A common injunction of self-help and management advice in recent years has been to "make more mistakes". But in the wake of the MPs' expenses scandal and the credit crunch, enthusiasm for such simplistic bravado has evaporated, with some justification. Making mistakes is neither a virtue nor a vice in itself. Everything depends on the nature of the mistake.

A decision can have negative consequences, without the person who made it having erred. Officials, doctors and even politicians have often chosen what the evidence before them indicated was best, only to find out later that it had unforeseen, negative results.

We need to tolerate such mistakes because we should encourage people to make decisions based on the best assessment of available evidence, and we should be prepared to support them when this generally reliable procedure goes wrong. To do otherwise would be unjust as well as counterproductive, as it would lead to the removal of talent and unnecessary disruption.

Sceptics will say this is all too permissive. Isn't it culpable negligence to ignore the law of unforeseen consequences? And what of those cases where negative outcomes are not entirely unforeseen, but simply judged to be improbable? Why make such inherently risky decisions?

This is the perennial protest of conservatives, who argue that we may not understand why things work as they do, and that society is a kind of ecological system with which we meddle at our peril. This challenge is not taken seriously enough by progressives, who see it as an excuse for maintaining historic injustices and privileges. But sometimes we are justified in running the risk of screwing up - because the status quo is screwed up as it is. So, how great a risk we take depends on just how bad things are, and the best assessment we can make of the risks of change, with consideration of how easy it is to reverse if it all goes wrong.

The current economic crisis has been a good example of how these variables apply. Governments around the world have been forced to take drastic steps to try to stabilise the financial system. Some will turn out to have been more effective than others. Some may even have ended up, in part, lining the pockets of already well-paid executives. But doing nothing has not been an option, nor has it been possible to think through all the consequences before jumping in. Reversibility has not been a critical consideration: whichever direction we go in, it cannot be back to what got us into this mess. When the dust settles, we will praise some people for making brilliant choices and condemn others for making stupid ones, but we should also excuse many others for making understandable mistakes.

In encouraging a more mistake-friendly culture, it is important that the whole idea of mistakes does not become meaningless. There is a difficult balance to be struck: we need to be as intolerant of unjustified failures as we are tolerant of justified ones.

Although the rational case for a more mature mistake culture is a rigorous one, there are daunting sociological, psychological and institutional barriers to overcome. The rhetoric, as well as the logic, of mistake culture has to be developed. Such a rhetoric needs to address head-on those who will claim the admission of mistakes as a sign of weakness. What is more, we need to insist that it is weak never to admit when things have gone wrong, or to do so too late, after too much damage has been done.

It is weaker to let things rot than risk a cure that might not completely work. It is the weak who neither learn from small mistakes nor own up to big ones. Taking responsibility for our mistakes and being prepared to make the right kind requires strength, and builds it.

Julian Baggini's essay "Making Mistakes and the Rightness of Wrongness" is published by Demos. Download it at www.demos.co.uk