Utilities — NAICS 221

A rollup investor primer for public-market and private investors

NAICS = North American Industry Classification System, the U.S. government's standard for grouping businesses by activity. NAICS 221 is a subsector (3-digit) — one level up from the three industry groups it contains: 2211 Electric Power, 2212 Natural Gas Distribution, and 2213 Water, Sewage and Other Systems. This note synthesizes the three child primers plus our ingested federal statistics for the combined 221 level. Core business statistics (revenue, firms, establishments, employment, payroll, concentration) are our authoritative ingested U.S. Census figures; physical scale, ownership mix, capital spending, and market data come from the Energy Information Administration (EIA), the Environmental Protection Agency (EPA), industry bodies, and company filings as cited via the child primers. Written for general investors, public-market and private alike.

---

1. Overview

NAICS 221 is "the utility bill" as a whole — every networked service that a wire or a pipe delivers to a building: electricity, natural gas, water, sewer, and district heating and cooling. Together the counted (private-sector) firms book about $765.95 billion of revenue across roughly 7,314 firms and 20,379 establishments, employing about 667,277 people [1][2]. That is a small headcount for the machinery that keeps the lights, heat, and taps running nationwide — the signature of an extraordinarily capital-heavy set of businesses (about $1.15 million of revenue per worker, and average pay near $127,000) [1][2].

Every business inside this code shares one investor-relevant piece of DNA: each individual system is a local natural monopoly. Once the wire, pipe, or main is in the street, no one builds a competing network alongside it, so the owner is granted a monopoly and, in return, a regulator caps what it can earn. The dominant economic model across the whole subsector is therefore the same regulated toll road: the utility earns an approved return on the capital it sinks into its network, which makes most of these businesses bond-like — visible, slow-growing, dividend-paying — rather than a bet on any commodity price.

But the reason to look at NAICS 221 as a group, rather than at one utility bill, is the contrast across its three children — the insight a top-down "utilities" summary averages away. These are not three versions of one business; they are on three different trajectories, owned very differently, and reached through very different instruments:

• Electric power (2211) is the growth engine — and, at ~74% of the group's revenue, it is most of the group. After roughly 15 years of flat demand, U.S. electricity load is inflecting up on data centers, electrification, and reshoring [13][14].
• Natural gas distribution (2212) is the stable middle — a ~24% slice of steady, regulated last-mile cash flow, but carrying a genuine long-term overhang: the same electrification wave that lifts the electric child is the terminal threat to the gas child.
• Water, sewage and district energy (2213) is the small, defensive tail — barely ~3% of counted revenue, grinding upward on a mandated, decades-long pipe-and-plant replacement cycle, and overwhelmingly government-owned.

So the single most useful thing this level tells an investor is that "utilities" is not one asset. It bundles a re-accelerating growth story (electric), a melting-ice-cube-on-a-long-fuse (gas), and a defensive mandated-capital compounder (water) — and one of the three (gas) is threatened by the very force that is re-rating another (electric). A broad utility index blends all of that into a single yield. This primer leads with the split, then covers the group.

Two ways in, common to all three. Public-market investors buy the listed regulated utilities, the merchant power generators, and the water and gas "pure plays" — though there is no pure-play stock for the subsector, and even within each child the listed names are mostly diversified. Private investors — infrastructure funds, pension consortia, and direct operators — buy whole systems, project equity, or tax-exempt municipal and cooperative bonds. Because so much of this infrastructure is government-owned, the private route reaches assets the stock market cannot.

---

2. What's inside — the three children, and how they differ

NAICS 221 contains exactly three industry groups. All three are regulated network monopolies, but they diverge on size, growth, ownership, and how you can touch them.

Comparison table — the three children at a glance

HHI = Herfindahl-Hirschman Index, a 0–10,000 concentration gauge; U.S. antitrust agencies treat anything below 1,500 as "unconcentrated." Share-of-group is of the group's counted Census revenue [1][3].

Five contrasts do the work.

• The size cliff is the first thing to internalize. By counted revenue the subsector is roughly three-quarters electricity, one-quarter gas, and a sliver water [3]. By employment it is even more lopsided — electricity is ~78% of the workforce. When you buy "utilities," you are ~74% buying the electricity story; gas is the junior partner and water is a rounding error at the group level (even though it is a large, vital service).

• Growth trajectories point in different directions — and two of the children are rivals. Electricity is accelerating on rising load; water and sewage are grinding up on mandated replacement; gas is stable now but faces structural displacement later. Crucially, electrification is a tailwind for the electric child and a headwind for the gas child — heat pumps and electric appliances move heating load off the gas main and onto the wire. The group literally contains a substitution rivalry between its two largest members. A single "utilities" number hides that internal tension.

• Ownership is government-heavy everywhere, but on a gradient that widens the smaller the child. Electricity is mostly investor-owned with large public pockets (federal hydro, the Tennessee Valley Authority, ~830 distribution cooperatives) [12][13]; gas is ~72% investor-owned with 900-plus city systems [16]; water and sewage are ~84% and ~98% government-owned [17][18]. So the federal-statistics undercount (see §3) gets worse as you move down the revenue ladder — the smallest child is the most undercounted.

• Commodity risk lives almost entirely in one child. Electricity's generation layer sells power into competitive wholesale markets and carries the group's commodity and merchant-price swings. Gas distributors pass the gas molecule through to customers at cost (a purchased-gas adjustment, with no markup), and water has no commodity at all. So all of the group's meaningful commodity volatility, and most of its policy-driven upside, sits inside electric generation [child 2211][child 2212].

• Concentration is a statistical artifact at the group level. The subsector's HHI is just 147.7 — below every child that reports one (electric 186, gas 386.5) [1][3]. Pooling three markets whose biggest firms do not compete with each other (a big electric utility is not a rival to a big gas or water utility) mechanically dilutes every firm's share of the combined revenue. Read the group's low concentration as a pooling effect, not as evidence of competition: the real structure underneath is thousands of local monopolies.

What all three share: each is a capital-is-the-business, rate-regulated network monopoly whose earnings grow by building more approved plant, pipe, or wire — which makes the entire subsector interest-rate-sensitive at once (see §5).

---

3. Size — the rollup figures, and how to read them

Our authoritative Census figures for NAICS 221 (the counted, private-sector subsector):

The rollup reconciles almost exactly against the three children. Revenue ($563.3 B + $182.7 B + $20.0 B = $765.95 B), establishments (12,735 + 2,434 + 5,210 = 20,379), employment (519,711 + 95,538 + 52,028 = 667,277), and payroll ($69.27 B + $11.58 B + $4.05 B = $84.90 B) are the clean sums of the child figures [1][2][3]. The one measure that does not simply add up is the firm count: 7,314 at the subsector versus 7,379 summed across the children — because a company that owns, say, both electric and gas utilities is counted once here but in two of the child groups. That gap of ~65 firms is a direct measure of how many owners are diversified across service lines (the Sempras, Dominions, and NiSources that own both a wire and a pipe) — and it is why the diversified holding company, not the pure play, is the typical listed name.

Water has the most firms but the least revenue — the fragmentation is real. Note the inversion in the table: water/sewage/district energy has the largest firm count (4,321) yet the smallest revenue (~2.6%) [3]. That is thousands of small, mostly local systems; electricity, by contrast, concentrates ~74% of revenue in fewer than 2,626 firms. The subsector's low national concentration coexists with intense local monopoly everywhere.

Two caveats before you take $765.95 billion as "the size of U.S. utilities."

1. The government undercount — the single most important structural fact about this subsector. Census business surveys generally exclude government-owned establishments [4]. That matters enormously here, and unevenly: federal and public power own ~73% of hydro generation and ~830 distribution cooperatives sit in electricity [12][13]; 900-plus municipal systems sit in gas [16]; and the majority owner in water (~84%) and sewage (~98%) is government outright [17][18]. So the $765.95 billion is best read as the private-sector baseline, not the size of the underlying service. There is no official consolidated total that adds government utilities back in — treat that number as not available rather than estimating it.

2. Receipts are not value added. The figure is the correct arithmetic sum of receipts, but it double-counts commodities that pass through more than one layer. Within electricity, distribution's retail bills already embed the wholesale cost of the power that generation booked [child 2211]; within gas, marketers' and utilities' receipts include the resold gas molecule [child 2212]. The number is right as reported; just don't read it as the subsector's economic size — the true value added is smaller and sits in the networks and generating margins, not in the pass-through cost of fuel.

Physical scale (the more reliable gauge of the real industry). Electricity: a fleet of about 1,230 gigawatts (GW; 1 GW = 1,000 megawatts) of capacity producing roughly 4,300 terawatt-hours a year, over 600,000 miles of high-voltage transmission and 5.5 million-plus line-miles of distribution, feeding a ~$514 billion retail market in 2024 [5][22]. Gas: about 2.4 million miles of distribution pipe serving ~79.6 million customer accounts [child 2212]. Water and sewage: an asset base the EPA sizes at ~$625 billion (drinking water) and ~$630 billion (wastewater) of 20-year need alone [19][20]. The physical inventories, not the receipts, are the honest gauge of what this subsector actually is.

---

4. The investable universe — where the group's value concentrates

Two facts govern the whole subsector. First, there is no pure-play public stock for the group — and even within each child the listed names are mostly diversified. Second, a large share of the physical asset base is not listed at all — it sits with federal agencies, public-power authorities, cooperatives, municipalities, and private infrastructure funds, reachable only through debt or through the equipment-and-services supply chain.

Where public-market value concentrates, by child:

• Electric power (the ~74% core). Three profiles: merchant / independent power producers valued on power and capacity prices — Constellation (CEG) (largest U.S. nuclear operator), Vistra (VST), NRG (NRG), Talen (TLN); renewable platforms with contracted cash flow — NextEra (NEE) (world's largest wind and solar owner), Brookfield Renewable (BEP/BEPC), Clearway (CWEN), AES (AES); and regulated utilities trading as bond proxies — Exelon (EXC) (the closest thing to a pure "wires" play), Southern (SO), Duke (DUK), American Electric Power (AEP), Xcel (XEL) [child 2211].
• Natural gas distribution (the ~24% middle). A handful of listed gas "pure plays" (none perfectly pure) — Atmos Energy (ATO), Southwest Gas (SWX), ONE Gas (OGS), Spire (SR), New Jersey Resources (NJR), Northwest Natural (NWN) — plus much larger gas rate base buried inside diversified parents (Sempra, whose SoCalGas is the largest U.S. gas-only utility; NiSource; CenterPoint). Canada's Enbridge (ENB) became the largest North American gas-utility platform after a ~US$14 billion purchase of three Dominion utilities in 2024 [21][child 2212].
• Water, sewage & district energy (the ~3% tail). A short roster of regulated water utilities that also carry the group's small wastewater leg — American Water Works (AWK), Essential Utilities (WTRG), American States Water (AWR), California Water Service (CWT), SJW Group (SJW) — but no listed pure play for sewage or district energy (district energy's closest name is Consolidated Edison's Manhattan steam system, a few percent of a large electric-and-gas utility) [child 2213].

The diversified holding companies span the children. The most common listed shape is not a pure play at all but a multi-utility parent — Sempra, Dominion (D), Southern, NiSource, CenterPoint, Ameren — that owns electric and gas (and sometimes water-adjacent) rate base at once. Valuing them means valuing segments, not a single business. And there is no exchange-traded fund (ETF) for the subsector or any single child; the broad utility funds (XLU, VPU) hold electricity, gas, and water mixed together, while thematic funds isolate slices (GRID for grid equipment, PHO/FIW for water) [child 2211][child 2213].

Where private, government, and public-power value concentrates (not buyable on an exchange): federal and public power (Army Corps and Bureau of Reclamation hydro, the Tennessee Valley Authority, the Power Marketing Administrations, large municipal systems, ~830 distribution cooperatives) [12][13]; municipal gas systems such as Philadelphia Gas Works [child 2212]; the government-owned majority of water and sewage systems [17][18]; and the infrastructure funds and private-equity platforms (Brookfield, KKR, Antin, LS Power, Stonepeak, Macquarie, New Mountain) that own merchant power, whole regulated utilities, district-energy platforms, and contract-operations businesses [child 2211][child 2213]. These are reached through municipal and cooperative bonds, project debt, and fund commitments — plus the equipment-and-contractor supply chain (Quanta Services, MasTec, Eaton, Hubbell, GE Vernova, Veolia, Jacobs) that captures the build-and-operate spend regardless of who owns the asset.

The pattern across the whole group: the public-equity universe is essentially the investor-owned slice of each child, and even there it is diversified rather than pure. The government, municipal, and cooperative owners — a minority of electricity and gas, a large majority of water and sewage — offer no common stock.

---

5. How the money works — one shared engine, one extra layer

The shared engine (all three children): rate base × allowed return. Wherever an asset — a power plant, a gas main, a water treatment plant — sits inside a regulated investor-owned utility, a regulator (usually a state public utility commission, PUC) approves a revenue requirement:

(rate base × allowed return) + operating costs + depreciation + taxes.

Rate base is the utility's prudent invested capital, net of depreciation; the allowed return blends the cost of debt with an allowed return on equity (ROE) on the equity-funded slice. Because the owner earns a return on the asset base, the growth algorithm is identical across the entire subsector: build more approved plant, pipe, or wire → grow rate base → grow earnings. Owners are paid for prudent investment, not for volume — which is why every child grows despite flat-to-falling usage per customer. Recent allowed ROEs cluster remarkably tightly across all three, near 9.5–10% (electric distribution ~9.7%, gas ~9.6–9.7%, water ~9.5–10.3%) [23][child 2212][child 2213]. The shared catches are also identical: returns are capped, the utility must actually spend to grow, there is regulatory lag before rate relief arrives, and a commission can disallow imprudent costs.

Because the model is capital-in, return-on-capital-out, the whole subsector is interest-rate-sensitive in the same direction at once — higher rates raise the cost of the capital that is the business and compress every valuation, in all three children together. This is why utilities of every kind trade as "bond proxies."

The extra layer (electricity only): the competitive / merchant model. This is what gas and water do not have. Uncontracted power plants sell into wholesale markets, earning a stack of energy, capacity payments, and long-term power-purchase agreements — increasingly with data-center operators. The recent capacity-price surge is the sharpest example: in PJM (the largest market), the capacity price jumped from $28.92 to $269.92 per megawatt-day for 2025/26 and hit the $325 cap for 2028/29 [8]. All of the group's commodity volatility and most of its non-regulated upside live here.

Where the other two children sit on the spectrum. Gas distribution runs the pure regulated model with the commodity passed straight through — the gas molecule earns no margin, so earnings track rate base, not gas prices [child 2212]. Water, sewage, and district energy run the same regulated model on the private slice, plus municipal break-even economics on the government majority (systems funded by cheap tax-exempt bonds) and, uniquely, a merchant/contract layer in district energy (long-term energy-service agreements) [child 2213]. So as you move across the children, the model goes from regulated-plus-a-large-merchant-engine (electric) to pure regulated pass-through (gas) to regulated-plus-municipal-break-even (water) — steadily lower-variance.

The policy-subsidy lever flows mainly to electricity. The 2022 Inflation Reduction Act's clean-electricity tax credits (production and investment credits, plus a nuclear credit) subsidize generation and storage, are transferable for cash, and reshape which power plants get built [10]. They do not flow to gas or water networks. So the subsector's single sharpest policy fault line — the 2025 budget law's wind/solar construction-start cliff around mid-2026, versus preserved credits for nuclear, hydro, geothermal, and storage — runs inside the electric child [10].

---

6. Demand drivers

Two forces lift the whole subsector; a third splits its members apart.

• Aging infrastructure and mandated replacement (shared). Every child is running a decades-long capital cycle to replace old assets that regulators let owners earn a return on: 70%-plus of transmission lines are over 25 years old [22]; gas utilities spend ~$37 billion a year replacing aging mains [child 2212]; and the EPA sizes drinking-water and wastewater 20-year needs at ~$625 billion and ~$630 billion, with an estimated 9-million-plus lead service lines to replace and new "forever chemical" (PFAS) treatment mandates [19][20][24]. This is a baseline of spending that continues even if demand growth disappoints.

• The electricity demand inflection (electric, and indirectly the group). After ~15 years of flat load, U.S. electricity demand is rising on data centers and artificial intelligence (AI) — Lawrence Berkeley National Laboratory estimates data centers used ~4.4% of U.S. electricity in 2023, rising toward 6.7–12% by 2028 [14] — plus broad electrification and factory reshoring. The North American Electric Reliability Corporation (NERC) now forecasts summer peak demand rising ~224 GW (~24%) over ten years [13]. Because electricity is ~74% of the subsector, this single driver dominates the group's growth.

• Electrification — the driver that divides the children. The same trend is a tailwind for electric and a headwind for gas: heat pumps, electric appliances, and building-electrification mandates move heating load off the gas main and onto the wire. Water and sewage are indifferent to it. So electrification is simultaneously the group's brightest growth story and the source of its one child's terminal risk — the clearest reason not to read "utilities" as one demand curve.

The shared counter-force: some of the data-center load underpinning the electricity re-rating may be delayed or double-counted, so the electric child (and any gas or water utility that builds ahead of it) faces the risk of investing ahead of demand that never arrives.

---

7. Regulation

Regulation is the business model, and the subsector's structure is consistent: an economic regulator that sets what the utility can earn, plus a safety/quality regulator that dictates what it must spend — with the gap between them being the source of regulatory lag.

• Economic regulation — mostly state, shared across all three. State PUCs set rate base, allowed ROE, and cost recovery for the investor-owned portion of every child — electric distribution, gas distribution, and water. Roughly 90%-plus of a typical distribution utility's economics are decided at this level [child 2212]. Municipal systems set their own rates through local boards.
• Federal economic regulation — the wholesale and interstate layer. The Federal Energy Regulatory Commission (FERC) governs wholesale electricity markets, the capacity auctions that drive merchant power economics, interstate electric transmission, and interstate gas transmission — but not retail gas, water, or sewer rates [child 2211][child 2212].
• Safety and environmental regulation — child-specific. The Nuclear Regulatory Commission (NRC) and EPA govern electric generation (reactors, emissions); the Pipeline and Hazardous Materials Safety Administration (PHMSA) and EPA govern gas pipelines (the mandated main-replacement that drives gas rate base) [child 2212]; and the EPA governs water and sewage under the Safe Drinking Water Act and Clean Water Act (lead, PFAS, discharge permits) [24][child 2213].

The through-line: all three children answer to a state PUC on rate-and-return terms; electricity and gas add a federal wholesale/interstate regulator; and each child bolts on its own safety-and-quality agency whose mandates create the very capital spending the utility earns a return on. Electricity is the most heavily and variably regulated (it alone has a competitive-market regulator layered on top).

---

8. Consolidation

The subsector is fragmented at the rollup (HHI 147.7) but consolidating within every child [1] — and, at the parent level, actively reshuffling which service lines the big holding companies want to own.

• Electric power is consolidating through merchant scale-up in gas and nuclear (Constellation/Calpine, ~$26.6 B; NRG's ~13 GW gas purchase from LS Power), transmission joint ventures and minority stakes, and infrastructure-fund buyouts of whole regulated utilities [child 2211].
• Natural gas distribution is consolidating through a clear rotation: electric-focused parents shed gas systems while gas specialists and infrastructure funds buy them for long-lived regulated cash flow — Enbridge's ~US$14 billion Dominion purchase (2024) and Spire's $2.48 billion purchase of Piedmont's Tennessee business (2026) are the marquee deals [21][child 2212].
• Water and sewage is consolidating through a roll-up of thousands of small, cash-strapped municipal systems by listed water utilities and infrastructure funds — accelerated by fair-market-value laws in ~13 states and headlined by American Water's pending ~$20.24 billion acquisition of Essential Utilities, creating a ~$40 billion platform [child 2213].

Two cross-cutting themes tie it together. First, infrastructure capital is the common buyer — private funds are consolidating merchant power, whole gas and water utilities, and district-energy platforms alike. Second, the diversified parents are reshaping their portfolios: shedding gas, spinning off generation (Exelon became a pure-wires company), and concentrating on the service lines the market rewards. The 65-firm gap between the summed children and the rollup (§3) is the current snapshot of that portfolio churn.

---

9. Risks

Shared across the whole subsector:

1. Interest rates. Every child is capital-is-the-business, so higher rates directly compress valuations and acquisition values across the group. Judge growth per share, not just aggregate capital spending.
2. Affordability backlash. Record capital spending across power, gas, and water pushes bills up, inviting rate freezes, adverse rate cases, and cost disallowances — the mechanism by which any of these regulated returns can be cut.
3. Regulatory lag and disallowance. Every child spends first and recovers later, and a commission can deny recovery of imprudent costs.
4. Cyber and supply chain. Control systems for grids, pipelines, and treatment plants are high-consequence security targets, and long-lead equipment (large power transformers exceed 36-month lead times) constrains the build [child 2211][child 2213].

Concentrated by child:

5. Electric (2211): commodity and merchant-price volatility; the 2026–27 wind/solar tax-credit cliff; nuclear and fossil fuel-specific tails; wildfire and catastrophic liability in distribution (PG&E's 2019 Chapter 11 and $25.5 billion wildfire settlement is the cautionary case); and the two-sided data-center demand-forecast / stranded-asset risk on which the whole group's growth narrative rests [child 2211].
6. Gas (2212): the existential long-term risk is electrification and stranded assets — if regulators shorten asset lives to match a declining-gas future, rate-base growth can reverse into a "fixed-cost spiral" (usage falls, fixed costs stay, bills rise, more customers leave); plus safety and catastrophic liability (the 2018 Merrimack Valley disaster) [child 2212].
7. Water/sewage/district energy (2213): PFAS, lead, and contamination-compliance liability (recoverable only after a rate case); merger-execution and multi-state approval risk; and, in district energy, fossil-plant stranding and bypass by building-level heat pumps [child 2213].

The unifying risk lens: because ~74% of the group is electricity, the single variable that matters most for NAICS 221 as a whole is whether the electricity demand inflection lands near consensus. If it does, the electric child pulls the group; if it disappoints, the group's growth thesis deflates and the gas child's electrification overhang looks nearer, not farther.

---

10. How to invest & outlook

Read the subsector as three risk profiles, not one "utilities" bloc.

• Electric power (2211) is where both the torque and the growth are. It offers the sharpest upside (merchant gas and nuclear on record capacity prices; solar, wind, and storage on the transition) and the sharpest risks (commodity volatility, the tax-credit cliff, wildfire liability, demand-forecast risk). Within it, the durable thesis is the scarcity and rising value of firm, already-built, dispatchable capacity meeting suddenly-rising load — and, for the lowest variance, the regulated wires riding a capital supercycle regardless of which fuel wins.
• Natural gas distribution (2212) is stable dividend-growth with a geographic split. Rate bases keep compounding on mandated safety spending and Sun Belt customer growth; the long-term overhang is electrification, and it bites unevenly — genuine terminal risk in restrictive coastal and Northeastern states, still-growing demand in the Sun Belt and Plains. Underwrite the state, not the sector.
• Water, sewage & district energy (2213) is the defensive compounder. Regulated water (with wastewater riding inside it) is the most investable and defensible corner — the clearest mandated rate-base growth, reached through a short roster of listed water utilities. Sewage and district energy have almost no direct equity and are naturally private-market or municipal-bond allocations.

Which parts are attractive vs. at risk (analytical judgment, grounded in the sourced facts):

• Attractive now: electric wires riding the decade-plus grid build-out; owners of existing, dispatchable generation (merchant gas and nuclear, plus firm-clean hydro/geothermal/storage with credits preserved to ~2033); and regulated water on its mandated lead/PFAS/pipe-replacement cycle.
• At risk / two-sided: electric wind and solar (strong demand, sharpest policy cliff — underwrite the construction-start date); gas utilities in electrification-forward states (terminal-value overhang); and any utility, in any child, that builds ahead of demand that never arrives.

How to get in. Public-market investors choose a profile, not a code: regulated electric and diversified utilities (EXC, SO, DUK, AEP; the multi-utility parents Sempra, Dominion, NiSource) for bond-proxy income with a grid-growth kicker; merchant power (CEG, VST, TLN, NRG) for torque to load growth; renewable platforms (NEE, BEP/BEPC) for contracted transition upside; gas pure plays (ATO, SWX, OGS) for stable dividend growth; water pure plays (AWK, WTRG) for defensive mandated-capital compounding; broad utility ETFs (XLU) for the whole mix; and the equipment-and-services suppliers (Quanta, GE Vernova, Eaton, Veolia) for the build-and-operate spend without the regulated-return ceiling. Private investors buy the assets the way the industry is actually owned: whole regulated utilities and minority network stakes, operating and development project equity, tax-equity and transferable credits, merchant plants and lines, and the tax-exempt municipal and cooperative bonds that are the only route into the government-owned majority of hydro, public power, gas, water, and sewage.

Outlook. After a lost decade of flat electricity demand, U.S. utilities as a whole have moved from managed maturity to a capital-led expansion — investor-owned electric utilities alone plan more than $1.1 trillion of capital spending across 2025–2029 [11], with parallel mandated cycles in gas safety and water replacement. The cleanest framing is the three-way split the subsector reveals: electricity is how you own the demand inflection (with the most torque and the most commodity-and-policy risk); gas is how you own steady regulated income (with an electrification fuse whose length depends on the state); and water is how you own the most defensive, mandate-driven compounding (with the least to buy directly). A broad utility fund blends all three into one yield — this level exists to un-blend them.

---

Sources

1. U.S. Census Bureau, 2022 Economic Census — Summary Statistics and Concentration Ratios, NAICS 221 (Utilities subsector) (receipts $765,952,367 thousand; 7,314 firms; CR4 14.7% / CR8 25.4% / CR20 48.0% / CR50 70.6%; HHI 147.7). Histometrics ingested ground-truth statistics. https://data.census.gov/table/ECNBASIC2022.EC2200BASIC
2. U.S. Census Bureau, 2023 County Business Patterns, NAICS 221 (20,379 establishments; 667,277 employees; annual payroll $84,904,374 thousand; first-quarter payroll $25,853,514 thousand). Histometrics ingested ground-truth statistics. https://www.census.gov/programs-surveys/cbp.html
3. U.S. Census Bureau, 2022 Economic Census and 2023 County Business Patterns, child industry groups 2211 / 2212 / 2213 — electric power receipts $563.3 B / 519,711 employees / 2,626 firms / HHI 186; gas distribution $182.7 B / 95,538 / 432 / HHI 386.5; water, sewage & other $20.0 B / 52,028 / 4,321 / HHI suppressed. Histometrics ingested ground-truth statistics; children reconcile to the 221 rollup on revenue, establishments, employment, and payroll, with the firm count slightly lower than the children summed because diversified owners are counted once.
4. U.S. Census Bureau, Understanding NAICS and Economic Census Coverage (government-owned establishments generally excluded from business statistics), 2022. https://www.census.gov/programs-surveys/economic-census/year/2022/guidance/understanding-naics.html
5. U.S. Energy Information Administration (EIA), Electric Power Annual 2024 (U.S. capacity ~1,230 GW; generation ~4,300 TWh; retail electricity market ~$514 B in 2024). https://www.eia.gov/electricity/annual/
6. Child primer 2211 — Electric Power Generation, Transmission and Distribution, and its 5-digit children 22111/22112 (physical scale, ownership, merchant economics, tax credits, company universe). Histometrics.
7. Child primer 2212 — Natural Gas Distribution (and its child 22121): last-mile economics, purchased-gas adjustment, physical scale (~2.4 M miles of pipe; ~79.6 M customer accounts; ~$37 B/yr investment), company universe. Histometrics.
8. PJM Interconnection, Reliability Pricing Model / Capacity Auction Results (2024/25 $28.92 → 2025/26 $269.92 → 2028/29 $325.00/MW-day cap). https://www.pjm.com/markets-and-operations/rpm
9. Child primer 2213 — Water, Sewage and Other Systems (and its children 221310/221320/221330): water/sewage/district-energy economics, ownership, company universe, consolidation. Histometrics.
10. Internal Revenue Service, Clean Electricity Production Credit (§45Y), Investment Credit (§48E), Zero-Emission Nuclear Credit (§45U), transferability; and U.S. Public Law 119-21 (2025) wind/solar construction-start cliff. https://www.irs.gov/credits-deductions/clean-electricity-production-credit
11. Edison Electric Institute (EEI), Industry Capital Expenditures / 2024 Financial Review (>$1.1 trillion planned IOU capex, 2025–2029). https://www.eei.org/resources-and-media/industry-data
12. U.S. Department of Energy / Oak Ridge National Laboratory, U.S. Hydropower Market Report and nuclear-ownership data (hydro ~73% federal + public; TVA and the Power Marketing Administrations). https://www.energy.gov/eere/water/hydropower-market-report
13. North American Electric Reliability Corporation (NERC), 2025 Long-Term Reliability Assessment (peak demand +224 GW / ~24% over 10 years); National Rural Electric Cooperative Association, Facts & Figures (~830 distribution cooperatives). https://www.nerc.com/; https://www.electric.coop/electric-cooperative-fact-sheet
14. Lawrence Berkeley National Laboratory, United States Data Center Energy Usage Report (4.4% of U.S. electricity in 2023; 6.7–12% by 2028), 2024. https://emp.lbl.gov/publications/united-states-data-center-energy
15. Federal Energy Regulatory Commission (FERC), wholesale power markets, interstate transmission, and interstate gas transmission jurisdiction. https://www.ferc.gov/
16. American Gas Association / EIA-176 municipal-systems data (~900+ city-owned gas systems; investor-owned utilities serve the majority of gas customers). https://www.aga.org/
17. U.S. Environmental Protection Agency, Safe Drinking Water Information System (community water systems ~84% government-owned). https://www.epa.gov/ground-water-and-drinking-water
18. American Water Works Company, 2025 Form 10-K (~98% of U.S. wastewater systems government-owned; fair-market-value laws; acquisition cadence; AWK–Essential merger). https://www.sec.gov/
19. U.S. EPA, 7th Drinking Water Infrastructure Needs Survey and Assessment (~$625 B 20-year need). https://www.epa.gov/dwsrf
20. U.S. EPA, 2022 Clean Watersheds Needs Survey (~$630 B wastewater 20-year need). https://www.epa.gov/cwns
21. Enbridge Inc., Enbridge Completes Acquisition of U.S. Gas Utilities (~US$14 B; East Ohio, Questar, PSNC), 2024; Spire Inc., Piedmont Tennessee acquisition ($2.48 B, 2026). https://www.enbridge.com/media-center/news
22. Congressional Research Service, Introduction to Electricity Transmission (IF12253) — ~600,000 miles of transmission line, 70%+ over 25 years old; ~5.5 million distribution line-miles. https://www.congress.gov/crs-product/IF12253
23. S&P Global Market Intelligence / Regulatory Research Associates — authorized ROEs clustering near 9.5–10% across electric distribution, gas distribution, and water utilities (2023–2024). https://www.spglobal.com/market-intelligence/
24. U.S. EPA, Lead and Copper Rule Improvements (Oct 2024; replace ~9 M+ lead service lines) and PFAS National Primary Drinking Water Regulation (2024). https://www.epa.gov/ground-water-and-drinking-water/lead-and-copper-rule
25. Company disclosures and transactions underlying the child primers — Constellation/Calpine (~$26.6 B), NRG/LS Power (~$12 B), Exelon generation spin-off (2022), PG&E Chapter 11 ($25.5 B wildfire resolution, 2019), American Water/Essential Utilities (~$20.24 B, 2025), 2022–2026 (SEC filings). https://www.sec.gov/

---

Prepared July 2026. Core NAICS 221 business statistics (receipts, firms, establishments, employment, payroll, concentration) are Histometrics-ingested U.S. Census figures and take precedence over third-party estimates; the three child industry groups (2211 electric power, 2212 natural gas distribution, 2213 water/sewage/other) reconcile to this rollup on revenue, establishments, employment, and payroll, with the firm count slightly lower than the children summed because diversified owners are counted once. Physical scale, ownership mix, capital-spending, tax, and market figures are drawn from EIA, DOE, EPA, EEI, NERC, FERC, and company filings via the three child primers, with reference years noted. Federal business statistics undercount this subsector wherever government / municipal / cooperative ownership dominates — a minority of electricity and gas, and a large majority of water and sewage — so the $765.95 billion is the private-sector baseline, not the size of the underlying service; and because receipts embed commodity costs that pass through more than one layer, that total is receipts as reported, not the subsector's value added. Both points are stated rather than estimated around. Forward-looking statements in §10 are analytical judgments, not guarantees.