What am I missing on Ricardian equivalence?

Paul Krugman appears to be saying that a temporary increase in government spending can provide a stimulus EVEN WHEN full Ricardian equivalence holds. All the comments on his blog are saying how simple his explanation is, and how the Chicago school is full of morons – but I don’t really think his argument water tight.

He says that when we have a permanent increase in spending, household’s foresee that they will have to pay back that same spending in tax increases – so that there is no stimulus in this environment.

He then says that if we have a 1 period increase of $100m there will be a stimulus – as the $100m is paid off over the infinite horizon, but the $100m stimulus is there now. However, it appears to me that he has forgotten that the households choice of consumption is endogenous.

If the government borrows to fund a $100m transfer to households then the household realises it has to pay off $100m in future taxes (in present value). As consumers want to “smooth their consumption path” through time they will save this $100m that turns up in order to maintain consumption levels in the future periods when they face higher taxes.

He seems to assume that the increase in savings occurs equally across periods – but in a world with perfect Ricardian equivalence I don’t think this is the case. This is because in a perfect Ricardian environment household treat government borrowing just like household borrowing and adjusts private savings rates accordingly.

When we have pure Ricardian equivalence the difference between a temporary and a permanent increase in government spending doesn’t matter – it still all gets saved.

Now, as soon as we assume liquidity constraints and the such the permanent vs temporary difference becomes VERY important – but that isn’t what Krugman is doing. Instead of attacking the idea of Ricardian equivalence (which is very justifiable) he says that the Chicago school is mis-understanding their own model – but this doesn’t seem to be the case to me.

What about when the spending is not a transfer to households?

This is the more interesting point. When the government spending is on something, instead of a transfer, the result is not as clear cut. In the case of a temporary stimulus we would expect to see real interest rates rise, consumption fall, investment fall – but (static) net output would increase. Even if we have perfect Ricardian equivalence.

However, this all stems from an effective increase in the size of government. Even though government was only larger in terms of spending for one period – the cost of funding that government exists for eternity. Furthermore, by crowding out investment and consumption it is far far far from clear that such a more is welfare maximising – even if it increases the immediate level of output in an economy.

If this is the argument Krugman and co were making – then it is essentially a very different beast to the assumptions of the Chicago school. One is looking at a larger government, and one is looking at a situation with no lifetime change in the size of government. Making this difference in assumption clear would be more useful than arbitrarily acting like other schools of economics are silly.