*Even low levels of air pollution can damage health , study finds.*

Don't panic, though — a tiny concentration of air pollution will normally do a tiny amount of harm, which is hard to measure without a sample of millions.

In the absence of any standard of perfection (keep air pollution under level L and it will do exactly zero damage), how should we decide how much to work to reduce air pollution? It's a matter of how the danger from air pollution stacks up against all the other dangers we face. Suppose we decide to spend a billion dollars on public health, which allocation of funds would we rationally expect to make the biggest improvement? Part of that plan could be spending to reduce air pollution.