Tuesday, 14 May 2013

The misuse of the fiscal multiplier

The misuse of the fiscal multiplier effect is beginning to annoy me and it's about time that the record is set straight on the issue.

You are seeing classic lines like this :
Tax cuts are only one third as effective as government investment in new capital spending projects
and the propaganda is then backed up by some magic number calculation that shows the multiplier is 1.5 for capital spending and 'only' 0.5 for tax cuts. Obviously the bigger number is better because, well, it always is isn't it.

On the other side you have this belief that if the multiplier is less then 1 for any activity then you shouldn't undertake it.

Let's expose the assumptions those statements are based upon:
  1. There is a fixed amount of money available
  2. If the government spends from this fixed pot of money, it automagically crowds out an alternative activity with a multiplier of one.
  3. That you can ignore the effect on the distribution of savings.
All of those beliefs are rubbish.

Let's look at what the multiplier is first so you can understand what is underneath the numbers. 

The multiplier is fairly straightforward. If the government gives money to somebody then they will spend some of it and save some of it. What they spend likely causes a real transaction to occur - and an amount of taxation. What they save will either go to reducing debt or building up financial savings. And in either case the banking system retains what isn't taxed away as bank reserves.

What is spent becomes somebody else's income. Rinse and repeat. 

The calculation of the multiplier however only adds up the GDP impact of the transactions over a time period. It is silent on the impact of extra savings caused by the process. And that is where the problem starts.

Firstly it biases the multiplier against tax cuts. The mathematics of the multiplier demonstrates that clearly. It's standard knowledge in simple Keynesian Models that the tax multiplier = 1 - spending multiplier.

So obviously the fiscal multiplier for a tax cut is going to be less than equivalent government spending. That's because you are, as a matter of policy, allowing extra savings first before any consumption happens. Duh!

Secondly the debate is always based upon an amount of government expenditure. Functional finance thinking shows this is putting the cart before the horse. Rather than saying how much output can we get for £10bn of government action, we should be saying given we have a GDP output gap of £80bn how much do we need to deploy on the various policies to get that - given the multipliers and savings/income distribution of those policies.

Thirdly there is this idea that government actions inevitably prevents private sector action. This is crowding out again. Endogenous money theory demonstrates that the money supply is elastic. It goes up and down as required by the demands in the economy. You won't run out of money. So if there is stuff stood idle, it can always be brought into use without affecting anything else.

So suggesting that government action with multipliers less than one shouldn't happen is nonsense. It ignores the value of net-savings. It ignores the obvious point that a do-nothing alternative means that nothing additional happens.

Relieving the burden of debt on individuals is a valuable action. It makes them feel more secure and therefore more likely to spend in the future. It reduces the size of bank's balance sheet in the economy and allows you to narrow them more easily - diluting their economic power so that they are no longer too big to fail.

I think that's something we should be looking at doing. Yet you'll still get people pointing out that it has a low fiscal multiplier and we'd be better off building bridges to nowhere instead.

There's more to this game than GDP I feel. The distribution of net-financial assets matters.