The Quiet Reinvention of Government Buying

Baseline Advisory   ·   Perspective   ·   September 2026

Local government never announced a change in how it buys technology. It simply started buying differently. The numbers say the shift is real — and that the decisive moment has moved upstream, to a place almost nobody is watching.

There was no announcement. No reform act, no landmark ruling, no keynote that marked the turn. But somewhere in the last five years, the way counties and cities buy software and appraisal services stopped resembling the way they bought it for the thirty years before. Five readings tell the story.

$168.5B → $209.8B

U.S. state, local and education IT market, 2026 projected to 2029 — roughly 7.5% average annual growth (Deltek | GovWin)

Start with what did not happen. Federal stimulus wound down, budgets tightened, and the market did not contract — it is projected to grow about 7.5% a year through 2029. What is changing is how often an agency reaches into that pot. Bid volumes fell in both 2024 and 2025 even as the dollars climbed, and buyers are shifting from fragmented one-off solicitations toward durable, longer-term vehicles. Fewer at-bats, bigger swings. A requirements document that was merely adequate in 2015 is load-bearing today.

$45B → $61B

Projected growth in national cooperative contract sales, 2019 to 2025 (GovWin)

This is the largest structural change of the four, and the least discussed. A single competitively awarded cooperative contract can be used by tens of thousands of agencies, and the largest national programs together move tens of billions a year on top of direct agency spending. For a growing share of purchases, the competitive event already happened — somewhere else, years earlier, against someone else’s requirements. It is the difference between commissioning a house and buying one already framed. Both can be right. They are not the same decision, and they do not reward the same preparation.

$609 billion

Property taxes collected by local governments in a single year — about 30% of all local general revenue, and the largest tax source in more than 90% of localities (Tax Policy Center; ITEP)

This is what makes assessment and tax procurement a different category of decision. The system under evaluation is not a back-office convenience. It is the machinery that produces most of the money the jurisdiction runs on, and four in five public-sector enterprise implementations still finish over budget and late. Elsewhere that means an awkward year. Here it means a certified roll against a statutory deadline, an appeal calendar that does not move, and a levy that has to be set whether the software is ready or not. Tolerance for a mediocre requirements document is lower here than almost anywhere else in local government.

Every 3 to 6 years

Typical statutory reassessment interval across most states — some jurisdictions revalue annually with physical inspection on a rolling cycle, a few let a decade pass (state revaluation statutes)

Software is the procurement everyone remembers, because it happens roughly once a generation. Appraisal and reassessment services are the one that repeats. Each cycle is its own solicitation — data collection, valuation modeling, commercial appraisal, informal review, appeal defense — and a county may buy a CAMA system twice in an assessor’s career while buying reappraisal services eight or ten times. The requirements problem here is not a rare event. It is the recurring one, and scope language left unexamined in one cycle is usually copied straight into the next.

80% over 50

Share of U.S. property appraisers and assessors over age 50; about 5,200 openings a year, most of them replacing people who retire or leave the field (industry analysis; U.S. Bureau of Labor Statistics)

The capacity gap is not about procurement staff. It is about the appraisers, deputy assessors and tax office veterans who can explain why an exemption is coded the way it is, or which shortcut from the last conversion everyone has worked around since. That knowledge almost never exists in writing. So when a jurisdiction replaces a system that has run for twenty years, the requirements get reconstructed from memory — at exactly the moment that memory is walking out the door.

What is actually being reinvented

From capital project to standing relationship

Subscription models mean the agency is not buying an asset once. It is choosing a vendor it may still be working with two decades from now, long after everyone who signed the contract has moved on. From one office to a shared table

Selection rarely belongs to a single department anymore. Security, integration, data governance and the operational result all have to be answered in the same document, by the people accountable for each of them.

This reinvention is real and, on balance, healthy. But it has quietly relocated the moment that determines success — out of the evaluation, where the process is well governed and heavily documented, and into the definition, where it usually is not. The agencies that do well over the next decade will not be the ones with the best RFP template, but the ones who know precisely what they are buying before they ask anyone to sell it.

Tony Henderson   ·   Founder, Baseline Advisory

Sources: Deltek | GovWin; Tax Policy Center; ITEP; Panorama Consulting; U.S. Bureau of Labor Statistics.