clock menu more-arrow no yes

What we know about the death toll in Puerto Rico

(Hint: Trump is wrong.)

People walk across a flooded street in Juana Matos, Puerto Rico, on September 21, 2017.
People walk across a flooded street in Juana Matos, Puerto Rico, on September 21, 2017, the day after Hurricane Maria slammed into the island.
Hector Retamal/AFP/Getty Images

President Trump late Friday tossed off two tweets again rejecting Puerto Rico’s official Hurricane Maria death toll, following assertions he made Thursday that the official count of 2,975 deaths was a plot by Democrats to make him look bad.

The president again provided no evidence to contradict the now widely accepted death toll, which was calculated after months of painstaking analysis of death records and expected mortality rates by researchers at George Washington University at the behest of the governor of Puerto Rico, Ricardo Rosselló. (Rosselló said in a statement on Thursday “it is a fact” that 2,975 people died following the hurricane.)

And falsely, Trump claimed Friday that “this method was never done with previous hurricanes because other jurisdictions know how many people were killed.”

In fact, the GWU researchers found that the death count remained low at 64 for many months after the storm largely due to the fact that doctors in Puerto Rico had been poorly trained in and were confused about the guidelines of the Centers for Disease Control and Prevention for attributing deaths to a disaster in death certificates. “The analysis shows that physician unawareness of appropriate death certification practices after a natural disaster and the Government of Puerto Rico’s lack of communication about death certificate reporting prior to the 2017 hurricane season substantially limited the count of deaths related to María,” they wrote.

Which meant that in the first few weeks and deaths after the storm, doctors failed to record many deaths that were indirectly related to the storm — but still worthy of inclusion in the death toll, per CDC guidelines — such as kidney patients who couldn’t get treatments because their dialysis center didn’t have generator fuel or fresh water. In the absence of reliable data from death certificates, researchers then used a statistical method approved by the World Health Organization to measure excess mortality to come up with the 2,975 deaths.

Make no mistake: The figure of 2,975 deaths is an estimate. But given that it has been so politicized, it’s worth walking through the research that gave us estimate, and why we can be confident in it, even though we don’t have all the names or know exactly how they died.

A brief history of the Hurricane Maria death toll

When the storm made a direct hit on the island on September 20 and knocked out 80 percent of the island’s power transmission lines, it was clear that many lives would be at stake. Thousands of sick, elderly, and small children were left without electricity, water, and medical care. And as the response progressed painfully slowly (in part because FEMA was stretched responding to two other hurricane disasters and wildfires in California), thousands continued to be without power, water, or medical care for months.

In the first week after the storm, the government announced an official death toll of 16. Immediately, to observers on the ground, this number seemed much too low. The first person to recognize it was Omaya Sosa Pascual, a reporter from the Center for Investigative Journalism, a local investigative journalism group. Sosa Pascual called the 69 hospitals around the island and learned that there were an estimated 60 confirmed deaths linked to the hurricane and possibly hundreds more.

Several other media outlets, including Vox, and independent researchers then began analyzing government data and reports from the ground. We all found additional evidence that even in the first month after the storm the death toll was much higher than the official count — likely to be more than 1,000. For instance, Vox reported in October 2017 that the death toll was likely more than 450.

The official death toll, however, was eventually updated to 64 and remained there until August, even after several groups of researchers published their higher estimates in peer-reviewed journals. It took the release of the GWU report, which Gov. Rosselló commissioned himself and was the most comprehensive assessment to date, to give the government the confidence to put out a new official toll.

But still, even the 2,975 figure is not a list of names but rather a best estimate, according to Rosselló. “This number can change,” Rosselló said in August. “It could be less, it could be more, as time passes.”

Why we should trust the GWU estimate

If it feels like there have been a lot of different death tolls mentioned over the last six months or so, there have been. In the absence of a reliable figure from the government, three different groups of public health and social science researchers, along with several teams of journalists, attempted in the months since the storm to calculate the “excess mortality.”

Researchers were interested in calculating not only the people who died during the peak of the storm, but also from the resulting loss of power, water, medical, and emergency services.

Several in the first wave of estimates beginning in December put the toll at about 1,000 deaths, about 15 times higher than the official count. The government also acknowledged in a report to Congress this summer (before it released the new official toll) that the storm caused 1,427 more deaths than “normal.”

The ideal way to calculate the death toll from a hurricane, disaster researchers say, generally, is to count all the deaths in the time since the event, and then compare that number to the average number of deaths in the same time period from previous years. Subtract the average number from the current number and that’s the death toll.

In Puerto Rico, there was, as the GWU researchers note, significant confusion among doctors about attributing deaths to the storm. And given how long the power outages lasted, there were storm-linked deaths even into February, five months after the storm hit.

“The official government estimate of 64 deaths from the hurricane is low primarily because the conventions used for causal attribution only allowed for classification of deaths attributable directly to the storm, e.g., those caused by structural collapse, flying debris, floods and drownings (see below),” they wrote. “During our broader study, we found that many physicians were not oriented in the appropriate certification protocol. This translated into an inadequate indicator for monitoring mortality in the hurricane’s aftermath. Verification of attribution takes time, while excess mortality estimation is a more immediate indicator.”

They then looked at available death certificates and other mortality data from September 2017 through the end of February 2018, following the Centers for Disease Control and Prevention’s guidelines on mortality reporting in disasters. They fed that data into a mathematical model approved by WHO that also adjusted for age, sex, and migration from the island, and ultimately found that there were an estimated 2,975 excess deaths related to Hurricane Maria, or 22 percent more deaths than would have been expected during that period in a year without the storm.

And it also provided a clearer picture of who died. The risk of death, they wrote, was 45 percent higher for people living in poorer areas. It was also higher for the elderly men. (For more on the victims, do read this excellent collaboration between the AP, Quartz, and the Center for Investigative Journalism.)

Researchers at GWU issued a statement Thursday after the Trump tweets standing by their study, calling it “the most accurate and unbiased estimate of excess mortality to date.” Indeed, though it doesn’t give us the names of the dead or how they died, the study does appear to be the most comprehensive statistical attempt to measure the deaths from the storm in part because it was able to include the deaths from January and February.

“Gathering data, including death tolls, during and after a catastrophe is often difficult but researchers have developed a number of approaches over the years,” Samantha Montano, a disaster expert and writer, tells Vox. “The methodology used in the GWU study was, in my opinion, an appropriate approach given the constraints of the situation.”

But it’s possible that the death toll is even higher. A study by researchers at Harvard, published in May in the New England Journal of Medicine, estimated a range of 793 to 8,498 deaths, with a midpoint of 4,645 deaths.

In January and February, those researchers surveyed about 3,300 randomly selected households across Puerto Rico and asked residents how many people from those households had died in the months after the storm made landfall. They found 38 deaths in the households, and they used that figure to calculate a mortality rate and then compared it to the mortality rate from the same time the previous year.

The difference was then used to come up with a very broad estimate for the number of people who died above what you’d expect to see in a normal year: the range of 793 to 8,498 deaths.

Alexis Santos, a Puerto Rican demographer at Penn State who conducted his own analysis of mortality following the hurricane, also published a paper in the journal JAMA on August 2, with Jeffrey Howard of the University of Texas San Antonio, estimating the death toll at 1,139.

Trump (and other officials) failed Puerto Rico. But he won’t take responsibility for it.

Again: All of the death toll estimates, from reputable, epidemiological statistical analyses, number in the hundreds and thousands. The estimates all show that the government initially underestimated the toll of the storm. And they show there’s no way, as Trump claimed today, there were not much more than 16 or 64 deaths.

Many of the deaths that ensued likely could have been avoided if more attention and resources had been offered early on. And Puerto Ricans remain bitter too: A recent Washington Post-Kaiser Family Foundation poll finds that 8 in 10 residents give negative marks to Trump, 7 in 10 give negative marks to the the Puerto Rican government’s efforts, while two-thirds are displeased with Rosselló’s response.

And as American University researcher Morten Wendelbo writes at The Conversation, “the fatality count alone is woefully inadequate for understanding the depth of the disaster still unfolding in Puerto Rico. Considering disaster in terms of fatalities does not capture the hardship experienced by survivors. Although survivors did not pay with their lives, many lives were still changed in devastating and often permanent ways.”

As Vox’s Matt Yglesias wrote: “The carnage in Puerto Rico is the most severe manifestation of Trump’s basic unfitness for the job he currently occupies. ... If you put a telegenic demagogue in office, you will get some choice moments of televised demagoguery. You won’t get an adequate response to a hurricane, and that means you will get a sky-high death toll.”