Cost, construction type and the labor package — Texas

Four passes, but the locked number lives in a different place

The four-pass structure — napkin, concept estimate, application budget, final reconciliation — is universal to the industry, not California-specific, and it applies in Texas unchanged. What differs is which document holds the locked number and how it gets re-tested.

Two structural differences from California are worth flagging before anything else. Texas runs a single annual competitive 9% cycle governed by the QAP's program calendar, while 4% Housing Tax Credit applications are accepted on a rolling, non-competitive basis all year, gated only by Texas Bond Review Board private-activity-bond allocation rather than by a competitive round. And a 4% deal can be structured with either TDHCA itself as bond issuer or a local Housing Finance Corporation (Texas Local Government Code Chapter 394) as issuer — a choice with real timeline and cost consequences that has no California parallel.

Who is actually in the room is close to California's cast — developer project manager, architect, GC preconstruction estimator, LIHTC consultant — but a Texas deal only adds a labor-compliance specialist when a specific federal funding source is actually layered into the stack, not by default and not because of a general state prevailing-wage law.

Cost of Development per Square Foot — a scoring item, not a gate

10 TAC §11.9(e)(2) of the 2026 QAP awards up to 12 competitive points for staying under a published cost-per-square-foot ceiling. The calculation divides Eligible Building Costs — voluntarily included in Eligible Basis, excluding structured parking or non-includable commercial space — by Net Rentable Area as shown on the Rent Schedule.

This is the single most important structural contrast in the domain. An application that exceeds every one of these figures is not disqualified — it simply scores zero points on this item and competes for an allocation with a weaker application. There is no Texas mechanism that reads like CTCAC's §10325(d) "staff shall not recommend the project for credits." A Texas cost calculator has to report points at risk, not eligibility at risk.

The 5% reconciliation rule — Texas's real cost check

10 TAC §11.302(e) is explicit: "For New Construction Developments, the Underwriter's total cost estimate will be used unless the Applicant's Total Housing Development Cost is within 5% of the Underwriter's estimate." The underwriter builds an independent number using the Marshall & Swift Residential Cost Handbook and historical cost-certification data; if the applicant's figure diverges by more than 5%, the underwriter's number replaces it for underwriting purposes. For Rehabilitation and Adaptive Reuse, the comparison point is the Scope and Cost Review required under §11.306 instead.

The consequence of an unreasonably high cost estimate in Texas is a smaller feasible deal — eligible basis, and therefore the credit request, gets sized to the lower of the two numbers — not outright disqualification of the application. That is easy to underestimate: because it is a reconciliation rather than a hard stop, it is tempting to treat it as a soft warning, and it is not.

Feasibility itself is tested separately, under 10 TAC §11.302(i), and it is a market-and-debt-coverage test, not a cost-ceiling test.

A Texas deal can be "feasible" under this test at almost any absolute cost level, provided the rents and debt service pencil. Cost discipline in Texas runs through the fee caps and the 5% reconciliation rule, not through a basis-limit ceiling the way it does in California.

Fee caps and structural limits — a sliding scale that inverts by deal size

Four caps at 10 TAC §11.302(e) bind the budget independent of the scoring item above, and one — the General Contractor fee — is genuinely unlike anything in California.

Developer Fee runs the opposite direction — small deals get a higher ceiling, large deals a lower one.

Contingency is capped at 7% of Building Cost plus Site Work plus Off-Site Construction for New Construction and Reconstruction, and 10% for Rehabilitation and Adaptive Reuse (§11.302(e)(5)). Eligible construction-period interest is capped at the lesser of actual interest or one year of a fully-drawn construction loan at the term-sheet rate — extended to 24 months for tax-exempt bond transactions (§11.302(e)(8)). Operating reserves are capped at 12 months of stabilized operating expense plus debt service, 24 months for USDA- or HUD-financed rehabilitation, with a 5-year minimum maintenance requirement documented in the partnership agreement or loan documents to count at cost certification (§11.302(e)(9)).

Why prevailing wage almost never reaches a Texas LIHTC deal

Texas Government Code Chapter 2258 is Texas's own general prevailing-wage statute, and it is narrow in a way that changes the entire shape of this phase. §2258.002(a) applies the chapter only to a public work "paid for in whole or in part from public funds," and §2258.001 defines the trigger by reference to a "public body" — the state or a political subdivision — that is itself awarding the construction contract. A typical LIHTC deal is structured with a private developer or owner entity as the party to the construction contract, not a state agency or political subdivision. That structural fact — not a funds-source exemption — is what keeps most Texas LIHTC construction outside Chapter 2258's reach even where a city or housing authority is genuinely involved through a land donation, a soft second, or a fee waiver.

The controlling case illustrates how fact-specific the boundary is. In San Antonio Building & Construction Trades Council v. City of San Antonio (Tex. App.—San Antonio, Feb. 21, 2007), a Texas appellate court held Chapter 2258 did not apply to a $307 million convention-center hotel built by a private developer under a 75-year ground lease from the City, financed with tax-exempt and taxable bonds issued through a city-created nonprofit financing corporation and secured by pledged hotel-occupancy-tax and sales-tax revenue — because the bond proceeds and the tax pledge were contingent security, not public funds actually spent to construct the building. Tax-exempt bond proceeds, a tax-revenue pledge used as backstop security, and a land lease from a city are three financing features that are routine in Texas 4% LIHTC deals, and each was individually held insufficient to make a privately built, privately operated project a public work. One honest caveat: the specific language attributed to the court here was read this session from a case-summary digest, not the full slip opinion — treat the holding as directionally reliable and confirm against the actual opinion before relying on it for a specific deal.

Where Chapter 2258 does apply, the penalty is $60 per worker per calendar day underpaid, but a municipality may only collect it if its population exceeds 10,000 — and a contractor is not in violation if the public body itself failed to determine and specify the rate.

House Bill 2127 (88th Legislature, 2023), the "Texas Regulatory Consistency Act," closes off the escape hatch — and the trap — that exists in California. It bars a municipality or county from enforcing an ordinance that conflicts with a field occupied by the Labor Code, among others, and creates a private right of action with attorney's fees for an affected business to enjoin it. A Texas developer cannot be surprised by a city-level prevailing-wage or project-labor-agreement mandate the way a California developer increasingly can — there is also no Texas counterpart to California's Public Contract Code §2601 skilled-and-trained-workforce requirement.

One edge case is genuinely unresolved: a development owned by a public housing authority or public facility corporation — relevant to RAD conversions in the At-Risk set-aside — might itself be "a political subdivision of the state" awarding its own construction contract, which could bring Chapter 2258 back into play on a project-by-project basis. Confirm ownership structure before assuming the general private-developer exemption applies.

Texas's own 2025 by-right zoning reform underlines the contrast. Senate Bill 840 (89th Legislature), signed June 20, 2025 and effective September 1, 2025, requires qualifying municipalities — population over 150,000, wholly or partly in a county over 300,000, covering Houston, Dallas, Fort Worth, Austin, San Antonio and their major suburbs — to permit multifamily development by right at a density of at least 36 units per acre and a height of at least 45 feet on sites currently zoned for office, retail, warehouse or mixed use. Unlike California's AB 2011/SB 423 wave, which ties streamlined ministerial approval to mandatory prevailing wage and, at scale, apprenticeship and healthcare-expenditure requirements, none of the reporting on SB 840 describes any labor condition attached to using it. This detail is drawn from law-firm summaries rather than the enrolled bill text itself, but the absence of a labor condition is corroborated across every summary read and fits the structural pattern above: an entitlement-pathway screener for Texas should not carry a labor-standards flag, because there is nothing to flag.

Davis-Bacon is federal, program-specific, and easy to get backwards

Where Chapter 2258 falls silent, federal Davis-Bacon can still apply — but only through specific funding programs, and the two TDHCA direct-loan sources a developer is most likely to layer under a 9% or 4% deal have opposite answers.

TDHCA's own Davis-Bacon guidance page confirms exactly four HUD Community Planning and Development programs as covered for its purposes: HOME, HOME-ARP, the Neighborhood Stabilization Program, and Section 811 Project Rental Assistance, which the page says follows HOME's labor standards. NHTF is conspicuously absent from that list, for the reason above. Where Davis-Bacon does apply, TDHCA requires a project-designated Labor Standards Officer who cannot be employed or paid by the construction contractor, a mandatory TDHCA-run preconstruction conference, and a Notice to Proceed that TDHCA will not issue until labor-standards documentation is on file, backed by a Contract Work Hours Safety Standards Act overtime penalty the page states as $29 per violation — a figure the page dates to January 16, 2022 and that adjusts periodically; confirm the current Department of Labor rate before citing a specific dollar figure.

The practical consequence for a feasibility tool: the trigger matrix cannot be "does this deal have a HUD program in it" — it has to be program-specific, because HOME and NHTF are structurally parallel TDHCA products that a developer might mentally file under the same "soft money" bucket, with opposite labor-standards consequences.

Building codes are jurisdiction-specific, and the cost benchmarks are thinner than California's

There is no statewide-current equivalent to the California Building Code. Texas Local Government Code §214.216 adopts the International Building Code "as it existed on May 1, 2012" as a municipal commercial building code floor — each municipality then decides for itself, through its own local-amendment ordinance process, whether and how to adopt a newer edition. A construction-type or story-count calculation cannot assume a single statewide code table the way California's CBC Table 504.4 supports; the applicable edition and any local amendments have to be confirmed city by city.

The county picture is narrower still. Local Government Code Chapter 233, Subchapter F reaches only "a single-family house or duplex," is opt-in even for that (a county must adopt its own resolution and either sit within 50 miles of an international border or have a population over 100), and where triggered borrows the code of the nearest city's county seat rather than setting an independent standard. For a LIHTC apartment building sited on unincorporated Texas county land outside any city's extraterritorial jurisdiction, there may be no applicable building code at all under this chapter — an open question requiring site-specific confirmation, not a settled answer either way.

Houston is the clearest illustration of a distinction worth keeping straight: the city famously has no zoning, but it is not without a building code. Multifamily siting is governed by Houston's Chapter 42 development ordinance and by private deed restrictions, while permit review still checks structural, electrical and plumbing compliance under whatever IBC edition the city has adopted under §214.216. "No zoning" and "no building code" are not the same claim.

The cost benchmarks available for Texas are thinner than California's, and it is worth saying so plainly rather than filling the gap with an estimate. The one figure that carries forward with real confidence is the cross-state comparison already verified for the California guide: California hard costs run roughly 2.3 times Texas hard costs on a per-square-foot basis. A frequently repeated companion figure — that California municipal impact fees average roughly $29,000 per unit against under $1,000 per unit in Texas — could not be independently confirmed against a primary Terner Center table and should be treated as directionally plausible rather than audit-grade.

No Texas-specific econometric study of a prevailing-wage cost premium exists, for the reason the rest of this phase makes obvious: if a Texas LIHTC deal essentially never carries a state prevailing-wage premium, there is no comparable natural experiment to study. The nearest Texas-specific figure — roughly $21,996 in tax credits per unit for recent TDHCA awards, from Kinder Institute reporting — is the tax-credit subsidy amount per unit, not total development cost per unit, and should never be substituted for one.

Modular construction fares the same as in California: no Texas-specific, credibly sourced quantitative savings study exists, only vendor marketing and general trade-press material, and the QAP contains no modular-specific cost adjustment. Because Texas's prevailing-wage exposure is federal-program-specific rather than a blanket state law, whether factory labor counts as covered work matters less by default here than in California — but matters just as much, unchanged, the moment a Davis-Bacon-triggering federal source such as HOME enters the stack, since Davis-Bacon's offsite-fabrication coverage rules are federal and do not vary by state. Do not build a Texas modular-savings feature on unverified numbers.