Researchers at the London School of Economics recently published a report assessing the increased social spending of the 1997-2010 Labour government. They concluded that "there is clear evidence that public spending worked". Enthusiasts for government spending, including the Guardian columnist Polly Toynbee, celebrated the finding. The wisdom of government spending had been scientifically confirmed.
It has not. The supposed discovery is derived not from the statistical analysis the researchers LSE conducted but from a play on words. When the LSE academics say that Labour's spending worked, they do not mean what an economist or consumer would mean by "worked". They mean something that makes their "finding" almost tautological.
On the standard view, spending works when it buys something for which the consumer would be willing to pay more: that is, when it creates a "consumer surplus". Suppose you would pay up to £110 for shoes that actually cost £100. Then your consumer surplus is £10. This provides a measure of how well spending works. The greater the consumer surplus, the better the spending.
The Labour government's spending provided people with all sorts of things: healthcare, education, bus passes, TV licences and so on. Would the consumers of these things have willingly paid more than Labour spent on them? If so, the spending worked. If not, it didn't. Consumers would have been better off with the money.
So you might expect the LSE researchers to have tried to find out what people would have willingly paid for the things Labour bought them. In fact, they asked whether Labour's spending bought what Labour wanted it to buy. Did the extra spending on education provide people with more education? Did the extra spending on healthcare provide more healthcare? And so on.
To see the absurdity, imagine I taxed ten people £1,000 altogether and bought each of them a pair of shoes costing £100. A social scientist wonders if my spending worked. Instead of asking how much my putative beneficiaries would have willingly paid for the shoes, he asks what I wanted them to get. Because I wanted them to get the shoes I bought, he concludes that my spending worked.
Once the LSE researchers took this approach, their "discovery" that Labour's spending worked was guaranteed. Admittedly, government spending does not always deliver its intended result. But Labour nearly doubled real spending on education, to take one example. How could this have failed to produce more education?
Many politicians believe they can help us by coercing us. It is an implausible idea. But, these paternalists claim, the evidence proves it; they have science on their side. Take a closer look at this "science", however, and you usually find that what they have on their side is a gross conceptual or methodological error.