If you talk to an economist about taxes, then he will probably frustrate you by criticizing the method of collection and the rate of taxation. But this is not because economists are free market ideologues. They’ll criticize your tax dollar spending too. Economics is the natural study for people who are natural contrarians. So what about the economists who dissent from the popular economist opinions? Today, I am one of those crazy dissenters. It is a popular position among economists to advocate for low taxes on capital. Better yet, some argue for no taxes on capital. My opinion is not that these economists are wrong. My opinion is simply a reminder of the affects of such policies and the costs that lie therein.
First, let’s make clear what we mean by capital. Capital is something which helps to produce additional output which is not a person’s labor. The point of this post is does not hinge on philosophical arguments about all consumption being investments in labor. Capital is often synonymous with investment, savings, and future consumption. Capital is most simply defined as “product of this period which is not consumed this period”. We don’t need to consider risk, depreciation, or inflation. They only make the picture more complicated than it needs to be for our current purpose. So all output/product can be described as being consumed now, or being for consumption in the future. This future consumption and foregone present consumption is capital.
When people talk about capital gains taxes, they are talking about taxing the nominal profits from savings and investment. For example, if I save my money in a security which pays 5% interest after one year, then my capital gain at maturity is that 5%. When capital gains are taxed, it is this little 5% which is taxed. So if the capital gains tax is 20%, then my real take home gain is 4% instead of 5% [5-(5*0.2)=4]. For the tax savvy, we’ll assume that there is single marginal rate for all capital gains. Tax on labor works a little differently. Rather than the tax being levied on the nominal return of the investment, the tax is levied on the up front cost. So if labor costs $100 to employ and the tax on labor is, then it doesn’t matter what the nominal return on labor is, the cost of labor is what gets taxed. So it does us absolutely no good to compare the marginal tax rate of capital gains and the tax on labor wages. One gets levied on the returns and the other gets levied on the up front cost. We can’t really talk about effective tax rates, because we have no idea what the cost is to a worker to maintain his labor – and therefore we don’t know what his gains are from that labor. It is an exercise in futility to compare the two tax rates.
For the implications of the taxes, we can look at the world from the point of view of the entrepreneur with increasing marginal costs. She can invest in labor or in capital. Her mix of investment among the two will depend upon the marginal benefit and marginal cost of each. If both a unit of labor and a unit of capital produce $150, then the costs of each are what drive the entrepreneur’s mix decision. If the cost of capital is $120 and the cost of labor is $100, then the entrepreneur will make more profit from producing with labor ($50 profit > $30 profit). The entrepreneur will continue to invest in labor until the cost of employing one more laborer is greater than the cost of capital. Then she will adjust her mix and begin using capital rather than labor until the relative costs switch again. This happens until the profit from investing in either capital or labor is zero.
Before, we were talking about the returns to the laborer. With the entrepreneur in mind, now we can compare the different taxes on labor and capital because we know the cost and gains from each.
We can compare the different taxes on labor costs and capital gains (see above). If both capital gains and labor are taxed at a rate of 20%, then the entrepreneur will invest in labor and capital until their costs reach $125 and $150. That is to say, the entrepreneur invests in both until the real marginal return, after taxes, is zero. In “real life” the entrepreneur pays multiple taxes that include taxes on costs and gains for both labor and capital. The above chart is simplification. What we can say is this. There is no reason for the tax on gains and the tax on costs to be identical. We could tax them both the same way and at the same rates, but why would we want to do that? The result would not be an ‘equally’ balanced mix of capital and labor because the marginal costs of the two methods of production increase at differing rates. Therefore, an equal tax would not result in equal quantities of labor and capital being employed.
There is no reason at all for a policy maker to aim at a specific ratio of capital and labor. What would that even mean? How much capital is equal to one unit of labor? The entrepreneur doesn’t care what the final mix of labor and capital is. The ratio is simply an accidental statistic of her profit maximizing behavior.
In a totally free market, the ratio would reflect the productivity of labor in contrast to capital. The return on either is what governs that ratio. If both capital and labor produce the same revenues, then it is their cost which governs the return and the ratio. The cost is determined by real world costs of production and the opportunity cost. For capital, the cost of production would be the cost of getting ore out of the ground, harvesting crops, smelting iron, assembling machines, transportation, etc. The opportunity cost is the alternative use of that capital for some other productive purpose. The purpose which produces higher returns is the one which is able to outbid those other purposes. In this way, capital is used for its highest productive purpose. The same is true of labor. The costs include food, sleep, clothes, leisure, medicine, etc. The opportunity cost is the highest valued alternative to that labor. The entrepreneur can get out bid by other purposes for the marginal unit of labor. Maybe the laborer prefers leisure instead of the $125 that he would be paid. Or maybe he can get a higher price elsewhere.
Relative prices are what matter. We don’t know what the total mix of labor and capital should be. But we do know what the value of the capital and labor stock are in terms of their costs. And we know that a change in the tax rates will change the mix of their use in production. Although a tax on gains doesn’t affect the ultimate cost of the last unit of capital, it does affect the quantity at which that cost exists (because all real returns of the lower marginal quantities are positive). The tax on labor affects both the cost of the last marginal unit and the quantity at which it occurs. So if the tax on one or the other method of production decreases, then the quantity in production should increase. This is because the relative price will have changed. And when the relative prices changes, the entrepreneur switches methods of production.
Some economists see capital accumulation as the road to growth and increased future consumption. The Solow growth model, and others, contradict this view. It turns out that capital accumulation accounts for very little of the differences in output across times and countries. Increased capital investment does increase output – but the effects are relatively tiny in contrast to the effects of increased knowledge and technology.
We’ve already discussed what the taxes would be in the most efficient economy, above, when the market is totally free from government interference. For some reason, I don’t think that most policy makers care whether labor and capital are working in their most productive capacity.
From a governing standpoint, I see the unemployment rate as the most salient topic on the minds of voters, in regards to the economy. That is, if I were a politician, I would love to take credit for a falling unemployment rate. The only way that the use of labor will increase, short of public employment, is to decrease its price relative to capital so that businesses and entrepreneurs voluntarily change their production mix toward labor in order to maximize their own profit. A relative price fall for labor can be achieved in two ways: increase the cost of capital or decrease the cost of labor. Policy can’t very well decrease the real costs of production – that requires innovation. But the government can affect the effective costs to businesses by changing the tax rates.
An increase in the capital tax would increase the proportion of labor used in production – but this is not the same as saying more absolute labor is used. It would more likely be that less capital simply reduces the total output (due to the newly unproductive capital) and keeps the costs of labor constant (also, taxing capital reduces real return to capital and therefore reduces the price. With mark to market accounting, a sharp capital tax increase would be an imprudent assault on leveraged assets and cause more financial market uncertainty. Gradual increases would avoid the financial shock – but it would also fail to increase employment quickly enough for a politician to claim credit in time for re-election). A decrease in the tax on labor will increase total output, increase the proportion of labor in the production mix, and increase the total labor employed. Including the empirical income and substitution effects only strengthen this logic.
So here’s my dissenting opinion: Decreasing the capital gains tax, although increasing efficiency, should not be our primary economic focus. Yes, there would be a positive income effect, but the substitution effect would be negative on the quantity of labor. Why not choose a policy which would have positive income and substitution effects? This would be a fall in the after tax relative cost of labor. One might argue that the government can’t afford to decrease revenues through tax decreases. That is a completely separate argument. It is not irrelevant. But it has nothing to do with decreasing the unemployment rate.
The growth effects of increased capital investment are small. Increasing the mix of capital in production will not “grow us out of our problems”. If unemployment is focal, then we should decrease the effective cost of labor. Policy makers should decrease the tax on labor rather than capital.