Consider the regression. log(rate) it = a + ß × diversity it + 7 × education;t + d × experience it + Eit a)Suggest a confounding factor that is different across firms but constant over time that makes β not a causal effect of diversity on productivity of workers b) Suggest a confounding factor that is the same across firms but change over time that makes β in (a) not a causal effect of diversity on productivity of workers. Variables in the dataset are: • firmid: firm id. Data from the same firm id is from the same firm • year: year that data is collected. Data are collected from 2001 to 2007 every two years. • rate: average hourly wage rate of workers in a firm, as a proxy for productivity - diversity: proportion of workers who are female or within small ethnic groups (less than 2.5% of Australian population) in a firm • education: average years of education of workers in a firm • experience: average years of experience of workers in a firm