Old sayings like "If you can't measure it, you can't manage it" are not always strictly true but it is true that, without either direct or indirect measurement, it is hard to know whether the actions you are taking are actually working. This is why governments and business are so keen to set targets. But, as soon as you do so, you modify the behaviour of the people who must meet the targets. We read stories about the British National Health Service where patients are denied formal appointments in order to keep them off the waiting lists, or they are given pre-appointments with the same objective. This is because government has set targets for the length of waiting lists and the maximum time until a patient is seen or treated. If they are not formally on the list then they do not contribute to the statistics. And if those statistics are examined to determine the funding of an organisation or the remuneration of an individual then there are powerful incentives to make them say what is required.
A recent report showed that police forces in the UK are routinely misreporting crime data by redefining classes of crimes so that, often, what we think is a crime turns out not to be. Of course once targets affect this behaviour there is a knock-on effect on public behaviour since they stop reporting crimes to a police force that does nothing about them...what's the point? I have a recent personal experience arising from credit card fraud perpetrated against a business: what is the point in reporting the fraud to a non-police agency when there appears to be no action resulting from this: no attempt to catch the fraudster? That, in turn, has a corrosive effect on general public confidence in the police.
But let us not imagine that targets only change behaviour in the public sector. Try creating a fair and effective bonus scheme for a salesforce! Make it a monthly target and sales from months either side mysteriously move into one month in the middle. And what about off-balance sheet financing? The use of leasing and similar instruments in business is often used to make it look as if the organisation has a lower debt level than it really has. Whilst the debt level is usually not a formal target, though lenders may make it one, there are generally accepted unofficial targets set by bank and stock market analysts - so an incentive exists to adjust the published figures.
We must not despair - there is a way of both setting targets to guide behaviour whilst also protecting against unintended consequences. But that way is expensive and requires continual evolution to match the evolving ducking and diving of those who try to cheat. There must be an audit function that checks the reported figures and there must also be punishment for those who cheat. Unfortunately that punishment rarely exists in the public sector. False police statistics came to public attention because a young police officer testified to a committee of MP's but instead of being a national hero he is currently on a disciplinary charge and there is little chance that those MP's will step in to help him. It seldom works out well for whistleblowers. Fortunately there remain people who answer to moral outrage which makes them speak out against things that are plain wrong, regardless of the consequences for themselves. We should be grateful.