Republican State Legislatures are Positively Correlated with Obesity, Democrat Legislatures with Death from #Coronavirus

Note from Charles. The original title of this piece was:

Do Lockdowns Reduce the Spread? Maybe a Little. But what else do they do?

However, the data also lead to the new title.

End note.

Guest post by David Stienmier, PhD (Mechanical Engineering)

Case and mortality data from Johns Hopkins (https://github.com/CSSEGISandData/COVID-19/tree/master/csse_covid_19_data/csse_covid_19_time_series)

Population estimates for 2019 from the US Census (https://www.census.gov/data/datasets/time-series/demo/popest/2010s-counties-total.html#par_textimage_70769902)

Unemployment data from the Bureau of Labor Statisics (https://www.bls.gov/)

Partisan makeup of state legislatures from the National Conference of State Legislatures (https://www.ncsl.org/Portals/1/Documents/Elections/Legis_Control_2020_April%201.pdf)

I’ve had this data for a month or so now and I’ve been trying to decide what to say about it. I assumed someone else would show this somewhere first and I could resume my quiet observation. But I haven’t seen it anywhere and with everyone talking about locking down again I decided I should at least put it out there with minimum comment. So here it is.

Lockdowns are intended to reduce the spread of cases and therefore the number of deaths. A known side effect of the lockdowns is an increase in unemployment. So we should be able to use the increase in unemployment as a proxy for how hard a state locked down. Figure 1 shows how cases relate to lockdown intensity as measured by unemployment. There appears to be very little relationship, but maybe it reduces the number of cases slightly.

Figure 1: Cases as of 10/7/202 vs Increase in Unemployment from the Three Months Ending March, 2020 to the Three Months Ending August, 2020. Case Data from Johns Hopkins University, Unemployment Data from the Department of Labor Statistics

Unemployment, however has long been associated with poor health outcomes so it’s reasonable to ask if this is the case with Covid. Figure 2 shows total Covid deaths vs unemployment increase. Do I need to comment?

Figure 2: Covid Deaths as of 10/7/202 vs Increase in Unemployment from the Three Months Ending March, 2020 to the Three Months Ending August, 2020. Mortality Data from Johns Hopkins University, Unemployment Data from the Department of Labor Statistics

While reducing total deaths is the goal, the known relationship with unemployment is with health (all cause mortality). So it’s reasonable to look at the case fatality rate. Figure 3 shows that 47% of the variance between states in case fatality rate can be explained by a simple linear relationship with how hard states locked down (and it’s not in the direction intended by the states that locked down hard). I know, “correlation is not causation”. But negative correlation is at least an indication of lack of causation.

Figure 3: Covid Case Fatality Rate as of 10/7/202 vs Increase in Unemployment from the Three Months Ending March, 2020 to the Three Months Ending August, 2020. Mortality and Case Data from Johns Hopkins University, Unemployment Data from the Department of Labor Statistics

And one is compelled to ask who is closing down so hard and failing to prevent if not causing so much death. I don’t think anyone will be surprised by Figure 4. Figure 5 through Figure 7 show how that worked out for the other variables.

Figure 4: Increase in Unemployment from the Three Months Ending March, 2020 to the Three Months Ending August, 2020 vs Democratic Control of the State Legislature. Unemployment Data from the Department of Labor Statistics, Legislature Data from NCSL.

Figure 5: Cases as of 10/7/202 vs Democratic Control of the State Legislature. Case Data from Johns Hopkins University, Legislature Data from NCSL.

Figure 6: Total Covid Deaths as of 10/7/202 vs Democratic Control of the State Legislature. Mortality Data from Johns Hopkins University, Legislature Data from NCSL.

Figure 7: Case Fatality Rate as of 10/7/202 vs Democratic Control of the State Legislature. Mortality and Case Data from Johns Hopkins University, Legislature Data from NCSL.

And just because I had the data Figure 8 confirms what we probably all knew. The improved mortality is NOT because republicans are generally healthier than democrats.

Figure 8: Obesity Prevalence vs Democratic Control of the State Legislature. Obesity Data from the CDC, Legislature Data from NCSL.

Table 1 shows all of the data for the above graphs. It’s helpful for locating a state since the labels are sometimes overlapping.

Table 1: Data in above Figures. Mortality and Case Data from Johns Hopkins University, Unemployment Data from the Department of Labor Statistics, Legislature Data from NCSL, and Obesity Data from the CDC

State	Unemployment Increase	Cases per Million 10/7/2020	Deaths per Million 10/7/2020	Case Fatality Rate 10/7/2020	% Dem in Legislature	Obesity
Alabama	4.03%	32679.9	497.4	1.52%	25.7%	36.2%
Arizona	4.00%	30462.6	752.6	2.47%	46.7%	29.5%
Arkansas	3.43%	29159.6	396.6	1.36%	24.4%	37.1%
California	9.30%	21121.5	381.0	1.80%	75.0%	25.8%
Colorado	4.93%	12961.2	350.4	2.70%	60.0%	23.0%
Connecticut	5.60%	16646.8	1260.8	7.57%	60.4%	27.4%
Delaware	5.83%	22095.4	643.9	2.91%	59.9%	33.5%
District of Columbia	3.83%	22177.9	879.9	3.97%	92.9%	24.7%
Florida	6.30%	33400.8	620.0	1.86%	40.0%	30.7%
Georgia	3.60%	30400.0	622.0	2.05%	40.7%	32.5%
Idaho	2.00%	25375.2	250.1	0.99%	20.0%	28.4%
Illinois	8.57%	24408.7	686.0	2.81%	64.4%	31.8%
Indiana	5.43%	19003.5	521.7	2.75%	28.7%	34.1%
Iowa	3.67%	29930.0	407.0	1.36%	43.3%	35.3%
Kansas	3.73%	21147.2	206.3	0.98%	31.5%	34.4%
Kentucky	0.77%	17035.2	248.9	1.46%	34.1%	36.6%
Louisiana	4.10%	36514.8	1156.6	3.17%	32.6%	36.8%
Maine	4.60%	4168.2	104.2	2.50%	58.6%	30.4%
Maryland	4.37%	21216.4	642.3	3.03%	69.1%	30.9%
Massachusetts	11.83%	19779.6	1351.8	6.83%	80.0%	25.7%
Michigan	7.07%	14495.1	699.0	4.82%	45.9%	33.0%
Minnesota	4.00%	18817.9	358.4	1.90%	53.2%	30.1%
Mississippi	4.37%	33918.9	944.2	2.78%	35.6%	39.5%
Missouri	3.43%	20416.2	277.4	1.36%	28.4%	35.0%
Montana	3.07%	14752.7	149.7	1.01%	41.3%	26.9%
Nebraska	1.27%	25036.9	233.7	0.93%	25.0%	34.1%
Nevada	9.57%	26796.8	497.1	1.85%	65.1%	29.5%
New Hampshire	5.80%	6421.2	322.1	5.02%	58.5%	29.6%
New Jersey	9.67%	23615.6	1809.1	7.66%	64.2%	25.7%
New Mexico	5.97%	14909.1	405.9	2.72%	64.3%	32.3%
New York	10.73%	17899.0	1264.9	7.07%	64.6%	27.6%
North Carolina	4.13%	21007.0	309.6	1.47%	44.7%	33.0%
North Dakota	3.00%	32118.8	253.3	0.79%	17.7%	35.1%
Ohio	5.30%	13884.9	395.5	2.85%	35.6%	34.0%
Oklahoma	3.23%	23497.5	239.6	1.02%	21.5%	34.8%
Oregon	6.00%	8401.0	125.4	1.49%	61.1%	29.9%
Pennsylvania	7.63%	13370.0	623.7	4.67%	45.1%	30.9%
Rhode Island	8.73%	24287.4	1035.5	4.26%	87.6%	27.7%
South Carolina	5.10%	29567.7	623.8	2.11%	37.1%	34.3%
South Dakota	2.73%	29283.6	228.3	0.78%	15.2%	30.1%
Tennessee	5.80%	30001.6	327.0	1.09%	23.5%	34.4%
Texas	3.90%	27531.7	521.7	1.89%	43.1%	34.8%
Utah	1.40%	24979.4	137.6	0.55%	21.2%	27.8%
Vermont	4.77%	2927.9	93.0	3.17%	64.4%	27.5%
Virginia	4.23%	17748.7	353.7	1.99%	54.3%	30.4%
Washington	4.90%	11856.1	269.1	2.27%	58.5%	28.7%
West Virginia	4.80%	9420.0	175.8	1.87%	41.0%	39.5%
Wisconsin	4.03%	23688.0	213.7	0.90%	37.9%	32.0%
Wyoming	2.70%	11920.3	86.4	0.72%	13.3%	29.0%