How to Count the Victims of Hurricane Maria in Puerto Rico

Alberto A. Martinez
4 min readJun 18, 2018


Controversy flares over how many Puerto Ricans died because of Hurricane Maria. On May 29, the New England Journal of Medicine published an article arguing that Puerto Rico’s Department of Public Safety misreported the number of deaths as 64 people on December 9, 2017. Instead, the researchers estimated that roughly 4,645 people died, from September 20 until December 31st.

Since they work at Harvard and other top institutions, their article caused a media storm. As a Puerto Rican professor of history, I was stunned by the headlines.

Apparently, more people died in Puerto Rico’s post-hurricane wreckage than in the destruction of the Twin Towers on September 11, 2001.

Is it true? I’ve been to the island repeatedly, post-hurricane, so I know the crisis is disastrous and prolonged. And now, in the news I see photographs of a public memorial to the dead.

On the north coast of Old San Juan stands the Capitol of Puerto Rico. Hundreds of Puerto Ricans placed empty pairs of shoes in front of it, as a memorial to those who died. Some brought shoes in solidarity. Others brought shoes of their family members who died.

New Balance sneakers, rubber flip flops, worn-out boots, Air Jordan basketball shoes, Rebels sandals, golden high heeled shoes, Aldo leather shoes, FILA sports shoes, Converse, faded Skechers, dirty Crocs, baby shoes, Rage shoes.

People stared in silence at those shoes. I was struck by photos of one girl: wavy bronze hair, jeans and a backpack. It was a bright day, but she had an umbrella. And she wore a big cape: the Puerto Rican flag. She placed white flowers on shoes while holding a rolled-up poster. In a gripping photograph, she sits alone in front of those shoes, facing the Capitol, and squinting with a stern painful face, holding her large white sign with handwritten words: “4,645 Deaths and daddy was one.”

Everything real is on her face. I don’t know how many people died. I don’t believe the article by the researchers at Harvard. But I believe her face.

The researchers randomly surveyed 3,299 households in Puerto Rico. The respondents reported only 3 deaths “directly related” to the hurricane, plus 12 “indirectly related.” Based on merely those 15 deaths, the researchers extrapolated to the entire population of three million people. They called their estimate of 4,645 deaths “conservative,” and “likely to be an underestimate.” They calculated that actual deaths may be as high as 9,889, or as few as 793.

Experts bashed the study for having huge margins of error. Journalists pleaded for data from Puerto Rico’s Department of Health (PRDH). Then it finally announced that in the same time frame as the Harvard study, from September until December 31st, there were 1,397 excess deaths.

But 1,397 was not an official recount of deaths caused by Maria.

Instead, it’s a blind tally of excess deaths in late 2017 compared to late 2016, with no analysis of how unusual is that excess compared to fluctuations in previous years. The PRDH just did not specify the cause of such deaths.

Therefore, this estimate is defective too. It cuts off the excess of deaths at December 31st, as if the crisis ended right then. And it compares mortality only to 2016.

The number of deaths is gravely important. It affects how we respond. Did local and federal agencies provide enough aid in timely ways? How does this crisis compare to others? Did the government lie? Were news reports fair? How can we save people next time?

According to official monthly records from 2009–2014 and 2015–2018, it seems that the number of deaths in Puerto Rico regained a semblance of normalcy in April and May 2018, compared to past years.

Therefore, to estimate “excess of deaths” for when Maria struck, September 2017, until March 2018, I tabulated the average deaths for each month in previous years. For example, for September 2009, September 2010, etc., until September 2016, the average number of September deaths in Puerto Rico was 2,364. In contrast to all those September months in eight years, in September 2017 many more people died: 2,928.

That’s a huge excess of 565 deaths in one month. I cannot imagine that it was caused by anything other than Maria. Because, in 2017 Puerto Rico had the lowest population in decades: roughly 400,000 fewer people than in 2009. And no month of September in the previous eight years is even within 430 deaths of September 2017.

Yet the government falsely claimed that only 64 people died.

Did the hurricane’s aftermath kill 500 people in just ten days? We must now inspect day-by-day statistics, so that we may know how many people died from September 20 to 30.

Similarly, in October 2017 there were 557 more deaths than the average for every October from 2009 until 2016. Next, there were 253 excess deaths in November 2017, followed by 167 in December. And finally 182, 81, and 91 excess deaths in the first three months of 2018.

Summing up, from September 2017 until March 2018, I estimate an excess of 1,896 deaths. The majority of those deaths were caused by the lack of electricity which obstructed heath care, stopped ventilators, communications, etc.

Puerto Rico’s government delayed the reporting of deaths. Because, they know that most deaths were not caused by the hurricane itself, but by the two political parties that run the island. For decades they did not invest enough on infrastructure and hospitals.

Instead, they incinerated public funds on wasteful government contracts, kickbacks, grossly overpaid consultants, bribes, frauds, absurdly high salaries, expensive lawyers, bankers, embezzlement, political campaigns, needless advertisements, and corruption. That’s why we stare at the Capitol, not the sky.

Alberto A. Martinez is a professor of history at the University of Texas at Austin and a Public Voices Fellow with The OpEd Project.



Alberto A. Martinez

Professor of history at the University of Texas at Austin. Author of six books, plus articles in Scientific American, The Hill, USA Today news, SALON, etc.