« Dogs can, monkeys and wolves can't | Main | The Fed's Dirty Laundry and Yours »
It is worse than you think
Here is Marty Feldstein, through some vias:
[F]inancing additional government spending by an across the board rise in all marginal tax rates would make the cost per dollar of government spending equal to $1.76.
These two facts — that the actual revenue is only 57 percent of the static gain and that the deadweight loss is 76 cents per dollar of revenue — should be central to any consideration of tax policy. And yet they are not.
What is talked about even less is that most government programs are, for better or worse, irreversible investments. They don't fade away very easily, even if they have been proven ineffective or harmful. Do you really think that the Departments of Energy and Education have to pass a market test every year? In fact government programs, on average, grow almost automatically.
Do you know the literature on irreversible investment and option value? Under plausible parameter specifications, you need a benefit-cost ratio of 3 to 1 or more before it makes sense to proceed with the irreversible investment. Otherwise it is better simply to wait. That's not 1.3 to 1, that's 3 to 1.
The main argument for not waiting is simply that the political process cannot do much useful with any new information that is generated by waiting. But viewed more broadly, that is hardly an argument for proceeding in the first place.
The bottom line: Cram all those numbers together into your noggin' and keep them there.
Posted by Tyler Cowen on September 11, 2007 at 06:31 AM in Economics | Permalink
Comments
Yikes, that is some pretty scary stuff. There is more dead weight loss associated with public outlays than previously thought. In light of all of the empirical evidence which is finding that most government expenditures have a benign or even harmful effect on society, why dont more people support rules for spending? I am thinking about something along the lines of what Buchanon argues?
Posted by: john pertz at Sep 11, 2007 7:15:28 AM
Feldstein must be a relative of Mr Spock to know this with such precision, $1.76 not even rounded to the nearest 10, as in over a $1.70 or almost $1.80.
Posted by: joan at Sep 11, 2007 7:47:06 AM
A few questions about Marty Feldstein's calculation.
1) Isn't it possible that government is serving a role that people want served? And thus, if you removed the government program, people would want the gap filled with private giving? And if people divert their cash flow, even through a private decision, it will still affect propensity to work?
If the program still needs the same cash flow whether via government or charity, what is the difference with charity which permits no dead weight loss? I can think of a few possibilities but they seem fairly subtle. I think there must be some effect on propensity to work.
2) Isn't much of this deadweight loss "regained" in terms of utility by people enjoying more leisure? I understand that the tax distorts incentives compared to laissez-faire, but suppose people are almost indifferent at current levels between working and leisure. In that case the picture is much more positive than Feldstein suggests.
Posted by: mk at Sep 11, 2007 9:17:07 AM
mk: "...what is the difference with charity which permits no dead weight loss?"
Charitable giving is voluntary, while fund-raising is competitive. Neither can be said for taxation or government programs.
Also, charities are less susceptible to abuse than government benefits. Recipients know that what they're collecting is the result of other people's good will and benevolence, and not just a welfare entitlement.
Posted by: Russ R at Sep 11, 2007 9:25:27 AM
Charitable giving is voluntary, while fund-raising is competitive
OK, the voluntary aspect I can imagine. If I control where my money goes, I get a bit more utility out of the marginal buck, thus I work a little harder for it. I think it's very much an open, empirical question how big this effect is, however.
The competitiveness you mention would seem to be an argument about administrative efficiency, not dead weight loss. There are many interesting empirical questions here, but the question earlier was particularly about dead weight loss.
Susceptibility to abuse is another interesting point.
I would prefer that all these effects be experimentally measured. Plausibility, as we empiricists know, doesn't guarantee it actually happens like that.
Posted by: mkayser at Sep 11, 2007 9:40:32 AM
In fact government programs, on average, grow almost automatically.
Compared to what? Population? GDP? I don't see why government programs shouldn't grow if the country is growing.
What really bothers me about this line of argument ("government just gets bigger!") is that it's used by "starve the government" advocates to argue for lowering taxes without concern for which programs get cut. If you really care about economic efficiency, you should be arguing which programs have outlived their usefulness.
I understand it is much harder to convince people that a government program is unnecessary than it is to convince people that taxes are too high, but I don't come to MR to hear inuitive arguments that sit easy with the masses. I come here for tight, fact-based, insights I can't get elsewhere.
Posted by: dstevens at Sep 11, 2007 9:57:44 AM
These two facts — that the actual revenue is only 57 percent of the static gain and that the deadweight loss is 76 cents per dollar of revenue
That sounds like only one "fact" to me, stated two different ways.
Posted by: Bernard Yomtov at Sep 11, 2007 10:01:44 AM
mk,
You make some excellent points regarding an empirical comparison of the marginal utility of money passing through the government via through voluntary institutions toward the same goal. The dead-weight loss must account for:
- the difference in utility for same-goal spending (the administrative cost you cite)
- the utility of politicians choosing the goal versus the market (including the market for charity)
- the loss associated with the coercion necessary to enforce the government goal (both administrative costs and personal utility loss for the schmuck earning the taken dollar)
I'm not sure which of these Feldstein's calculation refers to, but all of them have been documented as contributors to dead-weight loss.
And for you fans of empiricism, I would suggest you do a tour of your local government office compared to a private company, or even non-profit, and experience the difference. It's not a difference in people so much as in management and internal incentives. It's huge.
The uber-incentive, of course, is the fact that if my spending your dollar means 57 cents for the economy, but I get to spend it and you don't, then stick-em up--it's mine.
Posted by: M. Hodak at Sep 11, 2007 10:03:34 AM
And for dstevens who thinks "the people" are deciding on the existence, growth, or viability of government programs, you need to check your public choice premises. The point Tyler makes about public choice results versus a market tests is probably much more significant than you realize.
Posted by: M. Hodak at Sep 11, 2007 10:08:30 AM
I would think the deadweight loss of private giving is pretty close to zero. If I give money to a homeless shelter than I get to see the money actually benefit the people. Since I gave the money to the charity voluntarily, my utitlity for giving that money must have been higher than the equivalent amount of consumption.
When I pay taxes I have no way of knowing if the money is going for a welfare queen's crack habit or to buy bombs to blow up little kids.
Is there something I am missing here?
Posted by: sourcreamus at Sep 11, 2007 11:43:42 AM
No one is proposing an across-the-board increase in marginal rates. If you want to raise taxes, the trick is to raise the ones that have the smallest deadweight loss, alongside whatever equity concens you might have. Undoing W's tax cuts for rich folk would be a good start on both counts.
mk, the deadweight loss (if it is calculated correctly) is the loss *after* the benefit from the increase in leisure has been taken into account.
Posted by: David J. Balan at Sep 11, 2007 11:59:30 AM
What really bothers me about this line of argument ("government just gets bigger!") is that it's used by "starve the government" advocates to argue for lowering taxes without concern for which programs get cut. If you really care about economic efficiency, you should be arguing which programs have outlived their usefulness.
But it does just get bigger, every year, regardless of whether anything works. When was the last time the government spent less money than the year before? Here's a short list, just 10 ways to save money.
Just cut their budgets. They'll figure out how to spend the money efficiently. Necessity is the mother of invention.
Posted by: 8 at Sep 11, 2007 12:05:56 PM
M Hodak- For those who don't have the lingo down, what is the "point Tyler makes about public choice results versus a market tests" Is it basically that a contribution to a charity is not an irreversible investment, but, say, subsidies for dairy farmers has proven to be irreversible?
Posted by: guy in the veal calf office at Sep 11, 2007 12:26:49 PM
David, can you cite where the highest tax brackets have the smallest deadweight loss. It doesn't seem to me to be perfectly intuitive. I could certainly imagine a scenario where someone who is already at the highest tax rate would be willing to expend much more effort to avoid the coercive elements of taxation and be in a better position to dial up his leisure hours. Conversely, it would be the person who files a 1040-EZ and overpays his withholding that would have the lowest administrative cost. Last I checked, the IRS does cost-benefit analysis and sets their audit sights accordingly.
Posted by: Aschkan at Sep 11, 2007 12:32:02 PM
Aschkan is right and David_J._Balan is wrong. The DWL is highest for the rich: their decision to earn
additional income is much more sensitive to tax rates becausebeing rich, by definition, they don't
*have to* work. The DWL is the loss in a) shifting to leisure instead of lower tax + work, plus b)
the would-be recipient of that work who has to find someone else.
Posted by: Person at Sep 11, 2007 12:40:28 PM
Government bureaucrats don't participate economically if a government program they are running is operated more efficiently. They receive economic benefits if their program becomes larger with greater financing. The easiest way to gain power, job security and pay raises is to increase the number of employees and departments under your control. A larger, better financed government is much more attractive than a small, more efficient government.
Posted by: Steve Roberts at Sep 11, 2007 12:44:44 PM
To use these estimates to argue we should cut taxes would entail comparing apples and oranges – government does things that the private sector doesn’t. Saying that the dead-weight loss of a tax amounts to 80% means that instead of getting an orange for an apple you are getting 20% of an orange for an apple. How is that an argument for reducing consumption of oranges?
Posted by: datacharmer at Sep 11, 2007 12:58:55 PM
And for dstevens who thinks "the people" are deciding on the existence, growth, or viability of government programs...
Spare me your cynical wisdom, please. I did not say I think "the people" decide this. What I hoped to say was that cutting funding to for a specific program is harder than cutting taxes. I think this is because cutting taxes can only hurts people indirectly, but cutting taxes benefits people directly.
When was the last time the government spent less money than the year before?
From 1992 to 2000 government spending as a percentage of GDP decreased. See
here. I suspect that real government spending per capita has decreased at some point in our nations history, but I have absolutely no proof of this.
Why did the period 1992 to 2000 see a decrease in government spending? This was probably due to the policies of George H. W. Bush to balance the budget, which were continued by Clinton. Keeping a balanced budget probably had something to do with it--without that cutting taxes might result in increased deficits rather than decreased spending. There was probably something about the political dynamic at that time that helped, too, but I'm not really able to speculate.
I really appreciate the link to 10 ways to save money. I like it when people stick their necks out and makes real proposals where to cut spending. Thanks for that.
Posted by: dstevens at Sep 11, 2007 1:02:36 PM
*In itself. I'm not talking about the price going from 1 to 5, I'm talking about the price staying constant at 5 all along.
Posted by: datacharmer at Sep 11, 2007 1:04:44 PM
There is disagreement over the size of deadweight loss. But it's at least $1.20 per dollar at the margin, and likely quite a bit more. That does change the cost-benefit analysis for providing public goods. It also points up the importance of getting rid of bad government (ag subsidies) if you want more credibility in selling your notion of good government (at least to me).
In the case of social welfare programs, you can eliminate or reduce deadweight loss by making the benefits from the program depend on the taxes paid. This is how unemployment insurance and social security work. Note that this means that your social insurance program is for insurance, and NOT redistribution from rich to poor (i.e., redistribution based on expected income - insurance is by definition redistribution based on shocks).
So it would be easy, for instance, to design a very efficient national health care system with a relatively low deadweight loss. People have to go to a randomly-assigned government doctor once or twice a year to be eligible. The doctor diagnoses them. In the event of a diagnosis of something bad, the person sees another randomly assigned doctor for a second opinion. If that diagnosis holds, they receive a direct payment from the government based on their illness. This payment consists of a fixed amount plus a function of the taxes the sick person has paid for the health system. They then spend the money any way they please.
Posted by: Keith at Sep 11, 2007 1:26:03 PM
Wouldn't the Hayekian inference be that static analysis and deadweight loss are pretty much unrelated to reality?
Posted by: Lee A. Arnold at Sep 11, 2007 1:39:34 PM
"Wouldn't the Hayekian inference be that static analysis and deadweight loss are pretty much unrelated to reality?"
No. Hayek had no problem with neoclassical economics, he just found it incomplete and not open to other approaches. But the supply-demand framework works well for modeling the distortions introduced by non-lump-sum taxes.
Maybe you could make a hyperbolic discounting argument that decoupling future social security payments from earnings wouldn't increase the deadweight loss from Social Security. That is, if you are genuinely interested in policy implications from bounded rationality and not just trying to make some lame "I hate libertarians" argument.
Posted by: Keith at Sep 11, 2007 2:15:21 PM
One policy argument from bounded rationality would be that institutions are necessary. Therefore, taxes may perform the function of reducing future costs, (and not just compounding interest on the government debt, which on balance 80% of us pay to the other 20%.) And certainly there are other clear distortions, such as tax-avoidance behavior. But if you look at it dynamically, deadweight cost has a long way to go to justify much complaint. Let's call it the "dynamic Coasian inference," since if you go beyond the initial "transaction" of the taxation, institutions (some sorts of which, taxes are used to pay for,) perform their good functions, (if and when they do!,) by some sort of organizational innovation that decreases the costs of other transactions -- transactions that are in the arena to which the institution is applied. It's analogous to the savings from innovations in general -- even the simplest hand-tools save force translation in spacetime. Up at our macroscopic level, economic transaction costs take up around 50% of the GDP, according to Coase. It could be that taxes usually show a dynamic "liveweight gain," all things considered -- which they never are. Ideas, innovations, institutions are the ways in which bounded rationality leaps its bounds. They ARE economic growth. Coase was clearly aware of the seed of this argument -- he thought it would turn economics upside-down. And it is at the basis of Hayek's spontaneous organization, of course.
Not to say that an institution can't go wrong -- much as an idea or innovation can. They are selected against in different ways, of course: markets vs. voting, and all that, and people become habituated and so on. Still, Tyler Cowen might clarify whether his phrase "most government programs" refers to the cardinal count, or to the spending amount. Because by expenditure, the two biggest programs by far are: Defense (it should be listed as Numero Uno: if you include all defense-related items from other programs, it is close to twice the regular budget figure) and of course Social Security. Put together, if we include the expanded definition of defense spending, that's over half of the federal budget. Add Medicare, Medicaid, and debt service, and you're over 80% of all federal spending. Surely it is no one's contention that these are (entirely) "ineffective or harmful." (Unless perhaps, it is a libertarian's!)
Posted by: Lee A. Arnold at Sep 11, 2007 3:37:59 PM
The point I made above that one would not want to raise all marginal tax rates equally, but rather would want to choose which taxes to raise so as to minimize deadweight loss while still paying attention to distributional concerns is correct. And it is true that raising taxes on rich people would be appealing distribution-wise. I had thought I read that dwl was relatively small for high-income workers (i.e., labor supply for such workers was relatively inelastic), but I'm not sure about this, so I have to partially retract what I said above. Does anyone know what the relevant literature says?
Posted by: David J. Balan at Sep 11, 2007 4:26:25 PM
Tyler, do you realize that these numbers were arrived at through a tax simulator, which Mr. Feldman programmed to say that for every 10% reduction in marginal tax rates, the tax base would increase by 16%?
The rest of it's just mathematical masturbation - he made up a number for economic growth as a function of tax rates, and he's calling the output of the simulation, based on his made-up elasticity number, a "fact."
Garbage in, garbage out. By the way, the paper where he presents his "evidence" is riddled with typos. It's also pay-to-read, but a little bit of Googling can find it for free.
Posted by: Roger at Sep 11, 2007 6:58:49 PM
