Chandra, Gruber, and McKinight (2010) studies the effect of copayments on health care use. The study setting has two types of health insurance plans: (1) Health Maintenance Organizations (HMOs) and (2) Preferred Provider Organizations (PPOs). The HMOs experienced a rise in physician office visit copayments from $0 to $10 in 2002. But there was no corresponding change for PPOs in 2001-2003. In 2001, the average number of physician visits per person per month is 1.07 and 0.75 for PPOs and HMOs, respectively; in 2003, the average number of physician visits per person per month is 1.19 and 0.75 for PPOs and HMOs, respectively.

a. Suppose that we want to study the effect of the copayment increase on physician visits by a differences-in-differences strategy. What are the control group, the treatment group, the pre-treatment period, and the post-treatment period in this context?

b. Write down the regression equation for the differences-in-differences estimation. Clearly define each variable.

c. Based on the numbers provided above, what is the effect of a $10 increase in physician visit copayments on patient use of physician office visits? Please include the following in your answer: (1) does the copayment rise increase or decrease physician visits? (2) What is the magnitude of the increase/decrease?