Did your raise beat inflation?
A raise that only kept up with prices is not ground gained. This is how a reading tells the two apart, using your country's own price index.
Three steps, all arithmetic
Every step runs in code over published data. No model writes a figure.
1. Take prices out
Your income a year ago and your income now are each divided by your country's consumer price index for their year, so both are in the same year's money.
2. Measure your change
Your real change is the ratio of the two, less one. A raise smaller than the rise in prices comes out below zero, however large it looked on a payslip.
3. Set it against the middle
The same is done for your country's median income over the same year. Your ground is your real change less the median's, so a year in which everyone's money bought less is not counted as ground you lost alone.
Why the answer is a band
Income is asked as a band, never as an exact figure.
- A band stands for one fixed income: the middle of its range, or at the open top and bottom the average the source publishes for that segment. The same band always stands for the same income.
- A raise that keeps you inside one band does not register as movement: the method would rather say nothing than invent precision.
- Within 2 percent either side of no change, your ground counts as held.
Whose prices
The price index is your country's own national consumer price series: the whole national basket, not one city's.
In 21 of the 149 countries a reading covers, part of that series had to be bridged rather than read from the national figures, and every reading there carries low confidence for it:
What it does not do
- It uses one national basket. Rent in a capital can rise far faster than the national price level, and the reading cannot see that.
- It compares a year ago with today, and says nothing about why the year went the way it did.
- It gives no advice about pay, and does not compare you with anyone but yourself and the middle of your country.