What a wage buys now compared with a decade ago

This works through real wages in the order the parts actually depend on each other.

Nominal and real are different questions

A pay rise below the inflation rate is a reduction in purchasing power, however positive the number looks. Comparing pay across years without adjusting for prices produces conclusions that are simply wrong.

Statistics offices publish deflators that make this calculation straightforward. Run the comparison on net pay rather than gross, because tax thresholds that do not move with prices quietly increase the share taken as nominal wages rise.

The basket is an average of nobody

Headline inflation weights a representative basket, and your own weighting depends on how you actually spend. Households spending a large share on housing, energy and food have experienced higher effective inflation in recent periods. Several statistics agencies now publish inflation by income decile precisely because the average conceals this.

In practice, how housing enters the index differs between countries and between measures, and whether rent, mortgage interest or an imputed cost is used can move the headline figure while nothing has changed in any household.

Essentials are less substitutable

Discretionary spending can be traded down; rent, energy and transport to work largely cannot. That makes price rises in essentials bite harder than the headline figure suggests for lower-income households.

It is also why two people with the same pay rise can have very different experiences of it. The same goods often cost more in the sizes an irregular income allows you to buy, so the unit price paid rises as the amount available at any one time falls.

Wage growth is unevenly distributed

Aggregate wage growth figures blend sectors, regions and seniority levels that have moved very differently. Sector-level data is more informative for an individual than a national average. It is also the more useful benchmark when negotiating.

Averages are pulled by the ends of a distribution, so a median describes a typical worker better than a mean does, and which one a headline has used is worth checking before repeating it.

What to do with the number

Knowing your own real-terms position tells you whether to negotiate, change employer or change sector. It also converts a vague sense of things being harder into a figure that can be argued with.

Put simply, that is more useful than the feeling on its own. The number does not always point anywhere: where an entire sector or region has lagged, no individual has a lever that fixes it, and reading that as a personal failure misdescribes what the data shows.

Adjust the size of it until it is something you would actually do tired.

Comparing across years without misleading yourself

Use the same month in each year rather than comparing a January with a June, because seasonal price movement can swamp the trend you are trying to see. Compare like roles as well, since a promotion between the two dates makes the figures describe two different jobs rather than one wage changing. Where a currency has moved sharply or a country has been through a period of very high inflation, indices are revised and rebased, and the series you started with may not be the one you finish with.

In practice, national statistics offices publish their methods openly, and reading how a measure is constructed is the difference between using a figure and repeating one.

The takeaway

Convert your pay history to real terms once. It usually explains the feeling.

Small and repeatable beats ambitious and abandoned, almost every time.