What districts paid, and what it bought.

The consolidation decision is being made blind.

Sixty-one percent of school districts expect to consolidate spending with fewer vendors now that federal stimulus money is gone.¹ Every one of those decisions rests on a question that sounds simple: which of our vendors has a record worth keeping?

Nobody can answer it. Not because the answer is hard, but because the record does not exist in one place. What a district paid its vendors sits in a procurement system. How its students performed sits in an assessment file. In most states, no person and no institution has ever put the two side by side. Renewal decisions run on relationships, inertia, and memory.

The Ledger exists to close that gap. It is a reconstruction of the vendor spending record from state disclosure systems, joined to the public record of student outcomes, at the level where decisions are actually made: the individual district, year by year.

This first issue resolves one state completely. Maryland publishes sixteen consecutive fiscal years of vendor payments for all twenty-four of its school systems. We matched every payment line to a master vendor registry, paired the result with sixteen years of independently measured reading and math outcomes, and put both on a single timeline. Every number in this report traces back to a public source file and can be reproduced on demand.

Every dollar now has to defend itself.

The market context is not in dispute, because the most-read research in the sector has already established it. McKinsey's September 2025 survey of 386 district leaders found total K-12 funding flat at roughly $969 billion while inflation erodes real spending power, federal funding down 22 percent (a $24 billion loss) as ESSER expires, and 34 states facing enrollment declines through 2031.¹

The same survey describes where the pressure lands. Supplemental curriculum, tutoring, assessment, and professional development sit in the categories most vulnerable to cuts. Districts told McKinsey they are "focusing on getting the core curriculum right" because they "no longer have the funds for as many supplemental curricula." And the federal picture adds a structural threat: the proposed K-12 Simplified Funding Program would collapse 18 federal grant programs into a single block grant, a $4.5 billion reduction that would also scramble the program-level signals the market uses to read district budgets.¹

The conversation is global. McKinsey's April 2026 report with the International Education Funders Group frames the same question for education philanthropy: not just more capital, but "improving the effectiveness of existing capital."² From federal block grants to global aid, education finance in 2026 has converged on one question: what did the money actually buy?

McKinsey closes its district survey with advice we agree with completely: districts should concentrate spending on "the highest-quality external providers that offer solid evidence of their results." What that advice needs, and what a survey cannot supply, is the instrument: the record of what each district paid each vendor, next to what happened. That record is what follows.

Maryland, resolved.

Maryland is the first state where the full record can be put on one timeline: every school system, every disclosed vendor payment, every year of measured outcomes, sixteen years deep.

Start with the wallet. Over sixteen fiscal years, Maryland's 24 school systems disclosed just over 1.12 billion dollars in payments to the named instructional and assessment vendors in PILLAR's vendor registry: core curriculum publishers, supplemental and intervention providers, and assessment companies. Exhibit 1 shows the ten largest. Identities are de-identified in this public edition and available, named, inside PILLAR Vertical Intelligence.

Now the flagship question. Take the single largest family on that chart: a top-three literacy and core-curriculum incumbent, present in the Maryland record in all sixteen fiscal years, paid by 21 of the state's 24 school systems, $128 million in disclosed payments. Put that spending on a timeline. Under it, put the state's reading outcomes for grades 3 through 6 on Stanford's national scale, where zero is the national reference point.

Read the shape, not a verdict. Payments to this incumbent step up around the state's curriculum transitions in the mid-2010s, surge through the stimulus years, and reach their highest levels in FY2024 and FY2025, after outcomes had settled well below where they started. Over the same sixteen years, Maryland reading moved from clearly above the national reference (+0.17 at its 2011 peak) to clearly below it (−0.10 in 2025). Math moved further, from +0.19 to −0.20.

The finding is not that any vendor failed. The finding is that a district could pay one vendor family for sixteen consecutive years, through three renewal cycles and one historic decline in outcomes, without anyone on either side of the transaction ever seeing this chart. The villain is the disconnection.

A statewide chart can hide as much as it shows, so The Ledger's standard is to work at least one district completely. Montgomery County is Maryland's largest system, around 160,000 students, and one of only two systems whose disclosure record covers all sixteen years. Its full ledger against this incumbent:

The cuts were real. The retreat was not.

Survey research says supplemental programs sit first in line for cuts. The checkbook confirms it, and then shows something a survey could not.

Splitting Maryland's named instructional spend into three market categories tells the story of the last five budget cycles. Supplemental and intervention spending septupled from its pre-stimulus level to an $80.7 million peak in FY2022, then fell 43 percent in three years. That much is exactly what district leaders described to McKinsey.

What a survey cannot see is where the money went. Total spending on named instructional and assessment vendors did not fall with it. FY2025 is the highest year in the entire sixteen-year record, at $167 million. Core curriculum rose from $60 million in FY2022 to $106 million in FY2025, a 76 percent increase and a sixteen-year high, while assessment rose 77 percent over the same three years. Maryland's post-stimulus adjustment was not a contraction. It was a reallocation, and the bands that gained are the ones districts meant when they said they were focusing on getting the core right.

For a vendor in the middle band, this is the harder version of bad news. A shrinking market recovers when budgets recover. Maryland's budgets did recover, FY2025 set the record, and the money went somewhere else. Winning it back means competing for core adoption dollars, not waiting for a supplemental cycle to turn. For a district leader, the same chart says the consolidation has already happened, on your signature, and the only open question is whether the concentration was chosen on the record or by which contract renewed first.

Every district, on the record.

A statewide chart is an average, and averages are where detail goes to hide. The Ledger's unit of account is the district, so here is every system's sixteen-year ledger on one page: what each paid the named vendor families, by category, and which direction its students moved.

Three things stand out. Prince George's County is the largest wallet in the record at $297 million, though its supplemental share of 37 percent is mid-pack; six systems lean harder on the category. Calvert County bought almost three quarters of its named instructional spend in the supplemental and intervention category, the highest share in Maryland, and remains one of the state's stronger performers. And the four systems whose outcomes improved arrived there on very different spending mixes, which is the honest finding: the mix alone does not decide the outcome, and anyone claiming otherwise from this table is reading past the data.

What the next dollar has to survive.

The record shows what was bought. The balance sheet decides what happens next. PILLAR scores every district's budget pressure from its own public financial filings, and Maryland's twenty-four systems enter the post-stimulus cycle in very different shape.

The index reads four things from public filings: whether a district spends more than it takes in, how much of its revenue was federal stimulus that has now expired, which way enrollment is moving, and how much reserve it holds where that figure is published. Federal audit findings are marked separately. Higher means more pressure. Pressure is not a prediction of failure: Baltimore City carries the state's deepest stimulus exposure, 23 cents of every revenue dollar, and one of its strongest reserve positions at the same time, and the index shows both.

Six systems enter the cycle in the Pressured tier, and four carry open federal audit findings. The four systems whose outcomes improved across the sixteen-year record all sit in the Stable tier. Capacity and results traveled together in Maryland. That is an observation, not a causal claim, and it is one more thing that was invisible until the records were put on the same page.

The Maryland file

Three more entries from the same record, each one already shaping the state's next buying cycle.

Renewal as a decision, not a reflex.

This report was written for the administrator first. The record it reconstructs is your record, and it changes what you can ask.

A superintendent or chief academic officer holding this ledger can put three questions on the table that were previously unanswerable:

What have we actually paid this vendor, in total, over its full tenure? Not this year's license fee. The sixteen-year figure, which in Maryland runs to eight digits for a single family in a single mid-sized system, changes the seriousness of the renewal conversation.

What moved while we paid it? Not "did the vendor cause it," which no fair reading of this data can answer. Simply: over the tenure of this incumbent, did the outcomes we bought it to improve go up or down? If the answer is down for sixteen consecutive years, the burden of evidence in the renewal meeting should shift.

What would we expect a challenger to show? The same chart. Any vendor asking to displace an incumbent should be asked to show its own spend-and-outcome record from districts it already serves. The data to check the claim now exists.

None of this requires a district to buy anything. Maryland's disclosure files are public. SEDA is public. What PILLAR adds is the reconstruction: the master vendor registry that recognizes all 30 spellings of one company's name as one company, the exact district matching, and the standing analyses that keep the record current. The methodology section at the end of this issue is complete enough to check our work.

Consolidation favors whoever can show a record.

If 61 percent of districts consolidate vendors, the winners will be decided by evidence or by inertia. Vendors should prefer evidence, because inertia favors nobody's best product.

The Maryland ledger holds three lessons for the companies serving this market:

Incumbency is measurable now. The flagship incumbent's position, sixteen consecutive years, 21 of 24 systems, is visible in public data. So is everyone else's. Footprint claims, tenure claims, and "trusted by" claims are checkable, which means they are also provable for vendors whose record is real.

The record is a defense as much as an exposure. A vendor whose districts show durable results has, for the first time, third-party public evidence to bring to a consolidation review. The same chart that indicts inertia can defend a product that works. The record cuts both ways, which is what makes it credible.

The reallocation already happened. Exhibit 4 is drawn from disclosures rather than surveys, and it shows the supplemental band down 43 percent from its FY2022 peak while total spending on named vendors reached a sixteen-year high. The dollars did not leave the market; they moved to core and to assessment. A vendor waiting for supplemental budgets to recover is waiting for something that already came and went elsewhere. The contest now is for core adoption dollars, and the districts spending them have a sixteen-year record of what their last incumbent delivered.

Vendor identities are de-identified throughout this public edition. Inside PILLAR Vertical Intelligence, the same analyses return named vendors, district by district, year by year, with the full payment trail. That is deliberate: the public record belongs in public, and the competitive detail belongs to the people doing the work of serving districts well.

How much of the record exists yet.

The Ledger's recurring commitment: each quarter, this index reports how much of the national spend-to-outcome record has been reconstructed, and what remains structurally out of reach.

Reconstruction depends on what states choose to disclose. Six states currently publish district-level vendor payments in usable form. Maryland is the deepest and is fully resolved as of this issue. Utah is next: thirteen years, every district and charter, $26.7 billion, queued as the flagship of Issue No. 002.

One more trajectory belongs in this issue, because it anchors the outcome side of everything above:

How this was built, and what it cannot say.

The Ledger's only durable asset is that its numbers are true and checkable. This section is therefore the most important one in the report.

1. Correlational, not causal. Spend and outcomes are placed on a shared timeline. Nothing in this data isolates any vendor's causal contribution, and no exhibit or sentence in this report claims otherwise. The correct reading is always: outcomes moved X over the period during which Y was the recorded incumbent.
2. Disclosure floors. Maryland discloses payments of $25,000 and above; totals are floors. Utah applies a $1,000 threshold and masks payroll pseudo-vendors. An empty cell means no disclosed payment, not zero dollars.
3. Vendor families are corporate entities. The flagship family aggregates 30 payee spellings including a consulting division and a post-2023 acquisition. The Ledger traces companies, not individual product lines, and says so wherever it matters.
4. A district checkbook under-observes assessment. The largest assessment contracts are signed by state education agencies, not districts, so they never appear in a district payment record. Cambium Assessment, for example, delivers statewide summative testing to roughly a third of the nation's students, and across all 555,880 payment records in six states it receives nothing: every Cambium-family dollar districts pay is intervention or supplemental literacy. The assessment band in every exhibit is therefore a floor on district-purchased assessment, not a measure of the assessment market.
5. First year of record is not selection year. A vendor's first appearance is bounded by each system's disclosure start. Twenty of Maryland's 24 systems enter the record in FY2018–19; only Montgomery and Howard County span all sixteen years individually. Statewide claims use the statewide record; single-district claims use only full-span districts.
6. Outcome gaps are shown, not smoothed. No statewide testing occurred in 2020–21, and Maryland's assessment transitions create gaps at 2014 (partial) and 2018. Lines break at every gap. Grades 3–6 are used because Maryland's grade 7–8 math series is structurally truncated by middle-school algebra course-taking.
7. Matching is exact and documented. Districts are identified by their official federal identifiers through a published reference table (24 of 24 Maryland systems, no ambiguous cases). Vendor names resolve through a documented list of 75 reviewed matching rules; every match records which rule made it. Payments that cannot be matched with certainty are excluded and counted, never guessed.
8. Everything can be reproduced. Each exhibit carries a reference number tied to a saved analysis in PILLAR Vertical Intelligence. The same figures are live in the PILLAR product and were verified against this issue on July 23, 2026. Issue No. 002 repeats and extends this work rather than rebuilding it.
9. What we will not do. No estimated or interpolated figures, no filling gaps with outside guesses, no named vendors in public editions, and no survey data presented as records. Where the record does not exist, the coverage index says so.