FILE In a sem inal study, researchers docum ented race-based hiring in th e Boston and Chicago labor m arkets (American Economic Review, Septem ber 2004). They sen t o ut identical resum es to em ployers, half w ith traditionally African-American nam es and th e o th er half w ith traditionally Caucasian nam es.

Interestingly, th ere w as a 53% difference in call-back rates betw een th e tw o groups of people. A research fellow at an institute in Santa Barbara decides to repeat the sam e ex p e rim e n t w ith n a m e s along w ith a g e in th e Los A ngeles labor m arket. She rep eated ly se n d s o u t resu m es for sales positions in th e city th a t are identical e x cep t for th e difference in th e n am es a n d a g e s o f th e applicants. She also records th e call-back rate for each can d id ate. The acco m p an y in g ta b le show s a portion o f d a ta on call-back rate (%), ag

e, a n d a C aucasian d u m m y th a t e q u als 1 for a C a ucasian-sounding n am e; th e co m p lete d a ta set can b e fo u n d o n th e te x t w eb site, labeled Hiring.

Call-back Age Caucasian 12 60 1 9 56 0

⋮ ⋮ ⋮

15 38 0

a. Estim ate a linear regression m odel w ith call-back as th e response variable, and ag e and th e Caucasian dum m y as th e explanatory variables.

b. C om pute th e call-back rate for a 30-year-old applicant with a C aucasian-sounding nam

e. W hat is th e corresponding call-back rate for a non-Caucasian?

c. C onduct a test for race discrim ination at th e 5%

significance level.