Wage Determination Model

WAGE DETERMINATION MODEL
Name
Due Date
1.0 Introduction
The human capital theory is an economic framework that emphasizes the importance of investing in education, training, and other forms of human capital as a means of increasing productivity and economic growth. According to this theory, individuals and societies can improve their economic outcomes by investing in the development of human capital, which refers to the knowledge, skills, and abilities that people acquire through education, training, and experience. Human capital is seen as a key determinant of wages, as individuals with higher levels of human capital tend to be more productive and earn higher wages. In addition to individual benefits, investing in human capital is also thought to have positive spillover effects for society, as it can lead to innovation, higher economic growth, and improved living standards. The human capital theory has been influential in shaping policies related to education and training, as well as labor market policies aimed at reducing inequality and improving economic outcomes for individuals and societies. In this report, a wage determination model based on human capital theory is developed and evaluated.
2.0 Wage Determination Model
The proposed wage determination model will consider five key variables that play a role in shaping access to and development of human capital. First, the individual age will be considered an essential part of the model. Age is an important factor in human capital theory as it represents the accumulation of experience and knowledge over time, which can increase productivity and wages. Research suggests that as individuals get older, their wages tend to increase due to increased experience and skills (Strober 1990). Second, the model will include the education level of an individual. Education is a significant factor in human capital theory as it increases an individual's skills, knowledge, and productivity. Research has consistently shown that education is positively correlated with wages (Nafukho, Hairston, and Brooks 2004). Third, the model will consider an individual’s disability status as it has been shown to play a role in determining access to earning opportunities. It is hypothesized that disabled workers may experience wage discrimination in the labor market (Strober 1990). Thus, this variable could be included to capture any wage penalty associated with disability status. Fourth, the model will consider the ethnicity of the participant. Like disability, Ethnicity affects wages through discrimination and differences in human capital. Research suggests that ethnic minorities face wage discrimination in the labor market, which can lead to lower wages (Leping and Toomet 2008). In the same vein, Sex is an important factor in human capital theory as it can affect labor market outcomes through differences in education, experience, and discrimination. Research suggests that women earn less than men, even after controlling for human capital factors such as education and experience (Blau & Kahn, 2017).
The model would hypothesize that the individual hourly pay is likely to be determined by age, education level, disability status, ethnicity, and gender. The influence of each variable within the model is hypothesized as indicated in the table below.
Variable

Hypothesis

Age

Positive coefficient - as individuals get older, their wages tend to increase due to increased experience and skills.

Education

Positive coefficient - as education increases, wages tend to increase due to increased productivity and skills.

Disability

Negative coefficient - if disabled workers experience wage discrimination, then the variable could capture this effect and result in lower wages.

Ethnicity

Negative coefficient - if ethnic minorities face wage discrimination, then the variable could capture this effect and result in lower wages.

Sex

Negative coefficient - if women face wage discrimination, then the variable could capture this effect and result in lower wages.



3.0 Data Preparation
To test the wage model using the provided model, it will be essential to recode the categorical variables in the data. The dependent variable in the model will be hourly wage. To this end, the variable will not be recoded as it is already a continuous variable. Among the independent variables within the model, only age will not be recoded as it is already reported as a continuous variable. Education will be recoded as it is currently reported as an ordinal variable with values ranging from 1 to 5. To simplify the analysis, the variable will be recoded into two dummy variables; 1 (degree or higher”) and 0 (others). The disability variable will also be recoded into a binary variable, where 1 represents a respondent without a disability and 0 represents a respondent with a disability. In the original data, the ethnicity of the respondent is coded as an ordinal variable ranging from 1 (White) to 5 (Other). To simplify the analysis, this variable will be recoded into a binary variable, where 1 represents a respondent who is White, and 0 represents a respondent who is not White. Sex is coded as a nominal variable, with 1 representing the male and 2 representing the female. The variable was recoded as a binary variable, where 1 represents a male respondent and 0 represents a female respondent. By recoding the relevant variables into more meaningful forms, the analysis will be simplified and make it easier to interpret the estimated coefficients.
4.0 Model Discussion
The model is a multiple linear regression model used to predict the hourly wage of individuals based on their demographic and personal characteristics. The dependent variable is the gross hourly pay of the respondents. The independent variables include age, whether the respondent is disabled, ethnicity, gender, and whether they have a degree. With regards to explanatory power, the R-squared value of 0.198 means that approximately 20% of the variation in hourly wage is explained by the independent variables in the model. This may not be a very strong explanatory power, but it is still useful in predicting hourly wages based on the available information. The F-statistic of 365.443 with a p-value of 0.000 in the ANOVA table indicates that the overall model is significant, meaning that the independent variables are jointly useful in predicting hourly wage. Therefore, we can conclude that the model has some degree of explanatory power, although there may be other factors not included in the model that can also affect hourly wage.
Concerning model specification adequacy, the model meets the assumptions of linear regression. The Durbin-Watson value of 0.446 suggests there may be some positive autocorrelation in the residuals. An analysis of the Mahalanobis distance returned a value of 1.844. The Mahalanobis distance of the residuals is a measure of how well the model's assumptions fit the data. Specifically, it measures the distance between each data point and the regression line, considering the covariance structure of the variables in the model. Therefore, a small Mahalanobis distance indicates that the model fits the data well and that the residuals are normally distributed and have constant variance. The coefficients table shows the estimated regression coefficients for each independent variable and indicates that age, disability, race, gender, and having a degree are all statistically significant predictors of hourly wage at the 0.05 significance level. The resulting model can be presented as follows:
Gross Hourly Pay = 3.320 + 0.151(Age of Respondent) - 1.934(Does not qualify for disability) + 1.544(What is your race?) + 3.354(What is your gender) + 8.094(Do you have a degree)
Where:
(Age of Respondent is the age of the employee in years.
Does not qualify for disability is a binary variable, where 1 indicates the employee does not qualify for disability and 0 indicates they do.
What is your race? is a binary variable, where 1 indicates the employee is White and 0 indicates they are of another ethnicity?
What is your gender is a binary variable, where 1 indicates the employee is male and 0 indicates they are female.
Do you have a degree is a binary variable, where 1 indicates the employee has a degree and 0 indicates they do not.)
The coefficients represent the impact of each variable on the gross hourly pay.
The coefficient for age in the regression model is 0.151, and it is positive, indicating that older individuals tend to have higher hourly wages. This coefficient is statistically significant with a p-value less than 0.001. This result is consistent with past research on the relationship between age and wages, which suggests that there is a positive correlation between age and wages, up to a certain point. This is consistent with the human capital theory, which suggests that individuals acquire valuable skills and knowledge as they gain experience in the labor market, which increases their productivity and wage-earning potential. However, the positive effect of age on wages may plateau or even reverse after a certain age, due to factors such as declining health or retirement. The positive coefficient for age in this model suggests that age is a significant predictor of hourly wages, even after controlling for other factors such as gender, ethnicity, disability, and education.
The coefficient for education (coded as having a degree or not) in the regression model is 8.094, and it is positive, indicating that individuals with a degree tend to have higher hourly wages compared to those without a degree. This coefficient is statistically significant with a p-value less than 0.001. This result is consistent with past research on the relationship between education and wages, which suggests that higher levels of education lead to higher wages, on average (Miki and Yuval 2011). This is also consistent with the human capital theory, which posits that education is a form of human capital that enhances an individual's productivity and earning potential in the labor market. The positive coefficient for education in this model suggests that education is a significant predictor of hourly wage, even after controlling for other factors such as age, gender, ethnicity, and disability status. Holding all other variables constant, the model predicts that individuals with a degree earn approximately $8.09 more per hour than those without a degree. However, it's important to note that this is just a prediction based on t...