Industry Case Studies and Application of R&D Tax Law

The practical application of the United States federal and Michigan state Research and Development (R&D) tax credit laws requires a nuanced understanding of specific industry operations, historical context, and technical challenges. The city of Ann Arbor, Michigan, presents a unique economic geography. Long recognized as an academic powerhouse due to the presence of the University of Michigan, the region has cultivated a highly diversified technology and research ecosystem. To illustrate the complexities of statutory tax compliance, the following five case studies examine specific industries that have developed in Ann Arbor, detailing their historical roots and analyzing hypothetical scenarios against prevailing tax administration guidance and legal precedent.

Case Study 1: Advanced Automotive and Autonomous Mobility Systems

The Ann Arbor region has systematically transformed itself from a peripheral beneficiary of Detroit’s traditional automotive manufacturing into a global epicenter for advanced mobility and autonomous vehicle (AV) research. This regional evolution is anchored by the University of Michigan’s massive research expenditures, which exceed $1.77 billion annually, and the presence of major original equipment manufacturer (OEM) research centers, such as the Toyota Motor North America R&D Headquarters. The most critical catalyst for this specific industry’s localized development was the establishment of the Ann Arbor Connected Vehicle Test Environment and the Mcity Test Facility. Encompassing twenty-seven square miles, this infrastructure represents the world’s largest operational real-world deployment of connected vehicles and smart intersections, funded in part by the Federal Highway Administration. This unparalleled testing environment naturally attracted a dense cluster of over two hundred mobility technology companies and startups focusing on sensor fusion, pedestrian detection, and autonomous logistics.

Consider the scenario of a hypothetical Ann Arbor startup, Arbor-Autonomy Dynamics, which operates on the periphery of the Mcity testing grounds. The company is engaged in the development of a proprietary, multi-modal sensor fusion system that integrates Light Detection and Ranging (LiDAR), radar, and high-definition optical cameras. The core objective is to create a predictive algorithm capable of anticipating the erratic movements of pedestrians in severe winter weather conditions, a recognized vulnerability in Level 4 autonomous robotic taxis. During the development phase, the engineering team designs and fabricates multiple custom pilot models of the sensor housing using advanced, 3D-printed thermoplastic polymers to manage the extreme thermal output of the onboard processors. These prototype housings are subjected to rigorous testing on closed tracks; upon thermal failure, they are discarded and redesigned. Ultimately, the finalized, custom-fitted prototype units are sold to a regional OEM for integration into a beta test fleet.

Under United States federal tax law, specifically Internal Revenue Code (IRC) Section 41, Arbor-Autonomy Dynamics must prove that their activities satisfy the statutory four-part test. The development of predictive algorithms inherently involves resolving high levels of technical uncertainty through computational modeling, thereby satisfying the requirement that the research be technological in nature. Furthermore, the construction and evaluation of the sensor housings constitute a structured process of experimentation. The legal precedent established by the United States Tax Court in Intermountain Electronics, Inc. v. Commissioner (2024) dictates that expenses incurred in the development of a “pilot model” qualify as an eligible process of experimentation if the model is utilized to evaluate and resolve uncertainties regarding the product’s fundamental design and performance.

A critical tax controversy often arises regarding the material costs of prototypes that are eventually sold to customers. The Internal Revenue Service (IRS) frequently challenges these costs, arguing they represent depreciable assets rather than consumable supplies. However, Arbor-Autonomy Dynamics can rely on the precedent set in TG Missouri Corp. v. Commissioner (2009). In this case, the Tax Court ruled that production molds purchased from third-party toolmakers and later sold to automotive customers were not “assets of a character subject to depreciation” in the hands of the taxpayer, primarily because the taxpayer did not bear the economic risk of exhaustion over time. Because Arbor-Autonomy Dynamics bore the economic risk during the iterative development phase, and the prototypes were heavily modified and consumed during testing before the final sale, the $150,000 spent on specialized polymers would qualify as supply Qualified Research Expenses (QREs) under federal law.

For state tax purposes, Arbor-Autonomy Dynamics is highly positioned to benefit from the newly enacted Michigan R&D tax credit. Because the engineering wages and the physical testing occurred entirely within Washtenaw County, the expenses satisfy the strict statutory requirement that the research be “incurred in this state”. Assuming the firm employs fifty individuals, they are classified as a small business under Public Act 186 of 2024. Consequently, they calculate their tentative claim utilizing the accelerated 15% tier for qualifying expenses that exceed their historical base amount, subject to the $250,000 maximum annual credit limit.

Case Study 2: Biopharmaceutical Therapeutics and Protein Engineering

Ann Arbor possesses a rich, multi-generational legacy in the life sciences and biopharmaceutical sectors. This history traces back to 1866 with the founding of Parke-Davis in nearby Detroit, a company that pioneered the standardized, systematic method of clinical drug testing and constructed the first dedicated pharmaceutical research laboratory in the United States in 1902. Throughout the twentieth century, Parke-Davis (and later Warner-Lambert) maintained massive research operations in Ann Arbor, developing historic therapeutics including the polio vaccine (with Dr. Jonas Salk) and the cholesterol-lowering medication Lipitor. In 2000, Pfizer acquired Warner-Lambert and subsequently closed the Ann Arbor research facilities in 2007. Far from destroying the local industry, this closure dispersed thousands of highly skilled bioscientists into the local economy, seeding a robust startup ecosystem. Today, the region supports over three hundred life science firms, bolstered by extensive incubator infrastructure such as the Michigan Innovation Headquarters (MI-HQ), which provides essential wet lab space for emerging biotech ventures.

A representative firm in this modern ecosystem is Huron BioSciences, a mid-sized enterprise operating within the MI-HQ facility. The company is dedicated to developing a novel recombinant protein therapy targeted at a rare autoimmune disorder. The R&D team utilizes advanced CRISPR/Cas methodologies to evaluate new guide ribonucleic acid (RNA) designs. The project faces immense biological uncertainty regarding the structural stability of the protein expression vectors and the overall yield optimization. To execute the clinical trials, Huron BioSciences employs a dedicated internal team of geneticists while simultaneously entering into a contract with a specialized clinical research organization (CRO) based in Detroit to conduct extensive in-vitro toxicity assays.

The federal R&D tax credit framework is uniquely accommodating to the biopharmaceutical industry. The fundamental requirement that the research be “Technological in Nature” is inherently satisfied, as the entire enterprise is grounded in the hard sciences of biology and chemistry. The iterative processes of directed evolution, mutagenesis, and the adjustment of bioreactor parameters (temperature, pH levels, and feed strategies) represent a textbook execution of a process of experimentation designed to evaluate alternatives and eliminate technical uncertainty.

The inclusion of the Detroit-based CRO introduces the complexities of Contract Research Expenses (CRE). Under IRC Section 41(b)(3), a taxpayer is generally permitted to claim 65% of the amounts paid to third-party entities performing qualified research on their behalf. However, to defend this claim under rigorous IRS audit scrutiny, Huron BioSciences must navigate the “funded research” exclusion. The legal agreements with the CRO must unequivocally demonstrate that Huron BioSciences bears the economic risk of the research—meaning they are contractually obligated to pay the CRO regardless of whether the protein stabilizes or the toxicity assays yield favorable results. Furthermore, Huron BioSciences must retain substantial rights to the underlying intellectual property and the resulting RNA designs generated by the CRO.

Under the Michigan R&D tax credit statutes, Huron BioSciences benefits significantly from its decision to utilize an in-state CRO. Because both the internal scientific wages and the contracted CRO activities occurred within the state of Michigan, both expenditure categories feed directly into the Michigan unadjusted credit calculation. Assuming Huron BioSciences has grown to employ three hundred individuals, they are categorized as a large taxpayer under the state framework. If their overall QREs amount to $5 million above their historical base amount, they would calculate their state credit at 10% of that excess, positioning them to claim a substantial $500,000 credit, well within the $2,000,000 statutory cap for large businesses.

Case Study 3: Cybersecurity and Enterprise Software Development

Ann Arbor’s prominence as a national hub for cybersecurity and enterprise software is a direct architectural byproduct of its role in the foundation of the internet. In 1966, the Michigan Educational Research Information Triad (MERIT) was established by state universities to investigate resource sharing by connecting mainframe computers. In 1987, the Merit Network, operating out of Ann Arbor, secured a $39 million federal award to manage and upgrade the National Science Foundation Network (NSFNET) backbone, which served as the critical catalyst for the commercial internet. This monumental technological achievement anchored a deep talent pool in computer science, networking protocols, and cryptography within the region. As the internet commercialized, the imperative to secure these open frameworks created massive demand for cybersecurity solutions. This environment incubated spectacular commercial successes, most notably Duo Security, a firm founded in 2009 by University of Michigan alumni that broke state venture capital records before being acquired by Cisco Systems for $2.35 billion in 2018.

In this environment, consider Liberty Network Defense, an Ann Arbor-based software engineering firm developing a proprietary, artificial intelligence-driven anomaly detection algorithm. This software is engineered for commercial sale to financial institutions to preempt zero-day ransomware attacks. Simultaneously, to manage its rapidly expanding workforce, Liberty’s internal IT department is developing a separate, highly customized software system to automate their own human resources and multi-state payroll accounting, as existing commercial off-the-shelf software cannot accommodate their unique, non-standard compensation structures.

Federal R&D tax law mandates a bifurcated analysis for software development, imposing significantly different evidentiary burdens depending on the software’s intended use. The commercial threat detection software is evaluated under the standard four-part test. To qualify the wages of the software engineers, Liberty must demonstrate that the coding process involved the systematic evaluation of alternatives—such as testing different neural network architectures or routing protocols—to resolve objective uncertainty regarding the algorithm’s latency and threat-recognition accuracy. Liberty cannot merely rely on the assertion that they followed a standard software development life cycle (SDLC). As established in the 2024 Tax Court decision Phoenix Design Group, Inc. v. Commissioner, the court explicitly rejected the argument that following a multi-stage design or development process automatically constitutes a qualifying process of experimentation. Liberty must maintain contemporaneous documentation identifying the specific technical failures encountered and the subsequent code iterations implemented to overcome them.

Conversely, the HR and payroll software is classified as Internal Use Software (IUS) because it is developed solely to support Liberty’s general administrative functions. Under Treasury Regulation § 1.41-4(c)(6), IUS must satisfy the standard four-part test plus an additional, rigorous three-part “High Threshold of Innovation” test. Liberty must prove that the internal software is highly innovative (resulting in a substantial reduction of cost or improvement in operational speed), that its development involves significant economic risk due to technical uncertainty, and that the software is not commercially available for use without extensive modifications.

From a tax planning perspective, the enactment of the federal One Big Beautiful Bill Act (OBBBA) of 2025 radically altered the treatment of these expenses. The OBBBA restored immediate expensing for domestic research and experimental (R&E) expenditures under IRC Section 174A, permanently affirming that software development costs are classified as qualifying R&E costs. Consequently, Liberty can immediately deduct the millions spent on software engineer salaries on their federal return, significantly reducing federal taxable income. However, the State of Michigan enacted legislation in late 2025 specifically decoupling the state tax code from this federal provision. For Michigan Corporate Income Tax purposes, Liberty must add back these immediate federal deductions and amortize the software development costs over a restrictive five-year period. This decoupling underscores the vital strategic importance of capturing the Michigan R&D tax credit, as it serves as the primary mechanism to offset the increased state tax liability generated by the amortization requirement.

Case Study 4: Medical Device Manufacturing and Surgical Instrumentation

The medical device and surgical instrumentation sector in Ann Arbor is inexorably linked to the clinical demands and innovative environment of the University of Michigan Hospital. Established in 1869 as the first university-owned and operated hospital in the United States, the institution has fostered a culture where engineering and clinical practice intersect. This proximity to world-class clinical environments allows biomedical engineers to observe surgical procedures firsthand, identifying physical inefficiencies that spawn device innovation. Historical breakthroughs in the broader region include the oscillating electric bone saw developed by Dr. Homer Stryker, and the vacuum-assisted closure device for wound management. Today, initiatives such as the U-M Medical Device Sandbox facilitate interprofessional collaboration, where biomedical engineering students and attending physicians simulate surgical scenarios to identify design flaws, use errors, and improve patient safety through iterative physical prototyping. Commercial entities, such as Micro-Tech Endoscopy, have established major R&D headquarters in Ann Arbor specifically to leverage these localized physician relationships for product development.

Washtenaw Surgical Innovations (WSI) is a specialized engineering firm operating in this sector, currently designing a next-generation, articulating laparoscopic grasper. The design mandate requires a novel mechanical linkage system capable of operating within the restrictive confines of a 5-millimeter surgical trocar while maintaining immense, precisely calibrated grip strength to manipulate delicate tissues. The firm’s Chief Operating Officer (COO), a mechanical engineer by trade, spends 40% of his time directly supervising the engineering team, reviewing complex computer-aided design (CAD) models, and evaluating metallurgical alternatives for the linkage pins. The remaining 60% of his time is dedicated to general administrative duties, investor relations, and sales. Furthermore, WSI is developing this grasper under a specific contract for a large, multinational medical distributor.

When calculating the federal R&D tax credit, the mechanical engineering activities undoubtedly qualify, as they seek to eliminate uncertainties regarding the physical capabilities and material strength of the new surgical device. However, WSI faces a severe audit risk regarding the inclusion of the COO’s wages in the QRE calculation. According to the stringent precedent established by the Tax Court in Scott and Gayla Moore v. Commissioner, the IRS will aggressively disallow wage QREs for executive officers if the taxpayer fails to provide task-specific documentation proving direct supervision or direct support of qualified research. The court explicitly ruled that generalized payroll records indicating total hours worked are entirely insufficient. To survive an audit, WSI must maintain rigorous, contemporaneous documentation—such as project-specific timesheets, design review meeting minutes, and technical correspondence—to substantiate the exact 40% allocation of the COO’s salary to the R&D claim.

Furthermore, because WSI is an engineering firm performing work for a third party, they must navigate the treacherous waters of the “funded research” exclusion under IRC Section 41(d)(4)(H). If the contract with the multinational distributor guarantees payment to WSI for their engineering hours regardless of the grasper’s ultimate functionality, the IRS will deem the research “funded,” entirely disqualifying WSI from claiming the credit. The recent affirmation by the U.S. Court of Appeals for the Eighth Circuit in Meyer, Borgman & Johnson, Inc. v. Commissioner (2024) emphasized that taxpayers must clearly demonstrate that payment is contingent upon the successful technical development of the design, thereby transferring the economic risk of failure to the taxpayer. To capture the credit, WSI’s legal counsel must ensure the development agreement is structured as a fixed-price performance contract rather than a time-and-materials engagement.

Assuming WSI successfully structures the contract and substantiates the wages, the documented engineering salaries—including the substantiated portion of the COO’s compensation—will qualify for both the federal and the newly enacted Michigan state credits. WSI must ensure their tentative claim is filed with the Michigan Department of Treasury utilizing these actual, verified expense metrics before the statutory deadline to secure their pro-rata share of the state’s $100 million aggregate cap.

Case Study 5: Agricultural Bioinformatics and Environmental Sustainability

While the broader Michigan and Midwestern economies rely heavily on traditional agriculture and agribusiness, Ann Arbor has successfully merged its high-technology capabilities with this agricultural heritage. The University of Michigan possesses a long, documented history of environmental activism and research, dating back to responses to the catastrophic Midwest fires of the 1870s and the landmark environmental protests against water pollution in March 1970. Today, multi-disciplinary, action-based partnerships—often channeled through academic initiatives like U-M’s Business+Impact center and the School for Environment and Sustainability—merge data science with pressing ecological and agricultural challenges. This unique intersection has birthed the localized field of agricultural bioinformatics, where advanced computational models are deployed to optimize crop yields, monitor water systems via real-time DNA sequencing, and ensure sustainable land use.

Huron Valley Agri-Tech is a quintessential startup operating in this space. The company is developing a predictive machine-learning model to optimize the breeding methods and significantly increase the yield of whole teff grain, a highly resilient crop, within Michigan’s unique climate profile. Recognizing the immense complexity and sheer volume of soil microbiome data required to train the model, the startup enters into a formalized, written collaborative research agreement with the University of Michigan’s bioinformatics department. Together, the corporate engineers and university researchers conduct massive computational modeling to correlate the genomic sequences of the teff grain with variable soil acidity levels and precipitation patterns.

From a federal tax compliance standpoint, the project represents a complex amalgamation of computer science and the biological sciences. If the IRS were to scrutinize the claim, examiners would likely apply the “shrink-back rule” detailed in the IRS Audit Techniques Guide (ATG). If the broader agricultural optimization project is deemed too general to meet the strict four-part test, the shrink-back rule dictates that the test be applied to the next most significant subset of elements. In this scenario, the specific, proprietary machine-learning algorithm generating the microbiome models would be isolated and evaluated as the distinct, qualifying business component. The iterative process of tuning the algorithm’s hyperparameters to improve predictive accuracy would solidly establish the required process of experimentation.

At the state level, this scenario perfectly aligns with the specific legislative intent of the Michigan R&D credit’s university collaboration provisions. To reposition the state as a leader in innovation, Public Acts 186 and 187 of 2024 included a highly lucrative bonus structure to incentivize exactly this type of academic-industrial partnership. Because Huron Valley Agri-Tech incurred QREs pursuant to a written agreement with a recognized Michigan research university, they are entitled to an additional 5% tax credit strictly on those collaborative expenses. If the firm spent $1.5 million on the university collaboration, they would receive a $75,000 bonus credit. This bonus is calculated completely separately from their standard, size-based tier calculation, provided the bonus alone does not exceed the absolute statutory cap of $200,000 per taxpayer per year. This mechanism not only reduces the tax burden on the startup but directly subsidizes the retention of academic research talent within the state.

Comprehensive Analysis of United States Federal R&D Tax Law

The United States federal government employs the tax code as a primary macroeconomic instrument to incentivize domestic innovation, technological advancement, and the retention of high-skill employment. The framework governing these incentives is exceedingly complex, reliant on specific statutory definitions, evolving Treasury regulations, and a vast body of judicial interpretation. The primary mechanisms are codified in Internal Revenue Code (IRC) Section 41, which governs the Credit for Increasing Research Activities, and IRC Section 174, which governs the amortization and deduction of Research and Experimental (R&E) expenditures.

The Architecture of IRC Section 41: The Four-Part Test

To secure the federal R&D tax credit, a taxpayer’s activities must strictly adhere to the statutory parameters of IRC Section 41. The IRS evaluates eligibility at a granular level, requiring that activities be tied to a specific “business component” rather than a generalized project or departmental initiative. A business component is legally defined as any product, process, computer software, technique, formula, or invention that is to be held for sale, lease, license, or used in a trade or business of the taxpayer. Once the business component is identified, the activities related to its development must satisfy all four criteria of the rigorous four-part test.

Classification and Calculation of Qualified Research Expenses (QREs)

If a technical endeavor successfully navigates the four-part test, the financial expenditures directly allocable to those specific activities may be aggregated as Qualified Research Expenses (QREs). The precise categorization and calculation of these expenses form the basis of the ultimate tax credit. Federal law recognizes three primary classifications of QREs:

• Qualified Wages: This encompasses the W-2 taxable compensation paid to employees for engaging in qualified services. Crucially, the statute defines qualified services as not only the direct performance of research (e.g., the software engineer writing code or the chemist running assays) but also the direct supervision of the research (e.g., the laboratory director reviewing experimental protocols) and the direct support of the research (e.g., the machinist fabricating a prototype part for the engineer). General administrative, executive, or human resources wages are excluded.
• Qualified Supplies: This category includes tangible property consumed, utilized, or destroyed during the experimental process. Examples include laboratory reagents, raw materials for prototypes, and specialized polymers. The statute strictly prohibits the inclusion of land, any property of a character subject to depreciation (as heavily litigated in TG Missouri), and general administrative supplies (e.g., office paper, standard computers). Notably, the rental costs of computers, specifically off-site or cloud-based server space utilized exclusively for software development and testing environments, are permissible QREs.
• Contract Research Expenses (CRE): Recognizing that modern innovation is highly collaborative, the law allows taxpayers to claim a percentage of the amounts paid to third-party entities performing qualified research on the taxpayer’s behalf. Generally, the statutory inclusion rate is 65% of the total contract amount. However, IRC Section 41(b)(3)(C) provides a specialized tier, allowing a 75% inclusion rate for amounts paid to a “qualified research consortium,” defined as a tax-exempt organization (such as certain 501(c)(3) entities) organized and operated primarily to conduct scientific research on behalf of multiple unrelated taxpayers. To claim any CRE, the taxpayer must bear the economic risk of the research and retain substantial rights to the results.

Legislative Restructuring: Section 174, Section 280C, and the OBBBA of 2025

The landscape of federal R&D tax administration underwent a seismic shift with the passage of the One Big Beautiful Bill Act (OBBBA) on July 4, 2025. To understand the impact of the OBBBA, one must examine the preceding legal environment established by the Tax Cuts and Jobs Act (TCJA) of 2017. The TCJA contained a delayed revenue-raising provision that fundamentally altered the treatment of IRC Section 174 Research and Experimental (R&E) expenditures. Effective for tax years beginning in 2022, businesses lost the historic ability to immediately deduct these expenses; instead, they were forced to capitalize and amortize domestic R&E costs over five years, and foreign R&E costs over fifteen years. This resulted in artificially inflated taxable income and severe cash flow constrictions for innovative enterprises.

The OBBBA of 2025 effectively reversed this policy for domestic activities. The legislation introduced IRC Section 174A, which restored the option for businesses to immediately expense domestic R&E costs in the year they are paid or incurred, applicable to tax years beginning after December 31, 2024. Crucially, the law explicitly affirmed that software development costs are permanently classified as R&E expenditures under this provision. Foreign R&E costs, however, remain subject to the fifteen-year amortization schedule.

The interaction between the Section 174A deduction and the Section 41 credit requires highly sophisticated tax planning due to the anti-double-dipping rules contained in IRC Section 280C.

Comprehensive Analysis of Michigan State R&D Tax Law

The State of Michigan possesses a complex history regarding the incentivization of research and development. From 1976 until 2007, the state operated under the Single Business Tax (SBT), which included specific provisions and credits to encourage localized innovation. Following the repeal of the SBT, the Michigan Business Tax (MBT) took effect in 2008, carrying forward various R&D and historical preservation credits, though it was heavily criticized for its Byzantine complexity. In 2012, Michigan transitioned to the Corporate Income Tax (CIT), a much simpler regime that systematically eliminated the vast majority of business tax credits, including the R&D credit, leaving a significant gap in the state’s economic development portfolio.

Recognizing the necessity to compete nationally—particularly against aggressive incentive programs in neighboring states—the Michigan legislature enacted House Bills 5100 and 5101, codified as Public Acts 186 and 187 of 2024. This legislation reinstated a robust, targeted state-level R&D tax credit effective for tax years beginning on or after January 1, 2025.

Statutory Framework and Decoupling from Federal Law

The fundamental architecture of the Michigan R&D credit relies heavily on federal definitions. The Michigan statute explicitly adopts the definition of “qualified research expenses” as outlined in IRC Section 41(b), ensuring that the four-part test and the categorization of wages, supplies, and contract research apply uniformly at the state level. However, Michigan imposes a strict geographic limitation: the credit is exclusively available for expenses “incurred in this state”. Research conducted outside Michigan’s borders, even by a Michigan-headquartered company, is wholly disqualified.

A critical divergence between federal and state tax strategy occurred in late 2025. While the federal government enacted the OBBBA to restore the immediate expensing of Section 174 R&E costs, Michigan Governor Gretchen Whitmer signed House Bill 4961 on October 7, 2025. This legislation updated the state’s conformity date with the Internal Revenue Code but intentionally decoupled Michigan from the new federal Section 174A. Consequently, for Michigan state tax calculations, businesses are prohibited from immediately deducting their research costs; they must add back the federal deduction and continue to capitalize and amortize these expenses over a five-year period. This decoupling artificially inflates state taxable income relative to federal taxable income, positioning the new Michigan R&D tax credit as an indispensable tool to offset this increased state tax liability.

Credit Calculation Mechanics and Taxpayer Tiers

The Michigan credit is structurally designed to provide disproportionate support to small businesses and startups, while still offering substantial dollar-value caps for major corporations. The credit is available to both entities subject to the Corporate Income Tax (CIT) and flow-through entities (such as LLCs and S-Corporations) that are employers subject to Michigan income tax withholding.

The calculation is predicated on a “Base Amount,” defined as the average annual qualifying R&D expenses incurred within Michigan during the three calendar years immediately preceding the year for which the credit is claimed. For new businesses with no prior history, the base amount is legally set at zero. Taxpayers are stratified into two distinct tiers based on their total employee headcount, utilizing the IRC Section 3401(c) definition of an employee for withholding purposes.

Furthermore, the state introduced a unique University Collaboration Bonus. Any taxpayer, regardless of size, that incurs qualifying R&D expenses pursuant to a formalized, written agreement with an eligible Michigan research university (public universities described in the 1963 state constitution or independent nonprofit colleges within the state) may claim an additional 5% credit on those specific expenses. This bonus is capped at an additional $200,000 per year and requires the taxpayer to present the written agreement upon request by the Department of Treasury.

Aggregate Caps, Tentative Claims, and Proration Administration

Unlike the federal R&D credit, which is an open-ended entitlement, the Michigan program operates under a strict budgetary ceiling. The total aggregate amount of all R&D credits allowed across the entire state is capped at $100,000,000 per calendar year. To manage this cap, the Michigan Department of Treasury instituted a mandatory “tentative claim” process.

Taxpayers seeking the credit cannot simply claim it on their annual tax return; they must first file a tentative claim with the Treasury utilizing actual, not estimated, expense data. For expenses incurred during the inaugural 2025 calendar year, this tentative claim must be submitted no later than April 1, 2026. For all subsequent years, the statutory deadline is accelerated to March 15 of the following year. The Treasury has explicitly stated that tentative claims submitted after the statutory deadline will be categorically rejected.

Once all tentative claims are received, the Treasury evaluates the total requested amount against the $100 million aggregate cap. If the total exceeds the cap, complex statutory proration rules are triggered to reduce the allowable credits.

• Small Business Protection: The $100 million cap is sub-divided, establishing a $25,000,000 protected pool for small businesses (under 250 employees) and a $75,000,000 pool for large businesses. If the aggregate claims from small businesses do not exceed $25 million, their claims are not prorated, and the entire remaining balance of the $100 million cap is distributed among the large businesses on a pro-rata basis.
• Standard Proration: If small business claims exceed their $25 million allocation, they are prorated downwards to fit exactly within that $25 million share, while large businesses are simultaneously prorated downwards to fit within their $75 million share.
• Global Proration: In the event that small business tentative claims account for more than 25% of the total aggregate claims submitted statewide, the sub-pools are dissolved, and all claims, regardless of the taxpayer’s size, are prorated globally based on their pro-rata share of the absolute $100 million total cap.

Following the proration calculations, the Treasury issues tentative claim adjustment notices to the taxpayers, dictating the finalized, adjusted credit amount they are legally permitted to claim on their annual Corporate Income Tax return or flow-through withholding tax return. Notably, the Michigan credit is non-assignable and non-transferable, but it is refundable; if the final adjusted credit amount exceeds the taxpayer’s tax liability for the year, the state will issue a cash refund for the excess portion.

Tax Administration, Audit Defense, and Strategic Documentation

The financial magnitude of both the federal and Michigan R&D tax credits virtually guarantees rigorous compliance scrutiny and audit activity. Taxpayers operating in high-innovation ecosystems like Ann Arbor must transition from a posture of simple tax preparation to one of aggressive, contemporaneous audit defense.

The IRS Audit Techniques Guide and the Burden of Proof

The Internal Revenue Service evaluates R&D claims utilizing the specialized Audit Techniques Guide (ATG) for the Credit for Increasing Research Activities. This comprehensive manual directs field examiners to meticulously verify every element of the four-part test and the nexus between the claimed expenses and the specific business component. The defining characteristic of unsuccessful R&D claims is a failure of documentation.

The U.S. Tax Court has consistently ruled that the burden of proof rests entirely on the taxpayer to substantiate their claims. In Scott and Gayla Moore v. Commissioner, the court unequivocally disallowed wage QREs because the taxpayer relied on generalized payroll records indicating total hours worked, rather than task-specific documentation detailing the exact hours dedicated to qualified research activities. To survive an ATG-driven audit, companies must integrate specific R&D time-tracking directly into their operational and payroll architecture. This documentation must explicitly identify the business component being developed, the specific individuals performing the work, the exact technical uncertainty faced, and the nature of the experimentation utilized to resolve it.

Defending Contract Research and the Funded Exclusion

A highly litigated area of R&D tax law involves the treatment of third-party contracts and the “funded research” exclusion under IRC Section 41(d)(4)(H). Engineering, software, and contract manufacturing firms—which constitute a significant portion of Ann Arbor’s commercial base—must meticulously structure their client service agreements.

As demonstrated in Smith v. Commissioner and recently affirmed by the Eighth Circuit in Meyer, Borgman & Johnson, Inc. v. Commissioner, if a taxpayer performs research under a contract where payment is guaranteed regardless of the project’s technical success, the IRS will deem the research “funded” by the client, and the performing taxpayer will be entirely disqualified from claiming the credit. To protect their eligibility, contracting firms must ensure their agreements are structured as fixed-price or milestone-based performance contracts, legally demonstrating that the taxpayer assumes the financial loss if the technological development fails or requires extensive, unbillable rework. Concurrently, the taxpayer must ensure the contract explicitly retains their substantial rights to the underlying intellectual property, source code, or resulting technological know-how generated during the engagement.

Navigating Michigan Procedural Ambiguities

Because the Michigan R&D credit is newly reinstated, taxpayers face unique procedural hurdles and administrative ambiguities. The absolute rigidity of the tentative claim deadlines (April 1, 2026, for the 2025 calendar year) requires exceptional internal accounting coordination, as failure to file on time results in total forfeiture of the credit, regardless of technical merit.

Furthermore, because the Michigan statute cross-references the federal IRC Section 41 definition of qualified research, any fluctuations in federal case law or IRS interpretations will instantly and inherently impact state eligibility. Michigan tax practitioners are currently awaiting further binding Revenue Administrative Bulletins (RABs) from the Department of Treasury to clarify complex operational nuances. Critical unresolved issues include the precise mathematical treatment of the historical base amount when a member departs or joins a unitary business group, and the exact mechanics of how federal audit adjustments to QREs will flow through to state amended returns in subsequent years. Given Michigan’s historical propensity for litigation over the transferability and valuation of legacy Single Business Tax (SBT) and MEGA credits, meticulous corporate structural planning is essential for maximizing the new R&D incentive.

Final Thoughts

The intersection of federal and state tax policy presents a lucrative, albeit highly complex, economic landscape for innovation-driven enterprises. The federal government’s restoration of immediate domestic expensing under Section 174A via the OBBBA of 2025, when strategically coupled with the IRC Section 41 credit, provides massive federal tax liquidity for engineering and software firms. Simultaneously, the State of Michigan’s targeted reinstatement of the R&D credit—complete with protective carve-outs for small businesses and lucrative bonuses for university collaborations—serves as a vital counterbalance to the state’s restrictive requirement for R&E capitalization.

For the autonomous vehicle developers testing at Mcity, the bioscientists operating out of MI-HQ, and the cybersecurity architects born from the legacy of the NSFNET, these tax incentives are not merely administrative accounting mechanisms; they represent critical sources of non-dilutive capital essential for sustaining aggressive R&D lifecycles. However, the realization of these immense financial benefits is entirely contingent upon strict statutory compliance, rigorous contemporaneous documentation, and an acute, defensive understanding of evolving case law. By fundamentally aligning technical project management with proactive tax strategy, the industries of Ann Arbor can continue to leverage their historic foundation of innovation into sustainable, tax-optimized growth on a global scale.

The information in this study is current as of the date of publication, and is provided for information purposes only. Although we do our absolute best in our attempts to avoid errors, we cannot guarantee that errors are not present in this study. Please contact a Swanson Reed member of staff, or seek independent legal advice to further understand how this information applies to your circumstances.