Cases 4 and 5: Where the Money Goes

Greg delves a little deeper into the question of taxation and brings up one of my constant refrains. When analysing tax policy it is critical to think about what actually happens to the revenues.

Traditionally, economists simply assume that the revenue is given right back to the population in general. This is to avoid the multitude of murky issues that arise when the government starts to actually spend money.

Greg's three cases, however, miss what I think are the two most important real world outcomes.

For those who came straight here he mentions

1) The government spends the money on a useless war. I always say "Throws it down a rat hole", but the example of war is more realistic and timely. Its important to remember that most federal employees are still soldiers.

2) Rebates it back to the people in the traditional way economists assume. This is in fact what Greg does in his paper on dynamic scoring. The assumption that underlies this case is that the people are no poorer than they were before and hence there is no reason to expect income effects to be important - more on this later.

3) The government spends on some program for the poor, which is on net basically like spending it on everyone but then increasing taxes on the rich again to take the money right back. This makes the substitution effect of a tax worse.

There are at least two more cases, however, that I think or relevant.

4) The government uses the money to reduce the budget deficit. The argument today is not over whether or not the government spending grows the economy. Most people even on the left would probably concede that government spending on the poor and elderly doesn't grow the economy.

The often debated question, however, is whether deficit financed tax cuts can grow the economy. So here we have to assume that the revenue goes to national savings. At this point the effects become complex. However, I think it is easiest to think about this as case (1) plus some sort of subsidy on investment. Then think about the tax interaction effects between the investment subsidy and labor supply.

In the real world this means that perhaps mortgage rates are lower. The question is does this make you want to make more, since now you have a shot at a really nice house or less because you current house is cheaper?

However, I think it is difficult to weave a tale where a) overall investment is not increased b) the incentive to work today is not increased.

5) The government spends the money on services or transfer payments that are for the most part unrelated to income. Here I am think of Medicare or Social Security or K12 for the states. On net there is zero income effect but this does not mean that the net income effect is zero.

Huh?

For ubernerds this means that the income effect is not homomorphic with respect to summation. For regular nerds it means that the "income effect function" is non-linear. For everyone else it means this:

The top 20% of the income distribution earns a lot more than the bottom twenty percent. So suppose you took away an equal percentage of income from everyone. Then you gave everyone back exactly the same amount of money.

For the rich this would mean that on net they are loosing money. For the poor it means on net they are gaining money. This means that the income effect will encourage the rich to work more to make up for there lost earnings and the poor to work less.

Because, however, the poor earn less than the rich, each hour of work the poor give up is worth less than each hour the rich work more. Thus even though on net the population got back what it paid in taxes the result is that there is less low income work but more high income work and a richer economy.

The take home of all of this nerdiness being that income effects will tend to be important when looking at changes in tax rates.