Four Levers That Decide Where a Campus Lands

It helps to hold the whole terrain in view before walking any single path. Four distinct levers shape the tax economics of a data center, and they operate at different levels of government with different clocks.

The first is state and local incentives: property-tax abatements, sales-and-use tax exemptions on equipment, and negotiated agreements between a developer and a county or municipality. This lever is the most decisive for siting because property and sales taxes are large, recurring, and local — a county assessor and a state legislature control them, and they compete openly for projects.

The second is federal tax policy. Washington does not usually pick your county, but it shapes the after-tax return on the whole asset class through credits and depreciation rules. The clean-electricity credits enacted in the 2022 Inflation Reduction Act, the semiconductor manufacturing credit created by the CHIPS Act, and a quieter change to how research costs are deducted all feed into the model. These do not move a campus across a state line, but they move the hurdle rate for building anything at all.

The third is international tax. For operators with global footprints, where profit is booked — and how it is taxed across borders — is its own discipline, now reshaped by a coordinated push for a global minimum tax. A campus in Ireland or Singapore is partly a tax-planning decision, not only an infrastructure one.

The fourth is deal structuring: the choice of legal and capital wrapper around the asset. Whether a portfolio is held in a real-estate investment trust, financed with tax-advantaged debt, or moved through a like-kind exchange changes the tax outcome without changing a single server.

Faraday's read: most public commentary collapses these four into one word, 'incentives,' and loses the plot. They have different decision-makers, different time horizons, and different failure modes. State and local relief is negotiated and clawback-prone. Federal credits are statutory and stroke-of-the-pen fragile. International structure is exposed to multilateral policy. Deal structure is the operator's own choice. Keep them separate, and a confusing headline resolves into four readable stories. The rest of this course takes them one at a time.

A note on reading across the four levers at once, because the interactions matter. A federal credit that improves the economics of on-site generation can make a state with abundant land but a weak grid suddenly competitive, because the operator plans to build its own power and the credit helps finance it. A global minimum tax that erodes a low-tax jurisdiction's appeal can push marginal capacity toward markets chosen on power and land rather than profit-booking. And an operator's decision to hold assets in a real-estate investment trust changes how much a given incentive is even worth to it. Faraday's read: the four levers are not a checklist to be summed but a system to be read, where a move on one changes the value of the others. The analyst who keeps them separate for clarity, then studies how they interact, will consistently see siting outcomes before the market does — because most observers are still treating 'incentives' as a single undifferentiated number. A final habit worth building early: for any campus in the news, sketch which of the four levers the story is really about before forming a view. Most coverage names one lever and implies it is the whole picture. Naming the lever correctly is half the analysis.

As a working exercise, take any campus in the headlines and try to assign its story to one of the four levers before reading further: is this a county-abatement story, a federal-credit story, an international-structure story, or a capital-structure story? You will find that the correct label often contradicts the framing of the coverage, which tends to reach for whichever lever makes the cleanest narrative. Getting the label right is not pedantry; it determines which decision-maker, which clock, and which failure mode you should be watching. A story mislabeled as 'incentives' in general is a story you cannot actually forecast, because you do not know whose pen controls the outcome.