Tyler Cowen's Policy Implications of "Capital-Biased" Technological Change

1/05/2013 05:12:00 PM
Perhaps we should start calling it "capital-intensive" technology, because it has absolutely nothing to do with change per se. The point is that Krugman is upset--understandably--about the inequality caused by the fact that our current production technology is highly capital-intensive. So with that in mind, lets look at Cowen's policy recomendations:
  1. Try not to tax capital, unless they are substitutes
  2. Weaken intellectual property rights, because doing so is like a tax on capital
  3. Increase inflation, to force firms to hike wages more.
  4. What Matt Yeglesias said
  5. Drill, baby, drill
  6. No minimum wage
  7. Eliminate mandated employment benefits
  8. Convert Medicare into cash-equivalent social security.
I will address these in order:
First, on taxing capital. Cowen is right that Krugman should be more worried about the efficiency implications of taxing capital. Most optimal tax models suggest that revenue should be raised primarily from income, and that capital should be tax at near-zero rates. The way I explain it to people, when capital and labor are complements, every dollar raised from taxing capital has a greater than one for one impact on the revenue potential from taxing income. That is, a capital tax actually lowers, not increases, the maximum amount of redistribution you can do. Sure, there are benefits in the fact that capital and labor are highly correlated with the people we want to take from and give to, respectively, but this is an argument for a "one-and-done" redistribution of the existing capital stock, not for taxing capital gains. As for Cowen's comments about taxing substitutes, strictly speaking this is also inefficient, its just that if they are substitutes, we can tax the capital without hurting the incomes of those who don't own that capital. But then he names his examples--software is a substitute?--which generally have two problems: one, they are probably actually complements not substitutes, and second, they are owned by many of the people we want to redistribute to, rather than from.

On property rights. This point directly contradicts the first point. Now, I believe that in many areas current patent and copyright protections--which are government subsidies to owners of capital--are inefficiently large. So I agree with the point that they should be reduced. But there's something a little loony about his reasoning that relaxing these protections is good because its like a tax on capital--we've already established that taxing capital is bad. I consider this point a non-sequiter. Sure, it is good policy, but it really has little to do with redistribution from owners of capital to labor. Keep in mind that when a drug patent expires, its the capital-intensive rival pharma companies that profit.

On inflation. I had a little trouble seeing his reasoning on this one. If we think of Calvo pricing where capital prices can adjust more frequently than wages, that implies that the welfare losses from deviations between the optimal and nominal wage is much higher than the welfare loss from the deviation between optimal and nominal capital rents. I don't know what exactly kind of " downward pressure on wage shares" he has in mind, but his seems to actually be an argument for less inflation. Perhaps I am misreading him. At anyrate, nominal rigidities aside, there is a well-known effect whereby surprise inflation redistributes from savers to borrowers. So we can achieve some redistribution by enacting surprise inflation spurts. But there are two problems. First, this is just a form of taxing capital, so it contradicts the first point. Second, it requires increasingly erratic hyperinflation to keep going for long. This is clearly a loser idea when it comes to trying to manage the distribution of income over the long run.

On land rent. Ok, I'm all for rent liberalization policies. But, I don't get this point. Why not just tax land and pay the money to the renters directly? There needn't be any efficiency losses from taxing that kind of capital, because it can neither be created nor destroyed (ok, I need to qualify that remark). Again, the points about liberalization are certainly correct, but not specifically relevant to the problem at hand. This is mostly another non-sequiter.

On resource owners. Again, non-sequiter. Maybe we are over-regulating natural resources--that has nothing to do with distribution of income. The real proposal here would be to tax the mines and pay the money to the miners.

Cut the minimum wage!?! Actually, that's not as crazy as it sounds, once you note two things: the people who would be hired for less than minimum wage are the people we most want to give money to, so that checks out. Also, minimum wage jobs are the ones that are the most labor-intensive, meaning that this would increase labor's share of income, on average. I will dock points for solvency, however. Very few people actually make minimum wage, and under current law there are actually far more people exempt from it than affected by it.

Mandated Benefits programs. This is mostly the same logic of minimum wage. We'd be better off taxing the rich and giving the poor those benefits rather than have them get neither the benefits nor a job. But, this doesn't really go far enough--we should be mandating that employers can't offer benefits like health insurance, which would force insurers to compete for households, and better incentivize preventive medicine and long-term care.

Medicare. I think Cowen's point is completely negated by the fact that, to provide the same level of wellbeing, we'd have to payout two dollars extra in social security for every dollar in medicare we eliminate. I just don't think that the option value of that money could possibly be that high. Also, Cowen's comments are, again, a complete non-sequiter, since that is not redistributive.

Those are some of my thoughts on Cowen's proposals, at anyrate.