Why does paid vacation exist?

Amidst the discussion by Ezra Klein, Matthew Yglesias, and Reihan Salam about paid vacation, I think it's useful to ask a more fundamental question: in countries like the United States that don't require it, why does paid vacation even exist?

This is not a trivial question. Employers could simply offer an equivalent amount of unpaid leave, providing an option to employees who prefer to take time off while rewarding workaholics for continuing to contribute value to the company. Why doesn't this happen? As far as I can tell, there are four plausible stories:

1. Per-hour productivity declines when stressed employees don't take any vacations. In an ideal world, this might be written directly into the contract ("if you work 52 weeks a year, you only get paid 50% for the 52nd week"), but for various reasons that isn't possible, and the second-best solution is to provide paid vacation.


2. Paid vacation resolves a bad signaling equilibrium, where everyone feels obligated to put in full hours to signal that they are hard-working and deserve advancement in the company. Offering paid time off convinces enough people to take vacations to break this equilibrium, making virtually everybody happier and better off. (Of course, this could also make the signaling problem worse. How do you signal that you're really hard-working and really devoted to your employer? By not taking paid leave!)


3. Paid leave makes leisure time more enjoyable, since you're not incessantly bothered by the fact that you're losing money by being away from work. Employees are willing to sacrifice wages and flexibility for this psychological comfort.


4. By encouraging every employee to be away from work for a brief period each year, paid leave allows a "dry run" to see how the workplace might function without that person, ensuring that it's not overdependent on the contributions of a single individual. This reduces the risk from employee attrition and diminishes the bargaining power of workers who might otherwise claim they were indispensable.

Of course, signaling acts as a powerful force against paid leave as well. Suppose that two otherwise identical companies, A and Z, differ in their vacation policy: company A offers higher wages but no paid vacation, while Z has slightly lower salaries and paid leave. In general, who will choose to work at company Z instead of A? People who like to take time off! To the extent that this is correlated with general laziness (not easily detectable in other ways at the hiring stage), it will make company Z's labor pool less effective, discouraging it from offering paid leave in the first place.

Define "economists"

Really, Greg Mankiw? He writes on whether the Bush tax cuts should be extended:

This seems to be the economic policy question of the hour. It might be worth recalling that last month, the Wall Street Journal polled economists about this question. Of those who expressed an opinion, here are the results:

• 6 percent said no, all the tax cuts should be allowed to expire,
• 24 percent said yes, but only for those making less than $250,000 a year,
• 70 percent said that all the tax cuts should be extended.

I'm not so much upset at the conclusion—temporarily extending tax cuts in the middle of a recession doesn't seem like a terrible idea, although the estate tax should be reinstated and some other rates should be allowed to climb upward. Rather, I'm upset that polls from the Wall Street Journal are presented as being surveys of "economists" when they're really surveys of a very particular subset of conservative private-sector economists, who believe things like:

Despite the continued challenging conditions, 30 out of 48 economists who answered the question said the economy didn't need any more fiscal or monetary stimulus. Six economists said more fiscal stimulus was necessary, while five want more monetary stimulus from the Federal Reserve and seven said that the economy could use both.

If the Wall Street Journal included academics (like Mankiw!) in the survey, I guarantee that this result would be different.

The actual effects of inflation

Via Matthew Yglesias, Josh Barro makes the case for indexing capital gains taxes:

But one likely barrier to a higher inflation target is a quirk of tax policy: non-indexation of capital taxation means that higher inflation causes a stealth rise in the real tax rate on capital gains and interest income...

Inflation has large impacts on the real capital gains tax rate. To understand why, consider an investor who buys an asset for $100 and earns a 40% real gain over seven years. In a world of 2% inflation, the investor’s nominal gain is $60.82, and he pays $9.12 in capital gains tax at a 15% nominal rate; this is a 19.9% effective tax rate on his real gain of $45.95 (equivalent to $40 at the start of the investment period).

But if the inflation rate had been 4%, the nominal gain would have grown to $84.23 and the real tax rate to 24.0% -- a 21% relative increase in the tax rate. This effect would be even larger if the real rate of return were lower, meaning the inflation penalty on capital gains grows when inflation is high or when asset markets are anemic.

I'm glad to see Josh Barro addressing this issue: although it's little understood, it's arguably the most important economic effect of inflation.

To see why, it's useful to look at the more commonly cited deleterious effects of inflation. These include:

1. It hurts people on fixed incomes.
2. More generally, it redistributes from lenders to borrowers.
3. It lessens the incentive to save.
4. It causes people to hold an inefficiently low level of real money balances.
5. Adjusting prices incurs a transactions cost—for instance, in updating a menu.

Upon closer examination, none of these except the last two are really problematic effects of "inflation." Rather, they arise from unexpected and variable inflation. If high inflation is expected, it will be built into pension plans and interest rates; assuming it stays on target, no economic damage is done. Of course, if inflation moves in an unanticipated direction, we see an arbitrary redistribution from lenders to borrowers (or vice versa), and the resulting uncertainty will lessen the incentive to save. But the damage here is from the unexpected component of inflation, not inflation itself.

The direct effect of inflation on money balances is also quite small. 4% inflation won't cause any serious inefficiency in the use of cash for transactions; we carry so little actual cash that the small amount eroded by inflation is almost irrelevant, and when we pay via checking accounts inflation only matters for the fraction of our deposits that banks keep as reserves. (The required reserves ratio is at most 10% in the US for demand deposits and zero for certificates of deposit, money market funds, etc.) Moreover, inflation has no clear impact on the real returns to investment: the future nominal returns from an equity investment scale up along with inflation, with zero net real effect. And most firms change prices often enough that a moderate level of inflation doesn't lead to any serious logistical inconvenience.

To sum up: the commonly cited downsides of inflation are not really problems with inflation at all, but rather a lack of precision in our expectations of inflation.* Other negative effects are close to zero in practical terms, at least when we restrict ourselves to inflation levels that are plausible in the United States. So what is so bad about inflation—assuming it doesn't unpredictably lurch around?

As far as I can tell, interaction with capital gains taxes is actually the key issue here. Starting in the late 1970s, Martin Feldstein convincingly argued that the era's high inflation imposed an extraordinary burden on capital investment—which, of course, is critical for economic growth. This wasn't because inflation was inherently problematic. It was simply a result of the distortions imposed by a tax system that failed to index properly.

Given the central role of inflation in discussing macroeconomic policy, it is astonishing that this effect is so little understood by the commentariat.

---

*As Arnold Kling points out, the distinction blurs a little if high inflation itself causes expectations to be less accurate.