18 Mar 2013

Structural failure at the Corps of Engineers

The US Army Corps of Engineers Primer on water economics [pdf] highlights two sources of dysfunction that have been imposed on USACE.

First, USACE evaluates the costs and benefits of projects according to national and regional accounts [p. 3]:
The National Economic Development (NED) Account displays changes in the economic value of the national output of goods and services.

The Regional Economic Development (RED) Account displays the regional and localized economic impacts that result from each alternative plan.
But RED accounting is biased [p 6]:
From the federal perspective transfers are a zero sum game. That does not mean that RED benefits are not significant or important to the non-federal partner. The distinction between RED and NED is a matter of perspective and policy, not economics.
RED accounts, in other words, facilitate political, zero-sum games among political actors seeking to rob Peter to pay Paul.

Second, the Corps mixes up two different methods of comparing costs and benefits [p. 17]:
Net benefits [NB] represent the amount of total benefits less the total costs. This analysis is used to select and scale a recommended course of action from an array of alternatives.

A benefit-cost ratio [BCR] tells us which alternative produces the most benefits for every dollar of cost (total benefits/total costs). The benefit-cost ratio is useful for comparing or ranking different projects. Once the optimal scale of the alternative is identified by measuring net benefits the benefit-cost ratio can be used to rank among competing investments.
In other words, the Corps chooses one project in one region according to NB and then compares that project to other NB winners in other regions according to BCR to choose which projects to fund.

This method is both unfair and inefficient. It's unfair (assuming "fair" means projects in all regions) because a NB winner can still be defeated by BCR. It's inefficient because high BCR projects with low NB will be eliminated at the regional level before they can be considered/chosen at the national level.

Even worse, the NB method assumes unlimited budgets, since a $1 billion project creating $100 in gains is "more attractive" than a $150 project creating $50 in gains. Fail.

What the Corps should do* is take all projects in all regions, rank them by BCR, and then fund/implement them for as long as their budget lasts.

There's still the problems of bad BCR (regional BCR estimates ARE biased; it's a big game of liar's poker) as well as the problem of buy in from politicians who want to see Corps projects in their back yards. My solution for the former problem is to use the cost-sharing/signalling method I proposed for the stimulus. My solution for the political problem is to eliminate RED accounts.**

I could say a lot more on Corps operations (there was a rash of articles on the cessation of Mississippi shipping due to low water levels and everyone complaining that the Corps was failing to manage water levels, channels, etc.), but my point here is that political bickering prevents the Corps from doing its job.

Bottom Line: The Corps will fail as long as it uses Congressionally distorted measures.

* I have no doubt that they know they should do this, but they are required to follow these procedures established by Congress, so the blame for Corps failure can easily be shifted to a feckless Congress.

** If the goal is to spread spending evenly, then don't collect the taxes OR spend the money. Federal programs that transfer money from one region to another MUST create national/regional (>state) benefits -- unless they're theft.