 | 5.62 |
| $60,001 to $75,000 | 90 | 8.43 |
| $75,001 to 150,000 | 153 | 14.34 |
| $150,01 or more | 32 | 3 |
| Total | 1067 | 100 |
From the table above, a total of 247 respondents representing 23.15 percent of the total sample skipped the question on the range of their yearly earnings. The highest number of respondents earn between $75,0001 and $150,000 representing 14.34 of the sample. The second highest group in the sample in terms of wages earn between 0 to $10,000, and are represent 11.62 percent of the entire sample. Sixty individuals responded that they earn between $50,001 and $60,00 representing 5.62 percent of the entire sample. The lowest category had 32 individuals who indicated that they earn more than $150,000 representing three percent of the sample.
Correlation
The cross table below shows the relationship between the sex of the respondent and the amount of wages received.
| Sex | 0 to $10,000 | $10,001 to $20,000 | $20,001 to $30,000 | $30,001 to $40,000 | $40,001 to $50,000 | $50,001 to $60,000 | $60,001 to $75,000 | $75,001 to 150,000 | $150,01 or more |
| Females | 31 | 25 | 28 | 26 | 35 | 16 | 33 | 57 | 9 |
| Males | 93 | 59 | 73 | 60 | 55 | 44 | 57 | 96 | 23 |
| Total | 124 | 84 | 101 | 86 | 90 | 60 | 90 | 153 | 32 |
From the table above, there were more male respondents on the question on earnings than females. Men receive more salaries in all the categories of wages. A relationship between the level of education attained and the amount of earnings received by respondents in the sample is shown in the table below.
| Highest level of education | Earnings per Year | ||||||||
| 0 to $10,000 | $10,001 to $20,000 | $20,001 to $30,000 | $30,001 to $40,000 | $40,001 to $50,000 | $50,001 to $60,000 | $75,001 to 150,000 | $75,001 to 150,000 | $150,01 or more | |
| No High sch. | 8 | 6 | 5 | 8 | 1 | 2 | 2 | 1 | 0 |
| High sch. | 27 | 18 | 24 | 14 | 22 | 7 | 12 | 12 | 1 |
| Alternative high sch. | 5 | 6 | 2 | 3 | 3 | 1 | 1 | 3 | 0 |
| Less 1 yr college | 7 | 8 | 11 | 9 | 8 | 5 | 6 | 10 | 0 |
| 1 or more yrs college | 33 | 13 | 16 | 18 | 12 | 7 | 16 | 16 | 3 |
| Ass. Degree | 16 | 14 | 8 | 11 | 6 | 8 | 11 | 15 | 0 |
| Bachelor’s | 19 | 16 | 25 | 15 | 31 | 15 | 25 | 48 | 11 |
| Masters | 6 | 3 | 10 | 7 | 5 | 10 | 13 | 28 | 10 |
| Proff. Degree | 1 | 0 | 0 | 1 | 2 | 3 | 3 | 9 | 4 |
| Doctorate | 2 | 0 | 0 | 0 | 0 | 2 | 1 | 11 | 3 |
From the table above, most of the respondents in who fall in the category of 0 to $10000 have attained one or more years college credit as the highest level of school completed. Only one responded falls in the category and has attained a professional degree beyond a bachelor’s degree. Two individuals holding a doctorate degree fall also fall in the group. The highest number of respondents who earn between $10001 and $20000 have a high school diploma as their highest level of education completed. There are no individuals either having a professional degree or a doctorate degree who fall under the category. Most individuals who are under the category of respondents earning between $20001 and $30000 have a bachelor’s degree as the highest level of education completed. No individual in the sample holding a doctorate’s degree falls in the category. The highest number of the sampled study participants who fall under the category of $30001 and $40000 have a bachelor’s degree, whereas the lowest number of individuals in the category have a professional degree as the highest level of schooling achieved.
Eighteen individuals having one or more years of college credit earn between $40001 and $50000 are the highest in the category. On the other hand, no respondent who has a doctorate degree earns in the salary bracket. Two individuals holding a professional degree beyond a bachelor’s degree fall in the category. Similarly, the highest number of respondents falling under the category of earning between $500001 and $60000 have a bachelor’s degree as the highest level of education completed. The fewest number of individuals in the group have an alternative of a high school diploma. The highest number of respondents falling under the categories of earning $60001 and $75000, $75001 and $150000, and $150001 or more have attained a bachelor’s degree as the highest level of education. On the other hand, the fewest individuals in the three categories have attained, No high school diploma and alternative to high school diploma as their highest levels of education. The graph below shows the relationship between the two variables.
The graph shows that the higher an individual goes in education, the more likely one is able to receive higher wages. The table below shows the relationship between the field of study and earnings per year.
| Field of Study | Earnings per Year | ||||||||
| 0 to $10,000 | $10,001 to $20,000 | $20,001 to $30,000 | $30,001 to $40,000 | $40,001 to $50,000 | $50,001 to $60,000 | $75,001 to 150,000 | $75,001 to 150,000 | $150,01 or more | |
| General | 40 | 30 | 31 | 25 | 26 | 10 | 15 | 16 | 1 |
| Accounting | 12 | 8 | 11 | 6 | 8 | 7 | 6 | 11 | 0 |
| Administrative support | 0 | 1 | 1 | 5 | 5 | 2 | 7 | 8 | 3 |
| Agriculture | 0 | 2 | 2 | 3 | 2 | 1 | 1 | 1 | 0 |
| Broadcasting | 0 | 2 | 0 | 1 | 1 | 0 | 0 | 0 | 0 |
| Business admin | 1 | 3 | 1 | 1 | 2 | 1 | 1 | 1 | 0 |
| Communication | 8 | 7 | 10 | 3 | 3 | 9 | 9 | 22 | 9 |
| Computer science | 0 | 1 | 2 | 0 | 2 | 3 | 2 | 4 | 0 |
| Manufacturing | 4 | 4 | 4 | 3 | 4 | 3 | 5 | 10 | 1 |
| Cosmetology | 2 | 2 | 1 | 2 | 3 | 1 | 3 | 8 | 0 |
| Education | 1 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 |
| Engineering | 11 | 3 | 9 | 7 | 3 | 5 | 9 | 4 | 0 |
| English language | 1 | 1 | 3 | 5 | 7 | 3 | 1 | 8 | 5 |
| Fine arts | 1 | 2 | 1 | 4 | 1 | 0 | 2 | 2 | 0 |
| Healthcare | 4 | 1 | 0 | 3 | 2 | 2 | 4 | 2 | 0 |
| Law | 16 | 5 | 5 | 6 | 4 | 6 | 13 | 16 | 2 |
| Law enforcement | 3 | 0 | 1 | 1 | 2 | 1 | 2 | 6 | 1 |
| Liberal arts | 0 | 1 | 2 | 0 | 0 | 0 | 0 | 3 | 0 |
| Psychology | 2 | 2 | 2 | 2 | 3 | 1 | 1 | 3 | 1 |
| Theology | 2 | 2 | 3 | 0 | 0 | 0 | 0 | 2 | 1 |
| Mathematics | 3 | 0 | 2 | 0 | 0 | 1 | 0 | 0 | 1 |
| Social services | 6 | 5 | 2 | 3 | 3 | 2 | 5 | 12 | 4 |
| Social sciences | 2 | 1 | 1 | 1 | 2 | 0 | 1 | 5 | 0 |
| History | 2 | 1 | 2 | 1 | 5 | 2 | 2 | 5 | 3 |
This study’s findings does not show the comparison between those who reported to be from an unspecified college major (general) and other specified fields of study. From the table, the highest number of individuals from the sampled data having annual returns of between 0 to $10000 are from law followed closely by Engineering. The highest number of respondents falling in the category $10001 and $20000 have accounting as their field of study followed by those who studied communication. The category of annual returns ranging between $20001 and $30000 is highly represented by individuals who studied communication followed by those who studied engineering. The category of earnings between $30000 and $40000 has the highest number of individuals who took engineering as a major followed by those who studied accounting, and law respectively.
The highest number of individuals who earn between $40000 and $50000 studied accounting followed by those who took English as a language. The group with annual returns of between $50000 and $60000 has the highest number of respondents in the sample who studied communication, followed by those who studied accounting, law, and engineering respectively. The highest number of respondents earning between $60000 and $75000 studied law followed by those who studied engineering and then engineering. The majority of the respondents in the category earning between $75000 and $150000 are those who studied communication followed by those who took law and accounting respectively. Finally most individuals earning above $150000 in the sampled dataset studied communication. Below is a bar graph showing the relationship between the field of study variable and the earnings received.
The bar chart shows the distribution of the categories of wages against the educational field of study of individuals in the sampled data. From the graph, most individuals earning between 0 to $750000 fall in the categories of between 5 and 10 in the field of education variable.
Regression
A multinomial logistic regression analysis gave out a log likelihood of -2216.0942. The chi square likelihood ratio was 215.13 with a p-value<0.00001, implying that the model is statistically significant in fitting the data. The 3rd, and 7th, categories of annual earnings showed a statistical significance in their individual models. Below is a table showing the regression coefficients in the 3rd group.
| Earnings | Coefficient | Std. Err | Z | Pr>|Z| |
| Sex | .1471899 | .3469593 | 0.56 | 0.0575 |
| Eduattn | .0443281 | .0741923 | 0.60 | 0.0550 |
| Edufos | .0050481 | .021161 | 0.24 | 0.0811 |
| cnmain | -.9221545 | .3590793 | -2.57 | 0.010 |
| constant | .3469593 | .9141382 | 0.38 | 0.0704 |
The p-values are less than the significance value, implying that the model is statistically significant. The model to predict that an individual falls in the category of those earning between $20001 and $30000 is given below.
The table below shows the regression coefficients of predicting that an individual will fall under the category of earning between $50000 and $75000 in a year.
| Earnings | Coefficient | Std. Err | Z | Pr>|Z| |
| Sex | -.291434 | .2653201 | -1.10 | 0.272 |
| Eduattn | .2353459 | .0769414 | 3.06 | 0.002 |
| Edufos | -.0226386 | .0221483 | -1.02 | 0.307 |
| cnmain | -1.741823 | .3357005 | -5.19 | 0.000 |
| constant | 1.598566 | .8907978 | 1.79 | 0.073 |
The p-values are less than the significance value, implying that the model is statistically significant. The model to predict that an individual falls in the category of those earning between $50001 and $75000 is given below.
CONCLUSIONS
The purpose of the present study was to examine the variation of salaries according to college majors. It sought to prove the traditional theory that engineering majors have high returns than other majors of education, such as education, arts, and humanities. This study went further to estimate the influence of the experience of individuals in different fields of study on the wages received by comparing the starting median salaries and mid-career wages. This study used data that has been pre-processed for the analysis. From an observational analysis, it can be concluded that individuals in fields such as engineering, accounting, and law are likely to earn higher levels of wages than people in other fields of study. This study’s results concur with the evidence provided by research studies by Van de Werfhost (2002), Altonji et al. (2012), and Ransom (2014) that different fields of study attract different pays.
Recent times have seen an increase in students enrolling in higher learning institutions. There is a need to have studies to enable people to make the right choices on which courses to study in relation to the number of returns received. Several scholars prove that there is a strong relationship between the number of payoffs and the college major studied. This study sought to prove the social theory that engineering majors earn more than other fields of study, such as education, humanities, arts, and sciences. The findings of this study indicate that indeed individuals who study engineering majors are likely to earn more than those in other fields. The call by governments to have reforms in the education sector to have more entrants in the STEM courses is, therefore, supported by this study.
Works Cited
Altonji, J. G., Blom, E., and Meghir, C. (2012). Heterogeneity in human capital investments: High school curriculum, college major, and careers. Annual Review of Economics, 4(1):185–223.
Altonji JG, Pierret CR. 2001. Employer learning and statistical discrimination. Q. J. Econ. 116(1):313-350.
Arcidiacono, P. (2004). Ability sorting and the returns to college major. Journal of Econometrics, 121(1–2), 343–375.
Arcidiacono, P., Hotz, V. J., & Kang, S. (2012). Modeling major college choices using elicited expectations and counterfactuals. Journal of Econometrics, 166(1), 3–16.
Blau, Francine D., and Lawrence M. Kahn. 2017. “The gender wage gap: Extent, trends, and explanations.” Journal of Economic Literature 55: 789-865.
Berger, M. C. (1988). Predicted future earnings and choice of college major. Industrial and Labor Relations Review, 41(3), 418–429.
Böhlmark, A., & Lindquist, M. J. (2006). Life-cycle variations in the association between current and lifetime income: Replication and extension for Sweden. Journal of Labor Economics, 24(4), 879–896.
Carnevale, A. P., Cheah, B., & Hanson, A. R. (2015). The economic value of college majors. Washington: Georgetown University Center on Education and the Workforce.
Freeman, J. A., and B. T. Hirsch (2008). “College Majors and the Knowledge Content of Jobs.” Economics of Education Review 27(5): 517-535.
Haider, S., & Solon, G. (2006). Life-cycle variation in the association between current and lifetime earnings. American Economic Review, 96(4), 1308–1320.
Kan Yao (2019) Heterogeneous skill distribution and college major: evidence from PIAAC, Journal of Applied Economics, 22:1, 504-526.
Kirkeboen, L. J., Leuven, E., & Mogstad, M. (2016). Field of study, earnings, and self-selection*. The Quarterly Journal of Economics, 131(3), 1057–1111.
Robst, J. (2007). Education and job match: the relatedness of college majors and work. Econ. Educ. Rev.26(4):397407.
Van de Werfhorst HG. 2002. “Fields of Study, Acquired Skills and the Wage Benefit from a Matching Job.” Acta Sociologica 45: 287-303.
Zafar, B. (2013). College major choice and the gender gap. The Journal of Human Resources, 48 (3), 545–595.