Thursday, 31 October 2013

A Very Simple Error By Researchers That Should Know Better

The LSE (the London School of Economics) recently published a drastically flawed report on Labour's spending from 1997 to 2010 entitled Did Labour's social policy programme work?.  Their conclusion is that it did work - and they proffer several examples of why this is the case (a conclusion supported by The Guardian's Polly Toynbee in this article). Now I fully expect Polly Toynbee to be getting things like this wrong - she is very prone to this - but the drastic flaw that underlies the report is not something one should expect from the researchers at the LSE.

Here’s what's wrong with it (and I had to read it twice to make sure I hadn't missed it somewhere in the text - but I hadn't) - in asking whether Labour's spending from 1997 to 2010 'worked' it fails to mention what should be the most important part of the inquiry - what it actually means to say spending 'worked'. Reading between the lines, all the report says is that Labour spent more on X, Y and Z, and as a consequence had more of X, Y and Z.  Well 'No kidding!' Sherlock! - but that doesn't tell us whether the spending worked, and it is certainly no justification for claiming that the period from 1997-2010 was successful spending by Labour.

The way to measure whether spending has brought about value for money is by what is called ‘consumer surplus’ - which is the amount spent against the perceived value of the thing purchased. Suppose that tomorrow Pete gives Ted £75 and tells him to buy something on which he thinks Pete would have willingly spent the money. With the £75 Ted spends all the money on a portable DVD player. If Pete would have paid £100 for the portable DVD player and he actually paid £75 (via Ted) then the consumer surplus is £25. When a consumer has £25 consumer surplus that means £25 of consumer surplus has been contributed to society, and the spending has therefore 'worked'. If Pete would only have paid £30 for the £75 portable DVD player then Ted's spending hasn't 'worked' for Pete.

To apply that to government spending and taxpayers' benefits; the more the societal consumer surplus the more the Labour government's increased spending has 'worked'. It's true the Labour government spent a lot on education, health, transport, home affairs, and so on, but the spending could only be deemed to be successful if what was spent on these things was less than, or at the very least equal to, what would have been spent by consumers.  In the same way that I asked whether Pete had consumer surplus on the portable DVD player transaction, the LSE reasearchers should not even begin to assess whether Labour's spending 'worked' without asking the most important question of whether the taxpayers had consumer surplus on the Labour spending.

In other words, it just won't do to ask whether Labour's spending provided increased education, health, transport and home affairs benefits - the only important thing is whether the spending provided a level of improvement that would have been optimally chosen by the taxpayers, or whether they'd have been better off with lower spending and more money in their pockets.

Let's return to Pete's situation to show how silly the inquiry is without considering consumer surplus; if I take £150 out of Pete's bank account, and buy him a portable DVD player on Amazon for which he would have only paid £100, then the spending hasn't worked for him. It would be ridiculous of me to then publish an article saying my spending had 'worked' for Pete because he increased his expenditure on Amazon and got a brand new portable DVD player in return. That's what the LSE have done - it's an error one wouldn't expect from them. Instead of asking whether the 'beneficiaries' would have willingly paid for the things on which Labour spent all that extra money the LSE asked what Labour wanted the taxpayers to get from government spending, and then concluded that because Labour wanted them to get the extra expenditure on education, health, transport, home affairs, then that spending must have worked. That is the kind of inquiry that is beyond ineffectual.

Two further things. In the first place, I actually think a lot of Labour's spending between 1997 and 2010 was wasteful - too much of it went to costly PFI groups, over-expensive consultants and general profligacy, and not enough of it went on the front line services.  And in the second place, it seems clear to me that the enquiry to work out whether the taxpayers attained an overall surplus on Labour's spending is an all-but-impossible pursuit - given the level of, and presumably largely differing, opinion needed to be solicited, and then measured up against actual costs and benefits of the spending in every area - no research group would ever get a handle on such a complex nexus.

So in all probability, even if the LSE had sought to include the very essential 'consumer surplus' factors on which the inquiry should have hinged, they would have found that such an enquiry was too complex to elicit even near-precise conclusions, which renders their research project "
Did Labour's social policy programme work?" ineffectual and inadequate to the task.

 * Photo courtesy of