The next two questions refer to the following information.   A textile company has installed air conditioners in some of its buildings, but not others. It is considering installing air conditions in the remaining facilities, and wants to see how air conditioning affects output. You therefore conduct a difference-in-differences analysis, using data from facilities that do and do not have air conditions, before and after air conditioning installation. The plot below shows output per facility over time, for each group.   

Do the data appear consistent with the assumptions necessary for difference-in-differences?

Did you turn on sharing for your notebook, so that I can view it if you give me the link? 

The graphs below shows data from the company.  Roughly what is the effect of an additional $1,000 of salary offer on the probability of job acceptance?

Setting aside any possible concern from the prior question, what would you say is the difference-in-differences estimate of the effect of air conditioning on output?

The graphs below shows data from the company.  Roughly what is the effect of an additional $1,000 of salary offer on the probability of job acceptance?

Setting aside any possible concern from the prior question, what would you say is the difference-in-differences estimate of the effect of air conditioning on output?

For many years, a company had a very generous health insurance plan, with no deductible. In 2023, the company introduced a $3000 deductible for all employees, and the company has no information about non-employees. The company wants to know how the deductible affected health care spending. What identification strategy would be most appropriate?

What is the key assumption for difference-in-differences to be valid?

Here are data from the experiment:   Group # People Average cost per windshield replaced Assigned to hourly wage 1000 82 Assigned to treatment, chose hourly wage 500 90 Assigned to treatment, chose piece rate contract 500 54       What is the effect of the piece rate contract, relative to fixed hourly wage, on the cost per windshield replaced?