Marginal Revolution

Thought for the day

A non-nuclear method of increasing the national self-esteem of the large (either population-wise or economically) non-nuclear countries could do a lot of good.

That is from Matt Yglesias.

How good are auto hybrids for the environment?

I am not qualified to judge the science, but this article increased my skepticism.  Thanks to Orin Kerr for the pointer.

Paying for Performance

The field of education is littered with reforms designed to increase student performance - everything from the "new math," to more teachers to better pay.  Yet the most obvious reform of all has hardly been tried - pay the students to learn.  That's the simple idea of an impressive young economist, Roland Fryer (earlier I posted on Fryer's controversial work with Steven Levitt on the causes and consequences of distinctively black names).

Fryer was here on Monday and he told me of a large scale experiment he is running in 24 of the poorest performing New York schools.  Every three weeks students are tested and if they improve they are paid on the order of $20.  Control groups are also tested.  Early results are very encouraging.  No other reform has anywhere near the bang for the buck as paying the students.

As Fryer said to me, 'for years white parents have been giving their kids money for As, now we are trying the same system for black kids.'   

Awesome efficiency!

This new vehicle is slow but it gets 4.7 million miles to the gallon.

Will reforming social security yield an equity premium?

U.S. stocks yield six to eight percent on average, while T-Bills yield just a bit over one percent.  Might we reap higher social returns by investing a social security trust fund in equities?  That of course is one motivation behind many plans for social security reform.  Even if our government is borrowing more, the higher equity returns might more than compensate.

Under one common view, stocks yield more because they are riskier.  Yet stocks outperform bonds for any thirty year period in U.S. history.  So perhaps a sufficiently long-lived institution can ignore the short-term risks.  The return differential could instead exist because a) equity markets are not perfectly liquid, and b) many of us are irrational fraidy-cats, or have excessively short time horizons.

To make the case as strong as possible, let us put risk aside.  But even then the measured returns differential does not measure the gain from investing social security funds in equities:

1. Once the plan is announced, the price of equities will rise, eliminating some or all of these returns.  Current equity holders will be the beneficiaries on a once-and-for-all basis.  Of course if equity markets are not perfectly liquid -- one of the assumptions -- the current equity holders won't capture all of these gains.

2. Higher stock returns, driven by a change in liquidity, might just redistribute wealth rather than creating more wealth.  The real question concerns not measured nominal returns but rather the real output of goods and services.  Say that one group of people suddenly earned far more on their investments.  Without a corresponding rise in productivity, the economy as a whole does not have greater opportunities.  We have simply divvied up the pie in a new way.

3. So what is the real gain?  When stock prices rise, it becomes more profitable for a company to issue new stock shares.  This will increase investment and eventually output.  But the magnitude of this effect is not given by the measured differential in stock and bond returns.  The real questions concern a) the elasticity of investment with respect to equity values, and b) the value of the next marginal investment to be made; we should not confuse average returns with marginal returns.   

Note that companies do not usually prefer to finance investments from new equity issues.  Retained earnings and debt are more popular.  The value of a company typically falls on the stock market when new equity issues come forth.  This suggests that markets do not think so much of new investments financed by new equity issues.

The bottom line: There may be some equity premium to be captured, but it is much less than the measured return differential between stocks and bonds.  Most likely, the equity premium is not a strong argument for investing the trust fund in equity or social security privatization; Alex agrees.  The more relevant argument is simply that savings will increase.