The exclusion for research in the social sciences, arts, or humanities within the Texas tax code dictates that any inquiry rooted in non-STEM disciplines is ineligible for state-sponsored research and development incentives. This mandate restricts the application of the research and development tax credit and sales tax exemption specifically to advancements fundamentally relying on the physical or biological sciences, engineering, or computer science.

Statutory Foundations and the Evolution of the Research and Development Incentive in Texas

The architecture of the Texas Research and Development (R&D) tax incentive program is characterized by a deliberate convergence of federal standards and state-specific economic objectives. Established to bolster the state’s competitive position in the global technology marketplace, the program offers two primary forms of relief: a franchise tax credit under Tax Code Chapter 171, Subchapter M, and a sales and use tax exemption under Tax Code Section 151.3182. These incentives are not merely administrative benefits but represent a significant fiscal policy aimed at correcting market failures associated with R&D investment. Economic theory suggests that private companies often underinvest in research because the resulting knowledge spillovers—positive externalities such as technological innovation and process improvements—benefit society more than the individual firm can capture through profit. By providing these incentives, the state of Texas effectively subsidizes the gap between private costs and social benefits, encouraging a level of R&D activity closer to a socially optimal equilibrium.

The legal definition of “qualified research” in Texas is inherently tied to the federal definition found in Section 41 of the Internal Revenue Code (IRC), specifically the version in effect on December 31, 2011. This “frozen” conformity ensures a level of stability for taxpayers, even as subsequent federal tax acts modify the national landscape. However, this conformity also imports the categorical exclusions established by the federal government to narrow the scope of the credit to “technological” pursuits. Among these, the exclusion for research in the social sciences, arts, or humanities under IRC Section 41(d)(4)(G) stands as a primary gatekeeper. This exclusion is reinforced by Texas Administrative Code (TAC) Rule 3.599, which provides the administrative framework for the franchise tax credit and explicitly lists these fields as non-qualifying.

The Dual-Track Incentive System

The Texas R&D program operates through two distinct mechanisms, and the social sciences exclusion applies with equal force to both. Taxpayers must navigate the nuances of each to maximize their benefit, though they are generally prohibited from claiming both incentives for the same qualified research expenses in a given period.

Conceptualizing the Exclusion: The Boundaries of Social Sciences, Arts, and Humanities

The exclusion of the social sciences, arts, and humanities is not merely a list of prohibited subjects but a fundamental defining characteristic of what the state considers “technological” research. Under the definitions provided by the IRS and adopted by the Texas Comptroller, the social sciences encompass disciplines focused on human society and social relationships. This includes, but is not limited to, economics, business management, behavioral sciences, sociology, and psychology. The arts and humanities exclusion is similarly broad, covering literature, history, philosophy, linguistics, and aesthetic or creative endeavors.

The Exclusion of Economic and Behavioral Research

In the context of modern business, research often straddles the line between technological development and social science. For instance, a company may invest heavily in understanding consumer behavior to optimize a new software interface. While the data collection and analysis might be rigorous and “scientific” in a general sense, it is categorized as social science research if it relies on psychological or sociological principles rather than the principles of computer science. The Comptroller’s guidance indicates that studies relating to management functions, market research, and efficiency surveys are categorically excluded. This prevents firms from claiming the credit for activities that are essentially business strategy or marketing, even when those activities involve significant investment and data analysis.

Aesthetics, Style, and the Arts

The exclusion for the arts and humanities is closely linked to the prohibition of research related to style, taste, cosmetic, or seasonal design factors. Qualified research must be undertaken for a “qualified purpose,” which is defined as a new or improved function, performance, reliability, or quality of a business component. Research that focuses on the visual appeal or cultural context of a product is deemed non-technological. For example, research into historical textile patterns or the cultural significance of certain color palettes in different geographic markets would be excluded as humanities or arts research, even if the ultimate goal is to improve a commercial product.

The Four-Part Test as a Mechanism of Exclusion

To understand how the social sciences exclusion is applied, one must examine it through the lens of the “Four-Part Test” that all qualified research must satisfy. This test acts as a filtration system, ensuring that only activities meeting specific technological and experimental criteria are rewarded.

The Section 174 Test

The first threshold is that the expenditures must be treatable as expenses under IRC Section 174. This requires that the costs be incurred in connection with the taxpayer’s trade or business and represent research and development costs in the “experimental or laboratory sense”. Crucially, the regulations under Section 174 explicitly exclude expenditures for research in connection with literary, historical, or similar projects. This provides the first statutory hook for excluding the humanities. If a project’s primary uncertainty is interpretive or narrative rather than technical, it fails at this initial stage.

The Technological in Nature Test

The second part of the test is where the social sciences exclusion is most frequently invoked. To be “qualified research,” the activity must be undertaken for the purpose of discovering information that is “technological in nature”. Information is technological in nature only if the process of experimentation used to discover it “fundamentally relies on principles of the physical or biological sciences, engineering, or computer science”.

This requirement creates a clear divide between “hard” and “soft” sciences. Research that relies on principles of economics (social science) to determine the pricing elasticity of a new product is not technological, even if it uses sophisticated mathematical models. Conversely, research that relies on physics (physical science) to test the structural integrity of a new composite material is technological. The “Technological in Nature” test ensures that the state’s fiscal support is directed toward activities that expand the technical capabilities of a firm rather than its market insights or organizational behavior.

The Business Component Test

Research must be intended for use in developing a new or improved “business component,” which includes products, processes, computer software, techniques, formulas, or inventions held for sale, lease, license, or used in the taxpayer’s trade or business. Social science research often results in reports, studies, or strategic recommendations rather than functional business components. While a “process” can be a business component, it must be a technological process. An improved hiring process based on behavioral science remains a social science activity and thus fails to qualify as a technological business component.

The Process of Experimentation Test

The final requirement is that substantially all of the activities (generally interpreted as 80% or more) must constitute elements of a process of experimentation. This process must be designed to evaluate one or more alternatives to achieve a result where the capability, method, or appropriate design is uncertain at the beginning of the research.

The Comptroller and the IRS clarify that a process of experimentation involves:

1. Identifying uncertainty concerning the development or improvement of a business component.

2. Identifying one or more alternatives intended to eliminate that uncertainty.

3. Identifying and conducting a process of evaluating those alternatives (e.g., through modeling, simulation, or systematic trial and error).

Social science research frequently involves “testing” and “evaluation,” such as focus groups or A/B testing of marketing copy. However, because these evaluations do not fundamentally rely on physical or biological sciences or computer science to resolve the uncertainty, they are not considered a “process of experimentation” in the laboratory sense required by the statute.

Administrative Guidance from the Texas Revenue Office

The Texas Comptroller of Public Accounts serves as the state’s primary revenue office, providing critical guidance on the application of the R&D credit through administrative rules, letter rulings, and the State Tax Automated Research (STAR) system.

Interpretation of TAC Rule 3.599

TAC Rule 3.599 provides the most detailed local guidance on the franchise tax credit. It clarifies that “qualified services” performed by an employee include the actual conduct of research or its direct supervision and support. However, the rule explicitly excludes general administrative services and other services that only indirectly benefit research. For example, the wages of an accountant tracking R&D expenses or payroll personnel preparing checks for scientists do not qualify, as their work is rooted in business administration (a social science) rather than the conduct of the research itself.

The Role of the STAR System and Hearings

The STAR system contains a repository of Comptroller decisions that illustrate how the social science exclusion is applied in practice. These documents often reveal a strict adherence to the technological requirement. In various hearings, the Comptroller has denied claims where software development was found to be primarily for “internal use” or focused on “management functions” rather than resolving a computer science uncertainty.

A recurring theme in STAR decisions is the burden of proof. Taxpayers are required to establish their eligibility for the credit with “clear and convincing evidence”. This means that the taxpayer must provide contemporaneous records—generated at the time the research was conducted—proving that the work met all parts of the four-part test. If the records suggest that the primary objective was to improve user engagement or understand market trends, the Comptroller will likely reclassify the work as excluded social science research.

Refund Procedures and Carryforwards

For entities seeking to claim the credit retroactively or manage unused portions, the Comptroller provides specific procedural guidance. A refund claim must be submitted in writing, identifying the period of overpayment and stating all grounds for the claim. Unused credits may be carried forward for up to 20 years. However, a taxpayer cannot “create” a credit for a tax year that is closed by the statute of limitations. While the Comptroller may verify past expenses to confirm a carryforward amount for an open year, they cannot adjust the tax liability for a closed year.

The Internal Use Software (IUS) Boundary

The development of internal use software represents a significant area of conflict between taxpayers and the revenue office. Software is considered “internal use” if it is developed for the benefit of the taxable entity primarily for its own internal operations—such as financial reporting, human resources, or inventory management.

Under the 2011 IRC standards adopted by Texas, IUS must meet a higher “three-part test” in addition to the standard four-part test to qualify:

1. Innovativeness: The software must be intended to be unique or novel and involve significant cost reductions or speed improvements.

2. Significant Economic Risk: The taxpayer must commit substantial resources with uncertainty that those resources will be recovered.

3. Commercially Unavailable: The software must not be available for purchase, lease, or license in the commercial market.

The “Innovativeness” requirement often overlaps with the social science exclusion. If the “innovation” is merely a more efficient way to manage business operations (management science), the Comptroller may argue that the project is not “technological” in nature. For a software project to transcend the social science exclusion, the developer must prove they were solving a fundamental computer science problem—such as improving data structure efficiency or overcoming a hardware limitation—rather than just automating a business process.

Legislative Transformation: From HB 800 to SB 2206

The regulatory environment for the Texas R&D credit has undergone major shifts, most notably with the passage of Senate Bill (SB) 2206 in 2025, which takes effect on January 1, 2026. This legislation marks a move toward “rolling conformity” with the federal tax code and aims to simplify the compliance burden for Texas businesses.

The Previous Framework (2014-2025)

Prior to SB 2206, the credit was based on a 5% rate and relied on a complex determination of qualified research expenses that often led to disputes with the Comptroller over what constituted a “qualified expense”. Controversy particularly surrounded 2021 regulations (TAC 3.599) that limited the credit for internal use software and were applied retroactively.

The SB 2206 Paradigm (Post-2026)

The new legislation fundamentally changes the calculation and application of the credit:

1. Direct Alignment with Federal Form 6765: The new Texas credit is calculated based on the total qualified research expenses reported on line 48 of federal Form 6765, to the extent those expenses are attributable to Texas. By “tying” the state credit to a federal form, the legislature aims to eliminate the need for a separate factual determination of “qualified research” at the state level.

2. Increased Credit Amount: The credit rate increases to 8.722% of the difference between current-year QREs and 50% of the three-year average.

3. Refundability: Taxable entities that do not owe franchise tax can now receive the credit as a refund, whereas previously, it could only be used to offset tax liability or carried forward.

4. Preservation of Carryforwards: Unused portions of the credit from the previous statute can be carried forward for up to 20 years.

Despite this alignment, the social sciences exclusion remains a core component of the federal definition (IRC § 41(d)(4)(G)). Therefore, while the administrative burden of claiming the credit may decrease, the substantive barrier against social science and humanities research remains just as high, as those activities are excluded from the very federal line (Line 48) on which the Texas credit is now based.

Comparative Analysis: Franchise Tax Credit vs. Sales Tax Exemption

A taxpayer engaged in qualified research must choose between the franchise tax credit and the sales tax exemption for a given period. This choice is often driven by the nature of the research expenditures.

For a company whose primary R&D cost is employee salaries (e.g., a software developer), the franchise tax credit is typically more valuable. For a capital-intensive research firm (e.g., a biotech lab requiring expensive microscopes), the sales tax exemption on equipment may be more beneficial. In either case, if the equipment or the personnel are dedicated to a social science project, neither incentive will be available.

Comprehensive Example: The “EduTech” Development Lifecycle

To illustrate the application of these rules, consider a Texas company, “KnowledgeFlow Systems,” which is developing a new AI-driven adaptive learning platform for K-12 students. The project involves multiple phases, some of which are qualified and some of which fall under the social science exclusion.

Phase 1: Algorithmic Development (Qualified)

KnowledgeFlow hires computer scientists to build a proprietary machine learning model that predicts student learning gaps. The team must overcome significant technical challenges regarding data processing speed and the accuracy of the neural network. They test various mathematical models to find the most efficient way to process millions of data points in real-time.

– Analysis: This phase qualifies as research. It is technological in nature (computer science), involves a process of experimentation to resolve a technical uncertainty, and is intended to develop a functional business component.

Phase 2: Pedagogical Efficacy Study (Excluded Social Science)

Once the algorithm is built, the company hires educational psychologists to conduct a year-long study in a local school district. The goal is to determine if the platform actually improves student test scores and to understand how different classroom settings affect student motivation. They use surveys, standardized tests, and behavioral observations.

– Analysis: This phase is excluded. Although it is “research” in a general sense, its fundamental principles rely on behavioral sciences and pedagogy (social sciences). It does not resolve a technological uncertainty about the software itself, but rather a behavioral uncertainty about how students learn.

Phase 3: User Interface Aesthetic Redesign (Excluded Arts/Humanities)

The company hires a graphic designer to refresh the app’s look, choosing “soothing” colors and “fun” animations to make the platform more appealing to young children.

– Analysis: This phase is excluded. It relates to style, taste, and cosmetic design, which falls under the arts/humanities exclusion.

Phase 4: Integration Architecture (Qualified)

The company needs to integrate its platform with existing school district databases. They encounter a technical problem because the databases use legacy protocols that do not easily communicate with the new AI model. They spend several months developing a new “wrapper” and translation layer to ensure data integrity and low latency.

– Analysis: This phase qualifies. Developing new platforms or integrations and tackling back-end performance issues are cited as qualifying activities, provided they meet the four-part test.

Compliance and the “Clear and Convincing” Burden

Given the high bar for qualification and the strictness of the exclusions, the Texas revenue office places an immense premium on documentation. The Comptroller’s guidance suggests that the following practices are essential for a successful claim:

Contemporaneous Record-Keeping

Documents must be generated as the R&D activity occurs, not reconstructed years later during an audit. This includes project plans, test results, laboratory notes, and code repositories. If a company cannot produce records showing that a software project was fundamentally about computer science rather than management, they will likely lose a challenge under the social science exclusion.

Personnel Allocation

Taxpayers must carefully track which employees are working on which projects. For the franchise tax credit, only “qualified services” count. If a lead researcher spends 50% of their time on Phase 1 (Qualified) and 50% of their time on Phase 2 (Excluded Social Science), only half of their wages can be included in the credit calculation.

Annual Reporting Requirements

Entities claiming the sales tax exemption must register with the Comptroller to receive a Texas Qualified Research Registration Number. They are then required to file an Annual Information Report (AIR) by March 31 of each year. Failure to file the AIR can result in the cancellation of the registration and the loss of the exemption for subsequent purchases.

Final Thoughts: Strategic Implications for Texas Taxpayers

The research exclusion for the social sciences, arts, and humanities represents a deliberate policy boundary in the Texas tax code, separating “innovation” from “insight.” For businesses, this requires a nuanced understanding of their own development cycles. Projects that focus on the “how” of technology—the engineering, the physics, and the computer science—are the intended beneficiaries of these state incentives. Conversely, projects that focus on the “why” of human behavior—the economics, the psychology, and the aesthetics—are intentionally left out of the R&D credit framework.

The shift towards federal alignment via SB 2206 provides a clearer path for compliance but does not lower the technological threshold. By tying the state credit to federal reporting, Texas has outsourced the “qualified research” determination to a more established set of federal rules, but those rules are equally hostile to social science research. Taxpayers must therefore continue to rigorously distinguish between their technical experimentation and their social or aesthetic studies, ensuring that every dollar claimed as a credit is backed by clear and convincing evidence of a technological endeavor. In the competitive landscape of the Texas economy, the ability to navigate these exclusions is as critical as the innovation itself.