Thursday, November 18, 2010


The idea is rapidly spreading that a ban on earmarks doesn't affect spending, since earmarks are a way of distributing what's already been appropriated.

This is just true enough to be clever, and marks the speaker as being more sophisticated than those Tea Party rubes. But it's basically false, for three reasons.

First, it is more expensive to do things inefficiently than to do things efficiently. Building the Ted Stevens Bridge To Nowhere or the Robert Byrd Gold-Plated NORAD Auxiliary High Command Of West Virginia means that money has simply been wasted, and that all the needs that weren't met this year will arise again next year. If the real needs exert at least some pull on appropriations levels, then wasting money rather than spending it wisely at time 1 does affect appropriations at time 2. The U.S. gets very bad value per dollar of federal infrastructure spending, in part because earmarks screw up the ability to prioritize projects. That doesn't increase the appropriations at time 1; but it does tend to drive them up in every later year. Similarly, when earmarks keep alive weapons systems that the Pentagon wants to cancel, because the defense appropriators in Congress view the defense budget as a jobs program, the Pentagon shrugs its shoulders and increase its request the following year; it's not going to let the wasteful jobs-program part of the budget displace its own military priorities.

Second, bills often emerge out of House-Senate committees with higher appropriations levels that have the express aim of smoothing passage with earmarks.

But third, and most important: the earmarking members of Congress are the same people who set the appropriations level. And by this I don't only mean that they're members of the House and Senate; I mean that they're powerful members of the relevant committees. Ted Stevens and Robert Byrd took turns chairing the Senate Appropriations Committee. The knowledge that they were going to have a chance to start shoveling pork a little bit later in the process affected how much they appropriated at the beginning.

The idea that earmarks don't affect spending levels rests on a crazy image of how appropriations levels are set. We don't have one set of legislators who are dispassionate, disinterested judges of how much money needs to be allocated, who are then later on replaced by a bunch of grubby politicos deciding how to divvy up the spoils. Neither do we have legislators who, during their initial appropriations deliberations, somehow forget that earmarking comes later. Instead, we have normal human legislators throughout, responding to their incentives and environment. It would take a kind of saintly self-denial for them not to increase the initial size of the pool knowing that they were going to get a chance to give themselves a share later on.

Republican earmark supporters have been saying that abolishing earmarks transfers allocation authority to the dreaded Obama Administration. Well, yes. And if you tell a bunch of Republican legislators to decide how much money to give to the Obama administration to allocate, they'll come up with a smaller number than if you tell them to come up with an amount for them to divide up among themselves to allocate. Indeed the same holds true for members of the President's own party.

Earmarks aren't themselves a lot of money in the grand scheme of things, and abolishing them entirely would only make a tiny dent in the deficit. But they do indeed affect appropriations-- and my hunch is that they affect appropriations for more than their actual cash value, because they create a system that attracts appropriators like Byrd and Stevens, who err on the side of spending too much to make sure there's enough to go around.


Anonymous said...

I absolutely refuse to believe point two without seeing empirical evidence. You have any peer-reviewed studies that demonstrate over-appropriation as a result of earmarking behavior?

Jacob T. Levy said...

Anonymous, are you objecting to point two or to point three?

Anonymous said...

Without data, this is pretty meaningless. You have to prove that it does indeed raise spending, and then find the magnitude of the increase.

Boonton said...

What about in terms of smoothing spending cuts? For example, consider the health care bill that, at one point, gave one state extra money for Medicaid in exchange for on Senator's vote.

Consider one of the most potent political arguments against the bill was that it would cut Medicare spending. Assume for the sake of argument tha the Medicare cuts are sensible, but unpopular and the extra help for the one state is less than the savings from the cuts. The earmark facilitates reduced spending that would have otherwise been unpopular.

Likewise it's hard to imagine a bill that does something really unpopular like raising the retirement age not sailing through without some earmarks of some sort easing its path.

taw said...

The post is based on premise that on average earmarks are dumber that what would come out of executive. I see no evidence for it - both have about as many really stupid allocations.

Unknown said...

If Anonymous' desire for evidence is directed at the concept that bills are passed out of committee with the plan for passage directly related to pork, there is tons of evidence. Here is one: Ben Nelson of Nebraska achieved one of the largest pork payoffs in history in explicit exchange for his vote to pass ObamaCare.

If Anonymous directs his criticism at the concept that initial appropriations are influenced by (explicit or implicit) strategic planning for later earmarks, that is a harder point. However, this can be extracted to a series of variants of the Ultimatum Game. For instance, a repeated UG with multiple proposers and responders. In this case, the results will skew toward increasing pie size.

Jacob T. Levy said...

taw, consider university research funding. "Executive branch" means that grants go through peer review at NSF, NIH, or NEH, and then must still pass through a round of priority-setting by a committee of agency official and scholars comparing all of the peer-review-approved projects.

Earmarks mean that research money is allocated such that four of the top ten university recipients are part of the Mississippi state university system because Thad Cochran is a powerful appropriator.

You don't have to think that peer review is perfect to think that the peer reviewers are a lot more likely to fund promising and important research by capable scholars at universities with appropriate support infrastructure than is Thad Cochran.

Anonymous 1 and 2, you're right; I have no evidence for this, other than casual observation of the dynamics of the appropriations process, and the annual process by which the initial budget resolution gets turned into actual appropriations by the committees-- how, for example, the big transportation bills grow over the course of their passage. It's not clear to me how one would get suitably powerful evidence, without, say, abolishing earmarks for half the budget categories in a given year and seeing whether the dynamics differed in those categories.

But I do not think that the claim that earmarks *don't* increase appropriations is entitled to any presumption in its favor, any more than the claim that increasing the price of bread has no effect on quantity demanded. The incentives and the logic of the system give us pretty good reason to expect an increase; and what we see over the course of the appropriations process is compatible with that logic.

All the "no increase" argument has going for it is the statement that earmarking happens later in the process than appropriations. The aim of my post is to show that that procedural fact doesn't really matter. If you want to view the matter as empirically unsettled, that's fine-- but please acknowlegde that Inouye, McConnell, et. al. haven't provided any evidence for the "no effect" claim either.

TGGP said...

Garett Jones thinks the opposite. We're going to need more data.

Jacob T. Levy said...

Jones' model described here assumes a Congressional leadership that is determined to rein in entitlement spending and looking to bribe members to cast such a hard vote.

But why should we think that the Congressional leadership places a high priority on limiting entitlement spending?

He's offered a thought experiment about what such anti-spending leaders would need to do; he hasn't actually claimed that the dynamic holds now. "At some point, this or the next decade, Congress is going to wrestle with entitlement spending," and *then* they'll need earmarks.

steve said...

You should go through the process of how money is allocated that is not controlled by earmarks. Without offering readers a fuller picture, you are being disingenuous here.

Query-During the 60s and 70s when there were many fewer earmarks, was there less porkbarrel spending? How was money distributed to political allies during that time?

Finally, why not concentrate on real spending? Here is how I will know Republicans are serious about spending. When they repeal Medicare Part D, which they passed. The largest, IIRC, unfunded spending bill ever passed.



Heron said...

Abolishing earmarks would also be a rather despicable abandonment of responsibility on the part of the Legislative Branch. It is their job, not the president's, to determine how money is collected, and how it is spent. To give that authority to the Presidency; to throw their hands in the air and say "fine, we'll just raise money and had it over to the Admins, let them sort it out," would unbalance our government even further towards tyranny (in the old Greek sense of centralized, solitary rule) than it already is.

Jacob T. Levy said...

The relationship between the legislature and the executive is *always* one of general-to-particular. The legislature passes general criminal statutes; the executive detects and prosecutes particular violations. That's why we prohibit bills of attainder. According to classical separations-of-powers doctrine, it is precisely this feature that prevents tyranny: the same agency doesn't make general rules and apply them to particular cases.

Your budgeting model generalized-- if the whole budget were made up of tax earmarks on one side ("Tax Bill Gates for $15 billion, Warren Buffet for $10 billion, Joe Smith for $234") and spending earmarks on the other side would mean a catastrophic breakdown of the rule of law and the separation of powers. The tax earmark side would represent a way around the bill of attainder prohibition; the spending earmark side would mean that the legislature abandoned the idea of having *purposes* for its spending, and that it spent only as a means of favoring particular persons.

There's no chance of things getting that bad; earmarks are a marginal part of the budget on both sides, though "marginal" doesn't equal "unimportant." But that's the separation-of-powers logic of your model.

By contrast: "Out of the general nationwide tax of 20% on income, we allocate $10 billion for cancer research, to be allocated in accordance with an independent system of peer-review for the most meritorious proposals"-- that's perfectly good general lawmaking, accords with the rule of law, and requires a separate stage of executive allocation.

From Montesquieu to Rousseau to Hayek, rule of law theorists have *always* been clear that legislating with proper nouns, legislating for the benefit of punishment of particular persons rather than for the promotion of general rules and purposes, is the great evil. And "special legislation" was a term of abuse in federal and state constitutional law for three generations after the founding-- the antonym of "equal protection of the laws."

There is *absolutely* no constitutional obligation to earmark, and the system of earmarks exists in clear tension with the underlying rules and norms of the constitutional order.