A week ago I posted about the difference between red and blue states when it comes to COVID-19. It attracted quite a bit of attention, including requests to update the results. Happy to oblige.
The “red” states have republican governors, the “blue” states have democratic governors.
This blog is made possible by readers like you; join others by donating at My Wee Dragon.
Open Mind 
A demurral: The plot understates the variability among Red states or Blue states, something which might be better portrayed using a prediction interval about them. I only say this because the point of the post is, ultimately,
vs
for a fixed risk level
.
See their estimated
values, in contrast. Note that chart provides the Bayesian Highest Probability Density Intervals (HPDI) at 0.80. They are wide enough so there could be some inversions. But it also tells a tale which differs from the above: While Massachusetts and New Hampshire might be lumped into the Red state category, their 
values are pretty good, even compared with New York State.
This understates the case. “Red” state Massachusetts is not a red state in any meaningful sense of the word. It has a moderate Republican governor who is doing very well. Look at the numbers:
373 new cases yesterday.
Numbers in the small hundreds since June 2 and going down (looks like a small uptick now from the protests,* maybe). There was a major adjustment in counting on June 1 which looks like a spike but is a correction. Have a look:
https://www.worldometers.info/coronavirus/usa/massachusetts/
This results in a distortion that skews the comparison.
*Interestingly, when I think about that, the fact that overall cases are down made the spike from the protests much much smaller. And people wear masks!
Similar with Maryland, not quite as good stats as MA though.
@Susan Anderson: Perhaps, it would be interesting to classify states on criteria whether they elected Trump or not in 2016 (and I doubt if it is worth extra work). But you want to have simple criteria in the end and you want to avoid using too complex scenarios as you can easily introduce personal confirmation bias and you would start classifying states on criteria that they are successfull or not – which is an extremely bad idea.
Susan,
I agree and kinda said so, too. But Massachusetts needs to be watched. They are having a hard time enforcing distancing and the like on Cape Cod, and people from all over flock there. The example of New York State is telling: An outbreak occurred as a result of a high school graduation in Oswego where a relative came in from Florida, did not quarantine for 14 days. Don’t know if it’ll get enforced, but that relative now owes New York State a fine of $10,000.
Massachusetts has a $300 fine for breaching the 14 day quarantine rule which is unevenly enforced. Yarmouth just had an instance where the Commonwealth liquor commission found a restaurant operating their bar without social distancing, and they closed them down and fined them. Turns out the Selectboard of Yarmouth, under some business pressure, told them all they could take their chances with the Commonwealth, and they would not enforce the Governor’s dictate. It could be political gossip, but there’s some claim the Selectboard pressured Yarmouth Health Department on the point. Health Departments are supposed to be independent of local government. The Health Department has re-grown a spine, let’s say. That may, in part, be due to Lt Governor Polito being made aware of the situation: She has a summer home in Yarmouth.
The Town of Sandwich also tends to not like being “pushed around” by Boston.
We’ll see.
Yes, I have traveled regularly between Boston and Princeton, due to family affairs, and nobody has challenged me in any way. My mother’s nursing home in Princeton has had 24 deaths and 80 active cases (now COVID free) and I’ve been in regular contact with her aides (outdoors only). I’ve been very careful, but have yet to have any contact with anyone on the enforcement side.
And I am aware of the urban/rural MA divide and the resentment therefrom, not to mention people’s desperation to make a living, given the federal government’s preference for/insistence on giveaways to those who have rather than helping those who don’t.
But my point is, personal anecdotes notwithstanding, that if you remove the good citizens from the “red” side, the contrast is even more dramatic.
Oops. ERROR, big ‘un. 80 positive tests, not “active cases”.
I’m introduced here from the dataisbeautiful subreddit. I agree with others that Baker and Hogan are not taking the typical republican approach to covid. Perhaps it might be interesting to group Massachusetts, Maryland, New Hampshire, and Vermont with the blue states for the next update in addition to the normal blue state/red state selection. Results may be interesting to compare.
Neoliberal politics (Republican politics now) kills people.
They like to say that they will “die for their freedom” but the right to put everyone else’s life in jeopardy on a matter of whether or not to wear a bit of cloth over their faces is not exactly what the founding fathers had in mind.
It really is that simple.
Yes, stupid is stupid, no matter how legal. And even stupider to then claim in the face of innumerable extant counterexamples that regulation is impossible or invalid because “freedom.”
We really do not want to be in a position where “everything not forbidden is compulsory,” but that would be the logical outcome of this anti-masker nonsense–in so far as the term ‘logical’ can be made to apply in this context at all.
@BJChip,
I also sense, given correspondence and conversations with town health agents in contrast to epidemiologists at state level and in university, that the agents are donning a mix of:
(1) “What do you expect me to do?”
(2) “It’s only an advisory. People have to make up their own minds.”
(3) “We can’t regulate what people do in their private spaces.”
(4) “Cape Cod is doing a tremendous job at social distancing and mask wearing! *wink*, *wink*”
On #3, well, there are plenty of other things within homes health agents and committees are empowered to regulate (sanitation, water quality, vaccinations, dealing with dead bodies, etc).
On #4, fill in your local case in place of “Cape Cod”.
O-T comment. I failed to find how to contact, so here goes.
Today (30th June, 2020) on BBC website:
Flu virus with ‘pandemic potential’ found in China
A new strain of flu that has the potential to become a pandemic has been identified in China by scientists.
It emerged recently and is carried by pigs, but can infect humans, they say.
The researchers are concerned that it could mutate further so that it can spread easily from person to person, and trigger a global outbreak.
While it is not an immediate problem, they say, it has “all the hallmarks” of being highly adapted to infect humans and needs close monitoring.
As it’s new, people could have little or no immunity to the virus.
The scientists write in the journal Proceedings of the National Academy of Sciences that measures to control the virus in pigs, and the close monitoring of swine industry workers, should be swiftly implemented.
Continues …………..
https://www.bbc.co.uk/news/health-53218704
Refers to:
“Prevalent Eurasian avian-like H1N1 swine influenza virus with 2009 pandemic viral genes facilitating human infection”
https://www.pnas.org/content/early/2020/06/23/1921186117
livinginabox,
But you don’t complete the story so I’m not sure of the reason behind your comment.. You leave unsaid that, even with the somewhat-late delivery of vaccines that were less than perfect, 2009 Swine Flu proved to have a low Case Mortality Ratio of 0.03%.
I suspect your comment may be uncalled-for, because:
Part of the Abstract from the article linked-to by Honglei Sun, et al. states:
“Moreover, low antigenic cross-reactivity of human influenza vaccine strains with G4 reassortant EA H1N1 virus indicates that preexisting population immunity does not provide protection against G4 viruses. Further serological surveillance among occupational exposure population showed that 10.4% (35/338) of swine workers were positive for G4 EA H1N1 virus, especially for participants 18 y to 35 y old, who had 20.5% (9/44) seropositive rates, indicating that the predominant G4 EA H1N1 virus has acquired increased human infectivity. Such infectivity greatly enhances the opportunity for virus adaptation in humans and raises concerns for the possible generation of pandemic viruses.”
I would think running the numbers by county would be more accurate, though I guess that’s a much bigger task. CA, for example, is obviously predominantly blue, but there are a lot of small red counties which are now seeing a dramatic rise in cases.
Sounds like a logistic regression problem, although I’d/I’ll do it as a classification problem for a random forest. Anyone know where I can get a tabulation of counties vs Red or Blue status? Otherwise I’ll dig around on the Five Thirty Eight GitHub and see if they have one.
@Rob Honeycutt,
I didn’t find at Five Thirty Eight, but I did find at MIT. Also pulling county income data.
re Massachusetts
https://www.boston.com/news/coronavirus/2020/06/30/massachusetts-out-of-state-travel-quarantine-rules
Hi
Very interesting analysis. Can you change the y-axis to % vs 1st peak and also analyze changes among matched states based on absolute deaths or infected?
Yesterday (7/1/20) in the northeast (Worldometers version, not perfect but reasonably accurate), new cases/deaths:
NY 769/14
PA 636/39
MD 359/15
NJ 285/41
MA 261/27
CT 58/2
ME 41/0
DE 36/0
NH 20/2
VT 2/0
I have completed a random forests look at county responses and some economic, political, and explanatory variables for COVID-19 incidence. The details are available here.
In the conclusion most of the covariates did not show a monotonic response. Of those which did, some had surprising effects upon response, contrary to what would be thought. In all cases, it shows that whatever is happening at state levels, as Tamino originally hypothesized, the story at county levels are far more nuanced.
According to the regression, the increases in the following covariates are associated with increases in COVID-19 incidence:
* Number of votes for non-major party presidential candidates in 2016.
* Number of votes for non-major party House candidates in 2016.
* Hispanics or Latinos as a percentage of total population, 2016.
* Personal per capita income, 2018.
* Percent change in per capita income from 2017.
According to the regression, increases in the following covariates are associated with decreases in COVID-19 incidence:
* Number of votes for Democratic party House candidates in 2016.
* Non-Hispanic blacks as a percentage of total population, 2016.
* Unemployed population in labor force as a percentage of total population in civilian labor force, 2016.
* Population with an education of less than a regular high school diploma as a percentage of total population, 2016.
* White population with an education of less than a regular high school diploma as a percentage of total population, 2016.
* The ratio of votes in the county for presidential candidate Trump in 2016 to the sum of Obama and non-major party candidates in 2012.
The remainder have non-monotonic effects upon response, with the exception of votes for Trump in 2016, votes for Romney in 2012, and the ratio of Trump votes to sum of Clinton and non-major party Presidential candidates in 2016. The latter three had no significant effect upon response.
These are difficult to interpret, certainly from any common wisdom conventional frame. A story could be told about propensity for supporting non-majority political parties as contrarians. A story could be told about wealth and especially recent wealth. And a more conventional story might be told about support for Democratic House candidates. But the analysis is too weak to build much of a case here. For why is county historical support for Trump over Obama and others associated with decreases in COVID-19?
Another analysis, perhaps using a Bayesian additive regression trees or a generalized additive model might reveal more. I have included the result of a straight regression at the bottom of the post.
The evidence indicates that factors contributing to increases in COVID-19 across counties are more complicated than state political trends suggest.
A short follow-up to the above cited post … When counties having the most extreme upturns and downturns in case rates of COVID-19 are chosen, the
for a standard linear regression improves from 0.16 to 0.47, and that for the random forests regression improves from 0.54 to 0.71.
There was no change seen in the predictors which contributed to increases in COVID-19 incidence, but there were changes in the predictors associated with decrease in COVID-19 incidence, measured as numbers of confirmed tests.
The former were, in both cases, for such counties, when
* Number of votes for non-major party Presidential candidate in 2016
* Number of votes for non-major party House candidate in 2016
* Percentage of population that’s Hispanic, where this is defined in the database’s data dictionary
* Per capita income of county, 2018
* Percent change in personal income in county from 2017 to 2018
go up, there is an associated propensity for COVID-19 incidence to go up as well.