SAGE Research Methods Cases

Politics & International Relations

Submission for Consideration

Case Title

Inside Two Parties: Ethnographic Analysis of Partisan Differences in Political Candidate

Recruitment

Author Name(s)

Joshua J. Weikert, PhD

Author Affiliation & Country of Affiliation

Immaculata University, USA

Lead Author Email Address

Email: [email protected]

Discipline: D6 [please do not alter]

Sub-discipline within Politics & International Relations

American Government and Politics [SD-POLIR-2]

Academic Level of intended readership

Intermediate Undergraduate

Contributor Biographies

Joshua J. Weikert is Assistant Professor of Politics in the Department of Civic Engagement at

Immaculata University in Malvern, PA. His research focuses on political communication and

political behavior, primarily in the American political system. He is also an active political
consultant, advising campaigns on messaging, policy positions, communications, and

fundraising.

Published Articles

None (one under review)

Abstract
This case discusses the process and benefits of ethnographic research of candidate

recruitment by local political parties. Ethnography is especially useful in this research area,

as candidate recruitment is a closed, “inside” process that incorporates several veto-holding

stakeholders and a range of considerations that are challenging to disentangle using other

methods. Participant-observer data is richer and more substantively significant, especially in

this case, as the ethnographer was the subject of candidate recruitment by both major political

parties (Republicans and Democrats) in the same twelve-month period. This unusual “co-

partisan” recruitment effort, paired with a bottom-up methodological view provided by the

participant-generated data made for a unique research opportunity. The resulting data make

robust comparison and contrast possible, even given limits in record keeping and the limited

generalizability inherent to ethnographic analysis. The case also highlights the challenges of

confirming the findings in question, but recommendations for future research provide some

options to explore.

Learning Outcomes
By the end of this case, students should be able to:

 Identify and describe the circumstances in which ethnography is preferable as a

research approach.
  Analyze the benefits and limitations of participant-observer data in single and

comparative case studies.

 Formulate workable data recording approaches for ethnographic research.

 Identify the advantages and disadvantages of engaging in research from atypical

perspectives.

Case Study

Project Overview and Context

Candidate selection and recruitment has significant effect on political results in terms

of elections and public policy, which makes it interesting to note that there is far greater

emphasis within political science on the study of election campaigns, voter behavior, party

organization and behaviors, and legislative behavior. The raw material of political parties

and institutions – the proto-candidates which are brought into the process (or opt into it) – fall

somewhat through the cracks in the campaigns and elections literature. It is not so much that

candidate recruitment is not studied, but rather that it is swamped by the substantial (and

justifiable) number of studies focusing on the end product – people pay for more attention to

the laws that govern them and the personalities of political elites than on how those

personalities reached a position to create or influence those laws. While this enthusiasm

and/or production gap in the literature is understandable, it is nevertheless still something that

should be addressed. As Pippa Norris (1997) notes, the end products of electoral outcomes,

legislation, and public policy are, after all, informed by the individuals creating those

products. This project sought to add to the discussion of candidate recruitment by utilizing an

atypical perspective, both in terms of the theoretical viewpoint of the study of candidate

recruitment and in terms of the methodological approach employed. Structurally, it examined

the question from the prospective candidate’s perspective, looking from the bottom-up rather

than the top-down. Methodologically, it employed an ethnographic approach as I was, in

fact, the candidate being recruited. Making it even more interesting was that between the

2017 and 2018 election cycles, I defected from one party to the other, which meant that I was

able to document an “inside” look at the functioning of not one, but both major American

political parties in short order.

The ethnographic, participant-observer approach is especially valuable here. First, it

helps to shed light on what is usually a “closed” process: candidate recruitment is largely the

province of political parties, and while outside, independent, “wildcat” candidacies are not

uncommon, the party limits access to conveniences, information, and individuals that make

running easier for their “recruited” candidates. Second, the process is “closed” in another

way: the decision to run (or not) is made by both the recruiter and the recruited, which means

that the decision-making criteria and considerations are found inside the prospective

candidate’s head. While it’s true that interviews are an option to access that data, we can

never be quite certain what is influencing the respondents’ statements. At least, within my

own head, I can vouch for the validity of the responses (with due allowance for personal

biases, of course).

As noted in one recent paper by Fox and Lawless, “despite its pivotal role in the

candidate emergence process, we know little about political recruitment from the perspective

of the potential candidates who are well positioned to enter the electoral arena" (2010, 311).

The wants, needs, perspective, and behaviors of the party feature prominently in most

previous work in this area, which fails to recognize that candidate recruitment is, after all, a

bilateral negotiation. It is not simply the case that parties are selecting candidates who are

then drafted into political service: the prospective candidates, even after entering the

recruitment process, are themselves decision-makers with regard to their acceptance of the
recruitment effort prior to their candidacy. Not all recruited candidates advance through the

gate which is opened for them. One important decision-making factor (in this case, and

potentially in others) is the interaction the prospective candidate has with the recruiters and

their party apparatus. Investigation of this phenomenon is well-suited to a participant-

observer method. That I was recruited by both major parties in successive years adds to the

utility (and, arguably, interest) of the study. I received an inside look at an “inside” process

in both parties during one of the most politically interesting periods in recent history. Such a

moment deserves more than a quantitative, “outside” research approach: it is perfectly suited

to an ethnography.

Research Design

From the outset, I intended to use my prospective candidacy as an opportunity to

develop data suitable for some form of academic application. As a political scientist, this was

simply too good a chance to pass up! This was reinforced by the natural experiment that

resulted from my partisan defection and subsequent recruitment by the other party in the next

election cycle. Design decisions evolved logically from the circumstances. The general

model of candidate recruitment (see Norris 1997 for a detailed description) maps out a series

of narrowing gates, beginning with formal legal (who is eligible to run?), electoral (who can

win?), and party (who is in our party?) barriers to entry which limit the pool of prospective

candidates. Those that pass through this initial screening may then be subjected to a

recruitment process which further limits the “supply” of aspirants to office and which

incorporates the demands of partisan gatekeepers. The interaction between gatekeepers and

aspirants leads to an outcome in which either actor may play a veto role: in other words,

either the candidate or the party can say “no.”

 Since this process takes the practical form of a series of personal and group meetings

(we might conceivably think of them as “interviews”), I would seek to record as much data

on the content of these interviews as possible, which would yield a large contemporaneous

set of observations about the individuals, conversations, thoughts, and emotions in question.

This would not only serve me well as I decided whether or not to stand for election as a

candidate, but also preserve the data for scholarly analysis.

Observational studies make design decisions along four dimensions:

 Direct/indirect: is the phenomenon being observed in action, or are we

examining evidence/residue of action? For example, studying paths worn

through the woods is using indirect evidence of foot traffic by animals to

“observe” movement patterns – or, we could directly watch animals as they go

about their daily routines.

 Participant/non-participant: is the researcher an active participant in the

phenomenon being observed, or a passive observer? For example, one might

study restaurant culture by working as a cook (participant), or the behavior of

a school board by simply sitting in the audience during meetings (non-

participant).

 Overt/covert: do the research subjects know they are being observed as part of

a scholarly study?

 Structured/unstructured: Is the observation undertaken in a systematic way

(such as with the aid of a predetermined list of questions), or do the

interactions play out organically (with no predetermined guidance from the

observer)?
 I would be a covert participant-observer engaging in direct observation of an

unstructured encounter. Covert observations reduce “observer effects” and yield more

natural responses, but it should be noted that many (Warwick, 1982; Erikson, 1995) consider

this to be ethically dubious. In this case, I considered the risk to be worth the reward in the

context of yielding useful data, and I wasn’t concealing either my identity or the fact that I

was gathering information – only that the information might be used in academic research.

The ethical considerations were noted, and will be addressed in the next section.

Research Practicalities

From my perspective, there were three principal practical considerations in play:

 Which recordkeeping method was best suited to this project?

 How would subject selection/inclusion be handled? In other words, would I be

content including interactions only with those I came into contact with organically, or

would I affirmatively seek out additional subjects to expand my data set?

 What ethical considerations were intrinsic to this ethnography?

Recordkeeping is essential to any unstructured observation or interview, but it also

calls for balancing the value of the recordkeeping method and any reactions it may cause

during the course of observations. Recording video of every one of my interactions would

provide perfectly preserved data, but would also be likely to “freak out” the individuals and

groups I would be meeting with! Audio recording was considered, but would have required

the consent of the recorded party (laws vary by state). This would have created two

concerns:

 Knowing the conversation would be recorded would create observation effects –

maybe less so than video recording, but still…

  Recordings would almost certainly yield data which could result in the subject

being identified, which raises further ethical concerns.

Instead, I chose to take in-meeting notes on an old-fashioned paper pad. I would expand on

these notes immediately following my meetings/interviews, adding in relevant dialog and

observations, to the best of my memory. The net result gave me a reasonable degree of

sensitivity in terms of what could be recorded in real time mid-interview, while also

expanding on the records contemporaneously once the subject was no longer present.

Subjects were part of a purposive sample. Each recruiting cycle was led by a primary

recruiter, who made recommendations about which individuals and groups to contact, and I

followed those recommendations. In some cases, those subjects then referred me to others,

though these referrals were never solicited (for example, I never asked, “is there anyone else

you think I should speak with about running for office?”). In this way, the sample was not

biased by my own selection as it may have been if, for example, I contacted local political

actors based upon my own selection criteria and geographic considerations.

Privacy was the principal ethical concern, especially as the subjects were unaware of

the potential use of the information being provided. The practical considerations were dealt

with in conventional ways. Pseudonyms were used in all notes and in any documentation,

including any papers or presentations (Smith, 2017). Gender was preserved in these

pseudonyms, however, since gender-based considerations are relevant in the candidate

recruitment literature (see, among others, Bernstein, 1996; Carroll, 1994) and I wanted to be

able to consider issues of gender in my own work. Papers and presentations identified the

type of office being recruited for (“municipal board,” “state legislature”) but not the district

or municipality, which would further obscure the identities of those who provided the data

collected. All communications with subjects were either in-person, by phone, or via secure e-

mail. Barring an external data breach, there would be no feasible risk to the privacy of the
individuals in question. I endeavored to ensure that none of the conversations I engaged in

would ever, in any form or fashion, be traceable to specific individuals of either party.

The lack of disclosure about my research activity was less-easily addressed, but I

remain confident that my covert observation was on solid ethical ground. Researchers

consider ethical considerations of covert and overt observation within ethnography as a

matter of scholarly necessity, and a useful case described a researcher working as a bartender

at an establishment known to be popular with local homosexual residents (Lugosi, 2006).

The researcher wanted to engage in ethnographic research of establishments of this type, but

worried that engaging in a full description of the work and its purpose with every individual

with whom he interacted would create a sense of artificiality that would be

counterproductive. A “soft landing” or “incremental disclosure” approach was often used,

whereby the researcher would indicate through subtle but increasingly more-explicit

comments that the conversations in which he was engaged were to be used in a research

project (555-557). This was, generally, the approach I adopted.

Every subject and group knew already (or was advised when I introduced myself) that

I am a political scientist. I also sought to interject, as early as possible in the conversation,

the idea that the recruitment experience was a “fascinating opportunity to watch politics in

action” and that it would inform my research going forward. I also asked if it was alright that

I was taking notes. These three elements – disclosure of profession, professional interest, and

recording – plus the fact that all of these meetings were held in public spaces (an important

distinction noted in the literature; see Punch, 1994 for more) satisfied any ethical concerns.

To further enhance that position, I decided in advance that, if the subject inquired, I would

confirm that it was likely that our discussion would be used as data for some formal research

project, but that it would remain entirely anonymous: only one subject asked, and was, in
fact, eager to participate (and offered to provide comments for attribution, rather than

anonymously).

Method in Action

Broadly speaking, the method was effective. Each interview/interaction lasted

between thirty and ninety minutes, during which time I produced roughly one page of

handwritten notes per thirty-minute interval. In all but two cases I was able to add my

“augmenting” notes detailing specific reactions, quotes, and more immediately following the

interactions, and managed to preserve several high-value exchanges in their entirety.

However, there is no doubt that the absence of a compete transcript was at least a

moderate handicap. In meetings that sometimes ran to an hour or more in length, my

stenographic, bullet-pointed recording was effective but necessarily truncated and

incomplete. This was complicated further by the (apparent) fact that politicians and

politically-motivated individuals are a chatty, talkative lot! My subsequent reflections on

each meeting (whether minutes, hours, days, or weeks later) sometimes left me wanting a real

transcript of each exchange, especially as interactions with new individuals proceeded and

added more data – cross-referencing and pattern-seeking was made much more challenging

without a word-for-word account of previous meetings. By the second recruitment cycle I

was “primed” for the kinds of questions I had been left with in the first recruitment cycle, and

became a more-sophisticated note-taker. This “learning effect” creates a potential for bias in

the subsequent analysis, since the comparisons between the two parties’ efforts and my

responses to them are based on a better record of the process on one “side of the aisle.”

Despite this concern, though, I was left with a substantial data set which recorded a

useful and surprisingly large amount of information. Key comparisons and contrasts were

supported by the product of my data collection, even considering the “unknown unknowns”
that might exist. The lack of a complete transcript makes it impossible to know which dogs

did not bark – but their silence did not seem to be a substantial hindrance as I wrote up the

research report based on my observations.

Practical Lessons Learned

Whatever its limitations – and they are not insurmountable, if I should find myself

presented with a similar set of circumstances in the future – an ethnographic approach as a

participant-observer was undoubtedly the best choice to study political actors and their

behaviors.

First, being the subject of their attention and behavior yielded information about the

candidate recruitment process that would ordinarily remain obscured. The individuals doing

the recruiting, screening, and/or gatekeeping showed no signs of hesitation or reserve, which

is a commonly-sought benefit of participant-observer observations. Recruitment is a form of

“courtship,” and so I am confident that I was not always given the complete and honest

accounting of every participant’s motivations, but I am likewise confident that I experienced

an honest representation of the recruitment process. Since the goal was to see “inside” the

parties, this was an important conclusion.

Second, experiencing candidate recruitment as a subject/participant allowed for me to

experience the emotional and cognitive “ride” that a recruited candidate does, since I was

building to a personal decision point on whether to run for election. I noted immediately the

differences in personal thinking and evaluation that my participant-observer perspective

created, which were all the more apparent when I was shifting gears post-meeting to analyze

as a scholar what I had just heard, said, and thought. The distance between those two

“selves” – the prospective candidate and the academic observer – was startling, and

instructive.
 Last, I found the richness and depth of detail developed by this method to be ideally

suited to the subject of the research (candidate recruitment), which is an area about which we

know relatively little. The few “hypothetical” expectations I had going in were rapidly

dispelled; for example, there is a theory that local politics is “insulated” from national

political dynamics (Williams and Adrian, 1959), which took a real beating in my experience,

as national partisan politics was front and center. At the same time, new areas of inquiry

were uncovered. As a theory-generating exercise, ethnographic participant-observation

proved to be a robust and valuable approach.

The elite-led, “cartel” process common to political party activity was a great fit for

this observation method. I believe that it can be employed in like circumstances – decision-

making in a corporate setting, college search dynamics from the perspective of the student,

any number of microeconomics applications related to purchasing or selling decisions, the

development of marketing or messaging plans in commercial or political communication – as

well as within any group, culture, or sub-culture that exhibits higher-than-usual levels of

suspicion or a desire for privacy. The developed trust and “in-group” identity of the

participant-observer becomes a potent tool for the researcher and improves the quality of the

data collected.

One significant limitation, common to ethnographic research, is in confirming the

general nature of the findings. Interviewing other recruited candidates (especially those who,

like me, ultimately decided against running for office) may be the only viable option to

broaden the sample size to a sufficient degree that findings can attain a degree of external

validity. Without that additional data, it is simply impossible to know if these two

observations are truly representative, or if the process varies substantially based on timing,

geography, candidate demographics, or other variables. Ethnography is sometimes, as it is

here, simply setting the stage for other qualitative and/or quantitative research.
Conclusions

At the end of the day, in neither case did I decide to pursue a candidacy for elected

office. Far from being a waste of time, though, my ethnographic research provided an

invaluable counterpoint to the high-altitude quantitative research that I primarily engage in.

It is one thing to aggregate thousands of observations of multiple variables and analyze the

output of a statistical software program; it is quite another to interact with the people and

personalities that are making the decisions in the real world, and with obvious stakes for

one’s own career and life. I can now, confidently, state that I know what candidates for

public office are thinking about when presented with that potentially life-altering opportunity.

I can further report on the subtle changes in thinking and behavior that accompanied that

choice, as a new level of consideration of something as simple as walking through a grocery

store in “my” district changes into a situation in which I would have to be ready to greet

voters and constituents. Living a phenomenon is far different than studying it. Doing both is

a wonderful opportunity for a researcher.

I was somewhat unprepared for the emotional costs. I experienced like, dislike,

respect, rejection, approval and disapproval, and more in the course of this experience. I was

patently unable to fully divorce my “academic” observer self from my “pragmatic”

participant self. Field work, generally – and ethnography, particularly – call for a different

kind of patience and attention than many social scientists tend to experience in the modern

era, but it has been an invaluable one to me. I doubt you’ll ever see my name on a ballot, but

I absolutely anticipate delving into ethnographic participant-observation again in my

academic career.
Exercises and Discussion Questions
1. Why is politics a particularly useful research area for participant-observer research,

especially the study of political elites and candidates?

2. Should ethnographers engage in covert observation? When and why might it be

ethically acceptable? Unacceptable?

3. Write a reflection, based on your own “public” persona (social media presence,

appearance, public interactions), about how you would be perceived by voters if you

announced today that you were running for office in the next election.

4. In this case, the author chose not to run for office. How might that limit the

usefulness of the findings compared to, for example, interviews of candidates who

decided to run?

5. What advantages do participant-observations have over nonparticipant-observations?

6. Can ethnographers’ observations be trusted, given that they are often members of the

community being studied? What biases might they be prone to?

Further Readings

Lugosi, P., (2006). Between Overt and Covert Research: Concealment and Disclosure in an

Ethnographic Study of Commercial Hospitality. Qualitative Inquiry, 12(3), 541-561.

Punch, M., (1986). The Politics and Ethics of Fieldwork. Newbury Park, CA: Sage.

References

Bernstein, R., (1986). ‘Why Are There So Few Women in the House? Western Political
 Quarterly, (39)1, 155–164.

Carroll, S., (1994). Women as Candidates in American Politics. Bloomington, IN:

Indiana University Press.

Erikson, K., (1995). Commentary. American Sociologist, 26(2), 4-11.

Lugosi, P., (2006). Between Overt and Covert Research: Concealment and Disclosure in an

Ethnographic Study of Commercial Hospitality. Qualitative Inquiry, 12(3), 541-561.

Norris, P., (1997). Passages to Power: Legislative Recruitment in Advanced Democracies.

Cambridge: Cambridge University Press.

Punch, M., (1994). Politics and Ethics in Qualitative Research. In N. K. Denzin & Y. S.

Lincoln (Eds.), Handbook of Qualitative Research, 83-97. Thousand Oaks, CA: Sage.

Smith, R., (2017). “Don't Let the Illegals Vote!”: The Myths of Illegal Latino Voters and

Voter Fraud in Contested Local Immigrant Integration. RSF: The Russell Sage

Foundation Journal of the Social Sciences, 3(4), 148-175.

Warwick, D. P., (1982). Tearoom Trade: Means and End in Social Research. In M. Bulmer

(Ed.), Social Research Ethics: An Examination of the Merits of Covert Participant

Observation, 38-58. London: Macmillan.

Williams, O. and Adrian, C., (1959). The Insulation of Local Politics under the

Nonpartisan Ballot. American Political Science Review, 53(4), 1052-1063.