Understanding Inflation: CPI, Core, and Why Your Experience Differs

Few statistics are simultaneously as important and as widely misread as the inflation rate. It determines interest rate policy, pension uprating, wage negotiations and index-linked contracts, and it is routinely described in ways that obscure what it measures.

What the headline number is

A consumer price index tracks the cost of a fixed basket of goods and services over time. Statistical agencies survey household spending to determine what goes in the basket and how heavily each item is weighted, then collect hundreds of thousands of prices each month and compute the change.

Two things follow immediately. First, the rate is a change, not a level: inflation falling from six per cent to two per cent means prices are still rising, just more slowly. Prices only fall in absolute terms during deflation, which is rare and generally regarded as more dangerous than moderate inflation. Second, the basket is an average, and no household buys the average basket.

Why your inflation rate differs

The weights are what create the gap between the published rate and personal experience. Housing costs, energy and food carry weights derived from national average spending. A household spending a much larger share of its income on energy and food than the average — which lower-income households systematically do — experiences a different rate when those categories move sharply.

This is not a flaw in the statistic but a limitation of any single number. Several statistical agencies now publish inflation estimates by income decile or household type precisely because the aggregate conceals distributional differences that matter enormously for policy.

Core inflation and why it exists

Core inflation excludes food and energy. This strikes many people as excluding exactly the things that matter, and as a measure of household experience the criticism is entirely fair.

But core is not designed to measure household experience. It is designed to identify the persistent, domestically generated component of inflation, and food and energy are excluded because they are volatile and driven largely by global supply conditions outside domestic policy influence. A spike in energy prices caused by a supply disruption will reverse; raising interest rates in response would tighten policy into a shock that is already resolving itself.

Related measures try to achieve the same goal differently. Trimmed mean inflation discards the largest increases and decreases each month rather than fixed categories. Median inflation takes the middle of the distribution. Services inflation excluding energy is watched closely because it correlates strongly with domestic wage pressure.

The quality adjustment argument

A persistent complaint is that official inflation understates the true cost of living because of hedonic quality adjustment — the practice of adjusting recorded prices to account for improvements in what is being bought. If a laptop costs the same as last year but is substantially faster, the index treats this as a price decline.

The reasoning is defensible in principle: an index should measure the cost of a constant standard of living, not of a constant product. The criticism has force where quality improvements are not things consumers chose or value, and where a cheaper equivalent is simply no longer available. The practical effect on the headline rate is generally estimated as modest, but it is not zero and it is concentrated in categories with rapid technical change.

Shrinkflation and skimpflation

Shrinkflation describes a reduction in package size at an unchanged price. Contrary to widespread belief, competent statistical agencies do capture this, because prices are collected per standardised unit rather than per package. A cereal box shrinking by ten per cent at the same price registers as a ten per cent price increase.

Skimpflation is harder. This is a reduction in service quality at an unchanged price — longer waits, fewer staff, removed features. It is genuinely difficult to measure and is generally not captured well. Where it is widespread, official inflation understates the deterioration in what consumers receive for their money.

Different indices, different purposes

Most countries publish several measures. Some include owner-occupied housing costs and some do not, which matters enormously given housing’s weight in household budgets. Some use a formula that accounts for consumers substituting towards cheaper alternatives when relative prices change; others assume a fixed basket. The difference between formulae typically runs to a few tenths of a percentage point annually — small in a single year and substantial when compounded over a working life in a pension uprating formula.

Producer prices as a leading signal

Producer price indices measure costs at the factory gate rather than at the till. They tend to lead consumer prices, since input cost changes take time to pass through, and they are considerably more volatile because they lack the buffering effect of retail and distribution margins. As an early indicator of turning points they are useful; as a predictor of magnitude they are not, because pass-through is incomplete and varies with competitive conditions.

The practical reading

For understanding policy, watch core and services inflation, since those are what central banks respond to. For understanding household pressure, look for distributional breakdowns rather than the headline. And treat any single month’s figure with caution: seasonal adjustment, base effects from the comparison period a year earlier, and one-off category movements make individual readings noisy in ways that three-month averages are not.