Garbage and Non-Garbage Estimates: Puerto Rico Edition

Since the Puerto Rican government ceased publishing mortality data in February, there has been a debate over the death toll arising from Hurrican Maria. The official death toll, focusing on direct deaths, remains at 64. However, starting in November, a number of scholars attempted to gain further insight into the extent of the human disaster in the Commonwealth. One commentator has labeled another study “garbage”. What is the import of these competing analyses?

First consider the estimates over time of the death toll, both direct and indirect.

Figure 1: Estimates from Santos-Lozada and Jeffrey Howard (Nov. 2017) for September and October (calculated as difference of midpoint estimates), and Nashant Kishore et al. (May 2018) for December 2017 (blue triangles), and Roberto Rivera and Wolfgang Rolke (Feb. 2018) (red square), and calculated excess deaths using average deaths for 2015 and 2016 compared to 2017 and 2018 using administrative data released 6/1 (black square), and Santos-Lozada estimate based on administrative data released 6/1 (large dark blue triangle), end-of-month figures, all on log scale. + indicate upper and lower bounds for 95% confidence intervals. Orange + denotes Steven Kopits 5/31 estimate for range of excess deaths through September 2018; Orange triangle is Steven Kopits estimate for year-end as of June 4. Cumulative figure for Santos-Lozada and Howard October figure author’s calculations based on reported monthly figures.

The Harvard School of Public Health led analysis was a survey based study, which – because of the relatively small number of households surveyed relative to population resulted in fairly large confidence intervals.

The calculations based on administrative data — the Santos-Lozada and Howard study, the Rivera and Rolke study — both had confidence intervals, but much narrower. That’s because the uncertainty came from sampling error involved in estimating the average death rate per month under normal circumstances; the observations for 2017-18 post-Hurricane were treated as known, and without measurement error. As was highlighted by Mr. Kopits assertion as of 5/31 that the mortality data through December was well measured, that is not necessarily a good assumption.

In other words, confidence intervals are important, and it’s important to know what assumptions underpin the construction of the confidence intervals. Assuming away measurement error in say administrative data (either unbiased — the best case — or biased) leads to overconfidence in the precision of one’s estimates. Unfortunately, in the absence of more information, there’s not much one can do (Santos-Lozado and Howard calculate differences of 2017 reported data from upper 95% confidence interval for pre-Maria averages, in order to be conservative.)

Interestingly, Mr. Kopits’ current (as of 6/4) estimate of cumulative excess deaths by end-December is well within the 95% confidence interval cited by the Harvard School of Public Health-led research team. (Journalistic accounts were often misleading, saying at least 4600 excess fatalities occurred; but that is not the fault of the research team, and in fact the NY Times reported the confidence interval correctly in their article.)

An interesting account of the difficulties confronted by those relying on administrative data is to be found in FiveThirtyEight’s 2015 piece on the post-Katrina count:

By its own admission, Louisiana never finished counting the dead. Its last news release on the topic, from February 2006, put the statewide toll at 1,103. Three months later, it added hundreds of state residents who’d died in other states. Three months after that, in August 2006, Louisiana counted 1,464 victims, with 135 people still missing. Today, when asked about the Louisiana death total, the health department cites a 2008 study that reviewed death certificates and concluded that there were 986 victims. But that study said the total could be nearly 50 percent higher if deaths possibly linked to the storm were included.

One year after Katrina, the state’s medical examiner pledged to keep working until every victim was identified. Four years after that, he told the Houston Chronicle that he didn’t get the time or resources to finish the job.

Among federal agencies, the National Oceanic and Atmospheric Administration has been the primary one focused on determining how many people died because of Katrina regionwide. It reported 1,833 deaths in 2006 but has continually revised the number downward, to 1,100 at last count. Yet the 1,833 number has made it into news articles and the congressional record in the past month. The agency’s count remains as uncertain as it was in 2005, when NOAA researchers wrote that “the true number will probably not ever be known.”

John Mutter, a geophysicist at Columbia University, was more familiar with earthquakes than hurricanes before Katrina. After the levees failed and the official death counts kept rising, Mutter began looking into it. There had not been a storm with a comparable death toll since 1928, when a hurricane pushed the waters of Florida’s Lake Okeechobee up and over the levees at its southern end, drowning thousands. Mutter was interested in what standards existed for counting how many people were killed by a hurricane. He discovered that there aren’t any. “They made their own,” he said in a telephone interview.

Standards are extremely important for the grim task of counting the dead. They settle questions with no obvious right answer — for example, whether to include deaths that occurred immediately before a storm hit (such as someone who died in a fall while cutting down tree branches to mitigate anticipated damage).

What are known as indirect deaths are the most confounding to the count. Direct deaths are those that occur from drowning or an injury sustained during the storm or post-storm flooding, while indirect deaths occur from some other cause that might be linked to the storm, such as an inability to access medical care to treat an illness.

After Katrina, government counters in Louisiana chose to include indirect deaths based on an arbitrary time cutoff — people who were evacuated from New Orleans and died after Oct. 1 were not included, while those who died before were. The authors of the 2008 study that counted 986 Louisiana deaths took a different approach, counting only deaths that could be directly attributed to the storm. “I do think we’re likely an underestimate,” said Joan Brunkard, an epidemiologist with the Centers for Disease Control and Prevention and the study’s lead author. NOAA, meanwhile, has reviewed death reports and removed indirect deaths from its count, a major reason that its total went down.

Hence, my view is that contentions that a “firm” count has been achieved within a year of a major and ongoing disaster is unwarranted. I also think that people should learn the nature of the data sources they are using, and what survey design involves, what sampling error is, and what measurement error is, before undertaking the difficult task of policy analysis.

Addiitional discussion, here, here, and here. Gelman’s commentary on the Harvard School of Public Health led study, here. Useful thread on the study by way of Alexis Santos-Lozada here.

Additional conversation with some of those “on the ground”, NPR’s 1A show aired yesterday.

47 thoughts on “Garbage and Non-Garbage Estimates: Puerto Rico Edition

  1. Not Trampis

    As Mr Spock would say fascinating. Or he might have said its statistics Jim but not as we know it.

    in all seriousness well done Menzie..

  2. ilsm

    It is keeping score with none of the umpires on the field!

    What is an estimate worth? What is it offering?

    1. ilsm

      I heard a radio report while driving this PM. The Massachusetts Fish and Game is trying to find out how many wild turkeys there are in the Commonwealth of Massachusetts, they have set up a web site for citizens to report sightings……. that reminded me of the Harvard study.

      Are fish and game standards of confidence different for estimating a wild turkey population?

  3. CoRev

    Menzie, since you are relying on the importance of the confidence levels of the various studies, why not show that more clearly on your chart? Those little+ are too easy to miss. As I did on thei first viewing. I would recommend a background color change for the Harvard range. From your Alexis Santo-Lozada link: “Don’t get me wrong, their number is the same as ours, based on overlap of Confidence Intervals, but it is important to create ways to explain the meaning of CIs to the press.”

    With a revised graph I think we will see the value added importance of the CI of these estimates. IIRC your original complaint about Steven Kopits 1st estimates provided in his blog comments was that he did not include a CI. I don’t remember if he included a CI in his first article, and haven’t yet had enough coffee to do the research.

    Off to get my next cup.

  4. pgl

    “the government of Puerto Rico released island death statistics as of May 31, 2018. These statistics, which can now be taken as materially final through year end, indicate 1,397 excess deaths following Hurricane Maria for the Sept. 20 – Dec. 31, 2017 period covered by the MPR Study. ”

    I could care less if the official reported has been tagged “final”. That does not give it any semblance of being accurate. Look we had this issue during both the Vietnam War and the 2003 invasion of Iraq. All sorts of official counts of how many people were killed in battle. And in most cases, the counts were gross underestimates in part because the government agency had every incentive to low ball the number.

    If Steve is so gullible to trust such official counts, then he has no right to call other more serious analyzes “garbage”.

    1. Steven Kopits

      To be clear, pgl, the government listed these numbers as ‘as of May 31’. I deemed them as to be ‘taken as materially final’ based on revisions seen to earlier vintage data sets. Bear in mind these deaths are ostensibly old people who died in their beds of pre-existing conditions. If you think the PR official numbers are wrong, then you are accusing the PR authorities of gross incompetence or deliberate mendacity. That’s the implication of the Harvard study, if the authors continue to hold to their central estimate and confidence interval. Do the authors believe that, or not?

      1. pgl

        “Bear in mind these deaths are ostensibly old people who died in their beds of pre-existing conditions.”

        Your evidence for this fact free assertion? Of course had the electricity been restored in a timely manner, some of your alleged old people would likely have survived.

  5. dilbert dogbert

    I think arguing over deaths is a diversion. The real issue is that most of the mainlanders think Puerto Rico is a “shithole” place and can’t be helped. Sort of like Charles Murriay’s “The Bell Curve” makes the argument that Blacks are genetically dumb and can’t be helped.
    Statehood would help with two senators and a representative in the congress. That would help get the “Horse Trading” and “Log Rolling” in congress going and there would be more focus on help for Puerto Rico.

  6. Bruce Hall

    Attributing post-event effects to the event itself is difficult and I think everyone agrees that people who would not have died under normal conditions do die as a result of the disruption to resources and services from these catastrophic events.

    The real question should be: what wasn’t done that should have and could have been done given the pre-existing conditions and the post-event conditions? How much bureaucratic ass-dragging occurred by whom (political obstructionism at federal and local levels, FEMA unpreparedness)? What obstacles could and could not be addressed (e.g., temporary suspension of the Jones Act, lack of contractors capable of restoring power quickly)? What resources could have been diverted from other sources that were not tapped (e.g., military aid to civilians; Medicare, Medicaid, and welfare funds diverted from the states to provide emergency health care)?

    The data estimates of excess deaths do not shed light on the political, financial, and logistical issues.

    1. pgl

      All the right questions and as I recall, these were covered by certain reporters much to the dismay of Team Trump. Alas the White House dismissed most of this coverage as “fake news”.

      1. CoRev

        Pgl, from your reference: “In mid-October, a month after Hurricane Maria, most of Puerto Rico remains
        without electricity or safe drinking water, and still has serious food shortages, due to a combination of insufficient disaster response from the federal government and the sheer magnitude of the storm.” These deplorable conditions before leading to worse after Maria are solely the blame of the Federal Response and the magnitude of the storm, with no acceptance of PR Government’s responsibilities before, during and after Maria?

        As I have already pointed out more experienced state Governments request aid before they are struck by a forecast disaster. PR? After.

        Furthermore it was the PR Government that let the infrastructure get so bad before.

        Nobody is exemplary on their response for this disaster. Blaming God, Mother Nature on the size of the storm is just another cop out. So, ignoring the cop out claim, what they are saying it was only the Federal Response that caused the post Maria deplorable conditions.

        1. Menzie Chinn Post author

          CoRev: Yes, and all that damage in Houston was because of the Houstonians own fault for paving over land and building in a flood plain and for not planning for once-in-a-thousand-year flood (under the no-climate-change calculation). Floridians should suffer for building on land that is within a foot of sea level. Jersey shore deserved to be leveled because no building should’ve been allowed there. Shall we go on?

          1. Moses Herzog

            I believe we have another illustrious 5-star comment to be handed out by our panel of judges, to the blog co-host himself.

          2. CoRev

            Menzie, a simple yes! The same example works for Ellicott City, MD with damages from two 1,000 year storms within 2 years. Above the city has been heavily developed reducing or removing the absorption from the natural forest and grassland, and allowing the drainage free access to the city below. The list can go on for days.

            There is no excuse for human greed and poor judgement.

          3. Menzie Chinn Post author

            CoRev: And the 200 injured/14 killed in West, TX, deserved it because they or their parents voted for lax regulation of explosives. They “brought it upon themselves”. Got it.

          4. Moses Herzog

            @ Menzie
            Nearly forgot all about that West Texas one. Small town near a railroad tracks yes?? I watched it live that night in the AM hours on stream from one of the local stations. It was amazing. It caught my attention in more ways than one because it was in essence the same material used for the Murrah building bombing. Remember them saying the federal government was going to clamp down hard on that stuff after Murrah?? Apparently not……. Makes you wonder where that stands now in 2018, doesn’t it??

            I’d tell you another one that scares the sh*t out of me, and I think of it a lot, but I don’t want to give anyone any more ideas than are out there now.

        2. baffling

          “As I have already pointed out more experienced state Governments request aid before they are struck by a forecast disaster. PR? After.”
          PR is NOT a state, its a territory. different resources.

          1. CoRev

            Baffled, no state or territory have identical resources, or needs as far as that goes. Regardless, it has nothing to do with waiting/delaying their request for the aid to which they were entitled.

      2. Bruce Hall

        pgl, I’m certain that some aspects of the federal response could be called into question, but the real questions are as I outline them (and the suspension of the Jones Act was simply one small aspect of the total response or non-response).

        Since most of the island was annihilated, one could paint some possible scenarios for better response:
        1. Amphibious landings of supplies (food, clothing, building materials, generators, fuel, etc.) rather than use of ports. Would there be enough amphibious vehicles to bypass ports that were clogged? Would these landings have access to roads that were usable? How long would it take to offload supplies via hundreds of amphibious ships rather than a few cargo vessels?
        2. Relocate federal medical personnel to Puerto Rico and establish military-style hospitals? Would there be enough available generators, fuel, medical supplies, doctors, nurses, administrators available and could these be located around the island with support logistics?
        3. Bring in a squadron or two of Army Corp of Engineers to rebuild the power grid. Would they have the necessary resources, expertise, and understanding of the existing grid to make it work? Would they have the necessary logistical support?
        4. Evacuate most of the residents of the island who were in unsupported conditions. This might have been the easiest option, but would have required large housing encampments for temporary resettlement in the states. Who provides support once in the states? For how long? What if they won’t go and other resources to help them are not readily available?

        So, I’ve given you a few alternative to address the lack of a real Federal response. I’d be interested in reading your specific recommendations.

        1. 2slugbaits

          Bring in a squadron or two of Army Corp of Engineers

          A squadron or two??? It would probably take more than a couple dozens folks. I think you meant a battalion or two.

          1. Bruce Hall

            2slug> Yes, that’s more the scope. I wonder how many would be trained and useful in disaster reconstruction?

          2. CoRev

            Bruce & 2slugs, there’s a lot of discussion of the delayed response, but this article shows how much was already pre-deployed or deployed early. At Least 6 Dead in Puerto Rico as Military Readies Hurricane Response. The date of issuance, 2017/09/22 is in the URL:

            It recognized PR’s concern that the number 6 deaths was already known to be low: “Hector Pesquera, head of Puerto Rico’s Department of Public Safety, said at least six were killed in the direct hit by Maria. “We are aware of other reports of fatalities that have transpired by unofficial means, but we cannot confirm them,” he said.”

            By 09/22 the San Juan Airport was open to military flights, and they were at that time: “…the military response thus far has been mostly limited to overflights to assess damage and gauge where relief and rescue efforts are most needed.”

            Most here forget this was not the first hurricane of the season to hit the region, and response units were already deployed to the Caribbean Islands already damaged by Irma and Harvey: “On Friday, the amphibious assault ship Wasp, which had already responded to the Virgin Islands after Hurricane Irma, arrived off twice-devastated Dominica in the Leeward Islands to support State Department assistance and aid in the evacuation of U.S. citizens, SouthCom said in a statement.”

            1st Responder Communications was an early issue: “In Puerto Rico, one of the first priorities is to re-establish communications, which were decimated in the onslaught by Maria, the National Guard Bureau said in a statement.

            The Guard is setting up Joint Incident Site Communications Capability (JISCC) systems as a stopgap communications bridge for first responders, the Bureau said.

            The Bureau said 18 JISCC systems were deployed to support relief efforts after Hurricanes Harvey and Irma made landfall.

            In response to Hurricane Maria, there are currently four JISCCs on-site in Puerto Rico and the U.S. Virgin Islands, with more on the way, the Bureau said.

            NorthCom said in a statement that its Navy component, U.S. Navy North, is continuing damage assessment flights of airfields, cities and coastal regions in Puerto Rico and the U.S. Virgin Islands, using aircraft from the amphibious assault ship Kearsarge and the dock landing ship Oak Hill.”
            I highlighted Caribbean Island because of the snark associated to a reference to the area in an earlier article.

            Another article shows this about the impact of PR’s financial situation and planning:
            ““They might call me paranoid,” Rosselló said. “I anticipated this could happen, and seven days before the storm, we started working on it.”

            But he was still hamstrung by Puerto Rico’s measly coffers: Expenses incurred before the White House approved a Sept. 20 major disaster declaration had to be paid in full by the state, which is $72 billion in debt and under the control of a federally appointed fiscal board.

            So asking other states for help before Maria, which might have lined up resources for Puerto Rico more quickly, would have been an expensive undertaking without knowing for sure what havoc the storm would wreak.

            In contrast, six days before Irma hit Florida, the state filed its first request through the Emergency Mutual Aid Compact available to states and territories. Florida ultimately made 99 requests before landfall.

            The number of requests Puerto Rico made before Maria: Zero.
            Read the entirety of this article explaining PR’s preparations and their impacts.

            This whole series of Puerto Rico-centric articles started as an attack at the Trump administration response to Hurricane Maria, while the reality most of the commenters are forgetting magnitude of the damages received across the whole area including other US territories from Hurricanes Maria, Irma and Harvey. Clearly, no one was exceptional in its emergency response.

        2. ilsm

          @Bruce Hall,

          We haven’t several score Higgins* boats on stand-by! I do not know how many combat engineer battalions or SeaBee units are loaded and equipped these days.

          If needed assets were known, available and port facilities and transport could be arranged………..

          Military assistance in the aftermath of a [natural disaster] is developed with the [state] national guard through DoD and coordinated with FEMA. Military commands assign Emergency Preparedness Liaison Officers (EPLO, I was one for my command in the early 1990’s) to work with the state Adjutant Generals to do two things: prepare the DoD units for surviving and prepare to develop planning for military disaster assistance for the state. All this is coordinated with FEMA and the DoD.

          There is no major active military installation in Puerto Rico an indicator that EPLO interaction with the territory national guard was limited in advance……..

          Logistics was a big issue. Lacking a military airfield; airlift to San Juan airport was used…….. that airport had limited electric power for a long time and its fuel stocks and other aviation resupply may have been a very big problem. Who could get an oiler with the right fuels to San Juan?

          How things got from San Juan to point of need out in the country was a big problem, I presume. I noted at the time military sealift (if any ready/empty capacity were in the Atlantic) was days away even if needed supplies could be marshaled and loaded quickly, and unloaded and distributed in San Juan.

          Putting a few battalions of soldiers and marines in Puerto Rico would have been a drain on supplies, transport and port capacity to provide for the troops and their equipment.

          DoD or FEMA could probably “contract” for studies, and some company winning the bid could hire some one like me to figure out what happened.

          It took the US months to marshal assets to free Kuwait………

          *Tomorrow is 74 years since D Day and Higgins boats were a key planning factor, Eisenhower knew exactly how many Higgins boats he had for the invasion and Ike competed with the Mediterranean area for them.

          1. Bruce Hall


            I laid out the “alternatives” as a thought exercise for those who (like pgl) felt that the federal government did not respond properly. I waited for his “alternative” which didn’t appear. Criticism is easy; response is hard.

  7. Moses Herzog

    For whatever it’s worth, I did think the CBS reporter overplayed his hand a little. i.e. “we” may have gotten into the self-promotion realm a little. Having said that, I still don’t think he quite presented that as “hard fact”. But then, I very rarely watch CBS News. If I watch regular antenna TV then I’m watching Judy Woodruff on PBS, and I linked their story up here a few days ago. It’s minus the melodrama and “coincidentally” mostly minus the commercials.

    It’s ludicrous, to proclaim the local government as incompetent, which it largely is, view how Trump basically gives them the middle finger at the federal level, and not think there’a an extreme number of deaths not getting “catalogued”.. To me that’s common sense. Now a person can argue “we don’t empirically measure things with ‘common sense’ ” and that’s fine and I don’t think they should be. BUT…. I think those without an agenda (i.e. just wanting to defend “the orange menace” in the White House), can clearly see we’re not talking 200-400. That range is farcical on the face of it.

    1. pgl

      I do watch CBS News perhaps to temper my watching the socialists at MSNBC and those hard core commies at TalkingPointsMemo. OK – Josh Marshall is no commie but he does know the politics of my town (New York City) even better than I do. And of course real New Yorkers loathe Trump and mock Rudy G! On 9/11, on 9/11!

      1. Moses Herzog

        Both Rudy and Hillary were taking their “star turn” after the 9-11. Neither of the two of them were any different morality wise on that one—painfully transparent in their true intentions, wandering around “ground zero” like two lost French tourists with a bad map. Hillary and Rudy—two lost Frenchmen on their way to a costume party to pretend they were part of the NYC fireman’s union. 9-11 was “photo-op” for Hillary and nothing else. Her cynical maneuver (as cynical or more cynical than McCain’s choice of Palin as a running mate) to Xerox her husband’s “I feel your pain” moment. Those were the little moves that Hillary did to doom herself, nothing else.

        Hillary can cry all day and clear to her grave about the FBI and Comey. The fact of the matter is, if she had followed the advice of intelligence officials, followed White House protocol, and strictly used the government email instead of the personal server, Comey would have had nothing to hold a press conference about. But what else do you do when people get tired of candidates who want to be patted on the head like children because of their gender while simultaneously screeching they wanted to be treated “equal”. I guess all those dirty misogynist men in rural America must have been grappling their white wives wrists as they checked Trump’s name on the ballot. It sure as hell couldn’t have been because a large cross-section of America disliked Hillary or viewed her as a pathological liar. Just Ask Debbie Wasserman Schultz.

  8. baffling

    i find it fascinating that we have folks arguing about the inaccuracy of the harvard study. doesn’t the official death toll of 64 sound fishy at all? and it is used as the basis to criticize the harvard study. think about this for a minute. the harvard study was based upon boots on the ground. people actually went out and tried to get an estimate of what occurred. the official number? a passive collection of data based upon the hope those affected turned in reliable and accurate numbers. it is simply fascinating how much animosity is directed towards the harvard work, and such passive acceptance of truth for the “official” numbers. is it because the “official” numbers have been approved by trump?

    1. pgl

      Yes – our government now tells the truth. Just like Putin’s government tells the truth! Any news to the contrary is “Fake News”!

    2. Steven Kopits

      Baffs –

      The direct deaths were those which could be directly attributed to Hurricane Maria by a medical examiner. These would include drowning and wind-related accidents like death by falling branches, for example. Such deaths would have occurred on the day of the hurricane, or from injuries sustained on that day. A total of 64 deaths have been attributed directly to the storm so far.

      Indirect deaths are essentially imputed as the difference between recorded deaths in 2017 and the same in 2016 (less the 64 attributed deaths). These people did not, however, die of the direct impact of the hurricane, and mostly not on the day of the hurricane.

      These died subsequently due to heart, kidney or respiratory failure, among others. In such an event, the medical examiner’s job is to determine the proximate cause of death as heart failure, kidney, etc. Otherwise, attribution is imputed, but not known. The indirect cause may have been manifold or remote.

      Of course, the elevated deaths themselves suggest the hurricane as a plausible cause. But consider: Based on the latest numbers, for November 2017, only 7% of the 2671 deaths can be attributed to the hurricane. But very possibly more than the calculated 187 people died of vital organ failure, and the vast majority of those who died may have lost power. So which of these do we allocate to the hurricane?

      This matters, because the government pays the funeral expenses of those who died in a natural disaster, but not otherwise. So which of the 2671 deaths do we attribute to the hurricane, when only 187 of those qualify? Or do we allow more than 187? And if we do that, should we not really list the cause of death as the hurricane for all those who lost power and died of natural causes?

      And this gives us a clue as to the likely mandate of the Milken Institute. They are trying to piece together the picture to permit an allocation of the cause of death. The govt may have resisted publishing updated numbers in order to maintain maximum flexibility — but the Harvard number was so large that it effectively forced the government’s hand.

      What a mess! No matter how the issue is adjudicated, someone’s going to be unhappy.

      1. CoRev

        Bffled and Steven, “…the Harvard number was so large that it effectively forced the government’s hand. ” Evidently there were several court cases and court orders in play by the or as a result of the Harvard study. I read several references to the court orders forcing the PR release of the latest data and Baffled I don’t intend to chase those URLs now.

        What I have noticed after these many days, articles and comments, there have been few if (caveat due to old memory cells) any recommendations for improving the response then and in the future.

        1. Steven Kopits

          I think the controversy surrounding the Harvard study will material in the timing of the news release.

      2. baffling

        steven, the big issue folks are trying to get you to understand, is that for the coroner to make an “official” declaration, they need access to the body as well as information related to the persons demise. for months after the storm, resources were not available to make that happen. power and water were not available, and most of the transportation system was shut down due to loss of roads and hundreds of bridges. it is not reasonable to think that “official” numbers could be remotely accurate given the inability to fulfill the needs of the coroner to make a hurricane death declaration. hundreds, if not thousands, are still considered missing on the island. i doubt any of those missing will ever end up in the “official” numbers. look, harvey is credited with about 100 deaths in the us. maria was a much more formidable events. i have my doubts that it had a lower loss of life than harvey.

  9. Steven Kopits

    About the graph:

    1. The 5/31 number was withdrawn based upon updated figures from PR authorities last Friday
    2. You can show the PR authority’s line all the way through March — these are known numbers presumed to be firm
    3. The correct number for PR official M4 is, I believe, 1600, not 2000.

  10. Steven Kopits

    “Hence, my view is that contentions that a “firm” count has been achieved within a year of a major and ongoing disaster is unwarranted. I also think that people should learn the nature of the data sources they are using, and what survey design involves, what sampling error is, and what measurement error is, before undertaking the difficult task of policy analysis.”

    This statement is pure nonsense.

    The Harvard study closed its analysis period at Dec. 31, 2017. That is the period under examination.

    To that date, we have 1400 excess deaths officially.

    If there are additional bodies to be counted then, they must

    1. Yet to be discovered
    2. Yet to be processed
    3. Yet to be recorded
    4. Yet to be reported
    5. Yet to be disclosed

    Steps 2-4, as we know from revisions to the historical data, take about 10 days. That leaves undiscovered bodes and conspiracy by the PR government to hide the body count. Which is it, Menzie?

    1. Steven Kopits

      To clarify: If excess deaths do not materially occur within a year, then will be hard to see in the data.

  11. Steven Kopits

    Does Harvard stand behind the study, or not?

    That is, does Harvard SPH believe that the central estimate of excess deaths to 12/31 is 4645, or not? Does it stand behind the confidence interval, or not? Is there still a 50+ probably that the death toll comes in over 4600? If there is, then the people of PR need to start looking for the 3,250 missing or the press needs to assume PR authorities are lying. Those are the implied action items.

    Or should we just take whatever number HSPH publishes in the future and divide by 3 to get a realistic estimate of actual?

  12. joseph

    Kopits is like most Republicans — he hates data. Whether it’s economics or climate or hurricane deaths.

    The Harvard study is what it is. Unless you can document deception in the methods, then there is nothing to withdraw. The “official” estimates are now within the confidence interval.

    It is fine to discuss the weaknesses of the study, which the authors do themselves. The sample size was small, so the confidence interval was large — but as seen, within the official estimates. There could be sampling bias (and I hesitate to use the word bias because people like Kopits don’t understand what the means in statistics). But there is nothing nefarious going on. The data and methods are fully transparent. The data speak for themselves.

    The way to improve the results is to take more data.

    And never forget, Kopits estimate of 200-400, on which be based his “garbage” statement was much worse than the Harvard results.

    And also never forget that Kopits maintains the perverted view that killing off the weak early provides the benefit of negative excess deaths in succeeding months. He’s monstrous.

    1. pgl

      “Kopits is like most Republicans — he hates data. Whether it’s economics or climate or hurricane deaths.”

      Amen brother. Memo to self – skip all comments from Kopits going further.

Comments are closed.