What You Are Actually Breathing

Air pollution kills more people than any other environmental exposure, and the dominant contributor is not a gas. It is particulate matter: solid and liquid particles small enough to stay suspended, measured in micrograms per cubic metre, with the crucial threshold at two and a half micrometres of diameter.

The threshold is set by deposition physics. Larger particles are trapped in the nose and upper airway and cleared by mucus. Below two and a half micrometres, particles reach the alveoli and pass into the bloodstream. That is why the strongest associations are with cardiovascular disease rather than with respiratory illness alone, which surprises people who assume that breathing dirty air is primarily a lung problem.

There is no established safe threshold. Studies find associations with mortality continuing down to the lowest concentrations observed, including levels well below current guideline values in relatively clean countries. The World Health Organization has lowered its guideline repeatedly as evidence accumulated at lower exposures. Regulatory limits are practical compromises between health and cost, not statements about where the harm stops, and reading them as the latter is a common and consequential error.

Ozone illustrates how much location matters. The molecule is identical in the stratosphere and at street level. High up it absorbs ultraviolet radiation that would otherwise damage living tissue, which is why its depletion was an emergency. At ground level, formed when sunlight acts on traffic and industrial emissions, it is a powerful irritant that inflames airways and damages crops. Good up there, bad down here, same substance.

The clearest demonstration that any of this is worth acting on is lead. It is a potent neurotoxin with no safe exposure level, and it was deliberately added to petrol for decades after the hazard was documented, against the advice of the scientists who documented it. Blood lead levels fell sharply everywhere the phase-out happened. Measured population cognitive scores rose, and the timing of subsequent declines in violent crime across many countries matches the exposure of affected birth cohorts reaching adulthood. It is among the largest public health returns ever recorded from a single regulatory decision.

The exposure that receives least attention relative to its burden is indoor. Cooking and heating with wood, dung or coal in an unventilated room produces concentrations far above almost any outdoor urban level, and people spend the large majority of their lives indoors. It falls disproportionately on women and young children in low-income households. Clean cooking is one of the highest-return health interventions available and attracts a fraction of the attention that vehicle emissions do.

A closing caution about the most-cited success story in atmospheric policy. The Montreal Protocol worked, and it worked because a small number of firms made a small number of compounds for a small number of uses, and substitutes existed at modest cost. Carbon dioxide comes from essentially every economic activity in every country, with no drop-in substitute for the activity itself. Reading the ozone agreement as proof that international cooperation solves atmospheric problems has misled a great deal of argument about a problem that shares almost none of its structure.