Yesterday Neutral Source explained why inflation forecasting is like risk assessment. In a very important way, of course, they are very different.
The Bureau of Labor Statistics is the federal agency responsible for estimating inflation. They collect vast amounts of data and devote extraordinary resources to complex modeling. The BLS website has a treasure trove of information about how it’s done. Read the FAQ sheet on the CPI to learn more.
As a matter of policy, BLS tries to estimate inflation objectively — that is, without under- or overstating it. They develop different indices for different purposes, and these estimates can differ by quite a lot. But they do not base their estimates on “reasonable” upper bounds. In contrast, much risk assessment is designed to yield so-called precautionary estimates of risk — that is, estimates that are almost certainly greater than the true (but unknown) level of risk.
Some might say that risk estimates are more important than inflation estimates because risk estimates provide the basis for health protective regulation. But the CPI also affects people just as surely. For many people, their wages and benefits are tied to the CPI. For that reason, it is essential that the CPI be free from policy-driven bias. The same case can be made for health risk assessment.