Data. This study used the 2012 National Longitudinal Survey of Youth 1979 (NLSY79). The cohort of NLSY79 initially contained a sample composed of 6,403 males and 6,283 females, totaling 12,686 respondents. The respondents were 14 to 22 years old at the time of the first survey. Currently, 9,964 respondents remain in that sample. In respect to race, the initial survey contained 7,510 respondents who were non-black/non-Hispanic, 3,174 black, and 2,002 Hispanic or Latino. Respondents were interviewed annually until 1994 and biennially thereafter. After adjusting for invalid responses, valid skips, people who answered don’t know, refused to answer, and were not interviewed, the number of observations decreased to 7,301 respondents.
Empirical models. To determine the use of education saving vehicles, the following question was explored: Does an outstanding student loan balance of the parent affect utilizing saving vehicles such as Coverdell ESAs and 529 plans in order to save for their child(ren)’s education?
A binary logistic regression was used to obtain the likelihood of a parent using education saving vehicles to finance their child’s education based on their student loan balance
To determine whether parental student loan debt affected their views on loans for their children, the following question was explored: Does an outstanding student loan balance of the parent affect taking out loans on behalf of the child(ren) for higher education?
A second binary logistic regression was used to assess the likelihood of a parent taking out loans on behalf of their child based on their own student loan balance.
Dependent variables. Education saving vehicles. Education saving vehicles is a dichotomous variable that assesses two outcomes: whether the parents are using education saving vehicles or not. If either the Coverdell or 529 plan were used, then it would indicate that a parent using education saving vehicles is true, otherwise, it is false.
If the loan balance for child variable equals greater than zero, then the loan for child variable takes a value of 1; otherwise it gets the value of 0
Hypothesis variable. Parent’s student debt is the hypothesis variable for this research. It is made dichotomous by making the numerical value equal to 1 when the parent’s student debt balance is greater than 0; otherwise, the numerical value is equal to 0.
Control variables. Sex. The sex of a parent may influence their views toward financing the education of their child. The human capital theory would predict that women are more likely to invest in their child’s education since they have, on average, longer life expectancies than men and would be able to gain more in the long run from the investment. To investigate if sex affects the dependent variables, the dummy variable male is included in the model (the reference variable is female).
Race. Past research has concluded that race is significant in determining degree completion, student loan defaults, and the probability of borrowing for school. Past research indicates that this variable is correlated to educational financial decisions. It is divided into two dummy variables, black and Hispanic, and one reference variable, non-Hispanic, non-white.
College degree. Based on the status attainment theory, there is a positive correlation between college attainment of the parent and the parent’s aspirations for their child to go to college, thus the parent is more likely to provide financial support for college. This variable is a dichotomous variable that assesses two outcomes, whether the recipient has a college degree or not. If the recipient has a bachelor’s degree or a higher degree, then the college degree variable takes numerical value 1; otherwise, it takes value 0.