29 January, 2008


I was going to post this as a comment on Elliott's post below, but it was long so I made it a real post.

Let's deconstruct Cassady's comment:

Isn't that one of the main economic conundrums? I seem to recall that basically no matter what we do, the marginal propensity to spend goes down even if that lowest quintile (you and me, friends)gets the highest proportion of rebates. I could be wrong about this, but I also think that marginal propensity to spend is one of those quirky statistics that is very difficult to accurately predict. I mean, economists rarely take into account such fantastic products as "Jump to Conclusion" mats and Tickle-me Elmo's. Sign me up for two!

(As an aside, Cassady's conception of the income distribution of the United States is woefully inadequate. Those of us who aren't students are certainly not in the lowest quintile, and I wouldn't be surprised if most of us fall into the middle 20%. Yes Virginia, 40% of households (not even individuals) in the United States make less than $34,738 per year.)

There are a number of competing theories about the marginal propensity to spend (or consume, as it is usually put). One is known as Ricardian equivalence. Basically, the government has a budget constraint, at least in the long run. They can't spend more than they tax. So, if the deficit increases (via, say, a stimulus package) and everyone gets more money, they realize that the government will have to raise taxes by exactly the same amount in the future and so they save the money. Under this theory, stimulus has no effect.

The rest of the theory is basically a melange of tweaks that are made to this basic model to eke out a positive effect on consumption from government spending or tax cuts. If you include a constraint on saving for example. Then there are behavioral models, which don't even try to rationalize behavior, but try to use insights from psychology to create a model that is better at predicting behavior. Predictions from these models are mixed.

So basically, we know nothing.


Elliot said...

But wouldn't the Ricardian equivalence theory support rebates for lower income households? If people are making their decision whether or not to consume more based on the realization that the government will have to tax more in the future to compensate, than wouldn't those who will get taxed less in the future spend more immediately?

Elliot said...

Also, I will say that I would not follow the Ricardian model. I was just thinking about what I would do were I to receive a $600 cheque in the mail tomorrow. I can say pretty safely that I would 1) take Eremita out to a very nice restaurant 2) buy some frivolous stuff like booze and books and 3) put the rest towards groceries and or rent. If I saved anything, it would not be much.

spencer said...

A few points. First, you're right, it would support rebates to lower-income households to some extent. But this would have to take the form of a dynamic transfer from rich to poor (the rebate would have to be more progressive than the following tax increase). So the theory would say that you may as well just increase the progressivity of the tax system today while keeping the change revenue-neutral and it would have an identical effect.

However, if you believe in Ricardian equivalence, you probably also believe in the permanent-income hypothesis, which says that people try to smooth income over time. That is, a rational actor will borrow or save in such a way that she can consume the same amount each year. ($40,000 and $40,000 is better than $50,000 and $30,000). Under the permanent-income hypothesis, a $600 windfall today should be amortized over her lifetime, say 60 years, leading to an increase in consumption this year of only $10. So what you are saying doesn't really rescue the policy prescriptions of this theory, taken as a whole.

You're also right that Ricardian equivalence does not hold empirically. But to some extent, it depends on people's expectations. You are a student with a low but certain stream of income. If you were a manufacturing worker or retail worker or someone else with a high chance of being laid-off in a recession, would you blow the $600 on booze and books or would you save it for the bad times ahead?

Under this view, recession is a coordination failure: if only everyone believed that there would not be a recession, there wouldn't be one. In this world, the Fed's role is to organize expectations, which they have not been doing a great job of lately.