Describing it as a simple loss of $3 million is definitely a simplification, for the reason you mention. But saying this is "very misleading" is perhaps even mo

Describing it as a simple loss of $3 million is definitely a simplification, for the reason you mention. But saying this is "very misleading" is perhaps even more of an oversimplification. Unless you are so cynical that you think the normal tax rates are a complete sham -- i.e. that the official tax rates have no basis in reality, such that there is no impact whether a city charges $5M or $0. But if we assume that that $5M is somewhat tied to the actual costs to the city to administer services and regulation, then offering a $3M tax break is done with understanding the potential opportunity cost, and believing that Amazon's business will more than make up for that $3M that the city might have reaped otherwise. The point of these stories is that, in the past, there was basically no easy way for the public to evaluate these tax breaks. A situation like Amazon seems like a sure net win. But there are many, many more American jurisdictions than the city that happens to get Amazon's business.