Beverage and Tobacco Product Manufacturing (United States) — NAICS 312

A Histometrics rollup primer for public-market and private investors. This is the subsector (three-digit) level. It synthesizes the two child primers beneath it — beverages (3121) and tobacco (3122) — plus our ground-truth federal statistics for NAICS 312.

NAICS (North American Industry Classification System) 2022 code 312 is the subsector that pairs two things Americans consume for pleasure rather than nutrition: drinks (industry group 3121 — soft drinks, water, ice, beer, wine, spirits) and tobacco products (industry group 3122 — cigarettes and other tobacco). The classification staples them together because they share a common DNA: both turn cheap agricultural inputs into branded, heavily-regulated, excise-taxed consumer goods sold through licensed channels. But economically they are almost opposites, and that contrast — not the combined total — is why this level is worth reading.

1. Overview

At the factory gate this is a $175.0 billion U.S. subsector employing about 291,000 people across roughly 12,580 plants [2][1]. Almost all of that headcount and plant count is beverages; almost a third of the money, made by a rounding error's worth of plants, is tobacco. The single most useful thing to grasp about 312 is that its two halves could hardly be more different in shape:

• Beverages (3121) are ~71% of the receipts but ~99% of the plants and ~96% of the workers [1][2][3] — a sprawling, labor-heavy, largely private and fragmented world of thousands of brewers, wineries, distillers, bottlers, and water/ice operators. Concentration is low; ownership runs from public giants (on the non-alcoholic side) to almost entirely private family houses (on the alcohol side).
• Tobacco (3122) is ~29% of the receipts from ~1% of the plants and ~4% of the workers [1][2][4] — a tiny-footprint, near-monopoly cash machine of about 125 highly automated factories run by a handful of corporations, almost all of them reachable through public stock. It is one of the most concentrated and most profitable-per-worker industries in U.S. manufacturing.

Three threads tie the halves together despite the divergence. First, both are indulgence categories in structural per-capita decline. Gallup's 2025 reading put the share of U.S. adults who drink at 54% and self-described abstainers at 44% — the lowest and highest readings in its series [3][8] — while the adult cigarette smoking rate hit a record-low 9.1% in 2025, down from 9.9% in 2024 and 42.4% at the 1965 peak [4][6]. Second, the money is in brands, packaging, and distribution — not in the liquid or the leaf. In every child here the durable edge is a trademark, a route, or shelf placement, and the physical input is nearly free.

Third — and this is what has changed since the last revision — the shared defense against decline (raise price faster than volume falls) worked on one side of this subsector in 2025 and cracked on the other. Tobacco's version held: pack sales fell ~8.2% and net pricing more than covered it, with Altria's smokeable segment turning $17.4 billion of net revenue into $11.0 billion of operating income, a 63% operating margin [4][11]. Beverages' version slipped: U.S. spirits supplier revenue fell 2.2% to $36.4 billion even as case volume rose 1.9%, and wine's highest-margin direct-to-consumer channel fell 15% by volume [3][9]. If you read only one thing at this level, read that split — the same playbook, tested in the same year, with opposite results.

2. What's inside — the two child industries and how they differ

The reason to read this level is the divergence between its two children. Start here [1][2][3][4]:

Read across the rows and the split jumps out:

• Money and footprint invert. Tobacco earns nearly a third of the subsector's money from one plant in a hundred and one worker in twenty-five. Its ~$407M of shipments per plant and ~$4.5M per worker are roughly 40× and 10× the beverage figures — the fingerprint of heavy automation, brand pricing, and the tax-inflated value of each unit [4]. Beverages are the opposite: many small plants, modest revenue each, real people making physical product. (These per-unit figures mix 2022 receipts with 2023 counts and are directional, not exact.)
• Concentration runs from fragmented to oligopoly. Beverages are a genuinely competitive field (CR4 27.1%); tobacco is a textbook near-monopoly (CR4 90.6%, with the top 8 at 95.6% and the top 50 at 99.8%) [3][4]. Pooling the two is what makes the parent's numbers so misleading (Section 3).
• Ownership — and therefore how you invest — inverts too. In beverages, the public market gives you deep access to the non-alcoholic child and a short list of scaled alcohol names, while most of the alcohol side is private or foreign. In tobacco, almost the entire industry is public, and the harder problem is that many funds won't own it. So the same investor faces "the good stuff is private" in one child and "the whole thing is public but screened out" in the other.
• Shakeout is now visible on one side only. All three alcohol children are shrinking their operator counts at the same time — breweries, wineries, and craft distilleries alike [3][5]. Tobacco has no such churn: its structure was settled decades ago, and its recent deals repositioned the same few owners rather than culling them [4].

Note on 3122's internal structure. The tobacco child is itself a single-child chain: NAICS 3122 rolls up one five-digit industry (31223) and one six-digit national industry (312230), all numerically identical [4]. So when this primer says "tobacco," the group, the industry, and the leaf are the same thing.

Scope and the exclusions that matter most. Both children carve out fast-growing neighbors, and two exclusions shape how you read the totals:

• The soft-drink profit engine is booked outside 312. Flavoring syrup and concentrate — the secret formula the cola majors sell to bottlers, the fattest margin in all of beverages — sits in NAICS 311930, not here. So 312 captures the capital-heavy bottling half while the profit half is reported elsewhere [3].
• The nicotine growth engine is partly outside 312. Electronic cigarettes / vape devices are NAICS 339999 and e-liquid is 325998; tobacco farming is 111910. So "tobacco manufacturing" statistics increasingly understate where nicotine demand is actually heading [4].
• Also excluded across the subsector: grape growing (agriculture, 111332), industrial ethanol (325193), tap-water utilities (221310), juice/dairy drinks (311421/311511), dry ice (325120), self-serve ice vending (retail, 445132), leaf wholesalers (424940/425120), and all beverage/tobacco wholesalers, liquor stores, and bars (424/445/722) [3][4].

3. How big it is (rollup figures and undercount)

Our ground-truth federal statistics for the 312 level:

Average pay works out to about $58,700 per worker (annual payroll ÷ employees) — but that blends two very different labor markets: tobacco's ~125 automated plants pay roughly $87,000 per worker, well above beverages' ~$58,000 [1][3][4]. CBP = County Business Patterns; the CR ratios and HHI measure market concentration (the HHI is the sum of every firm's squared market share — higher means more concentrated).

The rollup reconciles perfectly — including firms. Every headline sums exactly across the two children: plants (12,452 + 125 = 12,577), employees (279,743 + 11,362 = 291,105), annual payroll ($16.1B + $0.99B = $17.09B), and receipts (~$124.1B + ~$50.9B = ~$175.0B) [1][2][3][4]. The firm count is the interesting one. Inside the beverage child, firms do not sum — its four children add to 11,323 against a reported 11,304, because a company that both brews and distills is counted once at the parent [2][3]. Yet across beverages and tobacco they sum exactly (11,304 + 121 = 11,425), which tells you something real: essentially no company makes both drinks and tobacco, so no parent is double-counted here. (Tobacco conglomerates once owned brewers and food companies — Philip Morris held Miller Brewing for decades — but those beverage arms were divested long ago, leaving the two sets disjoint today.)

A concentration paradox worth pausing on — sharper here than in any single child. The subsector's blended CR4 is just 34.2% and its HHI only 391.8 — a level antitrust agencies treat as effectively unconcentrated [2]. Yet one of its two children is a 90.6% near-monopoly [4]. That is not a data error: pooling a fragmented beverage industry (CR4 27.1%, HHI 263.1) with a tobacco oligopoly whose leaders do not compete with one another — Coca-Cola does not compete with Altria — mathematically dilutes measured concentration and buries the tobacco monopoly inside a benign-looking average [2][3][4]. The dilution repeats one level down: every beverage child is more concentrated than 3121 itself (beer's CR4 63.6%, spirits' 45.3%, soft drinks' 43.5%, wine's 37.8%), and where the Census publishes a child HHI it runs roughly 2.6× the beverage parent's [3]. The 312 aggregate makes this subsector look far more competitive than either market a buyer actually shops in. Read concentration at the child level, never here.

Undercount caveats — and where they do and don't apply. This is the sharpest structural contrast in the whole level:

1. Beverages are badly undercounted — but count carefully. Three of the four beverage children understate the number of operators, because small owner-operated activity is coded elsewhere or too small to capture. The Brewers Association counts 9,724 total U.S. breweries in 2025 (9,578 of them craft) against ~5,200 federal brewery establishments, the gap being brewpubs and taprooms coded as restaurants/bars [3][5]. Trade sources counted about 11,450 wineries heading into 2025, easing toward ~11,100 in 2026, versus ~4,560 federal establishments — CBP counts only businesses with paid employees, while thousands of wineries are nonemployer, "virtual," or custom-crush clients [3]. Spirits shows the trap in the other direction: the TTB recorded 5,069 active domestic beverage-alcohol permits at year-end 2024 against ~1,349 employer establishments, but 2,746 permits reported no taxable removals at all and just 50 permits accounted for ~91% of proof gallons removed [3]. So treat the federal establishment count as a floor on beverage businesses and permit or trade-association counts as a ceiling — a licensed still is not a shipping producer.
2. Tobacco is not undercounted — the distortion runs the other way. A few large corporations dominate, so the plant and firm counts are fairly complete [4]. Two other gaps matter instead. The ~$50.9B understates the industry's reach: Marlboro's international volumes are made abroad (outside the U.S. code), and the surging vape/pouch categories fall partly in adjacent NAICS codes. And the output figure itself is vintage-dependent — a 2025 Bureau of Labor Statistics sectoral-output series puts the same industry at ~$61.6 billion versus the 2022 Economic Census's ~$50.9 billion; different concepts and different years, best read as bounds rather than as a contradiction [4]. We keep $50.9B in the rollup table because that is the measure that is comparable with beverages.
3. Factory-gate, not retail — both halves. The $175.0B is what plants ship, not what consumers spend. Excise taxes and retail/hospitality margin sit on top of everything here, and are especially heavy in this subsector: federal-plus-average-state cigarette tax alone runs about $2.97 per pack [4], and alcohol carries layered excise and three-tier distributor markups. The gap is large wherever the children measure it — U.S. beer is ~$113B at retail, and bottled water goes from ~$28.2B at the producer level to ~$50.6B at retail [3]. Wine shows how uncertain even the retail number is: Silicon Valley Bank puts the 2025 U.S. wine market at about $74.3 billion while other trade estimates run $107–115 billion, and the 3121 primer reports both rather than picking [3][10]. The consumer-facing beverage-and-tobacco economy runs into the hundreds of billions above these shipment figures [3][4].

Net: the federal counts capture the subsector's economic mass (payroll, shipments, employment) well and its firm count cleanly, but they understate the number of businesses on the beverage side by roughly half, while capturing the tobacco side nearly completely.

4. The investable universe — where value concentrates

The defining fact of this level for investors: the two children require opposite strategies. Reserve tickers for this and the next section.

• Beverages (3121) — access is uneven by child. The deepest, most liquid public menu is in the non-alcoholic child — concentrate/brand owners The Coca-Cola Company (KO), PepsiCo (PEP), Keurig Dr Pepper (KDP); growth challengers Monster Beverage (MNST), Celsius Holdings (CELH), National Beverage (FIZZ), Zevia (ZVIA); the pure bottler Coca-Cola Consolidated (COKE); and the bottled-water pure-play Primo Brands (PRMB). On the alcohol side the public options narrow sharply: beer via Anheuser-Busch InBev (BUD), Molson Coors (TAP), Constellation Brands (STZ, now ~91% beer after its wine exit), craft pure-play Boston Beer (SAM), and Heineken (HEINY) via depositary receipts; spirits via one clean U.S. pure-play, Brown-Forman (BF.B), plus contract distiller MGP Ingredients and foreign majors (Diageo, Pernod Ricard, Suntory Global Spirits, Campari); and wine only as thinly-traded micro-caps (Willamette Valley Vineyards, Crimson Wine Group), which is to say not meaningfully at all. Everything else — Niagara and Red Bull, most of the ~9,580 craft breweries, the big private wine houses (E&J Gallo, The Wine Group), and large private spirits families (Sazerac, Bacardi, Heaven Hill) — is private or PE-owned [3][5].
• Tobacco (3122) — nearly the whole industry is public. Value concentrates in a few large brand owners: the U.S. income pure-play Altria (MO), ~$23.5 billion of 2025 net revenue; the global smoke-free growth story Philip Morris International (PM); and the foreign parents of U.S. businesses — British American Tobacco (BTI), which owns Reynolds American, Imperial Brands (IMB), which owns ITG Brands, and Japan Tobacco, which took Vector Group/Liggett private in October 2024. Smaller listed plays include niche brand house Turning Point Brands (TPB) and the leaf-supply chain Universal Corporation (UVV) [4].

Where value concentrates — and the ETF/ESG catch on both sides. For a public-equity investor the paradox is that the two richest pools of listed value here sit at opposite ends of the acceptability spectrum: non-alcoholic beverages (widely held, defensive staples) and tobacco (extraordinarily cash-rich but screened out by many funds). There is no dedicated U.S. beer, wine, spirits, or tobacco ETF (exchange-traded fund); passive exposure to either child comes only inside broad consumer-staples funds — and ESG-screened funds (environmental, social, and governance criteria) typically exclude both tobacco and alcohol, structurally capping the buyer base for the very names that generate the most cash [3][4]. For a private-market investor the map inverts on the beverage side (wine, most craft beer, spirits houses, bottling/ice routes are where private and PE capital actually operate) but stays thin on tobacco, where the multinationals own the core and only the edges (leaf, cigars, contract manufacturing, independent pouch/vape startups) are privately reachable.

5. How the money works

Both children run the same core equation — volume × price × mix, minus packaging and distribution cost — and both lean on the same survival lever in flat-or-declining categories: premiumization, selling less product at a higher price. But the economics sit at different intensities, and 2025 was the year the lever visibly worked in one child and slipped in the other:

• Tobacco is the purest cash-cow model in the subsector, and it still works. Cigarette pack sales fell ~9.2% in 2024 and ~8.2% in 2025, yet manufacturers raise net prices faster than volumes fall, so revenue and profit hold or grow [4]. The arithmetic that results is the reason the sector exists in a portfolio: Altria's smokeable segment turned $17.4 billion of net revenue into $11.0 billion of operating income in 2025 — a 63% operating margin — while sticks shipped kept declining [4][11]. Cigarettes are cheap to make and highly automated — the value is the brand and the addiction, not the factory — so free cash flow is strong with little reinvestment, funding high dividends and buybacks. Excise taxes and 1998 Master Settlement Agreement payments act like per-pack costs that makers largely pass through in price. The new growth engine is smoke-free — nicotine pouches (fastest-growing, led by Zyn), heated tobacco, and vapes [4].
• Beverages spread across a wider range of models, and the pricing lever is straining. Non-alcoholic splits into high-margin concentrate makers (KO, ~60% gross margins on formula) versus capital-heavy bottlers (Coca-Cola Consolidated posted a 39.7% gross margin in FY 2025 on plant, fleet, and packaging). Beer is capacity-utilization-driven and mix-led (imports, light, non-alcoholic beer), and its premiumization trade still works. Wine ties up cash in real assets and time — vineyard land takes ~4–5 years to reach full yield and wine ages before sale — and is not converting gross margin into profit: in SVB's premium-winery benchmark the 2024 gross margin was 59.1% while the pretax margin was negative 2.6%, with only about half of sampled wineries profitable, down from 76% in 2021 [3][10]. Spirits is a working-capital machine — aged whiskey sits in barrels for years, losing a few percent a year to evaporation (the "angel's share") — but in 2025 it sold more cases for less money: supplier revenue −2.2% on volume +1.9%, which is adverse mix or discounting, not premiumization [3][9].

The common tax across most of this subsector is regulation-as-cost. Alcohol producers must sell through the three-tier distribution wall (below), booking revenue at the first, lowest price point; tobacco makers carry excise and settlement payments; only non-alcoholic beverages escape a "sin"-style regime. Metrics that matter at this level: unit/pack/case/barrel volume and organic growth; price/mix and net pricing power; input costs (aluminum, glass, PET resin, grain, grapes, sweetener, tobacco leaf, diesel); capacity utilization; and — for aged spirits, wine, and any inventory-heavy line — working capital tied up before sale.

6. What drives demand

The two children share a striking macro rhyme — long-run decline in a habit, defended by pricing and format-switching:

• Secular consumption decline is the master force on both. Gallup's 2025 survey found 54% of U.S. adults drink and 44% call themselves abstainers, the weakest reading in its series, and per-capita wine fell from 3.16 gallons per resident in 2021 to 2.54 in 2024 [3][8]; the adult cigarette smoking rate hit a record-low 9.1% in 2025 — about 24 million adults — down from 9.9% in 2024 and 42.4% at the 1965 peak [4][6]. Neither habit is being replaced by younger cohorts at the old rate.
• Addiction and inelastic demand cushion tobacco more than beverages: remaining smokers are relatively price-insensitive, which is exactly what sustains tobacco's price-led revenue despite falling volumes [4].
• Format substitution reshuffles profit rather than eliminating demand. In beverages, dollars rotate from sugary soda to water/energy/sparkling, non-alcoholic beer/wine/spirits are the fastest-growing edge on the alcohol side, and ready-to-drink cocktails grew 16.4% to $3.8 billion in 2025 [3][9]. In tobacco, consumers migrate from cigarettes to nicotine pouches, vapes, and heated tobacco (U.S. pouches ~$4B in 2024, growing 30%+ a year; adult e-cigarette use was 6.7% in 2025) [4][6][7].
• Down-trading showed up on both sides in 2025 — a new shared signal. Tobacco's discount segment reached 31.8% of measured cigarette retail volume, up 2.2 percentage points on discretionary-income pressure [4]; in beverages, the reversal of premium mix in spirits and the 15% volume drop in wine's direct-to-consumer channel are the same consumer behaving the same way [3][9]. Both children's defense assumes buyers keep trading up; that assumption is now being tested in both.
• Taxes and price set volumes at the margin in both — soda taxes and alcohol/tobacco excise raise shelf prices and, for tobacco, push some buyers toward cheaper or illicit product.
• Shared wildcard — GLP-1 weight-loss drugs, a still-unproven but plausible drag on both calorie-dense drinks and alcohol appetite, a rare risk that touches most of the subsector at once [3].

7. Regulation

This subsector is among the most heavily regulated corners of U.S. consumer manufacturing, and it splits into three regimes, only one of which is light-touch:

• Non-alcoholic beverages are regulated as food by the U.S. Food and Drug Administration (FDA) — ingredient, labeling, and good-manufacturing-practice rules, plus bottled-water standards. Live fights: soda taxes; SNAP (Supplemental Nutrition Assistance Program) purchase restrictions, where USDA's March 2026 listing showed approved food-restriction waivers in 22 states, most limiting soda or sweetened beverages; sweetener/dye reformulation; packaging law (California requires 25% post-consumer recycled content in plastic beverage containers from January 2025, rising to 50% in 2030); and PFAS ("forever chemicals") scrutiny for water [3].
• Alcohol shares one heavy regime under the Alcohol and Tobacco Tax and Trade Bureau (TTB): producer permits, label approval, federal excise tax (tilted to help small producers via the Craft Beverage Modernization Act — beer at $3.50 per barrel on the first 60,000 for small brewers versus $16–18 for the majors, spirits at $13.50 per proof gallon cut to $2.70 on the first 100,000), and the post-Prohibition three-tier system forcing most producers to sell through licensed distributors to retailers, with 17 "control states." Cross-cutting risks: a January 2025 U.S. Surgeon General push for alcohol cancer-risk warning labels (not yet law), and the July 2025 U.S.–EU framework's 15% U.S. tariff on EU wine and spirits, which has since taken effect [3].
• Tobacco is regulated by the FDA's Center for Tobacco Products via the Premarket Tobacco Product Application (PMTA) process, where average review times now stretch toward ~700 days; the FDA granted first-of-its-kind reduced-risk marketing to 20 Zyn nicotine-pouch products in January 2025; federal excise runs $1.01/pack atop state taxes averaging ~$1.96; and the 1998 Master Settlement Agreement (MSA) adds perpetual per-pack payments to the states. The proposed menthol-cigarette ban was never finalized and was withdrawn in early 2025, removing one overhang, but a proposed cap on cigarette nicotine remains the biggest open question for combustibles, and an enforcement gap on illicit imported flavored vapes rounds out the picture [4].

The through-line: alcohol and tobacco are classic "sin"/excise regimes — permits, excise, and restricted distribution — while non-alcoholic drinks face ordinary food regulation. Regulatory change (a soda tax, a nicotine cap, a new tariff, a warning label) is a first-order swing factor for value here in a way it is not for most manufacturing.

8. Consolidation

Scale, brand, and distribution compound in every child, so the long-run direction is more consolidation — but the two arrive from opposite starting points, and for opposite reasons:

• Beverages consolidate within each child, and increasingly out of distress. The soft-drink majors continually refranchise bottling and buy high-growth challengers (PepsiCo–Poppi, 2025); bottled water crystallized around the November 2024 Primo Water + BlueTriton merger into Primo Brands; packaged ice combined its two largest players when Reddy Ice closed its Arctic Glacier acquisition in February 2026 — though the Department of Justice required divestitures in five states plus the New York City and Boston metros, capping how far one national roll-up can go. On the alcohol side the shakeout is now running in all three children at once: total U.S. breweries fell from 9,922 to 9,724 as the craft decline entered a second year [3][5]; Constellation exited mainstream wine (selling to The Wine Group for $846.5M in 2025), Duckhorn went private ($1.95B), Vintage Wine Estates went bankrupt, and California growers pulled out ~38,000–40,000 acres of vineyard; and the craft trade group's active-distiller count fell from 3,069 to 2,282 in a single year while Suntory paused Jim Beam production at Clermont, Kentucky for a full year from January 2026 [3].
• Tobacco is already consolidated — the deals are about pivoting, not scaling. With a CR4 of 90.6% the structure is set, and in cigarettes specifically the skew is sharper still: Altria reported 45.2% of measured U.S. cigarette retail volume in 2025, including 40.5% for Marlboro alone [4][11]. Recent M&A repositioned the majors toward smoke-free: British American Tobacco bought Reynolds American (~$49B, 2017), Philip Morris International acquired Swedish Match/Zyn (2022), and Japan Tobacco took Vector Group/Liggett private in a ~$2.4B deal completed October 2024. The battleground has shifted from cigarette share to nicotine pouches and heated tobacco [4].

The contrast is the point: beverage consolidation is capacity leaving the field — brewery, winery, and distillery counts falling together — while tobacco consolidation is the same owners changing what they sell. Because the two children's leaders don't compete, either way the 312 aggregate keeps looking benign, reinforcing the Section 3 point that concentration must be read at the child level.

9. Risks

• Secular volume decline — the shared structural risk. Moderation shrinks beer/wine/spirits; the smoking rate keeps falling with cigarettes down ~8–10%/yr. The whole subsector's model depends on pricing power outrunning volume loss — and it breaks if that ever fails [3][4].
• Pricing power is the model, and it has already slipped on one side. This is the sharpest change since the last revision: spirits' 2025 revenue decline on rising volume and wine's 15% drop in direct-to-consumer shipments show that premiumization can fail precisely when it is most needed [3][9]. Tobacco's version still worked in 2025 [4][11] — but the beverage half is the live demonstration that this defense is not guaranteed.
• Regulatory and tax pressure. Soda taxes, SNAP restrictions in 22 states, and plastics rules on non-alcoholic beverages; excise, three-tier franchise law, the now-effective 15% EU tariff, and a proposed cancer-warning label on alcohol; a possible nicotine cap, flavor bans, and steep excise increases on tobacco. Any of these can cut volumes or margins sharply [3][4].
• GLP-1 weight-loss drugs — an unproven but plausible drag spanning calorie-dense drinks and alcohol and possibly nicotine appetite [3].
• ESG exclusion and a capped buyer base — many institutional funds cannot own tobacco or alcohol, structurally limiting demand for the subsector's most cash-generative equities and pressuring their multiples [3][4].
• Litigation, settlement, and illicit competition (tobacco-specific) — perpetual MSA payments and product-liability exposure are permanent costs, unauthorized disposable vapes ran an estimated $2.4B+ of U.S. sales in 2024 (roughly 35% of convenience-store e-cigarette sales), and the smoke-free pivot carries execution risk (Altria's ~$2.2B NJOY impairment) [4].
• Oversupply, capital intensity, and illiquidity (beverage-specific) — the whiskey glut, the California grape glut, and falling brewery counts signal distressed capacity, while wine and aged spirits tie up cash in land, barrels, and multi-year inventory that is slow to sell in a downturn; roughly half of SVB's premium-winery sample was unprofitable in 2024 [3][10].
• Input-cost and tariff inflation — aluminum, glass, PET resin, grain, grapes, sweetener, tobacco leaf, and diesel hit the capital-heavy plants first; U.S. spirits exports already fell 3.8% in 2025 on trade friction [3].
• Leverage on the beverage side — post-merger and PE-legacy balance sheets can consume most of the operating profit; Primo Brands turned a 21.7% adjusted EBITDA margin into net income of just 1.2% of sales in 2025 after $326.5 million of interest and financing expense [3].

10. How to invest, and the outlook

Match the route to the child — the two halves demand different playbooks:

• Public investors get very different menus. In beverages, the richest and most liquid options are non-alcoholic — the concentrate compounder (KO) for defensive quality, the diversified staple (PEP), the pure bottler (COKE) as a leveraged volume bet, the water pure-play (PRMB), or growth challengers (MNST, CELH, FIZZ, ZVIA) — with only a short list of scaled alcohol names (BUD, TAP, STZ, Brown-Forman, MGP as a cyclical supplier bet, plus foreign majors via depositary receipts) and essentially no public wine. In tobacco, almost the whole industry is investable: Altria (MO) for U.S. income, Philip Morris International (PM) for smoke-free growth, British American Tobacco (BTI) or Imperial Brands (IMB) for U.S. exposure via a parent, plus niche TPB and supply-chain UVV. There is no dedicated ETF for either child, and ESG funds exclude both tobacco and alcohol, so passive exposure comes only through broad consumer-staples funds [3][4].
• Private investors own what the stock market doesn't. On the beverage side that is a deep opportunity set — contract/private-label bottling, regional water and ice routes, founder-led craft breweries and taprooms, wineries and vineyard land (including farmland REITs for the dirt, not the winery equity), spirits houses, and craft distilleries — though the current shakeout argues for buying assets rather than counting on the operator to survive it, and barrel or cask investment schemes deserve particular caution given the glut. On the tobacco side the private edges are narrow — leaf growing/processing, cigars, contract manufacturing, and independent pouch/vape startups (which sit in adjacent NAICS codes and face the PMTA gauntlet) [3][4].

Outlook (forward-looking judgment). Treat 312 as a low-growth, high-cash-generation subsector of two indulgence businesses moving on the same underlying tide — declining per-capita consumption — but at different intensities and from opposite structures. Tobacco is the extreme case: a shrinking, concentrated, extraordinarily profitable near-monopoly whose price-over-volume model was still delivering 60%-plus segment margins in 2025, betting its future on converting a fading combustible profit pool into leadership in pouches and heated tobacco — a high-yield holding for income investors who can accept secular decline, regulatory overhang, and ESG exclusion [4][11]. Beverages are the milder, more diversified case, but the milder case is where the model is being tested: the non-alcoholic child should keep grinding out low-single-digit growth as water and functional formats offset sugary soda, while the three alcohol children are all shrinking their operator counts at once and the premiumization defense has stopped reliably converting flat volume into rising value. Silicon Valley Bank does not expect a wine bottom before 2027–2028; spirits must work off its whiskey glut, and capacity is already coming offline, before scaled brand-owners resume compounding [3][10]. The common near-term swing factors across both halves are taxes, tariffs, and regulation (soda politics, alcohol labels, a possible nicotine cap) and the GLP-1 wildcard; the common structural direction is more consolidation, because in every business here the durable edge is brand and distribution, and both reward scale. Because the subsector's blended statistics hide a competitive beverage industry inside the same total as a tobacco monopoly, do your analysis one child down — read the 3121 and 3122 primers for company-level depth.

---

Sources

1. U.S. Census Bureau, County Business Patterns 2023 (NAICS 312 and children 3121/3122: establishments, employment, payroll). https://www.census.gov/programs-surveys/cbp.html
2. U.S. Census Bureau, 2022 Economic Census — Concentration by Largest Firms (NAICS 312: receipts, firms, CR4/CR8/CR20/CR50, HHI). https://www.census.gov/programs-surveys/economic-census.html
3. Histometrics rollup primer, Beverage Manufacturing (NAICS 3121) — federal figures, four-child structure (soft drinks/water/ice, beer, wine, spirits), investable universe, economics and premiumization strain, alcohol regulation and three-tier system, consolidation and shakeout, and undercount caveats.
4. Histometrics rollup primer, Tobacco Product Manufacturing (NAICS 3122) — federal figures, single-child chain (31223/312230), oligopoly concentration, cash-cow economics, FDA/PMTA/excise/MSA regulation, smoke-free pivot, and investable universe.
5. Brewers Association, National Beer Statistics (9,724 total U.S. breweries in 2025, 9,578 craft; 9,922 in 2024). https://www.brewersassociation.org/statistics-and-data/national-beer-stats/
6. Centers for Disease Control and Prevention / NCHS, "FastStats — Smoking" (9.1% adult cigarette smoking and 6.7% e-cigarette use in 2025; 9.9% in 2024; 42.4% in 1965). https://www.cdc.gov/nchs/fastats/smoking.htm
7. GlobeNewswire / Research and Markets, "United States Nicotine Pouches Market Forecast Report 2025-2033" (~$4B in 2024, ~30%+ annual growth). https://www.globenewswire.com/news-release/2025/04/21/3064762/0/en/United-States-Nicotine-Pouches-Market-Forecast-Report-2025-2033.html
8. Gallup, U.S. Drinking Trends (2025: 54% of adults drink, 44% abstain — lowest/highest in series). https://news.gallup.com/poll/467507/percentage-americans-drink-alcohol.aspx
9. Distilled Spirits Council of the U.S. (DISCUS), Annual Economic Briefing 2025 ($36.4B supplier sales; revenue −2.2% on volume +1.9%; RTDs +16.4% to $3.8B). https://distilledspirits.org/news/distilled-spirits-council-annual-economic-briefing-2025/
10. Silicon Valley Bank, 25th Annual State of the U.S. Wine Industry Report, January 2026 (market size $74.3B; premium-winery margins; bottom forecast 2027–2028). https://www.svb.com/news/company-news/silicon-valley-bank-releases-25th-annual-state-of-the-us-wine-industry-report/
11. Altria Group, Inc., "Altria Reports 2025 Fourth-Quarter and Full-Year Results" (smokeable segment $17.4B net revenue / $11.0B operating income; 45.2% cigarette retail share, Marlboro 40.5%; discount segment 31.8%). https://investor.altria.com/press-releases/news-details/2026/Altria-Reports-2025-Fourth-Quarter-and-Full-Year-Results-Provides-2026-Earnings-Guidance/default.aspx

Federal business statistics (establishments, employment, payroll, firms, receipts, concentration ratios, HHI) are from the U.S. Census Bureau County Business Patterns 2023 and 2022 Economic Census (Concentration), as ingested into Histometrics. Child-level figures are drawn from the 3121 and 3122 primers. Where a value is suppressed in the federal source (e.g., the tobacco-child HHI) it is not reported here. Derived per-plant and per-worker figures mix 2022 receipts with 2023 counts and are directional.