A collaboration between Lewis McLain & AI

Before you read this, consider these facts:

The Net Audit Adjustments have always been positive to local governments for the last 25 years and have totaled $2,102,648,997. That is good news! The reasons could be many, including the efforts of the Audit Division. We don’t know the breakdown.

The total Service Fees the Comptroller has collected from local governments for the last 25 years is $4,717,757,160. We don’t know the degree to which that covers or exceeds the cost of provide the services to local governments. We just know it is a ton of money.

On August 1, Don Huffines becomes Texas Comptroller of Public Accounts, appointed by Governor Abbott to fill the vacancy created by Kelly Hancock’s resignation. Huffines is also the Republican nominee for the office in November, where he faces State Senator Sarah Eckhardt of Austin. In an interview with CBS News Texas shortly before taking office, he described his intended reorganization of the agency: he would grow the audit staff and shift “a lot” of the office’s roughly 600 auditors away from auditing businesses for revenue and toward auditing government, in the service of a program he has branded “DOGE Texas.”

The premise is not controversial. Fraud, waste, and abuse in government are real, they are corrosive, and rooting them out is a legitimate and valuable public purpose. The controversy, to the extent there is one, lies entirely in method and structure — in what kind of audit is actually being proposed, by whom, under what authority, and at what cost to the functions the office already performs. On each of those questions, the plan as described raises concerns that have less to do with partisanship than with the settled principles of public-sector auditing and internal control.

The Connotation of the Brand

Branding is a reasonable place to begin, because Huffines chose his brand deliberately. He did not campaign on adopting Government Auditing Standards or strengthening the office’s forecasting. He campaigned on “DOGE-ing” Texas — an explicit reference to the federal Department of Government Efficiency, created by executive order in January 2025, led for a period by Elon Musk, and formally terminated on July 4, 2026.

By the time it closed, DOGE had accumulated a record against which the brand can now be measured. It had set out to cut $2 trillion in federal spending; its final public tally claimed approximately $215 billion, or roughly eleven cents on each dollar targeted, and even that figure has not withstood scrutiny. Independent reviewers documented duplicate entries, arithmetic errors, and a pattern of counting contract “ceiling” values rather than the amounts the government actually expected to spend.

The Government Accountability Office reported that it could not determine how much was ultimately saved or lost. Personnel reductions in 2025 were followed by substantial rehiring in 2026; at the Department of Health and Human Services, projected new hires exceeded the prior year’s layoffs. And when the House Appropriations Committee requested a comprehensive accounting, OMB Director Russell Vought replied that the administration had “no plans to do kind of a closing DOGE report.”

The final point is the most instructive. An initiative founded on the premise of accountability declined, in the end, to account for itself. By the ordinary standards of the field — where an engagement culminates in a report whose figures can be independently reconciled — an effort that produces no such report is not an audit in any meaningful sense. It is an exercise in publicity. When the incoming chief financial officer of one of the world’s largest economies adopts that effort as his organizing model, it is fair to ask which of its features he intends to replicate.

A Record Weighted Toward Rhetoric

An official’s prior public statements are legitimate evidence of how that official is likely to exercise discretion, particularly when the office in question controls revenue estimation, the treasury, and tax administration. Huffines’ record supplies several relevant data points.

At an April 2026 Texas Tribune event, he indicated he would withhold state funds from agencies or universities he judged to be engaged in diversity, equity, and inclusion activities; asked how he would define a qualifying violation, he said he would study the applicable laws and, if the matter remained unclear, consult the Legislature.

The sequence is notable: the commitment to enforcement preceded the definition of the offense. That ordering is difficult to reconcile with the role of a comptroller, who is a largely ministerial officer charged with executing appropriations made by the Legislature. Conditioning the release of appropriated funds on an official’s own evolving policy judgments is not auditing; it is a form of governance by disbursement — a practice conservatives have, in other contexts, criticized when federal agencies engaged in it.

More broadly, Huffines’ public posture has favored the vocabulary of confrontation over that of stewardship — the “bully pulpit,” the pledge to “expose,” the framing of fiscal questions through prevailing culture-war concerns. Texas Monthly reported that his 2022 gubernatorial rhetoric was sufficiently combative to render him unwelcome in much of the state’s Republican establishment.

None of this is disqualifying as a matter of politics. It is, however, a poor predictor of audit quality, because auditing is a discipline in which restraint, documentation, and the willingness to qualify one’s own conclusions are the core professional virtues. An audit program whose conclusions — pervasive fraud, savings sufficient to fund tax relief, vindication of a political movement — are announced in advance has inverted the analytical process before the first workpaper is opened.

The clearest single illustration of that tendency is an earlier statement, one that shows how Huffines weighs competing costs. In an April 2020 op-ed opposing the state’s pandemic restrictions, he argued that the Governor had “effectively shut down the 10th-largest economy in the world.” His case rested on a specific calculation: at that point 318 Texans had died out of some 29 million residents, which he expressed as a fatality rate of roughly 0.0000109 percent and characterized as too small to justify the shutdown, adding that unemployment and financial ruin were the more consequential harms.

Whatever one’s view of the reopening debate, the analytical maneuver is worth noting: a human death toll was reduced to a decimal small enough to discount, so that the quantifiable economic figure on the other side of the ledger would prevail. A disposition to treat the measurable dollar as more real than the harder-to-quantify harm is a consequential one to carry into an office whose central task is determining which costs are to count.

A Judgment Rendered at the Foot of the Curve

The pandemic statement warrants closer examination on its own, because its timing — not merely the sentiment behind it — is what makes it relevant to the office Huffines is about to hold. Placed on a timeline, the figure he dismissed was not a settled toll but one of the earliest readings on a curve that had barely begun to climb.

The chronology is unambiguous. Texas recorded its first death associated with COVID-19 on March 17, 2020, in Matagorda County. The op-ed appeared roughly four weeks later, in the middle of April, and the 318 cumulative deaths it cited represented, in effect, the first month of the state’s experience with the disease. Whatever that number was, it was not an endpoint. It was a starting point.

What followed establishes how premature the judgment was. Daily deaths in Texas did not crest in the spring of 2020; they continued upward. The state absorbed a summer surge severe enough that, on July 27, 2020, a change in reporting methodology added more than 400 previously uncounted deaths in a single day. The peak in daily deaths did not arrive until January 2021 — roughly nine months after the op-ed — when the state was recording on the order of 352 deaths per day. And the cumulative toll kept compounding well beyond that: by May 2023, Texas had recorded approximately 92,400 deaths attributed to COVID-19. The 318 figure that Huffines characterized as too small to justify concern ultimately grew to roughly 290 times that size. The Governor was aware of the steep curve already happening real time. Mr. Huffines took the approach that lives are expendable. Just statistics.

The significance of this is not epidemiological. It is methodological, and it speaks directly to the aptitude the Comptroller’s office actually requires. Huffines did not weigh a completed death toll against a known economic cost and arrive at a contestable conclusion; that would be an ordinary policy disagreement, and reasonable people reached different conclusions about reopening. What he did instead was declare a still-rising figure insignificant at the exact moment when the only defensible posture was uncertainty about where the line was heading — and the line then rose for nine more months and ended orders of magnitude higher. The failure was one of a faulty method rather than values: an early data point was treated as though it were a final one, and a provisional number was converted into a definitive judgment.

That habit is a specific and relevant liability in this particular office. The Comptroller’s defining statutory functions are revenue estimation and the projection of future spending — disciplines that depend entirely on the opposite instinct, on the professional discipline to treat an early reading as provisional and to withhold conclusions until a trend is genuinely established. A demonstrated willingness to mistake the foot of a curve for its endpoint is precisely the tendency a forecaster cannot afford, and it appears here in the public record, applied to a curve whose eventual shape is now fully known.

Three Distinct Activities Called “Audit”

The central conceptual difficulty in the plan is that the word “audit” conceals three different activities, and the program conflates them. Distinguishing them clarifies what is actually being proposed.

A financial audit addresses a narrow question: whether an entity’s financial statements fairly present its position in conformity with applicable standards. It is recurring, it is conducted by certified public accountants under generally accepted auditing standards, and it yields an opinion. Every municipality, county, and school district in Texas already undergoes such an audit annually, performed by an independent firm.

A compliance or performance audit examines whether funds were spent lawfully and whether a program achieved its objectives economically and effectively. For state agencies, this function resides principally in the State Auditor’s Office, supplemented by single audits wherever federal funds are involved.

A fraud examination, sometimes termed a forensic audit, is categorically different. It is neither recurring nor exploratory. It proceeds from predication — a specific, articulable basis for believing that a particular wrongdoing has occurred. The examiner does not audit “government” in the abstract; he investigates a defined allegation, traces an identified series of transactions, preserves a chain of custody over evidence, and assembles a file capable of supporting prosecution. The work resembles criminal investigation more than accountancy, and its terminus is a courtroom rather than a management letter.

Huffines is describing the third activity. He has stated that the objective is to find fraud, that a matter “could be a criminal case,” and that pursuing it requires “a real, distinct process.” Those characterizations are accurate as far as they go.

But it is precisely the investigative and predicated nature of fraud work that makes it unsuited to deployment as a standing force. A body of 600 auditors directed at “government” as a general category, and instructed to locate wrongdoing, is not conducting fraud examinations; predication is the very feature that distinguishes an examination from a dragnet. An examiner who begins from the conclusion that extensive fraud exists and must be proven has abandoned the method, and investigations that begin with their conclusions tend to produce them.

This is the structural error at the center of “DOGE Texas”: it is promoted with the moral urgency of fraud enforcement — criminality, prosecution, incarceration — while being organized at the scale and permanence of routine auditing. Genuine fraud work is narrow, evidence-driven, and comparatively rare. A program that is broad, conclusion-driven, and continuous is a different instrument altogether, and calling it fraud enforcement does not make it so.

The Independence Problem: Auditing the Auditor

The most serious objection is structural, and it is visible simply from the organization of Texas government.

The state already maintains a State Auditor’s Office, which audits state agencies for financial accuracy, compliance, and performance. Critically, that office reports to the Legislature through the Legislative Audit Committee — not to the Governor and not to the Comptroller. The reporting line is deliberate. It embodies the foundational principle of the field: that those who spend and hold public funds must be examined by a party that does neither. Independence of the auditor from the audited is not a refinement of the practice; it is the practice.

The Comptroller occupies the opposite position. As the state’s chief financial officer, the Comptroller estimates revenue, certifies the budget, holds the treasury, administers and collects taxes, and maintains the accounts. In the terminology of internal control, the office performs the custody and recording functions for Texas government as a whole. It is, therefore, among the principal subjects that an independent auditor exists to examine.

Two questions follow. First, what would a Comptroller-run audit of state government accomplish that the State Auditor’s Office does not already perform? And under what authority, given that the Legislature has assigned agency oversight to that independent office and, as reported during the primary, has over time reduced the Comptroller’s own audit powers over other governments? A parallel audit function situated within the CFO’s office does not close a gap in oversight; it duplicates an existing function while removing the independence that gave the original its legitimacy.

Second, and more fundamentally: the proposal has the chief financial officer auditing the government of which his own office is the largest financial component. An official who records the transactions, holds the assets, and estimates the revenue, and who then also renders judgment on how those funds were spent, is not exercising oversight; he is examining his own work.

This contravenes the most elementary rule of internal control — that the party responsible for custody and recording of assets cannot also be the party that independently examines them. The concentration of custody, recording, and audit authority in a single elected officer is not an advance in fraud prevention; it is precisely the control deficiency that fraud-prevention frameworks are designed to eliminate.

In this light, Huffines’ statement that he is “very much aligned” with the Governor who appointed him is not reassurance but disclosure: independence is asserted, not alignment.

The Direction of Scrutiny: Aligned With the State, Aimed at the Localities

If alignment with the appointing Governor is the impairment, the question it poses is directional: on whom will an aligned auditor actually train the apparatus? The plan supplies an answer, and it is not the state government to which the Comptroller himself belongs. It is local government.

This is stated in plain terms. In describing the reorganization, Huffines identified his targets as the federal programs the state administers, the state agencies, and — explicitly — local governments. Local entities are not a peripheral consideration in the program; they are named among its objects.

It is true that all local governments are subdivisions of the state, but how can a department of the state anoint itself to be ruler of those subdivisions without Legislative authority?

His campaign materials sharpen the emphasis, pledging to “shine a light on local spending and regulations” and casting local government as a principal driver of the housing costs he proposes to attack. The theme ran across the primary field: rival candidates likewise promised to scrutinize local governments and to bring cities and school districts into line with state law. One would have to be completely on another planet to not see the encroachment on local government to function as they were designed to serve the local populace.

The structure of the plan makes the local orientation not merely one element but the organizing one. The stated purpose of auditing government is to generate property-tax relief. Texas, however, levies no state property tax; property taxes are imposed entirely by local entities — cities, counties, school districts, and special districts. A program that promises to fund tax relief by auditing “government” is therefore, by construction, a program directed at local government spending. The local target is not incidental to the property-tax rationale; it is entailed by it.

Placed beside the alignment disclosure, this resolves into a single coherent asymmetry. An auditor aligned with state leadership is structurally disinclined to direct scrutiny upward — at the state agencies, and at the priorities of the administration that appointed him, including the school-voucher program the Comptroller’s own office is charged with administering. The path of least institutional resistance runs downward, toward the units of government with which state leadership has no political stake in solidarity.

Independence exists precisely to prevent audit authority from flowing along lines of political alignment in this manner; its absence is what permits the flow. The alignment statement and the local-government target are thus two aspects of one arrangement: the first explains why the audits are unlikely to point at the state, and the second identifies where they will point instead.

Two qualifications preserve accuracy without disturbing the conclusion. The first concerns fairness to the enterprise itself: local governments are not, and should not be, exempt from examination; they expend public funds and are audited annually for exactly that reason. The objection here is narrower and structural — that the direction of scrutiny is being determined by political alignment rather than by demonstrated risk, which is the only proper basis on which finite audit resources are allocated.

The second concerns authority. The Comptroller’s formal power over local governments is limited; as reported during the primary, the Legislature has over time reduced the office’s general audit powers over other governments, so it cannot audit a municipality at will. Or can it if the Governor and Legislature has plans to not only provide the authority but also the specific instructions?

Yet limitation is not impotence. The office retains substantial local-facing instruments — the School District Property Value Study conducted under Government Code Section 403.302, the local sales-tax data reported under Tax Code Section 321.3022, and the Truth-in-Taxation machinery through which local rates are calculated and disclosed. An officer inclined to press local governments has footholds independent of any general audit authority. The operative gap, then, lies between the breadth of the stated ambition and the narrowness of the lawful authority — and such gaps have historically been closed through pressure and publicity rather than through statute.

The Opportunity Cost of Reassignment

Beyond the structural concerns, the reassignment carries a direct fiscal cost that the announcement does not address.

The Comptroller’s audit division exists primarily to audit taxpayers — for sales, franchise, motor fuels, and mixed-beverage taxes, among others — for revenue compliance. This function is unusual within state government in that it reliably returns a multiple of its cost. A productive tax auditor generates assessments substantially exceeding salary, and, at least as importantly, the visible presence of audit coverage sustains voluntary compliance among the far larger population of taxpayers who are never audited. Compliance is, in significant part, a function of perceived enforcement.

Redirecting a substantial share of that staff toward auditing government therefore thins the enforcement presence and, with it, the deterrent — a cost that appears in no press release because it materializes as revenue that is quietly never collected. The reassignment moves skilled personnel from the one activity in the agency with a demonstrable, measurable yield to a mission the office lacks clear statutory authority to conduct. Characterized accurately, that is not an efficiency gain; on the office’s own accounts, it is a net loss.

This is not to suggest that existing arrangements are beyond improvement. A motivated Comptroller could do genuinely useful work in sharper revenue forecasting, in expanded data transparency, and in rigorous analysis of the cost drivers behind property tax burdens. But the imperfection of the current system is an argument for strengthening the established, standards-based machinery, not for constructing a politically branded parallel apparatus modeled on a federal effort that concluded without issuing an accounting.

Principles of a Methodical Alternative

A serious program against fraud and waste would be organized around several principles that the “DOGE Texas” model, as described, does not reflect.

Predication and risk should precede investigation. Fraud work begins with a documented reason to look. Programs should be prioritized by dollar volume, control-weakness history, and prior findings, and inquiry should follow evidence rather than publicity.

Professional standards should govern findings. Government Auditing Standards exist so that conclusions survive adversarial scrutiny. A methodology that treats contract ceilings as realized savings, or that double-counts cancellations — the documented DOGE pattern — does not produce findings; it produces assertions.

The three audit functions should not be conflated. A financial audit yields an opinion, a performance audit yields recommendations, and a fraud examination yields evidence for prosecution. Importing the investigative urgency of the third into the routine scope of the first two produces an instrument that discharges none of the three functions well.

Fraud belongs with prosecutors. Where genuine fraud is identified, the appropriate destination is the justice system. Huffines’ own record supplies the correct template: his role in exposing the Dallas County Schools scandal followed evidence through a legal process and concluded with criminal convictions and dissolution of the agency. That outcome was achieved because the matter was predicated, documented, and referred — not because it was announced.

Waste and disagreement should not be recharacterized as fraud. Waste is a management failure, properly addressed through management letters and the budget process. Spending with which one disagrees — typically the largest category in any inventory of purported “waste” — is a matter for the Legislature, which holds the appropriations power. Neither is criminal, and presenting either as fraud diminishes the term for the cases that genuinely warrant it. Doesn’t that somewhat describe the problem with the Federal DOGE program?

Every effort should conclude with a reconcilable accounting. The measure of an accountability initiative is whether it publishes figures that an independent, even hostile, reviewer could verify. DOGE conspicuously failed that test. Any state analogue should be evaluated, from the outset, by its commitment to meeting it.

Conclusion

The office Huffines is assuming is a significant one: Texas spends more than $500 million per day, and a rigorous, independent, and professionally skeptical examination of that spending would serve every taxpayer in the state.

But the value of such an examination is measured in reconciled dollars, not in public attention, and the model Huffines has chosen as his template promised $2 trillion, claimed $215 billion, could not substantiate the claim, and declined to issue a final report. To the extent “DOGE Texas” imports that method, the predictable result is the federal pattern reproduced at the state level: a conspicuous beginning, a contested middle, and no accounting at the end.

The stronger course would be quiet, predicated, standards-based work of the kind that actually sent six Dallas County Schools officials to prison — which is to say, ordinary auditing performed well, a practice considerably older than any initiative named for an internet meme. And it is work that, by the logic of internal control, cannot properly be performed by the same officer who keeps the state’s books. Fraud and abuse are real problems that deserve a real remedy. Pursuing them recklessly, or pursuing them through the very office that ought itself to be audited, is the surest way to discredit the effort — and the cause — for a generation.