The Wall Street Journal quotes Alan Auerbach--one floor down and eight offices north--on the likely effects of the dividend tax cut.
WSJ.com - Dividend Plan Straddles Debate Among Academics: ...Alan Auerbach began arguing a new theory. He said the traditional view of dividend taxes told only half of the story. Lowering dividend taxes did make it easier for companies to raise capital, increasing a company's incentive to invest in new projects, he said. But it also made shareholders more demanding. With lower dividend taxes, investors would expect executives to pay out more of their earnings in the form of dividends rather than pour that cash into new projects. "The company's incentive to invest for the future is offset by the investors' desire to have a dividend today," says Mr. Auerbach, now a professor at the University of California at Berkeley. "The two effects basically cancel." He concluded that overall spending by businesses would go nowhere with a cut in dividend taxes...
Alan is almost surely right not to expect anything in the way of higher corporate investment from the dividend tax cut. The tax cut makes shareholders richer--and so they want to spend more today. The tax cut also makes saving and investing for the future cheaper--and so they want to spend less today and save more. Given that both effects are first-order and that they oppose each other, the net effect is going to be small.
But there is one--very important--point that the Wall Street Journal missed. Alan is talking about the effects of a revenue-neutral shift away from the dividend tax to some other form of taxation. The Bush administration is talking about a revenue-losing deficit-expanding cut in taxes. The fact that the government is soaking up more of saving to finance its deficit puts additional downward pressure on investment, which makes it overwhelmingly likely that the total net effect will be a reduction in investment, and a slowdown in economic growth over the next decade.
Posted by DeLong at January 12, 2003 06:57 PM | TrackbackCreate a corporation, use as many tax dodges as possible (since corporations that are profitable on the balance sheet regularly have no profits according to the IRs) then pay yourself in dividends. I already know people who are looking into this.
But isn't that second-order effect on the deficit balanced by the second-order effect of what households do with the dividend income?
Being an Econ BA may not give you deep knowledge on the economy, but it does allow you to BS:)
Posted by: roublen vesseau on January 12, 2003 08:12 PMIt's a lot more complicated than that.
For one thing, profits are calculated on the income statement not the balance sheet.
More importantly, the only way companies that aren't earning "real" (as opposed to paper accounting profits) profits can grow their dividends over time is by borrowing. The old saying that "dividends don't lie" applies. This is actually a decent first step in restoring confidence in the stock market, and as importantly, in helping lower the cost of capital of companies with conservative accounting and helping raise the cost of capital of companies with liberal accounting.
And Dr. Auerbach seems to believe that the dividends payed out to investors will disappear somewhere. They won't. Lots of investors (ERISA pension plans, mutual funds*, etc) just take the dividend payout and reinvest it in the stock of some other company. Which will lower the cost of capital of companies that investors think deserving and thus encourage those DIFFERENT companies to expand their best capital projects at the margin.
There's always been tangible friction between the extra capital generated by many companies - the management wants to keep it within the company for many reasons, while professional shareholders want to see companies run with as little capital as possible. Lots of professional investors (such as myself) see the possibility of increased dividends as a very good thing - because it encourages companies that don't have great investment prospects to ship the capital back to shareholders through dividends.
* In the short run, equity mutual funds retain all dividends paid. On a fiscal year basis, tax law then requires mutual funds to pay out virtually all realized capital gains and dividends to shareholders . . . . . but at the 30 mutual funds I've worked on, about 85% of the dividends "paid out" to mutual fund shareholders were automatically reinvested back in the mutual fund at the current NAV per share. There was a little Leakage, but it was quite small.
Posted by: Anarchus on January 12, 2003 08:25 PMSo does Brad believe that increasing dividend taxes would be a good way to increase investment?
Posted by: ed johnson on January 12, 2003 11:06 PM"Given that both effects are first-order and that they oppose each other, the net effect is going to be small."
Hang on, that doesn't follow. That there are two effects working in different directions means the outcome is theoretically indeterminate. But theoretically indeterminate doesn't imply empirically small - we have to measure to see if in real life the two effects are of similar size.
Or do you hold that when raising taxes the income effect can always be expected to approximately equal the substitution effect (eg income taxes have always have little net effect on labour supply)?
Posted by: derrida derider on January 12, 2003 11:36 PMIt would be a shame if rationalising the taxation of corporations was to become a victim of larger questions about taxation. Favoring interest payments and sharebuybacks over dividends helps make investment in corporations more volatile especially when combined with incentivisation through options.
Lack of dividends also deprives the market of a signal about the health of many companies. Also the build up of large piles of cash in corporations unable ot return it to investors is a great destoyer of value. Microsoft's investments beyond Office and Windows have been woeful for example.
The other problems are real but maintaining a bizarre distortion of the tax system is surely not the best way of solving them.
Posted by: jack on January 13, 2003 04:48 AMderrida derider makes an interesting point. In my (hazy) understanding of such things, retained earnings are far and away the largest source of investment capital for most US corporations. I have read assertions (haven’t looked into them) that equity is a net drain on capital over time, as even under current tax law, more eventualy goes out in the form of dividends than comes in through stock purchases. If we make a small source of capital larger and a large source smaller, questions of the magnitude of change in each become very important. It would be nice to think that the answer won’t really matter, that corporate managers will always decide to invest up to the point that returns are no longer high enough to justify further capital outlay. The intrusion of corporate agents’ agendas makes me doubt that. I don’t know which way the first order effects will cut, but the difference in magnitudes between retained earnings and equity as sources of capital makes me thing derrida has a strong point – income and substitution effects very likely won’t come close to balancing.
Posted by: K Harris on January 13, 2003 05:26 AM>>So does Brad believe that increasing dividend taxes would be a good way to increase investment?<<
Increasing other taxes would be much better: increasing dividend taxes gives companies an artificial incentive to leverage up, and higher debt-equity ratios are not something we should be encouraging.
Posted by: Brad DeLong on January 13, 2003 06:50 AMJust to point out (derrida derider and Anarchus) that it is pretty much orthodox corporate finance theory that retained earnings are a cheaper source of finance than external finance, because they do not create the same agency problems. If one "recycles" dividends from one firm to another, the fact that insiders to a have have asymmetric information with respect to the investing public puts a fairly substantial wedge between the cost of (internal) capital and the rate which investors demand on new equity issues.
Posted by: dsquared on January 13, 2003 07:44 AMIncreasing taxes that reduces the deficit or produces a surplus increases domestic savings available for investment, and lowers risk-free interest rates. All that is good and positive for investment.
But why not at least consider removing some of the powerful incentives to consumption at the individual level? At least part of the reason the American savings rate is so low is that the tax code has been strongly biased against saving for a couple of generations . . . . .
Posted by: Anarchus on January 13, 2003 07:52 AMAnarchus -- The "dividends don't lie" maxim is overrated. It's easy enough to run a Ponzi scheme with a 400% per annum payout on "invested" capital ... until the pyramid topples or the schemers get caught red-handed. How long before a 4% per annum dividend payout would scrape bottom in the absence of genuine value-added activity?
On a slightly different tack, how many dividend-paying corp's are debt-free? Most of them are, in effect, paying dividends with borrowed money.
On a third tack, debt is not entirely unhealthy as an incentive to good stewardship of productive capital. The debt holder is sensitive almost exclusively to downside prospects, whereas option holders are almost entirely insensitive to downside prospects, and the Street (not just "the market", but diverse touts and hangers-on) is disproportionately sensitive to upside prospects.
Worth noting, major debt ratings agencies annouce plans to take more aggressive postures with regard to aggressive accounting. (2003-01-11 NYT, etc.)
Posted by: RonK, Seattle on January 13, 2003 09:27 AMThe basic point is the corporations pay less tax on income than individuals do. So, to get that lower effective rate you use a corporation. Moreover there appear to be more loopholes available for corporations. Small corporations, moreover, pay less than the 35% base tax. Getting rid of the dividend tax provides even stronger incentives to incorporate to avoid taxes than already exist. In effect, beyond a certain income level, no one will pay more than the corporate rate for their income (which I've listed below).
Small Corporations.
http://www.ctj.org/hid_ent/part-2/part2-11.htm
Smaller corporations are taxed at lower rates than the 35% regular corporate rate,with the tax savings from the lower rates phased out for larger companies.As a result, statutory corporate rates bounce around a lot. Specifically,they:
start at 15% of the first $50,000 of taxable income,
go to 25% on the next $25,000,
then go to 34% on the next $25,000,
then go to 39% on the next $235,000,
then go back down to 34% up to $10 million,
then go up to 35% on the next $5 million,
then go to 38% on the next $3.33 million
and finally go to 35% on income above $18.33 million.
On the loopholes:
http://www.law.wayne.edu/mcintyre/text/DCJ_state_avoidance-July18_02.pdf
"The effect of tax shelters on federal revenue is examined in a study that the I.R.S. will publish today in the quarterly Statistics of Income Bulletin.
"Untaxed corporate profits rose to 24 percent of the profits reported to shareholders in 1998, up from 14 percent in 1996, an increase of two-thirds. This increase meant that the amount of corporate profit not taxed rose by $66.5 billion
from 1996 to 1998."
"The I.R.S. analysis follows a March study by a Harvard economist, Mihir A. Desai, who sifted through reports to shareholders to determine whether the use of tax shelters was a significant issue and found evidence that the problem was much
larger than the I.R.S. study."
"In 1998, less than 62 cents of each dollar of shareholder profit turned up on tax returns, down from 84 cents in 1996, Professor Desai said. Untaxed corporate profits totaled $247 billion in 1998, Professor Desai found, but deductions for stock options and other known legitimate deductions explained only about $88
billion of this amount."
"His study suggests that in 1998 $155 billion or so of corporate profits were hidden from the I.R.S. in tax shelters, costing the government as much as $54 billion in taxes. That figure is more than five times the $10 billion cost of abusive tax shelters cited by in 1999 by Lawrence H. Summers, who as Treasury secretary then started a campaign to crack down on tax avoidance and evasion by large companies."
"Professor Desai said his analysis of shareholder and taxable profits from 1998 to 2000 found that until 1995 the relative difference was stable and followed expected shifts caused by tax rules. But starting in 1996, he said, the gap between
taxable profits and those reported to shareholders began to widen apparently because of tax shelters."
RonK: we disagree on this: "The "dividends don't lie" maxim is overrated. It's easy enough to run a Ponzi scheme with a 400% per annum payout on "invested" capital ... until the pyramid topples or the schemers get caught red-handed."
Ponzi schemes only work when there's more money coming in than going out. IF investors have truly learned a big lesson about accounting, and I think they have, then they can very easily track inflows/outflows just by watching the balance sheet over time, or by looking at the cash flow statement. Companies that try the Ponzi game of borrowing to pay dividends should get caught quickly in the new environment.
In the bad old days, it was easy for an Enron or a Worldcom to leverage up, hide debt-like liabilities off-balance sheet and/or do slimy pooling acquisitions to hide the deterioration of one's own financials. But given the indictments coming down, CFO's and auditors are scared to death, and investors look askance at any company with impenetrable accounting (GE, take note!).
This doesn't work for me either: "On a slightly different tack, how many dividend-paying corp's are debt-free? Most of them are, in effect, paying dividends with borrowed money."
Debt capital is cheaper than equity capital, and there's other good reasons for many companies to have some debt (funding a new plant or building cheaply with secured bonds, for instance) . . . . . . . . so most companies should have a capital structure made up of some debt and equity whether they pay dividends or not.
If a company is in a capital intensive business and really has a bunch of extremely high return projects to invest in, then of course they shouldn't pay a dividend - but in my opinion, there just aren't that many companies like that out there - American business has gotten much less capital intensive over time.
In fact, the last really huge capital intensive new industry to come along was Jet Transportation . . . . . . . . . compared to the capital intensity of the oil business in the early 20th century, or the railroad industry of the late 19th century, or the steel industry or auto industry or farm machinery industry, all the great new age modern stuff (of which I am a huge fan) is NOT CAPITAL INTENSIVE.
Look at Microsoft's balance sheet sometime. AWASH in cash, because the software business takes huge amounts of human capital but very little machinery, plant, equipment, etc. Cisco, for gosh sakes, makes almost NOTHING. It's a research, design and marketing firm that subcontracts out practically 100% of its manufacturing to third parties. Drug companies would be another example of a low capital, knowledge intensive industry.
that's all for now.
Dear er, whoever: When you say "Companies that try the Ponzi game of borrowing to pay dividends should get caught quickly in the new environment.", aren't you saying "dividends as tangible evidence of real earnings are superfluous in the new environment"? My point, either way.
Re dividends and debt, I'm not arguing that corp finance should avoid debt ... merely that "dividends as proof" (of anything) is proof of nothing, in cases where capital is coming in the back door as it is distributed out the front door.
As for your assertion of declining capital intensity ... a declining share of GDP goes to labor ... a declining share goes to capital ... and a declining share goes to the tax man. Hmmm.
Posted by: RonK, Seattle on January 13, 2003 02:41 PMLook at Microsoft's balance sheet sometime. AWASH in cash, because the software business takes huge amounts of human capital but very little machinery, plant, equipment, etc.
I thought Microsoft was awash in cash because (a) they have a quasi-monopoly in the OS and office applications markets, (b) a good chunk of the demand for computer software is generated by folks who have no understanding of computer software and are willing to pay a hefty premium.
Posted by: Stephen J Fromm on January 13, 2003 02:45 PMMicrosoft was just one example. Here's a few others: Oracle, with no monopoly, has $5.5 billion in cash versus $313 million in debt; Intuit has $860 million in cash and $32 million in debt; acquisitive Computer Associates has $3.1 billion in debt versus only $776 in cash; Adobe has $620 million in cash and no debt; Symantec has $1.2 billion in cash and $600 million in debt and Seibel has $2.1 billion in cash and $320 million in debt. With the exception of Computer Associates which made some acquisitions for cash, almost every major software company is awash in liquidity.
And it's because the software business is not capital intensive, because there are almost no diseconomies of scale in the software business (so marginal costs are close to zero as far up the production scale as you can see) and also because R&D is 100% tax deductible each and every year, while traditional capital expenditures are not.
Posted by: Anarchus on January 13, 2003 02:58 PMOne can argue intelligently for dividends and interest on debt to be treated equally in corporate tax policy. This can be done fairly easily and, with minor adjustments in corporate tax rates, be revenue neutral if that is the objective. This is quite different from the dividend tax proposal, which requires complex calculations and monitoring to put into effect and would be quite difficult to make revenue neutral.
However, one of the most telling arguments against the current proposal was made by my oldest son, a resident of Napa, CA in a recent letter published in the Chronicle:
"Editor -- As if anyone needed a reminder that the taxation system in this country is completely irrational, we in California are faced with the dual probabilities of a federal cut in dividend taxes and a state increase in the sales tax.
I leave it to the policymakers, accountants and attorneys who make their livings off the system to postulate and posture as to what the cumulative effect on the average Californian will be. It doesn't take a think tank to show that the net effect of a higher sales tax will be disproportionately on those whose incomes are derived from wages and whose wages purchase, primarily, the necessities of life."
But aside from arguing how regressive our system is becoming, can't anyone see the lunacy, waste, bloat and irrationality of our tax system today? Forget about explaining away the differences among federal, state and local taxes. How does the combination impact our citizenry? What is the bottom line we're willing to endure?
Death and taxes, yes. But mental cruelty and slow torture, no. "
SAM TAYLOR
Napa
I would be curious to see comments from other residents of California as well as economists who were (I'm sure) inadverdently omitted in my son's category of those who make their livings off the system.
Sam Taylor (the elder)
One can argue intelligently for dividends and interest on debt to be treated equally in corporate tax policy. This can be done fairly easily and, with minor adjustments in corporate tax rates, be revenue neutral if that is the objective. This is quite different from the dividend tax proposal, which requires complex calculations and monitoring to put into effect and would be quite difficult to make revenue neutral.
However, one of the most telling arguments against the current proposal was made by my oldest son, a resident of Napa, CA in a recent letter published in the Chronicle:
"Editor -- As if anyone needed a reminder that the taxation system in this country is completely irrational, we in California are faced with the dual probabilities of a federal cut in dividend taxes and a state increase in the sales tax.
I leave it to the policymakers, accountants and attorneys who make their livings off the system to postulate and posture as to what the cumulative effect on the average Californian will be. It doesn't take a think tank to show that the net effect of a higher sales tax will be disproportionately on those whose incomes are derived from wages and whose wages purchase, primarily, the necessities of life."
But aside from arguing how regressive our system is becoming, can't anyone see the lunacy, waste, bloat and irrationality of our tax system today? Forget about explaining away the differences among federal, state and local taxes. How does the combination impact our citizenry? What is the bottom line we're willing to endure?
Death and taxes, yes. But mental cruelty and slow torture, no. "
SAM TAYLOR
Napa
I would be curious to see comments from other residents of California as well as economists who were (I'm sure) inadverdently omitted in my son's category of those who make their livings off the system.
Sam Taylor (the elder)
As to this observation, "It doesn't take a think tank to show that the net effect of a higher sales tax will be disproportionately on those whose incomes are derived from wages and whose wages purchase, primarily, the necessities of life."
California's state sales tax exempts many necessities such as food and medicine. It at least attempts to avoid regressivity that way . . . . . . .
Posted by: on January 13, 2003 07:42 PMRegarding sales taxes...not only are they regressive...they are more expensive to collect!
Posted by: Sam Jackson on January 14, 2003 01:58 AMStephen,
Microsoft has earned a lot of cash for the reasons you cite but that does not explain why it has kept the cash nor whether what it does with the cash is valuable to anyone.
Derrida write:
>Or do you hold that when raising taxes the >income effect can always be expected to >approximately equal the substitution effect >(eg income taxes have always have little net >effect on labour supply)?
There is a rather large empirical literature on precisley this question and it indicates that the effect is indeed zero. Saying an effect is indterminate generally means you can't "assign a sign" that is positive or negative effect. In this case some of the magnitudes asociated with this are known- And the overall effect is indeed about as close to zero as you can get..
Oops, Sorry about the misspellings. I have a hard time believing anyone could seriously argue that a dividend cut, especially like the one proposed will have anything other than distributional effects. The oft misued term "rent seeking" applies. Money in exchange for no productive activity. In terms of the effect on assett prices I find it interesting that Canada does not offer an interest deduction on home mortgages but the U. S. does. This is very similar in some ways to the dividend tax, in that it is something like a tax on an assett. The interesting thing is that housing prices, and the level of home ownership in Canada are equivalent. to that in the United States..
Posted by: Lawerence Boyd on January 14, 2003 07:09 PMLawrence -
You're kidding!
First, the empirical literature is not large - its bloody massive. So massive, its hard to know where to start with references.
Second, estimates of compensated labour supply elasticities vary greatly, but the median estimate for most groups is NOT zero (except perhaps in some of the older, econometrically less sophisticated) literature. In particular, women (esp women with children) and high-income people seem to be quite sensitive to after-tax returns (tho in the case of women, much of this takes the form of change at the extensive margin - ie the decision to seek the corner solution of non-participation, rather than change in the intensive margin of hours worked). It is true that prime age blue collar males seem to vary their working behaviour in response to taxes less than other groups - which may explain why they are so heavily taxed in amny countries!
I'm no supply-sider, but I don't think you can tenably hold that labour supply is unaffected by taxes and benefits.
On company taxation, I still think theory gives little a priori guidance on the effects. US economists need to shrug off their insularity and look at the empiric evidence from the many countries which have played with this question over the past twenty years. You might start with my own country, Australia, and its neighbour New Zealand, both of which moved away fom the "classical system" of company tax in the 1980s and both of which have had quite a few academic studies and policy evaluations of the effects.
Posted by: derrida derider on January 14, 2003 09:44 PMDerida,
Could you cite your sources. In general males, and especially elderly males in the United States have been found to have negative elasticities. These are the ones most effected by the dividend cuts and previous tax hikes. Summaries can be found in Killingsworth _Labor Supply_ and other articles.
Derrida,
I forgot to mention this but Killingsworth in his Nobel Lecture also takes up this topic and that is a good source of references as well. BTW I don't disagree that their are large positive elasticities for females, blacks, hispanics etc. but in general this holds for lower paid workers, precisely those not effected by these tax cuts.
Derrida,
I forgot to mention this but Killingsworth in his Nobel Lecture also takes up this topic and that is a good source of references as well. BTW I don't disagree that there are large positive elasticities for females, blacks, hispanics etc. but in general this holds for lower paid workers, precisely those not effected by these tax cuts.
Derrida,
Point well taken on insularity. I realized that I generally follow American and British journals and might be unaware of publications on this topic from other countries. I do think that especially New Zealand has had coverage in these Journals. So if you could cite some summary articles on this I really would be happy to learn (and change my mind). I think you are reading more into this proposal than is there. It really is nonsensense, and the more closely you read the fine print the more nonsensical it becomes.