Suppose that managers of the manufacturing company run a field experiment to test a potential way to improve productivity in the shipping department. Managers believed that employees in the shipping department would work more quickly and effectively if they were allowed to listen to music using their own Apple AirPods.

To this end, they designed an experiment for 2020. At one factory (the "control" factory), employees in the shipping department were not allowed to listen to music at all during the year. At another factory (the "treatment" factory), employees in the shipping department were not allowed to listen to music for the first six months of the year (through June 30), but from July 1 until December 31 were allowed to listen to music on their own AirPods. For all of 2020, managers measured the percentage of orders shipped on time by each shipping department. The data follow:

SEMIANNUAL AVERAGES

1ST HALF 2nd HALF

TREATMENT 84% 87%

CONTROL 82% 78%

MONTHLY AVERAGES

J F M A M J J A S O N D

TREATMENT 83% 84% 84% 84% 85% 86% 86% 86% 87% 88% 87% 88%

CONTROL 83% 83% 82% 82% 81% 80% 79% 80% 78% 79% 77% 77%

1. Calculate the difference-in-differences "DiD" estimate for the effect of the treatment. Interpret the estimate.

2. Is there any evidence to suggest the DiD estimate may be a flawed estimator in this case? Explain what evidence you would consider to make this determination and why it does or does not suggest the DiD estimate may be flawed.