President Trump provoked outrage on Twitter and in the media in mid-September with his tweets that denied the death toll in Puerto Rico from Hurricane Maria in 2017 and blamed Democrats for artificially elevating it.

Measuring the mortality in epidemic or emergency periods has long been controversial. Authorities often deny reports of high mortality because they seek to avoid being blamed for mismanagement. But all too often, the reports are indeed accurate—and understanding why is essential to helping save lives in future disasters.

During the brutal Chicago heat wave of July 1995, for example, Mayor Richard M. Daley’s office rejected the medical examiner’s death toll. A subsequent study eventually confirmed the medical examiner’s figure, indeed finding that mortality was even higher than initially reported. 

When unprecedented extreme heat consumed France for two weeks in August 2003, the country’s health ministry and prime minister initially denied reports of thousands of heat-related deaths. Yet again, an official study produced by France’s national medical and public health agency confirmed that nearly 15,000 French men and women had died from the heat.

The president appears to argue that the figure of 3,000 deaths in Puerto Rico—confirmed by independent studies from George Washington University and Harvard—is false because initial reports indicated just a handful of deaths and the larger number appeared “a long time later.” 

Yet by definition, disaster mortality reporting takes time. The accepted method for estimating disaster deaths involves a calculation of what epidemiologists call excess mortality, which involves subtracting the average number of deaths that would have occurred in the absence of a disaster from the number of reported deaths during and in the aftermath of the disaster. Scientists can then attribute any excess deaths to the disaster itself. While the calculation is simple in principle, it cannot be performed overnight. 

These excess mortality estimates often include deaths that are due to the disaster yet are indirectly linked. For example, rates of drowning go up during heat waves, as many inexperienced swimmers take to the water to find relief. In the case of hurricane reporting, indirect deaths include not only those killed by flooding or being struck with debris, but also those who died in the storm’s aftermath as a result of power failures, a lack of potable water, disease linked to flooding and contamination, and an inability to access health care. These deaths are inextricably connected to the disaster that precipitated them since they would not have occurred in the absence of extreme weather.

More importantly, effective management of disasters can minimize both direct and indirect deaths. The federal government cannot alter weather patterns, but it can help communities prepare for disaster and rebuild in the aftermath. When an administration is negligent and mismanages the response, it can cost lives.

For nearly a year after Hurricane Maria, parts of Puerto Rico remained without power, straining the island’s capacity to provide secure water, food, and health care to its population. The Trump administration has contributed to the problem, awarding rebuilding contracts to inexperienced and unqualified contractors. This has exacerbated the difficulties of recovery, delayed the ability to account for disaster-related deaths, and elevated the death toll linked to the hurricane’s devastation. 

As the East Coast braces for the peak of another hurricane season, and Hurricane Florence wreaks havoc on the Carolinas, it is critical to our disaster preparation efforts that we acknowledge their potential for destruction. Denying death tolls trivializes disasters and discounts the value of human lives. Contrary to the president’s assertions, natural causes did not kill Hurricane Maria’s victims; vulnerability to disaster and a delayed response did. And a political party did not invent these figures to make the president “look as bad as possible.” Scientists calculated them to establish the truth of a disaster’s human toll.