JFO modeling: changing Vermont property tax credit would shift $153 million and redistribute tax burden

Julia Richter, a fiscal analyst with the Joint Fiscal Office, told a legislative committee that Vermont’s homestead property tax credit was estimated to cost about $153 million in fiscal year 2025 and that proposals to change the credit’s income sensitivity would reallocate that money across taxpayers.

"We’re nonpartisan legislative office producing unbiased fiscal analysis," Richter said, explaining the office’s role and the limits of the briefing. She walked the committee through the mechanics of the credit, the modeling assumptions and three illustrative scenarios intended to show bounds of possible outcomes.

The nut of the briefing: the homestead property tax credit and the so‑called circuit breaker interact with education fund tax rates in an iterative, "self‑leveling" way. Richter said the office models the credit by estimating the credit earned in one year (the JFO cited FY2025 figures), then calculating how much education tax revenue must be raised the next year to offset that credit. Because changing the credit changes rates and rates change the credit, the estimates require repeated recalculation to converge on a final figure.

Richter described three modeling approaches the office ran as examples. In the first, she modeled a theoretical removal of income sensitivity entirely — effectively eliminating the credit for eligible filers and applying the $153 million instead to lower statewide homestead and nonhomestead rates. In that scenario, some higher‑income or high‑house‑value taxpayers would see decreases in their net bills, while some low‑income households with relatively high house values would face higher bills. Richter said, for FY2025 modeling limits, the breakeven household income where property tax becomes always cheaper than income tax was about $115,000.

In a second scenario Richter modeled keeping the existing circuit breaker for households up to $47,000 of income but removing income sensitivity above that threshold. Because the state’s FY2025 circuit breaker cost was about $70 million of the $153 million total, that left roughly $83 million to reallocate. Richter said she ran both a uniform rate reduction (homestead and nonhomestead) and a homestead‑only buy‑down. Both approaches concentrated most bill reductions around moderate‑income, moderate‑house‑value households; at the same time some low‑income, high‑house‑value cells would face increased liability.

Richter also flagged sample aggregates the model produced: when the office applied all available savings to homestead taxes only in one run, the net increase in liability for the eight highest‑affected income/house‑value cells summed to roughly $55 million in additional tax for those households; in another accounting she cited about $90 million as a sum of increases in a different partitioning. Richter emphasized those were example outputs to illustrate distributional effects rather than policy endorsements.

Committee members raised practical and equity concerns. Several members noted households that inherited homes or have seen property values appreciate while household income remained fixed — often older or single‑earner households — could be harmed by measures that tax by house value alone. Members also discussed alternatives such as tweaking income and house‑value thresholds, using occupancy or square‑footage measures, or pursuing a targeted wealth attestation; speakers cautioned each option presents data, administrative and political challenges.

Richter flagged two data and modeling limits repeatedly cited during the briefing: the office’s current data set covered filers who reported household income below the modeled cutoff (about $115,000 for FY2025), and one illustrative run temporarily ignored the one‑year lag between when credits are earned and when they affect rates in order to show an immediate comparison. She said JFO staff are working with the tax department to obtain better data for higher‑income households and that iterative recalculation is required to converge on a final fiscal estimate.

The presentation also clarified how funds are carried: the education fund bears the state portion of the circuit breaker while the municipal circuit breaker component is paid from the general fund, so changing statewide income sensitivity would not automatically change municipal general‑fund payouts.

Committee members asked Richter to run additional permutations — for example, modestly raising or sloping the $47,000 circuit‑breaker threshold, or combining a house‑value exemption with other adjustments — and several members said the office’s modeling would be useful to compare against outside groups’ proposals, including work referenced from the Public Assets Institute and consultants such as Jake Felder and Austin Davis (both cited to illustrate alternative designs or past testimony).

Richter closed by reiterating the trade‑off: increasing the property tax credit for some households requires raising more revenue elsewhere to keep the education fund whole; decreasing the credit reduces that revenue need but shifts liability to other taxpayers. She urged committee members to treat her scenarios as starting points for further analytic work rather than finalized policy calculations.

Next steps recorded during the meeting included committee requests for refinements to the JFO model, more complete taxpayer data above the $115,000 illustrative cutoff, and scenario comparisons that factor the lag and iterate to convergence.