We are interested in the causal effect of having a university degree on earnings. Data from...

 

Transcribed Image Text:

We are interested in the causal effect of having a university degree on earnings. Data from a 2012 survey of n = 7500 full-time workers in the US, aged 25-34, was used to estimate the following regression: where wagei (and age (in years). = wage; 1.87 +8.32degree, - 3.81female; +0.51age; + Ûi is the hourly wage rate (in 2012 dollars), degree, = 1 if worker i has a degree 1 if worker i is female (and 0 otherwise), and age, is = = 0 otherwise), female, (a) Why do you think age was included in this regression? Give the descriptive interpret- ation of its estimated coefficient. Do you think it can be given a causal interpretation, i.e. do you think it consistently estimates a causal effect? (b) Discuss the estimated coefficient on female. (c) Do you think that the estimated coefficient on degree gives a reliable estimate of the causal effect of having a university degree on earnings? If so, explain why; if not, We are interested in the causal effect of having a university degree on earnings. Data from a 2012 survey of n = 7500 full-time workers in the US, aged 25-34, was used to estimate the following regression: where wagei (and age (in years). = wage; 1.87 +8.32degree, - 3.81female; +0.51age; + Ûi is the hourly wage rate (in 2012 dollars), degree, = 1 if worker i has a degree 1 if worker i is female (and 0 otherwise), and age, is = = 0 otherwise), female, (a) Why do you think age was included in this regression? Give the descriptive interpret- ation of its estimated coefficient. Do you think it can be given a causal interpretation, i.e. do you think it consistently estimates a causal effect? (b) Discuss the estimated coefficient on female. (c) Do you think that the estimated coefficient on degree gives a reliable estimate of the causal effect of having a university degree on earnings? If so, explain why; if not,

Answer rating: 100% (QA)

SOLUTION a Age is included in the regression because it is a potential confounding variable that could affect both the likelihood of having a universi... View the full answer