The debate behind The Holiday Heat Test Cities Are Still Failing

Advocate A’s weakest move is treating retrospective mortality accounting and emergency activations as proof of siloed non-adaptation. That framing confuses two different functions. Excess mortality—the statistical estimate of deaths above an expected baseline—is inherently retrospective because it depends on comparing observed deaths with modeled counterfactual deaths. But operational heat response does not wait for final death certificates: it uses forecasts, alerts, emergency-department syndromic data, 911/EMS signals, outreach lists, and utility coordination. In the U.S., CDC’s Heat & Health Tracker uses near-real-time emergency department data from the National Syndromic Surveillance Program, and CDC says communities can use these local heat-health data to prepare and respond during extreme heat events. That directly rebuts A’s claim that inability to count final deaths in week one means inability to allocate resources in week two. (cdc.gov)

A is right that gaps remain, and the Colorado-Utah firefighter deaths are severe evidence that systems can be overwhelmed under red flag conditions—weather warnings indicating warm temperatures, low humidity and strong winds that can create rapid fire spread. But a fatal burnover does not by itself show no pre-positioning or no compound-risk planning; it shows that planning has limits when fire weather, terrain, wind shifts and fuel conditions interact. The better question is whether agencies are building compound-risk architecture faster than critics acknowledge. On that, the evidence is stronger for adaptation than A allows.

First, heat warning and response systems are becoming more integrated. CDC and NOAA launched a national HeatRisk Forecast Tool providing a seven-day forecast of heat levels that may harm health, supplementing traditional heat advisories with health-relevant risk categories. (cdc.gov) New York City’s May 2026 Heat Emergency Plan is explicitly multi-agency: it opens cooling centers, expands outreach to older adults, people with chronic illness and unhoused people, uses targeted alerts through the Advanced Warning System, monitors health impacts in real time, coordinates with utilities, and works with healthcare and community organizations. A single Chicago weekend gap is real, but it is not representative proof that cooling-center systems are static. (nyc.gov)

Second, Europe is scaling wildfire capacity before peak season, not merely reacting afterward. The European Commission announced its largest-ever 2026 wildfire response, including continuous risk forecasts through EFFIS, Copernicus emergency mapping, a new Cyprus regional firefighting station able to pre-position six aircraft, and nearly 800 firefighters pre-positioned in high-risk countries. (civil-protection-humanitarian-aid.ec.europa.eu)

Third, grid planning increasingly addresses compound summer stressors. FERC’s 2026 summer assessment explicitly evaluates high temperatures, drought, low snowpack, wildfire and hydropower constraints; it finds adequate reserves under normal conditions while identifying specific regions needing mitigation under extreme scenarios. That is not complacency—it is risk budgeting with named contingencies. (ferc.gov)

Definitions matter here. Wet-bulb temperature measures heat plus humidity and indicates evaporative-cooling limits; a cooling center is an air-conditioned public refuge; the wildland-urban interface is where development meets burnable vegetation. Authorities are not keeping up perfectly, but A undercounts measurable institutional learning.

A’s rebuttal concedes the central ground B needs: authorities are in fact learning and deploying operational tools before final mortality counts arrive. A grants CDC/NWS HeatRisk, CDC syndromic emergency-department surveillance, and city heat emergency plans are real and operationally useful. That concession matters because the core question is not whether climate risk is fully neutralized; it is whether agencies are adapting fast enough relative to critics’ portrayal. Once A admits active forecasting, surveillance, outreach and coordination, their remaining argument depends on a much stronger claim: that any emergency invocation, fatal incident, funding fight, or uneven local implementation proves adaptation is “behind the curve.” That standard is too close to unfalsifiable. Compound summer risk cannot be “designed out”; competent systems will still use emergency authorities during extremes, just as hospitals use surge protocols without proving hospital planning is fake.

A’s specific empirical weakness is reliance on older or selective baselines while ignoring newer movement. They cite NERC’s 2025 six-region elevated-risk finding, but the current 2026 NERC snapshot says record resource additions have strengthened summer readiness, that more than 58 GW of resources were added since summer 2025, and that at-risk areas decreased from six in 2025 to three regions plus one locality in 2026. That does not eliminate grid risk, especially under heat, drought, wildfire, data-center load and maintenance overlap, but it directly contradicts A’s claim that planning is merely emergency improvisation. It shows measurable risk reduction year over year.

The same pattern holds for heat response. A’s Chicago example is important, but a single weekend gap cannot outweigh evidence that large jurisdictions are expanding playbooks: CDC’s Heat & Health Tracker uses near-real-time ED data; New York City’s 2026 plan opens cooling centers, targets older adults, chronically ill and unhoused residents, monitors health effects and coordinates with utilities; New York State is funding new cooling centers and heat emergency plans while expanding free air-conditioner eligibility and vulnerable-worker outreach. Those are precisely the targeted outreach and cooling-network adaptations the premise asks us to evaluate.

On wildfire, A treats the EU’s largest-ever deployment as an admission of failure. It is better read as evidence of scaling: nearly 800 pre-positioned firefighters, aircraft ready to deploy, EFFIS forecasts, Copernicus mapping and a Cyprus regional station represent anticipatory allocation across borders. On budgets and insurance, EIOPA’s natural-catastrophe protection-gap dashboard and C40 climate budgeting show institutions increasingly pricing climate risk into governance and finance. A is right that recent deaths reveal limits. But their rebuttal never shows the counterfactual pace required, and it discounts concrete improvements because the hazard is worsening. The more defensible conclusion is B’s: adaptation is incomplete but faster and more compound-aware than critics acknowledge.

A’s strongest case is that “adaptation” should be judged against worsening compound hazards, not against agencies’ press releases or tool launches. On that standard, current excess deaths in Europe, the June 27 Colorado-Utah burnover deaths, uneven cooling-center access, and emergency grid interventions show that real-world outcomes remain dangerous during the exact heat-fire-holiday overlap under debate. A also rightly warns that localized best practices do not prove national or continental sufficiency.

My position is still stronger because A converts evidence of residual harm into evidence of inadequate adaptation without establishing the counterfactual pace a competent system should achieve under intensifying extremes. The question is not whether heat deaths, firefighter fatalities, outages, or emergency orders still occur; no serious emergency-management framework can promise that. The question is whether authorities are increasingly managing the overlap rather than treating hazards as isolated. On that empirical question, the evidence favors B.

First, operational heat response has moved beyond retrospective mortality. CDC/NOAA HeatRisk gives a nationwide seven-day health-based forecast, and CDC’s Heat & Health Tracker provides local heat-health information for community preparedness and response; these tools do not solve mortality, but they directly address A’s claim that agencies are blind until final excess-mortality estimates arrive. (cdc.gov) New York’s June 2026 update is a concrete state-level example: it expands public cooling via resilience-hub grants, provides cooling-center guidance for consistent services, broadens free air-conditioner eligibility, and expands site visits and vulnerable-worker outreach. (dec.ny.gov)

Second, grid planning shows measurable risk reduction, not mere improvisation. NERC’s 2026 assessment reports that more than 58 GW of resources were added since summer 2025 and that summer elevated-risk areas decreased relative to 2025, while still identifying remaining vulnerabilities under extreme heat, hydrology, wildfire, and load scenarios. (nerc.com) That is exactly what compound-risk planning should look like: not denial of extremes, but quantified reserves, named risk regions, and mitigation triggers.

Third, Europe’s wildfire posture is anticipatory. The European Commission’s 2026 plan pre-positions nearly 800 firefighters in high-risk areas, keeps aircraft ready, uses EFFIS forecasts and Copernicus mapping, and adds a Cyprus regional firefighting station. (civil-protection-humanitarian-aid.ec.europa.eu) Finally, adaptation is entering finance and budgeting: EIOPA’s natural-catastrophe protection-gap dashboard is designed to identify drivers of uninsured climate losses, while C40 describes climate budgeting as integrating climate risk and resilience into spending decisions. (eiopa.europa.eu) A proves the systems are imperfect under severe stress; B better explains the observed institutional trajectory: faster, more integrated adaptation with remaining limits.