The Six-Pillar Audit Framework

Every Local Government Accountability Institute investigation follows a single, replicable framework. We call it the Six-Pillar Audit. It is the operating manual for how the Institute converts citizen concerns and public records into published findings that the median voter can use.

This paper documents that framework in full. It is intended for three audiences: (i) the citizens, neighborhood associations, and watchdog groups who refer communities to LGAI for investigation; (ii) the local journalists, civic leaders, and elected officials who want to apply the framework themselves; and (iii) the donors, counsel, and stakeholders who fund and oversee our work. The methodology is public on the principle that any framework worth applying is a framework worth publishing.

The framework rests on a single conviction: a local government cannot be evaluated by its budget alone. A government can balance its budget while quietly hollowing out its reserves. It can claim record service delivery while quietly tripling its administrative overhead. It can pass a balanced ordinance while quietly steering procurement to its donors. The only way to surface what is actually happening is to look at six dimensions simultaneously, in writing, with the underlying records open for inspection. That is what the Six-Pillar Audit does.

Each pillar produces a defined deliverable. Together, the pillars produce the body of work that LGAI publishes for a community. No pillar is optional, and no pillar is sufficient alone.

The fiscal pillar establishes the trajectory of the government's finances over a multi-year window. The minimum window is five fiscal years; the standard window for an LGAI investigation is ten. The pillar is not about whether the current year's budget is balanced — that is a low bar that most governments can meet on paper. The pillar is about the direction the government is moving in, measured across the data series that bond-rating analysts, forensic accountants, and serious investors actually use.

The full data series and all underlying calculations must be published in a public, machine-readable format. We do not accept the standard local-government practice of burying figures inside a 400-page PDF where they cannot be searched, sorted, or recomputed. If the government in question publishes only PDFs, LGAI extracts the data, normalizes it, and republishes the spreadsheet alongside the source documents. Citizens, journalists, and challengers should be able to reproduce every claim.

Three patterns surface most often in this pillar. First, growth that is structural rather than cyclical: spending increases that staff acknowledge will recur in future budgets, rather than one-time pressures attributable to a hurricane, a pandemic, or a federal grant cycle. Second, debt that is being issued to finance current operations or to mask reserve depletion rather than to fund discrete capital projects with defined revenue offsets. Third, structural deficit projections that assume revenue growth which is no longer plausible given state-level tax reform, demographic trends, or property-tax exemption expansion.

The headcount pillar establishes what the government's workforce actually looks like. It is not enough to know the total number of full-time-equivalent positions. The pillar asks: who are those people, what do they do, what do they earn, and how does the supervisory layer compare to the line-service layer?

The full position-level dataset is published in machine-readable form, with names redacted where state law requires it and exposed where state public-records law permits. Where redactions occur, the rationale is published alongside the data.

Three patterns. First, headcount growth that outpaces population growth and service-demand growth. Second, compensation ratios that have widened materially over the audit window — most often visible as an administrator's compensation rising 4–6x while front-line worker compensation rises 1.5–2x. Third, supervisory layers that have proliferated faster than the line-service workforce, producing manager-to-worker ratios that no peer-quartile private employer would tolerate.

This pillar is the operational core of LGAI's Principle IV — Productivity Parity with the Private Sector. The Gartner Group projects that twenty percent of organizations will eliminate at least half of their current middle-management positions by year-end 2026 through AI-driven flattening. The federal civilian workforce contracted by approximately nine percent in the ten months between January and November 2025. Local governments have, with rare exceptions, not begun this adjustment. The headcount pillar makes that gap visible in numbers.

The demographic pillar separates two questions that local governments routinely conflate: “Is the cost of government rising?” and “Is demand for government services rising?” The two are not the same. Population can grow ten percent over a decade while police calls per capita fall, fire-rescue volume stays flat, and per-capita road-mile maintenance declines. In that scenario, a government that grows its headcount by twenty percent has not kept pace with demand — it has overshot it. The pillar surfaces the difference.

The full per-capita dataset is published. So is the methodology for each metric — for example, whether police calls are defined as dispatched calls, officer-initiated stops, or both. Where state law mandates a particular reporting methodology, that mandate is cited; where local methodology departs from peer practice, that departure is flagged.

The dominant pattern is the “we are growing” justification used to defend administrative expansion that has no demonstrable line-service correlate. A county whose total budget has doubled while population grew thirty percent is not “growing” — it is overshooting growth by a factor of three. The decoupling ratio makes that visible, and it is one of the most powerful tools in this framework for cutting through the rhetorical defense of incumbent spending.

The procurement pillar follows the money out of the government and into the private sector. Every dollar spent by a local government either pays a public employee or pays a private vendor. The headcount pillar covers the employees. This pillar covers the vendors.

The cross-referenced contract-and-contribution dataset is published. We do not characterize the relationships ethically; we publish the records and let voters draw their conclusions. Some relationships are entirely lawful and proper; some are not. The public record is the record. LGAI presents it.

Patterns that surface in this pillar include perpetual sole-source designations on contracts that should be competitively rebid; vendor concentration in single departments that exceeds peer norms; and contribution-to-contract clustering that, while individually lawful, cumulatively suggests a pay-to-play system the public has a right to see.

The technology pillar measures the gap between the productivity tools the government in question is using and the productivity tools that the private sector, the federal government, and peer state and local governments are using. It is the operational pillar that connects most directly to Principle IV of the LGAI Guiding Principles.

Every comparator cited is hyperlinked to its source. The methodology used to project savings for the local government under audit is published. Citizens are entitled to reproduce the calculation; elected officials are invited to dispute it on the merits.

The dominant pattern is the absence of any meaningful automation strategy in a budget environment where automation has become standard practice at the state and city level. A local government that has not commissioned an AI-readiness audit, has not piloted any automation in permitting or financial operations, and has no published technology strategy beyond replacing aging hardware is, in 2026, a government that has chosen to fall behind the productivity curve. That choice has a cost. The technology pillar quantifies the cost.

The political and disclosure pillar establishes the connection between the decisions made by the governing body and the donors, lobbyists, and interests that fund the decision-makers. It is the most legally sensitive pillar; it is also the one that voters most directly recognize as accountability.

We publish the cross-referenced record. We do not allege misconduct on the basis of contribution-to-vote patterns alone, because lawful donors are entitled to vote for candidates who share their views, and lawful candidates are entitled to vote in ways their donors approve of. What we do — and what we believe voters are entitled to — is the side-by-side record. The voter draws the conclusion.

Patterns that surface in this pillar include systematic non-attendance during major fiscal decisions; recusals that occur rarely despite frequent contribution-to-vote overlap; and voting patterns where the contributor-aligned position prevails on close votes within a particular sector (most often development and procurement).

Each pillar produces a deliverable: a sourced narrative, a normalized dataset, a side-by-side cross-reference table, and a public dashboard. Together, the six deliverables constitute a community investigation file.

A community investigation file produces three published artifacts:

1. The Findings Document — a sourced narrative of approximately five to fifteen thousand words, written for the median voter, that summarizes what the records show. Every factual claim is footnoted to a primary public record. Pre-publication, the document is reviewed by a second LGAI researcher; the named officials in the document are notified and given a ten-business-day window to submit factual corrections.
2. The Public Dashboard — a permanent, searchable, machine-readable presentation of the underlying data. Citizens, journalists, and challengers can replicate every calculation in the Findings Document.
3. The Citizen Briefing — a private, structured briefing delivered to the citizens or organization who referred the community to LGAI. What they do with the briefing is their decision; LGAI does not coordinate political activity, and the briefing is not a campaign deliverable.

Each artifact has a permanent home on a dedicated community research site. Updates are published as new records become available. Corrections, when they occur, are logged transparently with the date, the original claim, the corrected claim, and the source of the correction.

A common objection from local officials, and occasionally from sympathetic journalists, is that the framework is too comprehensive — that LGAI should focus narrowly on the budget, or narrowly on procurement, and let voters sort the rest. We reject that framing for a simple reason: every prior generation of local-government oversight has done exactly that, and the result is the present condition. A budget-only focus produced governments that balanced their general funds while quietly tripling their off-budget debt. A procurement-only focus produced governments that competitively bid every contract while quietly steering the underlying specifications to a preferred vendor. The six pillars exist because anything less than six pillars produces blind spots.

The framework is also designed to be replicable by any citizen group with the discipline to apply it. The pillars are not proprietary. The methodology is publicly documented in this paper. The data sources are public records. The output formats — sourced narrative, machine-readable dataset, public dashboard — are standard. LGAI's role is not to monopolize this work; it is to demonstrate it at sufficient quality and scale that citizens in other communities can demand the same of their elected officials, or do the work themselves.

Every pillar in the framework derives from a single principle, articulated in the Institute's Guiding Principles: every elected official, every appointed administrator, every public employee carries a stewardship obligation that exceeds anything required of a private employee, because public funds are taken under threat of legal compulsion from people who had no choice in the matter. The standard of care owed to those funds is therefore higher — not lower — than the standard a private fiduciary owes to a willing investor.

The Six-Pillar Audit Framework is what enforcement of that standard looks like in practice. The records exist. The methodology is public. The pillars are not negotiable. Every local government in the United States can be audited under this framework. LGAI begins with the communities that ask for it; the rest is a question of time, resources, and the discipline to keep publishing what the records show.

The government works for the people. When it forgets that — and most local governments have — it falls to the people to remind it.

The Six-Pillar Framework is not proprietary to LGAI. It is fully public, fully documented, and intended for use by any citizen group, journalist, or local official with the discipline to apply it. The steps below describe how to begin in your own community.