It has been well over a century since Alexis de Tocqueville first hypothesized a relationship between the institution of jury service and civic engagement, yet this appealing claim has gone untested. In the mid-1970s, political theorist Carole Pateman restated Tocqueville’s idea as a more general participation effect, whereby any form of civic participation can promote others.[2] In the years that followed, there emerged no compelling investigation of the proposition. Reflecting on this dearth of evidence, both Pateman and fellow political scientist Jane Mansbridge declared that the participation effect—now an act of faith among those advocating participatory forms of democracy—might remain untested in perpetuity, as they doubted that one could design and implement a suitable test.[3]

            Meanwhile, the proposition that the jury can change the juror has shaped law and public policy without having an established basis in fact. In 2004, Japan rewrote its laws to develop a “quasi-jury” system based partly on the belief that the jury “promotes a more democratic society.”[4] A decade earlier, the U.S. Supreme Court affirmed the constitutional right to serve on juries by invoking Tocqueville’s claim that juries are essential to democracy.[5] The Court’s source for this belief was, simply, Tocqueville himself.

            Not long after the Court handed down its opinion, the authors of this article were sitting in a dusty New Mexico plaza trading stories over caffeinated beverages when the conversation turned from baseball to Tocqueville…

Tea sipper: “Why have juries?”

Coffee drinker: “They’re good for you.”

Tea sipper: “How do you know? What good comes from having twelve ordinary citizens deliberating when you could have one trained judge making informed choices?”

Coffee drinker: “You tea drinking elitist, Tocqueville told us juries are schools for democracy.” 

Tea sipper: “And Fourier said the sea would turn to pink lemonade after we all lived in phalansteries. How would you measure democratic citizenship, anyway?”

Coffee drinker: “We know people are being good citizens if they participate in political society,” the coffee quaffer replied, savoring another swallow of his inky drink.

Tea sipper: “So people are more likely to vote after serving on juries?”

Coffee drinker: “Why not?”

Tea sipper: “It’s testable. Jury records must be public, and we know that voting records are. Care to make it a friendly wager?”

Ten years later, we are ready to tell you who won that bet.

Designing a Hard Test for the Participation Effect

            Studies on juries have often relied entirely on self-reported data (surveys) or a handful of in-depth interviews. In either case, the skeptic can doubt the truth claims about the jury based on this evidence. After all, many addicted smokers assert that they can “quit tomorrow.” Are they any less reliable than others who self-report? To be fair, surveys and interviews have their place as legitimate research tools, but we did not feel they were definitive enough to decide our real money wager.

            An ideal study would also test the impact of reluctant participation. Normally, when citizens choose to engage in a civic activity, they do just that—they choose to participate, either spontaneously and voluntarily[6] or in exchange for honoraria and other positive incentives.[7] Public deliberation might benefit those who seek it out, but an ideal study of the participation effect would determine whether a civic activity can spark future public engagement even when some participants would rather be elsewhere.

            One more common problem in social scientific research is that effects are too often measured only hours, days, or a few weeks after the researcher’s intervention. Short-term assessments of behavior change leave open the question of the how long an impact lasts, yet in the laboratory, researchers often have access to study participants only for a limited time. Even if there is a follow-up survey separate from the experimental manipulation, it usually comes promptly, lest too many of the study’s participants drift out of the study’s reach. If you bet money against an effect you don’t want to lose merely because it lasts a few hours or days.

            The test we proposed to resolve our wager about the civic impact of jury deliberation avoids all of these measurement problems by linking official jury records with the actual democratic act of voting as recorded by the county’s registrar of voters. This is not the best way to understand the robust and varied nature of civic engagement, but it can demonstrate whether an effect exists under the most rigorous of testing conditions. Although subject to human error, the public records we gather do not depend on fragile human memory or anyone’s desire to impress an interviewer. Given that the researchers and recording clerks only meet long after the court and voting documents have been produced, bias in the recording of the data is implausible. In addition, this approach can yield as close to a perfect response rate as the archives will allow, with little chance that jurors’ experience at the courtroom could influence the odds of their appearing in our final dataset. Because jury service is mandatory, it also brings into our study many people who embark on their public service journey at the county courthouse only reluctantly. Finally, our approach allows us to collect data spanning a period of years, such that we can measure the impact of jury service on voting in elections held months or even years after the study participants completed their service.

Making a Hard Test Fair           

            Though our research design provides a hard test of the participation effect, it is just as important to ensure that the test is fair. The rules of statistical inference are designed to encourage modesty. When one hears of a “margin of error,” for instance, that is an acknowledgement that our estimates are not precise. If a survey finds that fifteen percent of Hawaiians would like to take a holiday in Alaska, the more careful scholar might add that the correct estimate is really “ten-to-twenty percent.” In this same way, when we make a comparison, such as whether more Hawaiians or Texans would rather travel to Alaska, we require that there be no significant overlap in the two estimates. If, in this case, we found that fifteen-to-twenty-five percent of Texans hear Anchorage beckoning them, we would conclude that there was “no statistically significant difference.” The margins of error, in these cases, overlap, and that makes a modest researcher reluctant to announce with confidence that a real difference exists in the vacation predilections of the two states’ residents.

            The smaller the differences one hopes to detect, the more likely it is that these margins of error will overlap. The same is true for our pursuit of the participation effect. We expect this effect to be relatively small, given the stability in people’s voting behavior, the varied ways in which people respond to the jury experience, and the multitude of random factors influencing the likelihood of voting on any given day. Even expensive and much ballyhooed Get-Out-The-Vote (GOTV) campaigns, for instance, have only minimal effects on voters (if any at all), with the most efficacious face-to-face canvassing yielding less (often far less) than a 10% increase in a citizen’s voting likelihood.[8]

            Fortunately, one can counterbalance the difficulty of finding small effects by shrinking the margin of error. By the laws of statistical inference, the larger the sample, the smaller the error margin. Creating a large sample can be an expensive proposition but with funding from the University of Washington and the National Science Foundation, we were able to do precisely that.

The National Jury Sample

            In 2004 and 2005, a team of researchers set off to gather data in counties from diverse regions across the United States.[9] This undertaking required not only a large travel budget but also a leap of faith. What if the study produced nothing interesting or even refuted the findings of the earlier work? Then the researchers would have nothing for their troubles save a merry trip to a brightly-lit courthouse in Omaha, Nebraska or Fayetteville, North Carolina. For those whose heart races at the cool touch of manila folders and metal file cabinets, the trip would offer its own rewards, plus frequent flyer miles and a few extra cups of tea.

            The principal objective of this larger study was to determine whether the jury-voting link would appear across a diverse sample of jurors and jurisdictions with varied trial rules, jury facilities, and court practices. Going out of our way to gather a disproportionate number of civil jurors, the larger National Jury Sample would also test whether this participation effect appeared equally strong for criminal and civil juries.

            Collecting thousands of jurors across a broader range of trials also permitted the creation of a more precise comparison group consisting solely of “cancelled trials,” in which the juror is empanelled, the trial begins, and then the trial ends prematurely—even before the jury can begin its deliberations. This includes mistrials, dismissals, withdrawn cases, settling out of court, or waiving the right to a jury after the trial began. This study works as a natural experiment where cancelled trials are the control group against which we compare four distinct jury experiences: serving as an alternate, defendant pleading guilty after the trial begins, hung jury (in part or full), and complete verdicts.

            Based on the pilot study,[10] we expected that reaching a verdict would have a positive effect on future voting relative to the experience of merely sitting in the jury box for what would end up as a cancelled trial. In addition, we anticipated that hung juries would have a net positive effect compared to cancelled trials.

            The hung jury prediction puts the focus squarely on experiencing deliberation, rather than the personal satisfaction of convicting a defendant or setting one free. A review of contemporary data on hung juries found they typically result from a complex case for which neither side can predict the trial outcome; juries typically hang only after intensive, prolonged jury deliberation.[11] Moreover, a hung jury often results in a decisive finding for the defendant (or respondent), so it serves as a final jury verdict in that sense. In any case, the hung jury constitutes an unambiguous experience of deliberation, regardless of outcome, and a core claim in this study is that participation in deliberation itself is the most critical component of the jury experience. To presume otherwise would require all deliberative forums to yield conclusive group decisions or fail to inspire their participants.[12]

            The other two outcomes—sitting in the jury box as an alternate or witnessing a guilty plea before beginning jury deliberation—were less clearly distinct from a cancelled trial. Alternates do not participate in deliberation, therefore when their jury reaches a verdict, they play no direct role, but they share the experience of sweeping oratory and numbing tedium in the trial. If the act of witnessing a trial plays a role in promoting civic engagement then it should apply to alternates as well. Similarly, a guilty plea yields a final verdict for all jurors, akin to a “conclusive” outcome, but it involves no jury deliberation. To test our expectations, these two outcomes were also compared to cancelled trials to test whether either would produce the same positive effect predicted for hung and full-verdict juries.

            In addition, consistent with the aforementioned pilot study, we predicted that the number of charges against the defendant in criminal trials would provide an additional boost in post-jury voting rates. More charges are one indicator of the complexity of the decision task, with multiple counts against the defendant requiring the jury to reach multiple verdicts. Deliberative theorists and practitioners alike have stressed the importance of the depth of deliberation when considering its potential benefits for participants.[13] In the context of the jury, the number of charges provides one glimpse of such depth.

            The National Jury Sample was further refined by distinguishing between historically frequent and infrequent voters, based on pre-jury voting behavior. A commonsense consideration of the participation effect is that it can draw into public life those citizens who are relatively less engaged, rather than merely reinvigorating those who are already regular participants. For instance, in his study on campaign participation, political scientist John Freie writes, “The popular panacea offered by some to reduce alienation is often participation itself. Political participation, it is hypothesized, will alleviate feelings of alienation and result in future political involvement.”[14]

            To test the jury effect for less versus more electorally active citizens, we split the National Jury Sample between habitually infrequent and frequent voters, based on their pre-trial voting rates. The study still controls for variation in voting rates within these two groups, but we separate low- from high-turnout juror groups to determine whether the participation effect is visible across both populations.

            Our expectation is that less active citizens are more likely to experience a cognitive and behavioral shift toward greater future public engagement than those who have already caught the civic spark. For infrequent voters, jury service offers entrée to a largely unexplored world—that of citizen participation and self-government. If jury service makes citizens, in Tocqueville’s language, “feel the duties which they are bound to discharge towards society,”[15] this feeling is newer for those previously less inclined to recognize and fulfill such duties. Regular voters may still experience the participation effect, but its impact should be stronger for habitually infrequent voters than their steady-voting counterparts.

Building the Sample

            Collecting a large and diverse sample of jurors required identifying a variety of counties in different parts of the country that had publicly accessible court archives, legible and complete jury records, and cooperative administrative staff. The eventual merger with electoral data also required access to complete and digitally archived voter histories dating back to at least 1994. To test the generalizability of the pilot study findings, we also aimed to assemble a set of counties that were demographically and politically diverse.

            It was not possible to construct a fully-representative national random sample of jurors for technical and logistical reasons. Chief among these was that only some courts make their jury records readily available for public inspection, and among those, many do not consistently record jurors’ full names, which are necessary for matching jury lists with voting records. In addition, counties above a modest size (e.g., those hosting a city larger than Seattle) would produce too few unique matches between full juror names and the corresponding county list of registered voter names. With these limitations in mind, the goal was, once again, to create a broad and diverse sample—not a perfectly representative one. Following these guidelines, the final set of data collection included Boulder County (Colo.), Cumberland and Swain Counties (N.C.), Douglas County (Neb.) El Paso County (Tex.), Orleans Parish (La.), Summit County (Ohio), and Thurston County (Wash.). Figure 1 shows the geographic distribution of the counties contributing to the National Jury Sample.

            For each of the eight counties studied, we employed the same general procedures. In each case, the patient court administrators obtained for us a list of trials during a two-to-three year period, then for hundreds of hours, tireless researchers and assistants pored over files by hand to record the full names of jurors from written lists enclosed in the files. Juror names were then used to laboriously merge data with electronic voter registration lists provided by the county’s election office, with voter histories extending back roughly five years before and five years after the period of jury service (see Figure 2).

            The final result was a dataset with 13,237 empanelled jurors, including 8,573 who were seated in the jury box for a criminal trial and 4,664 who sat for a civil trial. Of these jurors, 10,300 served on juries that reached complete verdicts, 554 were hung on some or all charges/claims, 818 were excused from the jury box after the defendant changed his/her plea to guilty, 904 were dismissed for various other reasons (mistrial, withdrawn charges, out-of-court settlement, etc.), and 576 served only as alternates, never joining in jury deliberation. Of these jurors, 65% matched voter files to produce 8,614 jury records with matching voter histories. (See the Journal of Politics web appendix for this article to see figures by county/parish.) A full sample of this size was necessary because we were pursing a relatively small effect size and breaking the sample down into smaller sub-samples for comparisons.


            We conducted regression analyses to judge which of our predictors explained changes in jurors’ voting behaviors. That meant conducting a total of four analyses: For persons with a history of infrequent voting, we ran separate analyses for criminal versus civil jurors, and we then ran the same analyses for more steady voters.

            The left-hand column in Table 1 shows what group was indicated as responsible for increased election turnout among infrequent voters serving on criminal trials. Criminal jurors reaching a verdict were more likely to vote in later elections. In the National Jury Sample, the likelihood voting in a future election rose by an average of 4.3% over the span of several years.

            Those previously infrequent voters who had the intense deliberative experience of a hung jury showed the largest effect. These jurors did not reach a conclusive verdict, but they did all have the opportunity to deliberate at length about the trial with their fellow jurors. On average, a participant in a hung criminal trial experienced a 6.8% increase in their voting rate in the years after completing their jury service.

            Consistent with this emphasis on the intensity of the deliberative experience, the number of charges against the defendant again proved a significant contributor to post-service voting. Each additional charge added a 1.3% increase in the likelihood of voting. For a complex criminal trial with, say, four charges against the defendant, this would amount to an average increase in voting of roughly 4%.

            Figure 3 puts these findings in comparative terms. For previously infrequent voters, the effect of deliberating on a criminal jury is comparable to the civic boost a high school student gets from taking a mandatory civics course for a semester, and it exceeds the impact of service on student council.

            The other results in Table 1 are unremarkable. As expected, neither of the other trial outcomes—serving as an alternate or having a defendant plead guilty before deliberation began—had any significant effect on the likelihood of voting. Also, in the National Jury Sample, the effect of a better-estimated pre-jury voting history was even greater than in the pilot Project, but again, this was to be expected and simply reinforces the fact that voting is habitual behavior.

            In fact, the rest of the analyses in this investigation are all non-findings. The results confirmed suspicions from interview data and the pilot project: Civil trials really are different. In the National Sample, the right-hand column of Table 1 shows that no significant participation effect was found for the civil jury experience among infrequent voters. The same column in Table 2.4 shows that civil jurors with a history of regular voting also experienced no civic benefit from their civil jury service.

            More generally, there were no participation effects for either civil or criminal trials among frequent voters. Aside from the obvious impact of pre-jury voting service, the likelihood of voting after jury service was not significantly related to any feature of that service experience.


            These data provide strong evidence for a pervasive and enduring effect of criminal jury deliberation on civic engagement, at least for those entering jury service with a relatively spotty voting record. The effect amounts to roughly a 4- to 7% increase in average turnout. On top of that, we find a participation effect that increases more than one percent per additional charge against the defendant. These effects were present even with the post-jury voting rate being measured over roughly five years. The National Jury Sample also yielded these results with a diverse sampling of courthouses and jury pools.

            The effect of criminal jury deliberation on voting, however, does not hold for those voters who are already active. We had expected this population to experience less voting change, but the data make it clear that the jury does not affect these individuals’ voting behavior. This finding substantially qualifies the extent to which one should expect the participation effect to occur in juries.

            The National Jury Sample made clear that the participation effect comes from all deliberating jurors, not just those who reach verdicts. We anticipated that hung jurors would experience the participation effect, and as it turned out, it had an effect slightly greater than that obtained for jurors who agreed on verdicts. Again, the nature of the deliberation may be a factor. Post-hoc analysis of the court records showed that hung jurors, on average, were more likely to have asked a judge for assistance during deliberation (79% of those on hung juries did so, compared to just 28% of jurors reaching verdicts). Also, hung juries deliberated for a considerably longer time—roughly nine hours—than did their verdict-reaching counterparts, who were usually done in four. For many jurors, long and challenging deliberations that end in a contested deadlock leave a stronger an impression than does reaching a comfortable verdict with one’s peers.

            In addition, the absence of any effect whatsoever for jurors in the alternate and guilty plea groups across the four sub-samples means that these experiences do not have the same impact as full-fledged jury deliberation. A spontaneous guilty plea in the courtroom may feel, to the juror, more like a cancelled trial than a conclusive experience. Serving as an alternate also appears to leave no strong impression on jurors, regardless of their prior voting history or the type of trial to which they were assigned.

            Finally, the clear impact of the number of criminal charges reinforces the significance of jury deliberation, but in a complex way. In the pilot study, we speculated that the number of criminal charges reflected the seriousness of the charges, but a post-hoc coding of charge severity in these data did not bear out this interpretation. We believe the best alternative interpretation is that juries weighing more charges simply face a more complex deliberative task as they make more decisions often involving interlocking judgments and mixed verdicts.

            Notice, however, that the number of charges has this effect independent of whether the jury was sent away to deliberate. As we explained before, statistical regression analyses separate out the effects of each independent variable, and the number of charges has its own heft as a predictor of future voting. This is consistent with the well-known fact that jurors begin processing a case cognitively before they sit down face-to-face to discuss it.


            This project has walked a long road from the dusty Albuquerque café to the collection of thousands of records in a large national jury sample. For us, the lesson is that you don’t get to the end without starting in the beginning. Returning to the wager that launched this escapade, we have now marshaled sufficient evidence to show that jury service can, indeed, contribute to an increase in civic engagement, at least in the form of voting. But we have learned many other lessons along the way.

            The National Jury Sample wielded a geographically and ethnically diverse sample that permitted a more definitive and fine-grained investigation of the link between jury service and voting. The national study produced three main findings. First, the main increase in voting rates results from the experience of jury deliberation, not merely from trials in which jurors reach a verdict. Second, cases involving multiple charges spark an additional, positive participation effect, which suggests the need to consider the complexity of a case and other features that might make jurors’ deliberative task more challenging or engaging. Third, none of these effects occur for jurors who are already frequent voters or who sit in the jury box for civil cases.

            The National Jury Sample also gave a more refined estimate of the participation effect. The study suggests that one can expect an increase of roughly four-to-seven percent, at least in the case of criminal jurors’ post-service voting rates. Figure 3 shows that this effect, for previously infrequent voters participating as criminal jurors, is comparable to a semester-long civics course or a full stint on the student council.[16]

            The results presented here have significance beyond the jury itself. Our findings suggest that other meaningful deliberative events may also spark a participation effect. A recent national survey found that Americans deliberate in a variety of ways in the course of their public lives, from attending public meetings to taking part in online discussions.[17] Our research suggests that the strength of the participation effect at such deliberative events probably depends on the quality of the deliberation and the gravity or complexity of the issue under discussion. It is important to understand that this research suggests a broader link between citizenship and deliberation and deliberation’s effect on voting may well appear in other venues beyond the courthouse.

            Undoubtedly, the main finding is simply that, as the U.S. Supreme Court assumed in Powers v. Ohio (1991), there is wisdom in Tocqueville’s claim that juries serve as a school for democratic citizenship. The objective measures in the pilot Project and the National Jury Sample demonstrate that, consistent with a broad range of theories on democracy, different forms of civic engagement are linked. Moreover, they show that the link carries a participation effect, whereby an intense—preferably deliberative—experience in one area of civic engagement can spur increases in another.

            It is also striking that the complexity of a case augments the participation effect in criminal trials. In our view, this underscores the importance of personal engagement as jurors work through a deliberative process toward a complicated set of agreements among themselves. It is striking that the technical complexity of a civil case—a challenge for jurors often noted in the legal literature[18]—does not have this impact. Perhaps technical complexity dulls engagement, whereas deliberative complexity enhances it.

            The transformative experience of jury service even on hung juries further underscores the importance of engaging citizens. The hung jury is not a case of successful citizen participation as a group; it may be the most personally frustrating and infuriating experience of failure in civic engagement. The struggle to avoid failure, however, lends intensity to the deliberative process and forces citizens to engage in a bitter conflict of individual opinions. A jury does not fail to reach consensus unless participants cling tenaciously to diametrically opposed views. Is it surprising, then, that from the ashes of a failed jury there should sometimes emerge a new citizen? Only a vigorous exercise of one’s sense of civic duty would keep twelve strangers struggling for hours, often days, only to find consensus impossible.[19]

             The importance of deliberative complexity speaks directly to Tocqueville’s conviction that “the jury teaches every man [sic] not to recoil before the responsibility of his own actions and impresses him with that manly confidence without which no political virtue can exist. It invests each citizen with a kind of magistracy…”[20] In deliberation on complex criminal cases and what Tocqueville might call, in the parlance of his time, the “manly conviction” of hung juries, the participation effect shines brightest. The jury administers justice on the state’s behalf, while promoting deliberation and democratic citizenship along the way.

            This should serve as a warning about the limits of research that focuses strictly on non-governmental civic engagement and democracy.[21] Although voluntary engagement is often the focus of research, there is more to citizenship than voluntary engagement. Jury service stands as a mandatory, government sponsored political institution. It is the one area where ordinary citizens are required to exercise state power as individuals. Some overlook the jury because it is mandatory, yet, with millions of participants each year,[22] the jury may serve a more powerful role in promoting democracy and citizenship than any voluntary association.



Ackerman, Bruce, and James S. Fishkin. 2004. Deliberation Day. New Haven: Yale University Press.

Anderson, Kent, and Mark Nolan. 2004. Lay participation in the Japanese justice system: A few preliminary thoughts regarding the lay assessor system (saiban-in seido) from domestic and historical and international psychological perspectives. Vanderbilt Journal of Transnational Law, 37, 935-992.

Bertelsen, Kristy Lee. 1998. From specialized courts to specialized juries: Calling for professional juries in complex civil litigation. Suffolk Journal of Trial & Appellate Advocacy, 3.

Crosby, N., and Nethercutt, D. 2005. Citizens Juries: Creating a trustworthy voice of the people. In The Deliberative Democracy Handbook: Strategies For Effective Civic Engagement in the Twenty-First Century, ed. John Gastil and Peter Levine. San Francisco, CA: Jossey-Bass.

        Fishkin, James S., and Robert C. Luskin. 1999. Bringing deliberation to the democratic dialogue: The NIC and beyond. In The Poll With a Human Face: The National Issues Convention                         Experiment in Political Communication, ed. Maxwell McCombs and Amy Reynolds. Mahwah, NJ: Lawrence Erlbaum.

Freie, John F. 1997. The effects of campaign participation on political attitudes. Political Behavior, 19, 133-56.

Fung, Archon. 2005. Deliberation’s darker side: Six questions for Iris Marion Young and Jane Mansbridge. National Civic Review, 93(4), 47-54.

Gastil, John, and Peter Levine. (Eds.) 2005. The Deliberative Democracy Handbook: Strategies For Effective Civic Engagement in the Twenty-First Century. San Francisco, CA: Jossey Bass.

Gastil, John, Eugene P. Deess, and Phil Weiser. 2002. Civic awakening in the jury room: A test of the connection between jury deliberation and political participation. Journal of Politics, 64, 585-595.

Gastil, John, Eugene P. Deess, Phil Weiser, and Jordan Meade. 2008. Jury service and electoral participation: A test of the participation hypothesis. Journal of Politics, 70, 1-16.

Green, Donald P., and Alan S. Gerber. 2004. Get Out the Vote: How to Increase Voter Turnout. Washington, DC: Brookings Institution Press.

Hans, Valerie P., Paula L. Hannaford-Agor, Nicole L. Mott, and G. Thomas Munsterman. 2003. “The Hung Jury: The American Jury’s Insights and Contemporary Understanding.” Criminal Law Bulletin, 39, 33-50.

Jacobs, Lawrence R., Michael X. Delli Carpini, and Fay Lomax Cook. 2004. How Do Americans Deliberate? Presented at the Midwest Political Science Association, April 15-19, Chicago.

Mathews, David. 1994. Politics for People. Chicago: University of Illinois Press.

Pateman, Carole. 1970. Participation and Democratic Theory. Cambridge: Cambridge University Press.

Pateman, Carole. 1989. The Disorder of Women: Democracy, Feminism, and Political Theory. Cambridge: Polity Press.

Powers v. Ohio, 499 U.S. 400, 407 (1991).

Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.


Thomas, Reuben, J., and Daniel A. McFarland. 2006. Bowling Young II: How Youth Voluntary Associations Affect Voting in Early Adulthood. Available at

Tocqueville, Alexis de. 1961 [1835]. Democracy in America. New York: Schocken.

Warren, Mark E. 1992. Democratic theory and self-transformation. American Political Science Review, 86, 8-23.


Table 1: Linear regression measuring the effect of jury experience on infrequent voters in the National Jury Sample



Served on a
Criminal Trial





Served on a
Civil Jury



B (SE)









B (SE)






Jury Verdict



.043 (.03)









.020 (.02)






Hung Jury



.068 (.04)









-.111 (.17)









.022 (.04)









.000 (.05)






Guilty Plea



.019 (.04)













# of Charges



.013 (.01)













Pre-Jury Vote



.640 (.06)









.769 (.07)











































Note. * p < .10, ** p < .05, *** p < .01. The reference group for the Jury Verdict, Hung Jury, Alternate, and Guilty Plea dummy codes was Cancelled Trial (mistrials, dismissals, withdrawn cases, settling out of court, or waiving the right to a jury after the trial began). The other variables entered in the regression equations were the dummy variables representing county and oversample; they are omitted for economy of presentation. B denotes unstandardized regression coefficient, SE denotes standard error of estimate, and b denotes standardized coefficient.


Figure 1: Geographic distribution of counties contributing to the National Jury Sample.

Note. Dots on map are place in the principal cities within each county, including, from left-to-right: Seattle, Wash. (Thurston County); El Paso, Tex. (El Paso County), Boulder, Colo. (Boulder County); Omaha, Neb. (Douglas County); New Orleans, La. (Orleans Parish); Bryson City, N.C. (Swain County); Akron, Ohio (Summit County); and Fayetteville, N.C. (Cumberland County).


Figure 2: Three-step process for creating a dataset of jurors with complete voting histories.


Figure 3: Comparison of estimated effects from National Jury Sample and
high school civic education on the likelihood of voting in future elections.

Note. Civic educational effects based on estimates in Thomas and McFarland (2007).

Perry Deess is the Director of Institutional Research and Planning at the New Jersey Institute of Technology. He was a co-primary investigator on the National Science Foundation sponsored Jury and Democracy project. He is also an author of numerous articles on juries and democracy. Deess is a notorious doomsayer, who believes the jury system will emerge stronger in the future, despite also loving the Detroit Lions for their ability to confirm the grimmest expectations.

John Gastil is a professor in the Department of Communication at the University of Washington. He is author of Political Communication and Deliberation (Sage, 2008), co-editor of The Deliberative Democracy  Handbook (Jossey-Bass, 2005), and the author of By Popular Demand (U California, 2000) and Democracy in Small Groups (New Society, 1993). Gastil continues to have optimism that the jury system will survive, but this is the same person who each year expects the Detroit Lions football team to win the Super Bowl. (They were 0-16 last season.)

[1] This essay is adapted from an early draft of the forthcoming book, The jury and democracy: How jury deliberation promotes civic engagement and political participation (New York: Oxford University Press), co-authored by John Gastil, Perry Deess, Phil Weiser, and Cindy Simmons.

[2] Pateman (1970).

[3] Pateman (1989) and interview of Mansbridge presented in Fung (2005).

[4] Anderson and Nolan (2004), p. 943.

[5] Powers v. Ohio (1991).

[6] Mathews (1994).

[7] Crosby and Nethercutt (2005); Fishkin and Luskin (1999).

[8] Green and Gerber (2004).

[9] For additional detail on the National Jury Sample, see Gastil, Deess, Weiser, and Meade (2008).

[10] Gastil, Weiser, and Deess (2002).

[11] Hans, Hannaford-Agor, Mott, and Munsterman (2003).

[12] Such a view is contrary to the spirit of most deliberative theory, which yields ideas like Deliberation Day (Ackerman and Fishkin, 2004) that aim to stimulate discussion without the guarantee of resolution. See generally, Gastil and Levine (2005).

[13] Crosby and Nethercutt (2005); Warren (1992).

[14] Freie (1997), p. 134.

[15] Tocqueville (1961 [1835]), p. 337.

[16] Thomas and McFarland (2006).

[17] Jacobs, Delli Carpini, and Cook (2004).

[18] Bertelsen (1998), for example, believes only professionally trained juries are competent to handle complex civil litigation.

[19] This also speaks to overemphasis placed on consensus in deliberative theory.  See, for example, Karpowitz and Mansbridge (2005).

[20] Tocqueville (1961 [1835]), p. 337.

[21] The touchstone for modern work in this area is Putnam (2000).

[22] Based on figures from the 2006 Annenberg Public Policy Center Survey on the Judiciary, one could extrapolate that 17 million Americans have served on juries during the past five years. A 2007 survey by the National Center for State Courts estimates that a full third of U.S. citizens are likely to have served on a jury at some point in their life. The absence of a comprehensive, national jury service reporting requirements make more precise figures impossible.


Chris Dominic responds to Deess and Gastil

Chris Dominic is President and Senior Consultant at Tsongas Litigation Consulting. Founded in 1978, Tsongas is a national, full-service, trial consulting firm.

When Perry Dees and John Gastil concluded in “How Jury Service Makes Us Into Better Citizens,” “The jury may serve a more powerful role in promoting democracy and citizenship than any voluntary association.” I found myself nodding my head, involuntarily. My nodding was the product of a feeling of both gratitude and satisfaction.

The gratitude comes from the authors assembling a dataset of 13,327 jurors to work with. This massive data set gives confidence to their findings that voting behavior increases after certain types of jury deliberation experience between 4 and 6.8 percent. It is good to see that there is a stronger backing for the participation effect than merely citing Alexis de Tocqueville who, while brilliant, died over 150 years ago.

The satisfaction portion of my reaction comes from three factors. First, because the results are generally consistent with my own experience; second because of the accessibility of the article; and third because of how this may have a positive effect on our jury system and indeed, our society.

Consistent With Experience

As a trial consultant, I am close to the jury experience as a part of my everyday job. Trial consultants get familiar with the way juries behave from research, which in practicality means watching mock jury deliberations over and over. However, I have to admit that my appreciation for the power and importance of the jury came years ago after I was seated on a jury (and yes, I told the attorneys about my occupation).

There are far more voters in a year than there are jurors and that means that “democracy” to most people means representative democracy. The only time you are a part of direct democracy in America is when you serve as a juror and the only time you really feel a part of direct democracy in America is when you deliberate. The government gives you the power, you go in a room, deliberate, and when you come back with your decision (as long as there is no misconduct) it is sacrosanct. I still have a distinct memory from my jury service experience feeling that I was a part of something nearly pure and very difficult to corrupt. The decision didn’t have to go through committees and risk vetos–it’s final. Appeals, after all, are really not much more than a complaints department for one side to make about the judge in the case. The phrase I heard growing up, “government of the people, by the people, for the people,” finally made sense.

Article Accessibility

The authors do a masterful job of drawing us in with a dialogue that illustrates one of the most common and enduring debates about our justice system (p.1.):

“Tea sipper: “Why have juries?”

Coffee drinker: “They’re good for you.”

Tea sipper: “How do you know? What good comes from having twelve ordinary citizens deliberating when you could have one trained judge making informed choices?”

This dialogue reminds us that this is no sterile research question, it’s a question important to all of us. The style in which the article is written is as egalitarian as the point the authors ultimately make in the article.

Positive Effect on the System

If people vote more as a result in participating in jury deliberation then they may be more civically involved in many other places that we haven’t thought of or haven’t figured out how to measure. Clearly more research in this area will shed more light on this phenomenon. In the meantime this has implications for how we treat our courts and our jurors. Jury service is even more important to a thriving, functional, democracy than some of us thought. If we make it easier and more welcome for people to serve then it’s reasonable to presume that more will serve.

The list of ways in which this study’s findings can be applied is long. A few initial possibilities are:

1)    Include some of the aforementioned concepts in the curriculum for high schools when teaching the judicial branch of the “Three Branches of Government.”

2)    Low juror pay is an issue that comes up in many jurisdictions. The findings of this study could be used as support in a legislator’s impassioned speech to help the next jury pay raise bill pass.

3)    Judges can improve their juror orientation presentation with some of the findings from the study. I’ve seen orientations that are sadly, almost apologetic and ones that actually draw standing ovations. Rousing applause comes when jurors are reminded that they are about to be a part of something that is their duty, that they can have pride in, and is truly greater than themselves.

We frequently use the term “access to justice” when referring to an important American right. Dees and Gastil show us that from the perspective of the juror, access to justice can also be “access to democracy.”


Response of Douglas A. Green to:  How Jury Service Makes Us Into Better Citizens

Douglas Green, Ph.D. has been a trial consultant for 25 years focusing on complex civil litigation. He is a past president of the ASTC and runs a national consulting practice based in Louisiana. 

Deess and Gastil present us with a very brief essay giving a view inside their upcoming book, The Jury and democracy: How jury deliberation promotes civic engagement and political participation.  I found this introduction interesting enough to keep my eye out for the full text.  But, I have to say that I have mixed reactions to the general proposition of the work. 

I am a firm believer in the jury system.  Juries are not just essential to democracy; juries are the essence of democracy.  The most direct participation in true democracy the vast majority of Americans will ever have is participating on a jury.  I feel a certain frustration that there exists a need to somehow justify the jury system in America and yet I recognize, as Judge William Young so emphatically stated at the ASTC conference in Philadelphia, that the American jury is dying.  So, the other reaction I have to this work is satisfaction that there exists the kind of momentum behind preservation of the jury system that this ambitious project could be completed. 

Readers of The Jury Expert come from diverse backgrounds so I feel a strong need to comment on the true scope of this research.  In “the real world” where trial consultant and lawyers live, sample size is a major consideration.  A few months ago I completed a major project for a client with a very large potential exposure in a civil case.  The study included about 150 mock jurors over the course of several days of research.  I believe it is safe to say that this is on the large size for most in-depth jury studies.  The initial sample in the work of Deess and Gastil included 13,237 individuals.  Moreover, the data was collected at eight geographically distant locations, including my home town, New Orleans.  Having spent many long days in the halls of our district courts, I don’t envy the researchers their experiences.  It was, no doubt, very hard work – the afterhours benefits of our fair city notwithstanding.  Make no mistake here – this was a massive effort that could have not have been accomplished without a great deal of dedication and major funding. 

Now, I have to agree with the authors on a couple of concerns.  First, as they indicate, “the democratic act of voting . . . is not the best way to understand the robust and varied nature of civic engagement.”  But, I understand the use of this measure.  It is an objective measure that should be highly reliable.  And since the authors knew they would be looking for small effects, reliability of all of the measures in the study was very important.  But, it was a hard test of the theory.  Voting behavior has shown strong resistance to change, as cited by the authors.  So, even a modest change in voting attributed to jury experience would be impressive.  Still, I find voting behavior to be unsatisfactory as a main dependent variable in a study so massive in scope.  I find myself wanting more even though I respect the thought that went into the design. 

My experience is certainly that jury service changes people.  But, I would have bet with the tea sipper on this one.  The changes are too complex to measure in a large, statistically powerful study.  These changes have to do with issues like self-esteem and empowerment that are very hard to measure.  Many people who sit on juries have never participated in a decision-making process with such importance attached.  They may find within themselves something that they did not know existed before.  This could express itself in many different ways.  Perhaps they agree to lead a discussion in their bible study group.  Or maybe they serve on a committee in a club or professional organization.  For other people, jury service is no different than what they do every day at the office, so no change is to be expected.  So, in looking at this work, I think it was a risky venture and I am impressed with the results.

Now, what does all of this mean for the practitioner?  Trial consultants and trial lawyers participate in the democratic process every day as advocates for their clients.  Sometimes I think that we take our participation for granted.  I think that we sometimes lose sight of the fact that the jurors sitting in the box are not regular participants in the process.  Are we consciously aware, in the middle of the stress and turmoil of the trial, that at least some of the jurors are going through a life altering experience?  If one stops to think about the results of this study, how would it change your behavior in court? 

Here are a few suggestions.  Respect the jury.  Engage jurors in the case and with the witnesses.  Make sure that you are communicating with the jurors not just talking at them.  Do not waste jurors’ time.  Any time the jury is excluded from the proceedings, they feel is wasted.  This includes side bars, out of the presence hearing, and even being late to the courtroom.  Keep your examinations focused, organized and concise.  This demonstrates not only respect for the jurors’ time, but also helps improve comprehension.  Finally, treat jurors as the most important people in the courtroom, not like those unfortunate souls who could not find a way out of jury service. 



Perry Deess responds to Chris Dominic and Doug Green

I offer my thanks to the respondents for their thoughtful comments and, more importantly, for their dedication to the jury system. Reading comments from those working in the heart of the judicial process gives hope and inspiration to the laborious process of research. This entire project would not have been possible without the dedicated assistance of many jury system advocates at all levels. They opened archives when roadblocks appeared and convinced reluctant colleagues to allow us a sometimes unprecedented level of access to jurors and jury records. We cannot adequately convey our gratitude to many of these unknown champions.

The suggestions Chris Dominic offers for promoting and supporting the jury process are well taken and we wholly agree. It is critical to the health of the democratic process that citizens, legislators, and legal practitioners understand and support the jury system and the critical role it plays in democracy.  For too many years the dedicated work of jurors has been overlooked and taken for granted. Their commitment and sacrifice should not go unacknowledged and largely uncompensated.

The comments of Douglas A. Green provide an excellent opportunity to expand on the research described in the article. He correctly observes that demonstrating a link between juries and voting, although useful, is ultimately unsatisfying. The ‘Tea Sipper’ and ‘Coffee Drinker’ agreed and, with substantial support from the National Science Foundation, they conducted a multi-wave ‘panel’ survey of jurors in the Seattle area to understand how the act of jury service changed their conceptions and habits of citizenship.

The survey design included a battery of questions about volunteerism, charitable contributions, interest in political news, faith in the judicial and political system, and belief in the empowerment of citizens. The surveys were administered to the same people at three intervals: 1) prior to the jury selection process; 2) two weeks after the completion of jury service; 3) several months after the completion of jury service. An additional series of questions about the subjective experience of jury deliberation was added to the second wave of the survey. Response rates were over 70%, and our study of county and municipal courts yielded a total of 2,872 jurors who responded to all three waves, including over 1,000 who ultimately served as empanelled jurors. This is the data set we are using to address some of the more subtle questions about citizenship, democracy, and the role of juries, and interested readers can see our first published findings in Policy Studies and Human Communication Research at

The jury-voting test described in this article is, what one might call, the hard ‘proof’ of a democratic effect. By comparing jury and voting records there is no question about potential bias or error in the research design. The survey studies, in contrast, give a more comprehensive view of the impact jury service has on citizenship. We will integrate these findings and other observations next year in our book from Oxford University Press, The Jury and Democracy: How jury deliberation promotes civic engagement and political participation.  –Perry Deess


Citation for this article: The Jury Expert, 2009, 21(3), 51-69.