State Taxation of Data Centers

Key Findings

• Data centers' state and local taxA tax is a mandatory payment or charge collected by local, state, and national governments from individuals or businesses to cover the costs of general government services, goods, and activities.burdens are heavily dependent upon policy choices surrounding sales and property taxation of data center equipment.
• Servers and other data center equipment are exempt from sales taxA sales tax is levied on retail sales of goods and services and, ideally, should apply to all final consumption with few exemptions. Many governments exempt goods like groceries; base broadening, such as including groceries, could keep rates lower. A sales tax should exempt business-to-business transactions which, when taxed, cause tax pyramiding. in most states, consistent with the principle of avoiding the taxation of business inputs, but the exemption is often contingent upon meeting economic development targets.
• Many states impose tangible personal property taxes on data centers' machinery and equipment, accounting for more than 20 percent of the total federal, state, and local tax burden on data centers across 12 jurisdictions examined in this publication.
• Data centers face high tax burdens and are particularly substantial contributors to local coffers, but poor tax structure can drive these operations to other locations and deprive local governments of a major revenue stream.
• Tax liability calculations for a model $1 billion data center in 12 representative jurisdictions across the country demonstrate how important tax considerations can be in location decision-making.

Introduction

The AI breakout is supercharging an already booming data center industry, with estimates that US data center investments will exceed $1 trillion over the next five years.[1] Even this may not be enough: one industry estimate suggests that the largest players in the industry-the so-called "hyperscalers"-will need to deploy an additional $1.8 trillion in capital by 2030 to keep up with demand.[2] These investments are coming. The only question is where they are located, and state tax policy choices can play a meaningful role in that decision.

Data centers represent sizable investments, with larger campuses costing billions of dollars. Investments of that size are serious business, and decisions about where to locate these facilities are not made lightly. Companies consider a variety of factors, including electricity costs and grid reliability, climate and cooling costs, land availability, colocation with or proximity to users, regulatory environments, and, of course, taxes.

None of these factors is sufficient on its own, but scoring poorly on one or more can take a region out of the running for major data center expansion. A cooler climate reduces equipment cooling costs, but if electricity is unreliable or unduly expensive, even at reduced loads due to favorable weather, an area will be unattractive for new data centers. Restrictive land use regulations can make it difficult to locate new facilities in some areas, even if utility costs and taxes are attractive. And aggressive taxation of data centers can harm a state's prospects of attracting data center investment even if other conditions are favorable.

States can't legislate their climate, and some regions lack the available open land for data center scalability. Lower electricity costs or greater grid resilience may flow from choices made by policymakers, but they can't just flip a switch to accomplish this. Tax structures, by contrast, are the policy lever most within reach of state lawmakers.

If lawmakers want data center investment in their states, they do need to ensure that the tax code does not uniquely penalize these operations, as is often the case. Decisions on whether to tax data centers' machinery equipment purchases or ongoing use, the degree to which capital investment costs can be expensed under corporate income taxA corporate income tax (CIT) is levied by federal and state governments on business profits. Many companies are not subject to the CIT because they are taxed as pass-through businesses, with income reportable under the individual income tax.regimes, whether and how to tax business utilities, and whether to impose taxes on data processing all influence businesses' decisions about where to locate new projects.

This paper examines how states competing for data centers approach these different tax choices and how tax policy influences data center investment. It outlines the proper tax treatment of the industry, then offers a 12-state comparison of state and local data center taxation.

Proper Tax Treatment of Data Centers

The appropriate tax treatment of data centers is no different than the appropriate treatment of any other industry, since, ideally, tax policy should be neutral toward different classes of investment. In some cases, states' tax treatment of data centers diverges from their policies toward other businesses. In other cases, however, shortcomings in the broader tax code take on outsized importance for data centers due to their economic profile.

For instance, sales taxes should not be imposed on business machinery and equipment (M&E). That is equally true whether the business in question is a data center, a retailer, or an insurance agency. The implications of that choice, however, will matter far more for the business with potentially a billion dollars or more in machinery and equipment purchases, and which needs to cycle in new equipment on a regular basis.

Sales Tax on Business Inputs

An ideal sales tax falls exclusively on final consumption, but no state's actual sales tax lives up to that ideal. To varying degrees, all sales tax bases include at least some business inputs. In doing so, states transform that portion of the sales tax from a tax on consumption to a tax on capital investment. It takes a tax that is neutral regarding in-state investment and turns it into a discriminatory tax on a state's own businesses.

Because most sales taxes are destination-based (meaning that they are imposed where a product is purchased or used, rather than at the location of the producer or seller), sales taxes on final consumption do not impede a business's ability to compete with out-of-state competitors, since out-of-state customers are taxed at their own local rate, or are legally required to remit use taxes at their own local rate, not at the rate in the business's jurisdiction. As soon as taxes are imposed on a business's own purchases, however, businesses in that jurisdiction are placed at a disadvantage against competitors not subject to such taxes in their own states. These taxes represent an additional cost of production that is not borne by their competitors based elsewhere, even if they sell into the same markets.[3]

For data centers, sales taxation of machinery and equipment (M&E) is particularly significant due to the sheer scale of capital investment and the frequency with which servers and network equipment must be replaced to keep up with technological developments. Data centers associated with so-called "hyperscalers" (the major cloud computing providers) begin at about $1 billion in initial investment and easily exceed $5 billion for large new AI-centered projects, with Meta breaking ground on-and breaking records with-a $10 billion data center in Louisiana.[4] Even investments not involving the industry's largest players can easily reach $1 billion per project now.

Some capital expenditures-electrical systems (including switchgear), mechanical systems (particularly cooling systems), batteries, and structures-have usable lives in excess of 20 years. Most of a data center's capital expense, however, is servers and networking equipment, which must be replaced much more frequently. Traditional cloud computing operations typically have a five- to six-year equipment replacement cycle; for AI-oriented builds, that cycle is often shortened to three years. A $5 billion data center could easily spend more than a billion dollars a year on machinery and equipment. At such scale, its equipment's exposure to sales taxation is a significant consideration.

In some states, machinery and equipment are exempt from sales taxation by default. Other states theoretically tax data centers' M&E-sometimes because their M&E exemption only applies to manufacturers-but provide specific exemptions for data centers' M&E, provided the operations meet investment or job creation criteria. And in still other states, data centers pay sales tax on their M&E purchases, increasing their cost of doing business.

When sales tax exemptions operate as a targeted "incentive" rather than default treatment, this introduces complexity and undermines tax neutrality. In South Carolina,for instance, the sales tax exemptionA tax exemption excludes certain income, revenue, or even taxpayers from tax altogether. For example, nonprofits that fulfill certain requirements are granted tax-exempt status by the Internal Revenue Service (IRS), preventing them from having to pay income tax.is contingent upon newly created jobs being maintained for three consecutive years,[5] while in Texas, only data centers with at least $500 million in capital investment, contracts for at least 20 megawatts of transmission capacity, and 40 qualifying jobs are eligible for local sales tax exemptions, though a wider range of qualifying data centers have their equipment purchases exempt from state sales tax.[6] Job creation, investment, wage rate, and even square footage requirements are common.[7]

Where sales tax exemptions operate as a targeted "incentive" rather than functioning as the default treatment of machinery and equipment, moreover, the exemption often comes with conditions.

States also vary in sales taxation of business utilities. As heavy energy users, data centers also face considerably higher costs of doing business if their electricity consumption is subject to sales tax or a parallel excise taxAn excise tax is a tax imposed on a specific good or activity. Excise taxes are commonly levied on cigarettes, alcoholic beverages, soda, gasoline, insurance premiums, amusement activities, and betting, and typically make up a relatively small and volatile portion of state and local and, to a lesser extent, federal tax collections..

One study on the sales taxation of business inputs found that if states were able to reduce their taxation of business inputs by even 25 percent while making up the revenue with a commensurately higher rate applied to the rest of the sales tax baseThe tax base is the total amount of income, property, assets, consumption, transactions, or other economic activity subject to taxation by a tax authority. A narrow tax base is non-neutral and inefficient. A broad tax base reduces tax administration costs and allows more revenue to be raised at lower rates.,they would boost capital accumulation (the amount of machinery, equipment, and overall capital investment in the state) by 1.2 percent and increase gross state product by 0.4 percent. Nationwide, that would represent an additional $115 billion a year in economic output.[8] With data centers, the choice may be even starker, as for many operators, it could be a binary choice: invest in a state that taxes these inputs, or in one of the many other viable alternative locations that do not.

Of states with a significant data center concentration, only California broadly applies its sales tax to data centers, and California is likely to be one of the few states that could get away with this. Even the attractions of Silicon Valley, however, have not been enough to keep many California-based businesses from locating many of their data centers elsewhere, to a far greater degree than would be explained by the benefits of geographic diversity alone.

At the national average combined state and local sales tax rate of 7.52 percent, an obligation to pay sales tax on servers and other purchases of tangible property would cost a $1 billion data center an estimated $58.3 million in the first year and $9.7 million each year thereafter. States usually exempt most data centers from this liability, but often only subject to meeting certain criteria.

Business Personal Property Taxation

Many states permit localities to include business tangible personal property (machinery, equipment, fixtures, vehicles, etc.) in the property taxA property tax is primarily levied on immovable property like land and buildings, as well as on tangible personal property that is movable, like vehicles and equipment. Property taxes are the single largest source of state and local revenue in the U.S. and help fund schools, roads, police, and other services.base, a significant consideration for any capital-intensive businesses, including data centers. Taxes on tangible personal property (TPP) comprise a small share of local tax collections, but create high compliance costs, distort investment decisions, and help influence location decisions for operations like data centers, where the site selection process is likely to be rigorous.

Personal property is valued based on the initial acquisition price as well as the age of the asset, as such property is depreciated according to schedules published by each state with a TPP tax. DepreciationDepreciation is a measurement of the "useful life" of a business asset, such as machinery or a factory, to determine the multiyear period over which the cost of that asset can be deducted from taxable income. Instead of allowing businesses to deduct the cost of investments immediately (i.e., full expensing), depreciation requires deductions to be taken over time, reducing their value and discois a measure of the "useful life" of a business asset, with the taxable value of the asset declining over time until it reaches a set floor or no longer has taxable value. Often, an asset is not permitted to depreciate below a particular threshold even after the end of its useful life according to the depreciation schedule.

Depreciation rules are set by state statute or local ordinance and are not uniform. Schedules vary considerably, but five years is about the average for the relevant asset classification (computer equipment), and typically such equipment is subject to some sort of accelerated depreciation rules.

Loudoun County, Virginia,which boasts a large concentration of data centers, offers accelerated depreciation for computer equipment. It is taxed at 50 percent of its value in the first year and depreciates a further 10 percent annually until year five, after which it remains at 10 percent of original cost in perpetuity.[9] In Silicon Valley (Santa Clara County, California), the standard California Board of Equalization valuation factors are used, with 73 percent of value counted in the first year, 47 percent in the second, and 30 percent in the third, declining to 2 percent of original value in year nine and after.[10] Charlotte, North Carolina,has a five-year schedule phasing out about 20 percent per year, with a 5 percent floor.[11]

Whereas land is inherently immobile, machinery and equipment are not-and perhaps more importantly, when companies scope out major capital-intensive investments like a new data center, they take TPP taxation into account. TPP taxes essentially function as stock taxes on the value of a business's assets and are particularly harmful to businesses where M&E costs predominate over labor and other costs of doing business.

Even when companies choose to locate data centers in jurisdictions with TPP taxes, they may sacrifice the efficiency of their operation to limit tax liability, by reducing the rate of replacement of aging servers and network equipment or by timing replacements to fall after assessment "snapshot" dates, even if earlier replacement would otherwise be preferable.[12]

Corporate Tax Exposure

Most data center operators-and particularly the corporate parents behind hyperscalers-are structured as C corporations and pay corporate income taxes. The degree to which data center location affects corporate income tax liability depends on (1) the apportionment factors used by the host state, (2) sourcing rules for services under the sales factor of those apportionment rules, and (3) provisions for the expensing of capital investment.

States may apportion their share of a corporation's net income (profits) for in-state taxation on the basis of their share of payroll, property, and sales in the state, though increasingly most states apportion primarily or exclusively on the basis of in-state sales. States that still include a property component in the base will source more income to states hosting data centers, since they are property (capital)-intensive. Data centers bring a certain amount of employment as well, increasing the payroll factor, but do not necessarily increase sales into the jurisdiction. Some data center operations may support tech clusters that increase in-state sales as well.

If states apportion exclusively on the basis of sales, this does not provide any disincentive for data center location. Retaining payroll and property factors, however, puts states at a competitive disadvantage in attracting data centers (and other businesses) because the domiciliary state will apportion a greater share of the company's overall profits for tax purposes, while most other states will still tax based on sales into their jurisdiction, leading to double taxationDouble taxation is when taxes are paid twice on the same dollar of income, regardless of whether that's corporate or individual income.of that income.

The higher the domiciliary state's corporate income tax is, the more significant an issue this is, since the profits apportioned would be those of the whole company, not just of the data center itself. Corporations may not want to build data centers in jurisdictions where apportionment factors would expose a larger share of their overall corporate profitability to a state's high-rate corporate income tax, above and beyond any amount apportioned based on actual sales into that state.

Even within the sales factor, states' rules matter. With tangible products, sales are sourced to a product's destination. With intangible goods and services, including data services, most states offer parallel treatment called market sourcing, but a few states still use what is called cost of performance sourcing, where the sale is sourced to where the income-producing activity takes place, not where its benefit is received. Under certain business configurations, this can increase taxability in the state hosting a data center, again yielding double taxation since other states would source the income to the location where the benefit was received.

This is not an issue for large enterprises or hyperscalers building out their own AI or cloud computing systems, since the data center itself would not be regarded as serving external customers. But for colocation and storage/hosting services, cost of performance sourcing increases the in-state tax base, typically on income that is also included in other states' tax bases. This sourcing method would also discourage other tech businesses from clustering around data centers, undercutting regional agglomeration effects that many states desire.

Finally, as capital-intensive businesses, data centers care about the way that capital investment is treated within the corporate tax code. Corporate income taxes are levied on net income (profits), so all ordinary business costs-capital investment, compensation, cost of goods sold, etc.-are ultimately deductible. But for capital investment, those deductions can be spread out over many years, using depreciation schedules. This penalizes capital investment by exposing high-investment businesses to greater overall tax liability due to delayed deductions that are eroded by inflationInflation is when the general price of goods and services increases across the economy, reducing the purchasing power of a currency and the value of certain assets. The same paycheck covers less goods, services, and bills. It is sometimes referred to as a "hidden tax," as it leaves taxpayers less well-off due to higher costs and "bracket creep," while increasing the government's spendinand discounted due to the time value of money.

At the federal level, the corporate income tax now features permanent first-year expensing of machinery and equipment purchases. This means that data centers can deduct the cost of their servers in the year in which they are purchased. Eighteen states conform to this treatment, while the remaining states with corporate income taxes require such equipment to be depreciated by schedule, typically following the old five-year federal schedule for such assets. Most corporations care about expensing policies, but the economic profile of data centers makes them particularly sensitive to differences in the treatment of capital investment.

Real Property Taxation

Data centers, like all businesses, owe property taxes on their land and structures. These costs are not inconsiderable, since large data centers can sometimes require hundreds of acres-potentially in areas with high land values-and the facility (shell) itself is quite expensive. Some places offer property tax abatements to qualifying data centers to defray these costs, though such targeted incentives are difficult to justify. Real property taxation, moreover, reflects the neutral application of a broad-based local tax to data centers, whereas other taxes discussed can represent discriminatory treatment.

However, jurisdictions with so-called split roll taxation, where commercial and industrial property faces higher rates or assessment ratios than other classes of property, render themselves less attractive to data centers and other businesses with substantial property values.

Other Taxes and Fees

Data centers often face significant license fees, impact fees, and proffers, frequently tied to related infrastructure, including water hookups and electrical connectivity. These costs can vary dramatically based on location. In some cases, they hew closely to actual costs, whereas in other situations they may serve as a substantial revenue driver for local governments. While important, these local costs are not considered here.

Do Data Centers Matter to States and Localities?

Data centers are big business, and with the AI boom, investment in additional data center capacity is skyrocketing. Data center employment was rising even before the integration of AI into workflows became de rigueur, rising 60 percent between 2016 and 2023.[13] New rounds of investment will make that prior growth seem trivial. With this much money on the line, data center operators have a strong incentive to make savvy site selection decisions, and state and local tax liability will feature in their calculations.

But if states adopt proper tax treatment of data centers, avoiding discriminatory taxes on their machinery, equipment, and operations, what can they expect to get out of the bargain?

Data centers create high-paying jobs, though not in large numbers. Direct employment is modest overall and particularly low in relation to capital costs. But that capital investment-ongoing, not just one-time, due to the speed of change in the industry-creates jobs for local contractors, suppliers, and construction workers. Data centers can serve as an anchor for clusters and tech ecosystems that bring economic diversification. The broader tech ecosystem supports information technology firms, electricians, specialty contractors, and managed service providers. These new jobs support existing businesses as well.

Investors are all-in on AI, to the extent that other industries might see reduced investment as companies battle for position in the AI arms race. Failing to create a competitive environment for data center investment doesn't just risk missing out on a new growth sector; it risks backsliding as capital flows away from established industries.

Finally, even without discriminatory taxes on data centers, these operations expand the tax base. Localities in which data centers operate receive tax windfalls that reduce burdens on, and support additional benefits to, residents.

Comparing Tax Liability

States differ in their current tax treatment of data centers. To demonstrate how states compete-and how some fail to keep pace-we examine the tax treatment of data centers in 12 locations, chosen because they have emerged as loci of data center activity in a variety of geographic and economic environments. Some, like California's Silicon Valley and Virginia's data hub radiating out from Ashburn, are already at or near the top of any data center investment list. Others, like Papillion, Nebraska,and Cheyenne, Wyoming, are smaller players, but important in understanding the emerging competitive landscape across a range of environments.

For each location, we model a $1 billion data center facility. For the sake of simplicity and comparability, we consider an independent standalone data center, not one affiliated with a large corporation with other tax factors. Our model firm has $220 million in annual revenue and 15 percent net income before tax. Further specifications are included in the appendix.

We calculate tax liability in each location twice, once for a new operation, where all initial M&E is purchased in one year, and where it is assessed at maximum value; and again for an established operation, where the business is already at the point of replacing equipment it had previously put into service. We run these tax liability calculations in the following jurisdictions, covering most states with a major data center presence and capturing a range of tax and economic environments:

• Phoenix, Arizona
• Santa Clara County, California
• Atlanta, Georgia
• Elk Grove, Illinois
• Papillion, Nebraska
• Charlotte, North Carolina
• Columbus, Ohio
• Hillsboro, Oregon
• Irving, Texas
• Loudoun County, Virginia
• Quincy, Washington
• Cheyenne, Wyoming

Regardless of where they are located, data centers will owe federal taxes (chiefly the corporate income tax) as well as taxes to other states, with state corporate income taxes largely apportioned based on sales into those states. These burdens do not change based on location. But most of a data center's tax burden arises from taxes levied by the state and locality in which it is sited.

Tax burdens tend to be highest in the first year, since all machinery and equipment must be acquired at once, yielding the highest concentration of sales tax liability, and all that property will be undepreciated, yielding the highest TPP tax liability. In some cases, however, abatements are available for the first years of operation, yielding lower effective rates initially than in later years. For each location, we calculate tax liability both in the first year and in a "steady state" once machinery and equipment are replaced on a regular schedule.

A standalone data center would not have a corporate income tax burden in the first year, however, because its first-year expenses would vastly outstrip its income. In our model, in addition to any corporate income or gross receipts taxes imposed by the domiciliary state, the data center pays just under $4.2 million a year in corporate income taxes to the federal government and to states other than the domiciliary state once mature.

In Table 1, effective rates are given as a percentage of net income (profits), and total steady-state tax liability, both to the domiciliary state and to the federal government and other states. All-in effective tax rates vary widely, from a low of 24 percent in Charlotte, North Carolina, to a high of 80 percent in Santa Clara County, California.

Effective tax rates can be much higher initially due to fully undepreciated property and (in some cases) substantial sales tax exposure. A data center is unlikely to be profitable in its first year given its high upfront costs, but comparing first-year taxes to long-run average profits yields an astonishing 263 percent rate in Santa Clara, where servers and other equipment are taxed under a high-rate sales tax.

Table 1. Data Center Tax Liability and Effective Tax Rates by Jurisdiction

Tax Liability and Taxation as a Percentage of Net Income for a $1 Billion Model Data Center

Note: Assumes $220 million in gross revenue and 15 percent net income before tax. Tax liability as a percentage of net income is based on long-run average net income for both first-year and steady-state calculations, as an independent data center is unlikely to be profitable in the first year, making all effective rates infinite. Calculations assume qualification for exemptions routinely available to data centers of this size. See paper appendix for model firm assumptions.
Source: Tax Foundation calculations based on state statutes, state revenue agency documentation, local government publications, and local utility costs.

All jurisdictions considered except California exempt M&E purchases from the sales tax, but most jurisdictions tax that M&E (including servers) under TPP taxes or their equivalents for established data centers, once any initial abatements have expired. Of the 12 locations reviewed, only two-Elk Grove, Illinois, and Columbus, Ohio-impose no tangible personal property tax. Both states eliminated TPP taxes in prior reforms, Illinois through business income surtaxes and Ohio through a low-rate gross receipts taxGross receipts taxes are applied to a company's gross sales, without deductions for a firm's business expenses, like compensation, costs of goods sold, and overhead costs. Unlike a sales tax, a gross receipts tax is assessed on businesses and applies to transactions at every stage of the production process, leading to tax pyramiding.,which also replaced prior corporate income and capital stock taxes.

Data centers are subject to gross receipts taxes in five states represented in our review: Ohio, Oregon, Texas, Virginia, and Washington. Additionally, Virginia's corporate income tax includes payroll and property apportionment factors, yielding corporate income tax liability in the state above and beyond any apportionment based on sales destination. The jurisdictions sampled vary on whether they subject data centers' electric utilities to the sales tax or, in some cases, apply a separate excise tax. Table 2 shows tax liability in the domiciliary state by tax category for a "steady-state" $1 billion data center.

Table 2. Data Center Tax Liability by Type of Tax

Annual Tax Liability for a $1 Billion Data Center

Note: Assumes $220 million in gross revenue and 15 percent net income before tax. Taxes on electricity include both sales and excise taxes on data centers' electricity usage. Corporate income tax liability is only included as a domiciliary tax where the apportionment factors include payroll and property. Calculations are for data centers in a "steady state" where property is depreciating and being replaced on a regular schedule. Calculations assume qualification for exemptions routinely available to data centers of this size. See paper appendix for model firm assumptions.
Source: Tax Foundation calculations based on state statutes, state revenue agency documentation, local government publications, and local utility costs.

For three tax types, tax liability can differ dramatically in the first year: (1) sales taxes, since all servers and other tangible property must be purchased new in the first year; (2) TPP taxes, since all property will be undepreciated initially, and also since some states offer initial abatements that are no longer available in later years; and (3) corporate income taxes, on the assumption that the business is not profitable initially.

In Nebraska, for instance, the ImagiNE Nebraska incentive can eliminate TPP liability for up to 15 years. In Georgia, development authorities regularly abate the tax for up to 10 years. And in North Carolina, a 90 percent reimbursement is available for the first five years.

The following table compares first-year and steady-state liability across taxes with such a differential. Additionally, we assume about $4.2 million in federal and other-state corporate income tax liability in later years, but not in the first.

Table 3. Data Center Tax Liability for New and Mature Operations, Select Taxes

Annual Tax Liability for a $1 Billion Data Center, by Tax Type and Maturity (Select Taxes)

Note: Assumes $220 million in gross revenue and 15 percent net income before tax. Table only includes taxes on which data center liability differs for first-year and steady-state firms. Calculations assume qualification for exemptions routinely available to data centers of this size. See other tables for full tax liability and see paper appendix for model firm assumptions.
Source: Tax Foundation calculations based on state statutes, state revenue agency documentation, local government publications, and local utility costs.

Conclusion

Data center taxation varies widely across states, but taxes are clearly a factor in site selection determinations. Silicon Valley, with its tech agglomeration and centrality to the new artificial intelligence boom, and Loudoun County, Virginia, the historic "capital of the internet," have built-in advantages that continue to attract data center operations even at high tax costs. But if companies were indifferent to tax liability and other costs of doing business, nothing approaching the growing spread of these facilities would be apparent. Companies might choose to locate facilities in some other strategic locations, but the growing popularity of places like Nebraska, North Carolina, and Wyoming testifies to the role of taxes in attracting these operations. Data centers can and do bring substantial tax revenues to the cities and states in which they are located, but excessive burdens make it harder to compete, forgoing all of the revenue along with the other economic benefits of regional investment in data processing capacity.

Appendix: Model Data Center Specifications

For purposes of our tax calculations, the following specifications are used for our model independent data center.

• Tangible Personal Property: $775 million, including "big plant" and "pure" tangible personal property, for which tax treatment sometimes varies.
  • Servers/chipsets: 75 percent of TPP, 5-year replacement cycle
  • Batteries/UPS: 8 percent of TPP, 10-year replacement cycle
  • Chillers: 8 percent of TPP, 20-year replacement cycle
  • Switchgears/electrical: 5 percent of TPP, 25-year replacement cycle
  • Diesel gensets: 3 percent of TPP, 25-year replacement cycle
  • Racks and miscellaneous: 1 percent of TPP, 25-year replacement cycle
• Shell Build: $225 million
• Land Area: 40 acres
• Electricity Usage: 32 MW utilization (~0.28 TWh/yr)
• Revenues: Gross revenue of $220 million with net income of 15 percent

References

[1] Perkins Coie, "Trends in the Growth of Investment in US Data Centers Under the New Administration," Feb. 4, 2025, https://perkinscoie.com/insights/update/trends-growth-investment-us-data-centers-under-new-administration.

[2] Vivian Lee et al., "Breaking Barriers to Data Center Growth," Boston Consulting Group, https://www.bcg.com/publications/2025/breaking-barriers-data-center-growth.

[3] For a more detailed analysis, see Jared Walczak, "Modernizing the State Sales Tax: A Policymaker's Guide," Tax Foundation, Sep. 5, 2024, https://taxfoundation.org/research/all/state/state-sales-tax-reform-guide.

[4] Louisiana Economic Development, "Meta Selects Northeast Louisiana as Site of $10 Billion AI-Optimized Data Center," https://www.opportunitylouisiana.gov/metadatacenter.

[5] S.C. Code Ann. § 12-36-2120(79)(B)(5)(iv).

[6] 34 Tex. Admin. Code § 3.335.

[7] Scott Wright, Alla Raykin, and Laurin McDonald, "Tricks and Traps of Data Center State Tax Incentives," Tax Notes, Jan. 1, 2024, https://www.taxnotes.com/special-reports/tax-technology/tricks-and-traps-data-center-state-tax-incentives/2023/12/28/7hmb7.

[8] Benjamin Russo, "An Efficiency Analysis of Proposed State and Local Sales Tax Reforms," Southern Economic Journal 72:2 (2005): 443-462.

[9] Loudoun County Board of Supervisors, "FY 2022 Budget Development: Analysis of Business Tangible Personal Property Taxes on Computer Equipment," Oct. 13, 2020.

[10] California State Board of Equalization, "Table 07: Non-Production Computer Valuation Factors," Assessors' Handbook, Section 581, Equipment and Fixtures Index, https://www.boe.ca.gov/dataportal/dataset.htm?url=AH581_07.

[11] North Carolina Department of Revenue, "2025 Cost Index and Depreciation Schedules," Nov. 15, 2024, https://www.ncdor.gov/2025-cost-index-manual-finalpdf.

[12] For more on tangible personal property taxes, see Garrett Watson, "States Should Continue to Reform Taxes on Tangible Personal Property," Tax Foundation, Aug. 6, 2019, https://taxfoundation.org/research/all/state/tangible-personal-property-tax/.

[13] Andrew Foote and Caelan Wilkie-Rodgers, "Employment in Data Centers Increased by More Than 60% From 2016 to 2023 But Growth Was Uneven Across the United States," US Census Bureau, Jan. 6, 2025, https://www.census.gov/library/stories/2025/01/data-centers.html.

Share this article