When the New York Post reported on March 31 that New York Gov. Andrew Cuomo announced 43,139 coronavirus cases in New York City, while Mayor Bill de Blasio was claiming 40,900, I decided to investigate this significant 2,239, or 5.5 percent, discrepancy.
Unlike many journalists who have been writing or speaking about this epidemic, I am a mathematician who taught in New York City public high schools between 1990 and 2009, and I am trained to detect flawed statistics.
The website of the New York City Department of Health (NYCDOH) has two data sets for coronavirus deaths. The first — in a table form that was released daily in the morning and afternoon until April 12 — provides the total number and demographics. On April 14, the NYCDOH switched to a much different format, which counts “confirmed” and “probable” coronavirus deaths.
The second data set is an interactive bar graph, which only displays the number of daily deaths when you click on a bar for a particular day in March or April.
While the twice-daily fatality tables are cumulative and don’t display how many New Yorkers died each day, the bar graphs just show the daily number of deaths, but not the total number.
Around April 3, when de Blasio somberly announced that in the last 24 hours “305 New Yorkers died from the virus,” I began to doubt the scientific integrity of the city’s official statistics. The tables for the morning and the afternoon of April 3 show an increase in deaths from 1,584 to 1,867, or 283 deaths, which are 22 fewer than the mayor announced that day.
But I suddenly realized that de Blasio’s NYCDOH was reporting the number of deaths that occurred from April 2 at 4:30 PM, which is 1,562, to 4:30 PM on April 3, which is 1,867, which does equal 305 deaths.
But when I saw the death toll increase from the afternoon of April 2, when it was 1,562, to the morning of April 3, when it was 1,584, I immediately realized that the NYCDOH was reporting, for this overnight 17-hour period, only 22 deaths. Indeed, by the next reporting time, 4:30 PM on April 3, the death toll had soared to 1,867, meaning 283 deaths had occurred in the next seven hours.
It is epidemiologically impossible for the coronavirus death toll in 24 hours to increase from a little more than one death per hour during the first 17 hours, to almost 38 deaths per hour over the next seven hours. A contagious-disease epidemic in one of the world’s greatest cities is not a wartime battlefield, where such extreme variations in hourly death rates often occur.
I next examined the death tolls between 5 PM on April 4 and 9:45 AM on April 5, and I found only two additional fatalities recorded, from 2,254 to 2,256, during these 17 hours. But at 4:45 PM on April 5, or seven hours later, the death toll had again skyrocketed, to 2,472, or a 216 increase.
Again, a deadly epidemic in a city with 8.4 million residents, whose living space is only 303 square miles, will not have two deaths in 16 hours and 45 minutes, and then in the subsequent seven hours and 25 minutes have 216 deaths.
Upon further scrutiny of the NYCDOH’s fatality data, I discovered an even more egregious flaw, this time in the interactive bar graphs. In the evening of April 3, I printed the NYCDOH’s bar graph and then recorded the number of deaths at the top of each day’s bar. Over the next several days, I printed other bar graphs and noted each day’s grim toll, since the first deaths — two — occurred on March 14.
But I soon recognized that on the bar graphs, which I printed between April 3 and 7, the daily death numbers for previous days had been significantly increased. Conversely, the twice-daily tables for fatalities were not adjusted between March 22, when they were first published and April 12.
The following table highlights the huge changes in daily fatalities from March 28 through April 2 on bar graphs I printed on April 3 and 7:
|
Evening, April 3 |
Evening, April 7 |
3/28 |
182 |
220 |
3/29 |
192 |
241 |
3/30 |
191 |
254 |
3/31 |
204 |
309 |
4/1 |
181 |
311 |
4/2 |
131 |
325 |
TOTAL |
1,081 |
1,660 |
Thus, between the evenings of April 3 and 7, the NYCDOH’s bar graphs added a mind-boggling total of 579 deaths for six consecutive previous days, with the additions ranging from 38 deaths for March 28 to 194 deaths for April 2. This massive alteration of the dates of death is medically incomprehensible, since almost all coronavirus deaths occurred either in New York City hospitals, nursing homes, or private homes and apartments, where relatives, friends and/or neighbors were present.
Furthermore, I printed two other fatality bar graphs, on the evenings of April 6 and April 11, to determine how many deaths were added to the previously published totals from March 30 through April 4:
|
Evening, April 6 |
Evening, April 11 |
3/30 |
220 |
275 |
3/31 |
262 |
324 |
4/1 |
225 |
363 |
4/2 |
225 |
392 |
4/3 |
217 |
410 |
4/4 |
190 |
398 |
TOTAL |
1,339 |
2,162 |
For these six days, de Blasio’s NYCDOH increased the total deaths by 823, and the largest adjustment is 208 deaths on April 4.
Having studied daily death tolls in the official tables and bar graphs for more than two weeks, I have concluded that they are worthless for rigorously documenting the trajectory of the coronavirus epidemic at its epicenter in America: New York City.
These dubious changes to death tolls didn’t only occur in New York City. On April 2, 2020, at 11:39 PM, I printed an article from The New York Times that listed New York State’s fatalities by counties, which was stated as 2,576 statewide and 1,562 in New York City.
I immediately questioned that 1,014 fatalities had occurred in the state outside the city. Indeed, when I totaled the counties outside the Big Apple, I found only 320 deaths. Adding these deaths to those that occurred in the city, 1,562, brought the total coronavirus deaths in New York State to 1,882. I immediately began to search in which counties the missing 694 fatalities (2,576 minus 1,882) had occurred.
I quickly struck epidemiological gold, as the coronavirus website of Johns Hopkins University had a ranking of the 6,069 deaths in America by county. New York City’s five counties (Bronx, Brooklyn, Manhattan, Queens, Staten Island) had the largest fatality toll, at 562;
No. 3 was Wayne County, 194, and No. 4, King, 175, in Washington State; and New Orleans, Louisiana, ranked No. 5 with 125 deaths.
Ranking No. 2 was New York State, with 516 deaths, which were “unassigned” to any county. Again, I was flabbergasted that there could be such uncertainty about where a large number of the coronavirus deaths occurred. However, the website of the NYS Department of Health has two geographic designations — county of death and county of the deceased’s residence — and there are understandable reasons for this dual designation. Many residents of Brooklyn, Queens, and the Bronx prefer hospitals in Manhattan, since they are affiliated with the borough’s medical schools: Columbia, NYU, Mount Sinai, and Cornell.
But, lo and behold, when I checked the state DOH’s website two days later, on April 4, I found that all of the state’s 3,565 coronavirus deaths, including the 516 “unassigned” from the Johns Hopkins website, were now “assigned,” and New York City had 2,624.
On April 2 at 4:30 PM, the NYCDOH fatality tables reported 1,562 deaths, which matched the total in the perplexing New York Times article, published on the same day, which gave the death tolls in each state county.
But the NYCDOH’s table for April 4 at 5:00 PM lists 2,254 deaths, which are 370 deaths fewer than the 2,624 than New York State’s official coronavirus website reported for New York City. This was another major discrepancy between the official city and state death tolls, as the one I mentioned in this article’s first sentence, that are epidemiologically unacceptable.
Finally, when I was almost finished with this article in the afternoon of Wednesday, April 15, I discovered an even more blatantly unscientific “adjustment” than the egregious ones made to the city’s daily death graphs.
The NYCDOH’s “COVID-19 Data” page now has created a truly Orwellian new type of coronavirus death, which is labeled “probable.” The NYCDOH explains a “probable” coronavirus death as:
“A death is classified as not known to be confirmed or probable COVID-19 if the decedent died in New York City (NYC resident or residency pending) had no known positive laboratory test for SARS-CoV-2 (COVID-19) and the death certificate does not list COVID-19 or an equivalent as a cause of death.”
In other words, anyone who died in New York City during the last month can now be reclassified as a coronavirus death.
Indeed, on April 15, the NYCDOH’s coronavirus data page listed 6,840 “confirmed” deaths and 4,059 “probable” deaths, or a total of 10,899. These totally unscientific death tolls were tallied as of 6 PM on April 14, and the “probable” deaths represent a 59% increase to the “confirmed” deaths.
Incredibly, the same data page also declares that, at 4 PM on April 14, “deaths in NYC reported by NY State” are 7,905. But as I already documented, the NYCDOH’s official website reported 2,254 coronavirus deaths in NYC as of April 4, while the same day the state DOH’s official website was reporting 2,624.
Thus, in 11 days, the official death tolls went from the state’s total being 370 more than the city’s, to the city’s total death toll reporting 2,994 more deaths than the state had (for the city).
President Donald Trump, in his daily press conference on Wednesday, April 15, discussed and accepted the recent large increase in recently reclassified coronavirus deaths in his and my hometown. But the president has not heeded the wise counsel of President Ronald Reagan: “Trust but verify.”
In conclusion, the death tolls for the epicenter of the 2020 coronavirus epidemic in America, which the NYCDOH has been disseminating daily since March 22, are epidemiologically worthless. If we just accept New York City’s 6,840 “confirmed” cases and exclude the highly suspicious 4,059 “probable” cases, the city has suffered 24% of the nation’s 28,529 deaths (as of 9:30 PM Eastern Time on April 15).
With 327 million American heroically battling this dastardly epidemic, President Trump and his advisers have unconscionably been deprived of precise reports of deaths in New York City, as a result of the abominable leadership of Mayor Bill de Blasio and, to a lesser extent, Gov. Andrew Cuomo.
Therefore, I am advising President Trump to end the national shutdown, and to return Americans to their normal, intensely productive lives. The president should also instruct Attorney General William Barr to appoint a special federal prosecutor to investigate the health departments of both New York City and State, for regularly releasing inexcusably shoddy data about the daily and cumulative death tolls during this catastrophic — in lives and livelihoods — coronavirus epidemic.
Mark Schulte is a retired New City schoolteacher and mathematician who has written extensively about science and the history of science.
© 2025 Newsmax. All rights reserved.