Perhaps the best single piece of news in the unemployment data isn't the reduction in the number of unemployed people, which, though a positive indicator, obscures a number of less positive results. It is instead the reduction in the number of unemployed people per vacancy, which now stands at 5.6, down from 5.9.

The data has also allowed the TUC to recompile their list of employment blackspots – those places where this ratio rises to unacceptable heights:

These figures should pour cold water on the government's triumph at the evidence that their work experience programme is effective at getting people back into employment. Yes, this is good news (although also, of course, the minimum that should be expected from a programme which, until recently, threatened people with loss of benefits if they did not volunteer their time). But unless the programme reduces this ratio of unemployed people to vacancies, then all it does is fill vacancies with one set of people rather than another.

When there are five people competing for every vacancy the best way to make unemployment go down is to create new jobs to fill. There are other ways to reduce the figure - for instance, the government could make seeking work so unpleasant that people stop claiming jobseekers' allowance and eventually stop looking for work altogether - but they are merely massaging the figures, rather than solving the problem. Helping people get what vacancies there when low growth means that there are no new jobs will do nothing for the unemployment figures.

It is the fundamental difference between acting on an individual level and a societal one.