Essay Grade

Journal of Policy Practice, 10:225–244, 2011
Copyright © Taylor & Francis Group, LLC
ISSN: 1558-8742 print/1558-8750 online
DOI: 10.1080/15588742.2011.582809
Social Workers in Politics: A National Survey
of Social Work Candidates and Elected Officials
SHANNON R. LANE
Adelphi University School of Social Work, Garden City, New York, USA
NANCY A. HUMPHREYS
Nancy A. Humphreys Institute for Political Social Work, University of Connecticut,
West Hartford, Connecticut, USA
This research identified 416 social workers who have run for local,
state, or federal elected office, the largest number of such social
workers identified and surveyed. The Civic Voluntarism Model was
used to describe the factors leading to their political engagement.
This model examines the resources, recruitment, and expertise of
individuals and how those factors play into their public involvement. This model begins to describe the path these social workers
took to elected office, and can help determine the opportunities for
social work education and professional social work organizations
who would like to influence this process. These results can be used
to increase the number of social workers who seek elected office.
KEYWORDS advocacy, campaigns, elected office, policy, political
engagement, political social work, social work education
Elected officials in the United States are likely to come from professional
backgrounds in law, business, higher education and politics (Lawless & Fox,
2005). The 111th Congress included 186 representatives and senators
with law degrees (Manning, 2010), while eight had social work degrees
(NASW, 2009).
Research suggests that changing the cast of characters in politics affects
the policies that are implemented (Burrell, 1996; Dodson, 1998; Swers, 2002).
For example, adding more women to office makes it more likely that policy
makers will consider issues that significantly and disproportionately affect
women, such as health research that prioritizes women’s health issues,
Address correspondence to Shannon R. Lane, Adelphi University School of Social Work,
1 South Avenue, Garden City, NY 11530. E-mail: Slane@adelphi.edu
225
226
S. R. Lane and N. A. Humphreys
foreign policy that emphasizes women’s rights, and domestic policies that
affect child care (Dolan, 2001). Legislative bodies with higher percentages
of women are also likely to support rehabilitative criminal justice policies
over punitive criminal justice policies (Swers, 2001). Specific policies that
have had heavy input by female officials include changes to Title IX of
the Education Amendments of 1972 that created opportunities for women
in high school and college athletics (Palley, 2001) and policies related to
domestic violence (Murphy, 1997).
Changing the face of elected politics to include non-Caucasians also
affects the resulting policies. Adding more African Americans to office has
been known to intensify the focus of a legislative body on civil rights
(Bratton & Haynie, 1999). Literature about Asian Americans in office suggests that they “have assumed the role of advocates for all Asian Americans,”
including leadership on legislation relating to redress for World War II
internment (Lai, Tam Cho, Kim, & Takeda, 2001, p. 613).
Political activity is specifically promoted among social workers in the
National Association of Social Workers (NASW) Code of Ethics in section 6.04,
where social workers are charged with “social and political action that seeks
to ensure that all people have equal access” to resources (NASW, 1999, n.p.).
Meetings of the NASW Delegate Assembly, NASW’s national representative
decision-making body, have encouraged social workers to get involved in
politics. NASW’s actions to increase social workers’ political activity have
included operation of political action committees, organization of political training programs, endorsement of specific candidates who advance
the profession’s agenda, encouragement of social workers seeking office,
and promotion of voter registration (NASW, 2009–2012). Despite this, social
workers have not traditionally been represented among legislators (NASW,
2008), legislative staff (personal communication, Dina Kastner, March 17,
2004), or those who influence the legislative process (Hoefer, 2000). By 2008,
only 201 social work-trained elected officials had been identified nationwide
(NASW, 2008).
The authors are part of a growing movement to define political social
work as a practice specialization. Political social work includes social workers who have run for or hold office, as discussed in this article, work for
elected office holders, are appointed by elected officials or must be confirmed by elected officials, and social workers who spend considerable time
lobbying elected officials as a volunteer or paid advocates. One other type
of political social worker is anticipated to emerge in the future. Some years
ago Ruth Messinger argued that social service agencies should have a political social worker on staff who would be responsible for organizing client
and staff voter registration activities, coordinating political empowerment
and advocacy activities and lead the agency’s legislative agenda, although
agencies may prefer not to use a “political” name to avoid concerns about
partisanship (personal communication, IAPSWP board meeting, 1999).
Social Workers in Politics
227
LITERATURE REVIEW AND HISTORY
The first social worker to hold elected office in the United States was the
late Jeanette Rankin, a suffrage activist who had studied at the New York
School of Philanthropy (Davidson, 1994; Josephson, 1974). Rankin was the
first woman to run for Congress, the first woman to be successfully elected
to Congress, and the first woman to be elected to any national representative
body in the world. In 1971, Ron Dellums became the second social worker
elected to the U.S. House of Representatives. Dellums was known for his
peace activities, his radicalism, founding the Congressional Black Caucus,
and serving on the Armed Services Committee (Dellums & Halterman, 2000).
By 1979, there were 51 identified social workers in elective legislative
office in the United States (Mahaffey, 1987). By 1991, NASW had located
113 elected officials who had been trained as social workers, and by 1993,
there were 165 (Weismiller & Rome, 1995). As reported by NASW, the 111th
Congress contained nine social workers: Senators Debbie Stabenow (D-MI)
and Barbara Mikulski (D-MD) and Representatives Susan Davis (D-CA),
Luis Gutierrez (D-IL), Barbara Lee (D-CA), Ciro Rodriguez (D-TX), Allyson
Schwartz (D-PA), Carol Shea-Porter (D-NH), and Edolphus Towns (D-NY)
(NASW, 2009).1 In addition to these, NASW was aware of 192 social workers
in state and local offices in 2008.
Only two studies of these elected social workers are known: a 1994
study of 41 social workers elected to state legislatures by the University
of Connecticut School of Social Work Nancy A. Humphreys Institute for
Political Social Work (NAHIPSW) (Humphreys, 1994) and a 1998 study by
Haynes and Mickelson (2006) that surveyed 84 social workers in federal,
state, and local office. Respondents to both surveys offered insight on topics
such as their backgrounds and education, the reasons they chose to run for
office, the challenges they met, and their recommendations for others who
might be interested in following their path.
THEORETICAL APPROACH AND CONCEPTUAL FRAMEWORK
Verba, Schlozman, and Brady (1995) developed the Civic Voluntarism Model
(CVM) to identify factors that contribute to individual political choices in
ways as specific as voting or as complicated as running for office. These
researchers hypothesized that interest in politics alone was not sufficient to
predict whether an individual would become involved in politics. Instead,
they focused on resources available to an individual interested in politics,
1
Two of these, Ciro Rodriguez and Carol Shea-Porter, were not re-elected to the 112th Congress
(NASW, 2010). Luis Gutierrez has worked as a social worker but did not graduate from a school of social
work (“Biography,” n.d.)
228
S. R. Lane and N. A. Humphreys
and considered whether the presence of those resources predicted an
individual’s political participation.
In this model, empirically tested nationally and longitudinally with several large samples of Americans through the Citizen Participation Study,
three main groupings of factors predict political activity. The first is psychological engagement, which includes political interest, political efficacy,
political information, family influences, and party identification. The second
is resources, including time, money, and civic or political skills. The third
is recruitment, defined as whether people are members of social networks
or groups that encourage members to get involved politically. Finally, the
research determined that a subset of political interest, interest in a specific
issue, often led to an individual’s decision to engage in a political activity
(Verba, Schlozman, & Brady, 1995).
Ritter (2008) used the CVM to compare social workers to the general public by examining predictors of political activity among licensed
social workers in 11 states across the country. Social workers differed
from the general public in four major ways. First, while the general
public was less likely to become politically active if they had less time
and money available to them, social workers in her sample participated
in political activity regardless of their available resources. Second, while
degree of partisanship predicted political activity in the general public,
it did not predict political activity among social workers. Ritter speculated that the lack of predictive ability for partisanship may have reflected
the relative homogeneity of social workers, who reported being primarily registered Democrats and fairly liberal. Third, in terms of recruitment,
Ritter found that while a majority of the social workers she surveyed
belonged to a church or another nonpolitical voluntary organization, very
few reported that they had been recruited into political activity through that
network.
Finally, within the context of civic/political skills, Ritter asked her sample of social workers an additional question: whether they felt that their
social work education had provided them with the skills they would need
to engage with the political system. Approximately half of the respondents
felt that they had learned these skills in their social work education, while
the other half disagreed or were unsure.
RESEARCH QUESTIONS
In order to expand the available knowledge about elected social workers, this research identified and contacted social workers who had run for
elected office in the United States. These individuals were surveyed about
their political beliefs, previous political participation, how they came to run
for office, potential future offices, their social work education, and their
Social Workers in Politics
229
demographics via an online and hard copy survey in 2008. The following
research questions were examined:
Research Question #1: Do Social Workers who run for office have
Access to Resources in the Areas of Time, Money, and Civic Skills?
The Civic Voluntarism Model found a statistically significant correlation
between civic engagement and income, with an increase in income correlating with an increase in civic participation. In addition, the examination of
several measures of individuals’ available time found that those who worked
fewer hours were more likely to be engaged in civic participation. In particular, those who were retired reported high levels of civic participation.
Finally, those who reported high levels of civic skills were also re likely to
be involved (Verba, Schlozman, & Brady, 1995). Based on those findings, this
research examined the income of this sample in contrast with other groups
of the general population and other groups of social workers. In addition,
the civic skills of the sample were examined through questions about their
preparation through social work education and their political activity prior
to runs for office.
Research Question #2: How were these Elected Social Workers
Recruited into Candidacy?
The original Civic Voluntarism Model shows a relationship between an
individual’s civic engagement and his or her affiliation with nonpolitical
associations and church attendance (Verba, Schlozman, & Brady, 1995). As
Ritter’s study of social workers did not find churches or nonpolitical voluntary associations to be significant political recruitment grounds for social
workers, this study examined recruitment through professional networks for
this sample.
Research Question #3: What Policy Issues were Emphasized by
Social Workers Running for or Elected to Office?
The Civic Voluntarism Model suggests that individuals tend to engage politically with issues that are either related to the individual’s experience or are
controversial (Verba, Schlozman, & Brady, 1995). Many other characteristics
also play into issue consideration. For example, Segal and Brzuzy (1995)
found statistically significant differences in voting among men and women
of the same party on four domestic issues: family leave, gun control, the
1993 savings and loan bailout, and gays in the military. This study asked
elected social workers to describe both the issues most important to them
and the issues about which they felt most knowledgeable.
230
S. R. Lane and N. A. Humphreys
METHOD
Participants
The objective of this research was to identify social workers who had run
for elected office at the federal, state, and local levels in the United States.
The criteria for inclusion included any run for an elective public office,
whether or not that candidacy had been successful. Eligible participants
held a BSW or MSW degree from a school of social work accredited by
the Council on Social Work Education. Since there is no definitive list of
elected social workers, the initial sampling frame was the list of social workers known to NASW’s national government relations staff, gathered by them
through contacts with NASW chapters throughout the country. This list was
expanded through purposive sampling. Internet searches combining social
work terms such as MSW, Master of Social Work, BSW, Bachelor of Social
Work, social work, and social worker with terms such as candidate and
names of specific offices (town council, mayor, city council, state legislator, etc.) were conducted. In addition, follow-up contacts were made with
NASW chapters and local social workers throughout the country, particularly
in areas where few elected social workers had been identified. Finally, snowball sampling was used as identified elected social workers were asked to
provide names of any other potential survey participants in their networks.
Individuals who were identified in any of these methods were reviewed to
ensure they met study criteria before inclusion. Potential participants were
contacted up to five times after the initial contact over a period of eleven
weeks via a mixed-mode survey (both e-mail and regular mail).
A total of 467 individuals who met the study criteria were identified.
Accurate contact information could be established for 416 persons, and
270 of those contacted responded to the survey, providing a 66% response
rate. While opinions vary on the acceptable response rate for a survey,
particularly given a population with an unknown overall size, a response
rate of more than 60% is considered a good response rate by Rubin and
Babbie (2008).
Design
The questionnaire used for this research was a modified version of the survey developed by Lawless and Fox (2005) to validate their theory of gender
and political ambition through the Citizen Political Ambition Study (CPAS).
Items about political attitudes and participation, attitudes about running
for office, demographics, family life, and upbringing were used from the
CPAS survey. Additional questions related to type of office, use of power,
social work education, and professional support were incorporated after
pilot testing by a group of elected officials and politically active social
Social Workers in Politics
231
workers. The final survey consisted of 32 questions, both open-ended and
closed-ended questions. The survey was available in hard copy and online
through SurveyMonkey.com.
Limitations
This response rate of 66% for this research allows for some confidence in
generalizing the results to all elected social workers. The number of social
workers who have run for elected office is not known. While there is no
way to know the total size of this population, the number identified here is
a significant increase over the 190 known before. In addition, conclusions
cannot be drawn about those who did not respond to the survey or those
who were identified but could not be contacted.
An additional limitation is the time frame between some of the events
respondents described and subjects’ completing the survey was significant. The social workers who responded to this study were most likely
to have received their social work degree before 1979 and the longestserving officeholder had been in office for 32 years. The time lag between
the events discussed in the survey and its completion could result in inaccuracy of retrospective recollections. These limitations suggest that while the
strong response rate provides a measure of confidence, care is needed in
generalization to all elected social workers based on this survey.
Consideration of Multi-Modal Differences
In order to assess any differences between those who responded in hard
copy and those who responded online, descriptive statistics of information
collected via the two methods were analyzed. In most respects, including
racial/ethnic background, likelihood of having children, age of children at
the time of first running, social work education, geographic area of the
country, size of town in which they current live, family encouragement,
recruitment, and previous political activity, the two sets of responses were
alike. There were statistically significant differences in only four areas. T-tests
revealed that those who completed the survey online were younger (the
mean age of online participants was 55; of hard copy participants, 62)
(t = 5.180, p < .001); more likely to be male (53% compared to 68%)
(t = 2.759, p < .01); and had higher personal income (average $75,001–
$100,000 compared to $50,001–$75,000) (t = −2.849, p < .01). Finally, they
differed in marital status: both groups were equally likely to be married,
but those who completed the survey online were more likely to identify
themselves as partnered but not married (8% compared to 4% of hard copy
respondents), more likely to identify as separated (3%, while none of the
hard copy respondents were separated), and less likely to be widowed (3%
232
S. R. Lane and N. A. Humphreys
compared to 13%) (χ 2 = 11.617, p < .05). Their overall similarities suggest
a lack of substantive differences between the two sets of results, and they
were therefore considered as one group for the remaining analysis.
Demographics
The elected social workers surveyed were 61% female and 39% male, with
one transgender participant. The most commonly reported racial/ethnic
background was Caucasian (74%), followed by Black, African American, or
African origin (13%) and Spanish, Hispanic, or Latino/a (6%). Respondents’
ages ranged from 25 to 87, with a mean of 58. Twenty-five percent were
under the age of 53 and 25% were older than 65. Only one respondent was
under the age of 30, with 13 (5%) between 30 and 39. Eighty-four percent
had children. Of those respondents with children, 49% had children over the
age of 18 when they first ran for office, while only 5% had children under
the age of one. Fifty-seven percent reported a personal income of $75,000 or
less, and the majority (77%) lived in a household with an income of $75,001
or more.
The sample was similar in many ways to social workers in general. They
were less likely to be female than other social workers (83% of NASW members are female and 81% of licensed social workers are female) and slightly
more likely to identify as non-white (14% of NASW members and licensed
social workers identify themselves as non-white). Members of this sample of
elected officials were older than the population of social workers in general,
as can be seen by comparing this sample’s average age of 58 with that of
NASW members (mean age of 45) and licensed social workers (averaging
45–54) (Arrington & Whitaker, 2008; Whitaker, Weismiller, & Clark, 2006).
Respondents to this study lived in 45 different states as well as the
District of Columbia, as shown in Table 1, at the time they completed the
survey. The regions with the largest numbers of respondents were New
England, the South Atlantic, the Eastern Midwest, and the Pacific West.
California, Connecticut, and New York had the largest numbers of respondents from individual states, while no elected social workers responded from
South Dakota, Tennessee, Montana, Nevada, or Virginia.
The most common social work degree held by respondents was an
MSW, held by 89% of those surveyed, followed by a BSW (29%). Twentythree percent of the total held both a BSW and an MSW, while 7% of the total
had an MSW and a PhD or DSW, and 2% held all three social work degrees.
Offices Held and Sought
Half of the social workers surveyed (51%) were holding office at the time
of the survey, while 39% had held office in the past and 10% had been
233
Social Workers in Politics
TABLE 1 Geographic Distribution of Respondents
Respondents
Census region
Included states
n
%
New England
Connecticut, Maine, Massachusetts, New
Hampshire, Rhode Island, Vermont
Delaware, District of Columbia, Florida,
Georgia, Maryland, North Carolina, South
Carolina, Virginia, West Virginia
Indiana, Illinois, Michigan, Ohio, Wisconsin
Alaska, California, Hawaii, Oregon,
Washington
Iowa, Kansas, Minnesota, Missouri, Nebraska,
North Dakota, South Dakota
New Jersey, New York, Pennsylvania
Arkansas, Louisiana, Oklahoma, Texas
Arizona, Colorado, Idaho, New Mexico,
Montana, Utah, Nevada, Wyoming
Alabama, Kentucky, Mississippi, Tennessee
53
21
40
16
35
35
14
14
29
11
23
18
14
9
7
6
7
3
South—Atlantic
Midwest—Eastern
West—Pacific
Midwest—Western
Middle Atlantic
South—Western
West—Mountain
South—Eastern
candidates for office but not (or not yet) elected. The largest group (69%)
had been officeholders or candidates at the local level, 29% were candidates at the state level, and 2% had been candidates at the federal level.
Unfortunately, no current federal officeholders participated. Both current
and former officeholders represented a wide range of experience: current
officeholders had been in their most recent office for an average of six
years, ranging from less than one year to 32 years. Former officeholders had
held their most recent office for an average of eight years, with a range
between two and 24 years.
Respondents were asked about all of their past runs for office, successful or unsuccessful, and their future offices they might consider. Social
workers in this survey were most likely to run for office on the city, town, or
county council. Over half (51%) had run for this office, 39% succeeding and
32% planning to run for this office in the future. They were also likely to seek
the positions of state legislator (38%) and school board (36%). Social workers surveyed were most likely to be elected when running for school board
(87% were successful) and judge (86% successful) and nearly as likely to be
elected to state legislatures (79%) and city/town/county councils (77%).
Candidates in this sample were least successful in their runs for U.S.
Senate, U.S. House of Representatives, and governor (of the 15 who had
run for these offices, none had yet been successful). While social workers
have certainly run for the federal legislature successfully, they were not
among these respondents. Table 2 shows all of the offices respondents had
sought, or had considered, and the offices to which they were successfully
elected. These numbers include multiple runs by the same individual, both
successful and unsuccessful for the same office. It should also be noted that
234
S. R. Lane and N. A. Humphreys
TABLE 2 Offices Past, Present, and Future
Ran for office
Office
City/Town/County Council
State Legislator
School Board
Mayor
US House of Representatives
Judge
Statewide Office (Attorney General,
Secretary of State, etc)
Governor
US Senate
Other
Elected to
office
Considering
running for
office
n
%
n
%
n
%
121
84
79
25
12
7
4
51
38
36
12
6
3
2
93
66
69
16
0
6
1
39
29
31
7
0
3

Purchase answer to see full
attachment