Five Texas cities’ reserves coexist with long-term deficits

Current resources and long-term promises use different accounts

The General Fund uses modified accrual for current financial resources. Government-wide statements use full accrual, including obligations that require payment in later years. The Austin policies and Dallas statements explain why adding the two columns would produce no coherent budget measure.

A city can remain liquid while current revenue falls short of the full cost of services. Deferring contributions preserves operating resources today and leaves more funding work later. Long-lived obligations can be financed over time, but the contribution path has to fit recurring revenue.

Austin illustrates the scale of the problem: its $3.116 billion governmental unrestricted deficit accompanies about $2.2 billion of net pension liability and $771.7 million of OPEB liability. Those amounts are not all due next year. They still belong in a decision about a service that will continue next year and thereafter. The relevant bridge is the schedule of contributions and benefits, the recurring revenue available for it, and the other commitments already made against that revenue. Austin FY2025 report.

The tax revenue Austin ISD keeps and its usable budget show why headline collections and resources available for current operations need different measures.

Population rescales the same accounting

Dallas reports about $301 per resident in unassigned General Fund balance and −$3,651 per resident in governmental unrestricted net position. Austin’s figures are about $192 and −$3,108. The statement values use July 1, 2025 Census estimates to scale the cities, and the chart’s panels use different axes.

Per-resident values put differently sized cities on a common scale. They are not individual tax bills. Visitors, commuters and businesses also contribute revenue and use services; a municipal utility can serve customers beyond the city limits. Census July 1, 2025 resident estimates provide the denominator, rather than a modeled service population.

Scroll horizontally to inspect the figure →

The utility can sit in a different place in the accounts

San Antonio reported $527.7 million of utility receipts in governmental funds in FY2025, 18.4% of their revenue. CPS Energy and SAWS are discretely presented component units. Austin’s electric and water utilities are within primary-government business activities; transfers into governmental funds appear as other financing sources rather than in the governmental-revenue denominator used here. San Antonio; Austin.

Tax and franchise revenue is about 53% of San Antonio’s governmental-fund revenue and 78% of Austin’s. San Antonio includes utility receipts in that denominator, while Austin records utility transfers as other financing sources outside it. I’d put those flows on the same basis before calling Austin more tax-dependent. A percentage cannot repair an inconsistent perimeter.

The utility’s sustainable contribution depends on customers, asset maintenance, and covenants. Business-type unrestricted net position is $1.124 billion in Austin, $617.9 million in Fort Worth, and $816.6 million in Houston. Dallas and San Antonio report negative amounts. These balances provide context for a transfer decision, leaving legal authority and future utility needs to be evaluated.

Scroll horizontally to inspect the figure →

Dallas’s contribution gaps totaled $89.7 million

Dallas ERF’s net pension liability fell from about $3.0 billion at December 2023 measurement to $1.9 billion at December 2024. Yet FY2025 contributions were $83.4 million against a $130.6 million actuarial benchmark. Police and fire contributed $205.6 million against $248.1 million. The two gaps total $89.7 million; they compare funding with the actuarial benchmark, without finding a statutory breach.

The reporting framework explains why both statements can be true. GASB’s 2017 guidance distinguishes the assumptions used for financial reporting from those used to set a funding policy. It also says that contribution history matters when projecting future contributions. Accounting and funding are different exercises, but the future cannot be financed merely by writing a more ambitious policy. GASB questions 4.22–4.23.

Dallas ERF’s liability rises from $1.857 billion to $2.506 billion when the discount rate falls one percentage point, while the same report describes an October 2025 contribution phase-in followed by a 30-year amortization policy, showing how improved reported pension stock can coexist with a larger future budget claim. The extra $649 million is a sensitivity, not an immediate bill.

There is a legitimate countercase to treating every unfunded amount as a reason to hoard cash. A durable employer can finance long-lived obligations over time, and a reserve also has a public purpose when it absorbs a temporary shock. The decision turns on whether the contribution path is credible under weaker revenue and returns. Across these plans, different measurement dates, assumptions and employer populations prevent a clean ranking by funding ratio. Houston’s municipal plan also requires an employer-allocation bridge before its full reported liability can be assigned to the primary city. Houston.

Scroll horizontally to inspect the figure →

A revenue decline also needs a decomposition

Governmental-fund revenue fell 1.4% in Austin, 4.5% in Dallas, 2.6% in Houston, and 2.8% in San Antonio from FY2024 to FY2025. Fort Worth rose 2.5%. In every declining city, intergovernmental revenue fell by more than the total, with other categories offsetting some of the loss.

Dallas’s total revenue fell $128.6 million while intergovernmental revenue fell $168.1 million; the remaining lines collectively offset $39.6 million. Austin, Houston and San Antonio show the same direction of offset. A grant expiry can create an acute service problem without implying that the local tax base is shrinking. Conversely, replacing an expiring grant with a permanent local promise can turn a temporary receipt into a continuing burden. The next step is a grant-by-grant bridge to the activities it financed and the revenue proposed to replace it. The aggregate table cannot supply that bridge.

Fort Worth’s total increased by $42.4 million, including $4.5 million more intergovernmental revenue. Its operating scope also changed: MedStar EMS operations transferred to the city on July 1, 2025 and had not previously been part of it. The table reports the published totals; it does not isolate the transfer’s revenue effect. Fort Worth Note A.15, physical PDF pages 87–88.

A decade of growth includes changes in the city itself

Real per-resident governmental revenue rose about 39% in Austin, 24% in Dallas, 19% in Fort Worth, and 15% in San Antonio over FY2015–2025, while Houston fell about 2%. The city histories, CPI, and Census denominators describe the changing reported footprint, including grants and transferred operations.

Some of that apparent growth follows changes in what the city reports. Dallas moved sanitation to enterprise funds in FY2016 and stormwater in FY2024. Its balance history restates FY2023 for stormwater; its flow history does not. Fort Worth’s EMS transfer changes its perimeter too. The population series crosses Census estimate vintages at 2020, which is why the chart breaks there. These are not fussy exceptions to an otherwise clean race. They determine which race the numbers could plausibly describe.

Scroll horizontally to inspect the figure →

Debt comparisons need the same issuer and payment dates

The Texas Bond Review Board reports issuer principal at August 31. City financial statements use their own fiscal ends and carrying-value definitions, which can include premiums, discounts, leases and other obligations. Utility and component-unit boundaries add another difference. A larger issuer total therefore needs an issuer-to-financial-statement bridge before it can support a comparison of city budget burdens. Texas BRB city records.

For FY2025, the cities’ reported governmental debt-service ratios range from 11.1% to 17.9% of their noncapital expenditure denominators. A ratio can move because debt payments change or because the spending denominator changes. The report-specific capital definitions matter too. The download reports these ratios beside the BRB principal series, with their own definitions.

The historical source check found a more immediate problem: Fort Worth’s printed debt-service percentages for FY2016–2023 do not match a direct calculation using the principal, interest/issuance costs, expenditures and capital figures printed in those schedules. For FY2016, the calculation gives 11.57%, compared with 12.41% in the source. Those eight observations remain flagged and are excluded from historical debt-ratio conclusions. FY2015, FY2024 and FY2025 reconcile at the printed precision. Fort Worth report.

The next promise must fit the payment path

A permanent service makes a claim on several future budgets. Its funding competes with debt service, pension and retiree-health contributions, and the capital needs behind any utility transfer. The recurring revenue left after those commitments is the city’s room to make another promise. Temporary grants and a favorable return year can enlarge a reported balance without permanently enlarging that room.

Reserves matter because adjustment takes time. Taxes, staffing and capital programs cannot all change when a forecast changes. A cushion can spare a city from a damaging permanent cut in response to a temporary shock. If the gap endures, however, spending the cushion postpones the revenue or service decision until there is less time to make it.

I’d support a permanent expansion when recurring revenue can pay existing commitments and leave room for the new service. Dallas can have both current reserves and substantial claims on future revenue. The contribution and payment path decides how much room exists between them.