Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Tax Policy topic
No spam. Unsubscribe anytime.
Study: Minnesota data‑center sales‑tax exemption created jobs but shows slight net fiscal loss
Summary
A University of Georgia analysis presented Jan. 15 to the Tax Expenditure Review Commission found the state's data‑center sales‑tax exemption generated construction and permanent jobs but produced a point‑estimate return on investment of about −2%; an alternate-use model that assumed the state collected and spent the revenue returned an estimated 18.6%.
At a Jan. 15 meeting of the Tax Expenditure Review Commission, University of Georgia economist Dr. Tommy Shepard told commissioners that Minnesota’s sales‑tax exemption for data centers—cited in the presentation as "Minnesota statute 29 978.68"—helped produce construction activity and some permanent jobs but resulted in a small net fiscal loss over the study period.
Shepard, who said he conducted the review with data from approved project applications, Department of Revenue refund reports and a commercial data set, described the exemption as covering construction materials and high‑value equipment (servers, routers, heavy‑duty cooling and related infrastructure) for new and refurbished data centers. He said the exemption’s electricity treatment differed historically and that the electricity portion was repealed in mid‑2025, which the analysis nonetheless includes in its look‑back.
"Our point estimate is that there's a very slight loss to that tax exemption," Shepard said, summarizing the fiscal calculation that combines gross foregone revenue and the economic offsets from construction and ongoing operations. The study identified 42 projects that received the exemption during the review period and reported gross foregone revenue that rose into the roughly $110–$130 million range annually during 2014–2023 before declining.
Why the gap between gross foregone revenue and net fiscal effect? Shepard noted that much of the equipment spending leaks out of the state because servers and related computer equipment are largely produced overseas; construction‑related spending, by contrast, sources more locally and creates larger multiplier effects.
The analysis separated job impacts into front‑loaded construction employment (a pronounced spike while projects were built) and smaller but accumulating permanent data‑center jobs. Shepard said construction jobs were large early in the program’s life and permanent jobs accumulated over time as facilities came online.
Commission members pressed for context and assumptions. Representative Robbins asked how many data centers were in the job totals; Shepard replied that 42 projects received the exemption. Representative Smith requested comparison with overall statewide construction employment; Shepard said he had limited his calculation to construction tied to data‑center projects and offered to run an IMPLAN comparison against statewide construction totals on request.
The report tested alternatives. In an "alternate use" scenario—where the state would have collected the taxes and spent them according to the state's historical spending pattern—the modeled return on investment was about 18.6 percent. Shepard emphasized the design differences between the scenarios: the alternate‑use calculation does not apply a "but‑for" adjustment, while the incentive analysis does, so results are not direct apples‑to‑apples comparisons.
Shepard also modeled a direct‑expenditure approach that assumed level spending into construction activity to maximize local job creation; under that set of assumptions the study estimated a much larger return to the state. The presenter and members agreed that the direct‑spend concept is hypothetical and that mechanics would require policy design work.
Commissioner Markwart asked how the Minnesota results compared with Georgia, where the institute had conducted similar work; Shepard said Georgia’s incentive (targeting hyperscale construction and structured differently) has produced a stronger positive return because the scale of construction activity has remained high.
Several members asked about externalities. Shepard said the study did not examine water use or other environmental externalities and that he assumed electrical capacity could be met for the operations modeled.
The commission’s next meeting is scheduled for Wednesday, Jan. 28 (12:00–1:30). The Legislative Budget Office will distribute compiled evaluation results and the commission will vote on whether to continue, repeal or modify tax exemptions previously reviewed.
The meeting record shows follow‑up actions requested (staff to provide additional contextual comparisons and the presenter to supply requested cross‑sector construction job context).
