What a Building Code Study Gets Wrong About Housing Affordability

A working paper from researchers at the Center for Risk and Insurance Research (CRIR) at the University of Alabama concludes that when a jurisdiction adopts new residential building codes, it has no effect on the sales prices of homes. But a close technical review reveals that the study’s policy conclusions go well beyond what its analysis can establish.

The study, funded by the International Code Council (ICC) and the insurance industry, does acknowledge that newer codes can add thousands of dollars to the cost of constructing a home. But the study implies that the construction industry absorbs additional costs.

The CRIR working paper is now being cited as evidence that the adoption of newer, more stringent codes does not increase home prices or constrain housing production, and that concerns about code-related costs are overstated.

Key Issues with the Working Paper

The study’s research design, however, cannot reliably isolate the effect of building codes on home prices for several reasons.  First, the analysis lacks direct data identifying whether each home in the sample was built under the old or new code. At the same time, the statistical controls do not adequately distinguish code impacts from differences associated with annual inflation, a home’s age, construction vintage, or whether it is newly built or a resale. As a result, the analysis cannot determine whether price increases following the adoption of new codes are attributable to the code changes or to other factors.

The study also claims that updated codes have no impact on housing supply, but such assertions are even less supported. The researchers do not directly analyze building permits, housing starts, or total housing supply, yet draw broad but empirically unsupported conclusions about whether modern codes constrain construction. The paper’s conclusions also extend beyond the versions of the code it examines. It examines amended codes in several states but draws broad conclusions about the effects of adopting unamended model codes.

These are not minor technical shortcomings; they make the study’s central policy claims unsubstantiated. Added construction costs from updated codes can make a home financially unviable to build, put it beyond what families can afford, and ultimately reduce the availability of attainable housing. An analysis that cannot reliably isolate code-related effects and does not directly measure these broader outcomes cannot be used to dismiss these potential consequences or justify additional code requirements without fully accounting for their costs.

Paul Emrath, PhD, from NAHB Economics, was invited to review the draft report and found that “the statistical model used for the analysis is poorly conceived, misapplied, and fails to meet the rigorous standards necessary to prove the broad, counter-intuitive, and far-reaching policy conclusion that added regulatory costs do not matter.” He added that “although NAHB provided feedback on the shortcomings of the paper, it is concerning that the authors chose to publish their research as a working paper without properly addressing the data and methodological flaws of the study and suppressing all descriptive statistics and regression results.”

It is essential that policymakers have evidence that supports the claims being made. This study does not provide it.

What the Study Doesn’t Capture

There are significant and important limitations to what the CRIR working paper can tell us about the real-world cost of added building code requirements due to its many flaws. Among them:

• The working paper does not reliably match homes to the building codes they were built to. To measure the effect of a code adoption, the analysis first needs to know which homes were subject to the old code and which were subject to the new code. The study does not have this information and instead uses the year a home was built as a proxy for its permitting date. Building codes are generally applied when a home is permitted, not when a home is completed. As a result of the proxy, the working paper misclassifies homes permitted before the code enactment but completed afterwards as new-code homes. This is incorrect. The study’s alternative specification, which arbitrarily shifts the adoption date by one year, does not solve the problem: changing the date does not establish which code actually governed the construction of an individual home. Transition periods, local implementation practices, and projects already in the permitting process further complicate the timing. Although the paper acknowledges that a home’s “year built” may reflect its initial permit, final inspection, or a point in between, this admission does not fix the underlying problem.

• The working paper cannot adequately isolate the effect of the building codes on home prices. Using the year a home was built to infer its exposure to a new code also creates an empirical identification problem because construction year is also embedded in the model through the interaction of home age and the year in which the home was sold. As a result, the model does not adequately distinguish the effects of building codes from broader changes in home prices over time, including inflation and other market conditions. Misleadingly, the paper attributes the persistent home price increases that follow the new code adoption to “year-specific” effects (including general inflation).

• The study’s broad conclusions do not account for how code-related costs may affect homes at different price points, particularly entry-level new homes. Examining these differences is important because there is less room to absorb added costs while keeping homes attainable. In her review of the CRIR working paper, Leah Brooks, PhD, of George Washington University’s Trachtenberg School identifies this concern directly: “In addition, at the low value end of the market, it is possible that the additional costs in new codes may make some homes unprofitable to build, and will lead to the construction of fewer new lower value homes.” Without examining effects at different price points, the study may miss how added code costs affect the affordability and availability of entry-level homes. If those costs put a new home out of reach, a family may be pushed toward older housing and lose access to some of the energy-efficiency and resilience benefits the study itself cites as reasons for adopting modern codes. Homes that become financially unviable to build will never appear in a database of home sales.

• The working paper misses the distinction between codes adopted by jurisdictions and model codes. States and local governments frequently amend model codes, including costly fire sprinkler and energy-efficiency provisions. As Brooks explains, “Because codes are not usually adopted in their entirety, results do not tell us about the correlation between home prices and the entire new code.” Findings based on amended codes therefore cannot support broad claims that adopting the full, unamended model codes would have no impact on home prices.

• The working paper doesn’t analyze the codes at the center of today’s affordability debate. The study examines the 2009 through 2018 editions of the International Residential Code but fails to consider the International Energy Conservation Code (IECC). That matters because NAHB estimates that compliance with the 2021 IECC can add between $9,600 and $21,400 to the cost of a new single-family home, with some costs taking up to 90 years to recoup through energy savings.

• The study fails to consider other implementation costs. The cost of adopting a new building code is not limited to the materials or features required by the code itself. Each new code can require changes in design, training, purchasing, construction practices, inspections and enforcement, all of which take time and resources to implement. Those real-world costs are another part of the affordability equation that the paper does not attempt to capture.

• The working paper does not properly evaluate code effects on housing supply. The paper relies on new home sales to establish whether new codes constrain home building. However, home sales are not a reliable measure of construction activity. The timing of sales, demand conditions, and data coverage (e.g., the sales data do not capture custom or contractor-built homes constructed on owners’ land) affect how well sales data reflect new home construction. Without an analysis of permitting activity or housing starts, the study cannot reliably establish to what extent updated codes constrain construction. Moreover, the isolated North Carolina–South Carolina border exercise presented in the paper does not provide a sufficient basis for the broad conclusions the paper reaches about housing supply nationwide.

• The working paper does not establish the causal effect of building codes on home prices. Although the paper acknowledges that its research design cannot identify causal effects, it nevertheless draws substantially stronger conclusions about the effects of building code adoption on home prices that contradict this limitation. As Brooks explains, “Because the authors estimate a correlation, the authors cannot conclude that adopting a new building code causes home prices to change.” Accordingly, claims that the paper demonstrates that building codes decrease home prices, increase them, or have no impact on them are unsupported by its findings.

The Responsibility to Address Affordability Cannot Be Dismissed

Building codes provide important health, safety, efficiency, and resilience benefits. Those goals can and should be achieved without losing sight of the need to keep homes affordable. New requirements deserve careful evaluation of what they will cost, what benefits they will provide, and whether there is a more affordable way to achieve the same goals.

There is no single policy change that will solve the nation’s housing affordability crisis. Progress requires addressing the many factors that make homes more expensive to build, including the cost of building codes. The working paper does not relieve policymakers of that responsibility. They cannot afford to ignore the real impact of building code costs on whether homes are built and whether families can afford them.