Supreme Court Denies UT’s Bid for Med Students FICA Refund
Client Alert | 1 min read | 05.05.15
The Supreme Court denied without comment the University of Texas' petition to recoup $11 million in Federal Insurance Contributions Act (FICA) taxes paid with respect to medical residents, letting stand a Fifth Circuit decision denying the refunds because the residents were not "students."
UT's argument was based on Texas's "§ 218 agreement" with the Social Security Agency (SSA), pursuant to which Texas opted into the Social Security system for its employees. The § 218 agreement excludes services performed by students, and therefore such services are exempt from FICA tax. Applying a contractual approach to Texas's § 218 agreement, the Fifth Circuit noted that at the time Texas entered into the agreement, the SSA clearly stated its position that medical residents did not fall within the meaning of the term "student" for purposes of the student exclusion. There was no evidence presented that Texas had a different understanding.
In its petition to the Supreme Court, UT argued that the Fifth's opinion conflicted with a similar Eighth Circuit case, Minnesota v. Apfel, which potentially caused a difference in law between the Eighth Circuit and the rest of the country. After Apfel, however, the IRS adopted regulations stating that a school employee who works 40 or more hours per week is not exempt from FICA taxes. Those regulations are applicable for services performed on or after April 1, 2005.
Crowell & Moring's Education Group will continue to monitor cases and legislation important to post-secondary education institutions, ensuring awareness of any changes.
Contacts
Insights
Client Alert | 13 min read | 10.30.25
Federal and State Regulators Target AI Chatbots and Intimate Imagery
In the first few years following the public launch of generative artificial intelligence (AI) in the autumn of 2022, litigation related to AI focused primarily on claims of copyright infringement. Suits revolved around allegations that the data on which AI models train, and/or the output they produce, infringe upon the intellectual property rights of others. (While some of these cases have settled or reached preliminary judgments, many remain ongoing.)
Client Alert | 3 min read | 10.30.25
Is Course Hero Heading to Summer School After Summary Judgment Loss?
Client Alert | 6 min read | 10.29.25
Enhancing UK cyber security resilience and leadership engagement
Client Alert | 9 min read | 10.28.25
Key Takeaways from a Consequential Month of Russia-Related Sanctions

