I spent five years on cargo ships, and one thing you learn at sea is the difference between a storm and a leak. A storm gets your attention. Everybody scrambles. A leak just sits there, dripping into the bilge, and if nobody checks the pumps, one morning you notice the deck is closer to the waterline than it used to be.
San Antonio has a leak. If you wanted to understand what climate costs actually look like when they compound quietly instead of arriving all at once, you could do worse than open the mail in Bexar County.
The insurance line
Texas homeowner insurance premiums climbed roughly 57 percent between 2015 and 2023. San Antonio saw a 29 percent jump in just three years, 2021 to 2024. The number of double-digit rate filings submitted to the Texas Department of Insurance has risen 560 percent since 2014.
Here is how the system works in Texas, and I want you to admire this: insurers file a proposed rate increase and start charging it immediately. If the state later decides the increase was unreasonable, they can disapprove it. The bill arrives before the review. Progressive has restricted selling homeowner policies in the state. First Street Foundation estimates 88 percent of properties in the San Antonio metro are at risk of further rate hikes or non-renewal.
The electricity line
CPS Energy is San Antonio's municipal utility. You cannot choose your provider. Rates went up 4.75 percent in 2014, 3.85 percent in 2022, 4.25 percent in 2024. The CFO told reporters he expects to come back every two years through 2030. CPS doesn't make it easy to reconstruct what you were paying a decade ago. Draw your own conclusions about that.
The rate is only part of the bill. Since March 2022, every customer has paid a Winter Storm Uri surcharge to cover fuel costs from the 2021 freeze. Every summer, a Peak Capacity Charge kicks in above 600 kilowatt-hours between June and September. As triple-digit days multiply, air conditioners run longer, more households cross the threshold, and the surcharge eats deeper. CPS itself identified the mechanism: consecutive hot summers forcing higher consumption. The utility enrolls 65,000 low-income customers in bill assistance and plans to grow that to 80,000. Sixty-five thousand people who need help paying for air conditioning tells you something the rate filings don't.
The health line
Bexar County averaged about 318 heat exhaustion cases per year from 2018 to 2021. (Consistent tracking only began in 2022, so even that baseline is short.) In 2023, the number hit 764. Recorded heat strokes rose from zero to 23 over the same period, though whether earlier zeroes reflect absence of illness or absence of tracking is exactly the kind of question this city's data can't answer. San Antonio logged 75 days above 100°F that year.
And the deaths? The Metropolitan Health District's heat dashboard lists mortality as "N/A." Not zero. Not available. Bexar County did not track heat-related deaths. The City Council voted in late 2025 to begin considering a tracking measure. Texas recorded 357 heat deaths statewide in 2023. How many were in Bexar County? N/A.
You can't have a crisis if you don't count the dead. Texas recorded 357 heat deaths statewide in 2023; Bexar County's share reads "N/A."
Eighty-eight percent of San Antonio's population lives in areas where urban heat islands add at least 8°F. The hottest districts, 2, 3, and 5, are on the South and West sides. Predominantly lower-income, predominantly Latino, with less tree canopy. Research has linked historically redlined areas to higher land surface temperatures and more heat-related ER visits. Hotter, poorer, sicker, and paying more for the electricity to survive it. The geography of the bill is never random.
The line nobody carries
San Antonio sits in Flash Flood Alley. About 19 percent of buildings face significant flood risk. Flood insurance penetration in Bexar County remains below 1 percent. In June 2025, the airport recorded 3.98 inches in a single hour, the highest hourly rate since 1934. What that exposure has done to property values in flood-prone and heat-island neighborhoods, nobody can tell you. Bexar County doesn't publish climate-adjusted appraisal data. Another line on a tab nobody's itemizing.
The name on the tab
Earlier this month, a Stanford study published in Nature calculated that U.S. emissions have caused $10.2 trillion in global economic damages, with roughly $3 trillion falling back on Americans. The study explicitly excludes health costs and mortality. It is a floor. And in San Antonio, the data beneath that floor has its own hole: a city that doesn't count its heat dead cannot appear in a national accounting that already excludes health. A floor beneath a floor.
The paper gave a number to what was already arriving in the mail. Insurance up 57 percent. A storm surcharge from 2021 still on every electric bill. Heat exhaustion cases more than doubled. A death count that reads N/A. Flood exposure that 99 percent of residents carry alone.
The tab just kept running, one line at a time, year after year, in a city where nobody declared an emergency because the emergency was the bills.
Things to follow up on...
-
Heat is shrinking lives: A March 2026 study found that climate change has doubled the hours per year that heat makes ordinary activity unsafe, with older adults in the southern U.S. now losing hundreds of hours annually to conditions too dangerous for something as simple as sweeping a floor.
-
Warming just accelerated: A Potsdam Institute analysis confirmed with 98% statistical certainty that global warming has nearly doubled its pace since 2015, reaching 0.35°C per decade and putting the 1.5°C threshold within reach before 2030.
-
Redlined neighborhoods, hotter now: As an early-season heat wave hit Southern California this week, UCLA research showed that emergency room visits during heat events cluster in formerly redlined, lower-income communities with less tree cover, a pattern San Antonio's own data mirrors.
-
Compound drought-heat events surging: Since the early 2000s, events where heatwaves trigger drought conditions have increased nearly eightfold in affected land area, a nonlinear jump that single-variable risk models aren't built to capture.

