Here’s an article from the Chicago Tribune in which Transportation Secretary Ray LaHood is quoted as saying “there is no favoritism” involved in the disbursement of the transportation stimulus funding to the states. The reason given was that the disbursements are “based on a long-standing formula for allocating highway funds to the states.”
Using their system eliminates some forms of bias, but there is no reason bias can’t be built into the system itself.
Apparently, if there is a formula involved we are getting an outcome that is impartial, effective, and fair. After all, a formula involves fancy math, and anything involving fancy math must be true. Right?
In this case the answer is “yes” – but only to a degree. The use of formulas has taken some of the arbitrariness and the worst forms of politicking out of the transportation spending process. Moreover, the application of existing formulas was certainly the only politically feasible way to push such a large and complex spending program through Congress with the lightning speed that was required of a stimulus measure.
But on the other hand, the formula was not handed to us on stone tablets: humans, with all their imperfections, devised it. Using their system eliminates some forms of bias, but there is no reason bias can’t be built into the system itself.
Here’s one of the two formulas that determine the way most of the highway funds are being distributed:
* 25 percent based on total lane miles of federal-aid highways.
* 40 percent based on vehicle miles traveled on lanes of federal-aid highways.
* 35 percent based on estimated tax payments (e.g. from fuel taxes) attributable to highway users in the states into the Highway Account of the Highway Trust Fund (often referred to as “contributions” to the Highway Account).
* No state can receive less than one half of one percent of the spending.
No doubt this descent into the minutiae of transportation finance allocation has sent many of you scurrying off to read about Marijuana Pepsi, electric cars, or Mike Zarren‘s take on basketball statistics by now. But for those hearty souls who remain, can you spot the (at least) five ways that ineffectiveness, inefficiency, and (depending on one’s values) inequity are built into the process?
My answers next time. Yours?
and 
. You can also watch it on 
Two papers you should have brought into the discussion:
Knight, Brian “Parochial Interests and the Centralized Provision of Local Public Goods: Evidence from Congressional Voting on Transportation Projects” JPubE 2004
Knight, Brian “Legislative Representation, Bargaining Power, and the Distribution of Federal Funds: Evidence from the US Senate” Economic Journal.
Looks like politics still plays a substantial role.
Here’s my 5:
(1) Rather than spending federal aid to repair roads, build new ones – which increases future aid-allocations based on criterion a.
(2) Widen highways that don’t need to be widened to increase the number of lane-miles of highway to increase aid based on criterion a.
(3) Under-allocate funds to local roads, thus encouraging additional highway usage which increases aid-allocations based on criterion b.
(4) Discourage state-level fuel-efficiency standards (and encourage the purchase of gas-guzzlers by residents) which increases the tax-payments per mile driven – thus increasing aid-allocations based on criterion c.
(5) States with very few roads (that are under the 0.5% minimum) have incentives to spend aid intended for transportation elsewhere rather than on roads, given that they won’t increase aid by spending more on roads given criterion d.
1. Since the funds are partly allotted based on the existing miles of federal-aid highways, previous inefficiencies are propagated.
2. Funds based on vehicle-miles traveled on federal aid highways are almost as bad as #1 (being dependent on existing highways), though they are a little bit fairer, since it is essentially a use-weighted average over existing federal highways.
3. Giving 1/2 a percent to each state, minimum, will create overspending in places without as much need for additional federal aid highways.
4. Estimated tax payment sources may or may not correlate with a need for additional highways.
5. This formula does not attempt to use the stimulus money where it will do the most to reduce the sum of all automobile transit times in the vicinity of the project, to decrease wear and tear on vehicles traveling in the vicinity of the project, or in any way to create something of economic value to people living in the vicinity of the projects.
6. This project doesn’t question that the best way to connect distant locations is through additional highways. In some areas, for example, bike paths might be a cheaper (in terms of reducing the sum of all transit times per federal dollar spent).
I’ll give it a shot:
1. “25 percent based on total lane miles of federal-aid highways.” Current funding is based on past funding, which does not reflect usage and which may have been under “unfair” rules.
2. “40 percent based on vehicle miles traveled on lanes of federal-aid highways.” These are estimates which may be very inaccurate. In addition, if estimates are made by state officials then they are incented to inflate the numbers.
3. “No state can receive less than one half of one percent of the spending.” Obviously favors very small states irrespective of need.
4. “35 percent based on estimated tax payments” States with more money may get more from the government. Not necessarily wrong, but not necessarily putting investment where’s it needed.
5. There’s the 25/40/35 weighting. A poorer, state with few miles, that might be considered “drive over country” may not get enough funds to maintain its roads because of the weighting system.
If you have received the benefit of massive pork in the past you have a higher percentage of federal highways in the present.
I can think of 3:
1. The minimum of 0.5% seems too high. If the funds were divided equally among the states, each would get 2%.
2. Compounding. If you have a high number of highway miles, you will get a lot of funds, allowing you to build more highway miles, increasing your share even more.
3. Highway miles vs. miles traveled. I can imagine a lot of states have motivation to build empty highways.
(knuckles cracking) Ok – I’ll have a go…
1. The percentages chosen
2. “Estimated” is never a good word to use in a formula
3. Who sets those tax payments?
4. If you add in the minimums states can receive, isn’t that more than 100%?
5. Does the fact that fewer vehicles might be using a highway make it less important?
6. Are longer highways more important OR cheaper?
- d
1. Since future funds are distributed according to lane miles (basically square miles) of asphalt, there is a significant incentive to pave.
2. With the vehicle mile requirement, mass transportation (non-highway) spending has a lower ROI, as future highway funds will disappear.
3. The tax calculation seems redundant with the vehicle mile requirement, effectively doubling that input, while lowering the revenue-generating (and emission reduction) benefits of higher state gas taxes.
4. The minimum funding requirement (0.5%) means low-density flyover states (think North Dakota) will be spent on more ineffective projects.
5. Appears biased to intra-state commerce (more lanes around cities and to suburbs) than interstate commerce, which is arguably where gains from trade are primarily generated.
While we can’t solve everything, an elegant solution to number 4 would be to allow North Dakota et al. to auction their highway fund allocation to other states for general fund cash.