Social Econ Blog

Returning to the topic of tax shifts from last week, Greg Mankiw starts an economics blogosphere debate:

In today's CEA op-ed, the sentence that will likely generate a blogosphere debate with the most heat and least light is this one: "The president's tax cuts have made the tax code more progressive."

The basic problem is that there is no single way to gauge changes in progressivity. As a result, people can take the same set of numbers, look at them from different angles, and reach very different conclusions.

He then goes on to provide a nice description of the different ways people might define progressivity.

Brad Delong disagrees with even the premise of the post:

The first problem with Greg's post--the first reason that Greg is generating more heat than light--is that Lazear and Baicker's op-ed doesn't say "the tax code is more progressive, even though the differences taes generate in after-tax take-home pay have grown." What Lazear and Baicker say is "The president's tax cuts have made the tax code more progressive, which also narrows the difference in take-home earnings." They claim--a claim that I think is simply false--that the Bush tax cuts are progressive no matter what definition of progressivity you adopt.

The second problem is deeper. Let me try an analogy. I, full professor Brad DeLong, am having lunch with lecturer Dariush Zahedi today. After lunch, I presume Dariush will say we should split the bill--$10 each. Suppose I say: "That isn't fair. Berkeley pays you less (a lot less: what we do to our lecturers is shameful) than it pays me. I should lay out more cash for this lunch. How about this: I put down $5 cash, you put down $0, and we put the balance on your credit card. That would be fairer, wouldn't it?"

Dariush would then be an unhappy camper. He would think--correctly--that I was mocking him.

Back in 2000 the U.S. government was running a surplus of some $200 billion a year--a broadly appropriate fiscal policy, given the state of the business cycle and the looming health care costs dilemma. Today we're running a deficit of $300-$400 billion a year. Relative to what would be a sane, reality-based, and appropriate fiscal policy, the Bushies are putting $500-$600 billion this year on our collective national credit card. That bill will come due: somebody has to pay it. To pretend that it won't--to pretend that you can talk about the progressivity of the burden of paying for the federal government without talking about the long-run incidence of the national debt--well, that would be the equivalent of me telling Dariush that only cash matters: that when we talk about who paid for lunch, we should count only cash put own now, and we shouldn't count the fact that his credit card bill will show an extra $15 due next month.

Alex Tabarrok likes Brad's description of tax shifts, and provides a link to his own op-ed on the subject.

Finally, Jane Galt points out that Brad's metaphor is wrong because it assumes that the debt must be paid by the poor. She argues that in the end the Bush tax cut will be progressive:

There are four ways that I can think of for the US to deal with its debt burden:

1) Raise taxes to pay it off. This will be a burden on America's relatively poorer members only to the extent that the tax burden falls on them. But while overall withholdings are only very mildly progressive in America (thanks to Social Security payments), income tax is extremely progressive; the effective federal tax burden on someone like Arnold Schwarzenegger is roughly 30%, while a median-income family of four pays less than 8% of its income to the federal tax ... Is the tax system likely to change? Colour me sceptical. "Raise taxes on the poor" just doesn't have much of a ring to it ...
2) Inflate/devalue the currency, so that the debt gets repaid in cheaper dollars. This largely hurts investors, aka the rich, whose assets become worth less, while helping debtors, aka regular joes.
3) Default. This is very bad for the economy, and everyone in it, but especially the wealthy. The wealthy, who, as mentioned above, hold t-bills, lose their whole investment. They also tend to see their incomes permanently hurt from big financial crises, at least according to this paper.
4) Roll it over. This presumes growth in GDP/tax revenues large enough to do so at attractive rates. The interest is financed by tax revenues, which, as noted above, tend to come from the rich--and will ever more so after 2010, as social security revenues are diverted to--crazily enough!--pay for social security.

Of course, if her reasoning is correct. It doesn't make sense for rich people to support tax shifts unless they expect to no longer be rich when taxes are finally raised or they are currently severely credit constrained (and thus the tax "cut" was basically the government borrowing on its behalf). I'd like to see the empirics on this. When the government choose tax shifts, who ultimately pays off the debt? Are the rich able to successfully increase taxes on poor people (through things like "sin taxes")? Are credit constrained rich people more likely to favor deficit inducing tax cuts? Among those who were "rich" at the time the tax cuts are passed how many are still rich when their taxes are, according to Galt, "inevitably" raised?

Update -- Mankiw reenters the frey and, among other things, says:

Because of technological progress, the income and consumption of a typical individual in the economy rises over time. Because budget deficits shift taxes forward in time, they benefit relatively poor current taxpayers at the expense of relatively rich future taxpayers. If reducing inequality is a goal of policy, shouldn’t budget deficits be applauded?

Tony V debunks this logic here:

It was suggested that one possible motivation for such a time-inconsistent policy is that due to economic growth, future generations would be better suited to handle paying the tax burden (i.e. they would be richer and thus able to handle it). The graph above plots the list 15 years of data of GDP growth rates and the annual average of monthly 5-Year Treasuty Bill rates (i.e. one measure of the interest rate paid on government debt). As you can see, only recently has the growth of GDP been higher than interest rate paid on debt, so it doesn't seem like that rationale holds any water.

In fairness, I should point out that some of Mankiw's other arguments are more reasonable.