38

I was reading about Ron Desantis on Wikipedia and was a little surprised by the section about his handling of COVID-19 in Florida. It lists a huge number of actions to oppose standard COVID preventative measures, including outlawing mask and vaccine mandates and minimal shutdowns of the state. It also claims that Florida had some of the highest death tolls in later parts of the epidemic.

However it also has these two statements:

Florida's death rate from COVID-19 (75,000 deaths) ended up being within the national average and Florida's economy fared better than many other U.S. states

and

in May 2022, a Bloomberg News op-ed claimed that, when adjusting state death tolls based on what they would be if age distribution were equal between the states, Florida's COVID-19 death toll would be less than the national average

I'm skeptical that Florida could average better than the national average while resisting all precautions. A quick search found Statistica.com which lists Florida's current per capita death toll as of this month as the 14th highest in the USA, with a death rate of 426/100,000. By contrast the John Hopkin's Coronavirus Resource Center shows the current average death rate for USA as 329.68/100,000, which would seem to refute at least the first claim I quoted, unless there are differences in how death rate or average is being measured.

The one potential discrepancy I could see is that the Wikipedia links are a bit older, so Florida may have caught up/exceeded the national average since then as it's still a hotspot for COVID as I understand? But it's hard to believe a few months could make that huge a difference in numbers.

So, is Wikipedia's claim that Florida managed as well, or better, then the national average true at the time the links were cited, and is it still true now?

1

4 Answers 4

60

The source provides no methodology.

tl;dr The claim that "if age distribution were equal" is not complete. The original source adjusts COVID mortality by age and an ill-defined "metabolic health". They provide no methodology for either. Without knowing the details, they could have done any amount of monkeying with the numbers.

The results are meaningless.

The Britannica Claim

Florida's death rate from COVID-19 (75,000 deaths) ended up being within the national average and Florida's economy fared better than many other U.S. states.

This is from a Britannica article about Ron DeSantis and provides no citations how they concluded Florida is within the national average.

The Bloomberg Claim

in May 2022, a Bloomberg News op-ed claimed that, when adjusting state death tolls based on what they would be if age distribution were equal between the states, Florida's COVID-19 death toll would be less than the national average

This is from an op-ed by Justin Fox. He is a Bloomberg Opinion columnist covering business, not health. The same op-ed was posted in the Washington Post with no paywall.

Hiding inequality in averages

The relevant claim appears to be this:

Adjust state Covid mortality statistics to what they would be if all the states’ populations had roughly the same age distribution as the nation — as conservative activist and policy analyst Phil Kerpen has been doing for much of the pandemic — and Florida’s rate drops to 275 per 100,000, well below the national rate of 302, while California’s rises to 267.

Phil Kerpen is a founder of the Committee To Unleash Prosperity. More on them in a moment.

We must address whether ignoring a state's demographics is a relevant statistic. It smacks of the "sacrifice the elderly" modest proposals of the early pandemic.

Population demographics are not a hypothetical. Those are real people, really getting sick, and really dying. Instead of arguing that Florida would have done better with average demographics, one can argue Florida failed to adapt their policies to their population.

It's very tempting to boil everything down into a single number for the "average person", but there is no such thing as an average person. Populations are messy. For example, (50, 50, 50, 50) and (-100, 100, 200, 0) are very different sets of numbers, but they average (mean) out the same. Abusing population averages is a way to hide inequalities.

I would argue that if we really want to compare states without being biased by their age distributions, we shouldn't be mashing all ages together into one statistic. We should be looking at COVID mortality rates by age. This removes the age distribution bias without artificially smoothing out bumps. We should compare people apples-to-apples, not applesauce-to-applesauce.

That's just what this chart already does (sorry about the poor quality)...

enter image description here

...we can see that, relative to California, Florida did as good or slightly better with their 75+ population. However, they did far worse with their younger populations. That is more meaningful information than one number for all ages.

One could argue that doing worse in younger populations is less relevant because their death rates are orders of magnitude lower than the 75+ rates, but this all teeters on the verge of devaluing lives. Devaluing the elderly by ignoring the demographic differences and devaluing the young by ignoring their higher death rates in Florida.

A Final Report Card on the States’ Response to COVID-19

This claim appears to be sourced from A Final Report Card on the States’ Response to COVID-19 by Phil Kerpen, Stephen Moore and Casey Mulligan of the Committee To Unleash Prosperity. The numbers don't exactly agree, maybe they were using an earlier version, but they are very close.

The Committee To Unleash Prosperity is an organization to promote supply side economics. They are anti-regulation. Its founders are all economists, not public health professionals. All the authors of the paper are economists, not public health professionals.

Critically, there's little discussion of their methodology; no formulas are provided. However, while the article only mentions age, the report also includes "metabolic health"...

We adjust COVID mortality (through March 5, 2022) for age and “metabolic health,” by which we mean the pre-pandemic prevalence of obesity and diabetes – as these are highly correlated with higher death rates from the virus.

There is no methodology about how they calculated or applied "metabolic health". How did they determine the prevalence? How did it affect the COVID mortality rate? Why only obesity and diabetes? Why not respiratory problems? Auto-immune diseases? Without that information, we can't check their math; their adjusted COVID mortality rates are meaningless. They could have used it as a fudge factor, tweaking the method until they got the result they wanted.

With all this in mind, and the lack of methodology, the source cannot be used.

1
  • Comments are not for extended discussion; this conversation has been moved to chat.
    – Jamiec
    Dec 22, 2022 at 10:42
14

Florida's economy fared better than many other U.S. states

No one addressed the economic angle, so I'll take a stab. The best source of economic performance in each state is the Bureau of Economic Analysis. If we evaluate the annualized change in GDP between Q4 2019 and Q2 2022, we get the following results:

  • US as a whole: +1.4%
  • Florida: +3.5%, ranked #4 among the 50 states
  • Mississippi (had the highest COVID death rate per capita): +0.8%, ranked #35
  • Hawaii (lowest death rate per capita): -2.3%, ranked #45
  • Vermont (second lowest death rate): +1.6%, ranked #19

Economically speaking Florida is the 4th best performing state in the country, so that part of the claim is definitely true. The three states ahead of Florida (Idaho, Tennessee, Utah) were also notable for having relatively lax COVID regulations.

Update: it was suggested in the comments that historical patterns matter, because it's possible that Florida was already growing very rapidly before the pandemic. Looking at the annualized change in GDP between Q2 2017 and Q4 2019 we see the following data:

  • US as a whole: +2.7%
  • Florida: +3.0%, ranked #15 among the 50 states
  • Mississippi: +0.6%, ranked #46
  • Hawaii: +0.4%, ranked #49
  • Vermont: +0.9%, ranked #40

So Florida went from #15 to #4 in the ranking of fastest growing states. So it becomes clear that Florida really did do great economically, even compared to its previous growth.

15
  • 8
    How does that compare to non-Covid years? It's hard to interpret the numbers without knowing if in other years they had a GDP growth of about the same, more, or even less.
    – Laurel
    Dec 16, 2022 at 21:16
  • 6
    @Laurel past performance is no guarantee of future results. We can't just extrapolate pre-COVID growth rates and pretend like they would've continued forever. Dec 16, 2022 at 21:23
  • 3
    It might not be a guarantee of future results but if Florida typically ranks in the top for growth then saying it ranked in the top for growth during covid doesn't mean as much.
    – Joe W
    Dec 16, 2022 at 23:02
  • 4
    @JoeW For growth over 2016-2019 it ranked at #11, so moving up to #4 is an improvement. If Covid deaths had any influence whatsoever, it’s too small to notice in the data. Dec 16, 2022 at 23:04
  • 4
    This is totally out of any context with the pandemic. The pandemic did not strike all states equally or at the same time. I also wonder if this is impacted by using RDP values which includes personal income. For a state like Florida with a high number of wealthy, retired people it's likely less volatile other states.
    – MikeJ
    Dec 17, 2022 at 19:09
13

I had enough comments that they ended up forming a supplementary answer on their own.

So, is Wikipedia's claim that Florida managed as well, or better, then the national average true at the time the links were cited, and is it still true now?

So, to start, it should be noted that the article claimed "within the national average", not "below", so when you state earlier on:

I'm skeptical that Florida could average better than the national average while resisting all precautions.

you're skeptical of a claim the source does not make.

We can try to verify its position relative to the national average by narrowing the dates. The Brittanica source's own article history shows that that claim was in an article posted on 31 August 2022. The same claim about being "within the average" notes that Florida had had "more than 75,000 [deaths]", and checking a chart of total deaths over time in Florida, the first time it had registered >75K deaths was 02 June 2022. That means that the "within the national average" claim would be based on stats from between 02 June and 31 August of 2022.

Using 2022 population figures of 22.1M for Florida and 338.3M for the whole of the U.S., and the same cumulative deaths source, Florida began that period with 75,016 deaths (339.4 deaths/100K), and ended on 31 Aug with 80,446 (364.0 deaths/100K). The U.S. as a whole grew from 1,036,634 deaths (306.4 deaths/100K) to 1,074,162 (317.5 deaths/100K). Florida's per-capita death toll was 10.8% higher than the U.S. average at the beginning of the window, and grew to 14.6% higher by the end of the window (likely due, at least in part, to Florida's choice to limit mitigations; at least today it's slightly more vaccinated than the U.S. average, at 69% fully vaccinated to the U.S. average of 68%). This doesn't put it outside a single standard deviation from the mean for "worst states", so it could be said to be "within the national average", but it was by no means "below average" at any point when that data could have been collected, and was definitely a non-trivial amount above average.

At the time I'm writing this, it has the 14th highest total per-capita death rate (out of 56; the 50 U.S. states, D.C. and the five U.S. territories with meaningful population). It was probably a little better back in the summer (it's had a persistently higher death rate than the average for a while), but still nowhere near below average. The most recent numbers for 15 Dec 2022 from the same source for the cumulative deaths over time number are larger, but the percentage gap between Florida and the U.S. is largely unchanged (still rounds to 14.6%, though NYT's data shows a greater gap, 387 vs. 330, 17.3% higher for Florida than the U.S.) so the accuracy now is not meaningfully different than it was at the latest possible data the claim could have based its claim on.

As others have already noted, Bloomberg's op-ed provides no methodology for their adjustments, so it's impossible to verify, and given the source, it's suspicious. On age alone, Florida is one of the oldest states:

Florida is one of the oldest states

but of the four states with older populations (from the map, WV, VT, NH and ME), only one has a higher per-capita COVID death toll (WV, with a much worse vaccination rate of 59%) and three of them have significantly lower death tolls (VT, NH and ME, with per-capita death rates 125, 209 and 208 respectively) (source). While VT and ME have much higher vaccination rates than Florida, NH is comparable to Florida (at 70% to Florida's 69%), so Florida merely having an older population is not sufficient to explain their worse performance.

1

According to reports in the media, Florida changed the way that deaths were reported to reduce the headline mortality rate.

The Miami Herald found that until three weeks ago, information gathered by DOH counted deaths by the date they were recorded. But on August 10, the state switched to tallying new deaths by the date an individual died, which can cause a lag in reporting because those deaths need to be evaluated and certificates need to be processed first.

Florida's New COVID Reporting Method Shows 'Artificial Decline' in Latest Data

10
  • 6
    The change seems for the better and shouldn’t impact the overall numbers. Dec 17, 2022 at 19:34
  • 4
    @JonathanReez, the changes included removing people whos deaths were handled out of state, even if they occurred in state. Dec 17, 2022 at 19:38
  • 4
    It might be helpful to add that last fact to your answer. Dec 18, 2022 at 20:50
  • 1
    @JonathanReez: The other issue the change could cause (besides the change excluding some deaths) is that, given people continue to die at meaningful rates, any comparison at a fixed point in time of reported deaths, where the other population being compared does not delay reporting, will favor Florida. The slant towards Florida asymptotically approaches zero (the delayed deaths become a smaller fraction of the total), but it matters; Florida has been around 24-30 deaths/day recently, so if the lag is 30 days, they'll exclude 720-900 deaths (~1% of their total deaths to date). Jan 6, 2023 at 15:53
  • 1
    @ShadowRanger cool but we’re so far into the pandemic now that it’s a negligible difference (1% as per your own calculation) Jan 6, 2023 at 15:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .