© 2018 by KlarnerPolitics

Determinants of Gubernatorial Response to COVID-19

By Carl E. Klarner

 

 

 

We examine the determinants of nine gubernatorial actions taken to fight the COVID-19 pandemic in an event history analysis of February 27th to March 26th, 2020.

 

The media narrative has been that Republican governors have been slow to respond.  But this notion ignores the fact that states first hit by the disease are much more likely to have Democratic governors.  In the time period examined, the average number of cases per 100,000 has been 3.00 in states with Democratic governors and 1.19 in states with Republican governors.

 

This analysis examines whether Republican governors have been slower to respond to the COVID-19 epidemic after taking the seriousness of the crisis in their states into account.

 

This report also provides a tool for advocates, by identifying states that have not yet adopted provisions but are most likely to do so.

 

All of the data, replication code and output from the analysis has been archived on Harvard’s Dataverse under my name.

 

FINDINGS

When the seriousness of COVID-19 in each state is taken into account, Republican governors have been slower to enact stay-at-home orders than Democratic governors.  The model indicates that on any recent day, there is a 1.4% chance a Republican governor will mandate a stay-at-home order, while there is a 5.1% chance a Democratic governor will do so.

 

Democratic governors have also been more likely to institute bans on gatherings of 25 people or more than Republican governors.  For the recent past, the chance of a Republican governor banning such gatherings is 7.6% a day, while the chance of a Democrat doing so is 14.9%.

 

Both of the above findings are statistically significant, although the latter only achieves conventional levels of statistical significance in a 1-tailed test (p<.029, 2-tailed / p<.081, 2-tailed).

 

For the other seven policies examined, there were no statistically significant differences between how Democratic and Republican governors acted.  This is true for policies other than social distancing policies where greater partisan differences might be expected, such as moratoriums on evictions, provisions to expand access to child care, and expanding access to unemployment insurance for the entire labor force.

 

An analysis of the role of race and ethnicity in the response to COVID-19 indicated that state responses have been fairly similar regardless of percentages of African-Americans and Hispanics for most of the nine policies.  A noteworthy exception is that governors have been more likely to call out the national guard in states with larger percentages of African-Americans (p<.001).  Although the evidence from this analysis is only suggestive, this finding is consistent with work that has found stricter law enforcement policies in states with larger percentages of African-Americans.  On the other hand, there were several instances of national guard mobilization that were focused on provision of assistance to medical facilities.

 

One other exception is that governors in states with larger percentages of Hispanics were more likely to declare states of emergency (also statistically significant), but again, these results are merely suggestive, and probably have more to do with region of the country.

 

POLICIES EXAMINED

 

We read through a list of 1,494 actions taken by state governors on the National Governor’s Association Website and coded them as fitting into one of the nine dichotomous categories reported here.  https://www.nga.org/coronavirus/#states.  The following tables report the date a particular action was taken, and the predicted probability that such an action will be passed within a one day period, if it hasn’t already.

State of Emergency

Some states have different categories of states of emergency, and what a state of emergency means varies from state to state.  If the governor declared any type of emergency, we coded the governor as taking this action.

National Guard

If the governor mobilized the national guard in any capacity, the governor was coded as taking this action.

Stay-at-Home Order

Merely mandating that all non-essential businesses close (such as in New York) isn’t included in the definition of this type of action.

Gatherings of 25 or More Prohibited

Almost all of these consisted of prohibitions on 10 or more people.  A stay at home order without a prohibition on gatherings was also counted as this type of action.

Restaurant and Bar Closures

This means that dine-in services have been prohibited.  This is generally a proxy for all non-essential businesses being closed.  A stay at home order without a prohibition against restaurants and bars was also counted as this type of action.

Close Schools

In some states, the governor doesn’t have the authority to close public schools.  We didn’t have time to figure out which states grant the governor this power.  For this reason, we coded a governor as having closed schools if they either explicitly made the order, or advised it.  When another entity closed schools—such as local school districts statewide—and no mention is made of the governor, the state is no longer “at risk” for gubernatorial action on closing schools and is taken out of the analysis but isn’t coded as a gubernatorial action.  We considered March 18, 2020 to be the date California schools closed.

Unemployment Insurance

Whether provisions have been put in place to broaden access to Unemployment Insurance benefits.  Such provisions have to involve the entire labor force, not merely those who are sick, to count as this type of action.

Child Care

Whether special provisions to expand child care to first responders have been made.  This also includes general provisions to expand access to child care.

Eviction Moratorium:

Whether a moratorium on evictions from houses and apartments has been put in place.  Again, if another entity acts—such as the courts—cases for that state are then removed from the dataset after that date, and the governor isn’t coded as having taken the action.

FACTORS INFLUENCING COVID-19 POLICIES

 

Each model predicting one of the nine policies included the following eleven independent variables.

 

Democratic Governor: Coded “1” if the governor is a Democrat, “0” otherwise.

 

Income Growth Per Capita %: States may be hesitant to close down businesses the worse their economies are compared to other states.  Unfortunately, this measure comes from the third quarter of 2019, before the current economic slowdown.

 

COVID-19 Cases / 100,000 Population: Measured in the prior day.  Daily data on positive cases was obtained from the COVID-19 Tracking Project on March 29, 2020 at 2:22 PM Eastern.

 

Neighbor States: COVID-19 cases per 100,000 population in the neighboring state with the highest infection rate, measured in the prior day.

 

Age 65 and Over %: More vulnerable populations may cause more timely responses.

 

Population Density: If more of a state’s population is in close proximity with each other, there may be more concern about the crisis.

 

Hospital Beds per 1,000 population: More concern may exist if hospital capacity is lower.

 

Time Counter: Starts at “0” on February 26, 2020, and goes up “1” each day.  As the crisis has unfolded, states all over the country have become more apt to respond, and public pressure to respond has plausibility increased.

 

Time Counter Squared: Allows the impact of time to slow down or speed up.

 

Weekend Variable: Coded “1” if a day was on a weekend, “0” otherwise.  Actions of any type are about half as frequent on weekends versus weekdays.

 

African-American %: Percentage of state population African-American.

 

Hispanic %: Percentage of state population Hispanic.

 

APPENDIX A: DATA DETAILS & SOURCES

 

Our analysis begins on February 27, 2020 because an examination of the dates of policy adoptions make that appear to be a natural cutoff.

 

Another possible way of measuring gubernatorial response would be to count the number of actions that governors have taken, but this would be an unwise strategy.  A reading of the 1,494 actions makes it obvious that the level of reporting in the source we used varies from state to state given the level of detail reported in some but not others.  For example, Connecticut reports around 100 actions, while the states with the second and third most actions have around 50 each.  Furthermore, disparate actions are often presented as one action.

 

Similar policy actions in different states are often different from each other in their details.  We made quick judgments about which policies warranted a code of "1," but interested parties can look at the entire list of state actions we've posted on Dataverse and now we coded them.

 

Judgments also had to be made about when a state that had incrementally adopted a response was considered to have entirely made that response.  We generally used the last date of such incremental adoptions as the date of adoption, although if we judged later adoptions were small enough, we might use the second to last date, etc.

 

We compared whether a state had adopted a measure by March 25, 2020 with Politico's report on Coronavirus policy responses as a check on our work, as well as other lists.  Discrepancies between our lists and others often hinged on differences of definition.

 

It’s easy to conceive of ways the independent variables might interact with each other.  Because of concerns about multicollinearity, we only examined one: the interaction between cases per 100,000 population and governor’s party.  Multicollinearity was a major problem for the interaction just mentioned in many of the models (see output posted on Dataverse), but not for any other independent variables (except “time” and “time squared” as expected).

 

Other Data Sources: Citations to the data sources for the other independent variables used in the analysis can be found in the Dataverse post associated with this analysis.

 

Population Density: was measured with a Herfindahl index of Census Combined Statistical Area-state intersections, with “0” input for all population living outside of metropolitan areas.  This means that if everyone in a state lives in the same metropolitan area—such as in Rhode Island—the state will receive a score of “1.”  If a state has no people living in what the Census defines as a metropolitan area—such as in Alaska, Montana, and Wyoming—it receives a score of “0.”

 

Neighbor States: This is an approximate way of tracking officials’ concern about infections spreading from other states.  It ignores the fact that some neighboring states have a great deal more interaction than others.  For example, New Jersey and New York are closely intwined, while Ohio and Pennsylvania aren’t.

 

The State Legislature: It’s possible that some governors didn’t engage in particular actions because the state legislature had already taken that action.  For the most part, we did not code actions taken by the state legislatures, although an analysis that does so would be an improvement over the current analysis.

 

Age 65 Years and Over as % of Voters: Older individuals are more likely to vote than younger voters generally, but are especially more likely to do so in some states.  This variable was removed due to multicollinearity.

 

Two sets of models are run.  One set has all of the listed independent variables.  The second excludes the party of the governor, state economy, and hospital beds variables.  This is done because race and ethnicity are plausibly causally prior to these variables.  The race/ethnicity variables perform much the same in each of the two types of models, but the effect sizes are more accurately assessed in the second type of models.

 

The finding about stay-at-home orders is robust to different model specifications, although the ban on gatherings becomes weakly statistically significant in other models.

 

The reported probabilities are for an imaginary state where all the other independent variables are held at their means, on march 26, 2020.

The State Legislative Election Returns database 

updated through the 2018 elections 

is now available for purchase

State by State Analysis of

Whether Non-Citizens Will be Excluded from Redistricting Population Counts

Here's a state by state run-down of the likelihood that eleven states will exclude non-citizens from population counts for the purpose of redistricting.  These are the top eleven states in terms of how much the change would harm the Democrats in state legislatures, with at least a one percent loss of seats in each.

 

In summary, excluding non-citizens from districting population counts is unlikely to occur in these states by legislative action, with the possible exception of AZ.  However, it may come about through the initiative in AZ, CO, FL, or NV. 

As of June 22, 2019, I see no constitutional amendments by initiative restricting redistricting population counts to citizens only on the AZ, CA, CO, FL or NV Secretary of State Web sites.

 

The table below shows the percentage of seats Democrats are predicted to lose with this exclusion of non-citizens, in both the state legislature and U.S. House for each of the 50 states.  These figures come from my analysis at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2699850, which was conducted using the results of the 2014 elections and the 2010-2014 American Community Survey.

Column three of the table below displays the percentage of seats the Democrats in both chambers of the state legislature (averaged) are estimated to lose if non-citizens are excluded from population counts for purposes of redistricting, while column four does the same for the U.S. House.    Column two shows the percent of a state’s population that is non-citizen for context. 

 

Note that the analysis presented in the table doesn't exclude children: it's doubtful anyone would make the strategic blunder of combining excluding children with excluding non-citizens.  The partisan consequences of excluding children are also much smaller than excluding non-citizens (see cited analysis). 

 

Columns five and six indicate the percentage of state legislators who are Democrats in the state house and senate, respectively.  Column seven reports whether a constitutional amendment is required to change redistricting standards, which is very often the case, especially in the top 11 states focused on, making excluding non-citizens from redistricting counts challenging.  Of the top 11 states, only CT can have redistricting standards changed by law.  Once a constitutional amendment passes the state legislature--sometimes in two separate sessions (see column nine)--it must then be ratified by the voters in a state.  In many cases, that's enough to make it impossible to pass such measures before the post-2020 redistricting is underway.

 

A comparison of the percentage of Democrats currently in the state legislature with the vote requirements for passing a constitutional amendment (column eight) provides insight into whether changing who is counted for redistricting will come about by legislative action.  If it doesn’t, the only other alternative is through the initiative process, and as the last column of the table indicates, this can happen in some key states.

 

AZ

It's easy to pass constitutional amendments in Arizona, with only a simple majority of legislators in one session being enough for passage, meaning the Republican controlled legislature might do so.  On the other hand, a few Republican legislators in the narrowly controlled state house could join with the Democrats to block passage.  If only a few Republicans lose out because of the electoral volatility caused by changing how population is counted, it would have a hard time passing. 

 

If the legislature doesn't do it, the public might via the initiative.  Such votes must be in November of even numbered years, meaning that if it isn't on the November 2020 ballot, it isn't going to happen.

 

CA

If it passed here, it's true it wouldn't hurt the Democrats' ability to keep the state legislature, but it might lose them a seat in the U.S. House, or effect whether they have a super-majority for the budget, and the public could do it through the initiative process.  Constitutional amendments by initiative can only be voted on in general elections, making the November 2020 election the only possible way this could happen in time for redistricting.

 

CO

Current Democratic control of the legislature makes this unlikely to happen and the Democrats are unlikely to be beaten down to less than one-third of legislative seats after the 2020 elections, since a legislative vote of two-thirds is necessary to pass a state constitutional amendment (but only in one session of the legislature).  In the unlikely event a measure passes in the 2021-22 biennium, it still has to be ratified in the November 2022 elections, too late for the 2022 redistricting cycle.

 

However, it could still happen via the initiative process.  Colorado allows constitutional amendments to be ratified in odd-year elections, which occurred as recently as 2013, meaning if it isn’t on the 2020 ballot, it could be on the 2021 ballot, theoretically in time for drawing maps before the 2022 elections.  It is also possible an attempt at re-redistricting through the courts could be made to redraw lines under the new standard if one were to pass in 2021 or even 2022. 

 

CT / MD / NY / NJ

It's unlikely the Democratic state legislatures in these states will pass such a measure, and the public can't since none of them have the initiative.

 

FL

A 60% majority is required in the legislature to pass a constitutional amendment, and since the Democrats have 42.5% of the seats in the State Senate, they can block passage.  They don't have enough in the State House to block passage, however.  Even if the Republicans pick up enough seats in 2020 to attain the super-majorities they need, the referendum approving the amendment would be held in the 2022 general election (barring a 75% vote approving it to be ratified in a special election), too late for redistricting. 

 

That just leaves the initiative process.  Initiatives amending the constitution must always occur in general elections, so if such a provision isn't put on the November 2020 ballot, excluding non-citizens from population counts isn't going to happen in FL. 

 

IL

The strongly Democratic legislature won't pass it, but the public could do it through the initiative process.  This seems doubtful in blue Illinois.

 

NV

Only a simple majority is needed to pass a state constitutional amendment here, but the Democrats control the state legislature, so it’s unlikely to happen.  Furthermore, since a constitutional amendment must be passed in two sessions of the legislature, it's too late for this to be done by the 2020 redistricting cycle.  If the Republicans take the legislature in 2020, pass an amendment in the 2021-22 biennium, they'd have to pass it again in 2023-24, and then have it ratified in the 2024 general election. 

 

It could very well pass via the initiative process, however.  Nevada doesn’t allow odd year votes on constitutional initiatives, so it would have to be in November of 2020.

 

TX

A two-thirds majority is required in the legislature to pass a constitutional amendment, and the Democrats have a large enough minority in both chambers to block passage: 44.7% in the House, and 38.7% in the Senate.  If the Republicans pick up enough seats in 2020 for the necessary super-majorities, it would be possible--theoretically--to have the amendment ratified by the public in time for redistricting.  This is because the legislature is only required to pass the measure once, and also because Texas allows votes on constitutional amendments in odd-numbered years.

 

Since TX doesn't have the initiative, in any form, this is the only scenario in which non-citizens are excluded from population counts in TX in time for the 2022 elections.  Add another reason that progressive groups will continue to send campaign resources to state legislative elections in this safely red state.

 

VA

Could be done with difficulty before redistricting, since an amendment to the state constitution would have to pass now and after the November 2019 elections, which could then be ratified in the November 2020 election.  Only a 50% vote would be necessary, but the Republicans currently have a bare majority. 

 

The initiative doesn't exist in VA.

 

Estimated Democratic Percentage Seat Loss if Non-Citizens Excluded from Districting Counts & Factors Relevant to the Likelihood States Will Do So

Cleaning Digitized Election Data Efficiently

This 25 minute video

  • Gives numerous suggestions about how to clean data faster.

  • Discusses four strategies for cleaning data when sub-totals are reported.

  • Assesses two of these strategies with a simulation.

2018 State Legislative Election Forecasts:
October 30, 2018 Release

The forecasts here indicate that seven chambers will flip from red to blue in the upcoming election.  

The ME & NY Senates are very likely to flip blue, while the AZ, CO, MI, NH, and NC Senates are essentially toss-ups.  Among state houses, AZ, MI, NH and WV are all toss-ups.  

Overview of my August 27, 2018 forecast

State Senates, 2018 Elections:Forecast Probability of Democratic Control

Percentages in each state indicate forecast probability Democrats will have a majority after the election. 

Circles shaded red to blue to reflect this percentage (truncated between 35 and 65%).  States red if currently Republican controlled, blue if Democratic. Chambers not up for election assigned 0 or 100%.  Nebraska has a non-partisan unicameral legislature, but is colored red.  NY Senate takes IDC into account.  AK, HI, MA & RI omitted to make the map easier to take in.

State Senates, 2018 Elections:Forecast and Current Percent Legislators Democratic

Top number indicates forecast Democratic percentage of legislators, reflected by red to blue shading of circle (truncated between 35 and 65%).  Bottom number indicates current Democratic percentage of legislators, reflected by red to blue shading of state (truncated as above). Nebraska has a non-partisan unicameral legislature, but is assigned a score of "25.“  AK, HI, MA & RI again omitted.

State Houses, 2018 Elections: Forecast Probability of Democratic Control

Percentages in each state indicate forecast probability Democrats will have a majority after the election. 

Circles shaded red to blue to reflect this percentage (truncated between 35 and 65%).  States red if currently Republican controlled, blue if Democratic. Chambers not up for election assigned 0 or 100%.  Nebraska has a non-partisan unicameral legislature, but is colored red.  AK, HI, MA & RI again omitted.

State Houses, 2018 Elections: Forecast and Current Percent Legislators Democratic

Top number indicates forecast Democratic percentage of legislators, reflected by red to blue shading of circle (truncated between 35 and 65%).  Bottom number indicates current Democratic percentage of legislators, reflected by red to blue shading of state (truncated as above). Nebraska has a non-partisan unicameral legislature, but is assigned a score of "25.“  AK, HI, MA & RI again omitted.

The district forecast table contains the district level forecasts made on October 30, 2018.  A more detailed version of this table (as well as a table with the August 27, 2018 forecasts) can be found on Harvard's Dataverse.

Chamber Forecasts

This table contains the chamber level forecasts made on October 30, 2018.  A more detailed version of this table (as well as a table with the August 27, 2018 forecasts) can be found on Harvard's Dataverse.