I am not sure who it was who said "All models are wrong but some of them are useful", but they had a point. I just made up the version that says "All models are wrong and some of them are harmful." I am thinking of austerity.
The model has always been wrong because it embraces the government as a household fallacy. It is true that if a household that is in debt cuts its spending it increases its chances of balancing its books. That is not true of a government for two reasons. First, government spending is national income. Cutting it does then only increase the chance of balancing the books if the spending adds less to income than it incurs in costs (meaning in economists' terms that the 'multiplier' is less than one). This is what the Office for Budget Responsibility has always assumed, except for investment. They are however wrong; when there is underemployment there is now a mass of evidence, from the IMF and others, that the multiplier is well over one. In other words, spending more than pays for itself and cuts impose more harm than the initial sum supposedly saved reduces spending by. Second, unlike a household a government with its own currency cannot have a debt crisis. It can always pay because it creates the currency required to make the payment and households can't do that.
What is the harm? This is becoming apparent. Data on dramatic increases in crime rates are clear indication that the police have now had resources cut beyond the point where their operations remain effective. They just add to the long list of services where this is apparent. From health, to education, to the rest of the justice system, to social care, and more, cuts are destroying much of what we value and leaving our society at risk. And all this to supposedly balance the books, which has never happened because cuts actually make the achievement of that goal harder, and not easier for reasons already noted.
This we knew though. Let me now add in something many are less familiar with. This is Baumol's law. What William Baumol suggested in the 1960s was that salaries in jobs that have not experienced growth in productivity increase in response to rising salaries in jobs where there has been productivity growth. So, for example, automation in manufacturing might increase productivity, and wages rise with it, but the salaries of string quartet players also rise as a result and there is no productivity gain to be had from playing a string quartet movement written to be played in six minutes in four minutes: in fact, you ruin it if you do. The example is of course contrived: the reality is that string quartet players have increased productivity through better transport, distribution methods and recording techniques. But the fact remains that their core product can't be compressed in time scale.
Nor, once admin gains are won, can policing, education, nursing, social care, social work, the justice system and many other services be compressed and still function. In fact, because they require face to face contact you ruin them if you do that. But the government, dedicated to an austerity programme that demands increasing productivity, demands that we do crush the time spent on these services, which undermines their value, and that the wages of those undertaking them be cut to break the relationship Baumol observed. But there's a problem with that too, because those who enjoy the higher pay set the cost of living e.g. by forcing the price of housing up. In that case those in the services where productivity gains have reached their limit suffer real pay cuts, a loss of job satisfaction as they're not allowed to do the job properly, and quit as a result. The exodus of teaching staff and the declining number of nurses is evidence of that.
And this matters. It matters because we literally cannot do without these people but apparently we can't afford them. I don't blame Baumol for this. Describing the phenomenon does not make him responsible for it. I blame governments, economists and politicians, none of whom have apparently appreciated that technology can only take us so far in many public services and then we reach the crunch point where dealing with people requires real time and irreducible interaction and there is nothing that can be done about this barring supplying a poor service, or none at all.
That's the conundrum at the real heart of modern government. We're phenomenally more productive than we ever have been. And apparently, as a result, we can't afford large numbers of things that were possible in the past. Unless, of course, we change the model. Remember all models are wrong. And what is now clear is that model we're using is not helping.
The model we're using is one that says the state has a limit to the size it can be in proportion to the private sector. And it says the state can supply all the services we need despite this constraint, even if Baumol's law applies. And that model is wrong. We cannot do that. What Baumol's law actually implies is that as we get richer in terms of the productivity of some labour (which, because of robotics, is likely to continue to grow even as total employment falls) the price of public services will rise and that we must accept the growth in the size of the state sector that follows if we are to still enjoy those services we once thought of as basic.
And of course, that means we will pay more tax too. But that tax will be paid out of the proceeds of the spending, because as I argued yesterday, spend comes first and tax comes second, and unless there is full employment at fair pay (which there clearly is not in the UK) the spend does not result in inflation. In fact, the tax ensures that is true. Indeed, because of the multiplier effect, we might even, perchance, balance the books.
But that's not the goal. The goal is to have the services that we are told, despite our apparent riches, that we cannot now afford but which we could have when we were much worse off. And of course we can have those services: it is glaringly obviously absurd to say that because we are richer we can no longer afford to teach people, have police or provide care. It's just that the economic model we are using is wrong. And that's the bit we have to change as a result, which is, unfortunately, the hard bit. Although heaven knows why.