Dividend tax delusions

Posted: Dec 22, 2005 12:05 AM

When Congress reconvenes for the New Year, few chores will be as urgent as reconciling differences between the House and Senate over keeping the tax on stock dividends as low as the tax on capital gains after 2008, preferably indefinitely (as Ed Prescott argued in The Wall Street Journal). Unfortunately, many reporters and pundits are trying to obscure all the key issues by making a big deal out of one tiny tidbit of substandard economics.

 Washington Post columnist Robert Samuelson wrote, "A new study by staff economists at the Federal Reserve finds little independent effect of the dividend tax cut on stock prices." Gene Sperling, former director of the National Economic Council under President Clinton, wrote that "economists at the Federal Reserve found no evidence that the dividend tax cut raised stock market prices as a whole."

 But this study by Fed staffers found nothing of the sort. It only claimed to find that news about the dividend tax cut was not instantly reflected in much higher stock prices during two amazingly brief periods before the law was passed. That is a far less interesting topic, and one that depends entirely on the authors' immature opinion that investors remain clueless about what is going on in Washington until the news finally appears in print.

 This study is not really new (it is dated Sept. 26), just newly leaked. It was prepared for one of the Fed's conferences by Gene Amromin of the Chicago Fed and Washington Fed staff economists Paul Harrison, Nellie Liang and Steve Sharpe. It is a "preliminary draft," and "the views expressed are those of the authors." This is no more an official Fed publication than a paper I presented at the Fed in 1989.

 The authors claim the market suddenly became aware that the tax rate on dividends might be cut on Jan. 3, 2003, because that is when some big newspaper first wrote about it. Yet I had already published three columns about plans to cut the dividend tax in 2002 -- the first appearing in September of that year and the last at year-end in The Washington Times and New York Post. "Dividend Taxes Are Going Down" was the unambiguous title of my fourth column in early February.

 I did not argue, by the way, that boosting stock prices was an important reason for having the same tax rate on dividends and capital gains (and perhaps estates). That's a red herring.

 The Fed staffers imagine investors had no idea what might happen to dividend taxes between Jan. 9 and May 14. "The issue became a prime news story again in early May," they say, "following reports that House Republican leaders had finally agreed on a specific tax package containing a provision to lower the top tax rate on corporate dividends to 15 percent." That was news?

  On April 16, I spoke at the U.C. Santa Barbara Economic Forecast Project, leaving no doubt that Congress would reduce the tax on dividends and capital gains -- probably to 18 percent. The 15 percent rate was unexpected in March and April, but those of us working with House Ways and Means Committee Chairman Bill Thomas knew he would succeed in getting dividends and capital gains taxed at the same low rate.

 The study's dubious dates are critical because its myopic authors imagine only two brief "windows" during which news about lower tax rates on dividends and capital gains could possibly have affected stock prices. The first was the "Jan. 3-9 window," which involved four trading days. The other was the "May 14-28 window" -- just 11 trading days up to and including the day the bill was signed.

 They conclude that "stock market gains during the two tax-cut event windows (were) relatively modest," although opening the second window six days sooner would double those gains. What about improvements in the economy and stock market after tax rates were actually reduced? Their approach is: Don't ask, and don't tell.

 In a November paper for the National Bureau of Economic Research (NBER), Clemens Sialm of the University of Michigan Business School took a somewhat longer view. Using data from 1917 to 2004, he demonstrates that "stock valuations tend to be relatively low when tax burdens are relatively high."

 The Fed economists theorize that cutting tax rates for U.S. investors should have boosted U.S. stocks more than European stocks, over those four-to-11-day windows, because "U.S. investors hold a relatively small fraction of foreign equities." What matters for asset pricing is the marginal trader, not average holdings. If marginal U.S. investors bid for Exxon-Mobil, BP and Shell stock because of their enhanced prospective after-tax dividends, then both U.S. and European stock markets would be affected.

 The paper goes on to say "only" two-thirds of firms that began paying dividends for the first time in the second half of 2003 also increased total payouts, because the other third cut back on stock buybacks. Yet inducing firms to switch from stock buybacks to dividend payouts was a very desirable reduction of a tax-induced distortion, and one that I and others predicted.

 The authors mention only one academic paper on the subject they purport to study. They may be unfamiliar with the literature, or may prefer to keep their readers in the dark. An unmentioned NBER paper by James Poterba, for example, estimated the lower tax "could ultimately increase dividends by almost 20 percent."

 Poterba did not say that would happen overnight, or that all stockholders would instantly capitalize all future benefits within four to 11 days. Another NBER paper by Raj Chetty and Emmanuel Saez showed that "the surge in dividend payments after the 2003 reform is unprecedented in recent years."

 Like Sialm, these are serious scholars who are well known for doing serious research. Don't expect to see the press pay them any attention unless and until they dream up something nice to say about higher tax rates on dividends and capital gains.