Who Benefits? A Parcel-Level Analysis of Municipal Tax Incentives in Lansing

Rhinoceros Media · Civic Intelligence Working paper, September 2026

Abstract

This paper assembles that ledger for the City of Lansing from public records only: the City's audited Annual Comprehensive Financial Report (GASB Statement No. 77 disclosure), the Ingham County 2025 assessment roll, City Council packets, and Michigan State Tax Commission filings. We estimate the city forgoes or diverts on the order of $18 million per year across all mechanisms (including roughly $8.5 million in downtown tax-increment capture, which funds authority debt and administration rather than a private recipient), of which only $7.65 million (FY2025) appears in the audited abatement note. We then evaluate the individual deals approved by Council in 2024–2026 using the standard incentive-evaluation lens from the economic-development literature (cost per permanent job, additionality, and who captures the value). The results do not support a blanket conclusion in either direction. Blight-rehabilitation abatements (OPRA, NEZ) are approved to rehabilitate buildings that were genuinely obsolete or long-vacant, on a near-dead tax base. Several brownfield tax-increment plans, by contrast, direct the majority of the captured increment to the developer at public costs ranging from about $11,800 to $510,800 per permanent job. We document one case in the primary record where a developer that holds a cluster of vacant commercial buildings on these corridors held two of them empty (by the assessor's dated notes) and then, as the developer, brought a brownfield plan to the Council to redevelop those same parcels. This is one matched parcel, and brownfield law is written to let the owner of a blighted parcel become its redeveloper; we document the single instance, not a proven playbook. The contribution of this paper is methodological: a reproducible, parcel-level method for assembling the full stack of municipal incentives, applied to one mid-sized American city.

1. Introduction

The policy question motivating this analysis is simple to state and hard to answer with existing data: when a city grants a property tax break, who benefits? The developer who receives it certainly does. The city gives up revenue that would otherwise fund services. The claimed public return, usually jobs or housing or the redevelopment of blighted property, may or may not materialize, and may or may not be caused by the incentive rather than merely coincident with it.

Three features of the Michigan incentive system make this question unusually opaque at the municipal level:

1. Fragmentation. The programs are created by separate acts of the legislature, carry separate certificate regimes, and are approved by different bodies (City Council, the Brownfield Redevelopment Authority, the State Tax Commission). A resident who wants to know the total cost must reconstruct it from several disconnected sources.

2. Partial disclosure. Since fiscal year 2016, Governmental Accounting Standards Board (GASB) Statement No. 77 has required governments to disclose taxes abated under qualifying agreements. But the standard covers only certain agreement types. Housing PILOTs (structured as exemptions with a service charge), OPRA, and NEZ certificates do not appear in Lansing's Note 18, and downtown tax-increment capture is a separate diversion mechanism outside the abatement framework entirely. The audited number is therefore a floor, not a total.

3. No delivery accounting. Approvals record promised jobs and investment. Few programs condition the benefit on delivery, and none publish a post-hoc reconciliation of promised against realized outcomes. The public record captures the grant, not the result.

This paper does not argue that incentives are categorically good or categorically wasteful. It argues that the question is empirical and answerable at the parcel level, and it demonstrates a method for answering it. We measure, per deal: what program, which recipient, which parcel, how much public revenue, and against what promised public return. We then apply the cost-per-job and additionality framework that the economic-development literature has used for decades, and we let the per-deal numbers separate the programs that deliver from the programs that primarily capture value.

2. Background: Lansing's tax-incentive instruments

Two distinctions matter for the accounting and recur throughout this paper:

• PA 328 is a state-enabled but city-granted exemption. It is frequently confused with the 2014 statewide elimination of the personal-property tax (PA 80–93 of 2014), which is a different policy. The 2014 reform removed small-taxpayer and eligible-manufacturing personal property from the rolls statewide. PA 328 of 1998 is a discretionary local exemption that a city grants to a specific company in a designated distressed area. Revenue forgone under PA 328 is Lansing revenue, and it appears in Lansing's GASB 77 disclosure.

• Brownfield TIF and downtown TIF are "captures," not line-item abatements. No check is written and no rate is reduced on the recipient's bill. Instead, the increase in tax revenue that follows redevelopment is redirected (to reimburse the developer's eligible costs, or to service authority debt) rather than flowing to the general fund and the overlapping taxing jurisdictions. The economic effect is forgone general-fund revenue; the accounting treatment differs.

3. Prior literature and the evaluation framework

The evaluation framework in this paper is not novel; it is the standard apparatus of the incentive-evaluation literature, applied at the municipal parcel level.

Additionality and the "but-for" problem. Timothy Bartik of the W.E. Upjohn Institute is the central authority. His review of the research ("But-For Percentages," Upjohn working paper 18-289, 2018; Making Sense of Incentives, 2019) finds that a typical incentive tips the location or expansion decision for only about 2 to 25 percent of recipient businesses; for at least three-quarters of them, the business would have made a similar decision without the incentive. Where an incentive funds activity that would have happened anyway, it is a transfer to the recipient rather than a purchase of new activity. Bartik also finds that customized business services (job training, supplier matching) typically cost less than one-third as much per job as tax incentives. The cost-per-job metric we use below is Bartik's framework.

Tax-increment financing redistributes more than it creates. The empirical TIF literature (reviewed in Merriman; Lincoln Institute of Land Policy) is mixed-to-negative: TIF districts grow, but the non-TIF remainder of the municipality tends to grow no faster, and some studies find slower municipal growth after TIF adoption. The consistent finding is that TIF relocates activity within a jurisdiction rather than generating net new activity. The "but-for" question (would redevelopment have occurred without the capture?) is again the crux.

Disclosure and the measurement gap. Good Jobs First (Greg LeRoy) built the model this analysis follows: its Subsidy Tracker and Tax Break Tracker compile incentive awards into searchable ledgers, and the organization's advocacy produced GASB 77 itself. Even so, local programs remain thinly evaluated: a 2022 Virginia statewide housing study (HB854) found that the "scale, scope, and effectiveness of these various provisions have never been systematically studied or evaluated," a conclusion drawn about local housing-tax provisions that applies equally to the broader local incentive stack this paper assembles.

PILOTs are a distinct instrument. Most PILOT scholarship (Lincoln Institute) concerns negotiated payments from tax-exempt nonprofits (hospitals, universities). Lansing's housing PILOTs are a different thing: a statutory affordable-housing service charge under MCL 125.1415a, not a negotiated nonprofit payment. We treat them separately.

Stance. We weigh both the deadweight critique (Bartik; the TIF literature) and the case that some abatements do pay for themselves, apply the cost-per-job and additionality lens, and let the per-deal evidence decide. This is what distinguishes, in our results, OPRA blight-rehabilitation (a near-dead tax base, a plausible but-for case) from several brownfield plans (high capture, few jobs).

4. Data and methodology

4.1 Sources

All figures in this paper are drawn from public records. The primary sources, with vintages:

• City of Lansing Annual Comprehensive Financial Report, FY2025, Note 18 (GASB 77 tax-abatement disclosure). The audited totals for IFT, PA 328, Brownfield, and Commercial Rehabilitation. FY2024 totals are available in the prior year's ACFR for comparison.
• Ingham County 2025 assessment roll (FOIA export). Parcel master list, names file, values, and the assessor's field-canvass comment file. Used for parcel-level taxable values, ownership, property class, and documented vacancy. The roll carries about 116,700 parcels; the field-canvass comment file, the source for documented vacancy, carries 91,482 notes.
• City Council and Brownfield Authority meeting packets and minutes (CivicClerk), 2023–2026. Staff memos and brownfield plans with promised jobs, investment, reimbursement, and term for each deal (e.g., the four housing PILOTs in the September 28, 2026 regular session, CivicClerk event 7887; the June 9, 2025 brownfield packet; Plan #84 in the October 30, 2023 session, event 5054 / minutes file 7164).
• Michigan State Tax Commission filings. The May 2026 exemptions agenda (the PA 198 industrial-facilities certificate 2026-008) and the 2015 revocation-denial notice (the PA 328 certificate 423-2013).
• Michigan corporate records (LARA). Resident-agent and registered-office records used to attribute each developer LLC to its principal (e.g. the Gillespie Group buildings to Patrick K. Gillespie).
• Ingham County Register of Deeds. Recorded property transfers used to confirm acquisition dates against staff-memo claims (e.g. the October 2022 deed conveying the Pacific Flatz parcels).

4.2 Estimation approach, by mechanism

• Audited programs (IFT, PA 328, Brownfield, Commercial Rehab): taken directly from the GASB 77 note. No estimation.
• Housing PILOTs: not in the GASB 77 note. We estimate forgone revenue parcel by parcel from the assessment roll. Where we have the actual taxable value (e.g. the SK Lansing portfolio), we apply the FY2025 non-homestead millage of 62.6 mills to the taxable value. Where we have only estimated true cash value, we apply the standard half-of-TCV assessment ratio and then the millage, and label the result an upper bound. The PILOT estimate is approximately $2.2 million per year, against roughly $73 million of market value held off the roll.
• Downtown tax-increment capture: taken from the companion transparency analysis of the FY2025/FY2026 budget (approximately $8.3–8.5 million per year). This is a diversion, not an abatement, and is reported separately.

The citywide total of approximately $18 million per year is the sum of the audited GASB 77 total, the PILOT estimate, and the downtown capture, with OPRA and NEZ acknowledged as additional but not separately quantified at the citywide level.

4.3 Per-deal evaluation

For each deal approved in the 2024–2026 window we record the program, recipient, parcel, public cost (reimbursement for brownfield; rehab investment and frozen value for OPRA/NEZ; estimated forgone for PILOT), and the promised public return as stated in the staff memo. We compute cost per permanent job where the deal is justified on employment, using the permanent/operational job count stated in the packet, not temporary construction jobs. (We verified the basis directly: the 2025-06-09 packet language reads "the completed project will create 13 full-time equivalent jobs" and "approximately 2 full-time jobs," confirming the figures are permanent.) Where a deal's promised jobs include positions retained rather than newly created, we report cost per new job, following the additionality principle that only new activity is a public return (416 N. Cedar, for instance, promises 5 new jobs and 20 retained, so its figure is cost per new job); deals with no retained component report cost per total permanent job. We classify each deal as delivers, captures, review, or pending based on the ratio of public cost to verifiable public return.

4.4 Precision commitments and what we do not claim

Consistent with a publish-first discipline, we hold several precision commitments:

• Exact identifiers match exactly. Certificate numbers, committee IDs, and EINs are matched on exact normalized equality, never by fuzzy similarity.
• Estimates are labeled. Every PILOT figure derived from the roll is an upper-bound estimate and is labeled as such; audited figures are labeled audited.
• Two breaks that look alike are kept apart. Niowave appears under two statutes: its PA 328 personal-property exemption (certificate 423-2013, part of the FY2025 $2.73 million PA 328 line) and a new PA 198 industrial-facilities abatement (certificate 2026-008, a $38,268,475 investment over 2026–2039). These are different certificates under different acts; the $38.27 million is an investment figure, not a state-equalized value and not a forgone-tax amount. We do not add or conflate them.
• Developer identity is verified by resident agent, not inferred. The eight vacant commercial buildings attributed to the Gillespie Group are each held by a single-asset LLC whose Michigan registered resident agent is Patrick K. Gillespie at 330 Marshall St. This was confirmed against corporate records, not inferred from the shared tax-mailing address. It is kept distinct from the separate company run by his brother, Scott Gillespie (The Gillespie Company, 1629 E. Michigan Ave); none of the eight buildings is Scott's.
• A documented vacancy is dated. Every vacancy finding carries the date of the assessor's field canvass; older notes are treated as "last canvassed vacant," not as a current-year assertion.
• We do not assert delivery we cannot verify. For 2026 approvals, jobs and occupancy are promised, not realized; we say so.

A false-positive finding we caught and discarded illustrates the discipline: an early pass computed that 64 percent of one developer's parcels were "vacant" using the roll's vacant flag, until validation against the full roll showed that flag is set on 88 percent of commercial-improved parcels citywide (it encodes an occupancy/homestead status, not an empty building). The finding was withdrawn. The vacancy results reported below rest instead on the assessor's dated field-canvass notes and the named-tenant field, which record vacancy explicitly.

5. Results

5.1 Citywide magnitude

The audited GASB 77 disclosure for FY2025:

The audited total rises to $7,654,357 in FY2025. These figures are the floor. Adding the housing-PILOT estimate (approximately $2.2 million per year) and the downtown tax-increment capture (approximately $8.5 million per year), with OPRA and NEZ additional but unquantified, yields a citywide total on the order of $18 million per year. The single largest component is not in the audited abatement note at all: it is the downtown capture.

For scale, the city's total taxable value in FY2025 was approximately $4.36 billion, roughly double the 2016 figure of $2.17 billion. The ten largest taxpayers account for about $425 million. The city's ACFR reports their share as 12.95 percent (Statistical Table 7); measured against the $4.36 billion taxable base cited above, the same $425 million is about 9.75 percent, so the two figures rest on different base definitions.

5.2 Who receives what, where, and for what

The per-deal ledger (full table in the Appendix) covers deals introduced or approved in 2024–2026. Selected entries:

PA 328 (personal property). The city's PA 328 distressed-community personal-property exemption is held by Niowave, Inc., a superconducting-radiofrequency and particle-accelerator firm at 1012 N. Walnut Street (the Clinton County portion of the city), under certificate 423-2013. Niowave's exact share of the $2.73 million FY2025 program total requires the assessor's personal-property exemption roll to isolate; we do not assert it is the entire line. The accountability history is documented in State Tax Commission records: in 2015 the Lansing City Council asked the Commission to revoke Niowave's certificate (423-2013), and at its August 25, 2015 meeting the Commission determined it had no statutory authority to revoke the exemption and denied the request. The city granted the break and could not claw it back. Separately, the State Tax Commission's May 2026 agenda approves Niowave for a new PA 198 industrial-facilities abatement (certificate 2026-008) on a $38,268,475 investment over 2026–2039. This is a different statute and a different break from the PA 328 exemption; the $38.27 million is the investment figure, not a state-equalized value and not a forgone-tax amount.

Brownfield TIF. The brownfield plans concentrate in one developer's portfolio and on specific corridors.

• Plan #88, Michigan Ave Revitalization LLC (Gillespie), 603–607 E. Michigan Ave (Brick Row): $1,021,517 developer reimbursement out of $1,415,011 total capture (72.2 percent to the developer), on an $8.19 million investment, for 2 permanent jobs plus mixed-use housing. Cost per permanent job: $510,759.
• Plan #92, 416 N. Cedar LLC (Gillespie), 416 N. Cedar St: $871,088 developer reimbursement out of $977,736 total capture, 20 jobs retained and 5 new, diverting $106,464 from schools, over a 24-year term. Cost per new job: $174,218. The rehabilitated building is tenanted by a lobbying firm.
• Plan #86, Novi Properties LLC, 820 W. Miller Rd: $439,524 reimbursement on a $3.84 million investment for 13 full-time jobs. Cost per job: $33,810.
• Plan #91, GTZ Properties (6420 S Cedar), a former car lot redeveloped as a fast-food outlet: $709,650 reimbursement for 60 jobs at $15–23/hour. Cost per job: $11,828.

OPRA and NEZ (blight rehabilitation). These deals are approved to rehabilitate obsolete or long-vacant buildings. (The packets document each building's pre-rehab condition; realized occupancy is not independently verified, per the Limitations.)

• OPRA, T.C.I. Properties LLC, 1515 W. Mt Hope Ave: $1.12 million rehabilitation of a 1954 building with uninhabitable upper floors into 11 commercial and 4 residential units.
• OPRA, Maximillian Boortz, 414 Baker St: $437,000 rehabilitation of a 1915 building vacant for decades with a failed roof into a mixed-use duplex.
• NEZ, Summit Street Development, 700 May St: 134 residential units in the rehabilitation of the historic Prudden Wheel building.

PILOT (income-restricted housing). The housing PILOTs are the largest off-the-books component. The SK Lansing portfolio (125 parcels) carries an estimated $436,197 per year in forgone revenue; a set of ten senior and family complexes carries an estimated $1.16 million combined; the four Simtob-family "Flatz" deals (440 units, introduced September 2026) carry an estimated $445,000 per year over a 15-year term. The staff memo for the Flatz deals dates the purchase of the Pacific buildings to 2020, a date the recorded deed (October 2022) and the sponsoring LLC's July 2022 formation contradict, and two of those buildings remained red-tagged "unfit for human occupancy" on the city's May 2026 monitoring list; the companion report documents this in full (Rhinoceros Media, "Four More Tax Breaks Before the Lansing Council, All One Operator").

5.3 Cost per job: does it deliver, or capture?

Ranking the employment-justified deals by cost per permanent job produces a clear spread:

For comparison, Bartik's work finds customized business services deliver jobs at well under one-third the cost of tax incentives, and that the activity would have occurred anyway for at least three-quarters of incented businesses. That is a general finding across many jurisdictions, not a specific additionality test of any Lansing deal. Measured against the other three deals in this table, though, a $510,759 public cost per promised permanent job is nearly three times the next-highest of the four ($174,218) and more than forty times the lowest ($11,828). The better lens for #88 may be its housing: the mixed-use plan ties 3 of its 15 units to a cap of 120 percent of area median income for fifteen years, through an MSHDA Housing TIF component, with the rest market-rate (see §5.5). Judged as a housing deal rather than a two-job one, the public return is that partial, time-limited affordability, set against the roughly $1 million reimbursement to the developer.

The share of the captured increment that flows to the developer is the second diagnostic. At #88, 72.2 percent of the total capture reimburses the developer. The plan shows no recapture: if the promised jobs do not appear, the #88 packet provides no mechanism to recover the reimbursement.

5.4 Land-banking, and one hold-empty-then-subsidize case

The assessor's field-canvass notes record commercial-building vacancy explicitly and with dates. In one documented case they line up with the subsidy record. For the Gillespie Group portfolio (eight single-asset LLCs whose registered resident agent is Patrick K. Gillespie of the Gillespie Group, 330 Marshall St, confirmed in Michigan corporate records), the notes document at least eight vacant commercial buildings, including:

The assessor's phrase, "STILL VACANT" in October 2024, denotes multi-year vacancy. The connection to the subsidy record is direct: 505–507 E. Shiawassee were recorded unoccupied in October 2022, and those are the parcels of Gillespie's Brownfield Plan #84 (approximately $234,000 in capture). Three Gillespie Group executives, by their Gillespie Group titles (the environmental consultant, the director of finance, and the director of project development), presented Plan #84 to the City Council at its October 30, 2023 public hearing. The sequence, hold the buildings empty then seek tax-increment financing to redevelop the same parcels, is documented across two independent public records: the assessor's field note and the Council's hearing minutes. The vacant buildings and the subsequent brownfield plans cluster on the same corridors (E. Michigan, Shiawassee, Larch, Cedar).

Gillespie's eight-building cluster is the largest, but it is not the only concentration of vacant commercial stock, and establishing that required looking past the owner name on the roll. An owner who places each building in its own differently-named LLC, as Gillespie does, appears in the assessment roll as several unrelated "single owners"; grouping by name alone therefore understates concentration. To measure it, we resolved 181 of the 204 corporate owners among the 287 vacant commercial buildings to a Michigan LARA record (resident agent and registered office), counting the 23 that returned no match or an error as single owners, and clustered on the resident agent, with one precision rule: a shared agent counts as common control only when the agent is an individual person. Commercial registered-agent services and law firms (CSC, CT Corporation, Cogency, and PLLC/P.C. firms) are shared by unrelated clients, so clustering on them would be a false merge; those were excluded.

After that tracing, 235 of the 287 buildings have an owner we found holding nothing else. The remaining 52 are held by 21 owners with two or more buildings: the Gillespie Group's eight, plus 20 owners holding 44 buildings between them. Eight of those 20 are concentrations the name-level view misses entirely, because the buildings are held under separately named LLCs that resolve to one principal: a second Gillespie (Scott P. Gillespie, three buildings on E. Michigan Ave, registered at the 1629 E. Michigan address of his company, The Gillespie Company), the developer Paul V. Gentilozzi (two downtown buildings), Bradley Simtob (two), and five others. Each of these clusters is corroborated by a second independent signal: the LLCs in it share not only an individual resident agent but a single registered office (Scott Gillespie's three at 1629 E. Michigan Ave #103, for example, and Pat Gillespie's eight at 330 Marshall St). The resident-agent and registered-office evidence is the same standard applied to the Gillespie eight, now extended to the whole roll.

One of these owners does appear in the subsidy ledger of §5.2: Bradley Simtob, whose family holds the four 2026 Flatz housing PILOTs, also holds two vacant commercial buildings here, though not the subsidized parcels. The others are not on record seeking a brownfield or other incentive on their vacant stock. Gillespie is the only owner in this sweep matched to a subsidy on parcels it had documented vacant (Plan #84). The full sweep establishes three things: vacant-stock concentration through multiple LLCs is a citywide pattern, it is measurable to a defined standard, and Gillespie ranks first in it, both in size and as the one case where a documented-vacant building and a brownfield plan fall on the same parcel.

We separate what the data supports from what it does not. The data supports that the buildings were vacant on the dates canvassed and that the same owner pursued subsidy on the same parcels. The data does not, by itself, establish intent to use vacancy as a tax shield; that is a hypothesis the sequence raises, not a fact the records prove. What the records establish is this one documented instance of the sequence, plus a broader concentration: the vacant commercial buildings this developer holds and the brownfield plans it wins cluster on the same corridors. That overlap is geographic, not proof of a repeated playbook; 505-507 E. Shiawassee is the single parcel where a vacant Gillespie building is matched to a Gillespie brownfield plan. The companion web application maps the overlap.

A methodological note bears repeating here, because it reverses an earlier error. An initial pass concluded the empty-building hypothesis was "not supported" because the developer had zero code-enforcement red tags. That was the wrong instrument: a red tag certifies a building unfit for occupancy (a safety failure), and a well-maintained but empty building is never red-tagged. The correct instrument was the assessor's occupancy record, which documents the vacancy plainly. The red tag was the wrong instrument for the question, not evidence the vacancy was absent.

Counter-arguments considered. Two points weigh against reading even this one instance as opportunism. First, the Brownfield Redevelopment Financing Act (PA 381 of 1996) does not bar the owner of a blighted or contaminated parcel from becoming its redeveloper, financed by the increment the redevelopment creates; an owner who holds a vacant building and then brings a brownfield plan for it is following a sequence the statute permits, and brownfield planning itself (environmental assessment, a multi-step authority approval) is not a quick process. Second, this paper applies the ownership-before-subsidy check to Gillespie's brownfield deal but not, with the same rigor, to the OPRA and NEZ recipients it credits in §5.2 with plausibly clearing the but-for test, several of whom also held an obsolete or long-vacant building before seeking the certificate meant to rehabilitate it. The line this paper draws between the two is a framing choice about one program versus another, not a distinction the assessor notes or the statutes themselves establish.

5.5 Housing: who bears the occupancy risk

From the city's perspective, the housing deals raise a structural question: does the developer profit from building regardless of whether the units fill, and is the "affordable" housing affordable?

Affordability varies sharply by program. Housing PILOTs under MCL 125.1415a are genuinely income-restricted, with statutory caps (and, in the Flatz memos, current rents running at 35–55 percent of area median income, affordable now even though the binding cap is looser). Brownfield residential does not carry a blanket affordability requirement, but a specific plan can attach one: the Brick Row component of #88, for example, ties 3 of its 15 units to a 120-percent-of-area-median-income cap for fifteen years through an MSHDA Housing TIF component, leaving the rest market-rate. OPRA and NEZ residential is market-rate unless separately restricted.

Bartik's additionality applies directly here: where the subsidy is not conditioned on delivery, it rewards the project (the developer fee, the tax credits, the reimbursement, the eventual refinance or sale) rather than sustained affordable occupancy. A developer can be made whole by building even if the units sit empty or the affordability lapses at the end of the term. Testing whether specific projects are "market-rate that does not fill" requires lease-up and occupancy data that no public dataset currently provides; it is the clearest next data requirement.

6. Discussion

The programs separate cleanly when measured:

• Blight-rehabilitation abatements (OPRA, NEZ) take buildings that were obsolete or vacant for years, with documented physical failure (failed roofs, uninhabitable floors), and are approved to rehabilitate them into commercial and residential use. The frozen or reduced value is on a building that was producing little tax and no use. These deals plausibly clear the but-for test (the rehabilitation of a decades-vacant 1915 building is not the counterfactual baseline), though, as with every deal here, realized occupancy is not independently verified (§7).

• Several brownfield tax-increment plans direct the majority of the captured increment to the developer at public costs per promised permanent job that range into the hundreds of thousands of dollars, with no recapture shown in the plans and, in one documented case, on parcels the same developer had held vacant. These are the deals where the additionality question is sharpest and the public return is weakest on the stated (employment) justification.

• The largest single mechanism, downtown tax-increment capture, is invisible in the audited abatement disclosure and services authority debt and administration rather than any project-specific job or housing promise. Its scale (approximately $8.5 million per year, larger than the entire GASB 77 total) is the strongest argument for assembling the full ledger rather than relying on the audited note.

The contribution is the method. Good Jobs First assembles award-level data nationally; GASB 77 produces an audited municipal total for qualifying agreements; neither produces a parcel-level, all-mechanism municipal ledger that pairs each deal with its promised return and its documented context (including whether the recipient held the parcel vacant before seeking subsidy). The Virginia housing study's observation that such local programs "have never been systematically studied or evaluated" names the standing gap. This analysis is a reproducible template for closing it in any Michigan municipality: the GASB 77 note, the county assessment roll, the council packets, and the State Tax Commission filings are public everywhere the programs operate.

7. Limitations

• PILOT figures are upper-bound estimates derived from the assessment roll, not audited amounts. Where only estimated true cash value was available, the half-of-TCV ratio introduces additional uncertainty.
• The PA 328 recipient's exact share of the $2.73 million program total is not isolated; it requires the assessor's personal-property exemption roll.
• Delivery is not measured for any deal. Promised jobs and occupancy, for the brownfield plans and the OPRA/NEZ rehabs alike, are not independently verified against realized outcomes (certificates of occupancy, tenant records, job counts). The cost-per-job figures and the delivers/captures labels rest on the packets' promised figures and on each deal's documented inputs (e.g., a building's pre-rehab condition), not on confirmed delivery.
• OPRA and NEZ forgone revenue is not separately quantified. We capture each OPRA/NEZ deal's rehab investment and pre-rehab value, but the per-certificate abated tax requires the assessor's special-act roll (an open records request). OPRA and NEZ are acknowledged in the citywide total but not line-itemed, so the "near-dead tax base" characterization rests on the pre-rehab taxable value, not a computed forgone figure.
• The $18 million total mixes fiscal years and mechanism types. It sums the FY2025 audited GASB 77 figure, the FY2025 downtown capture, and a roll-based PILOT estimate: different vintages and two different kinds of diversion (line-item abatement vs. tax-increment capture). It is an order-of-magnitude figure, not a single-year audited sum.
• The four Corridor Improvement Authority districts' tax-increment capture is excluded from the $18 million, which likely makes the total an undercount.
• Vacancy notes carry their canvass date. Older notes (some from 2008 and 2017) indicate the building was vacant when last canvassed, not necessarily in 2025–2026.
• Commercial occupancy is not fully quantifiable from bulk public records. The city maintains no vacant-commercial-building registry; the vacancy findings rest on the assessor's field notes and named-tenant field, which are thorough but not a purpose-built occupancy census.
• The analysis covers the City of Lansing only. The corridors and developers extend into adjacent jurisdictions not inventoried here.

None of these limitations affects the audited GASB 77 totals, the certificate-level PA 328 facts, or the dated vacancy-to-brownfield sequences, which rest on primary documents.

8. Conclusion

A resident of Lansing who asks "how much tax revenue does the city give up, to whom, and for what?" cannot answer the question from any single public source. Assembled from the audited financial report, the county assessment roll, council packets, and State Tax Commission filings, the answer is on the order of $18 million per year, most of it outside the one audited disclosure that exists. The deals that give up that revenue do not all fall the same way. Blight-rehabilitation abatements are approved to bring long-dead buildings back into use; several brownfield plans direct most of a parcel's growth in value to a developer for a handful of jobs, in one documented case on parcels that developer had held vacant first. The difference is measurable, deal by deal, and measuring it is the prerequisite for any honest claim about who benefits.

References

• Bartik, Timothy J. Making Sense of Incentives: Taming Business Incentives to Promote Prosperity. W.E. Upjohn Institute, 2019.
• Bartik, Timothy J. "'But For' Percentages for Economic Development Incentives." Upjohn Institute Working Paper 18-289, 2018.
• Good Jobs First. Subsidy Tracker and Tax Break Tracker. goodjobsfirst.org.
• Governmental Accounting Standards Board. Statement No. 77, Tax Abatement Disclosures. 2015.
• Kenyon, Daphne A., and Adam H. Langley. Payments in Lieu of Taxes: Balancing Municipal and Nonprofit Interests. Lincoln Institute of Land Policy, 2010.
• Merriman, David. Improving Tax Increment Financing (TIF) for Economic Development. Lincoln Institute of Land Policy, 2018.
• Virginia Department of Housing and Community Development and Virginia Housing. HB854 Statewide Housing Study: Current Efforts, Future Needs, New Strategies. 2022.
• City of Lansing. Annual Comprehensive Financial Report, FY2024 and FY2025, Note 18 (Tax Abatement Disclosures).
• Michigan State Tax Commission. May 2026 Exemptions Agenda (PA 198 IFT certificate 2026-008); September 8, 2015 revocation-denial notice (PA 328 certificate 423-2013).
• Ingham County. 2025 assessment roll (parcel master list, names, values, field-canvass comments). FOIA export.
• City of Lansing. City Council meeting packets, 2024–2026 (CivicClerk).
• City of Lansing / Lansing Tax Increment Finance Authority. FY2026 budget and the LTIFA budget table presented at the TIFA meeting of June 5, 2026 (source for the downtown-capture figure).
• Ingham County Register of Deeds. Recorded property-transfer instruments (e.g. the October 11, 2022 deed conveying the Shiawassee/Pacific Flatz parcels to Monopoly Apartments LLC).

Appendix A: Per-deal ledger

Every break in the registry (2023–2026 window). "Public figure" is the developer reimbursement for brownfield, the rehab investment or pre-rehab value for OPRA/NEZ, the audited program total or estimated annual forgone for PA 328 / PILOT; figures labeled as estimates in the text carry the same caveat here. Realized delivery is not verified (see §7). Full machine-readable data in the companion web application.

Prepared by Rhinoceros Media. All figures trace to the public records cited. Per-deal data and the interactive ledger are available in the companion web application.