Four years after 739 excess deaths, Chicago got a second chance. In late July 1999, another heat wave settled over the city. This time, the city had a plan.
The Extreme Weather Operations Plan had been developed in the aftermath of 1995. When temperatures climbed again, officials activated it. They issued strongly worded media warnings, opened cooling centers, provided free bus transportation, called elderly residents by phone, and sent police officers and city workers door-to-door to check on seniors who lived alone.
The death toll dropped. The CDC's 2003 MMWR report documented 103 heat-related deaths in the 1999 event, of which 80 were attributed to extreme heat as the underlying cause rather than a contributing factor. The CDC's tally for 1995 was higher than the medical examiner's contemporaneous count of 465: the 2003 report put the figure at 485 heat-related deaths, likely reflecting revised classifications over time. By any measure, 1999 was less deadly.
The question is how much of that difference the plan can claim.
The CDC's Causal Claim
The CDC's language was direct:
The agency stated that implementation of Chicago's plan "reduced the death toll" by increasing the number of daily contacts for elderly residents.
That's a causal statement, and it may well be true. But the evidence available doesn't isolate the plan's effect from everything else that differed between 1995 and 1999.
What Else Changed
The weather, for one. The 1995 heat wave was exceptional by any historical standard. Maximum daily temperatures reached 104°F, and the heat index peaked at 119°F. The 2003 MMWR report does not provide comparable peak temperature or heat-index figures for the 1999 event, making direct meteorological comparison difficult. What the report does say is that 1999 produced the highest annual heat-related death count in the 1996–2001 window, confirming it was a serious event. But if peak temperatures were lower or the event shorter, mortality would have declined independent of any intervention.
Public awareness had also shifted. In 1995, Chicago had not experienced a catastrophic heat wave in recent memory. Media coverage during the event was initially slow, and public awareness of indoor heat danger was low. By 1999, the city had buried its dead and endured a painful public reckoning. Residents who remembered 1995 may have taken precautions on their own. Families may have checked on elderly relatives without being prompted by a city worker.
Media coverage of the 1999 event was itself shaped by the earlier disaster. Newsrooms that had been criticized for slow response in 1995 were primed to treat the next heat wave as front-page news from the first day. That coverage functioned as a mass public health intervention, reaching people who would never encounter a wellness check or a cooling center bus.
And the plan itself was a bundle of interventions: media warnings, cooling centers, transportation, phone calls, door-to-door visits. Even if the package as a whole reduced mortality, the CDC's attribution to "increased daily contacts" doesn't tell us which component mattered most.
The Counting Problem
There's a deeper difficulty. The CDC's own 2003 report notes that deaths classified as hyperthermia represent only a portion of heat-related mortality, because cardiovascular and respiratory deaths also rise during heat waves. The 1995 excess-mortality figure of 739 was substantially higher than the certified count, suggesting the medical examiner's tally captured roughly two-thirds of the actual toll. If similar undercounting applied in 1999, the true excess mortality was higher than 103. But we don't know whether the ratio held. Improved awareness may have led to more complete certification in 1999, or the different temperature profile may have produced a different distribution of direct versus indirect heat deaths.
Comparing 485 to 80 looks like an 83 percent reduction. But comparing two events with different meteorological profiles, different levels of public awareness, different media environments, and a multi-component intervention makes it impossible to assign a clean percentage to any single cause.
What a Wellness Check Actually Is
The Extreme Weather Operations Plan consisted of cooling centers, phone calls, door-to-door checks, free buses, media warnings. These are institutional substitutes for the daily contact that, in neighborhoods like Auburn Gresham, happened without any plan at all. The block club president who knocked on a neighbor's door. The grocery store owner who noticed a regular customer's absence. The church member who called after a missed Sunday service.
The plan tried to manufacture, through municipal labor, what social infrastructure provides automatically. A police officer checking on a senior who lives alone is performing, for one afternoon, the function that a functioning neighborhood performs every day. A cooling center stands in for the diner or the church basement. A free bus route approximates the walkable commercial corridor.
A wellness check requires a list. Someone has to know which addresses to visit. In 1995, the people who died alone were, by definition, on nobody's list. A city plan can generate lists from whatever municipal records are available, but the list is only as good as the data feeding it, and the visit is only as useful as what happens at the door. The neighbor who knows your name, knows you take heart medication, knows your air conditioner broke last month is doing something a uniformed stranger at the door cannot replicate.
When the social infrastructure has been stripped away by decades of depopulation and disinvestment, institutional intervention is what remains. That reality describes both the conditions the plan was designed for and its inherent limits.
What Remains Unknowable
The plan probably helped. The lower temperatures probably helped. The changed media environment probably helped. Shifted public behavior probably helped. We cannot parse out how much each contributed. No city can run a controlled experiment on heat-wave response. You cannot randomly assign cooling centers to some neighborhoods and withhold them from others. You cannot hold temperature constant between events separated by four years.
That ambiguity is the nature of disaster response evaluation. Twenty-three years after the CDC published its comparison, cities across the country are investing in extreme heat response plans modeled on the same logic Chicago pioneered. Cooling centers, phone banks, wellness checks, transportation. These are real interventions, potentially life-saving. But the social infrastructure that protected people in Auburn Gresham and Little Village in 1995 was built over generations, not on a budget cycle. The commercial corridor, the block club, the church that notices when someone is missing: those accumulated across decades. And no evaluation framework yet exists that can measure the distance between what a city can deploy in a crisis and what a neighborhood provides when it's whole.