Question: In January 2016 Starbucks introduced incentive pay into its California stores, so managers get \(2 \%\) of revenue. In Oregon and Washington, stores pay managers a fixed wage.

You initially have cross-sectional data on annual revenue for 2016 for all California, Oregon and Washington stores. You observe that the mean revenue in California stores is significantly larger than Oregon and Washington stores.

1. Can we conclude that incentive pay raises revenue? Explain how location can cause omitted variable bias. If you know the location of every store, what analysis can you do to reduce this bias? Explain why this analysis may work, and why it may not.
2. Besides location, what other store characteristics may interfere with the comparison. What data would you like to help you to make this conclusion?
3. Suppose you also have annual revenue for 2015 . What regression would you run to test whether the new pay scheme raised revenue?
   Suppose we have concluded that incentive pay raises revenue (as best as we can).
4. Why should we be cautious in concluding that Starbucks should roll out the scheme across the US?
5. Can we conclude that incentive pay makes individual managers work harder? What extra data would you need to test this? What analysis would you perform?

Price: $2.99

Solution: The downloadable solution consists of 2 pages
Deliverable: Word Document