Infrastructure permitting reform is so hot right now…
And why not? From housing to hyperscalers, pundits to podcasters, the abundance agenda is IN.
This is a good thing. But I will say, as in all trends, the newbies Columbus-ing in think they’re the first to arrive, and the old-time scenesters like me can’t wait to get all “back in my day…” with it.
Background
In 2015 we passed S. 280 (Federal Permitting Improvement Act),2 better known as FAST-41 (i.e., Title 41 of a larger transportation bill). I don’t see it get mentioned much. Folks over at FAI have been doing yeoman’s work on this, especially state reforms, but even there a recent report gets a single mention on p. 7 in a table.
That said, I hope it was meaningful and wanted to know if we could claim any impact. The primary delay in federal permits is sequential reviews across multiple agencies. Among other things, we established a presumption of concurrent review with a “lead agency” responsible for keeping it moving.
(ed. Cut to the chase. Did it help?)
Results
I used Gemini heavily to gather/organize data, draw estimates for that data, test refinements, and iterate the best way to categorize results. Assumptions are mine built on that data.
These are strictly back-of-the-envelope calculations of the economic gains.
The intuition: reform benefits infrastructure investment (capital expenditures) through a Weighted Average Cost of Capital (WACC). Every month of delay reduces the potential return on investment by prolonging the beginning of it’s productive deployment. Inflation is effectively chipping away at the net present value of a given investment so every month you shave, it goes sooner to the return-generating use it is intended.
Of course FAST-41 did not operate in a stick-and-hold policy environment. I further attempted to break down the role of follow-on and other factors during the ten-year period.
Methodology
Assumptions:
Baseline drag (11%): Even if the risk-free interest rate is 0, there’s still opportunity and liquidity costs of standing idle.
First there’s some equity risk premium you need to clear. If I can make as much or more as an equity investment elsewhere the project is infeasible. ~6.5%3
Second, major projects are not start-stop in nature. I have to contract out for major equipment, expert and specialized labor, property deals, etc. (certainly lawyers on retainer, hah!). There’s some non-zero cost to all that. I put it, somewhat arbitrarily at ~3.0%.4
There’s a liquidity cost/friction where capital is booked as intended to be deployed in some given period, but now has to push back. There could be direct marginal borrowing costs, portfolio reallocations, normal cash flow stuff. ~1.5%
Financing Multiplier (1.75): A constant reflecting the scaling costs of borrowing. Some equity financing is often present, but financing is almost always primarily debt instruments. A multiplier of 1.0 would mean the project is financed at strictly the risk-free rate, which if it is, great! But no one’s lending at that price because of course there’s risk! So every increase in the risk-free rate shows up additionally here: credit spreads, rate swaps, and debt service terms all increase faster than the baseline rate.
The rest is verifiable from available data. The cost estimation looks like this:
C: Capital expenditure for covered projects = $200B based on the federal project dashboard.
t: net reduction in permitting delay (years) = 1.5 based on permit council data, McKinsey report, and recent GWU Reg Studies Center study.
Rf: risk-free cost of capital. Treated as one rate each over two periods: Low - 1.75% (2016-2021) and High - 4.10% (2022-2026).
Discussion
As I mentioned, the law established a presumption of concurrent reviews when possible. I’m skipping the specific incentives because it’s boring. The tweet-length version is the law did two major things:
A public federal project permitting dashboard w/ public data
A mechanism for interagency coordination, with accountability for a “lead agency” to drive the process
Before there was a lot of benign patience where one agency would wait for another to finish and then pick up there. Say a DOT project will abut a BLM property — both have to sign off or it’s not approved. What it didn’t do is change anything in the review standards (NEPA’s environmental impact studies). But there was a clear arbitrage opportunity where multiple agencies have overlapping data and analytical needs. One task imported to multiple destinations.
Additional Points
It also reduced a statute of limitations for challenges from 7 years to 2. Most importantly it established a proof-of-concept infrastructure within the government to build on. The law did not cover every project permitting need or category. But over time other laws expanded the scope and fed more opportunities into an efficiency machine, as in the second table.
Ultimately the interest rate environment is the single biggest proximate driver of economic gains, but remember it’s the costs saved from shorter approval time in that environment that is the ultimate driver.
My assumptions lean conservative but fairly easy to adjust accordingly. Though any benefits are justified because they come at essentially zero cost. The delays are deadweight loss: they do not benefit the investors, builders, local users, or even the environment. Essentially vaporized opportunity.
Something about bugs?
I’ve excluded some “under-the-hood” analysis resulting in the 1.5 years median reduction estimate. One finding is worth flagging.
Earlier I referenced a bugs v. bureaucrat story. As Thomas describes, a forest fire happened on Oregon tribe land. Working quickly the tribe decided to recover what was useable and clear the debris. Waiting means bugs take over and the wood loses value. Conditions also become more hazardous.
Well the Bureau of Indian Affairs (BIA) issued a cease-and-desist; eight months went by. The wood had very little value now.5
I mention this because in the data BIA is consistently a major bottleneck.
Ex: median right-of-way approval time for BLM is 11.9 months v. BIA is 43.8 months (> 3.6x longer).
The author is nonresident senior fellow at the Foundation for American Innovation.
Ok, not exactly the same thing. But relevant, as I’ll explain. In any case, Thomas Stratmann is one of the most careful empirical analysts I know — worth following regardless.
There’s my little committee report. Codified at 42 USC 4370m et. seq.
Admittedly I’m a little defensive:
‘Twas the night before Thanksgiving, and all through the Senate, not a staffer was working, #becauseRecess.
Sen. Portman’s 2x Supreme Court clerk and I sat in my office negotiating over speaker phone with the conference committee. Between callbacks, we’d pull from a righteous whiskey I’d been saving for final passage.
He went on to a high six-figure law firm signing bonus and I’m writing on Substack.
Canonical methodologies (historical-looking) put this in the 6-7.5% range. See, e.g., Mehra and Prescott (1985); Ibbotson and Sinquefield (1976). Finance field not my strength and admittedly these are pretty old. Open to adjustment and of course your preferred flavor of ERP is simple enough to input.
Again, relying on Gemini for data gathering on useful benchmarks, it breaks down roughly in equal parts to: (1) Option renewals on property and access rights (American Council on Renewable Energy and National Renewable Energy Lab) (; (2) Engineering, Procurement, and Construction (EPC) firm holds; and (3) other retainers and price escalation (BLS Producer Price Index - Heavy & Civil Engineering Construction).
Some went into a display at PDX, so not a total loss.