Inflation Rate

The percentage rate at which the general level of prices for goods and services rises over a given period.

Source: Mankiw 2016

The general price level is tracked by an index built from prices of a defined basket of goods and services, and the inflation rate is the proportional change in that index between two periods, usually expressed at an annual rate. Different indices give different answers because they cover different baskets: a consumer index covers household purchases, a producer index covers goods leaving factories, and a deflator derived from national accounts covers everything produced domestically. None is the true rate, since there is no single price level.

Source: Mankiw 2016

A nominal amount is expressed in the money of the day, and a real amount is expressed in the purchasing power of a chosen base period. Converting between them requires an index and a stated base year. The distinction matters because a nominal increase in spending can coexist with a real decline, and a series presented in nominal terms will show growth even where nothing changed but prices. Any statement about spending over time is uninterpretable without knowing which basis it uses.

Source: Mankiw 2016

Discounting adjusts for the timing of costs and benefits, and inflation adjusts for changes in the price level, and the two are separate operations frequently confused. Standard practice expresses all flows in constant prices of a base year and applies a real discount rate, which excludes general inflation entirely. Applying a nominal rate to real values, or a real rate to nominal ones, produces an error that compounds over the horizon and can be large in a long analysis.

Source: Briggs, Claxton & Sculpher 2006

General inflation moves all prices together, while relative price change describes one price moving differently from the rest, as health sector pay and medicines have done over long periods. Constant-price analysis removes the first and not the second, so an expectation that a particular input will become more expensive relative to others must be modelled explicitly. Doing so requires evidence, since an assumed divergence extended over decades can drive a result more strongly than any clinical parameter.

Source: Drummond et al. 2015

A fixed basket cannot capture consumers shifting towards goods whose prices rose less, which overstates the increase in the cost of living. Improvements in the quality of goods are difficult to separate from price rises, which overstates it further. New goods enter the basket after their prices have already fallen. In the other direction, indices may miss price increases concentrated in items households cannot avoid. The direction and size of the net error are contested and vary between countries and periods.

Source: Mankiw 2016