The debate over double taxation of dividends is rapidly
narrowing to a few minor details. There have been the usual dubious
complaints that a lower tax on dividends would result in higher interest
rates on municipal bonds. And there have been the usual budgetary anxieties
about estimated revenue losses. In the end, however, something significant
is going to happen.
Faced with senior backlash, a dwindling number of national
legislators appear willing to stand up for the sheer principle of taxing
dividends twice. On the other hand, state governors' efforts to steer
overtaxed investors toward municipal bonds have been a familiar but annoying
distraction. At a recent gathering of congressional Republicans, I was
asked, "Since there are always winners and losers from any tax change, who
would lose from eliminating the double taxation of dividends?" That
congressman echoed the common hunch that interest rates on municipal bonds
would have to go up to compete with single-taxed dividends. When Fed
Chairman Alan Greenspan testified before the Senate budget committee, he too
was asked what to do about this alleged effect on "munis".
Just as bond yields contain an "inflation premium" to compensate
for inflation, taxable bonds also contain a "tax premium" to compensate for
high tax rates. That is why tax-free bonds pay a lower rate of interest than
taxable bonds of comparable risk. What happens when high tax rates are
reduced, however, is not that the yield on tax-exempt bonds goes up, but
that the yield on taxable bonds goes down.
Issuers of tax-free bonds have lobbied against every reduction
of every tax rate since at least 1981. But look what happened when the top
tax rate dropped from 50 percent in 1986 to 37.5 percent in 1987 and 28
percent in 1988. The yield on 30-year Treasury bonds fell 3 percentage
points in 1986, from 10.8 percent to 7.8 percent, while the yield on the S&P
index of high-grade municipal bonds fell by "only" 1.8 percentage points.
What matters to states and cities is not the spread between their rates and
Treasuries but the absolute level they have to pay. And shrinking the tax
premium shrinks interest rates generally.
Since the tax hike of 1993, ironically, the spread between
taxable and tax-exempt bonds narrowed and nearly vanished, dropping to only
three-tenths of a percent from 1998 to 2001. The greater risk of default on
state and local securities has swamped any tax advantage.
Bringing the top tax back down to 35 percent will reduce the tax
premium in taxable bonds, as in 1986, not raise the yield on tax-exempts. As
for dividends, the public made half of them tax-exempt already by keeping
them inside pension funds and foundations. Making the other half tax-exempt
will increase the flow of dividends in taxable accounts, but that will not
affect municipal bonds yield a bit.
The only other politically potent objection to the president's
plan is the estimated revenue involved. Those estimates are static, and
presidential adviser Glenn Hubbard figures improved investment could pare
the actual cost by 40 percent. But that still leaves a 10-year tab above
$200 billion. Budgetary pressures may be pushing even taxpayer-friendly
congressmen to consider scaling back the dividend plan to conserve some
tax-cutting ammunition for other chores, like killing the alternative
There are four alternative dividend tax plans floating about,
whose effectiveness cannot be judged by the amount of revenue they
supposedly lose. Yet that may be a good place to start.
In a recent column, I mistakenly implied that a plan proposed by
Newsweek columnist Allan Sloan had the blessing of the Brookings
Institution, simply because he used their estimates. In fact, Brookings has
taken no stand in favor of any such proposal. And I, too, will use some
figures from the Tax Policy Center, a co-venture between the Urban Institute
and Brookings Institution.
The Center estimates that taxing dividends at the same rates
that we tax long-term capital gains (10 percent to 20 percent) would cost
the IRS $78 billion over 10 years. Cutting dividend tax rates by half (5
percent to 17.5 percent) costs twice as much -- $168 billion. Both of those
estimates include the president's clever basis adjustment to eliminate
capital gains taxes to the extent they reflect retained earnings. Although
the 50 percent deduction is a larger tax cut than capping the tax at 20
percent, I nonetheless favor the latter plan because it taxes dividends the
same way we tax capital gains and I believe we should tax income from
investments at a single, flat rate. I have explained in previous columns why
I believe either of these two plans would be static "revenue neutral," with
little or no revenue loss over time.
Another worthy alternative would be to allow corporations to
deduct dividend payments. Unfortunately, Congressional stinginess is an even
bigger problem here than with the president's plan. Steve Entin of The
Institute for Research on the Economics of Taxation notes that corporations
paid some $376 billion in dividends in 2000, but individuals reported only
$145 billion of dividend income, some of which was interest. Full deduction
at the corporate level clearly entails much bigger revenue losses than an
equivalent deduction for individuals because many dividends end up in
tax-exempt pensions and organizations.
One genuinely meaningless gesture is the idea of excluding, say,
the first $2,000 of yearly dividends from taxation, while taxing the next
dollar at ordinary income tax rates. That scheme would lose about $60
billion over a decade without the slightest beneficial impact on marginal
decisions. As soon as dividends reached the limit, investors would prefer
capital gains to larger dividend payments and might even sell
dividend-paying shares to avoid the nasty notch in their marginal tax.
Excluding the first few bucks of dividends is a plan that maximizes the
revenue loss while minimizing the economic benefit. It is no better than
doing nothing, maybe worse.
Something much better is going to pass. Meanwhile, standing in
the way will prove politically hazardous to legislators from either House
and either Party.
I have seen some journalistic speculations that Congress is
going to do little or nothing this year about double taxation of dividends.
If anyone has the courage to offer a wager on that, take the bet.