The Party Replies: Examining Local Party Responsiveness to Prospective Campaign Volunteers

Abstract

We examine the role that local parties play in responding to and equipping local volunteers to work during campaign seasons. We use a field experiment during the 2020 U.S. general election to investigate whether local parties are more likely to respond to certain types of volunteers and to examine what factors are associated with local parties’ responsiveness. We find that both Democratic and Republican local parties in competitive counties are more likely to respond to volunteers. Moreover, we find that both parties are more likely to respond to white volunteers and Democratic parties are more likely to respond to women. These differential response rates may be contributing to the increased demographic sorting between the parties.

Keywords

political parties representation and electoral systems American politics political participation

Introduction

Party organizations fulfill a critical role in American politics. They recruit candidates to run for public office, connect voters to elected officials at all levels of government, propose policy platforms and model legislation, and more. At the local level, parties provide solidary and purposive benefits to citizens by encouraging them to become politically active and provide state and national parties with important insights about local public opinion (Clark 2004). Central to this paper is the role local parties play in affecting the makeup of the electorate. Findings from get-out-the-vote (GOTV) field experiments consistently document the importance of interpersonal contact with party officials (Gerber and Green 2000; Green, Berger, and Nickerson 2003; Jacobson 2015) and canvassing by local citizens (Sinclair, McConnell, and Michelson 2013) to voter turnout. In this way, the maintenance and strengthening of local parties should be a priority for political parties that wish to win elections and ultimately get preferred legislation passed.

Yet even local party leaders who wish to grow their membership may also have incentives to include or exclude some groups. Local parties may take an exclusive approach if they think their electoral fortunes depend on appealing to moderate voters in the area (Downs 1957), fearing too many members that they perceive to be demographically or ideologically unrepresentative (Enos and Hersh 2015). For example, Doherty, Dowling, and Miller (2019) found that county party chairs from both parties viewed Latinx and Black candidates as less likely to win state legislative races. At the same time, partisan activists are not representative of the median voter and local party chairs (almost all volunteers) are more likely to reflect the characteristics of “intense policy demanders” than the characteristics of a typical local citizen (Bawn et al. 2012; Hassell Hans et al., 2020). In such cases, party leaders may seek out members who are demographically or ideologically similar (Wojcik 2018).

At the same time, all local parties face constraints on limited resources, and the ability to recruit and train new volunteers may be a function of organizational capacity. Local parties are tasked with important organizational tasks, such as participating in party meetings, supporting candidates, and registering voters (Clark et al. 1998; Feigert and Todd 1998). Local party organizations with accessible party headquarters, adequate staff, and stable budgeting (Cotter et al. 1984) may be best positioned to incorporate new volunteers into party operations. Moreover, while all party leaders typically devote more time to party activities during election years (Doherty et al. 2021), professional leaders spend more time on party business than amateurs. Those professional leaders may be better equipped to handle the influx of volunteers that comes during election season.

It is clear that understanding how local parties operate is fundamental to understanding who is and who is not involved in politics. In this paper, we set out to understand how local parties respond to prospective volunteers. Volunteers serve a critical role for parties; they can perform the labor-intensive work of voter registration and mobilizing voters (Hershey 2017). Volunteers and members of local political parties are also more likely to be recruited for political office in the future (Fox and Lawless 2010). Given the importance of volunteers to the health of political parties, we seek to answer the following questions: How does organizational capacity affect the responsiveness of local parties to prospective volunteers? And are local parties more likely to respond to some prospective volunteers over others?

We answer these questions through an experimental correspondence study in which party leaders received email requests from “prospective volunteers” who varied in their demographic characteristics. Our results indicate that local political parties’ responsiveness to volunteer inquiries varies both on volunteer-level characteristics and the larger context within which the parties operate. Our results also suggest important differences between Democratic and Republican party responsiveness. These differences in engagement may provide evidence of the continued demographic sorting between the two major parties.

Theories of Local Political Party Responsiveness

Even in an era of nationalized politics, there is a great deal of variation in political party organizations at the local level. In most states, local party organizations are formed at the county level. In some states, however, local party organizations form at the township or precinct level; still others are based on state legislative districts. The degree to which local parties work with, and receive assistance from, state party organizations also varies across the United States. A majority of local parties report receiving campaign training from state organizations (Roscoe and Jenkins 2016). Local party organizations also have a fair amount of autonomy in choosing to distance themselves from the national party, or focus on local issues (Doherty et al. 2021). This decision may reflect, in part, whether local party leaders want to solidify their existing base or broaden their appeal (Hansen 2016).

Regardless of how the local party is organized and its relationship to the state and/or national party, local political parties are often the first point of contact for people interested in getting involved in politics (Doherty et al. 2021). In this section, we present three theories to explain the variation in the level of responsiveness among local party organizations.

Organizational Capacity

In a survey of local party chairs, Doherty, Dowling, and Miller (2021) find that activity increases during election years, with more than half of survey respondents indicating they spend at least 30 hours a week on their chair duties (p. 1750). Local parties report organizing and voter registration as among their most important activities during elections (Clark et al. 1998). Meanwhile, party leaders also report that it is “somewhat common” or “very common” to be contacted by the media to provide quotes or background information (54.4%), to recruit (54.2%) or be contacted by potential candidates (51.5%), to consult candidates for Congress (26.3%) or state and local races (59.3%) (Doherty et al. 2021). Many of these activities are labor-intensive and rely on the help of volunteers (Hershey 2017).

The degree to which local parties organize prior to elections is a function of their capacity. Organizational capacity is seen in the party structure; more formalized structures with sufficient funding and professional staff can perform organizational maintenance and programming. Party organizations led by amateurs, for example, are generally weaker and may have fewer resources to mobilize during a campaign (Gibson et al. 1989). Rural counties, and counties with a greater percentage of nonwhite residents, also tend to have weaker party structures (Crowder-Meyer 2011; Hershey 2017). Rural counties may find it difficult to arrange meetings among their members. Moreover, we anticipate that there will be more contested elections and candidates for office in densely populated areas.

In her survey of local party organizations, Crowder-Meyer (2011) finds that Democratic party strength is associated with higher proportions of college-educated residents. This might explain why non-college-educated voters in battleground states were more likely to receive personal contact—often conducted by volunteers from local party organizations—from Republicans than Democrats during the 2016 election (Beck et al. 2018). We expect that local parties in wealthier areas will have more resources in terms of people who can afford to volunteer with the party and resources available to the party.

Organizational capacity theory: Local party organizations with more organizational capacity are more responsive to requests from prospective volunteers.

Political Environment

Competitiveness has the potential of either increasing or decreasing the strength of local parties. On one hand, increased competitiveness may mobilize voters in the area, providing the grassroots support necessary to maintain local parties. On the other hand, the national party is more likely to have a presence in competitive districts and battleground states, which in turn may weaken local parties. In presidential election years, local parties may find themselves at odds with the presidential candidate’s campaign (Doherty et al. 2021). This dynamic was particularly prevalent during the 2008 presidential campaign, when Obama’s grassroots campaign worked independently of local party organizations and clashed with local volunteers (Blumberg et al. 2011; Mayceko 2021). In contrast, Trump’s 2020 reelection campaign worked closely with state and local party organizations (Isenstadt 2018).

Partisan competition is likely to create an environment where Republicans and Democrats have a chance at winning some seats. While we cannot observe all races down the ticket, we expect that counties with close margins in presidential voting are likely to have some competitive federal, state, or local races. We also know that presidential campaigns focus their resources on competitive battleground states (Shaw 2006; Beck and Heidemann 2014). These states receive special attention by campaigns and the media, and so the local parties are more likely to receive an influx of resources and volunteers (Franklin, Richey, and Yonk 2013). In addition, local parties might feel more pressure to be active participants online in battleground states to support their presidential candidate.

Political environment theory: Local party organizations in more competitive political environments are more responsive to requests from prospective volunteers.

Solidifying the Base

In an ideal world local party leaders might like to bring in all types of volunteers to their local party, assuming that the volunteer supports the end goals of the party. But, with limited resources and time local leaders have to decide on what types of volunteers will be most useful for their local operation. We argue that two forces lead local party leaders to be focused on volunteers that are already solidly in the base of their party. First, volunteering represents a principal-agent problem that can be best solved by recruiting volunteers that are the most intrinsically motivated to support the party. Second, party leaders are interested in the solidary benefits that local parties create and these solidary benefits will be most beneficial when the local party is filled with likeminded individuals.

Although party volunteers can be an important part of political work they are not paid for their labor. In addition, their work, such as registering voters or going door to door, is often done with little monitoring. This creates a principal-agent problem for local party leaders. Party leaders, the principal, want to ensure that volunteers, the agents, complete whatever tasks that the volunteer has been assigned. The principal-agent problem of political volunteering has been documented in national campaigns (Enos and Hersh 2015; Chewning, Green, Hassell and Miles 2021). One solution to this problem is to recruit volunteers that are intrinsically motivated because of their ideological preferences. For example, Chewning and colleagues (2021) extremely or very important find that congressional campaigns are more responsive to ideologically proximate volunteers, although this effect decreases as a volunteer sends other signals to demonstrate their commitment to work (and so overcomes the principal-agent problem).

In addition to facing a principal-agent problem, local party leaders may be interested in simply creating an organization where they are comfortable. Historically parties overcame this principal-agent problem through the use of patronage jobs, rewarding those who worked for the party. Civil service reform removed this option though, leading parties to identify new ways to recruit volunteers (Sorauf 1960). Parties now have to rely on non-material benefits to recruit and keep volunteers and instead must use solidary and purposive benefits to attract people (Clark 2004). For example, in a survey of local party leaders, Roscoe and Jenkins (2021) find that 69.7% report that the friends and social contacts they have within the party are an extremely or very important reason for their involvement in the party. Since local leaders are interested in solidary benefits they will want to ensure that whoever enters the party has similar values and background as they do. The interest in solidary benefits then will also lead them to work hardest to bring in people who they perceive as already belonging to the party base.

An interest in building solidarity among party volunteers may lead party officers to work with those who are most like themselves. This pattern has been shown in candidate recruitment, particularly as it relates to women candidates (Niven 1998; Crowder-Meyer 2013). Brooks and Chatfield (2020) found a similar pattern with campaign staff, in which women candidates hired more women. This propensity to favor volunteers who share characteristics with party leaders may even take the form of strategic discrimination, in which party leaders choose not to engage with volunteers whom they believe will be perceived negatively by others (Bateson 2020).

The fact that parties are interested in recruiting volunteers that are within the base already helps explain some previous findings on who are party activists. In a study of national convention delegates from southern states, for instance, Hulbary and Bowman (1998) find that the vast majority (96%) of Republican delegates were white, whereas Democrats had more racial diversity among their activists (16). In addition, party activists are more extreme than the average partisan (Aldrich 2011). A Pew Research Center report in 2014 found voters at either end of the ideological spectrum were more likely to report having volunteered for a candidate or campaign. Thus, we anticipate that local parties will be more likely to respond to volunteers who are strong partisans, perhaps even the extremists within the party.

Solidifying the base theory: Local party organizations are more responsive to individuals who are perceived as being a member of the party’s base.

Correspondence Study

We use an experimental correspondence study to test our theories about local party responsiveness, which we model on other correspondence studies of political responsiveness (Broockman 2013; Butler and Broockman 2011; Butler 2014; Einstein and Glick 2016; Hayes and Bishin 2020). Emails from “prospective volunteers” were sent to local party leaders, we then recorded whether the volunteer received a response or not. We test the solidifying the base theory by varying characteristics of the volunteer. To test our political environment and organizational capacity theories we combine the response rates with observational data about the local environment of the party. In this section we outline how we collected email addresses for local parties, and the treatments used in the correspondence study. In the following section we identify our expectations for all three theories and how we operationalize the remaining two theories.

Data Collection

We collected email addresses for local county parties from 29 May 2020 to 18 August 2020 (Whitesell, Reuning, and Hannah 2022).¹ Table 1 displays the resulting sample and the source of the email addresses. In total we found email addresses for 80.3% of local Democratic parties and 69.0% of local Republican parties. Note that because not all state parties organize at the county level there is a difference in number of local parties across Democrats and Republicans. For example, in Minnesota both parties predominately use counties as local units, but in some instances use state legislative districts. Within the state, Republicans and Democrats differ in their division of local units. In addition, Connecticut and Massachusetts both use smaller township and precinct level units, which we took a sample of instead of searching for them all. We discuss these unique cases in the online appendix.

Table 1.

Source of local party organization email addresses.

	Democrat	Republican	Total
County website	27.8%	28.0%	27.9%
County Facebook	28.7%	35.2%	31.7%
County page on state website	8.4%	4.3%	6.5%
State website	33.2%	32.2%	32.7%
Other source	1.9%	0.3%	1.1%
Contact person
Chair	47.5%	51.3%	49.3%
General	49.1%	40.9%	45.3%
Other/Unknown	3.3%	7.8%	5.4%
N	2570	2225	4795

Most emails were found on county websites, county Facebook pages or state websites. For Democrats the most common source was state websites (33.2%) while for Republicans the most common source was a county Facebook page (35.2%). In addition to recording where the information came from, we recorded whether the email was for the chair of the party, a general email account, or something else. For Republicans, 51.3% of the emails were for a chair or similar leader of the party, while 40.9% were a general county email. For Democrats, 47.5% of the emails were for a party chair while 49.1% were a general email address.

We should point out that differences in online presence and email availability is correlated with organizational capacity. Whitesell, Reuning, and Hannah (2022) find that local parties have a larger online presence in counties with larger populations, higher per capita household income, higher proportion of population over 65 years old, higher proportion white population, and higher proportion of computer and smartphone users. Research on legislative audit studies has demonstrated that response rates can be moderated by other contextual factors like access to resources (Landgrave and Weller 2020).

At the end of this process, we found the email addresses of 2570 Democratic and 2225 Republican county parties. To avoid detection, we did not send emails to local parties in a county with a population of less than 2500. This left us with a sample of 2532 Democratic and 2186 Republican email addresses.

Treatment Conditions

Figure 1 provides the full text of each email sent to county party chairs. We sent one email to each chair during the week after Labor Day—Tuesday, September 8 through Thursday, September 10, 2020. We sent a random batch of 1600 emails on each day between 2:00 and 4:00 p.m. EST. The emails expressed a general interest in working for the party and did not specify a campaign. We did not want to prime the party chairs to think the volunteer was only interested in working on a specific race.

Figure 1.

Experimental email template.

We sent a roughly equal number of emails from four accounts that allowed us to randomize on race and gender. To select racially distinct first names, we used a dataset of 4250 first names with six mutually exclusive racial and ethnic categories (Tzioumis 2018). We further narrowed the list by using Social Security Administration (2020) data to identify the most popular first names in 1998 (approximately the birth year of a college student in 2020). From that final dataset we selected four names: Ayanna (Black female), Jermaine (Black male), Kaylee (White female), and Colton (White male). To select surnames, we consulted a Census Bureau publication with the most common surnames by race and ethnicity, according to the 2000 Census (Word et al. n.d).

In addition, we included a signal about the individual’s class, randomizing whether or not they were a full-time student or employed and needed flexibility around their work schedule. Finally, we included a signal of the volunteer’s ideology by including a quote in the signature space. We randomized treatments to Democratic county parties with either a quote by Jimmy Carter (moderate) or George McGovern (liberal) and to Republican county parties with either a quote from George H.W. Bush (moderate) or Barry Goldwater (conservative). Therefore, the emails included eight treatment conditions on race, gender, social class, and ideology. Only in the case of ideology was the treatment condition different based on whether the email was intended for a Republican or Democratic county party. From these conditions, we can determine whether county parties discriminated against volunteers based on race or gender.

Ethical Considerations

Audit studies have been conducted since the 1960s and are used primarily to detect discriminatory behavior, especially when standard survey questions would introduce social desirability bias. Gaddis (2018) defines the audit study as “a specific type of field experiment that permits researchers to examine difficult to detect behavior, such as racial and gender discrimination, and decision-making in real-world scenarios” (p. 2). This type of field experiment has become more popular in recent years as there are more opportunities to contact individuals and organizations online. Audit studies have been used in political science to learn more about public officials and their level of responsiveness to constituents (Butler and Broockman 2011; Costa 2017; Einstein and Glick 2016).

By examining how county political parties respond to prospective volunteers in the real world, we are learning about the organizational capacity and responsiveness of the parties as well as about whether the parties discriminate on the account of gender, race, class, or ideology. This type of information could not be captured through a survey of the parties due to social desirability bias. The audit study design allows for the observation of the true behavior of the party leaders. However, audit experiments require deception and put a burden on the recipient, so ethical concerns must be addressed (Landgrave 2020).

There are a few ways that we try to limit ethical concerns. We went to great lengths to try to minimize the use of the party leaders’ time. Our email is brief and our request for more information is relatively simple. In the most comprehensive survey of local party officials, Doherty and colleagues (2021) found that responding to inquiries from volunteers was one of the most common parts of their job. Responding to our request does not take much time and effort, nor does it preclude the organization from responding to other requests from volunteers or other interested parties. In fact, the average response to our inquiries was 41 words in length (including email signatures).

Expectations

In Table 2 we outline our expectations for response rates given the three theories we discuss above. Our expectations for the solidifying the base theory varies across the parties depending on who is traditionally perceived to be within their party’s base. For example, the Democratic Party is a multiracial coalition with a greater proportion of women. In 2016, 61% of Clinton voters were women and 40% of Clinton voters were nonwhite. In contrast, 47% of Trump voters in 2016 were women, and only 12% of Trump voters were nonwhite (Pew 2020). We expect then that local Democratic parties will be more responsive to female volunteers but not differentiate based on the perceived race of the volunteer. We expect local Republican parties to respond at higher rates to white volunteers compared to black volunteers but to respond equally to requests from male and female volunteers.

Table 2.

Expectations for county parties’ responsiveness.

Theory	Treatment type	Expectation/Operationalization
Solidifying the base	Experimental	Gender: Democrats respond more to female volunteers, Republicans respond equally Race: Republicans respond more to white volunteers, Democrats respond equally Class: Republicans respond more to working volunteers, Democrats respond equally Ideology: Both parties respond more to ideologically extreme volunteers
Organizational capacity	Observational	Parties in densely populated areas are more likely to respond Parties in high-income areas are more likely to respond Parties with more of an online presence are more likely to respond
Political environment	Observational	Parties in competitive counties are more likely to respond Parties in competitive states are more likely to respond

As it pertains to class, the Republican Party’s base is more likely to be non-college graduates (71% of Trump voters in 2016 did not have a college degree). The Democratic Party is more evenly split between non-college graduates and college graduates. In fact, Democratic voters in the 2018 midterms were divided 50-50 between non-college graduates and college graduates (Pew 2020). We expect then that local Republican parties will respond more to those who are working and less to those in college while Democratic parties will respond equally to each. Finally, given how ideologically extreme most activists are we expect that both parties will be more likely to respond to those who demonstrate ideological extremism.

Our next two theories, organizational capacity and political environment both depend on the local conditions of the party. In order to test the role of organizational capacity we use several variables that relate to the potential for organizational capacity and one that more directly measures it.² As discussed above, parties in richer and more densely populated areas tend to have higher degrees of organizational capacity. To measure local wealth we use estimates of Household Income (in $10,000) from the United States Department of Agriculture (USDA) (United States Department of Agriculture 2020). We use the proportion of county population that is classified as urban by the U.S. Census to account for more densely populated areas (United States Census 2010). Given the organizational capacity theory we expect that parties in areas with higher income and with a higher proportion of urban population to be more likely to respond to emails as they have higher capacity.

We also directly measure a type of organizational capacity through the local party’s online presence. This variable is derived by measuring a local party’s presence and activity on social media platforms (see Whitesell et al. 2022 for a full description of this variable). Higher values mean that a local party has more of an online presence. We again expect this to be positively related to whether a party responds. We also use an indicator of whether the local party donated to any state legislative candidates in the 2016, 2017, 2018, or 2019 election cycles from the National Institute on Money in Politics. Only 37% of local parties donated any amount. Whitesell and colleagues (2022) found that parties that had donated to legislative races had a larger online presence.

Finally for our political environment theory we account for the effects of state and local political conditions. We expect that parties in competitive areas will be more responsive than parties that are in areas where a single party dominates. To test this we use the Democratic presidential vote share in the 2016 election at the state and local level including both the direct proportion of votes and its square (County Presidential Election Returns 2016). We expect to find an upside-down parabola which peaks close to 0.5—where the parties are evenly split.

Analysis and Results

All emails were sent in the beginning of September 2020. Of the 2532 emails sent to Democrats, 40.1% were responded to within 1 week while 4.9% were undeliverable and bounced back. Of the 2186 emails sent to Republican parties, only 32.2% were responded to while 7.4% bounced back. Overall then, 55.0% of emails to local Democratic parties and 60.4.% of emails to local Republican parties were not responded to.

In order to test our theories though we need to examine who was responded to and which local parties responded. We start with bivariate analysis looking at just the treatment effects. Because this is a randomized experiment, we can use this to directly test our solidifying the base theory. To test the remaining two theories we use logistic regression and include the email treatments as well as the environmental variables discussed in the previous section. The logistic regression serves additionally as a robustness check for the bivariate analysis.

Bivariate Analysis

Figure 2 plots the response rates for both parties across the gender and race experimental treatments. Above the bars for each political party, we show the 95% confidence interval of the difference in response across the two treatments. Our first set of hypotheses focuses on whether local parties are interested in recruiting those that they already view as part of their bases. Overall, we find some support for this, with some important caveats.

Figure 2.

Experimental response rates based on characteristics of prospective volunteer.

For Democrats, we expected that they would be more likely to reply to an inquiry from women but equally likely to respond to emails from Black volunteers (compared to white volunteers) and from college students (compared to working volunteers). For Republicans, we expected that they would be more likely to reply to inquiries from white volunteers and volunteers that are working (compared to Black volunteers and volunteers in college) while there would be no differences across gender.

We find that both political parties are more likely to respond to women. Democratic party leaders responded to 43.7% of women volunteers, compared to 36.5% of emails from men. The 95% confidence interval (−11.02, −3.25) shows that the difference in response rates—7.2 percentage points—which is statistically significant. The gender difference among Republicans was smaller; Republican party leaders responded to 33.9% of inquiries from women and 30.5% of men. This difference—only 3.4 percentage points—is not statistically significant.

With respect to race, we find that both parties are more likely to reply to an email from a potential white volunteer over a potential Black volunteer. Democrats were 6.76 percentage points more likely to respond to white volunteers than Black volunteers. Party leaders responded to 43.49% of emails from white volunteers, but only 36.73% of Black volunteers. The difference was slightly larger for Republicans; party leaders were 8.38 percentage points more likely to respond to white volunteers (36.38% response rate) than Black volunteers (28.00% response rate). For both parties, the 95% confidence intervals around the difference between response rates to Black and white volunteers does not include 0; therefore, we can conclude that the difference is statistically significant.

In contrast to the race and gender treatments, our treatments for working class and ideology did not lead to statistically different response rates.³ The class treatment may present an informational equivalence problem (Landgrave and Weller 2022). In addition to signaling class, for example, attending college or working may also signal age or availability. Moreover, the ideology signal in the email signature might have been too subtle or complex to have a significant effect on response rate. Because of this we are hesitant to draw too much from the results presented here as the treatments might be testing more than we initially hypothesized. We have included them for transparency.

Overall then the bivariate analysis provides mixed evidence for our solidifying the base theory that local party organizations are more likely to respond to volunteers from groups viewed as being within the party’s voting coalition. In support of that theory, we find that Democrats are more likely to respond to women than men and Republicans are more likely to respond to white voters. As we move forward, we include the class and ideology treatments in the model, but focus primarily on race and gender.

Multivariate Analysis

Next, we use multivariate analysis to test our political environment and organizational capacity theories. We model whether a local party responded as a function of the sender covariates discussed above (each added as a dummy variable) and include the local covariates as well (Table 3).

Table 3.

Logistic regression results for predicting county party responsiveness.

	Republican	Democrat
Sender: Male	−0.13 (0.10)	−0.32** (0.09)
Sender: White	0.40** (0.10)	0.33** (0.09)
Sender: Working class	0.10 (0.10)	0.04 (0.09)
Sender: Moderate	−0.02 (0.10)	0.10 (0.09)
Online presence	0.13 (0.08)	0.46** (0.07)
Household income	0.03 (0.04)	0.09* (0.04)
Donated	0.13 (0.11)	0.10 (0.10)
Proportion urban	0.15 (0.20)	−0.19 (0.18)
Percent black	−0.80 (0.47)	−0.31 (0.42)
Dem vote (county)	4.49** (1.48)	2.86* (1.33)
Dem vote (county) squared	−4.67** (1.65)	−3.81* (1.48)
Dem vote (state)	4.41 (4.59)	9.69* (4.21)
Dem vote (state) squared	−5.34 (4.96)	−9.62* (4.50)
Intercept	−2.93** (1.07)	−3.85** (1.00)
N	2034	2367
AIC	2514.9	3051.5
BIC	2593.6	3132.3
Log.Lik.	−1243.5	−1511.8

* p < 0.05, ** p < 0.01.

The results for sender characteristics are broadly the same as what we presented in the bivariate analysis above. Both Democrats and Republicans were more likely to respond to a white volunteer than a Black volunteer; the coefficients on white volunteers were 0.33 and 0.39, respectively (p < 0.01). Democrats were more likely to respond to women with a −0.32 coefficient on the volunteer being a man for Democrats, while the coefficient for Republicans was −0.14 (only slightly larger than the standard error of 0.10).

Organizational capacity appears to have a larger effect for Democratic parties than Republican parties. Democratic parties with an online presence are more likely to respond to emails; among Republican parties, there is no significant relationship to online presence. In addition, Democratic parties in counties with higher household income are more likely to respond to volunteer inquiries. Republican parties are no more responsive to volunteer inquiries as the resources within the county change.

The local political environment is predictive of responsiveness, but competitiveness within the state, is not. For both Democrats and Republicans, local parties in competitive counties are most likely to respond to volunteer inquiries. Figure 3 shows that responsiveness decreases if either party dominates in the county. For example, in a county where Democrats received 50% of the vote in the 2016 presidential election, there is 39% probability of receiving an email response from the local Democratic party. In a county where Democrats received 75% of the 2016 presidential vote, that probability decreases to 29%. Similarly, in a county where Democrats received 50% of the 2016 presidential vote share, there is 34% probability of receiving an email response from the local Republican party. That probability decreases to 29% when the Democratic vote share decreases to 25%.⁴

Figure 3.

Probability of reply based on share of democratic vote in county.

The Conditional Effects of Race

In the results presented at this point, we have assumed our experimental effects would be consistent across different counties. Recent research though has indicated that the effects of race might vary in ways that are in line with our initial solidifying the base theory. Doherty et al. (2019) found that local Democratic party leaders’ interest in African American candidates is dependent on racial makeup of the county. Local Democratic party leaders preferred white candidates over Black candidates, but as the percentage of African Americans in a county increased this difference shrunk until it vanished. To test if the same pattern exists for our results, we include an interaction between percentage Black in a county and the race of the sender in our emails. In Figure 4 we plot predictions from these models across the range of percent Black population in a county and varying the race of the sender.⁵ For Democrats, a gap favoring white volunteers exists in counties where there are no (or very few) African Americans. This gap though shrinks as the percentage of African Americans in a county increases. For counties where at least a quarter of the population is African American, a white volunteer and a Black volunteer are equally likely to receive a response. In contrast, for Republican parties the gap widens (though not by a substantial amount) as the percentage of African Americans increases. Overall, this unhypothesized relationship provides additional evidence for the solidifying the Base theory and that the local conditions matter for Democratic party chairs, while local conditions do not matter for Republican parties.

Figure 4.

Predicted response rate by percent Black county population and race of sender.

Implications and Conclusions

Our results indicate that local party responsiveness is a function of both the political environment in which the party is operating and to whom the party is responding. Both Democrats and Republicans in more competitive counties are more likely to respond to prospective volunteer emails. In some ways, parties appear to be more responsive to their base—Democrats are more likely to respond to women, for example—but both political parties are more likely to respond to white prospective volunteers.

The results suggests that local political parties’ decisions may affect who is and who is not involved in politics. Parties may be missing key opportunities by ignoring volunteers in less competitive areas that could have an electoral impact. The progressive organization, Run for Something, found that Joe Biden performed 0.3 to 1.5% better in conservative state legislative districts where Democrats put forth challengers (Run for Something 2021). While challengers are different than volunteers, both parties would benefit from mobilizing volunteers to reach local voters regardless of the area’s competitiveness. Even if volunteers are working in an area with few likeminded voters, their efforts can be pivotal in tamping down margins that might swing closely contested statewide races.

Recent trends have shown not only increasing political polarization but also democratic sorting. Many explanations of these changes focus on ideological polarization as a primary cause (Zingher 2018; Gillion, Ladd and Meredith 2020). Our research though indicates that local parties can also be an important source of partisan sorting. If local Democratic parties are more interested in recruiting women, and more likely to respond in high-income areas, then demographic sorting will only increase. Who becomes involved in local party politics ultimately has consequences for who participates in politics long-term and runs for political office (Butler and Preece 2016; Fox and Lawless 2010).

There are a few limitations that might spur future research. Although we identified differences in responsiveness, the motivation behind the responses remains unclear. For example, it could be that local party volunteers are more likely to respond to requests from volunteers who are demographically like them. Research on state legislators has found this type of effect where Black legislators are more likely to respond to putatively Black constituents when they are outside their districts compared to white legislators (Broockman 2013). The difference in responsiveness could also be a function of more strategic decisions related to perceptions of what type of volunteer is likely to be helpful. An important step in untangling this is to identify who local party leaders are.

We also do not address the content of the responses which varied significantly. In some cases, the local party contact inquired as to why the prospective volunteer was not currently registered to vote (after looking their name up in county election data), while in other cases volunteers were immediately invited to become precinct leaders. The differences in the type of response are potentially dependent on both the characteristics of the county party and the volunteer. Less professionalized counties might have a greater need for volunteers and fewer resources available to check voter registration. Volunteers from groups outside the traditional party base might be viewed with suspicion especially given some recent high-profile examples of conservative activists infiltrating Democratic campaigns and related organizations (Mazzetti and Goldman 2020).

Fewer than half of local political parties responded to emails from prospective volunteers. This is an alarming and discouraging revelation. It is noteworthy, however, that not all counties even have readily accessible contact information or online presence. In other work we show that both Democrats and Republicans are less likely to have an online presence in heavily Republican counties as well as in less populated and poorer counties (Whitesell et al. 2022). This means that the variation in responsiveness based on county characteristics we find here likely understates the variation in responsiveness across all local party organizations across the country.

Replicating this study in future elections could help us to better understand which findings are broadly generalizable versus which findings are unique to the 2020 electoral environment. While the heightened interest in the presidential election may have led to greater responsiveness from the parties, it could have also overwhelmed local party leaders, making them less responsive to a singular email. Additionally, the 2020 election presented unique challenges to party organizations across the country. Prior to the pandemic and based on lessons learned from 2016, the Democrats were aggressively fixing their digital infrastructure (Lapowski 2019). But the pandemic forced both parties to focus on digital efforts (Harris 2020) although Republicans returned to traditional canvassing by the end of the summer while Democrats relied on phone- and text-banking (Khalid 2020).

As political scientists continue to study who participates in American politics and the increasing divide based on partisan lines, we must not overlook the influence of local party organizations. The aftermath of the 2020 election has demonstrated their importance, as local Republican party leaders promoted conspiracy theories (Burke et al. 2021), organized transportation to the January 6 “Stop the Steal” rally (Hananoki 2021), and later censured Republican members of Congress who voted for President Trump’s impeachment or conviction (Leonard 2021). The resources available to these organizations and the context in which they operate will continue to influence who gains access to the political system.

Supplemental Material

Supplemental Material - The Party Replies: Examining Local Party Responsiveness to Prospective Campaign Volunteers

Supplemental Material for The Party Replies: Examining Local Party Responsiveness to Prospective Campaign Volunteers by Lee Hannah, Kevin Reuning, and Anne Whitesell in Political Research Quarterly

Footnotes

Acknowledgments

The authors would like to thank Jessica Church, Jon Green, Hans Hassell, Seth Masket, Matt Miles, and Travis Ridout for their feedback and comments. They also thank the anonymous reviewers and editor for their helpful guidance.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

A Lee Hannah

Kevin Reuning

Anne Whitesell

Supplemental Material

Replication Files:

Notes

References

Aldrich

John

. 2011. Why Parties? A Second Look. University of Chicago Press.

Bateson

Regina.

2020. “Strategic Discrimination.” Perspective on Politics 18 (4): 1068–1087.

Bawn

Kathleen

Cohen

Martin

Masket

Seth

Noel

Hans

Zaller

John

. 2012. “A Theory of Political Parties: Groups, Policy Demands and Nominations in American Politics.” Perspectives on Politics 10 (3): 571–97.

Beck

Paul A.

Heidemann

Erik D.

. 2014. "Changing Strategies in Grassroots Canvassing: 1956–2012." Party Politics 20 (2): 261–274.

Beck

Paul A.

Gunther

Richard

Nisbet

Erik

. 2018. “What Happened to the Ground Game in 2016?” In The State of the Parties: The Changing Role of Contemporary American Parties, edited by Green

John C.

Coffey

Daniel J.

Cohen

David B.

, 263–277. Lanham, MD: Rowman & Littlefield Publishers, Inc.

Blumberg

Melanie J.

Binning

William C.

Lewis

Sarah K.

Green

John C.

. 2011. “Party on the Periphery: The 2008 Presidential Campaign in Mahoning County.” In The State of the Parties: The Changing Role of Contemporary American Parties, 6th Ed, edited by Green

John C.

Coffey

Daniel J.

, 101–114. Lanham, MD: Rowman & Littlefield Publishers, Inc.

Broockman

David E

. 2013. “Black Politicians are More Intrinsically Motivated to Advance Blacks’ Interests: A Field Experiment Manipulating Political Incentives.” American Journal of Political Science 57 (3): 521–536.

Brooks

John

Chatfield

Sarah

. 2020. “Mind the Gap: Examining the Role of Gender in Campaign Staffing and Compensation.” Electoral Studies 67: 1–10.

Burke

Garance

Mendoza

Martha

Linderman

Juliet

Fenn

Larry

. 2021. “Some Local GOP Leaders Fire Up Base with Conspiracies, Lies.” Associated Press. https://apnews.com/article/local-gop-leaders-conspiracy-theories-0ef88e6f3b2b418ae07b214131d7f128

10.

Butler

Daniel M

. 2014. Representing the Advantaged: How Politicians Reinforce Inequality. New York, NY: Cambridge University Press.

11.

Butler

Daniel M.

Broockman

David E.

. 2011. “Do Politicians Racially Discriminate against Constituents? A Field Experiment on State Legislators.” American Journal of Political Science 55 (3): 463–77.

12.

Butler

Daniel M.

Preece

Jessica Robinson

. 2016. “Recruitment and Perceptions of Gender Bias in Party Leader Support.” Political Research Quarterly 69 (4): 842–851.

13.

Chewning

Taylor K

Green

Jon

Hassell

Hans J.G.

Miles

Matthew R.

. 2021. “Campaign Voters as Faithful and Representative Agents.” Presented at the American Political Science Association Conference.

14.

Clark

Alistair

. 2004. “The Continued Relevance of Local Parties in Representative Democracies.” Politics 24 (1): 35–45.

15.

Clark

John A.

Lockerbie

Brad

Wielhouwer

Peter W.

. 1998. “Campaign Activities.” In Party Organization and Activism in the American South, edited by Steed

Robert P.

Clark

John A.

Bowman

Lewis

Hadley

Charles D.

, 119–133. Tuscaloosa, AL: University of Alabama Press.

16.

Cotter

Bibby

John F.

Gibson

James L.

Huckshorn

Robert J.

. 1984. Party organizations in American politics. University of Pittsburgh Press.

17.

Costa

Mia

. 2017. “How Responsive Are Political Elites? A Meta-Analysis of Experiments on Public Officials.” Journal of Experimental Political Science 4: 241–54.

18.

Crowder-Meyer

Melody

. 2011. “The Party’s Still Going: Local Party Strength and Activity in 2008.” In State of the Parties: The Changing Role of Contemporary American Parties, 6th Ed, edited by Green

John C.

Coffey

Daniel J.

, 115–134. Lanham, MD: Rowman & Littlefield Publishers, Inc.

19.

Crowder-Meyer

Melody

. 2013. “Gendered Recruitment Without Trying: How Local Party Recruiters Affect Women’s Representation.” Politics & Gender 9: 390–413.

20.

“County Presidential Election Returns, 2016” . 2021. MIT Election Data + Science Lab. https://electionlab.mit.edu/data

21.

Doherty

David

Dowling

Conor M.

Miller

Michael G.

. 2019. “Do Local Party Chairs Think Women and Minority Candidates Can Win? Evidence from a Conjoint Experiment.” Journal of Politics 81 (4): 1282–1297.

22.

Doherty

David

Dowling

Conor M.

Miller

Michael G.

. 2021. “The Conditional Effect of Local Party Organization Activity on Federal Election Outcomes.” Journal of Elections, Public Opinion and Parties 31 (3): 368–387.

23.

Downs

Anthony

(1957). An Economic Theory of Democracy. New York: Harper & Row.

24.

Einstein

Katherine Levine

Glick

David

. 2016. “Does Race Affect Access to Government Services?: An Experiment Exploring Street-Level Bureaucrats and Access to Public Housing.” American Journal of Political Science 61 (1): 100–116.

25.

Enos

Ryan D.

Hersh

Eitan D.

. 2015. “Party Activists as Campaign Advertisers: The Ground Campaign as a Principal-Agent Problem.” American Political Science Review 109 (2): 252–78.

26.

Feigert

Frank B.

Todd

John R.

. 1998. “Party Maintenance Activities.” In Party Organization and Activism in the American South, edited by Steed

Robert P.

Clark

John A.

Bowman

Lewis

Hadley

Charles D.

, 105–118. Tuscaloosa, AL: University of Alabama Press.

27.

Fox

Richard L.

Lawless

Jennifer L.

. 2010. “If Only They’d Ask: Gender, Recruitment, and Political Ambition.” The Journal of Politics 72 (2): 310–326.

28.

Franklin

Daniel

Richey

Sean

Yonk

Ryan M.

. 2013. "Battlegrounds and Budgets: State-Level Evidence of Budget Manipulation in Competitive Presidential Election States." State and Local Government Review 45 (2): 108–115.

29.

Gaddis

S. Michael

. 2018. “An Introduction to Audit Studies in the Social Sciences.” In Audit Studies: Behind the Scenes with Theory, Method, and Nuance, edited by Gaddis

S. M.

. Springer.

30.

Gerber

Alan S.

Green

Donald P.

. 2000. “The Effects of Canvassing, Telephone Calls, and Direct Mail on Voter Turnout: A Field Experiment.” American Political Science Review 94 (3): 653–663.

31.

Gibson

James L.

Frendreis

John P.

Vertz

Laura L.

. 1989. “Party Dynamics in the 1980s: Change in County Party Organizational Strength, 1980-1984.” American Journal of Political Science 33 (1): 67–90.

32.

Gillion

Daniel Q.

Ladd

Jonathan. M

Meredith

Marc

. 2020. “Party Polarization, Ideological Sorting and the Emergence of the US Partisan Gender Gap.” British Journal of Political Science 50 (4): 1217–1243.

33.

Green

Donald P.

Gerber

Alan

Nickerson

David W.

. 2003. “Getting Out the Vote in Local Elections: Results From Six Door-to-Door Canvassing Experiments.” Journal of Politics 65 (4): 1083–1096.

34.

Hananoki

Eric

. 2021. “Dozens of Republican Party Groups Used Facebook to Help Organize Bus Trips to DC for Pro-Trump Insurrection.” Media Matters. https://www.mediamatters.org/january-6-insurrection/dozens-republican-party-groups-used-facebook-help-organize-bus-trips-dc-pro

35.

Hansen

John Mark

. 2016. “Mobilization, participation, and political change.” Party Politics 22 (2): 149–157.

36.

Harris

Ricki

. 2020. “How the Pandemic Reshaped Election Campaigns – Maybe Forever.” Wired. August 21. https://www.wired.com/story/pandemic-reshaped-2020-election-campaigns-democrats-republicans/

37.

Hassell

Hans J. G

. 2020. “It’s Who’s on the Inside That Counts: Campaign Practitioner Personality and Campaign Electoral Integrity.” Political Behavior 42 (4): 1119–1142.

38.

Hayes

Thomas J.

Bishin

Benjamin G.

. Do Elected Officials Serve the Poor on Health Care? Evidence From a Field Experiment on Members of Congress and State Legislators. The Social Sciences Journal 2020; 1–17.

39.

Hershey

Marjorie R

. 2017. Party Politics in America: 17th Edition. Routledge.

40.

Hulbary

William E.

Bowman

Lewis

. 1998. “Recruiting Activists.” In Party Organization and Activism in the American South, edited by Steed

Robert P.

Clark

John A.

Bowman

Lewis

Hadley

Charles D.

, 13–31. Tuscaloosa, AL: University of Alabama Press.

41.

Isenstadt

Alex

. 2018. “Trump Launches Unprecedented Reelection Machine.” Politico.December 18. https://www.politico.com/story/2018/12/18/trump-machine-swallows-rnc-1067875

42.

Jacobson

Gary C

. 2015. “How Do Campaigns Matter?” Annual Review of Political Science 18:31–47.

43.

Khalid

Asma

. 2020. “Republicans are Knocking on Doors. Democrats Aren’t. Biden’s Campaign Says That’s OK.” NPR. September 13. https://www.npr.org/2020/09/13/911460651/republicans-are-knocking-on-doors-democrats-arent-biden-s-campaign-says-that-s-o

44.

Landgrave

Michelangelo

. 2020. “Can We Reduce Deception in Elite Field Experiments? Evidence from a Field Experiment with State Legislative Offices.” State Politics & Policy Quarterly 20: 1–19.

45.

Landgrave

Michelangelo

Weller

Nicholas

. 2020. “Do More Professionalized Legislatures Discriminate Less? The Role of Staffers in Constituency Service.” American Politics Research 48 (5): 571–578.

46.

Landgrave

Michelangelo

Weller

Nicholas

. 2022. “Do Name-Based Treatments Violate Information Equivalence? Evidence from a Correspondence Audit Experiment.” Political Analysis 30 (1): 142–148.

47.

Lapowski

Issie

. 2019. “Inside the Democrats’ Plan to Fix Their Crumbling Data Operation.” Wired. April 2. https://www.wired.com/story/democrats-fix-crumbling-data-operation/

48.

Leonard

Ben

. 2021. “Party Leaders Rip Republicans who Voted to Convict Trump.” Politico. February 13. https://www.politico.com/news/2021/02/13/republican-senators-convicted-trump-backlash-469008

49.

Mayceko

Melissa

. 2021. “Political Campaigns, Voter Outreach, and American Democracy: Socializing Effective Participation and Citizen Agency in the United States.” American Anthropologist 123 (3): 539–551.

50.

Mazzetti

Mark

Goldman

Adam

. 2020. “Erik Prince Recruits Ex-Spies to Help Infiltrate Liberal Groups.” The New York Times, March 7, 2020. https://www.nytimes.com/2020/03/07/us/politics/erik-prince-project-veritas.html

51.

Niven

David

. 1998. “Party Elites and Women Candidates: The Shape of Bias.” Women & Politics 19 (2): 57–80.

52.

Pew Research Center . 2020. “Democrats Made Gains From Multiple Sources in 2018 Midterm Victories.” September. https://www.pewresearch.org/methods/2020/09/08/democrats-made-gains-from-multiple-sources-in-2018-midterm-victories/

53.

Roscoe

Douglas D.

Jenkins

Shannon

. 2016. Local Party Organizations in the Twenty-First Century. SUNY Press.

54.

Roscoe

Douglas D.

Jenkins

Shannon

. 2021. "Amateur Hour: The dominance of Purposive Benefits Among Local Political Party Chairs." Social Science Quarterly 102 (4): 1602–1614.

55.

Run for Something . 2021. “Reverse Coattails – Run for Something.” November. https://runforsomething.net/wp-content/uploads/2021/03/RFSBlueLabs-Reverse-Coattails.pdf

56.

Shaw

Daron R

. 2006. The Race to 270: The Electoral College and the Campaign Strategies of 2000 and 2004. Chicago, IL: University of Chicago Press.

57.

Sinclair

Betsy

McConnell

Margaret

Michelson

Melissa R.

. 2013. “Local Canvassing: The Efficacy of Grassroots Voter Mobilization.” Political Communications 30 (1): 42–57.

58.

Social Security Administration . 2020. “National Data on the Relative Frequency of Given Names in the Population of U.S. Births Where the Individual Has a Social Security Number.” https://www.ssa.gov/oact/babynames/limits.html

59.

Sorauf

Frank J

. 1960. “The Silent Revolution in Patronage.” Public Administration Review 20 (1): 28–34. DOI: 10.2307/974064

60.

Tzioumis

Konstantinos

. 2018. “Demographic Aspects of First Names.” Scientific Data 5: 180025.

61.

United States Census . 2010. “2010 Census Urban and Rural Classification: Percent Urban and Rural by State and County.” https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural/2010-urban-rural.html

62.

United States Department of Agriculture . 2020. “Unemployment and Median Household Income for the U.S., States, and Counties, 2000-2019.” https://www.ers.usda.gov/data-products/county-level-data-sets/download-data/

63.

Whitesell

Anne

Reuning

Kevin

Hannah

A. Lee

. 2022. “Local Political Party Presence Online.” Party Politics, forthcoming.

64.

Wojcik

Stefan

. 2018. “Do Birds of a Feather Vote Together, or Is It Peer Influence?” Political Research Quarterly 71 (1): 75–87.

65.

Word

David L.

Coleman

Charles D.

Ninziata

Robert

Kominski

Robert

. N.d. “Demographic Aspects of Surnames from Census 2000.” https://www2.census.gov/topics/genealogy/2000surnames/surnames.pdf

66.

Zingher

Joshua N

. 2018. “Polarization, Demographic Change, and White Flight from the Democratic Party.” Journal of Politics 80 (3): 860–872.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.42 MB