Arkansas’s long history as a solid Democratic state ended some time ago. Today, with Republicans controlling the legislature and every statewide office, Arkansas Democrats are rowing against the tide. Data from the Associated Press’ 2022 Votecast survey show that only about 37% of likely Arkansan voters identify as Democrats or Democratic-leaning while 52% identify as Republican or Republican-leaning. Another 11% of likely voters identified as independent.
It’s easy for Democrats to look at these numbers and feel dejected. But elections are not predetermined. Yes, party identification plays an outsize role in vote choice, but so do other factors – and data from last year can help us understand how.
Overcoming party identification
Many Americans probably think the purpose of a political campaign is to convince voters that a candidate’s positions are correct. The reality is that a campaign works by focusing public attention on issues that provide the candidate with a competitive advantage.
Rather than trying to change hearts and minds, which is extremely difficult, successful candidates talk about issues about which large numbers of voters already agree with them. This is done to increase the importance (what political scientists call “salience”) of those issues in voters’ minds. If the public is inclined to agree with a candidate on education-related issues, for example, and the candidate can get people thinking about schools, that helps the campaign.
Of course, opposing candidates are trying to do the same thing. To win in a state like Arkansas, a Democrat needs to identify an issue or set of issues that can eat into existing Republican support and increase the salience of those issues in a big way. Republican candidates just need to push back against Democratic efforts to maintain their strong base of support.
This is not a simple one-off strategy. In every election in which the candidate from the weaker party (in Arkansas, the Democrats) gains a little, blind support for the stronger party erodes.
If a Democratic candidate in Arkansas can win an independent’s vote in one election, that person becomes more likely to vote for a Democrat in the future. If a weak-Republican voter can be convinced to cast a ballot for a Democratic candidate because of a specific issue, he or she becomes less connected to the GOP in the future. And first-time voters who vote for a Democrat are substantially more likely to vote Democratic their whole lives.
In essence, Democratic candidates can nibble away at Republican strength over time by identifying and talking about issues that draw voters to the Democrats. But which issues are those?
Issues in 2022
The AP’s national Votecast survey, conducted in late October and early November of 2022, includes a representative sample of nearly 3,000 Arkansans. With the help of statistical models, we can isolate the effects of certain attitudes and opinions about issues while controlling for variables such as age, income, race, gender and so on. This data allows us to understand what issues did and didn’t work for Democrats in the 2022 Arkansas gubernatorial election. For details on methodology, please see the note at the end of this story.
Figure 2 presents the issues Arkansans said were most important to them in 2022, broken down by self-identified Democrats, Republicans and independents. The economy was the most important issue to all three groups, though especially for Republicans and independents.
Significant percentages of Democrats (but smaller numbers of Republicans and independents) identified abortion, the environment or health care as their most important issue. Republicans were more likely than independents (and much more likely than Democrats) to say immigration was their most important issue.
Figure 3 examines voting patterns just among Arkansans who identify as Democrats or Democratic-leaning, based on whether they viewed a particular issue area as important. Obviously, a person identifying as a Democrat was far more likely to vote for Democratic candidate Chris Jones regardless of what issues they saw as most important.
But there are still clear differences. For example, Democratic-affiliated likely voters who feel strongly about immigration were more likely than their peers to cross over and vote for Sarah Huckabee Sanders.
The environment provided Jones the most leverage. Nearly 88% of Democratic-affiliated likely voters who said the environment is the most important problem in the country voted for Jones, and only 5% voted for Sanders (7% chose not to vote). Health care was a close second, and abortion, foreign policy and gun control rounded out the top five.
Figure 4 shows the inverse: The likelihood of a Republican-identifying likely voter casting a ballot for Jones, Sanders or neither candidate based on what issue that voter deems most important.
Predictably, we see high levels of support for Sanders among Republican-affiliated likely voters. But Sanders was not as strong on certain issues. For example, among those Republican-affiliated likely voters who believed health care to be the most important issue — admittedly a small percentage of all Republicans — only about 63% cast a ballot for Sanders. More than 14% voted for Jones, and nearly 23% chose not to vote.
Sanders also performed less well with Republican-affiliated likely voters who prioritized abortion, the environment and gun control, with between 11% and 16% of these voters supporting Jones.
The big prize for any campaign, however, is independents. The likelihoods of those who identify as neither Republican nor Democrat voting for Jones, Sanders, or not voting are displayed in Figure 4.
Independents were far less inclined to vote than those who identified with a party, regardless of their issue of choice. But Jones performed best in both absolute terms and relative to Sanders among independents who believed the environment to be the most important issue, followed by health care, abortion and gun control. Sanders bested Jones on the economy, immigration, crime and covid-19.
No matter the partisan identification of a voter, the voting patterns in the 2022 election appear similar. Jones’ performance was boosted among voters who cared about health care, abortion, the environment and gun control. Sanders performed best on the issues of immigration, crime and the economy.
Could Democrats have gained more votes?
Of course, most people are not single-issue voters, and a candidate must be mindful to not drive voters away when emphasizing particular issues. The question then becomes: Do those voters who did not identify a specific issue as the most important agree with Democrats on that issue? For example, do most people who did not say health care was most important to them agree with Democrats on how to improve health care?
The AP Votecast data includes four questions that help us understand attitudes on the issues that most benefited Democrats in 2022:
Figure 7 shows the percentage of Democrats, Republicans and independents who are likely voters and agree with Democrats on each of the four issues that Democrats did best with in 2022. The likelihood that a Democratic-affiliated likely voter agrees with the Democrats on health care, the environment, gun control and abortion is well over 80% in each category, which is hardly surprising.
There are, however, clear differences across the issues with regard to support among Republicans and independents. Though less than half of Republican-affiliated likely voters hold similar positions to Democrats on any of these issues, 40% agreed with the Democratic position on health care, 35% agreed with Democrats on the environment and 34% agreed with Democrats on gun control. Only 28% of Republicans supported Democratic positions on abortion.
We see a similar pattern among independents who are likely voters: They tended to agree more with Democrats on health care, the environment and gun control, but slightly less on abortion. But even on abortion, 48% of independent likely voters agreed with the Democratic position.
We can take this one step further. If we assume an imaginary election in which only one of these issues is important to voters, and party identification doesn’t matter, we can determine how much advantage Democrats have on a particular issue. These competitive advantages are displayed in Figure 7.
Importantly, one cannot take these figures as realistic. The numbers should be interpreted only relative to one another, telling us which issues confer more of an advantage on Democrats.
This model suggests a focus on health care would be particularly helpful for Democrats. Their position on this issue has appeal for their own base as well as a substantial share of likely voters who identify as Republicans and independents, or nearly 60% of the likely voters in Arkansas. The environment and gun control also provide an advantage, with 56% and 54% of likely voters holding similar positions to the Democrats.
It’s less clear whether abortion provides an advantage, as only about 50% of likely voters agree with the Democrats on the issue. However, given that Democratic vote share in recent elections has been in the mid- to low-30s, one could conclude that abortion would still provide Democrats some advantage.
Of course, one should note some of the limitations of this study. First, the question asked about the most important issue is very broad. Obviously, there are many facets to the economy, healthcare, immigration, abortion, etc. As an example, one person could support government run healthcare, while another simply supports maintaining Obamacare. I endeavored to take this into account by examining questions that are more detailed for the issues for which Democrats are strongest.
Second, and most importantly, one candidate or party cannot completely control the agenda. If Democrats decided to increase the salience of healthcare (something that the data suggest would be a good idea), they would have to contend with Republican messages which might be focused on a different issue, and real-world events which can sometimes supersede any issue a candidate or party might talk about. However, the fact remains, regardless of what Republicans do, or what events are transpiring around the world, Democrats are still best served to focus on those issues that they can increase the salience of most, and on which the public tends to agree with them.
A note on methodology
Why a person chooses one candidate over another cannot be parsed out by simply looking at what social scientists call descriptive statistics. Rather, we need to use some pretty sophisticated statistical models that allow us to isolate the effects of certain attitudes and opinions.
For other data geeks out there, here’s what I did. I created a dependent variable that is coded as a “1” if an individual said they did or would vote for Chris Jones, a “2” if an individual said they did or would vote for Sarah Huckabee Sanders, and a “3” if an individual said they were unlikely to vote. I then used a series of nine independent variables asking whether a survey respondent deemed a particular issue to be the most important. These issues included the economy, healthcare, immigration, abortion, crime, the environment, foreign policy, COVID-19 and gun violence. Using a multinomial logistic regression with robust standard errors, I examined the effect of viewing one of these issues as important on a person’s vote choice. I controlled for each survey respondent’s ideology, party identification, age, income, education level, race, gender, and church attendance. I then calculated the marginal effects, holding all control variables, with the exception of party identification, constant at either their median or means.
The post How can Arkansas Democrats win more voters? Data from 2022 holds clues. appeared first on Arkansas Times.