The Economist - USA (2021-07-10)

TheEconomistJuly10th 2021
Graphic detail Elections and the pandemic

81

Stampedout

J

ustascovid­19hasupendeddailylife,
so has it changed civic rituals. Historical­
ly, Americans have mostly voted in person.
But  in  2020  many  states  made  postal  vot­
ing  easier,  to  reduce  the  risk  of  the  virus
spreading. The share of ballots cast by mail
duly soared to 46%, from 21% in 2016.
Nonetheless, some 85m people still vot­
ed in person. Did this contribute to Ameri­
ca’s surge of covid­19 cases late last year?
Data from earlier in 2020 are inconclu­
sive. In Wisconsin 450,000 people voted in
person in a primary election in April. Two
studies later that month did not detect any
unusual increase in covid­19 cases; a third,
released in May, found a large effect.
The  general  election  in  November  of­
fered richer data. So far, 20 states have pub­
lished the number of ballots in each coun­
ty  cast  by  each  method.  Overall,  places
where  a  high  share  of  votes  were  cast  in
person  on  election  day—distinct  from
both postal ballots and votes submitted in

personbeforetheelection—alsohadhigh

covid­19 rates. However, this pattern could

arise for reasons besides polling queues.

To  help  rule  out  alternative  explana­

tions, we studied changes in the incidence

of  covid­19  within  states  over  time.  First,

we compared each county’s case rate with

its  state’s  average.  Many  factors  can  make

the disease more or less common through­

out  a  state,  such  as  super­spreader  events

or mask mandates. Examining the gap be­

tween a county’s numbers and those of its

state strips out the impact of such events.

Next, we tracked how these disparities

changed  between  the  pre­  and  post­elec­

tion  periods,  a  method  known  as  “differ­

ence  in  differences”.  Suppose  that  people

who  would  not  have  been  infected  other­

wise did catch the virus at polling places. If

so, then covid­19 cases in the counties with

the most in­person voting in a state should

either have risen unusually quickly or de­

clined unusually slowly after the election.

The  data  display  just  such  a  pattern.

From mid­October to early November, co­

vid­19  cases  in  counties  with  their  states’

highest in­person turnout fluctuated simi­

larly  to  those  in  areas  with  the  lowest  in­

person  voting  rates.  But  a  week  after  the

election, positive tests became more com­

mon  in  places  with  the  most  in­person

turnout  on  election  day.  The  gap  was  big­

gestafter20­25days,shortlyafterofficial

data would include people infected by peo­

ple who caught the virus while voting.

This  divergence  does  not  prove  that

polling sites were at fault. Places with lots

of in­person voting on election day tended

to  share  other  attributes  as  well,  such  as

having relatively low levels of income and

education and having voted in 2016 for Do­

nald Trump, a sceptic of masks and social

distancing. Such characteristics could also

have caused the striking “difference in dif­

ferences” in the incidence of covid­19.

To  isolate  the  impact  of  in­person  vot­

ing, we built a model to predict each coun­

ty’s post­election change in covid­19 rates,

relative to state averages. We tested 22 vari­

ables,  such  as  population  density  and  the

pre­election growth rate of covid­19 cases.

Many  of  these  factors  did  affect  the

spread of the virus. Yet after accounting for

all of them, in­person voting still had a sta­

tistically  significant  effect.  Holding  other

variables  constant,  the  gap  in  in­person

voting  on  election  day  between  the  state

with the highest rate in our data (Alabama,

at  41%  of  the  population)  and  the  lowest

(Arizona, at 6%) was associated with an ex­

tra  173  cases  per  100,000  people.  This  im­

plies that if no one had voted inpersonon

election  day,  220,000  fewer  peoplewould

have been diagnosed with covid­19.n

In-person voting really did accelerate
covid-19’s spread in America

-2

-4

0

2

3

1

-1

-3

4

12th 19th 26th 2nd

October November

Electionday

9th 16th 23rd 30th

Changeincounty ’scovid-19casesper100,000people*comparedwithelectionday 2020

Byshareofpopulationvotinginpersononelectionday*,in 20 stateswithavailabledata,seven-daymovingaverage

UnitedStates,shareofpeoplevoting ,bymethod,%

Decreaseinin-personvotingbystate,201-20

%points

→ After America’s election day, covid-1 spread unusually quickly in places where lots of people voted in person

*Relativetothestateaverage †20%ofpopulation living in

countieswiththeirstates’highestin-personvoting rates

Sources:Stateelectoraldata;Townhall;HealthResources & Services

Administration;USCensusBureau;Surveyofthe Performance of

AmericanElections;NewYorkTimes

Countieswiththelowest

in-personvoterturnout*

Eg,Philadelphia,PAand

Miami-Dade,FL

Eg,Sumter,FL(TheVillagesretirement

community)andNavajo,AZ

Countieswiththehighest

in-personvoterturnout*

Increased rate of covid-19 cases* ↑

Population-based quintiles

of in-person voting rates

1st (bottom 20%)

2nd

3rd

4th

5th (top 20%†)

05102040600 ME

VT NH

WA ID MT ND IL MI NY MA

OR NV WY SD IA

MN

OH PA CT RI

CA UT CO MO KY WV MD DE

NM KS TN NC SC DC

AL GA

HI FL

AK

IN

NE VA

OK

TX

AR

NJ

WI

AZ

LA MS

100

80

60

40

20

0

1996 20161208042000

Postal/absentee

In person, early

In person, on election day