Briefing the President

"Arguing in My Spare Time" No. 3.03

by Arnold Kling

January 13, 2000

May not be redistributed commercially without the author's permission.


A new nonprofit organization called the Internet Policy Institute has been formed. Its main project is called "Briefing the President," which is described on its web site ( as "a series of policy memoranda designed to identify and explore the fundamental issues that will affect the development and use of the Internet."

In my opinion, there is one policy issue that is likely to overshadow all others that the Internet might pose for the next President. The issue will be how to ameliorate a long, deep economic slump that results from a collapse of the Internet bubble.

Here are my opinions concerning the probabilities of various events:

--The probability of a stock market crash over the next four years, in which at least 25 percent of all stock market wealth is wiped out, is 90 percent.

--If such a crash occurs, the probability of a serious economic slump is 80 percent.

--Given an economic crisis, the probability that the President's economic advisers will know what to do about it is not better than 50 percent.

--Given that they know what to do about a crisis, the probability that their advice will be followed is less than 25 percent.

People who know me are aware that I am pessimistic by nature. These probabilities must sound extremely gloomy. However, let me try to make the case that they are reasonable, and perhaps optimistic.

First, consider the stock market. In the January 12 edition of the Washington Post, op-ed columnist Robert J. Samuelson cited estimates that the aggregate price-earnings ratio of the Nasdaq was about 200 as of the end of 1999. I have not confirmed this estimate, and if it happens to be very far off then the rest of this essay will have to be discounted accordingly. However, it comes from a reasonable source, Brian Rauscher of Morgan Stanley Dean Witter.

Let us place an aggregate price-earnings ratio of 200 in perspective.

1. Last September, James Glassman and Kevin Hassett published a book called "Dow 36,000" which purported to justify a higher market price-earnings ratio than has been observed in the past. They claim that investors no longer are demanding the high risk premium that in the past held down the price-earnings ratio. They argued that the price-earnings ratio should be 100. Since the Dow was at 9000 at a price earnings ratio of 25, they concluded that the Dow should be at 36,000. However, they only could justify a price-earnings ratio of 100. Using their logic, the Nasdaq needs to fall 50 percent to be valued correctly.

2. Most economists believe that Glassman and Hassett commit fundamental errors of arithmetic and accounting. Even if we assume that investors in stocks require no risk premium, economists would argue that the market price-earnings ratio ought to be the inverse of the real interest rate (the interest rate after inflation is subtracted). Today, the interest rate on the Treasury's inflation-indexed bonds is about 4 percent. That means that the risk-neutral price-earnings ratio ought to be 25 for the market as a whole. For stocks whose earnings are going to grow faster than the economy, the p-e ratio can be higher. For stocks whose earnings are going to grow more slowly than the economy, the p-e ratio ought to be lower. If 25 is the right p-e ratio for the Nasdaq, then the Nasdaq needs to fall 87.5 percent to be valued correctly.

3. Until 1996, the average price-earnings ratio for the stock market was about 14. If that is the appropriate ratio for the Nasdaq, then it needs to fall 93 percent to be valued correctly.

In conclusion, an 85 percent drop in the value of the Nasdaq seems entirely plausible. How would this drop in paper wealth affect the real economy?

According to estimates gived in Samuelson's column, the market value of the Nasdaq was $5.2 trillion, out of a total market value of $15.8 trillion for all stocks. Thus, an 85 percent drop in the value of the Nasdaq would reduce stock market wealth by about $4.4 trillion, or over 25 percent of all stock market wealth.

The economic links between stock market wealth and economic performance are somewhat tenuous empirically. We have two important historical examples in which major market meltdowns were followed by slumps--the 1930's in the United States and the 1990's in Japan. However, this does not prove that stock market crashes cause depressions.

There are a number of plausible linkages between wealth and economic activity. Moreover, because stock market wealth today is so high (it is much higher relative to the size of the economy than it was in 1929), even if these linkages are weak, the overall effect could be devastating.

Overall GDP in current dollars is running at a $9.3 trillion annual rate. About two-thirds of this is personal consumption expenditures. Estimates are that consumers spend between 1 and 5 percent of wealth each year. Therefore, a drop in wealth of $4 trillion would lead to a drop in consumption of between $40 and $200 billion.

Business Fixed Investment is just over $1 trillion at an annual rate. A collapse of stock prices would cause at least some of this investment to disappear. If investment were to decline by 10 percent, that would be $100 billion.

One of the most dangerous components of the economy from the standpoint of macroeconomic stabilization is the state and local government sector. This is a very large sector of the economy (almost twice the size of the Federal government in terms of GDP expenditures, although smaller in terms of transfer payments), and it is highly procyclical. When the economy is strong, this sector increases spending, and when it is weak it cuts back. State and local spending is just under $1 trillion at an annual rate. If capital losses in the stock market reduce tax collections by 5 percent, then spending will fall by almost that much, or $40 billion.

It is not difficult, therefore, to envision a drop in GDP of $200 to $400 billion, arising directly from a crash in the Nasdaq market. This would be a decline of 2 to 4 percent, which likely would be extended by multiplier effects. That is, as spending declines, employment declines, spending declines more, etc.

Beyond the straightforward linkages, one should be concerned with the impact of a crash on the psychology of consumers and businesses. For example, the personal savings rate, which averaged 6 percent for most of the post-World War II period, has fallen close to zero recently. What if the savings rate rose back to 6 percent? In that case, consumer spending would fall even farther.

Business investment has consisted of spending on technology and the Internet. What if a stock market crash leads to a general reconsideration of technology spending?

Another psychological issue is the effect of a crash of the Nasdaq and of an economic downturn on the New York Stock Exchange, where stocks that represent most stock market wealth are traded. So far, we have been assuming that these stocks escape unscathed. However, suppose that the prices of non-Nasdaq stocks fall by 25 percent. This would wipe out another $2.5 trillion in wealth, and cause nearly as much damage as the 85 percent decline on the Nasdaq that is assumed to be the trigger.

In short, it seems very plausible that a stock market crash would cause a very deep recession, with real GDP declining at a 5 percent rate or higher for two years or more. Could economic policy overcome this?

I believe that most economists continue to be Keynesians. That is, we believe that government spending and tax cuts, accompanied by stimulative monetary policy, would help to alleviate a slump. However, there is some probability that the economists who advise the next President will not be Keynesians. Or, if they are Keynesians, there is some probability that we are wrong, and that textbook fiscal and monetary stimulus will not work.

Even if the Keynesian prescription is correct, it may be difficult to carry out politically. For example, in an economic crisis the dollar could be falling sharply as foreigners attempt to sell their U.S. assets, and this could put pressure on the Fed to tighten monetary policy. For fiscal policy, the most reliable stimulus would be government spending, but the political pressure may be to use tax cuts and increases in transfer payments, which might go mostly into savings. Also, with the Federal deficit already increasing due to the economic slowdown, it may be impossible to vote for a further fiscal stimulus of more than $100 billion or so. In that case, the stimulus probably would not be sufficient to offset the collapse in demand coming from the private sector.

In the 1930's in the U.S. and in the 1990's in Japan, policy was not particularly successful. Much political energy was expended on "structural reform." In the U.S., this included deposit insurance and securities regulation. In Japan, various forms of financial and business reforms have been debated and, in some cases, adopted. If the Nasdaq crashes, we can expect to see similar energy focused on electronic trading, investor suitability, accounting rules, and other issues which at best will amount to closing the barn door after the horse has gone.

Economic depressions, like personal depressions, are frustrating, frightening events. I certainly hope that we can avoid one. However, it seems to me that such a scenario is at least plausible, and indeed highly probable.