A pharmaceutical company wants to test whether a new health campaign reduces instances of the flu in a household exactly one month after the campaign. To test this, they send out mailers to every other household on a single city block. Exactly one month later, which is coincidentally right after the neighborhood barbecue, they test each household for the flu. The researchers find that neither the household that received the mailers nor the households that did not receive them contracted the flu. They conclude that the health campaign is ineffective since the difference in flu instances between the groups is zero.
Law of Large Numbers
Constant Effect
Treatment Time
No Spillovers