The Economist's factchecking department is a place where legends are broken. Hacks speak of their work with wonder. Anne McElvoy, their Public Policy Editor, told an event recently "I have just suffered a Wednesday with a factchecker. It is an absolutely properly gruelling process and it is intended to be so."

Perhaps this attention to detail is why such elevated minds as Jonathan Portes, Director of the National Institute for Economic and Social Research, are moved to action when they spot a mistake.

Two weeks ago the Economist published an article which claimed that half of youngsters on the government's unpaid work experience scheme "found paid work soon after finishing."

That's not true. In fact, as Full Fact showed before the Economist went to print, that half is simply not claiming benefits. They may have jobs, they may not. They may be back in education or simply have dropped out of the benefits system.

Mr Portes concluded "@TheEconomist 0, @FullFact 1" and dispatched a letter to the paper to point out its mistake, which the paper duly published.

However, the online version of the article is unchanged and nowhere has the Economist said that they accept Mr Portes' point.

Full Fact has always said that simply printing a dissenting letter does not amount to a correction. Publications must acknowledge errors themselves so that readers can be sure of the distinction between fact and conjecture.

The unchanged website may well be an oversight but with the Economist increasingly distributed as a digital product in common with other publications, it is an important one.

The care that the Economist bestows on its factchecking may mean that it has never formalised its corrections process. Nevertheless, mistakes do inevitably creep into print occassionally, and it is important that these are properly corrected.