There is so much that is troubling and wrong with Harvard economist Greg Mankiw’s op-ed in Sunday’s New York Times that it is hard to know where to start. His piece warns that if the Bush tax cuts for those who earn more than $250,000 a year –as he does– are allowed to expire later this year, then he will work less.
To make his point, Mankiw compares what will happen if he earns an extra $1,000 (for writing an article such as his New York Times column) and invests the proceeds at an 8 percent real rate of return for 30 years, under two different circumstances. In the first case, he lives in a world where no one pays any taxes on anything. In the second case, we have the tax structure that will be in place after the Bush income tax cuts on top earners sunset and health-care reform related tax increases kick in.
Here are a few of the things that bother me.
(1) Remember, the piece is primarily about the scheduled increase in marginal tax rates for the top two tax brackets, which cover taxpayers with incomes above $250,000. (The highest bracket would revert to its pre-2001 level of 39.6 percent, up from 36 percent now; the next highest bracket, which kicks in at about $250,000 per year, would rise to 35 percent, up from 33 percent now.) Yet, Mankiw’s motivating example contrasts (a) a world with no taxes to (b) a world with the full 2011 tax system in place, including –almost incidentally– the incremental increase in taxes related to the scheduled sunsetting of the Bush tax cuts.
Could there be a less instructive counterfactual than a world with no taxes? No public schools, no Social Security, no Medicare, no Food and Drug Administration, no Environmental Protection Agency, no banking regulation, no Securities and Exchange Commission, no police, no courts. Mankiw’s zero-tax example absurdly assumes that in such a world, he would somehow be able to earn an 8 percent annual rate of return, keep the proceeds, and safely pass them on to his children.
(2) Even when Mankiw reels it back in and seems to focus more concretely on just the incremental effects of the Obama-related tax increases, the example he gives is confusing at best. Let’s go line by line:
First, assuming that the Bush tax cuts expire, I would pay 39.6 percent in federal income taxes on that extra income.
The 39.6 percent rate may sound like a lot, but the Obama-related portion of this only increases his marginal rate by 3.6 percentage points, from 36 percent to 39.6 percent.
Beyond that, the phaseout of deductions adds 1.2 percentage points to my effective marginal tax rate.
Fair enough. Cumulative increase in Mankiw’s marginal rate: 4.8 percentage points.
I also pay Medicare tax, which the recent health care bill is raising to 3.8 percent, starting in 2013.
Yes, but Mankiw is already paying 2.9 percent on Medicare. The increase in his marginal rate in 2013 will be only 0.9 percentage points. Cumulative increase in Mankiw’s marginal tax rate: 5.7 percentage points.
And in Massachusetts, I pay 5.3 percent in state income taxes, part of which I get back as a federal deduction.
Is Obama raising state taxes in Massachusetts, too? Mankiw pays these taxes either way. State taxes should have absolutely no bearing on his decision about how to respond to the federal tax increases (except in so far as he gets some of them rebated, which has already factored that in above). It is astonishing to me that someone who teaches economics at Harvard would make this argument. His teaching assistants would certainly take points off of their undergraduates for writing something like this in an exam. Cumulative increase in Mankiw’s marginal tax rate: (still) 5.7 percentage points.
Putting all those taxes together, that $1,000 of pretax income becomes only $523 of saving.
Let’s take Mankiw’s word on the $523 calculation. The relevant question, however, is how the Obama-related tax increases (counting the planned sunsetting of the Bush tax cuts as “Obama-related”) affect his marginal tax rate. Again, according to Mankiw’s calculation, if President Obama has his way, for every $1,000 article that Mankiw writes in 2011 (well, really, 2013 when the Medicare increase kicks in, but no need to nitpick), he’ll receive $523. That is 5.7 percentage points less than he would have under the “Bush” rates. On $1,000 that translates to a grand total of $57 –on a $1,000 paycheck, for someone whose income is already in excess of $1 million a year.
(3) One way that Mankiw gets this $57 dollars on every thousand to look like Swedish-style socialism is to walk his readers through a scenario where he earns 8 percent annual real rates of return and accumulates an estate large enough to be subject to the estate tax. Suppose, he writes, that he earned $1,000 tax-free for writing an article:
… [and] I invested it in the stock of a company that earned, say, 8 percent a year on its capital, then 30 years from now, when I pass on, my children would inherit about $10,000.
…Now let’s put taxes into the calculus.
…that [$1,000] savings no longer earns 8 percent. First, the corporation in which I have invested pays a 35 percent corporate tax on its earnings. So I get only 5.2 percent in dividends and capital gains.
The 35 percent rate is the marginal tax rate facing corporations, not the average (or effective) tax rate they actually pay. Mankiw would do better choosing to invest in a firm that pays a lower effective tax rate.
In any event, the argument about “double taxation” is as old as the Wall Street Journal op-ed page. Nothing compels Mankiw to invest his savings in corporations that are subjected to a 35 percent tax rate. Business owners are free to choose to incorporate –and be subject to a 35 percent corporate tax rate– or to choose other forms of legal organization such as partnerships where the owners, not the corporations, pay taxes on any profits. Despite the 35 percent tax rate, most large firms choose to incorporate because corporate status gives owners massive benefits –most importantly legal and financial liability limited to the individual’s total investment. How do we know that these government granted and enforced benefits are massive? Because owners are willing to pay up to 35 percent of their profits to the government in exchange for these legal protections. (Though, again, very few companies actually pay anything close to 35 percent of their profits in taxes.)
Then, on that income, I pay taxes at the federal and state level.
Again, the only amount that is relevant here is the additional federal tax resulting from the Obama-related tax increases.
As a result, I earn about 4 percent after taxes, and the $523 in saving grows to [only] $1,700 after 30 years.
Without the Obama-related tax increases, Mankiw would have received $580 for his $1,000 article; after 30 years at a 4 percent growth rate, this would have increased to about $1,880. So, 30 years from now, the actual cost to Mankiw’s estate of the Obama-related tax increases would be $1,880 minus $1,700, or about $180.
(4) Mankiw’s characterization of the estate tax is incomplete and probably misleading for most New York Times readers, who likely don’t have direct experience dealing with large estates. Mankiw notes that the $1,700 that his estate would hold 30 years from now as a result of him writing an article for $1,000 and investing the proceeds at 8 percent would would only be worth about $1,000 to his children –after the estate tax.
To arrive at this conclusion, Mankiw is assuming that his heirs will pay an inheritance tax of about 40 percent. (Because the US Congress is fairly dysfunctional, the estate tax for people who die in 2010 is zero. The previous tax, which expired at the end of 2009, was 45 percent on the portion of an estate above $3.5 million. If the Congress does nothing to change the law, in 2011 the estate tax will rise to 55 percent, with an exemption on (approximately) the first $1.3 million; Congress, however, will likely act next year to lower the rate and raise the exemption.)
In other words, while the specifics are a bit up in the air, Mankiw is, in general terms, telling his readers that he plans on bequeathing his three children at least $400,000 each –otherwise they would owe nothing in the estate tax under 2011 law. He is also saying that for each $1,000 article he writes once his estate tops $1.3 million, his children will lose about $180, as a result of the Obama-related tax increases (that is, after 30 years compounded growth at an 8 percent annual real rate, the cost of the Obama-related increase is smaller if the real return is lower).
(5) A great use of Mankiw’s column would be for him to reveal where he is getting an 8 percent annual real return over the next 30 years.
(6) Of course, incentives are important, but Mankiw paints a cartoon picture of them. He makes hair-on-fire predictions about the impact on his (and, by extension, other highly paid professionals’) output in response to a 5.7 percentage-point increase in their marginal tax rate. As Kevin Drum notes: “If you believe that taxes affect incentives — and I do — then you should also believe that small changes in tax rates affect incentives in small ways.”
Mankiw also ignores all of the non-monetary incentives at play here. As Mark Thoma argues the fact that Mankiw teaches, does research, or writes columns for the New York Times or anywhere else suggests that he is not particularly responsive to financial incentives. Mankiw could make substantially more as a consultant to the financial sector, for example, than he does at any of those activities. As Thoma writes: “I’d … guess that even if the New York Times stopped paying him for his column, he’d write it anyway. It’s a boost to his ego and reputation that he’d want even without whatever small payment he gets for each column.”
Economists are fairly consistently uninterested in what economic actors say and much more interested in what they actually do. Mankiw’s actual behavior suggests that, on the margin, rich, creative people do what they do for a lot of different reasons, and that when you are worth millions, percentages of that here and there don’t have a big impact on your behavior. (Did Mankiw work less –let alone a lot less– in the 1990s, when marginal tax rates at the top were a bit higher, and then more –let alone a lot more– in the 2000s when they dipped down? Would anyone but an economist believe that?)
(7) What about the children? Mankiw has dragged his poor kids into this, for which I hope that they do ultimately receive substanial financial compensation (very, very many years from now). But, doesn’t Mankiw worry about undermining his children’s incentives to work, the same way that his colleague Robert Barro worries that unemployment insurance benefits undermine the work incentives for the unemployed? From a social point of view, shouldn’t we all be concerned that Mankiw is acting in a way that will reduce the labor supply of his children?
(8) For me, though, the deepest issue here is that Mankiw presents the problem facing the American people as choosing between more articles by Greg Mankiw, on the one hand, and fewer articles by Greg Mankiw, on the other hand. As Princeton economist Uwe Reinhart put it in the comments section of Berkeley economist Brad DeLong’s blog:
… the taxes paid by the rich actually pay for stuff. They are not raised just to punish the rich.
Greg might have mentioned that, as Americans, we face the difficult trade-off between (1) having one more essay by Greg Mankiw and (2) the things government buys with tax revenues including, say, adequate armor for our warriors in combat (I recall my son’s unit being short flack jackets when it fought in Iraq in 2003) or health care for waitresses and cab drivers. Someone has to pay for that stuff.”
Mark Thoma wonders whether the Obama-related tax increases on the top two percent will lead Mankiw to quit his column at the New York Times (I am sure that there is someone who is almost as good who is willing to do the job for a little less). If so, I can only say that it is too bad that the tax hike didn’t kick in before he wrote this particular column.
This article originally appeared on John Schmitt's blog, No Apparent Motive.
(Only one link allowed per comment)