Offering a Higher Credit Card Limit: Quantifying Noncompliance Concerns

CreditCo, a large credit card company, decides to run an experiment. It sends an offer in the mail to a random group of 50% of its customers: those in the treatment group are invited to navigate to a webpage and opt in for a 10% higher credit limit. CreditCo wants to see how credit balances and late payments are impacted six months later as a result of the experiment.

Suppose that, of the group that received the mail offer, 40% of people opted in. Do you think that noncompliance will be a problem for CreditCo's analysis? Why or why not?

Possible answers

No, because people in the treatment group likely used a coin flip to make their opt-in decision.

Yes, because the opt-in rate is low.

No, because an opt-in rate of 40% is actually quite high.

Yes, because the set of people opting in are probably people with worse spending habits.

Exercise

Offering a Higher Credit Card Limit: Quantifying Noncompliance Concerns

Instructions

Possible answers