The Natural Rate of Unemployment: The Threshold Below Which Employment Is Paid For With Inflation

The natural rate of unemployment is the share of people out of work that remains in an economy even when it is running at full capacity and demand is not lagging behind supply in any respect. It is not a malfunction and not a failure of the authorities: some people are always between jobs, and some are looking for a position in an industry where their skill is no longer needed. The practical meaning of the indicator lies not in the figure itself but in its role as a threshold: when actual unemployment falls below the natural rate, any further growth in employment is paid for with accelerating wages and prices. That is exactly why central banks watch the gap between actual and natural unemployment as one of the indicators of overheating — and through the policy rate this view reaches the yields of OFZ and the valuation of stocks.

What the "natural" part is made of

The natural rate is the sum of two kinds of joblessness that cannot be cured by demand.

Frictional unemployment is search time. A person has quit voluntarily, moved, graduated from university, or turned down a first offer for the sake of a better one. While the matching goes on, that person is counted as unemployed, even though there are vacancies in the economy. The size of this part depends on how quickly worker and employer find each other: on the quality of job-search services, on the size and duration of unemployment benefits, on geographic mobility.

Structural unemployment is a mismatch between skill and vacancy. Demand has shifted to another industry or another region, while qualifications have stayed the same. Vacancies and unemployed people exist at the same time, but they do not match up. This part changes slowly and responds not to the policy rate but to education, migration policy and shifts between industries.

Everything above this sum is cyclical unemployment, a consequence of weak demand. This is the part that is "treated" by easing monetary policy. The natural part cannot be removed by easing: an attempt to do so ends not in employment but in prices.

Why it is a threshold and not an average

The early version of the Phillips curve promised a stable trade-off: accept higher inflation and you get lower unemployment. In 1968 Friedman and Phelps objected: the trade-off works only as long as people do not expect inflation. As soon as expectations have adjusted, wages demand indexation, the real cost of labour returns to its previous level, and employment returns to where it started. What remains is higher inflation with the same unemployment.

Hence the indicator's second name — NAIRU, the non-accelerating inflation rate of unemployment. The wording is more precise than the everyday one: this is not "normal" unemployment and not "good" unemployment, but the level at which inflation neither speeds up nor fades away by itself. The stagflation of the seventies was the practical confirmation: high inflation and high unemployment coexisted because the natural rate had itself shifted upwards by then.

The value cannot be measured — it is estimated

The natural rate is not observed directly: it is not published in the statistics, it is calculated with models, from the behaviour of inflation and wages. Different models give different answers, and all of them are revised after the fact, sometimes substantially. This has a practical consequence: when the regulator refers to an "overheated labour market", it is relying on an estimate with a wide band of uncertainty, not on a measurement. Building a trade on the presumed location of this threshold means betting on a model assumption.

On top of that, the rate itself drifts. Demographics, the share of people in salaried employment versus self-employment, the availability of retraining, the structure of demand for labour — all of this moves it in either direction, and the shift can usually be noticed only after it has already happened.

What this means for an investor in the Russian market

There is no direct trading idea here — there is the context in which rate decisions are made. The logic of the chain is as follows: persistently low unemployment alongside rising wages is read as a sign that demand is running up against the limit of labour supply; that is an argument for tight policy; tight policy means high yields on long OFZ and pressure on the valuation of companies whose profit is tied to consumer credit. The reverse picture is an argument for easing.

A separate, more down-to-earth channel is costs. A shortage of staff means competing for workers with pay. For labour-intensive businesses this is a direct hit to margins, and it can be seen not in a macro review but in issuers' financial reports — in the trend of personnel expenses. What is released and when is convenient to follow in the events calendar, and the regulator's decisions and the commentary on them are in the news.

A caveat for the sake of honesty: the platform does not keep its own labour market statistics. Data on unemployment and wages are published by Rosstat and the Bank of Russia — here it is possible to trace only the market's reaction to them, not the primary source.

A common mistake

The natural rate is often confused with a "level" in the stock-market sense — a price mark from which a quote bounces. In Russian the same word is used for both, and that word is all they have in common. A market support level is a point on a price chart, while the macroeconomic natural rate is a calculated state of the economy. The other meanings of the word are covered in the glossary.

The second mistake is to treat the threshold as a target. An economy can push unemployment below the natural rate for a short time; it cannot keep it there. The price of the attempt becomes visible with a delay, and it is paid not by the labour market but by holders of long bonds.