The jury system as it is understood in America appears to me to be as direct and as extreme a consequence of the sovereignty of the people as universal suffrage. (Tocqueville [1862] 1945:294)
Jury service has long been described as a political participation right that is “akin to voting,” in part because of the importance attached to it by the Founders (Reference AmarAmar 1995). Three of the first 10 amendments to the U.S. Constitution refer to citizens' rights to have their cases reviewed by a panel of citizens. However, no summons arrives to tell people that they must vote; no person is required to take time away from work or family to write to their elected representative. Jury service thus resembles no other democratic involvement because it is compulsory.Footnote 1 Also, unlike other forms of political participation, this one must be distributed: only some who are summoned and who appear at a courthouse will eventually sit on an actual jury.
Those who appear for service at a courthouse may be excused in one of two ways. First, judges may grant a “challenge for cause,” which means the jurors have a demonstrable conflict with the case (e.g., hardship, a financial interest in the outcome, a stated unwillingness to be fair). In addition, people can be excused through “peremptory” challenges, which permit attorneys to excuse a limited number of people at their discretion without providing a reason. Jury selection at this courthouse stage has received some empirical attention; however, the focus is typically on how challenges affect litigants' interests in an unbiased jury and the representativeness of juries (see, e.g., Reference Finkelstein and LevinFinkelstein & Levin 1997; Reference Johnson and HaneyJohnson & Haney 1994; Reference JonesJones 1987; Reference KerrKerr et al. 1991; Reference Narby and CutlerNarby & Cutler 1994; Reference Nietzel and DillehayNietzel & Dillehay 1982; Reference SeltzerSeltzer et al. 1991; Reference Zeisel and DiamondZeisel & Diamond 1978). Although these are important questions, we know little of how individual jurors themselves view the selection process, which means we do not know how this conscripted group regards their experience of jury participation. Reference Diamond and LitanDiamond (1993) reviewed the few, mostly unpublished, studies that have asked jurors for their perceptions of jury service and found “a generally accepting” attitude (1993:289). However, such work has focused almost exclusively on people selected for juries, which is a severe limitation if one is concerned about the effects of exclusion from political participation, especially when exclusion stems from the largely unregulated discretion of attorneys' peremptory decisions (Reference HoffmanHoffman 1997; Reference MarderMarder 1995).
The absence of data on jurors' reactions to jury selection is a significant omission given that Supreme Court rulings regarding permissible uses of peremptory challenges are framed in terms of jurors' (rather than litigants') interests. In its Batson v. Kentucky (1986) ruling, the Supreme Court forbade prosecutors from using peremptories to dismiss people on the basis of race in a single trial; previously, the Court had required that defendants prove systematic exclusion across multiple trials (Swain v. Alabama 1965). In deciding Batson, the Court considered the black defendant's equal protection rights, along with the rights of the (all-black) excluded jurors. However, the Court later expanded Batson to disallow the use of race in decisions by criminal defense attorneys (Georgia v. McCollum 1992), decisions involving white jurors rather than minorities (Powers v. Ohio 1991), and decisions in civil trials (Edmonson v. Leesville Concrete Co. 1991). In addition, gender may not be the sole basis for a peremptory (J.E.B. v. Alabama, ex rel. T.B. 1994). To rein in attorneys' behavior in all these areas, the Court has based its reasoning upon jurors' Fourteenth Amendment rights, i.e., the discriminatory use of the peremptory harms the juror, rather than necessarily undermining a litigant's right to a fair trial.
In short, the Supreme Court has identified specific instances in which jury selection practices can be “unfair” to prospective jurors. In this article, I consider the extent to which attorney decisions predict jurors' attitudes toward jury selection. However, besides the possibility of racial or gender bias in decisionmaking, I examine other possible sources of unfairness in the process of selecting jurors. To do this, I first note the ways in which jury selection is a distinctive distribution system, meting out a duty rather than necessarily a desirable end. I argue that trial court actors (judges, attorneys) seek to frame the procedure as a simple search for fair and impartial jurors—in other words, as a “cognitive conflict,” or an investigation into the true state of the world (Reference Thibaut and WalkerThibaut & Walker 1978). In such procedures, participants (jurors) are asked to align their own interests with those of the court and parties, and to relinquish both decisionmaking and process control. Nevertheless, despite such rhetoric, jurors' interests can and do conflict with decision makers, which introduces elements of Thibaut and Walker's other class of disputes, a zero-sum “conflict of interest.” I identify two potential conflicts as likely bases for evaluating jury selection: jurors' desired outcomes and their privacy interests. To test this perspective, I present data from 194 interviews of excused and selected jurors from 13 felony criminal cases.
The Distribution of a Duty
Social scientific research has amply demonstrated that perceived injustice in distribution systems leads to declines in perceptions of institutional legitimacy, respect for authority, and even law-abidingness (for reviews, see Reference Hegtvedt, Cook, Sanders and HamiltonHegtvedt & Cook 2001; Reference Lind and TylerLind & Tyler 1988; Reference MillerMiller 2001; Reference Tyler, Lind and ZannaTyler & Lind 1992, Reference Tyler, Lind, Sanders and Hamilton2001; Reference TylerTyler 1990). A sense of injustice can stem from perceptions of unfair decisions (distributive justice), unfair procedures (procedural justice), or, typically, the combination of the two (Reference Brockner and WiesenfeldBrockner & Wiesenfeld 1996). An uncomplicated examination of jury selection might simply seek to explore the extent to which traditional indicators of distributive and/or procedural justice predict perceptions of the fairness of distribution to a jury (e.g., how fair was the result? How did the result compare with what was expected? How fair was the procedure? How fairly were people treated?). Such a straightforward analysis, however, presumes that jury selection procedures are similar to other distribution systems. There are, however, several reasons why jury selection is distinguishable from distribution systems typically studied in the justice literature. (The discussion that follows describes several features of jury selection but presumes some knowledge of its basic practices. For those unfamiliar with these basics, the Appendix details one court's procedural structure, questions, and decisionmaking system.)
According to Thibaut and Walker's “Theory of Procedure” (1978), procedures are applied to two classes of “conflicts.” In most procedural justice studies, authorities (decision makers) distribute outcomes in settings broadly described as “conflicts of interest” (Reference Thibaut and WalkerThibaut & Walker 1978). In such situations, parties appearing before the authorities are considered to be in conflict with one another because two or more people seek that which the authority is distributing—be it a job (Reference BrocknerBrockner et al. 1990; Reference GillilandGilliland 1994), a monetary reward (e.g., Reference Van den Bos and Van Prooijenvan den Bos & van Prooijen 2001), or some other desirable but scarce resource (Reference Lind and TylerLind & Tyler 1988). Conflicts of interest are also present when the distribution involves something a person seeks to avoid: for example, a long prison sentence (e.g., Reference CasperCasper et al. 1988) or some lesser form of punishment, such as a traffic fine (Reference TylerTyler 1990). As nearly all people would prefer the lightest sanction possible, or none at all, their preferences are at odds with the society's interest in sanctioning misdeeds properly.
For several reasons, jury selection is not easily characterized as a conflict of interest. First, no matter jurors' personal preferences about service, the court asks people not to view jury selection in zero-sum, conflict of interest terms. Rather, the court frames jury selection as a simple search for qualified jurors, resembling what Thibaut and Walker termed a “cognitive conflict,” or an investigation into “the most accurate view of reality” (Reference Thibaut and WalkerThibaut & Walker 1978:541). The paradigmatic example of a cognitive conflict is a scientific dispute, such as whether a drug is safe. In such instances, interests are not viewed as conflicting because all parties should be equally interested in discovering the correct result (e.g., everyone benefits from drug safety). Just as a scientific panel might minimize any one person's preference for either the investigation's outcome, or whether a result is “fair” to any one person, the court minimizes the relevance of jurors' preferences regarding who should and should not be on the jury.
In the trials I observed, this framing occurred in several ways. Orientation and the oratory from judges reminded prospective jurors of their shared interests as community members. All should want good decisions about who should serve or not. Often, the judge's opening comments to the panel assembled included reminders about the importance of a working jury system. Throughout questioning, outright appeals to a sense of duty were commonplace. In one case, an elementary school teacher asked to be dismissed from a two-week trial because, she said, her schoolchildren needed her. The judge refused, lectured her briefly on jury service, and asked her to think instead about how much her students would learn about civic participation through her service. A peremptory challenge eventually dismissed her; however, the rhetoric seemed to have the desired effect. She stated during a later interview with the author: “After the judge said those things, he showed me I was being selfish.”
Attorneys also communicate the irrelevance of jurors' own preferences, as occurred in one exchange between a juror (who was ultimately selected) and the assistant district attorney (ADA) from an armed robbery case:
Juror: I knew someone involved in a similar crime. He was accused of it.
ADA: OK. Can you give both sides a fair and impartial trial?
Juror: It won't impede my abilities, no. But I would prefer not to.
ADA: Well, yes, but can you be fair?
Juror: Yes, if I have to be on this jury.
An additional way that the design of jury selection resembles a cognitive conflict is its explicit lack of “process control” (Reference Thibaut and WalkerThibaut & Walker 1975), what has come to be termed “voice” (Reference FolgerFolger 1977). Research consistently finds that in conflicts of interest, people express more support for outcomes, for authorities, and for institutions when they can contribute their perspective and feel as if they are heard (Reference Van den Bos and Van Prooijenvan den Bos & van Prooijen 2001). Recent work in political science likewise finds that belief about how much government should listen to individual citizens is a good predictor of public mood and governmental support (Reference Hibbing and Theiss-MorseHibbing & Theiss-Morse 2001). However, as Thibaut and Walker point out, when the goal of a procedure is to uncover the true state of affairs, minimal voice from participants is optimal:
Such a … strategy increases the likelihood of obtaining the relevant information, reduces the strain of assimilating and tracking information, and minimizes the risk of failing to reach the correct solution within a limited number of attempts. (1978:548)
Jury selection procedures quite explicitly minimize prospective jurors' opportunities to contribute to the decisionmaking process. Citizens do not and cannot know much about the case facts; indeed, people are supposed to be entirely ignorant of the case. Thus, they have few ways to gauge and contribute their own views of whether they are right for a case. The court never allows people to argue for why they would make especially good jurors and should be selected—e.g., no prospective juror ever says, “I'm bright, a hard worker, and I'm a team player!” Likewise, jurors cannot simply say, “Here's why I don't want to be on this jury…. May I try another?” Jurors can plead hardship or claim bias, but the court does not allow them to contribute their ideas about what constitutes a “good juror.”
Jurors may certainly question their own fairness; however, the court defines bias. Nearly every inquiry to prospective jurors eventually comes down to the singular issue of whether any opinion or characteristic (e.g., knowledge of the parties, prior information about the case, prior experiences) renders the person unable to hear the case with an open mind and to follow instructions (see also Reference DiamondDiamond et al. 1997 on this same point in federal trials). For jurors to directly and overtly participate in being eliminated from the jury, individuals must do more than merely protest or suggest that they have some “concerns”; they must publicly declare themselves unfair, close-minded, unwilling to follow the law, and/or insensitive to duty. For example, in the armed robbery case in my sample, a juror expressed reservations about “sending such a young man to jail.” His views were explored, but the ADA wound up summarizing:
If you think you can't fulfill your duty, this is your opportunity to tell us. But if you would be troubled, but you could still follow the law, this is your opportunity to tell us that, too. (emphasis added)
At another point, a different juror responded to a question about his feelings toward the state's having only one witness, saying, “I might have trouble if it's one person's word against another.” The ADA responded quickly:
Now you don't know the evidence yet. The question has to be answered independently of the evidence. You must consider all evidence carefully. We would like it if we could present 50 witnesses, but we can't. If your position is that you would listen to both sides, you could be fair. But if your position is, no matter what, you won't listen when there's only one witness, then that's not fair to the state.
In both of the above instances, with the issue framed in stark, simplistic terms of being fair or not, the jurors stated publicly, “I can follow the law,” or “I can be fair.” When pushed, most people promised to be fair, a common reaction in routine jury trials. As Reference GarfinkelGarfinkel wrote of his interviews with civil jurors:
In various ways the judge and others in the court invite the juror to see himself as a person who can act in accordance with the official line. Jurors [are] typically avid to accept this invitation. (1967:111)
Further evidence for this conformity stems from the fact that, in this sample of trials, the majority of the challenges for cause represented hardship issues or jurors' travel or work conflicts rather than admissions of bias. Likewise, no matter the circumstances, no juror who promised to be fair was excused for cause. The judge instead appeared to take jurors' own bias assessment at face value, and these people's fate lay in how the attorneys distributed their limited number of peremptory challenges.
Potential Conflicts of Interest in Jury Selection
Thus far, I have suggested several ways in which the court seeks to frame jury selection as akin to a cognitive conflict, or a simple search for whether prospective jurors are, in fact, capable of sitting fairly on the jury. The court makes frequent appeals to a sense of common interests and duty and explicitly tells people that their own preferences do not matter, and the procedure affords people little “voice” in decisionmaking.
Of course, just because the court frames jury selection as a cognitive conflict does not mean it will be perceived that way. People's own perceptions of fairness may not disinterestedly center on getting the “best” jury, and people may not feel that the court's and parties' concerns are their concerns. Instead, they may rebuff the court's requests regarding their time, personal preferences, and willingness to “do their duty.” In this situation, prospective jurors may well see their own interests as being in direct conflict with what the court wants from them.
Why should such a conflict between authorities' and jurors' interests matter? There is good reason to believe that it is precisely the areas of potential conflict between juror and authority that are most likely to lead to perceptions of fairness or unfairness. In experimental studies of bargaining situations between equal-status persons, sociologist Linda Molm and colleagues have shown that, compared to cooperative situations, zero-sum conflicts create opportunities for people to perceive unfairness (Reference MolmMolm et al. 2003). In their work, even though bargainers had voice in a given situation (i.e., they were engaged in direct negotiations with a partner), the fact that their outcomes were at odds with their partner led them to make more negative attributions about the motives and behaviors of the other person. That is, the conflict of interest tended to bias and lower participants' perceptions of fairness. Molm and colleagues noted that prior justice research has not focused on this issue because it has “primarily been concerned with procedures enacted by third parties whose own interests are not necessarily at stake” (2003:381). In contrast, as a distribution of a duty, jury selection does present an instance in which people's individual goals may be in conflict, or at least tension, with those of authorities and decision makers. One source of conflict has to do with the ways in which people's outcomes may not be desirable to them; the second has to do with the ways in which the information-gathering process is at odds with jurors' interests.
Conflicts With Decision Outcomes
Despite what the court asks of them, jurors may care less about how the jury as a whole is configured, or whether the litigants find the group to be satisfactory, and more about whether their being excused or selected is consonant with their preferences. Individuals may have a variety of reasons for wanting to be on or off a given jury. On the one hand, some scholars (e.g., Reference HoffmanHoffman 1997; Reference MarderMarder 1995) have highlighted the drawbacks to being excused from service, which include the stigma of being publicly deemed an “unfit” juror and, in particular, the lost opportunity to participate in one of the hallmarks of a democracy. People who come to the courthouse only to be excused may feel that their efforts were, in some ways, “for nothing.” On the other hand, there are clear downsides to being selected for a jury, including the sacrifice of time and the minimal compensation (about $12 per day in some locations), which can be especially burdensome for those whose employers do not cover their salaries or who are self-employed or working on commission (Reference BoatrightBoatright 1999). Caretakers of children, the sick, or the elderly may also find service difficult because the court does not assist them with alternative care arrangements. In addition, for any person, jury service can be stressful (Reference ShumanShuman et al. 1994), involving possible anxiety over the responsibility of deciding another's fate or even fear of retaliation for decisions—anxiety that is undoubtedly furthered via the jury-tampering plotlines in popular movies such as The Juror or The Runaway Jury, through the occasional television episode (e.g., Law & Order, The Sopranos), or through real-life anecdotes such as the Tyco trial, in which a holdout juror received harassing phone calls and letters.
Because the meaning of jury selection outcomes depends upon a person's ability to accommodate jury service to their schedule and activities, I would expect both a great deal of variability in people's desire to be on a jury and that people evaluate jury selection with an eye toward how being selected (or not) affects them. This is a starkly rational choice model of jury selection, and although such models have been largely discredited in the procedural justice literature (e.g., Reference LindLind et al. 1990; Reference MillerMiller 2001), interest in obtaining an outcome seems important to assess in detail when a duty, rather than an inherently beneficial or burdensome outcome, is being distributed.
Prospective jurors' instrumental interests in serving are considered in several ways in this study. First, I asked people to report on which result they had wanted. An instrumental model would predict that those who wanted to be on the jury and were selected, as well as those who did not but were excused, will have a more positive impression of the decision and of their experience—at least compared to those whose outcomes were not consonant with preferences (i.e., selected jurors with little interest in serving and excused jurors who wanted to be chosen). Second, justice researchers suggest that outcomes should be considered both in terms of decision satisfaction and decision fairness, because the two ratings can elicit a different pattern of results. Reference Van den BosVan den Bos and colleagues (1998), for instance, found that people can question the fairness of an outcome but nevertheless feel quite satisfied with it. If the instrumental model is correct, however, and people focus on their outcomes, decisions viewed as more fair and more satisfactory should positively predict reactions to jury selection.
Conflicts With the Decision Process: Privacy Protection
Procedural justice has come to mean many things. Early research focused largely on how a procedure is designed (i.e., whether it allowed for voice or not; Reference Thibaut and WalkerThibaut & Walker 1975; Reference FolgerFolger 1977). Later work additionally emphasized how an authority behaves, in particular the extent to which the authority is polite, honest, and even-handed (see especially Reference TylerTyler 1994; Reference Tyler, Lind and ZannaTyler & Lind 1992). As I have attempted to show, it would be difficult to assess the procedural fairness of jury selection in terms of the amount of voice it affords prospective jurors, for the structure affords them very little. Further, although jurors undoubtedly expect to be treated with politeness and respect (Reference MillerMiller 2001), there is no reason to expect an inherent tension between jurors and decision makers over this issue. Given that attorneys are interacting with and selecting those who may ultimately sit in judgment of the case, it is entirely in attorneys' self-interest to be as kind and polite as possible to prospective jurors. This in no way suggests that those who perceive rudeness or disrespect will not be affected by it; indeed, the current study controls for ratings of the attorneys on several procedural justice items, as well as for general ratings of how fairly people were treated. The issue is whether such measures reflect the aspects of the procedure that are either likely to vary or to which jurors are likely to attend when evaluating their experience with the courts.
If areas of conflict between juror and decision maker are especially likely to affect overall evaluations of jury selection, then I would expect conflicts over the type and amount of information jurors must share to predict impressions of their experience. As Reference Thibaut and WalkerThibaut and Walker (1978) noted, in cognitive conflicts decision makers want as much information as possible in order to make the “right” decision. From the lawyers' point of view, more information assists them in eliminating patently unfair jurors and, more strategically, in retaining people who will likely favor their side.Footnote 2 However, jurors have interests in restricting how much they must reveal about themselves. Even if the court must thoroughly survey jurors to find a fair jury, citizens naturally prefer this to be done with a minimal dredging of their private lives. Thus, in a situation of a duty, rather than valuing the extent to which authorities allow for the expression of information (e.g., through voice), jurors are likely to attend to how authorities manage and limit the exchange of information via attention to privacy protection, control of which is squarely in the hands of the attorneys and judges.Footnote 3
Tension over the appropriate amount of information jurors must share during voir dire implicates two aspects of privacy protection that should concern jurors and hence predict people's overall satisfaction with jury selection: intrusiveness and relevance.Footnote 4 First, jury selection can and does pose inquiries into matters that some may see as uncomfortable and overly intrusive (Reference RoseRose 2001). People must disclose in public experiences they may have kept even from close friends (e.g., that they committed a crime or a have been a crime victim; see Reference GloverGlover 1982; Press-Enterprise Co. v. Superior Court of California 1984). In rare instances (e.g., an anti-gay hate crime), trials could conceivably be concerned with the sexual orientations of jurors or their relatives (Reference LyndLynd 1998).
Second, even if disclosures are not particularly intimate or emotionally arousing, voir dire may also invade privacy if authorities ask “too much” by posing questions regarded as unnecessary for or unrelated to determining juror impartiality (Brandborg v. Lucas 1995; Reference HannafordHannaford 2001; Reference HoffmanHoffman 1997). In Brandborg, a juror refused to answer items on a questionnaire about her religious and political preferences, income, and television/reading habits. She did not argue potential embarrassment over the inquiries, but rather that they were simply “irrelevant” to her abilities to be impartial (Brandborg v. Lucas 1995:354). Both relevance and intrusiveness were measured in the present study.
The Present Study
To sum up, I argue that jury selection is exhaustively described to jurors as an information-oriented search for the true state of the world (juror impartiality), and that the procedure affords participants little control and voice. As with most cognitive conflicts, decision makers would prefer that all jurors be equally disinterested in whether they are selected or not, and authorities expect jurors to be open about their backgrounds, views, and life experiences. Nevertheless, despite the court's expectation, I suggest that jurors' interests can and will be in tension with those of the authorities. Jurors may want a different outcome than the decision makers desire, and they likely expect authorities to attend to juror privacy—considered in terms of both intrusiveness and relevance—even when decision makers would like to gather as much information as possible. As sites of tension between juror and authority, these domains are most likely to predict jurors' ratings of their jury selection experience.
To test these expectations, I draw on interviews of people who had undergone jury selection in the 13 trials. I observed the entirety of all voir dires and then contacted those who had been either selected for or excused from the juries. I here focus on predicting two global assessments: one invited respondents to look back and rate the overall jury selection experience, the other asked them to look forward and consider their willingness to return to jury service at another time. Apart from views of outcomes, privacy, and variables already mentioned, these data and analyses include several other controls and moderators to estimate effects.
All analyses include the juror's race as a predictor, not only because of the jury system's history of antipathy toward minority group members (see Reference FukaraiFukarai et al. 1993; Reference HoffmanHoffman 1997), but also because analyses of even this small sample of cases found that race was associated with peremptory decisionmaking, albeit in a complicated manner (Reference RoseRose 1999). Race did not predict selection, per se. However, a clear adversarial use of race contributed to this overall null finding: blacks were disproportionately likely to be dismissed by the state, and the reverse was true of whites (disproportionately dismissed by the defense). Although the sample of trials is small, the results are consistent with patterns observed in larger data sets from other areas of the country (Reference BaldusBaldus et al. 2001).Footnote 5
Next, the experience of excused and selected jurors fundamentally differs, not only because the latter must ultimately stay and take responsibility for deciding the case, but also because by doing so selected jurors gain additional information. Because the additional information and decision outcome are confounded, my goal is not to make causal claims about the effect of selection on perceptions of the court. Instead, results for selection status are necessarily descriptive. That said, potential differences in reactions to jury selection must be examined empirically, and the current study does this via tests of selection status as a moderator of the other variables. A significant interaction would indicate that the two groups require different sets of predictors to understand their reactions to their experiences. The two areas of conflict I have identified seem especially prone to the moderating effects of selection status. First, although excused jurors may have preferences regarding their outcome that can be undermined, decisions about selected jurors affect more than just their desires; they affect the use of their time and energy. All things being equal, views of outcomes are likely to be more strongly associated with the selected group's ratings because they must actually serve. Second, the additional information selected jurors acquire through service may alter privacy perceptions. Those serving may come to understand why some questions were posed; the excused group, in contrast, is more likely to persist in believing that some issues were irrelevant or that privacy violations occurred “for no good reason.” This would tend to make privacy a more powerful predictor for those excused than for those selected.
Last, in theory some types of trials are likely to either exacerbate or limit the effect of variables. For instance, perhaps case seriousness or trial length make outcomes particularly salient to people and therefore alter the weakness or strength of their relationship to the dependent variables. In statistical terms, a coefficient for a given predictor may not be “fixed” (the same value applies to all groups, in this case, trials) but is instead “random” (the coefficient's value varies across trials; see Reference Bryk and RaudenbushBryk & Raudenbush 1992; Reference SingerSinger 1998). The current sample size (n=13 cases) precludes a reliable analysis of this type (Singer, personal communication, October 2000). Nevertheless, I am able to provide an empirical assessment of the amount of trial-level variability in the dependent variables. The mixed models reported below allow the intercept value to vary across cases, even while the regression coefficients for the predictors remain fixed. This analysis produces an “intra-class correlation” for each dependent variable, reflecting the amount of variability in ratings attributable to trial effects (rather than to individual juror differences). This value is somewhat analogous to an upper bound for the “R-square” for trial-level effects (see Reference Snijders and BoskerSnijders & Bosker 1999 for a more detailed discussion). To the extent that an intra-class correlation is low, the dependent variable is only weakly associated with features of a given trial.
Method
The County and Its Court
The people interviewed for this study appeared for jury service in a North Carolina County courthouse (county population approximately 200,000). Home to several universities, the county has traditionally been a mostly biracial area (mostly white but about 35% black and 2% Hispanic during the 1990s). The area was and is known for having a strong black middle class, something reflected in the racial composition of courthouse personnel. In nearly every case I observed, African Americans served as either a judge, an ADA, a defense attorney, and/or court clerk. The defendants in this sample of cases, however, were starkly homogeneous: all, save one, were black; all but two were male. The cases involved serious felony charges, including four homicides (three charging second-degree murder, one, involuntary manslaughter); a felonious assault that included sexual offenses; two armed robberies; two felony drug offenses; and two cases each of breaking and entering/possession of stolen goods, and obtaining property by false pretenses. In North Carolina felony criminal cases, there are six peremptory challenges for each side, plus one each for an alternate. This is a modal number of challenges in state criminal courts (Reference MunstermanMunsterman et al. 1997:233). (See Appendix for more on the jury selection procedures.)
Recruitment Procedure
Across trials, 348 people underwent voir dire, of whom 181 were excused, most (81%) via the peremptory; 89% of this group (n=309) were eligible for the follow-up interview.Footnote 6 The court clerk provided me with jurors' names and addresses. I then contacted people with listed phone numbers (69% of all eligible), usually during evening or weekend hours. Attempts typically continued until I could reach the person, with an outside limit of six weeks post–voir dire. Second, because privacy was being examined, I worked to include those with unlisted numbers (31% of the eligible sample, n=96). Initially I sent a letter asking for their telephone contact information, which added 21 people (22% of those unlisted). I further conducted at least one “house call” to the address listed with the court; this in-person contact added 17 people, which was 33% of all attempts.Footnote 7
A total of 209 former jurors completed the interview; however, two participants' data (from one excused and one selected juror) were omitted due to their questionable reliability. In the excused juror's case, I had doubts that the person understood the questions. In the other, the person consented but became increasingly hostile during the interview and appeared not to be giving her honest opinions. The total response rate was 67%. The response rate varied across trials (39–88%); however, only two cases involved response rates below 50%, and these were largely nondistinct (e.g., one occurred near the holidays, when people were more difficult to reach). I contacted 75% of the sample within three weeks of jury selection (range: two to 46 days from voir dire).
Participant Profile
Women were 55% of respondents, and 72% were married. Blacks made up 24% of the sample (n=50) but were 42% of nonrespondents. A logistic regression analysis predicting participation suggests that this underrepresentation is mediated by blacks' greater tendency to have unlisted phone numbers (51% did, compared to only 24% of whites). In all, I recruited 80% of those listed but just 40% of all unlisted persons. The final sample was well-educated (54% had a college degree or higher) and middle- to upper-class (one-third had household incomes between $45,000 and $75,000; 29% had incomes in excess of $75,000). For 54%, the target trial was their first voir dire experience. More than half the interviews (51%) were from excused jurors, but only 13 of these were excused through cause challenges. Due to their small number and in order to focus more directly on the controversial peremptory challenge, analyses reported below omit these cause-challenged jurors.Footnote 8
Instrument
Dependent Variables
At separate points in the interview, participants were asked (1) “How satisfied were you with the jury selection experience?” (I described to them a scale ranging from 1=“very unsatisfied” and 7=“completely satisfied”), and (2) “How willing are you to serve on a jury in the future?” (1=“not at all willing” and 7=“very willing”).
Independent Variables
Ratings of the Questioning Procedure
Privacy. Respondents rated both the relevance and intrusiveness of questioning. For the former, I asked (1) “Thinking about all the questions as a whole, did the questions seem useful for deciding if people could be fair and impartial jurors?” (1=“not at all useful” and 7=“completely useful”), and (2) “How effective did the questioning process seem for deciding who can be fair and impartial jurors?” These two items were strongly related (r=0.55, p<0.0001) and had good reliability (Cronbach's α=0.71).
Jurors rated the extent to which they felt comfortable with questions (“How comfortable did you feel during the jury selection questioning?” 1=“not at all,” 7=“very”). In addition, people rated the extent to which attorneys attended to privacy: “How much did the lawyers seem to care about protecting your personal privacy?” (1=“not at all,” 7=“a great deal”); and “How much did the lawyers seem to care about protecting other jurors' privacy?” (same endpoints). These latter two items were nearly perfectly correlated (r=0.98); however, when combined with the comfort item, only modest reliability resulted (r=0.37, α=0.54). (These items also did not improve the reliability of the relevance construct.) Due to the low reliability of the composite, I report comfort separately from attorneys' care regarding privacy.
Fair treatment. Perceptions of fair treatment were measured by two items: “How fairly were you treated during selection?” and “How fairly were other people treated?” (1=“very unfairly,” 7=“very fairly”; r=74, α=0.85).
Attorneys. Participants rated the attorneys on the following procedural justice constructs: “How respectful was the [district attorney (DA)/defense attorney] during questioning?”; “How honest did the [DA/defense attorney] seem to be in what [he/she] said to you?”; “How interested did the [DA/defense attorney] seem to be in selecting jurors who could be fair and impartial to both sides of the case?”; “How hard did the [DA/defense attorney] try to obtain and fair and impartial jury?”; and “How polite was the [DA/defense attorney] toward potential jurors?” All ratings used the same endpoints (1=“not at all,” 7=“very”). These variables were added together to form a reliable composite for both the DA (α=0.82) and the defense attorney (α=0.86).
Due to the sequential manner of voir dire questioning (see Appendix), all participants rated the DA. However, those persons dismissed by the state (n=25) did not rate the defense attorney as this party asked them no questions. Missing data on this and other variables were handled through a multiple imputation method (Reference RubinRubin 1987), which generated a random sample of plausible values for the missing data, thus representing the uncertainty in these estimates and allowing for valid statistical inferences (Reference Schafer and GrahamSchafer & Graham 2002). The regression coefficients reported below stemmed from this procedure.
Ratings of Outcome
Participants were asked (1) “How satisfied were you with the decision to have you [excused from/sit on] this particular jury?” (1=“very unsatisfied,” 7=“very satisfied”), and (2) “How fair was the decision to [excuse you from/select you for] this particular jury?” (1=“very unfair,” 7=“very fair”). In addition, participants rated their desire to be on the jury through the question “How much did you want to be on the particular jury for which you were questioned?” (1=“very little,” 7=“very much”).
Selection Status and Race
Participants were coded 1 if they were excused from the jury and 0 if they were selected. Race was dummy-coded such that white=1, with 0 otherwise. Neither gender nor listing in the phone book correlated with either of the dependent variables and adding them to the models changed nothing; therefore, these characteristics are not discussed further. In addition, selection status was examined for confounds with either income or education; nonsignificant chi-square tests of association indicated there were none.
Results
Descriptive Statistics
Table 1 reports the overall means and standard deviations for all variables used in the analyses. Several ratings were skewed toward the top end of the scales: ratings of the district attorneys, a sense of fair treatment, and the extent to which the decisions were fair were all above a 6 on 7-point scales. Most of the remaining variables had means (M's) that were well above the midpoint of the scale (all M's>5, with 4 as the midpoint), including overall satisfaction, willingness to return, the relevance of and jurors' comfort with the questioning, perceptions of the defense attorney, and decision satisfaction. Jurors had a somewhat less positive view of the extent to which attorneys cared about protecting jurors' privacy (M=4.71). The lowest-rated item concerned people's reported desire to be on the jury (M=3.66); it was also the most variable (SD=2.15).
Table 1. Distributions of Variables, Overall and by Selection Status
Notes: Asterisks indicate significant differences between means for selected compared to excused.
* p<0.05;
** p<0.01;
*** p<0.0001.
Table 1 also considers ratings by selection status, which indicates simple relationships with some variables. Excused jurors reported less overall satisfaction with the experience of jury selection (t=2.95, p<0.01) and rated the decision regarding their outcome as substantially less fair (t=4.45, p<0.0001), although their satisfaction with their own outcome was equivalent to that of the selected group (t=1.63, p<0.15). Compared to those selected, those excused expressed somewhat less desire to be on the jury (t=2.20, p<0.05), were less likely to say they or others were treated fairly (t=2.69, p<0.01), and tended to see questions as less comfortable (t=3.10, p<0.01). The excused group also had lower ratings of the defense attorney (t=2.57, p<0.05) but rated the district attorney similarly to those selected. Separate analyses (not presented in Table 1) indicated that race predicted only one variable in this study: compared with blacks, whites expressed greater willingness to return for jury service in the future (M=5.82 vs. 4.77; t=−3.41, p<0.001).
In terms of trial-level effects, the intra-class correlations (ICC) for the two dependent variables indicated more trial-level variability for overall satisfaction (ICC=0.14) than for willingness to return for service in the future (ICC=0.03). In other words, 14% of total variability in overall satisfaction, and 3% in willingness to return, represented variability in jury selections, rather than in individual jurors. Thus, few differences in willingness to return to service were associated with features of any given jury selection experience; this was less true for overall satisfaction.
Pearson correlations by selection status appear in Table 2. The privacy variables typically correlated significantly and positively with the attorney variables and with a sense of fair treatment; however, the comfort item had stronger effects for those selected, whereas the other two privacy measures exhibited stronger relationships in the excused group. Taken as a whole, privacy protection was related to other procedural justice indicators but was by no means redundant with them. Fair treatment and most privacy protection variables were related to satisfaction with and fairness of the decision (attorney variables had just two significant correlations among the excused group and just one among the selected group). Satisfaction with the decision was highly correlated with desire to serve: the selected jurors who wanted to serve were more satisfied with the selection decision, and the excused jurors who wanted to serve were more disappointed in their outcome.
Table 2. Correlations Between Variables, by Selection Status
Notes: For those selected, all coefficients greater than 0.20 are significant at p<0.05; for those excused, all coefficients greater than 0.21 are likewise significant at p<0.05. Sel=selected jurors; Exc=excused jurors.
Fair treatment, the attorney variables, and privacy all showed relationships with overall satisfaction, although again there appeared to be differences in the magnitude of these effects by selection status. The other dependent variable of interest, willingness to return in the future, had far fewer correlates: comfort with questioning, desire to be on the present jury, and the decision variables for the selected group. The following set of analyses looked for predictors of the two dependent variables, while simultaneously controlling for other variables.
Predicting Overall Satisfaction
A test for interactions between selection status and the other variables supported the need for different models for those excused and those selected. All three privacy variables significantly interacted with selection status: relevance (t=2.87, p<0.01), comfort with questioning (t=−2.33, p<0.05), and the extent to which attorneys seemed to care about privacy (t=2.08, p<0.05). Selection status also interacted significantly with decision satisfaction (t=−2.55, p<0.05). The separate models by selection status are presented in the middle panels of Table 3.
Table 3. Mixed-Model Regression Results, Presented Separately by Selection Status
+ p<0.10;
* p<0.05;
** p<0.01;
*** p<0.001.
a As all variables reported here were measured at the individual level, the amount of variability accounted for is based only on the portion attributable to individuals; the phrase “Proportion-reduced error” distinguishes it from the more familiar R-square value.
The interaction between selection status and relevance was attributable to relevance being a positive and significant predictor of overall satisfaction for those excused from the jury (t=3.33, p<0.001), whereas it had no predictive value for selected jurors. Although the interaction tests suggested a similar pattern for attorney concern for privacy, when the excused and selected groups were separated out, the effect did not reach conventional significance in the excused group (t=1.47, p<0.15). No other variables reached significance, and this model explains 26% of the juror-level variability in overall satisfaction for those excused.Footnote 9
In contrast, selected jurors who reported more comfort with questioning and more satisfaction with the decision to select them also tended to report higher overall satisfaction (t=3.81, p<0.0001 and t=2.49, p<0.05, respectively). Along with decision satisfaction, decision fairness showed a trend as a predictor of overall satisfaction; however, contrary to predictions, its sign was negative (t=−1.79, p<0.10). A sense of fair treatment (t=3.19, p<0.01) also contributed positively and significantly to the overall satisfaction of those selected. This model accounts for nearly 27% of the juror-level variability in ratings.
Willingness to Serve on a Jury in the Future
Once again, there was evidence that excused and selected jurors require separate predictive models; decision satisfaction again significantly interacted with selection status (t=−2.01, p<0.05). The right part of Table 3 presents the models for each group. Decision satisfaction bore no relationship to the ratings of those excused, but it significantly and positively predicted the ratings of those selected (t=3.75, p<0.001). Those who were satisfied with the decision to be on the present jury were more likely to say they were willing to return for service in the future. In addition, race positively and significantly predicted the views of those selected; net of other variables, whites reported more willingness to return than did blacks (t=2.43, p<0.05). This model explains about one-quarter (24%) of the juror-level variability in the selected group's ratings. Within the excused group, willingness to return had few predictors, and the total model explains just 7% of individual-level variability. Comfort with questioning showed a near-significant and positive association with willingness (t=1.87, p<0.07). Race of the juror also fell just short of significance (t=1.84, p<0.07), with a pattern similar to those selected (whites were more willing than blacks).
Discussion
There are many reasons to be concerned, as the Supreme Court has been, over unfairness in the procedures for distributing jury service once people show up to the courthouse. Nevertheless, as theorists of procedural fairness have pointed out, the factors that determine views of fairness depend upon the type of “conflict” that a given procedure aims to settle (Reference Thibaut and WalkerThibaut & Walker 1978). The nature of the dispute—a cognitive conflict over the state of the world or a conflict of interest—and the associated role of individual interests will predict the relevance and forms of procedural and distributive concerns. Jury selection is a challenging procedure to classify in these dual terms because, I argue, the court asks jurors to view the forum as a pure cognitive conflict even while jurors' interests may conflict with the goals of the procedure or the decision maker. Research in other domains suggests that conflicts of interest between a person and another decision maker will alter perceptions of justice (Reference MolmMolm et al. 2003). I argue here that conflicts of interest between attorneys and jurors over privacy protection and decision outcomes should predict views of justice. The data speak to the relevance of these areas for understanding fairness in the distribution of a duty; however, inconsistencies in the results also indicate that jurors did not narrowly focus on their own needs.
Conflicts Over Privacy
Significant interactions with selection status revealed that the perceived relevance of questioning was a strongly positive—indeed the sole—predictor of excused jurors' overall satisfaction ratings. Selected jurors' ratings of their comfort with questioning also significantly predicted overall satisfaction, together with ratings of how fairly they were treated. Finally, other privacy variables demonstrated trends toward significance in the excused groups (attorney care for privacy as a predictor of overall satisfaction and comfort with questioning as a predictor of willingness to return). Results support the notion that jurors attend to how well attorneys manage the inherent tension between their desire for more information and jurors' interest in privacy protection. This is a mirror-image tension to the one normally examined in procedural justice studies—i.e., whether people are able to express themselves enough via “voice.”
The results are important for two reasons. First, note that without a consideration of privacy interests, the conception of procedural justice in this setting would have been biased. Simple behavior-oriented indicators of procedural justice (a sense of fair treatment and perceptions of decision makers' respectfulness, etc.) were not as helpful as privacy protection in understanding reactions to this distribution system, especially for those excused. This is largely attributable to the fact that jurors perceived themselves as having been treated quite well. My own court observation confirmed this; attorneys asked almost everybody the same questions (excepting individual follow-up), bent over backward to be appreciative of jurors' presence, and always addressed them with the Southern conventions of “Sir” and “Ma'am.” Nevertheless, attorneys still had to direct questioning to what was pertinent (i.e., what biases does this person possess?), and they still had to probe—sometimes deeply, sometimes confusingly—into jurors' lives. It was the perception of how well attorneys managed this questioning, above and beyond how polite or respectful they were, that proved predictive of jurors' reactions to jury selection.
Second, because all models controlled for perceptions of attorneys' politeness and respect, and a general sense of fair treatment, privacy protection should not have been viewed as a simple matter of attorneys' being nice. This result strengthens accounts of procedural justice that distinguish a sense of authorities' good behavior from other domains of fairness, especially those concerning an imbalance of power (see, e.g., Reference MikulaMikula et al. 1990). Several theoretical accounts of procedural justice focus on the role of vulnerability at the hands of authorities in making people care about fairness, as well as the need for authorities to signal that they are not exploiting their power (see review in Reference Tyler, Lind, Sanders and HamiltonTyler & Lind 2001). Jurors appearing for jury selection are indeed vulnerable, especially with respect to their personal privacy. These jurors had to rely on the attorneys (and the judge's oversight) to attend to their interests, even when such protection threatened attorneys' interests in the information-gathering task at hand. Lacking voice in either decisions or in how the procedure was to be conducted,Footnote 10 this sample's attention appears to have shifted to how well the decision maker wielded his or her inquisitorial control.
A reviewer of this work suggested that relevance might be a distributive, rather than procedural, concern. That is, perhaps relevance signals that the most deserving jurors were selected, and the excused jurors may have focused on this issue as a marker of good decisionmaking. However, there are several reasons to consider relevance as more an issue of procedural rather than distributive fairness. First, the zero-order correlation between relevance and either of the outcome variables was lower (r=0.20 and r=0.15 for decision satisfaction and fairness, respectively) than its relationship with a sense of fair treatment (r=0.30). Second, in other work I demonstrate that excused jurors' own guesses about the reasons why they were excused showed little relationship to overall satisfaction (Reference RoseRose 2003); thus, it does not appear that excused jurors' own assessments of the sensibility of the decision drive overall satisfaction. Finally, in order to more specifically address a distributive view of privacy, I ran additional post hoc models for those excused in this study, using decision satisfaction and decision fairness as the dependent variables and the three privacy variables as predictors. The relevance of questioning was never a significant predictor of either outcome rating (in contrast, an analysis using fair treatment as the dependent variable did find effects for privacy as relevance). In short, the expectation that attorneys limit their inquiries to relevant topics appears to implicate procedural rather than distributive concerns.
Even when conceptualized as an issue of procedural fairness, effects for privacy were not consistent. In no instance did all three variables significantly predict the views of both those excused and those selected. A larger sample would be useful in future studies to allow small, trend effects to emerge, and the multidimensional nature of privacy should be explored and refined. For example, the fact that comfort with questioning operated differently suggests that privacy violations should be distinguished on the basis of a subjective sense (i.e., the extent to which someone felt personally uncomfortable) and more objective assessments of the perceived relevance and attorney care. The reason for the relation of such differences to selection status is not immediately clear. I suggested at the beginning of this article that selected jurors may minimize the importance of relevance once they learn more about a case; however, note that the mean rating of relevance did not differ by selection status and, in any case, this does not explain why the effects for comfort were so much stronger for those selected than for those excused in predicting overall satisfaction. If anything, uncomfortable disclosures during jury selection should be more compelling and memorable for those just recently dismissed.
Despite these inconsistencies, in practical terms, these findings suggest that management of information-gathering is one route for increasing the legitimacy of jury selection. Attorneys and the courts would do well to keep questioning on track, to anticipate issues that will make jurors uncomfortable, and, where possible, to explain the need for odd-sounding or intrusive questions. Research in other domains suggests that even brief explanations can dissipate negative reactions to norm violations (Reference LangerLanger et al. 1978). When feasible, jurors would likely appreciate having a better understanding of an attorney's or judge's need for information which, from their perspectives, may seem intrusive or of little use. This is especially true for those excused, as jury selection is often the sole interaction they may have with the court. Although not directly assessed here, the attention to information may mean that overly superficial jury selections also could be seen as “unfair.”
Conflicts (or Lack Thereof) Over Outcomes
Apart from questions posed, the decisions themselves are another way in which jurors' interests may conflict with those of decision makers. In traditional conflict of interest situations, people are expected to respond negatively if they believe their own outcome was undeserved, i.e., if the result is deemed both undesirable and unfair. The interesting news from this study is the almost-total irrelevance of personal preferences and decision fairness to overall satisfaction with jury selection and willingness to return. With respect to desire to serve, the null finding is not due to limited variability. The measure had a low mean and fairly high standard deviation. Clearly, some looked forward to service and some loathed it, but this particular sentiment was not predictive of the dependent variables, except at the zero-order level (e.g., desire to serve now correlated well with willingness to return later). Decision fairness likewise proved a poor predictor in these analyses, save a strangely negative trend (p<0.10) in predicting a selected juror's overall satisfaction. In contrast to the variability in desire to serve, ceiling effects likely explain this unusual finding. Selected jurors overwhelmingly rated their own outcome as fair (M=6.49 on a 7-point scale); thus the negative slope likely reflects an undue influence of a few outliers.Footnote 11
The results for those excused are perhaps the most striking, given the serious concerns about the message that peremptory exclusion from juries may send to citizens (Reference HoffmanHoffman 1997; Reference MarderMarder 1995). Despite such concerns, outcome ratings were never predictive despite the fact that excused jurors had a more negative view of the fairness of their outcome and were somewhat less satisfied overall with their jury selection experience.Footnote 12 It must be emphasized that all findings with respect to outcome—null or otherwise—occurred in a sample that rated both the attorneys and their sense of fair treatment highly. Thus it is more correct to say that views of the decision did not predict overall satisfaction or willingness to return given that the attorneys were seen as behaving in an otherwise professional and respectful manner. With this caveat, there did indeed appear to be a “generally accepting” attitude (Reference Diamond and LitanDiamond 1993) toward the peremptory's role in jury selection among those so excused. This conclusion in no way negates efforts to diagnose and address discrimination where it occurs (Reference BaldusBaldus et al. 2001). It simply indicates that distribution to a jury, including a process that utilizes attorney discretion, is not inherently “unfair” to citizens (see also Reference RoseRose 2003).
Perceptions of outcomes were more relevant for those selected, although the effects centered around ratings of decision satisfaction, which significantly and positively predicted overall satisfaction and stated willingness to return in the future. There are several possible explanations for why decision satisfaction was uniquely significant. For instance, given the court's request of jurors to put aside their interests, decision satisfaction may have been comparatively easier to assess and hence more likely to be predictive. Thus, although the court discourages people from thinking in terms of what they want or what is fair, it cannot preclude them from feeling more or less satisfied with their result. Further, unlike decision fairness, a subjective sense of decision satisfaction requires no reference to a standard of what's “fair.”
However, the measurement explanation is less persuasive in light of the significant interaction, i.e., the fact that a satisfactory outcome predicted willingness to return and overall satisfaction only when the outcome entailed actually sitting on the jury. Perhaps, as I suggested in the beginning of the article, this effect reflects the relatively greater impact of the final decision on the time and energy of those selected. In other words, outcome ratings may become a more central consideration in evaluations when more is at stake, a somewhat controversial assertion in the justice literature (compare with Reference Tyler, Lind, Sanders and HamiltonTyler & Lind 2001, reviewing research that finds no effect for magnitude of an outcome on judgments). What are the “stakes” for those selected that might contribute to viewing the decision as satisfactory? In this sample, the actual time the trial consumed appears less of an issue. A post hoc analysis examining the ICC of decision satisfaction among those selected found that the amount of trial-level variability was only about 2% of the total (ICC=0.02). Further, due to last-minute plea-bargains and other factors involving four trials, about one-third of those selected did not actually have to decide the case. I examined differences in decision satisfaction between these two groups and found none (t=0.37). Naturally, the relationship between trial time and views of outcomes should be examined in a larger sample with a greater range in trial times, but the evidence here did not suggest that time is the jurors' main concern.
Apart from time demands, differences in decision satisfaction likely reflect variability in comfort with the responsibility of jury service. Reference BoatrightBoatright (1999), for instance, finds that people's expectations about whether they would do a good job at being jurors positively predicted obeying their jury summons.Footnote 13 In this sample, it is noteworthy that decision satisfaction predicted both overall satisfaction with jury selection and willingness to return, suggesting that some people possess a more general willingness to accept the responsibility of serving across circumstances. The idea of differences in comfort with service may likewise explain the sole race effect observed in this study. Even after accounting for several evaluative factors regarding jury selection—none of which varied by race—blacks reported less willingness to return, especially those who had been selected. That is, even if blacks perceived the court as treating them well and even if they felt satisfied with the decision to select them, they were simply less interested in returning. Note further that nearly all defendants in this study were black. Responsibility for the prospect of sending to jail an African American—one who, typically, is young and male—may evoke strong aversion to jury service among black jurors. Again, this is speculative and but one explanation; the small number of trials and of blacks precludes any strong claims. Nevertheless, the weak effects for procedural justice as a predictor of willingness to return for all respondents provide further evidence that people's ideas about service are not necessarily ameliorated by the court. Jury service may be one of the United States' most important forms of democratic participation, but actually serving on a jury remains, for many, an uncomfortable duty.
Jury Selection as a Quasi-Cognitive Conflict
This study does not present a simple portrait of what is at stake for prospective jurors and what is “fair” about jury selection. The data reveal that people were more satisfied overall if they had positive views of privacy protection (along with a general sense of fair treatment), and if they expressed satisfaction with the decision to have them serve on a jury. However, individual jurors also minimized the extent to which jury selection was “about them” and their particular interests. Indeed, the news from this study is how few variables predicted outcomes. Desire to serve and decision fairness were not particularly useful variables, and even though privacy had effects, these differed across selection status, conceptions of privacy, and dependent variables. In short, although not entirely blended with the interests of the court and parties, citizens' interests were not predictably at odds with what jury selection is trying to accomplish.
Thibaut and Walker addressed the possibility that there are “hybrids” of dispute-types, in which a decision maker faces a serious, zero-sum conflict of interest as well as a truth claim that must be settled (e.g., whether a power plant a community opposes causes environmental harm; Reference Thibaut and WalkerThibaut & Walker 1978). In such cases, they suggest the necessity of a “two-tiered” solution, one that adequately addresses the need to discover truth and the need to give voice to disputing parties. However, just as jury selection is not wholly a cognitive conflict or a pure conflict of interest, the hybrid conception and its remedy likewise seem ill-fitting. These data do not paint the conflict of interest aspect of routine jury selections as particularly intense or serious—although in very long or high-profile trials, which may have few willing jurors, it might be. Instead, the jurors' qualified deference to decision makers' judgments about questioning and outcomes suggests a better label: “quasi-cognitive conflict.” Massive procedural changes do not seem necessary in order to secure a sense of fairness about the process and the outcomes. Rather, justice here calls for a more quiet recognition of the ways in which the jurors' own concerns about privacy and reservations about service need to be acknowledged and, where possible, protected.
Appendix: Jury Selection in One State Courthouse
Jury service in this county ran on a one-day/one-trial system, which means that the result of a day's service was either selection for a single trial or dismissal from any remaining service; this is a popular, albeit not universal jury reform (Reference MunstermanMunsterman et al. 1997). Jurors were oriented to the one-day/one-trial system and to most jury selection procedures through a video that was narrated by a North Carolina native, the late Charles Kuralt.
The Structure
Following orientation, and usually after a few hours of waiting, a group of between 30 and 40 randomly selected people (for a typical trial) went to a courtroom for jury selection. Then a randomly selected subgroup of 12 took seats in the jury box. The structure of questioning and the exercise of peremptory challenges in this court followed a “sequential method” (Reference Bermant, Shapard and SalesBermant & Shapard 1981). After an introduction from the judge, the ADA initiated questioning and then excused jurors. Those excused were replaced, and the ADA began again on this new group. Only after the prosecutor said, “The state is satisfied,” with 12 jurors in the box did the defense begin its questioning and exercise challenges. After the defense eliminated jurors, the process repeated (i.e., questioning went back to the ADA) until 12 jurors acceptable to both sides, as well as one or two alternates, filled the jury box.
There are other structures. Were a “strike” method used, a clerk would summon as many jurors as it takes to both fill a panel and exercise all peremptory challenges. All prospective jurors would be interviewed at the same time, challenges for cause exercised (and those jurors replaced), and only at the end would both sides simultaneously exercise peremptories. The first 12 jurors in the box would become the jury. Note that in the sequential method, attorneys exercise challenges without knowing anything about who may take the seat of the excused person, whereas in the strike system, attorneys hear from everyone. Also in the strike system, the peremptory's source may be unclear; attorneys sometimes submit strikes to the judge, who simply announces those selected. In the sequential system, challenges follow the questioning of each side, and jurors always know who has dismissed them.
In this court, jury selection in routine cases (drug cases, assaults, property crimes) usually lasted between two and three hours; in trials expected to be lengthy or those involving multiple defendants, selection took two days or more. Unlike in federal court (Reference BermantBermant 1977), the bulk of voir dire questioning in this state courthouse came from the attorneys with oversight from the judge, who ruled on challenges for cause and the rare objection to a line of questioning. Thus, those dismissed from jury selection would sometimes hear little from the judge, who in some cases gave only introductory remarks.
The Questions
Although the 13 trials observed for this study varied in charges, questioning across trials was remarkably consistent. If judges asked questions, they would begin by having jurors state their names and say in which part of the county they resided, how they were employed (if so), whether they were married, how their spouses were employed (if they were), and whether the jurors had ever been on a jury. The group was asked whether anyone knew of a reason they could not “give this trial their undivided attention and render a fair and impartial verdict.” This quickly allowed people to offer up hardships or other conflicts (problems at work, upcoming travel, caregiving responsibilities, illness). Some other judges asked no questions and immediately turned questioning over to the ADA, who usually elicited similar basic information, or it came out later in the defense attorney's questions.
In addition to the above, prosecutors usually inquired about knowledge of the parties, outside information about the case, or strong feelings about a type of crime (especially in drug cases). ADAs also focused on issues that would make someone suspicious of the state. Thus, the prosecutors typically covered all of the following areas, often posed in terms of applicability to the jurors themselves as well as their family members:
• previous experiences in court, especially as a defendant
• problematic encounters with police officers
• prior legal training or relationships with attorneys or court personnel
• knowledge of the crime scene
• religious beliefs that forbid “sitting in judgment”
• willingness to follow instructions
Several prosecutors also attempted to understand the jurors' “ties to the community” and would ask whether jurors belonged to a religious community (and if so, which one), as well as whether they belonged to clubs or organizations to which they “give considerable time or consider important to who [they] are.” All questions were posed to the group, and attorneys then followed up individually with people who indicated that a question was applicable. In very serious cases, such as a death penalty trial or a trial receiving extraordinary publicity, questioning would be done one-on-one with individual jurors to avoid other people's influence on answers. These cases were rare and not part of this sample (although two of my cases did involve some media coverage).
Although the defense attorneys were also likely to be interested in some of the same issues as the ADA, they also sought information that tapped factors that might make someone have more allegiance to the state than to the defendant, such as:
• social or personal relationships with police officers or other law enforcement
• a history of criminal victimization (self or family)
• gun ownership, as well as “whether their lives have been impacted by guns or violence” (in trials for crimes involving guns)
• negative impressions of lawyers
• willingness to follow instructions on the burden of proof
Much as others have found (Reference BalchBalch et al. 1976; Reference DiamondDiamond et al. 1997), these jury selections included questions that transmitted rather than gathered information from jurors. For example, the defense might “ask”: “You would all agree, would you not, that the presumption of innocence is one of the most important rights a defendant enjoys?” or from the state, “Y'all understand that ‘proof beyond a reasonable doubt’ does not mean ‘beyond all doubt’?” The tone of questioning (seriousness, depth of follow-up) varied somewhat across cases, but the general pattern was that the ADA handled much of the questioning and tended to be more focused and direct. Some defense attorneys attempted to be more ingratiating and, because the state had already covered the basics, would occasionally use their time to ask more “folksy” questions, such as “What do you like most and least about your job?”“What do you do in your free time?”“What sort of hobbies do you do for fun or recreation?”“Do any of you have grandchildren?” These were, however, atypical lines of inquiry. Each was separately observed on just one occasion, except the hobbies question, which appeared in two trials.
