Protect Our Power SA Protect Our Power San Antonio

Sources & Methodology

Every figure on this site, where it came from, and how the comparison is calculated.

Why this page exists

An argument about public money should be checkable. Every number on this site traces to a primary document, and the comparison tool is built only from published tariffs and the state’s own offer data — no rate aggregator or affiliate site is used as a price source.

Claims across the site carry a Source link that lands on the exact section below, so you never have to guess which document a figure came from.

City budget

  • City of San Antonio budget and the OpenGov transparency portal — FY 2025 actuals for all revenue, expense, and departmental figures.

Supports: the $492,160,673 CPS Energy transfer and its 29.88% share of the General Fund; General Fund revenue of $1,647,012,323 and spending of $1,657,319,363; public safety at 62.04% of spending, including police at $609,081,560 and fire at $397,685,356; and the reduction scenarios on the City Budget page.

The cut scenarios are straight-line arithmetic, not a forecast. They reduce only the CPS line, hold every other revenue source constant, and allocate the shortfall pro-rata by each category’s share of spending. Real budgets have fixed costs, contracts and collective bargaining agreements that this does not model.

Electricity rates and offers

What customers actually paid

  • EIA Form 861, Sales to Ultimate Customers, 2024 — every utility’s residential revenue, sales and customer count, filed with the federal government.

Realized price is total revenue divided by electricity sold. It measures what customers were billed, not what was advertised: it includes the majority who never shop, contracts that lapsed onto a month-to-month holdover rate, variable plans, and fees.

Supports: competitive retailers at 16.66¢/kWh across 6,963,807 households and CPS Energy at 12.43¢/kWh across 865,914; the 4.23¢/kWh gap, which is 34% and about $571 a year at the Texas-average 13,508 kWh; the Texas residential average of 14.94¢; municipals at 12.55¢, co-ops at 12.68¢ and investor-owned at 13.33¢; and every per-retailer figure on The Retailers page, including Reliant at 17.38¢ and TXU at 18.70¢ and their combined 43% share.

The dollar figure moves with usage, so it is always computed rather than stored — the homepage prices the gap at the Texas-average household, the calculator at whichever home size or set of real bills you give it.

What deregulation has cost Texans

Supports the $28 billion figure. The Journal analysed U.S. Energy Information Administration data and found that Texans in the deregulated retail market paid $28 billion more for power from 2004 through 2019 than they would have at the rates charged to customers of the state’s traditional regulated utilities. Over the same period, traditional Texas utilities’ rates averaged 8% below the national average while retail providers’ rates averaged 13% above it.

This is the one figure on the site taken from journalism rather than from a document we can re-derive ourselves. It is a newspaper’s analysis of federal data, not a government publication, and the article is paywalled. We cite it because the underlying EIA data is public and the finding is consistent with what our own Form 861 work shows for 2024 — but it is a secondary source, and it is labelled as one. The comparison is against Texas’s own regulated utilities, not against a national average or a hypothetical.

How much electricity a household uses

The comparison tool asks for a home size rather than a kilowatt-hour figure, and turns that into a real twelve-month usage curve. Two independent federal sources build it.

  • EIA Residential Energy Consumption Survey, 2020 — public microdata, restricted to the 1,016 surveyed Texas households and grouped by total energy-consuming floor area. Annual kWh per size band is the survey-weighted mean of metered household electricity use. RECS measures consumption; it does not model it.
  • EIA Form 861M — monthly Texas residential sales divided by monthly Texas residential customers, averaged across 2021–2025, gives the share of the year billed in each calendar month. Form 861M reports sales as billed, which is the same basis both tariffs in the comparison use, so no shift is applied.

The two agree to within half a percent on the Texas average household: 13,440 kWh a year from RECS, 13,508 from Form 861M. That agreement is the check that the size bands are the right magnitude; if a future revision moves either figure materially, the bands should be re-derived rather than adjusted.

These are statewide figures used for a San Antonio question. CPS Energy’s own residential customers averaged 12,845 kWh in 2024 — 11,123,381 MWh across 865,914 customers, from EIA Form 861 — about 2% below the Texas average. Close enough to stand in, and disclosed rather than assumed. A household that enters its own twelve bills does not rely on any of this.

Who generates Texas electricity

  • EIA Form 923, Schedules 2/3/4/5, Page 1 — 2024 Final Revision. Plant-level net generation with the operating company named. Texas totals 566,502,688 MWh across 388 operators.

Supports: CPS Energy as the third-largest generating operator in Texas at 22,266,364 MWh and 3.93% of state generation; Vistra at 12.07%, NRG at 6.58%, Constellation at 2.53% and NextEra at 0.82%; and the finding that 39 of the 49 retailers selling on Power to Choose own no generation at all.

Generation shares are aggregated from that file by operator, then rolled up to parent company for the retailers whose ownership is a matter of public record (TXU and Ambit to Vistra; Reliant, Green Mountain, Cirro and Discount Power to NRG; Gexa to NextEra). The retail provider list is the PUCT Power to Choose export.

We used the 2024 Final release rather than the 2025 file, which EIA publishes as an early release carrying the instruction “not fully edited, use with caution, do not aggregate to state, regional, or national totals.” Aggregating it to a state total is exactly what this statistic would require, so it was not used.

One caveat on attribution: CPS Energy also holds a 40% ownership stake in the South Texas Project nuclear station, which appears separately under its own operating company. The 3.93% figure counts only generation CPS operates directly, so its economic share of Texas generation is somewhat higher than the chart shows.

Who is licensed to sell electricity

  • PUCT Directory of Retail Electric Providers — the state’s register of certificated providers. Each provider’s report gives its certificate number, the date it was certificated, its service area, its registered trade names, and the contact a complaint goes to.

The certificate link on each row of The Retailers table goes to that provider’s own record. A retailer is linked only where its EIA name matches a certificate unambiguously once legal suffixes are stripped; the rest link to the directory search, because a link that lands you on a different company’s certificate is worse than one that makes you type a name.

Campaign finance

  • Texas Ethics Commission — the complete bulk export of electronically filed campaign finance reports, covering filings from July 2000 onward. This is the primary record, not a republication of it.

The figures on The Retailers are built from that export under one rule: a political committee is credited to a company only where the committee’s own registered name establishes the connection. Nothing is inferred from a shared word, a shared address, or a treasurer’s employer, because a misattributed political donation would discredit every other number on this site. Every figure is published alongside the committee name exactly as filed and its TEC filer number, so the same record can be pulled and checked.

Two adjustments are worth knowing about. Reports the TEC marks as superseded by a later amendment are excluded, because the filings contain both versions and summing them would count an amended report twice. And the “to candidates” totals count only expenditures the committee itself categorised as contributions to a candidate, officeholder or committee — filings that predate those categories are not counted there, so those totals are a floor rather than a full accounting.

What CPS Energy returns to San Antonio

Supports: 3,654 employees including about 1,226 skilled craft, at 12.4 years’ average tenure; pension-covered payroll of $308,095,000 and the employer contribution of $72,774,000; the STEP efficiency programme at $57,094,000; the Affordability Discount and REAP figures; other payments to the City of $13,031,000; the $25.9 million returned from the City to CPS under the 2024 memorandum of understanding; and the bond-ordinance cap holding all City payments to 14% of gross revenues.

The ~$84,000 average wage is derived, not published: pension-covered payroll for the year ended 31 January 2025 divided by the 31 October 2025 headcount. The two figures carry different dates, and covered payroll is a pension-plan measure that need not equal total compensation, so treat it as an approximation of scale rather than a salary.

The proposal

  • San Antonio Report — coverage of the August 2026 announcement.
  • KSAT — the proposal and the 10–20% savings claim.

How the comparison is calculated

A counterfactual bill model. One identical kWh profile is priced under the CPS Energy residential tariff and under every qualifying competitive offer, and the annual totals are compared. Advertised cents-per-kWh headlines are never compared directly — they are only ever inputs to a bill.

  • CPS side. Service availability charge, energy charge, the summer peak capacity charge on usage above 600 kWh in June–September, and the actual monthly fuel and regulatory adjustment factor for the residential class.
  • Competitive side. Power to Choose publishes each offer’s all-in average price at exactly 500, 1,000 and 2,000 kWh. Those are converted to three bill totals and the underlying tariff is recovered from them, so any usage level can be priced. Bill credits are modeled as real thresholds; the small number of offers that are not two-part tariffs at all are priced by interpolation through the published points.
  • Excluded. Variable and indexed plans, and municipal pass-through fees on the CPS bill.

Known limits

  • Published offer prices are rounded for display, which puts a floor of roughly $0.20 a month on the accuracy of any reconstructed tariff. That is far below the differences being reported, but it is not zero.
  • The tool prices currently advertised offers to new customers. That is the best case for competition. It is not what a household pays after a contract lapses onto a holdover rate, which is a common and much worse outcome.
  • A home size is a state average applied to one house. It gets the magnitude and the seasonal shape right and the particular house wrong; entering twelve real bills removes that limit entirely.
  • The same seasonal curve is applied to every home size. Larger homes are probably somewhat more summer-weighted than smaller ones, but no primary source publishes a monthly shape broken out by house size, so none is invented.
  • Offer data is a snapshot. Rates change constantly; the collection date is shown with the results.

Found an error? That matters more to us than the argument does. Tell us — say which figure, on which page, and where the right one comes from, and it gets checked and corrected.