Job Creation in SBA Feasibility Studies: A Quantitative Framework

Methodology, multipliers, and asset-class benchmarks for SBA 504 and 7(a) feasibility analysis

On October 1, 2025, the Small Business Administration recalibrated the 504 program's central compliance test. The per-job debenture threshold climbed from $90,000 to $95,000 on the standard track, and from $140,000 to $150,000 for small manufacturers and projects meeting an energy public-policy goal. The change, codified at 90 FR 47117, was administrative, a routine inflation adjustment dressed up in a Federal Register notice. It also sharpened a question that the lending industry has spent the last decade managing around rather than answering.

How many jobs does a 504 project actually create? Not in the projections filed at credit approval. In the payroll records pulled twenty-four months later.

Every SBA-financed project tells two job stories. There is the regulatory story, anchored by the test at 13 CFR § 120.861, satisfied by direct, permanent, full-time-equivalent positions verified at the two-year debenture anniversary. And there is the economic story, told through the four cascading layers of impact that economists use to measure what a project does to its region. The two stories are routinely confused. Feasibility studies blend them. CDC credit memos cite the wrong multipliers. Lenders ask for "total jobs" without specifying which layer they mean. The confusion is consequential, because the gap between projected employment and realized employment in the SBA portfolio is large, persistent, and well documented.

This article reframes job creation as a quantitative discipline. The framework below covers the regulatory math, the four-layer methodology that should underlie every projection, the operational FTE benchmarks by asset class, and the compliance architecture that determines whether a project survives the 24-month checkpoint.

1. The Regulatory Foundation

The SBA 504 Job Opportunity test is older than most of the analysts who apply it. The structural logic, that federally subsidized financing should produce measurable employment, was baked into the program in 1980 and has survived every administrative recalibration since. The dollar thresholds drift with inflation. The framework does not.

Three sections of 13 CFR carry the weight. Section 120.861 sets the per-project test: the small business borrower must create or retain at least one Job Opportunity per dollar-threshold of SBA debenture. Section 120.829 sets the alternative, a CDC-portfolio-level average that allows individual loans to fall short if the development company's overall book of business clears the line. Section 120.862 enumerates eighteen "other economic development objectives" that substitute for the job test entirely.

After the October 1, 2025 update, the operative thresholds look like this. Standard 504 projects must produce one job per $95,000 of debenture. Small manufacturers in NAICS sectors 31–33 with U.S.-based production, and projects qualifying under the energy reduction or renewable energy public-policy goals, get a more generous $150,000 per job. CDCs operating in Special Geographic Areas (Alaska, Hawaii, state-designated enterprise zones, empowerment zones and enterprise communities, labor surplus areas, and Opportunity Zones) may average their entire portfolio at $150,000 per job rather than the $95,000 standard.

The regulation contains one stranded reference that practitioners should know about. 13 CFR § 120.882(g)(15), which governs debenture refinancing, still cites a $90,000 figure that was the prior threshold. SBA has not yet conformed the cross-reference. The substantive standard for new originations is $95,000.

Two procedural rules deserve special attention because they cause more compliance failures than the dollar thresholds themselves.

The first is the 24-month clock. It runs from debenture funding, not from loan closing, not from certificate of occupancy, not from the borrower's first month of operations. Because 504 debentures pool monthly and the SBA-guaranteed debenture typically sells to investors thirty to sixty days after closing, the clock effectively starts one to two months after the borrower takes possession of the financed asset. There is no extension mechanism. A project that hasn't ramped to projected staffing by the anniversary has three options: substitute a public-policy goal under § 120.862, ride the CDC portfolio average, or document the shortfall in the annual report.

The second is the 75 percent community rule, codified in SBA Information Notice 5000-1374 and carried forward into the current SOP 50 10 8 (effective June 1, 2025). A Job Opportunity does not have to be located at the project facility (a borrower with multiple sites can count jobs at other locations), but 75 percent of the counted jobs must be in the community where the project sits. Feasibility studies that aggregate FTEs across multi-location operators without verifying local-community share are the single most common documentation defect identified in OCRM reviews.

The eighteen alternative pathways at § 120.862 split into five Community Development goals and thirteen Public Policy goals. The total grew from fifteen to eighteen with the 2022 rule amendment (87 FR 38909, June 30, 2022) that added three Public Policy goals covering energy reduction, sustainable design, and renewable energy upgrades. Industry references that still cite fourteen or sixteen goals are working from the pre-2022 rule. The current five community goals cover improving or diversifying the local economy, stimulating business development, bringing new income into the community, assisting manufacturing firms, and assisting businesses in labor-surplus areas. The thirteen public policy goals cover business district revitalization (with a written redevelopment plan), expansion of exports, women-owned business development, veteran and service-disabled-veteran development, minority enterprise development, productivity and competitiveness improvements, modernization to meet health and safety or environmental requirements, assistance to firms affected by federal budget cuts or base closings, rural development, labor-surplus-area unemployment relief, reduction of energy consumption by at least 10 percent (or 15 percent renewable generation as a share of whole-building consumption under the newer formulation), plant upgrades for renewable energy and biofuels, and others.

These goals are not a fallback. They are the operational reality of the program. The Brown and Earle finding discussed later in this article, that realized job creation across the SBA portfolio runs roughly one-third of what the regulatory benchmark implies, means that public-policy-goal qualification is the path many deals take by design, not by exception.

2. The Four Layers of Employment Impact

A project creates employment in four distinct ways, and a feasibility study that doesn't distinguish them isn't a feasibility study. It is a marketing document.

The first layer is direct employment: the FTEs who work for the borrower at the financed facility (or, under the 75 percent rule, at the borrower's other locations). Direct jobs are what the SBA test counts. They are what the CDC documents at the 24-month anniversary using IRS Form 941, payroll registers, W-2 reconciliations, state unemployment insurance filings, and contractor 1099s. In input-output terminology this is "Type 0" or "final demand" employment, the labor consumed by the project's own operations.

The second layer is indirect employment: the jobs at the project's suppliers. A hotel buys laundry services, food, energy, HVAC maintenance, accounting, and marketing. Each of those purchases supports employment in upstream industries. A feasibility study that wants to talk about regional economic impact can include indirect jobs, but they do not count toward the SBA Job Opportunity test. Reporting them as if they did is the single error most likely to draw an OIG sampling finding.

The third layer is induced employment: the jobs supported by household spending of wages earned in the direct and indirect tiers. A hotel housekeeper earning $32,000 a year spends it at grocery stores, daycare centers, auto repair shops, and pharmacies. The labor at those establishments is part of the project's induced impact. Like indirect employment, induced jobs are useful for narrative and inappropriate for compliance.

The fourth layer is construction employment: the trades labor that builds the project. Construction jobs are temporary by definition, ending at certificate of occupancy. SBA Information Notice 5000-1374 confirms what § 120.861 implies: construction jobs do not count toward the Job Opportunity test, regardless of dollar magnitude. They belong in the regional economic impact narrative, properly labeled and isolated, and they should never be aggregated with operational employment in a single "jobs created" figure.

The two standard economic-impact frameworks, BEA's RIMS II and IMPLAN, handle these layers slightly differently, but the basic taxonomy is shared. RIMS II reports Type I multipliers (direct plus indirect) and Type II multipliers (direct plus indirect plus induced, with the household sector treated endogenously). IMPLAN reports Type SAM multipliers that close the system through a full social accounting matrix. The arithmetic differs at the margins; the conceptual cleanness is the same.

The BEA's own user guide contains a caveat that every feasibility-study author should commit to memory. Chapter 6 of the RIMS II User's Guide notes that "since RIMS II assumes that local workers can work on all types of construction projects, the construction multipliers may produce inflated impact estimates for projects that use specialized, nonlocal labor." Cold-storage refrigeration mechanics, hospital medical-gas specialists, modular pre-fab assembly crews: projects that depend on imported trade labor systematically overstate their local construction employment when run through unadjusted multipliers. The fix is to discount the indirect and induced portions, or to fall back to direct construction employment only, and to footnote the choice.

The other caveat worth carrying into every analysis comes from the same chapter: "If industries can increase their output without hiring as many workers as the model assumes, then using the model's multipliers will produce inflated impact estimates." Productivity gains, automation, and the substitution of capital for labor all mean that a multiplier estimated from 2017 input-output tables and 2022 regional data will systematically overstate jobs supported in 2026. Multipliers age. Directional rankings hold; absolute magnitudes drift.

3. From Project Plan to FTE: A Methodology

The discipline of translating a project's physical specifications into a defensible direct-FTE projection is a six-step sequence. None of the steps are exotic. All of them get skipped routinely.

• Step one is to lock the NAICS code. This sounds clerical and isn't. NAICS classification determines the small-manufacturer eligibility test (sectors 31–33 with U.S. production), the operational FTE benchmark to apply, the multiplier set to invoke, and the public-policy-goal pathways available. NAICS 2022, the current binding standard, includes a change that catches deals in the field constantly: gas stations moved from 447110 to 457110, and the convenience-store-with-fuel category moved from 447190 to 457120. A feasibility study citing the obsolete 447xxx codes is signaling that the analyst hasn't refreshed the methodology since 2022. CDC reviewers notice.

• Step two is to pull the operational FTE benchmark from a defensible primary source. This is where most projections go wrong. Borrower projections are aspirational; lender pro formas are deal-driven; the only honest input is industry-published staffing data anchored to operator disclosures, trade-association surveys, or government statistics. Section 5 of this article walks through the published benchmarks by asset class.

• Step three is to apply the FTE conversion correctly. Two thousand eighty hours per year is one FTE. The IRS-aligned standard incorporated into SBA Form 1253A is 130 hours per month averaged across the calendar year. The most common error in hospitality and food service projections is reporting peak-season or peak-shift headcount and calling it FTE. A resort hotel running 4.48 hours of housekeeping labor per occupied room at September peak occupancy is at 3.66 hours in March; the annualized FTE per room is calculated from the year-round average, not the September spike.

• Step four is to apply the 75 percent community filter. For multi-location borrowers, identify the community where the project sits and demonstrate that at least three-quarters of the counted FTEs are physically located there. A national restaurant operator using a 504 loan to expand a single store in Tulsa cannot count corporate headcount in Dallas toward the Tulsa project. This is where multi-location compliance fails.

• Step five is to decide owner-employee and contractor treatment. The unwritten industry practice is that owner-operators count as a Job Opportunity if they work full-time in the business. Best practice is to document their compensation structure (W-2 versus K-1 versus Schedule C), hours, and IRS reporting, and to disclose the treatment in both the feasibility study and the credit memo. Contractors count when they are dedicated to the operating business, work in a regular and continuous pattern that aggregates to at least 40 hours per week, and have a documented 1099 trail. One-off vendors do not count. Construction labor never counts.

Step six is to memorialize the projection in SBA Form 1253 at credit approval, signed by the borrower. This is the figure against which the 24-month verification compares. A projection that the borrower can't credibly attest to today will be a documentation problem in two years.

A note on the construction-employment layer, since it is the most-mishandled piece of the cascade. The standard anchor is the BEA RIMS II Industry Aggregation 7 (Construction, NAICS Sector 23) Type II Final-Demand Employment Multiplier, which the California High-Speed Rail Bakersfield-Palmdale Final EIR/EIS publishes at 10.9433 jobs per $1 million of construction outlay, in 2013 dollars, for the two-county Kern and Los Angeles County resource study area. The corresponding Direct-Effect Multiplier is 1.9576, which expresses total jobs supported per one direct construction job, a ratio rather than a per-dollar figure. The two values are not interchangeable. Applied together, the Type II final-demand figure implies roughly 5.6 direct on-site jobs per $1M of construction, 7 to 8 with indirect, and 10.9 with induced. Practitioners applying this multiplier to other regions or to nominal 2024 to 2026 dollars should obtain a region-specific RIMS II table from BEA or deflate to the underlying 2013 dollar base; the Bakersfield-Palmdale figure is illustrative, not nationally generalizable.

The Economic Policy Institute's Bivens (2019) update reports a closely tracking 2.26x total indirect multiplier on direct construction employment (226.1 additional jobs supported in supplier and induced channels per 100 direct construction jobs, for a 3.26x grossed-up total). These are job-years, one FTE for one year, not permanent positions. A $10M construction project supporting 60 job-years over an 18-month build supports 40 simultaneous FTEs on average, with peak labor during MEP and finishes running 1.3 to 1.5 times the average.

One regulatory clarification worth including in every feasibility study: SBA 504 projects are not subject to Davis-Bacon prevailing wage requirements. The Davis-Bacon Related Acts list at the Department of Labor's Wage and Hour Division, and at HUD Handbook 1344.1 Chapter 2, does not include SBA 504 or 7(a). Standard 504 commercial construction is priced at market wages, not DOL prevailing wages, unless a co-funding source (HUD CDBG, FHWA federal-aid, USDA Rural Development construction) independently triggers Davis-Bacon, in which case it attaches to the entire project scope. Thirty-two states have "Little Davis-Bacon" laws that may apply via state-funded layers (Points North, 2025). The market-versus-prevailing differential runs 0 to 30 percent or more depending on region.

4. Multiplier Discipline

The choice between RIMS II and IMPLAN is, for most SBA feasibility studies, a non-issue. RIMS II is government-published, free to cite, accepted in every federal environmental impact statement, and adequate for the level of precision the program demands. IMPLAN offers finer industry resolution and zip-code-level geography but costs money and adds little to a 504 narrative.

What matters is not the model. It is the multiplier values, the closure type, and the region.

The most-cited construction anchor is the RIMS II Type II Industry Aggregation 7 figure of 10.9433 jobs per $1M, sourced from the California HSR Bakersfield-Palmdale FEIR Appendix 3.18-A, Table 3.18-A-1. The Nuclear Regulatory Commission's Interim Storage Partners EIS publishes Type I and Type II tables for the Augusta, Georgia region that are equally citable. Federal EIS appendices are the cleanest publicly available source of region-specific RIMS II values; the BEA sells region-specific multipliers directly to feasibility-study authors for $500 per region or $150 per industry, with current data based on 2023 regional input-output relationships and the 2017 national benchmark.

For operational impact, the multipliers worth knowing are these. Accommodation (NAICS 721) runs about 10.97 Type II jobs per $1M, with a Direct-Effect of 1.28. Food services and drinking places (NAICS 722) runs 16.53 jobs per $1M, with Direct-Effect of 1.18. Nursing and residential care facilities (NAICS 623) runs 19.13 jobs per $1M, Direct-Effect 1.20. Food and beverage manufacturing (NAICS 311–312) runs 8.95 jobs per $1M but with a much higher Direct-Effect of 2.49, meaning manufacturing pulls more supplier employment per dollar of direct labor. Real estate (the self-storage proxy at NAICS 531) runs 7.87 jobs per $1M, Direct-Effect 1.32. IMPLAN values typically run 5 to 15 percent higher than RIMS II at the same industry level.

Three discipline points separate competent multiplier work from sloppy multiplier work.

First, distinguish Type I from Type II. Some lenders specifically want indirect impact excluded, in which case the Type I (Direct plus Indirect, no household closure) is the right figure. Reporting Type II without disclosing the induced component is the kind of analytical opacity that erodes credibility when a reviewer notices.

Second, watch the NAICS 2022 boundary. Gas stations, convenience stores, and several adjacent retail categories changed codes in the 2022 revision. A multiplier pulled from a 2017-vintage source applied to a 2022-vintage NAICS code is dimensionally wrong. The fix is to either restate the multiplier using NAICS 2022 mapping or cite the conversion explicitly.

Third, for specialized projects, drop to trade-level analysis. NAICS 238 specialty trade contractors split into electrical, plumbing/HVAC, foundation/structure, finishing, and other specialty trades. A cold-storage project depends heavily on 238220 refrigeration mechanics; a hospitality project on 238210 electrical and 238330 hardwood flooring. When trades are imported rather than local, the general construction multiplier overstates regional impact and a more conservative direct-only analysis is appropriate.

5. Asset-Class Benchmarks

This section is the empirical heart of the article. The numbers below are the operational FTE benchmarks that translate physical specifications (hotel keys, restaurant seats, storage units, car-wash format) into defensible direct-employment projections. They are drawn from operator 10-Ks, trade-association data, BLS QCEW and OEWS tables, and named industry sources. Borrower projections that materially exceed these ranges require operator-specific justification; projections that fall below them probably are not tight enough.

Hotels

The cleanest hotel benchmark in the field is HotelData.com's Hours Per Occupied Room (HPOR) series, drawn from Actabl Hotel Effectiveness data. The Q3 2025 chain-scale averages are: 1.30 HPOR for extended stay, 1.44 for select service, 2.57 for full service, and 4.48 for resorts. Converted to FTE per occupied room using the 2,080-hour standard, those figures translate to 0.228, 0.252, 0.451, and 0.786 respectively. At a typical 65 percent occupancy, FTE per available room runs about 0.148 (extended stay), 0.164 (select service), 0.293 (full service), and 0.511 (resort).

The American Hotel and Lodging Association's 2025 State of the Industry Report records 2.15 million direct U.S. hotel employees in 2024, with 2025 forecast at 2.17 million, both well short of the 2.37 million the industry employed in 2019. CBRE Trends® showed total hotel salaries, wages, and employee benefits rising 4.8 percent in 2024 across all departments.

Translated to per-million-dollar terms, hotels run roughly 1.5 to 2.5 FTE per $1M of project cost at the limited and select-service end (typical project cost $180 to $337 per SF, $167K to $223K per key), 2.4 to 3.6 FTE per $1M at the upper upscale tier, and 2.5 to 5.0 FTE per $1M for luxury and resort properties where labor intensity climbs but per-key construction cost climbs faster.

Multifamily

Multifamily property management operates to the longstanding "1:100 rule," one onsite FTE per 100 units, with Class A urban high-rise running closer to 1:75 and Class C garden product running 1:100 to 1:150 inclusive of maintenance. The NAA Income/Expense IQ 2023 benchmark places salaries and personnel at roughly $1,023 per unit per year for garden product, or about 9.3 percent of gross potential rent.

A 250-unit Class A garden property typically staffs a property manager, an assistant manager, one to two leasing consultants, a maintenance supervisor, one to two maintenance technicians, and a half-to-one groundskeeper: total 5.5 to 8.5 FTE, or 0.022 to 0.034 FTE per unit. At $200K to $300K per unit construction cost, multifamily generates only 0.08 to 0.15 FTE per $1M of project cost. It is the least labor-intensive of the labor-intensive asset classes and almost never carries a 504 project on direct jobs alone. (SBA 504 cannot finance investment multifamily, but mixed-use deals with multifamily components are common.)

Senior Living

Five distinct acuity tiers, each with materially different staffing intensity. Independent living runs 0.20 to 0.30 FTE per occupied unit; assisted living 0.55 to 0.70; memory care 0.85 to 1.10; CCRC blended (excluding the SNF wing) 0.50 to 0.70; skilled nursing 0.95 to 1.10 per occupied bed.

The CMS minimum staffing rule story is unresolved. CMS announced a 3.48 hours-per-resident-day standard on April 22, 2024, published in the Federal Register on May 10, 2024 at 89 FR 40876. Section 71111 of P.L. 119-21, signed July 4, 2025, blocked CMS from enforcing it through September 30, 2034. CMS published an interim final rule on December 3, 2025 (FR Doc. 2025-21792) formally rescinding the standards. Current national Payroll-Based Journal data shows average total nursing HPRD of 3.68 in Q1 2024 (LTCCC) and 3.85 in 2025 (KFF). For feasibility purposes, SNF projects should be analyzed against state minimum staffing laws plus the current PBJ benchmark, not the rescinded federal floor.

Argentum's 2023 Workforce Projections report puts direct senior-living industry employment at 888,500 jobs in 2021 across roughly one million units. ASHA's State of Seniors Housing 2024 places labor expense at approximately 55 percent of total operating expense industry-wide. Senior living is the densest job-creator per dollar of any common SBA asset class: AL economics at $350K per unit and 0.65 FTE per unit produces about 1.8 to 2.2 FTE per $1M of project cost; SNF at $300K per bed and 1.0 FTE per bed produces roughly 3.0 to 4.0 FTE per $1M, the highest job density of any asset class covered here.

Restaurants

The National Restaurant Association's 2024 State of the Restaurant Industry puts industry employment at 15.7 million workers across approximately one million U.S. locations, with the 2026 forecast at 15.8 million by year-end. The NRA's 2025 Restaurant Operations Data Abstract, drawn from more than 900 operators reporting FY2024 data, places median labor cost at 36.5 percent of sales for full-service restaurants and 31.7 percent for limited-service. These are operating medians for profitable restaurants, not the 30 percent benchmark often cited as a target.

By tier: QSR runs 15 to 30 FTE per location, 0.3 to 0.6 FTE per seat, with labor at 25 to 32 percent of sales. Fast casual runs 25 to 40 FTE, 0.4 to 0.7 FTE per seat, labor at 28 to 34 percent. Casual dining runs 40 to 80 FTE, 0.6 to 1.0 FTE per seat, labor at 33 to 38 percent. Fine dining runs 50 to 100+ FTE, 0.8 to 1.5 FTE per seat, labor at 36 to 42 percent or more.

The complication unique to restaurants is turnover. BLS JOLTS data for accommodation and food services shows total separations exceeding 75 percent annually in recent years, among the highest in any U.S. sector. New restaurants will not achieve top-quartile labor efficiency in their first 18 to 24 months. Feasibility projections that lean on aspirational labor cost percentages rather than industry-median performance are setting up 24-month documentation problems.

Wedding Venues and Event Centers

The structural pattern is a small W-2 core supplemented by per-event labor that is typically contracted or 1099. The Knot's Real Weddings Study 2023 identifies banquet halls (20 percent), farms/barns/ranches (18 percent), and historic buildings (14 percent) as the most common reception venues.

Industry-standard per-event ratios run one server per 25 cocktail guests, one server per 12 plated seats, one bartender plus barback per 50 to 75 guests, and one banquet captain per 150 guests. A 450 to 500-guest luxury wedding requires roughly 45 to 55 event-labor positions. A core W-2 team for a boutique wedding venue is 3 to 5 FTE; for a luxury multipurpose venue running 80 to 120 events per year, 10 to 20 W-2 plus annualized event labor totaling 15 to 30 FTE.

Self-Storage

The capital-intensity inflection point begins here. Public Storage's FY2024 10-K discloses 5,900 total employees across 3,073 consolidated facilities, of which 5,120 are customer-facing, 340 are field management, and 440 are corporate. Property-level FTE is therefore closer to 1.67 per facility, with the platform overhead spread across the portfolio. Extra Space Storage's FY2024 10-K shows 8,012 employees across the post-Life Storage merger portfolio, broadly tracking the same per-facility ratio. A typical 600-unit, 75,000-SF facility costs $5 to $8 million to build and supports one site manager plus a relief manager, 1.0 to 1.5 site-level FTE, backed by a district manager covering 10 to 15 sites.

At a $5M project cost with $2M of SBA debenture and 2 direct FTE, the math is $1 million of debenture per job. The standard $95,000 threshold requires roughly 21 jobs. The deal fails by 10x on direct employment alone. The compliance pathway is § 120.862, most commonly the energy reduction goal achieved through LED lighting plus a rooftop solar canopy. This is not a workaround. It is the design intent of the regulation as applied to capital-intensive real estate.

RV and Boat Storage

Thinner than self-storage. RecNation, the largest U.S. operator, runs many sites fully unmanned, supported by a 24/7 customer-success team. A 100 to 300-stall open-air facility is 0.5 to 1.5 FTE; covered or valet operations run 1 to 3 FTE. A typical 108-stall facility at $2M project cost supports 1.0 to 1.5 direct FTE, $1.0 to $1.5 million of debenture per job. Same compliance answer as self-storage: § 120.862, typically rural development plus energy reduction.

Car Washes

Format determines almost everything. Mister Car Wash's FY2024 10-K, the largest U.S. express tunnel operator, discloses 6,836 employees across 548 locations, or 12.5 FTE per location. Express tunnels run 8 to 15 FTE; in-bay automatic 1 to 3; flex serve 10 to 25; full service (a declining segment) 20 to 40; self-serve coin or card 0 to 2.

An express tunnel at $5.5M project cost with $2.2M of debenture and 12 FTE produces $183K per job, fails the $95K standard but lands within range. The standard route is the energy reduction PPG via water reclaim systems (most modern express tunnels reclaim more than 50 percent of water) and frequency-drive blowers. Combined with the $150K manufacturer/energy rate, 12 FTE clears compliance comfortably.

Convenience Stores and Travel Centers

The NACS 2025 State of the Industry data places industry employment at 2.75 million jobs across 151,975 stores, or 19.9 FTE per c-store on average. Average store-level hourly wage is $15.04. Adding a QSR concept (Subway, Hunt Brothers Pizza, Chester's Chicken) adds 4 to 8 FTE. A dual-QSR store can reach 25 to 30 FTE.

Travel centers operate at a different scale entirely. Pilot Travel Centers employs approximately 30,000 team members across more than 900 locations, about 33 FTE per location. Love's Travel Stops operates more than 670 locations and disclosed more than 40,000 employees as of its most recent corporate publications, implying roughly 60 FTE per location. TravelCenters of America's last standalone 10-K (FY2022, pre-BP acquisition) showed 15,000 full-time plus 2,500 part-time employees across 281 travel centers, about 62 FTE per location.

A standard c-store plus fuel plus dual QSR at $5 to $6M project cost with $2.4M debenture and 22 FTE works out to roughly $109K per job, fails the $95K standard but is close. Rural location qualifies for the rural development PPG almost automatically; EV charging qualifies under energy reduction. Travel centers clear the test on direct jobs alone in nearly every case.

Manufacturing

Manufacturing is the asset class where 504 financing works cleanly. Vertical IQ data citing BLS shows commercial bakeries (NAICS 311812) averaging 46 FTE on $14 million of revenue, or 3.3 FTE per $1M of revenue. The Brewers Association's 2024 report places craft brewing employment at 197,112 workers across 9,612 active breweries, 20.5 FTE per facility industry-wide, with microbreweries and brewpubs in the 5 to 15 FTE range. A $3M food manufacturing project with $1.2M of debenture and 20 FTE works out to roughly $60K per job, passes the standard $95K threshold and benefits from the manufacturer's $150K rate as belt and suspenders, plus the "Assistance to Manufacturers" community-development goal under § 120.862.

6. The Capital-Intensity Problem

Self-storage and RV/boat storage fail the SBA Job Opportunity test by an order of magnitude on direct employment alone. Express car washes fail by less, but they fail. Standalone gas stations without QSRs fail. Single-tenant industrial warehouse projects often fail. The pattern is structural. The asset classes that anchor a growing share of the 504 program (real estate-heavy, automation-enabled, low-headcount businesses) were not the businesses Congress had in mind when it wrote a 1980 statute oriented around manufacturing employment.

The math is easy. A $5M project funded with a standard 504 structure carries roughly $2M of SBA debenture. At $95K per job, the project needs 21 direct permanent FTEs to pass on direct employment alone. At $150K per job under the manufacturer/energy rate, it needs 13. Self-storage produces 2. RV storage produces 1 to 1.5. Express car wash produces 12. The gap between direct employment and threshold is not closed by tighter analytical work. It is closed by structural reliance on § 120.862.

This isn't a regulatory failure. The eighteen alternative pathways at § 120.862 exist precisely because the per-project job test was always understood to be one of several routes to program eligibility. What has changed is the proportion of the portfolio relying on the alternative routes. Real estate-heavy capital-intensive deals (storage, car wash, fuel and convenience, light industrial) have grown materially as a share of 504 originations over the past decade. The Job Opportunity test in its pure form would disqualify a meaningful slice of these. Public-policy-goal substitution makes them eligible.

The academic literature confirms what the practitioners already know. Brown and Earle, writing in the Journal of Financein 2017 using the universe of SBA loan recipients matched to U.S. Census employer registers, found that SBA lending produces "an increase of 3 to 4 jobs for each million dollars of loans." The post-October 2025 regulatory benchmark of $95,000 per job implies the program should produce roughly 10.5 jobs per $1M. The realized figure is roughly one-third of that. Some of the gap is definitional, since the regulatory test counts retained as well as created jobs, and Brown and Earle measure incremental employment relative to a control group. But the bulk of the gap is real. Across the portfolio, many 504 projects underperform their projections, and program eligibility is sustained through portfolio-level averaging and § 120.862 substitution, not through pure per-project job production.

The implication for feasibility analysts is direct: build a § 120.862 basis into every credit memo from the start. The dominant compliance failure mode is post-hoc PPG substitution that does not survive OIG sampling because the original credit memo identified only a job-test path. A premium feasibility study identifies both: a primary job-test pathway where the math works, and a belt-and-suspenders PPG basis where the math is tight. For self-storage, RV/boat storage, and express car wash, the PPG basis is primary and the job test is secondary or absent.

7. Compliance, Verification, and the Audit Trail

What happens at month 24 separates well-run CDCs from the rest. The mechanics are unforgiving. The CDC must obtain, in writing on borrower letterhead or in an email from the borrower, the actual FTE count as of the second anniversary of debenture funding. The supporting documentation package is specific: IRS Form 941 quarterly federal tax returns covering the verification period, payroll registers reconciled to W-2 totals, state unemployment insurance filings, contractor 1099s for any non-W-2 labor counted, and explicit identification of the community where each job is located to confirm the 75 percent rule.

The CDC reports actual job numbers to SBA on Form 1253A, which differentiates debentures less than two years old (estimated jobs) from those two or more years old (actual jobs confirmed in writing by the borrower). The CDC's portfolio-level dollar-per-job average is calculated, with Small Manufacturers separated as required and Special Geographic Area loans optionally separated. If the CDC's portfolio falls below the threshold, 13 CFR § 120.829(c) allows it to retain certification only if it justifies the failure to SBA's satisfaction and demonstrates how it intends to attain the required average.

The Office of Credit Risk Management's review framework, known internally as SMART, evaluates each CDC across solvency, management, asset quality, regulatory compliance, and mission. The post-GAO-14-233 baseline is an annual sample of five loans per CDC that have reached the two-year debenture funding anniversary, with documentation review embedded in both off-site annual reviews and on-site risk-based examinations. On-site reviews follow a risk-tiered cycle. CDCs with deteriorating portfolio metrics, audit findings, or financial-condition concerns get visited more frequently.

GAO-14-233, published in March 2014, is the report that drove most of the current verification architecture. GAO found that SBA had no standardized methodology for how CDCs should calculate jobs created and retained, and that SBA reviewers did not require supporting documentation. SBA closed the recommendations by 2016 through three actions: Information Notice 5000-1374 defining jobs created and retained with calculation examples, institution of the annual five-loan CDC sample, and embedding of documentation review into OCRM templates. The result is the framework practitioners work within today.

The OIG audit record is sparse on standalone job-verification audits but rich on the credit-analysis deficiencies that produce job-test problems. In OIG Report 10-10 (March 2010), OIG sampled 25 loans from 1,169 FY2008 disbursements at the three largest Premier Certified Lenders and identified 10 with inadequate credit analysis. Statistical projection from that sample yielded an estimate that approximately $56.4 million in 504 program funds may have been improperly approved that fiscal year, with more than half of the sampled deficient loans lacking adequate support for eligibility determinations including the job creation/retention basis. The pattern OIG identified then, credit memos that don't document the eligibility analysis sufficiently to survive sampling, remains the dominant compliance failure mode in the program.

Three compliance pitfalls are worth flagging explicitly because they recur in every audit cycle.

The first is the construction-jobs error. Construction jobs do not count toward the Job Opportunity test. Information Notice 5000-1374 is unambiguous. Feasibility studies that aggregate construction employment with operational employment in a single "jobs created" figure are setting up a compliance problem. Construction belongs in the regional impact narrative, labeled and isolated.

The second is the FTE conversion error. The IRS-aligned standard incorporated into Form 1253A is 130 hours per month averaged across the calendar year. Peak headcount, peak-season headcount, and total W-2 count all overstate FTE in heavily seasonal or part-time-heavy industries. Hospitality and food service feasibility studies routinely conflate the three. The fix is annualized FTE based on actual hours worked.

The third is post-hoc public-policy-goal substitution. The cleanest compliance posture is to identify the PPG basis at credit approval, in the credit memo, with supporting documentation. A project that fails its job test at month 24 and only then asserts a PPG basis, for the first time, without contemporaneous documentation, is exposed if the loan ends up in an OIG sample. Best practice in well-run CDCs is to identify both a primary and a fallback pathway at approval.

The compliance question that gets asked least often and matters most is what the CDC actually does when a project misses. The realistic answers are three. Substitute under § 120.862 if a goal was identified in the credit memo (clean). Ride the CDC portfolio average if the CDC's overall book of business clears the threshold (acceptable, but creates dependency on other loans). Document the shortfall and accept it (acceptable, but degrades the CDC's portfolio metrics over time and can trigger OCRM scrutiny if it recurs).

8. The Framework

A feasibility study that handles job creation rigorously performs eight discrete functions. None of them are difficult. All of them are routinely skipped.

It locks the NAICS code at NAICS 2022, the binding current standard, and confirms small-manufacturer eligibility and PPG pathway availability before any FTE math begins. It calibrates operational FTE projections to published asset-class benchmarks (operator 10-K disclosures, trade-association data, BLS QCEW and OEWS) rather than to borrower aspiration. It distinguishes Direct, Indirect, Induced, and Construction-phase employment cleanly, with each layer labeled and isolated. It applies the IRS-aligned 130-hours-per-month FTE conversion and documents seasonality and contracted-labor treatment. It explicitly addresses the 75 percent community rule for multi-location borrowers. It identifies a primary compliance pathway (job test or PPG) and a fallback, with both supported by contemporaneous documentation. It provides the borrower and CDC with a 24-month verification roadmap specifying the documentation package required. And it acknowledges where projections sit relative to the empirical realization rate observed across the program.

The reframe matters. Job projections in SBA feasibility studies are not pro forma decoration. They are the audit defense file. A rigorous feasibility study reduces the probability of a 24-month documentation problem from material to negligible, gives the CDC a credit memo that survives OCRM and OIG sampling, and gives the borrower a roadmap to verification rather than a surprise at month 23.

The October 1, 2025 threshold update was administrative. The underlying discipline is not.

May 13, 2026, by Michal Mohelsky, J.D. Principal of MMCG Invest, LLC, feasibility study consultant serving feasibility studies for SBA projects.

Reach out to discuss how our methodology supports your lending decision.

Michal Mohelsky, J.D. | Principal | mmcginvest.com 

Contact: michal@mmcginvest.com

Phone: (628) 225-1125

Disclaimer: This report is provided for informational purposes only and does not constitute investment advice. Data presented herein is derived from proprietary MMCG databases and third-party sources believed to be reliable; however, MMCG Invest makes no representation as to the accuracy or completeness of such information. Figures from third-party industry databases have been independently verified and, where appropriate, adjusted to reflect MMCG's proprietary analytical methodology. Past performance is not indicative of future results.

Sources

Primary Regulatory and Statutory Authority

13 CFR § 120.861, "Job Opportunity average a Project must create or retain." U.S. Code of Federal Regulations.

13 CFR § 120.862, "Other economic development objectives." U.S. Code of Federal Regulations. Eighteen alternative pathways (five Community Development goals and thirteen Public Policy goals), as amended at 87 FR 38909 (June 30, 2022).

13 CFR § 120.829, "Job Opportunity average a CDC must maintain." U.S. Code of Federal Regulations.

13 CFR § 120.882(g)(15), "Eligible Project costs for 504 loans" (per-employee 504 loan cap for stranded refinancing). U.S. Code of Federal Regulations.

90 FR 47117, "Notification of Changes to Development Company Loan Program: Job Creation and Retention Requirements; Additional Areas for Higher Portfolio Average; Request for Comments." Small Business Administration Notice published in the Federal Register September 30, 2025 (FR Doc. 2025-19072). Applicability date October 1, 2025; public comment period closed December 1, 2025.

89 FR 40876, "Medicare and Medicaid Programs; Minimum Staffing Standards for Long-Term Care Facilities and Medicaid Institutional Payment Transparency Reporting." Centers for Medicare and Medicaid Services final rule published May 10, 2024.

Section 71111 of Public Law 119-21, "One Big Beautiful Bill Act," signed July 4, 2025. Prohibits CMS from implementing, administering, or enforcing the minimum-staffing standards in 89 FR 40876 through September 30, 2034.

90 FR 55687, "Medicare and Medicaid Programs; Repeal of Minimum Staffing Standards for Long-Term Care Facilities." Centers for Medicare and Medicaid Services interim final rule published December 3, 2025 (FR Doc. 2025-21792).

SBA Standard Operating Procedure 50 10 8, "Lender and Development Company Loan Programs." Effective June 1, 2025; supersedes SOP 50 10 7.1.

SBA Information Notice 5000-1374, "Certified Development Company Reporting of Job Creation and Retention." Carries the 75 percent community rule and the 24-month verification clock from debenture funding.

Federal Oversight and Audit Reports

U.S. Government Accountability Office, GAO-14-233, "Small Business Administration: Actions Needed to Ensure Planned Improvements Address Key Requirements of the Development Company (504) Loan Program." Published March 6, 2014; publicly released April 7, 2014.

SBA Office of Inspector General, OIG Report 10-10, sample-based audit of 504 loan job creation documentation across FY2008 disbursements.

Federal Economic-Impact Methodology

U.S. Bureau of Economic Analysis, Regional Input-Output Modeling System (RIMS II). Final-Demand and Direct-Effect Employment Multipliers; 2017 national benchmark with regional data through 2023. Region-specific tables available from BEA at $500 per region or $150 per industry.

California High-Speed Rail Authority, Bakersfield to Palmdale Section, Final Environmental Impact Report / Environmental Impact Statement, Appendix 3.18-A: Regional Growth Methodology Memorandum, Table 3.18-A-1. May 2021. Source of the Construction (RIMS II Industry Aggregation 7) Final-Demand Employment Multiplier of 10.9433 and Direct-Effect Employment Multiplier of 1.9576, two-county Kern and Los Angeles County resource study area, in 2013 dollars.

IMPLAN Group, IMPLAN regional economic modeling system. Comparative input-output framework cited for methodology discussion.

Federal Labor Market and Workforce Statistics

U.S. Bureau of Labor Statistics, Quarterly Census of Employment and Wages (QCEW). NAICS-level employment and wage tabulations.

U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics (OEWS). Occupational wage estimates by industry and area.

U.S. Bureau of Labor Statistics, Job Openings and Labor Turnover Survey (JOLTS). Accommodation and Food Services separation rates.

U.S. Centers for Medicare and Medicaid Services, Payroll-Based Journal (PBJ). Skilled nursing facility staffing reports by HPRD.

Long Term Care Community Coalition (LTCCC), analysis of CMS PBJ data, Q1 2024.

Kaiser Family Foundation (KFF), analysis of CMS PBJ data, 2025.

Academic Research

Brown, J. David, and John S. Earle, "Finance and Growth at the Firm Level: Evidence from SBA Loans." Journal of Finance, 2017. Empirical analysis comparing realized to projected job creation across the SBA 504 portfolio.

Trade Association Publications

American Hotel and Lodging Association (AHLA), 2025 State of the Industry Report. Direct U.S. hotel employment figures: 2.15 million in 2024, 2.17 million forecast 2025.

CBRE, Trends in the Hotel Industry, 2024 edition. Hotel operating cost benchmarks.

National Restaurant Association (NRA), Restaurant Operations Data Abstract (ROPDA) 2025. Drawn from more than 900 operator participants for fiscal year 2024.

National Restaurant Association (NRA), 2024 State of the Restaurant Industry. Industry employment of 15.7 million across approximately one million U.S. locations.

NACS, State of the Industry presentation, 2026 Summit (April 14-16, 2026), covering calendar year 2025 industry data. Industry employment of 2.75 million across 151,975 stores nationally; 19.9 FTE per c-store baseline; average store-level hourly wage $15.04.

Institute of Real Estate Management (IREM), Income/Expense IQ National Summary 2023 (multifamily). Per-unit payroll trend data and personnel expense as share of gross potential rent.

National Apartment Association (NAA), Income/Expense IQ benchmarking program (Lobby CRE platform; discontinued December 31, 2025).

American Seniors Housing Association (ASHA), State of Seniors Housing 2024. Labor expense as share of operating expense industry-wide.

Argentum, Senior Living Workforce Projections 2023. Direct senior-living industry employment of 888,500 in 2021.

Brewers Association, Annual Craft Brewing Industry Production Report, updated May 6, 2025. Craft brewing employment of 197,112 workers across 9,796 operating U.S. craft breweries.

The Knot, Real Weddings Study 2023. Wedding venue mix data: banquet halls (20 percent), farms / barns / ranches (18 percent), historic buildings and homes (14 percent).

Industry Data Services

HotelData.com, Hours Per Occupied Room (HPOR) series, H2 2024 and Q1 2025 readings, drawn from the Actabl Hotel Effectiveness dataset.

Actabl, Hotel Effectiveness data platform. Underlying source for the HotelData.com HPOR series.

Vertical IQ, industry profiles for commercial bakeries (NAICS 311812) and related manufacturing sectors.

CoStar Group, commercial real estate analytics platform.

STR (Smith Travel Research), hotel performance data.

Operator SEC Filings (Form 10-K)

Public Storage (NYSE: PSA), Annual Report on Form 10-K, fiscal year ended December 31, 2024. Filed with the U.S. Securities and Exchange Commission. Discloses approximately 5,900 total employees (5,120 customer-facing, 340 field management, 440 corporate) across 3,073 consolidated self-storage facilities.

Extra Space Storage Inc. (NYSE: EXR), Annual Report on Form 10-K, fiscal year ended December 31, 2024. Filed with the U.S. Securities and Exchange Commission. Discloses 8,012 employees across the post-Life Storage merger portfolio.

Mister Car Wash, Inc. (NYSE: MCW), Annual Report on Form 10-K, fiscal year ended December 31, 2025; filed February 18, 2026. Discloses 6,836 employees across 548 locations.

Pilot Travel Centers, corporate disclosures. Approximately 30,000 team members across more than 900 locations in 44 U.S. states and five Canadian provinces.

Love's Travel Stops & Country Stores, corporate disclosures. More than 670 locations and more than 40,000 employees.

NAICS Classification References

U.S. Census Bureau, North American Industry Classification System (NAICS) 2022 Manual. Note the NAICS 2022 reclassification of gasoline stations from 447110 to 457110 and convenience-with-fuel from 447190 to 457120, material to the cross-walking of pre-2022 RIMS II multiplier tables.