Why Published Inflation Is Not Your Inflation

The number and the feeling

Meera reads that inflation is at some rate and does not believe it. Her school fees have risen faster than that, her medical bills certainly have, and the gap between the published figure and her sense of it has been widening for years.

She is not wrong and the statistic is not wrong. They are measuring different things, and understanding how the published number is built explains exactly why the two diverge — and, more usefully, tells her which of her own goals the national figure will underdescribe.

How the published figure is built

A price index tracks the cost of a fixed basket of goods and services. Statisticians decide what goes in the basket and how much weight each item carries, based on surveys of what households actually spend money on. Each month, prices are collected and the weighted total is compared to a year earlier.

Three features of that construction matter to a reader.

The weights are somebody's average. Food carries a large weight because a typical household spends a large share of income on it. If your share differs — and it will, because the share of income spent on food falls sharply as income rises — then the same price changes produce a different total for you than for the index.

The basket is fixed for a period. It is revised occasionally, not continuously, so it lags changes in how people actually spend. Between revisions the index measures the cost of a pattern of spending that is progressively less like anybody's.

Quality changes have to be handled somehow. When a product improves and its price rises, part of that rise is buying something better rather than paying more for the same thing. Statisticians adjust for this, the adjustment is a judgement, and it is one of the main reasons technically literate people argue about inflation figures.

None of this is a defect. An index that tried to describe everybody would describe nobody. It is a tool built for a purpose — monetary policy, contract indexation, national comparison — and your household budget is not that purpose.

Why your rate differs, specifically

Four things push a personal rate away from the national one, and they are worth knowing because they identify which goals are most exposed.

Your spending mix. The largest effect. Someone whose spending is concentrated in categories rising faster than average experiences faster inflation, by arithmetic, with no mystery involved.

Your life stage. Spending patterns shift with age and circumstance. A household paying school fees, one paying rent, one paying a fixed home loan instalment and one paying for regular medical treatment are exposed to four different sets of prices.

Where you live. Prices and their rates of change vary by city and between urban and rural areas, particularly for housing.

Whether you can substitute. The index assumes some ability to shift towards things that have become relatively cheaper. Where you genuinely cannot — one particular school, one specific treatment — you bear the full price change with no offset available.

That last one explains why the divergence is worst precisely for the goals that matter most. A concentrated, non-substitutable, long-dated commitment is the case the national average describes least well, and it is also the case where being wrong is most expensive.

What we can and cannot show you

Our inflation figures come from the published national index, and the spread of what it has actually done is laid out in choosing an inflation assumption. Note what that page shows: even the national rate varies enormously from year to year, so a single planning number is the wrong shape of answer before any personal adjustment is considered.

We hold no separate index for school fees, medical treatment, urban rent, or any other individual category, and this page will not invent one. Statements of the form "education inflation runs at some particular rate" circulate widely, and we have not sourced any of them. Repeating one would put a specific, load-bearing number into a reader's thirty-year projection on no authority at all.

What can be said without the data is the direction and the mechanism, which is what this page is.

Estimating your own

You can do better than either the published figure or a guess, and it takes an evening.

Take your last few months of spending and group it into a handful of categories — housing, food, transport, education, medical, everything else — with the share each takes. That is your personal weighting, and it is the part the index cannot know.

For anything concentrated and important, get the actual price history rather than an index. The fee schedule from the institutions Meera is considering, for the last several years, is better evidence about her goal than any national statistic — and she can obtain it by asking. The same holds for rent in her area and for a treatment with a known cost.

Then plan with a range rather than a point. A goal concentrated in one category should carry a wider range than the national figure implies, because the national figure is an average of things that move differently, and averaging is exactly what is not happening inside her goal.

What to take away

The published rate is an accurate measurement of a basket that is not yours. The mismatch grows with how concentrated and how non-substitutable your spending is, which means it is largest for the commitments you are least able to abandon.

Use the national figure for what it is good for — a starting point, and a sense of how much rates themselves vary. For the goals that dominate your planning, go and find the actual prices. They exist, somebody publishes them or will tell you if asked, and they beat any index.