Is government spending too low? It sounds like a silly question, but it’s the one raised by this new paper by Zohal Hessami. He estimates - based on 12 EU nations in the 1990s - that there’s an n-shaped relationship between the size of government and subjective well-being.
But here’s the thing - the peak in well-being comes when government spends around 70% of GDP. This suggests that, from the point of view of maximizing well-being, all governments spend too little. This chimes in with Sunder’s point - that the general public are in no mood for spending cuts.
Which raises the question: how can one argue against this?
We can quibble with the econometrics (we can always quibble with the econometrics); the fact that such huge spending is out of our sample means it’s dangerous to infer that it really would increase well-being. And the in-sample data shows that increases in government spending from high levels have only a small positive effect upon well-being.
But these are mere quibbles. A stronger argument is needed.
Many of you would say that this estimate holds GDP per capita constant but in practice, bigger government would depress GDP growth, and folk would become less happy on this account.
This, though, is an awkward argument. In developed economies, GDP has only a small effect on well-being, so it would require large falls to offset the benefit of higher public spending.
Instead, I’d suggest two other arguments.
First, this estimate holds corruption constant. In fact, though, as Mr Hessami shows, increased corruption causes a negative relationship between government spending and well-being. This matters, as it’s likely that once government exceeds a certain size, corruption increases. Stephen Byers could not have offered his “influence”* for sale had government been so small that it didn’t affect business much.
Secondly, there’s the question: why should well-being have priority over liberty? This philosophical question gains practical force from the fact that many of the public’s anti-liberal demands rest upon irrationality and cognitive biases.
And there might be such a bias at work here - the representativeness heuristic. Maybe support for higher government spending rests upon the belief that inputs (money) are representative of outputs (better health and education, more support for the worst off). But this belief might be mistaken; it is at least possible that these outputs could be achieved with lower spending. If so, support for higher spending might rest upon an error.
Whatever you think about the merits of these arguments, there is one strong implication here. There’s a flat contradiction between libertarianism on the one hand and utilitarianism and democracy on the other.
* Dsquared has the final word on this.
But here’s the thing - the peak in well-being comes when government spends around 70% of GDP. This suggests that, from the point of view of maximizing well-being, all governments spend too little. This chimes in with Sunder’s point - that the general public are in no mood for spending cuts.
Which raises the question: how can one argue against this?
We can quibble with the econometrics (we can always quibble with the econometrics); the fact that such huge spending is out of our sample means it’s dangerous to infer that it really would increase well-being. And the in-sample data shows that increases in government spending from high levels have only a small positive effect upon well-being.
But these are mere quibbles. A stronger argument is needed.
Many of you would say that this estimate holds GDP per capita constant but in practice, bigger government would depress GDP growth, and folk would become less happy on this account.
This, though, is an awkward argument. In developed economies, GDP has only a small effect on well-being, so it would require large falls to offset the benefit of higher public spending.
Instead, I’d suggest two other arguments.
First, this estimate holds corruption constant. In fact, though, as Mr Hessami shows, increased corruption causes a negative relationship between government spending and well-being. This matters, as it’s likely that once government exceeds a certain size, corruption increases. Stephen Byers could not have offered his “influence”* for sale had government been so small that it didn’t affect business much.
Secondly, there’s the question: why should well-being have priority over liberty? This philosophical question gains practical force from the fact that many of the public’s anti-liberal demands rest upon irrationality and cognitive biases.
And there might be such a bias at work here - the representativeness heuristic. Maybe support for higher government spending rests upon the belief that inputs (money) are representative of outputs (better health and education, more support for the worst off). But this belief might be mistaken; it is at least possible that these outputs could be achieved with lower spending. If so, support for higher spending might rest upon an error.
Whatever you think about the merits of these arguments, there is one strong implication here. There’s a flat contradiction between libertarianism on the one hand and utilitarianism and democracy on the other.
* Dsquared has the final word on this.
"This matters, as it’s likely that once government exceeds a certain size, corruption increases. Stephen Byers could not have offered his “influence”* for sale had government been so small that it didn’t affect business much."
If government was smaller, wouldn't there be less 'influence' for sale, pushing its price up?
Posted by: Adam | March 24, 2010 at 12:19 PM
Also spending as a % of GDP can mean many things, after all there's no theoretical reason why it can't be 150% of GDP. So perhaps 70% of GDP in which the lion's share of that is a citizen's income would be different for welfare than a more interventionist 70%.
Posted by: Matthew | March 24, 2010 at 12:41 PM
Nice post. But on this point:
"Secondly, there’s the question: why should well-being have priority over liberty?"
I don't think there's always a straight trade off here. A state-directed reduction in Alice's holdings (i.e. tax) might result in an increase in liberty for Bob (if it curbed Alice's ability to collect rent from Bob, say) additional to any effect of state spending.
Posted by: Charlie | March 24, 2010 at 12:55 PM
One of the objections to utilitarianism has been the calculus of felicity - or dificulty of it.
If this has solved this at all we can just concern ourselves with questions of happy pigs.
Posted by: alanm crisps | March 24, 2010 at 04:36 PM
Why stop at 70%? Let the State take all our income and have us all come on bended knee in front of it to supplicate for a little of it back. That presumably would result in the maximum amount of happiness, no?
Posted by: Jim | March 24, 2010 at 08:03 PM