Lower Tax Rates Spurred Dividend Growth

After a continuous decline in dividend payments over more than two decades, total regular dividends have grown by nearly 20 percent since the beginning of 2003 -- precisely the point at which the lower tax rate was proposed and subsequently applied retroactively.

In May 2003, President George W. Bush signed into law the Jobs and Growth Tax Relief Reconciliation Act. Among its main provisions, the Act cut the individual tax on dividend income to 15 percent; previously, dividend income was taxed according to the regular income tax schedule, with a maximum rate of 35 percent. In their recent paper, Do Dividend Payments Respond to Taxes? Preliminary Evidence from the 2003 Dividend Tax Cut (NBER Working Paper No. 10572), authors Raj Chetty and Emmanuel Saez assess whether the tax reform induced companies to pay out more dividends.

Chetty and Saez data come from the Center for Research in Security Prices (the CRSP tracks dividend, stock price, and share information for NYSE, AMEX, and NASDAQ companies). The authors examine quarterly data on corporate dividend payments spanning 1980 to the first quarter of 2004. Their core sample includes 431,379 firm-quarter observations, although they also examine a "selected sample" of 180,170 firm-quarter observations, taking into account only the firms listed in the 2004 CRSP.

The authors find that, after a continuous decline in dividend payments over more than two decades, total regular dividends have grown by nearly 20 percent since the beginning of 2003 -- precisely the point at which the lower tax rate was proposed and subsequently applied retroactively. However, data on aggregate dividends are highly volatile and often driven by the behavior of just a handful of firms. Indeed, the authors find more than 100 cases when a lone firm changes the sample's total dividend payments by more than 5 percent. To deal with this "extreme values" problem, Chetty and Saez analyze three measures of dividend payments that are less sensitive to outliers: 1) number of initiations and terminations of regular dividend payments; 2) number of increases in payment amounts by firms already paying out dividends; and 3) number of "special dividends," that is, dividends intended to be one-time distributions.

In 1980, the percentage of firms paying monthly, quarterly, semi-annual, or annual dividends stood at 60 percent. By the fourth quarter of 2002, this percentage had declined to 20 percent, only to rebound to nearly 25 percent in 2003. Of the 3,813 firms in the sample, 113 began paying regular dividends in 2003 -- a large increase from the average of 22 new dividend payers in prior years. "The fact that the decline in the fraction of dividend payers stops precisely in 2003 constitutes strong evidence that the 2003 tax reform induced more firms to start paying regular dividends," the authors explain. Notably, the increase in dividend initiations occurred across firms of all sizes and industries. This is true even after the authors control for levels and lags of profits, assets and cash holdings, and firm age.

Among firms already paying dividends, Chetty and Saez find that an average of 65 firms increased their dividend payments by 20 percent or more in each of the quarters that followed the tax reform enactment -- more than double the average of 31.7 firms in earlier years. The authors also find that the number of firms paying out special, one-time dividends rose substantially immediately after the tax reform, from an average of 11 firms per quarter in 2002 to 28 firms in each of the quarters following the tax reform.

Ultimately, the authors explain, their data "strongly suggest that the 2003 tax reform induced a large, widespread set of firms to initiate regular dividend payments or to raise the payments they were already making." This, they explain, is "unprecedented in the record of publicly traded U.S. corporations in the last three decades." However, Chetty and Saez caution that "it remains to be known whether the 2003 tax [reform] spurred investment and business activity."

-- Carlos Lozada