You experiment by offering free warranties for your product in market A but not in market B. Sales in A rise from 240 to 360 units per week while sales in B rise from 410 to 430. The Difference-in-difference estimate of the effect of the free warranty is:

Difference in difference estimate = 50 - 5% = 45 %

Explanation:

a) Data and Calculations:

                           Market A          Market B

Sales                       240                   410

Sales rise                360                   430

Rise difference       120                     20

Percentage of rise  50%                    5%

120/240 x 100 = 50%

20/41 x 100 = 4.878% or 5%

Therefore, the Difference in difference estimate = 50 - 5% = 45 %

One can then say that the free warranties in market A brought about a difference in difference of 45% in Market A when compared to the no warranties in Market B.  This can be seen from the presented data.  Sales in A rose from 240 units to 360 units, an increase of 120 units or 50%.  Sales in market B only rose from 410 to 430, an increase of 20 units or 5%.  This difference in difference estimator shows the effect of the free warranty on market A and market B.  This means that the firm could do better by introducing the free warranties for its product in market B, all things being equal.