Puerto Rico's governor says there will be 'hell to pay' if officials withheld mortality data

[DonViejo]

Source: CNN




By Emanuella Grinberg, CNN

Updated 8:46 AM ET, Fri June 1, 2018

(CNN) - Puerto Rico's governor defended his government's efforts to account for deaths on the island following Hurricane Maria after a new study called the official tally a "substantial underestimate."

The official death toll in Puerto Rico has been the subject of substantial controversy since Hurricane Maria hit the island, a US territory, on September 20. Gov. Ricardo Rosselló said he welcomed the results of a Harvard study published Tuesday in the New England Journal of Medicine.

Researchers surveyed 3,299 randomly chosen households across the island and compared it to official mortality statistics from 2016. They concluded that an estimated 4,645 people died in Hurricane Maria and its aftermath in Puerto Rico -- almost 70 times the government's official tally of 64.

Rosselló said he does not stand by the official death toll. But he said his government did the best it could with information-gathering protocol available to them at the time.



Read more: https://www.cnn.com/2018/05/31/us/puerto-rico-governor-responds-to-study/index.html

[Plucketeer]

1. So in THIS case

the "Crowd Size" is TOO BIG?

[Guy Whitey Corngood]

2. Oh yeah, I'm sure. People here in the mainland aren't the only ones electing useless tools. nt

[bucolic_frolic]

4. Actuarial science not even much help in a disaster

This fiasco could be repeated on the United States mainland if we don't count ALL the people in a census.

[Girard442]

5. No, there won't.

People sat on their hands while the now dead could have been saved. They’re not going to fly into a rage when they’re undercounted.

[keithbvadu2]

6. The Hell you say!

[Igel]

7. The first CNN article was better.

This one tries to pin a number on the researchers.

The "source" of this article is, of course, another CNN article, but at least that one tried a bit harder in stating, "Marqués and colleagues say 793 to 8,498 hurricane deaths occurred." https://www.cnn.com/2018/05/29/us/puerto-rico-hurricane-maria-death-toll/index.html

And that's the problem with statistics. It won't give a number. The top and bottom numbers themselves have error bars associated with them.

One problem is what a death "due to" the hurricanes would look like. If somebody has a heart attack because the day after the storm hit they couldn't get through a blocked road and dies, is that due to the hurricane? The person might have died anyway. The strictest definition requires that the death be immediately caused by the storm: floodwaters, something falling on them, etc. This hypothetical heart-attack case? After the storm, therefore not immediately caused by the storm.

Then, apart from the bookkeeping aspect, there's the blame aspect. People need to minimize or maximize the death count for political or personal reasons. In that hypothetical case, no government would have been able to transfer manpower and materiel in sufficient quantity to clear all the streets, restore power, etc. Esp. since parts of the territory were still flooded. Even with Ike, part of the quick response was because a lot of supplies were pre-positioned near the suspected path of the storm and the roads were clear fairly quickly. There were land routes and they could haul trailers with ice and supplies from those places to Houston in 4-6 hours. Even then, some people were without power for weeks. (On the other hand, my apt. lost power at 11 pm and had it back around 5 pm the next day.) With PR, it took a day to get the airports open, and the sea ports remained closed for a while longer undergoing repairs.

The best that'll happen is there'll be a generally accepted number of deaths. But we'll argue over that for a while and try to appeal to authority and redefine things as necessary because politics. Of course, every redefinition makes it impossible to compare the death toll from any other hurricane, but that won't stop advantageous comparisons or damning comparisons.