A number is not a labour market

When politicians, institutions and media communicate employment through one headline figure, several questions disappear.

Istat’s May 2026 data already show why these distinctions matter. Year on year, employment increased among people aged 25–34 and those aged 50 and over, while it decreased among 15–24-year-olds and 35–49-year-olds. Permanent employees and self-employed workers increased, while fixed-term employees decreased.
That is a considerably richer story than “employment is up”.

“A statistic can be perfectly accurate and still leave out most of what people actually need to know.”

The same applies to wages. On 29 July 2026, Istat reported that average contractual hourly wages in January–June 2026 were 2.6% higher than in the same period of 2025. But in the second quarter of 2026, after ten consecutive quarters in which contractual wage growth had exceeded inflation, that relationship reversed.

Employment quantity matters. Employment quality matters too.

What the headline does to the person reading it

This is not only an economic problem. It is a communication problem.

Imagine spending months applying for jobs. You rewrite your CV, learn new tools, complete courses, attend employment programmes, adapt to automated screening and perhaps even pay for platforms promising to improve your chances.
Nothing happens.
Then you repeatedly hear that unemployment is falling and employment is booming.

At some point, a structural problem can begin to feel like a personal failure.
If the market is doing so well, perhaps you are the problem. Perhaps you are not skilled enough, flexible enough, young enough, experienced enough, visible enough, optimised enough.
That is why the framing matters.

“When structural problems are communicated only through reassuring averages, people can end up interpreting systemic friction as personal failure.”

Measuring activity is not the same as measuring outcomes

The same communication problem appears in employment policy.

Those numbers tell us how much the system did.
They do not, by themselves, tell us what happened to the people afterwards.

How many found jobs? How stable were they? How long did placement take? What were the wages? Did the work use the skills people had been trained for?

These are not hostile questions. They are evaluation questions.

Ask better questions

The point is not that official statistics are fabricated, nor that every positive labour-market indicator should be distrusted.
The point is that communication selects reality.
What we choose to measure, highlight and omit determines what the public thinks is happening.

The same applies to job platforms. If a marketplace presents thousands of vacancies, the number of listings tells us very little unless we know how many correspond to genuine, active hiring.

That question has now become serious enough to attract government scrutiny in the United States. On 14 July 2026, the Texas Attorney General opened an investigation into allegations involving LinkedIn, Premium subscriptions and fake or misleading job opportunities.

So the next time somebody says employment is rising, my response is simple.
Good.
Now tell me: where, for whom, under what conditions, for how long, and for how much?

Until we ask those questions, we are not describing the labour market.
We are just reading a headline.