Creating a Culture of Voting in Direct and Generalist Practice: Training Field Instructors

Social workers have an ethical responsibility to be engaged in policy change, regardless of their practice area or specialization. Voter engagement and the importance of political power through voting is often overlooked in the literature as a valid and important component of social work practice. Creating a culture of nonpartisan voter engagement in practice settings can help empower individuals who have been historically and intentionally disenfranchised from our electoral system. Training for field instructors, faculty, and field staff is a key aspect of voter engagement in social work education. Unfortunately, social work education is unlikely to include substantive content on voter engagement or its connection to social work practice and impact. This article presents one component of a model for integrating voter engagement into social work education: the provision of training for field instructors on nonpartisan voter engagement at two universities over two years. Evaluation findings suggest that pre-existing levels of political efficacy affect the reaction of field instructors to nonpartisan voter engagement training. Furthermore, findings indicate that field instructors who receive voter engagement training are more likely to serve as resources for their students and to consider voter engagement as part of their own practice. We offer evidence on the important role field educators can play in the success of the larger national effort to integrate voter engagement in social work education. Increasing awareness of what social workers, nonprofit, and public agencies are allowed--or even required--to do is a critical first step.

social work's history (Mahaffey & Hanks, 1982). Yet social workers, even while voting at relatively high rates themselves, are likely to see voter engagement as either macro social work or outside of social work practice (Ritter, 2008).
Social workers can engage in the voting process in a variety of ways (Lane & Pritzker, 2018). The most basic involvement is to engage civically by voting; encouraging others to vote; and staying engaged and educated about policy, candidates, and the election process. The next level of involvement includes more active engagement of others, who could include colleagues, client groups, and community members. This level includes assisting others with voter registration; education about the process of when, where, and how to vote; and sharing information about candidates and the issues. These steps can be partisan or nonpartisan depending on the social worker's role and employment. A social worker who is employed by a political candidate's campaign will do partisan voter engagement, while a social worker in a nonprofit setting can do nonpartisan voter engagement. The final level of voter engagement for social workers includes leadership in voter engagement, ranging from running voter engagement drives or voter education events, working as an election official, training and mentoring others in voter engagement, and engaging with the media about related issues.
Increased levels of voting by clients and communities served by social workers could have a significant impact on the outcome of elections (Lane, Humphreys, Graham, Matthews, & Moriarty, 2007). Building political power of oppressed clients and communities is a key component of taking an empowerment approach in social work practice (Postle & Beresford, 2007), as it explicitly recognizes clients as "members of the community with certain rights and obligations" (Davis, 2010, p. 245) rather than as passive recipients of services. Voting addresses key empowerment concepts such as inclusion, mastery and control over key issues, self and community advocacy, and representation in the power structure (Davis, 2010;Gutierrez, 1999). Treating social work clients as full members of society recognizes their essential humanity and importance in the creation of a functioning democratic society. Recognizing clients as the "experts in their own lives", Postle and Beresford (2007) argue for connecting social work values of client empowerment and self-advocacy and recognizing the political power of clients and the need to create pathways to democratic and political participation (p. 146). The NASW Code of Ethics (2017) explicitly calls for social workers to empower clients. Integrating voter engagement into our practice is an effective, professional, and legal way to meet this ethical mandate.
The social work profession is well-positioned to engage oppressed and underrepresented communities in voting, but social workers do not always seize these opportunities. Studies show that many social workers still wrongly believe voter mobilization is partisan, illegal or unimportant (Rocha, Poe, & Thomas, 2010;Rome, Hoechstetter, & Wolf-Branigin, 2010). In social work practice political power remains an overlooked indicator of self-actualization and community health (Klar & Kasser, 2009). Social workers can create alliances with clients to create change. Indeed, fostering political engagement is "a crucial part of the social workers' role, consistent with building capacity within groups of people using services and thus enabling them to work autonomously" (Postle & Beresford, 2007, p. 153). Social workers can help to address barriers to voting, which include confusing registration processes and deadlines; lack of transportation; not knowing where, when or how to vote; long lines; little to no information about who is on the ballot; and a feeling that one's vote does not really matter (Piven & Cloward, 2000). Many people are disengaged and disconnected from voting; nearly 40% of nonvoters state they do not like the candidates, the issues, or believe their vote does not matter (Frint, 2018;López & Flores, 2016). Research suggests successful, ongoing engagement around voter registration, voter education, and encouragement to vote can increase social dialogue around voting and support voting as a social norm (Nonprofit VOTE, 2013). This is similar to findings that nonpartisan door-knocking campaigns to encourage people to vote are effective because of the social interaction between the volunteer and potential voter (Bedolla & Michelson, 2012). When nonvoters engage in dialogue about voting in general, not related to a particular campaign, it can influence their perceptions and increase their interest in voting (Rolfe, 2012;Sandler, 2017).
This article describes the implementation of a model of nonpartisan voter engagement within schools of social work. Leveraging the strength of social work's signature pedagogy of field education (CSWE, 2015;Wayne, Bogo, & Raskin, 2010) to reach students, field instructors, and educators, this model posits that training agency-based field instructors on, and working with these instructors to identify barriers to voter engagement, is pivotal to establishing a culture of voting.

Literature Review
A direct relationship exists between voting, power, resource allocation, and community well-being. Voting benefits individuals and their communities, including higher levels of civic participation, stronger relationships, and better outcomes for the individual voters themselves, in areas like health, social connections, mental health, and overall well-being (Avery, 2015;Blakely, Kennedy, & Ichiro, 2001;Griffin & Newman, 2005;Klar & Kasser, 2009;Martin, 2003). Voting influences political decision-making (Avery, 2015;Griffin & Newman, 2005), overall community health (Blakely et al., 2001;Martin, 2003), and overall health and well-being on an individual level (Klar & Kasser, 2009;Sanders, 2001). Studies show that elected officials award more attention and more resources to communities that vote in higher numbers (Martin, 2003;Martin & Claibourn, 2013). Thus, the advantages of engaging voters extend beyond partisan concerns of specific political offices into individual and community well-being, which are directly related to social work's values, professional mandate, mission, and impact.
Disadvantaged and oppressed groups served by social work face obstacles in voting (Rolfe, 2012) and their lack of political participation harms both their lives and our democracy. Many individuals and communities do not vote regularly. There are many reasons that individuals and communities may not participate in voting. In many states, there are both official and unofficial restrictions that lead to voter disengagement and voter suppression (McElwee, 2015). Low-income communities, people of color, and young voters are disproportionately affected by structural barriers (McElwee, 2015;Rolfe, 2012). These structural barriers include the disenfranchisement of people who have been convicted of felonies (The Sentencing Project, 2016), registration deadlines, requiring photo identification, and closing or relocating polling places in certain districts (Brennan Center for Justice, 2018;McElwee, 2015, Rolfe, 2012. Low voter participation, particularly by members of oppressed communities, reduces the likelihood of responsive governmental solutions to problems that are prevalent in these communities (Bartels, 2008;Frasure & Williams, 2009;Verba, Schlozman, Brady, & Nie, 1993).
In 2017, one qualitative study asked community college students who received meanstested government assistance about their approach to voting, both in terms of internal decision-making and structural factors (Sandler, 2017). In this study, a subset of respondents were very strategic about their approaches to voting and were aware of the relationship between electoral politics and government assistance. Interestingly, this includes people who chose not to vote because of their knowledge of political structures. Other members of the sample, both those who voted and those who did not, were less aware of the structural connections. This study suggests that multiple intervention strategies and tactics are needed to encourage voting among this group who are very affected by policy decisions (Sandler, 2017). More research is necessary to develop the connection between the experiences of those who receive government assistance and the political process.

Voter Engagement and Social Work Practice
Social workers have a long history of registering clients to vote beginning with the women's suffrage in the mid-19th century, the civil rights movement in the 1960s, and continuing with the work of Human SERVE in the 1980s (Piven & Cloward, 1988, 2000. Piven and Cloward founded the Human SERVE (Service Employees Registration and Voter Education) Project (Piven, 2011). The project mobilized social workers and successfully campaigned for the passage of the National Voter Registration Act (NVRA) of 1993, commonly referred to as the "motor voter" law. To this day, the NVRA enables public agencies (including social service agencies) to help clients register to vote (Piven, 2011). In the present day, the National Social Work Voter Mobilization Campaign is a national collaborative effort intended to train social work practitioners, students, and educators in voter engagement for the 2018 elections and beyond (Voting is Social Work, n.d.).
However, despite the documented need and historical roots, many social workers are not politically active within their professional roles. One reason for this may be due to popular misconceptions about what activities are acceptable within the structure of public and non-profit agencies. Many social workers believe that they are not allowed to engage in voter engagement within their organizations (Hardina, 2013;Lane et al., 2007;Rocha et al., 2010). Contrary to these common misperceptions, most agencies that employ social workers are allowed to engage clients in the voting process, and some are legally required to do so. For example, the 1998 Reauthorization of the Higher Education Act of 1965 has required institutions of higher education such as colleges and universities to conduct voter registration activities in years where there are federal or gubernatorial elections (American Association of Collegiate Registrars and Admissions Officers, n.d.). The National Voter Registration Act mandates that all state agencies which administer driver's licenses, welfare assistance (now Temporary Assistance to Needy Families), food stamps (now Supplemental Nutrition Assistance Program), Medicaid, and disability benefits offer assistance in registering to vote (Piven & Cloward, 2000). Additionally, nonpartisan voter engagement, education, and outreach, which stays away from support of individual candidates or political parties, is permitted in nonprofit organizations (Nonprofit VOTE, 2017).

Voter Engagement and Social Work Education
Most social workers are not provided substantive opportunities to increase their knowledge or experience of voting or other political activities while they are in social work schools, field settings, or in professional development training or continuing education activities once they enter practice. Consequently, the connections between these political activities and the social work profession are generally not learned. Past studies (Ritter, 2008;Rome & Hoechsetter, 2010) found that over half of licensed social workers are only minimally involved or not involved at all in political activities, including voting. Pritzker and Burwell (2016) found that voter registration efforts are included in fewer than half of all accredited social work programs. An increase in civic engagement in social work education is necessary in order to continue the social work profession's ethical commitment to creating and implementing just and inclusive social policies (Hylton, 2015;McCabe, Hylton, Kooreman, Mellinger, & Day, 2017).
Given that voting behaviors are largely habitual (Miller & Shanks, 1996) and developmental (Plutzer, 2002), strengthening participation habits early is important. Pritzker, Springer, and McBride (2015) suggest that both classroom-based and communitybased opportunities for political engagement and civic involvement in college are effective in laying the foundation for adults' ongoing engagement. Furthermore, a deeper understanding of specific "regulatory and legislative processes" (McCabe et al., 2017, p. 91) may be even more critical for social workers than the general public, given the expectation of competency in policy practice set forth by CSWE (CSWE, 2015). Several studies have indicated that professional social workers feel that they did not receive adequate opportunities for developing political advocacy skills during their Master of Social Work (MSW) programs (Hill, Donaldson, Fogel, Erickson, & Ferguson, 2017;Ostrander, 2016;Ritter, 2008;Rome & Hoeschetter, 2010). Integrating voter engagement activities into social work education can be an opportunity for students to practice political advocacy that is nonpartisan and impactful, while also building on and using their core social work practice skills (Hylton, Rhodes Smith, Powers, Ostrander, & Lane, 2018).
Field education is a critical component of social work education with long-term impacts on future social workers' preferences, competencies, and understandings of social work practice (Deal, Hopkins, Fisher, & Hartin, 2007;Mizrahi & Dodd, 2013;Wayne et al., 2010). Field settings provide students with hands-on learning, access to social work role models, and mentorship that they may carry throughout their practice. In addition, direct-service agencies are particularly effective in engaging underrepresented communities and individuals to participate directly with electoral processes (LeRoux & Krawczyk, 2012;Nonprofit VOTE, 2017). Unfortunately, little political or policy content is present in most social work students' field placements (Pritzker & Lane, 2014). Further, much of the emphasis on voter engagement is on the rules and process of registering people to vote versus how it connects to social work's organizations, missions, or impacts.
Lane and colleagues (2007) examined agency-based voter registration programs through their review of MSW students' voter registration projects within their field placements. Evaluation of this project found that "sustaining momentum, supplying userfriendly information, and providing open lines of communication for assistance" (Lane et al., 2007, p. 85) are all critical components to successful implementation of a voter registration project in a field setting. Challenges to implementing voter engagement projects within a field setting included the short timelines between the beginning of field placements and voter registration deadlines, competing demands in field placements, and client ineligibility to vote (Lane et al., 2007). Findings also indicated that voter registration work in agencies should be an ongoing activity, rather than something done only once or only immediately preceding an election. Finally, participants of this study recommended training on voter engagement for both students and field instructors (Lane et al., 2007).

Voter Engagement Model
Given that field education is social work education's signature pedagogy (Wayne et al., 2010), the current project is centered in field education at two public universities' schools of social work. In both schools, voter engagement training is integrated into courses (policy or macro practice) which are required of all students. BSW (Bachelor of Social Work) and/or MSW students are provided training on the importance of voting to community outcomes, human rights, anti-oppression work, empowerment practice, and social work's impact. Students are also provided tools and resources on basic civics, the voting rules in their state, and how to integrate nonpartisan voter engagement into social service delivery and social work practice. To reinforce these concepts and build students' efficacy, they are asked to complete a required assignment in field. This training model is designed to provide students and practitioners with a foundation of the importance of voting, an understanding of the barriers to being an informed and consistent voter, and the knowledge needed to create a culture of voting in all practice settings. The objective of the model is for field instructors to expand their knowledge of voter engagement and support student-driven change within their field placements related to the activities outlined in the students' assignments.
Although the training module was developed jointly by faculty at both sites, there are a few key differences in its implementation. All students are asked to develop a plan for their field agency, which suggests all the ways in which their agency could support clients, staff, and communities to vote. It is an exercise that addresses micro, mezzo, and macro touchpoints and interventions. At one school, the assignment is simply a planning exercise (as they are not required to implement their plans), and students are asked to discuss the training and their plan with their field supervisor (the individual employed by the agency who supervises the student at the field site). These students also write a reflection paper on the significance of voting to social work practice, the intersection between voter engagement and communities of color, and a policy they would change to make voting easier. Students at the second school are required to implement their plans within their field placements. After implementation, students submit final papers in which they reflect on the successes, barriers, and learning outcomes of their voter engagement projects. In order to clarify this project for field supervisors and to address any questions or concerns they may have, a shortened version of the student training is given to all new field supervisors at one site during the required orientation course. This additional training was not required at the second site. Students at both schools are given a letter to bring to supervision, which provides their field instructors with background information on the training, objectives, and assignment. This voter engagement model has been implemented and evaluated since 2015 at the two schools of social work, reaching about 250 students annually, and 75 new field supervisors in total. Preliminary evaluation findings from students indicate that students are more likely to recognize the importance of voter engagement to social work practice and gain valuable knowledge about best practices for implementing voter engagement in their practice after their participation in this training model (e. g. Hill, Lane, Powers, & Smith, 2019). This paper examines the evaluation data provided by field instructors on the impact, implementation, and efficacy of these student projects.

Research Questions
Initial data from 2015, the first year of implementation, illuminated that field instructors can serve as key support for students or as substantial barriers for implementation. Lack of support from agencies also emerged as a significant barrier for 44% of students at one university and 25% at the other (Lane et al., 2016). Training for field supervisors is a key aspect of this voter engagement model because the research discussed above, along with our professional experiences, suggests practitioners' social work education was unlikely to include voter engagement content. Based on this information and evaluation results from students after the first year of implementation, increased effort was made to include field instructors, faculty, and field office staff as important stakeholders in the voter engagement model. Specifically, a shortened (45-60 minutes) version of the training presented to students was added to the annual field instructor seminar training (required of all new field supervisors) beginning in the second and third year of implementation (2016 and 2017).
Similar to the student training, the objective of this field instructor training is for participants to connect voting and voter turnout to all social work practice and its overall impact. This includes voting as a social determinant of health; a basic human right; a central component to social justice and equity efforts; and a legal, effective, and ethical intervention at the macro, mezzo, and micro levels. Training also clarified the student assignment for instructors, including providing them with an opportunity to process how this might fit into their organizational context. A longer version of this training was also offered to all field instructors as a free professional development opportunity with CEUs (continuing education units) available. The research reported here asks the questions: 1. What is the impact of voter engagement training on field instructors' perceptions of the importance of voter engagement in social work practice?
2. What barriers do field instructors perceive in implementing student-driven voter engagement projects within field agencies?
3. What is the impact of participation in voter engagement training on self-reported future behavior of field instructors?

Sample
The sample for this study included all field supervisors who participated in the trainings at two schools of social work in 2016 and 2017. The training was also provided to faculty at one of the schools. This sampling frame was chosen in order to identify the effects of this training on individuals who were agency employees, held supervisory responsibility, and who would be in a position to help or hinder students from completing voter engagement projects. Both schools of social work are located in urban areas. The school in the Northeastern United States had an MSW and PhD program, while the school in the Western United States had a BSW and MSW program. Participants were full-time employees of agencies in the communities near the two schools of social work. The majority of trainees (for whom demographic data were available, approximately n=41) were women and identified as White. A total of 120 people attended the trainings across the two locations.

Measurement Instrument
The instrument used for this study was developed by the authors for work on political engagement. It includes questions about demographics, current political knowledge and activity, previous experience with voter engagement, future plans for voter engagement, and any barriers they experienced, as well as questions about the training overall. Questions about political efficacy were taken from the American National Election Studies (ANES; 2012) refined by Niemi, Craig, and Mattei (1991). These questions measure both internal and external efficacy. Internal political efficacy refers to the confidence individuals' have in their ability to intervene in the political system, whereas external efficacy refers to the confidence individuals' have in the system's capacity to respond to their intervention. These ANES scales of internal and external efficacy have each been tested for validity and reliability (Niemi et al., 1991).
All of the items have five Likert responses, with a total score ranging from 0 to 32; high scores indicate strong political efficacy. The internal political efficacy scale was comprised of four questions ("How often do politics and government seem so complicated that you can't really understand what's going on?"; "How well do you understand the important political issues facing our country?"; "How much do public officials care what people like you think?"; and "How much can people like you affect what the government does?"). Likert response options range from 0 (never, or not at all) to 4 (always or a great deal.). Some items are reverse-coded within this internal efficacy subscale, with a total range from 0 to 16. The external political efficacy scale also had four questions ("There are many legal ways for citizens to successfully influence what the government does"; "Under our form of government, the people have the final say about how the country is run, no matter who is in office"; "If public officials are not interested in hearing what the people think, there is really no way to make them listen"; and "People like me don't have any say about what the government does"), with Likert response options ranging from 0 (strongly disagree) to 4 (strongly agree)-some of which are reverse-coded-with a subscale ranging from 0 to 16.

Data Collection and Analysis
Data were collected at the start of each training session and at the completion of training. Surveys were offered in hard copy or electronically. Approval was received from the Institutional Review Boards at the participating universities and all research partners' universities. Data was imported from Qualtrics or manually entered into SPSS and analyzed using versions 24 and 26.

Limitations
The biggest limitation of this study is the small sample size and geographical areas. We are in the process of larger scale studies in a variety of geographic areas in order to assess whether these results are characteristic of other social work field instructors. Another limitation of the study is the timing-these studies were done during the contentious 2016 U.S. presidential election and in the year following. It may be that this time period has affected the results. In an effort toward minimizing potential non-response or attrition, the researchers decided to omit survey questions asking for sensitive information, such as political affiliations. Therefore, our data limits our understanding of the potential impact that participants' political affiliations may have on their voter engagement perceptions or behaviors. Social desirability is also a concern; as the training was designed to encourage voter engagement, participants may have chosen survey answers that they expected would meet the researchers' expectations. Finally, this data only reflects the participants' selfreports that are retroactive or their immediate reflections after the training. For example, this study would be strengthened by collecting data that reflects their voting behaviors more accurately than retroactive self-reports and their planned behaviors, such as participants' publicly available voting records. However, the authors were unable to collect participants' names (in order to preserve anonymity, as required within our IRB approval) nor dedicate the extensive time required to achieve this task. Relatedly, longer-term follow up with participants would provide more information about the impact of the training, both for individual behaviors as well as for changes in participants' work with agencies (for example, advocacy for changes in agency policy around voter engagement). More research needs to be done on the long-term outcomes, as well as the effect on students of their field instructors receiving this training.

Results
A total of 44 individuals completed the pre-test before training in 2017. (Pre-tests were not administered in 2016.) Forty participants were field supervisors, one universityemployed field advisor, and three faculty members. They ranged in age from 25 to 44 (M=38.50, SD=11.52). The majority of the 41 participants who identified a race were white (n=33, 80%), with the remainder identifying as Black/African-American (n=4, 10%), Latina/Hispanic/Puerto Rican (n=2, 5%), or bi-or multi-racial (n=2, 5%). The majority who identified a gender were female (n=38, 93%), with the remaining identifying as male (n=3, 7%) and no one identified as transgender.
We asked participants a set of questions to identify their baseline political activity. Forty respondents (91%) were registered to vote. The remaining four were not registered to vote and reported no plans to register. Eighty-six percent (n=31) reported voting in federal elections always or most of the time, with just over half (n=24, 54%) reporting the same for state elections and 41% (n=18) reporting the same for local or municipal elections. On the less politically active side, six participants (14%) reported rarely or never voting in federal elections, one-fourth (n=11, 25%) rarely or never voting in state elections, and 39% (n=17) rarely or never voting in municipal or local elections. One-third of this group (n=15, 34%) had engaged in behavior to promote voting prior to the training.

Barriers
Thirty-seven (47%) of the 2016 respondents and 18 (49%) of the 2017 respondents perceived significant barriers to implementing voter engagement in their agency settings. The barriers could be broken down into five sub-themes. These are: need for approval/concerns from higher-ups, needs of clients, agency policy, attitude, and other barriers. Each is discussed in more detail below.
Need for approval/concerns from higher-ups. One of the two most commonly reported concerns (n=13, 54% of concerns listed) related to the need for approval from those in authority and their potential disapproval. One school social worker wondered if voter engagement projects would need approval from the Board of Education and another wondered if the superintendent would need to approve projects. Another reported, "Administration would need to be on board; but I will assist in advocating for it." Another said that any project would need to be approved by the CEO, which would require additional time because that person "is hard to reach and takes a long time to decide things. By the time he decides, the election may have happened." One non-profit social worker commented, "My director may not be aware that this type of work is legal in our agency setting," and others in non-profit and public agencies wondered whether their administration would be supportive.

Needs of clients.
An equal group of respondents (n=13, 54%) described concerns about the setting and clients with which they worked, particularly those who generally worked with individuals in crisis. One noted a concern about "doing this in a way that makes sense with the current level of need vs. crisis our youth experience on a day-to-day basis," and others stated: "usually only see parents at time of crisis" and "difficulty with engagement during crisis situations." Another wrote that "clients are coming to the agency for a specific service and could feel blindsided by questions about voting" while another commented, "When people are trying to survive, voting isn't exactly at the top of the list." Concerns about conducting voter engagement with several specific client populations were identified, including young children, people with cognitive disabilities, students at a private university, and people who are homeless. A field instructor who worked with survivors of domestic violence listed concerns including "the homelessness, lack of IDs (though not technically required), PTSD, and safety needs of DV survivors." Agency policy. Concerns were also raised by seven participants (13%) about agency policies that might prohibit political action. Field instructors who were most likely to raise these concerns were those in state agencies and those working for federal elected officials. It was a concern that state agencies might see voter engagement as a "conflict of interest." Field instructors working for federal elected officials, such as members of U.S. Congress, have a firm line between official and political actions. Others described more general agency policy restrictions without specifying barriers.
Attitude. Six (11%) participants discussed mindset and attitudinal concerns. One reported that their agency "is not political and would not want to do this" while others describe bureaucracy and hierarchy as barriers. Another described "fear of change" or potential controversy. One social worker noted that "even though it is 'nonpartisan' people assume an agenda" and another stated that their agency was "not willing to be involved in engaging clients in political conversations." Other barriers. The remaining 15 (28%) barriers included issues such as lack of time (n=4), the current political climate (n=3), and lack of interest by intern or others (n=2). Furthermore, participants identified the need to know more about logistics, the need to be careful of stepping on the toes of other departments already doing voter engagement and concerns that clients would move into partisan territory as possible barriers.
Impact of training. The impact of the training was measured using several variables. First, participants were asked about how important they believed voter engagement to be to their social work practice, using a scale of 0 (low importance) to 4 (high importance) prior to the training (as a retrospective baseline) and at the end of the training. The 2016 respondents moved from 2.10 to 3.45 (df=108, t=-9.247, p<.001) with 22 (26%) reporting choosing the highest response before the training, compared to 53 (62%) after the training, and the lowest level dropping from 20 (23%) to 1 (1%). The 2017 participants' change was positive and statistically significant, though lesser compared to 2016, from 2.30 to 3.03 (df=52, t=-2.361, p=.024). In this iteration, 13 (30%) chose the highest response prior to the training, 18 (42%) after the training, and the number choosing the lowest response dropped from 7 (16%) to 3 (7%).
Sixty-six total respondents (81%) reported that they planned to change their work with students and/or practice as a result of the training. Planned changes included: discussing voter engagement with all students every year, bringing voter registration forms to class (from a faculty member), inviting local elected officials to their agency, being more creative about encouraging voting, providing education about voting, encouraging students to engage clients about the topic, putting registration information in intake packets, more in-depth discussions with students, and connecting with clients' parents.
Two variables that correlated positively with participants' likelihood to change their work with students were political efficacy and their rating of the importance of voter engagement after the training. As shown in Table 1, for the 2017 respondents who could have their pre-test and post-test responses matched, efficacy was positively correlated with likelihood of changing work with students. Matching was not possible with 2016 data. Correlations were also found between participants' rating of the importance of voter engagement to their social work practice and their stated intention of changing their work with students to include voter engagement. This was true for both 2016 (r=.315, p=.004, n=81) and 2017 (r=.456, p=.010, n=31). Those who saw barriers to voter engagement in their agencies were also less likely to report that voter engagement was important to their practice in 2016 (r=-.269, p=.017, n=78).

Discussion
Multiple prior studies have indicated that integrating voter engagement into nonprofit agencies, particularly social service agencies, is an effective way to engage traditionally marginalized populations in voting (Lane et al., 2007;LeRoux & Krawczyk, 2012;Nonprofit VOTE, 2017). Moreover, LeRoux and Krawczyk (2012) report that each votingrelated contact a client receives through a service agency results in an increase in likelihood of voter turnout by 11%, even after controlling for other factors that may influence voting behavior. Findings reported by LeRoux and Krawczyk (2012) also support previous research which indicates that many social workers do not integrate voter engagement into their practice settings due to various barriers (Ritter, 2008;Rome & Hoeschstetter, 2010).
Voter education programs such as the model discussed in this paper may help address many of the barriers to voter engagement in social work practice identified in previous studies (LeRoux & Krawczyk, 2012;Ritter, 2008;Rome & Hoeschstetter, 2010). Concerns about the legality and ethics of voter engagement in nonprofit and social service agencies are widespread. However, as previously discussed, voter engagement is both legal and ethical, as long as it is nonpartisan in nature and provided in a way that is respectful of clients' social and community locations. Increasing awareness of what social workers, nonprofit, and public agencies are allowed--or even required--to do is a critical first step. Allowing social workers to conceptualize and practice how they might integrate voter engagement into their own practice settings is an equally important second step.
The findings from this training evaluation indicate that integrating voter engagement training and activities into social work field education, specifically for field instructors, is an effective method for increasing the recognition of voter engagement within social work practice. Field instructors indicated intentions to take concrete steps toward integrating voter engagement activities into their practice settings, as well as in their work with students. They also recognized the importance of voter engagement to their social work practice at higher levels after the training.
Despite these positive impacts of the training, however, participants also continued to identify barriers to creating a culture of voting within their social work practice. Many of these barriers (i.e., agency policies, needs of clients, and concerns about the political environment) have been identified in previous literature (Hardina, 2013;Postle & Beresford, 2007;Rome & Hoeschetter, 2010). Indeed, previous literature also identifies possible responses to these perceived barriers, including more education for both practitioners and social work managers about the legal, ethical, and professional frameworks that support voter engagement in social work settings (Hylton et al., 2018;Lane et al., 2007;Postle & Beresford, 2007); greater emphasis on the positive impact of voter engagement for nonprofits and their clients (LeRoux & Krawczyk, 2012); and creating learning opportunities for social work students to practice voter engagement within supported settings (Hylton et al., 2018;Lane et al., 2007).
This study supports these recommendations of increased and ongoing education for social work practitioners, as well as a broader voter engagement model discussed below. Findings suggest that field instructors' increased knowledge of voter engagement, its level of importance to their practice, and their political efficacy related to their plans to practice more voter engagement in the future. The training aimed at addressing these knowledge, attitude, and behavior changes by including the what, why, and how of voter engagement within social work practice: What is nonpartisan voter engagement; and what is allowed, or even required, within various social work agencies and organizations? Why is voter engagement important to social work's mission and impact? And how can social work practitioners navigate barriers and advocate for practices aimed at voter turnout within their agencies? Findings from this training evaluation suggest that field instructors' increased knowledge of voter engagement increases their planned voter engagement behaviors. However, field instructors may need additional guidance in order to overcome potential barriers, both perceived and actual, to incorporating voter engagement within their professional practice.

Implications for Practice
Short-term trainings are helpful, but are not enough to overcome the barriers discussed. Evaluation is currently underway of the overall voter engagement model, of which this field instructor training is one component, that aims to empower social workers to create a culture of voting within their practice settings to increase voter turnout of marginalized communities and, thus, be represented within policy-making in order improve their outcomes. The overall model includes providing ongoing connection and support for voter engagement in social work settings. One method of continuous access to voter education includes providing field instructors, students, and schools of social work with a number of voter engagement resources, some of which are outlined in Table 2. The Institute works to increase the political power of all social workers and the communities they serve. To learn more about their projects and for contact information, go to their website: ssw.uconn.edu/political institute. Nonprofit VOTE (n.d.) A national partner for nonprofit organizations that fosters their engagement of the people they serve in voting and political engagement. Resources, including free webinars, can be accessed on their website: www.nonprofitvote.org.

TurboVote (Democracy Works, 2019)
Turbovote.org is an online application that aims to make voting easier and more accessible to all Americans.
Vote.org (2018) Vote.org is a clearinghouse of voter information by state.

Implications for Education
By linking voter engagement with field education, the overall voter engagement model described above builds on the pre-existing strengths of social work education. This education model is also an example of best practices in civic education, including leveraging pre-existing trusting relationships to share voter registration information (Bogard, Sheinheit, & Clarke, 2008); explicitly linking community-based (field) and classroom-based activities (Pritzker et al., 2015); and creating multiple opportunities for students to practice citizenship while they are enrolled in postsecondary or graduate education (McCabe et al., 2017;Pritzker et al., 2015). Increasing social workers' civic "toolkit" of skills such as civic knowledge, political efficacy, and participation in political activities will serve the training participants and our profession's mission and impact into the future (Hylton et al., 2018).

Implications for Policy
In order to effectively advocate for marginalized and vulnerable populations, social workers must be aware of discriminatory policies that contribute to voting disparities as well as policies and programs that increase voter turnout. Informing social work practitioners about these policies can increase the effectiveness of voter engagement strategies as well as help to alleviate concerns about the legality of nonpartisan voter engagement within social service agencies. Given that the field instructors involved in this training saw agency prohibitions to voter engagement as a barrier, social work practitioners may need to be educated about the rights of staff and clients in relation to nonpartisan voting activities. If unjust policies do in fact exist, training programs must empower social workers to advocate for policy change within their agencies. In order to justly serve their clients and communities, nonprofit and social service agencies-where nonpartisan voter engagement is legally permitted-must implement policies that encourage a culture of voting.
Similarly, in order to adequately advocate and support their clients, social work practitioners also need to be informed of the government-level policies that inhibit or support equal access to voting for disadvantaged populations. As previously noted, many of the communities that social workers work with face multiple barriers to voter engagement (Hill et al., 2019;Hylton et al., 2018). There are numerous laws that specifically limit the ability of many Americans to vote. This category of laws includes those that disenfranchise individuals with prior felony convictions, photo identification laws, purges of voting rolls, closing of polling places, or limiting the hours that polls are open. Each of these laws negatively impact the ability of many people to vote, particularly young people, people of color, people in poverty, and people who are highly mobile (McElwee, 2015).
Social workers should advocate for national, state, local, and organizational policies that seek to remove systemic barriers to voting. Social workers have been active in advocating for voting rights throughout history: the movements for women's suffrage from the mid-19th century to the early 20th century, the civil rights movement in the 1960s, and the work of Human SERVE in the 1980s (Piven & Cloward, 1988, 2000. In keeping with social work's historic and current commitment to social justice, access, and empowerment, social workers should advocate for policies that support engaging all of our clients in voting. For example, policies such as same-day voter registration, early voting, no-excuse absentee voting, and vote-by-mail have each been demonstrated to have positive impacts on voter turnout, particularly for marginalized groups (National Conference of State Legislators, 2017;Perez, 2017). Social workers should continue to advocate for policies that increase access to voting for all members of our communities, and ensure that they are being implemented effectively. The social, political, and economic justice of historically oppressed communities largely relies on the social work profession ensuring that policies and practices enhance the representation of all voters within local, state, and federal elections.

Conclusion
Voter engagement is a professional, legal, and ethical obligation for social workers. By building upon one of the primary areas of strength of social work education, we are able to help students and community agencies to connect voter engagement to the social work field. Field education as the venue for voter engagement education allows both current and future social workers to learn why and how to incorporate voter engagement into their practice, thus magnifying the reach of social work education and practice. Social workers are well-situated to play a critical role in engaging communities within the political process who have typically been disenfranchised and disconnected from the political arena. This begins with providing all clients and communities with equitable opportunities and access to increasing their participation in voting. Taking on this work is consonant with our profession's historic commitments and current ethical responsibilities to empowerment and social, economic, and political justice. The findings from the evaluation of these trainings evidences the important role field educators can play in the success of the larger national effort to integrate voter engagement in social work education.