Beginning in 2010, American political scientists, led by Colin Elman and Arthur Lupia, organized the Data Access and Research Transparency (DA-RT) initiative to make visible both the logic of inquiry and the empirical foundation of research, and to facilitate the testing and replication of results. The essential idea seemed reasonable and, following several years of discussion among APSA leaders, a joint statement (the Journal Editors’ Transparency Statement, known as JETS) was signed by editors of many of the discipline’s top journals, including the American Political Science Review, signaling their commitment to making greater openness and transparency a valued and central part of political science publications. Footnote 1 Other top journals, including The European Journal of Political Research, Perspectives on Politics, Political Psychology, and World Politics, did not sign the statement. Footnote 2 APSA’s leaders supported DA-RT and, despite increasing expressions of both enthusiasm and concern in 2015 and panel discussions at the APSA meetings in 2016 and 2017, DA-RT is now being widely implemented by JETS signatories. Footnote 3
Disciplinary uneasiness about the current DA-RT proposals centers on two of its provisions captured in the JETS implementing DA-RT in January 2016. The first provision requires authors to make their data available at the time of publication. It authorizes editors to refuse to review manuscripts when authors fail to satisfy this requirement. The second compels authors to delineate clearly the analytic procedures and, where possible, provide access to all relevant analytic materials. Footnote 4
Much of the extensive questioning and pushback on the rapid implementation of DA-RT has focused on these aspects of DA-RT and, in particular, DA-RT’s failure to distinguish and allow for critical differences between quantitative and qualitative data/methods. I take it as a given that some questions are better answered using qualitative methods and data, either on their own or in concert with quantitative methods. Unless one rejects this supposition, the discipline thus has no option but to develop a set of data transparency norms and practices appropriate for qualitative data. My goal here is to aid in developing such practices. It is both necessary and possible to develop such procedures, as witness by the revised guidelines issued in fall 2017 by the APSR, largely in response to public deliberations concerning qualitative transparency and the implementation of DA-RT for qualitative research. To encourage similar modifications by other journals, I would like to raise several concerns I became aware of only as my own scholarly work shifted from econometric modeling and survey research to narrative interviews with interlocutors in precarious situations, facing inherent moral dilemmas. Footnote 5 These concerns focus on four inter-related areas that—I believe—necessitate different rules for qualitative data: (1) space constraints that hinder full journal presentation of the analysis of qualitative data; (2) costs of data collection and the right of first usage; (3) ethical concerns about protecting human subjects, and the time needed to prepare such data before publicly sharing them; and (4) a potential chilling effect of DA-RT on certain types of research topics. I discuss each of these concerns using illustrations from my own work since a survey of even one person’s experience demonstrates why DA-RT needs additional safeguards for qualitative data and methods. Footnote 6 I conclude by arguing that implementing the current version of the DA-RT without developing workable procedures for qualitative work could hinder political science’s ability to address important political questions. Thus DA-RT should be modified to address the distinct needs of qualitative researchers.
Constraints on Space and Time
When I wrote The Heart of Altruism, Footnote 7 I came out of a rational choice background and wanted to examine the self-interest paradigm that underlines the rational actor approach. I organized my research around in-depth interviews designed to answer questions from a 14-page survey questionnaire I had developed, based on reading the literature on altruism in fields ranging from evolutionary biology and economics to political science and psychology. After a brief introduction, literature review, and description of research data and methodology, the book moved quickly to present full interviews with archetypes of the people I interviewed. Analysis began with entrepreneurs (quintessential self-interested, rational actors who take an idea and harness it for profit) and philanthropists (who give away much of their money but retain some for themselves). It then considered Carnegie Hero Commission winners (who risk their lives for strangers) and ended with rescuers of Jews during the Holocaust (whose efforts to save strangers risked not only their own lives but also those of their families). After presenting one chapter containing interview material from one of each of these archetypes, I concluded with a chapter analyzing all the interviews to test hypotheses about altruism/self-interest. Footnote 8
The Heart of Altruism followed traditional norms for post-behavioral social science: literature review, hypotheses formation, discussion of the 14-page survey instrument, methods, and data. In the field, I quickly realized, however, how awkward—even rude and impersonal—it felt to dive immediately into the survey questions when dealing with actual human beings. So I shifted and began asking: “Could we just get to know each other a bit by your telling me about yourself and, in particular, how you came to be an entrepreneur (philanthropist, etc.)?” Many of my survey questions were answered in this free-flowing narrative. When I began the project, in my naïveté, I did not know narrative interviewing even was a technique. I always circled back to re-ask questions, often more directly or in a slightly different form, in interviews that stretched over many days, even weeks, and years in some instances. I sent the transcribed interviews to speakers for approval/corrections. Only then did I analyze the interviews. Footnote 9
Even though my research plan followed the post-behavioral scientific model, I ended up re-inventing my methodology as I went along, as I knew of no other work on narrative interpretation in political science at the time. The bulk of the book thus consisted of stories—lengthy stories. My editor at Princeton, Mary Murrell, took her role seriously and gave me extensive substantive feedback. Mary asked me to cut the interviews and to do what was typical at the time: summarize most of the information about the person and highlight a few of the speaker’s critical remarks, especially those thoughts that spoke directly to the research design. I was not the kind of person who argues with an editor (although I suspect later editors might disagree), but we did wrangle politely about the cuts. We agreed to abide by the copy editor’s suggestions on this question. As luck would have it, the copy editor came down on my side of the debate: “Don’t touch the stories. They’re gold.”
She was right. I am convinced that whatever impact my work had on discussions of the self-interest model and altruism was due to the stories. Readers already convinced of the ubiquity of self-interest as an explanatory concept in social science could slide right by my “rigorous” analysis. But the stories—with their human interest—slowed readers down enough to let doubt creep in, and for some to change their mind. This conclusion, of course, confirms what Kahneman and Tversky Footnote 10 found in their experimental work: vivid examples often carry more weight than do statistical data.
My experience highlights another important difference between qualitative and quantitative work. Space constraints create huge problems for those looking to publish work that relies on in-depth analysis of qualitative data. Where do we get the space to produce the full examination of a hypothesis? This difficulty is especially acute in journals, which usually limit manuscripts to 8.000–12,000 words. I had to cut so many quotes from one journal article due to space constraints that the final argument feels—to me—disjointed and incomplete, an embarrassing, bastardized version of the original manuscript. Journal editors are aware of this issue and are trying to allow for the special needs of qualitative data. But can posting supporting data on a website accurately convey the intricate analysis through which an argument is constructed? Will readers pay sufficient attention to a web appendix, when it is not an integral part of the analysis itself?
My initial experience with qualitative data was serendipitous on many levels. I already had tenure. I had a flexible book publisher who cared about quality work. The altruism book won an APSA prize, and sales, reviews, and publicity were good. The press thus gave me greater leeway with the next book, The Hand of Compassion: Stories of Moral Choice during the Holocaust (2004). Footnote 11 That book addressed how real people make moral choices during difficult political times, through the stories of five people who rescued Jews. Again, the stories were central to the research. One of the anonymous reviewers said he had planned to skim the stories due to time constraints but found them too compelling to put down. This time the press exerted no pressure on me to summarize the stories and each story was a separate chapter.
I had the luxury of publishing books. In many departments, however, articles are key to tenure and promotion. Given the reasonable space limits in journals and the time involved in writing a book, untenured scholars analyzing qualitative data thus are at a disadvantage. Further, expectation that qualitative scholars post full field notes or interview transcripts is unrealistic. The time to fully clean and digitize notes is exorbitant. It will mean that only tenured faculty members, with time and resources to hire professional transcribers, will do qualitative work. Beyond that, will researchers take the time to review material with interview subjects, as I did? This critical issue demands serious reflection and setting standards that are fair, standards that do not put onerous burdens on qualitative researchers. So far, DA-RT has not done this.
It is important to remember that space constraints affect not only professional careers; they also may impact our ability to reach certain substantive findings. Insofar as DA-RT makes it harder to publish qualitative work in journals, such insights might not be published in journals. DA-RT does not block researchers from using qualitative techniques, but if it complicates scholars’ ability to publish their results in mainstream journals, the discipline as a whole is the poorer.
Right of First Usage
Beyond this, a scholar who invests the time to create and conduct interviews should have the right to analyze these data fully before having to release them publicly. Footnote 12 Sometimes this right also extends to the person interviewed, or to their family. For example, it was 1988 when 1 conducted my first interviews with Tony van Renterghem, a leader of the Dutch resistance during World War II and a prominent cinematographer and technical adviser in the film industry after World War II. I interacted with Tony and his wife in formal interviews and informal conversations over the next fourteen years. With their approval, I had published and shared much of Tony’s story before he died. But I have never made the filmed interviews public since Tony and his wife had discussed wanting to have the opportunity to do a movie about his life. I felt they had first right here since, as Tony’s wife expressed it, “We all heard his stories many times but you are the only one who got them on tape.” Footnote 13 So some of the “data” have been shared, but not all of it. What is critical here is that my decision about how the data got shared, and with whom, was informed by a personal and full appreciation for the complexities of the situation. My choice was not one decided according to rigid rules or the discretion of editors unfamiliar with the particulars of the case, on the basis of their notions of what constituted an acceptable level of research transparency.
Ethical Obligation 1: Sharing Data and the Costs of Cleaning It of Identifying or Personal Information
Data should only be shared ethically. While working on my books, several scholars asked about my data and the survey that underpinned my narrative interviews. I always provided them to anyone who asked for the surveys or the data. Eventually my data were deposited in an archive at the UCI Ethics Center and anyone—scholar or member of the public—has free access at www.ethicscenter.uci.edu. In doing this, however, I learned several key things relevant for any implementation of the principles inherent in DA-RT.
Sharing is good as a general moral principle and is a practical, if not essential, way to advance science. I certainly want future scholars to address the questions of moral choice that have been central to my research. I hope new perspectives and methods will allow them to address the holes and errors in my research, providing better answers to these abiding questions. To share the data on which past work was based is essential. Much of the data I collected are not reproducible; thus making them available to others is essential. Many of the people I interviewed—especially rescuers of Jews—gave me their stories as they neared the end of their lives. It was an honor and a privilege to be able to spend time with them. I treat their interviews as a gift, one that I need to bequeath to posterity. Future researchers might well draw new insights based on things I missed or did not pick up on. This is how science advances.
If I believe in sharing data for ethical and scientific reasons, why do I then have reservations about DA-RT? Essentially, I suspect DA-RT, generated out of good impulses, simply moved too fast. It was adopted by some journal editors before the profession as a whole had a chance to consider whether the principles inherent in DA-RT rules—at least as initially proposed—were practical. To be effectively implemented, the initiative should have paid greater attention to the actual experiences of scholars working on qualitative data, especially data on what we call “human subjects” from vulnerable populations, discussing sensitive topics.
Beyond posting my data on the UCI Ethics Center website, I also offered my data from the rescuer project—begun in 1988 but completed with the publication of the last book only in 2012—to the Holocaust Museum, clearly a logical repository. Footnote 14 The Museum wanted the data, but only if I provided the interview tapes. I could not do this, because the interview tapes contain personal conversations that could violate the privacy of others, both the speakers’ and their loved ones. People often said things like “I tell you this in confidence . . . ” or later asked me to “omit the comment about my ex-son-in-law” from anything published. I wish I had the time and resources to go through these tapes and remove all stories that are personal or hurtful but time is a serious constraint here. I can analyze data with these comments included but I cannot release them to others without violating the privacy of my research subjects.
Posting data publicly imposes extra demands. I cannot make interview materials available unless they have been “cleaned” of anything that will identify the subject or cause them embarrassment or harm. This is not the kind of data-cleaning work that I can hire someone to do. First, the stakes are higher than measurement error, as important as that is. Second, knowledge of the particular speaker, the memory of when someone broke down and cried or seemed upset or disconcerted, all the nuance and particularities of an interview, require interviewers to draw on their personal, intimate knowledge, to make the decision about whether such things are appropriate to share. These decisions cannot be “farmed out” to a graduate assistant. This makes the preparation of data for sharing exceptionally time consuming.
Ethical Imperative 2: Protection of Human Subjects
The issue of cleaning data before posting in order to protect the privacy of interview subjects is just one of the ethical dilemmas that qualitative researchers face. Even thornier is the imperative to protect vulnerable populations when they share their stories with researchers. I am currently interviewing Syrian refugees, and am just beginning to grapple with the complex issues affecting this vulnerable population, so my insight is incomplete. But already I can see that it is ethically untenable to ask—let alone demand—that scholars post such data without sufficient safeguards to protect vulnerable people. We have a special obligation to deserve and respect the trust people give us when they open up their lives to us, especially when discussing emotionally-charged or politically sensitive topics.
Consider one woman I interviewed for A Darkling Plain: Stories of Conflict and Humanity during War Footnote 15 . My speaker, the wife of a high official in the Shah’s Iran, told me her husband could not stay away from Iran. He loves his country, keeps visiting under an assumed name, quietly going in and out at personal risk. Although the woman is extremely well educated (a PhD) and had twice approved the transcripts I sent her, I feared her openness might endanger her husband’s safety. So with her permission, I modified identifying details, giving her more children, altering her professional field, town, etc. Wondering if these modifications would mask her identity sufficiently, I gave her transcript to her grandson, who happened to be a summer intern at the Ethics Center and someone especially interested in social science.
“Would you like to read a transcript from a forthcoming book, dealing with a woman from Iran?” I asked him.
“Sure.” He was enthusiastic. I sent him the transcript and the next time we met, he told me how much he had enjoyed reading her story. “She’s fabulous,” he said.
“Would you like to meet her? She lives nearby.”
“Awesome! That would be so great.”
“It’s your grandmother.” The kid was astonished, and I relaxed, confidant his grandmother’s identity was sufficiently protected.
I may be overly cautious, and I confess I find tampering with data, even in this instance to protect a speaker’s identity, feels like falsifying data. I spend hours going through the interviews, listening to the tapes or viewing the filmed interviews, trying to capture the sound of the voice, the inflections, weeding out the pauses that mean nothing and keeping the pauses that might signify. So a fully-edited transcript already represents hours—sometimes weeks or even months—of effort on my part. Taking the extra time to share a transcript with the speaker to ensure that their privacy is protected, however, is critical. That also takes extra time.
Ethical Imperative 3: Protecting the Privacy of Sources
Finally, some speakers do not want their data shared publicly. They specifically request that their interviews be used only as background. Even though they give informed consent, they specify that I use their interview only to interpret and understand the phenomenon that concerns me. They want their privacy protected. Sometimes these requests are formal; more often they are vaguely-worded requests, sometimes after the interview has ended and we are saying goodbye. They agree to have quotes used in the analysis but ask that their identity remain masked. At other times they allow me to use quotes anonymously. Sometimes interviewees are willing to have their stories summarized but do not want to be quoted. Occasionally speakers are famous people or individuals who have special insight that mean their names and opinions would carry weight. Yet they choose to remain private, anonymous, or off-the-record, and I must respect this.
For example, Albert Hirschman Footnote 16 spoke about his experience smuggling people out of Europe during WWII. (His actions are well known so I am not betraying any confidence here.) Albert asked that the interview be treated only as background. Further, he suggested that we conduct the interview over lunch at the Institute for Advanced Study, a setting I quickly realized was quite noisy and not conducive for quality recording. Even though I transcribed the interview that afternoon, I still had trouble retrieving from all the background noise the precise words on the tape. I know sometimes things do occur by chance but I concluded that in this instance, Albert’s choice of place was symbolic of his ambivalence about speaking about traumatic events in which he lost beloved family members and friends. I respected Albert’s wishes, never used any quotes from his interviews and never published the interview. Did I do the right thing? I recognize that there is historic value in Hirschman’s interview. Is it right to share Albert’s words and thoughts only through the filter of my analysis? Reluctantly, I concluded that my first responsibility is to the speaker, not the broader research community, or even to history.
As I read the current version of the DA-RT initiative, it seems insufficiently sensitive to the kind of important ethical issues that concern those of us who deal with human beings speaking about private matters (how Margot slept with an Amsterdam Gestapo commander to get information for the Resistance), of painful topics that are ethically sensitivity (a Dutchman’s admission that he was hired not just to hunt, but also to kill, Nazis after the war), painful personal memories (Margot’s story of how her fiancé was beaten to death), or touching on concerns of safety with vulnerable populations (not reporting someone involved in human trafficking). Footnote 17
Why Protecting Subjects Is Essential: An Illustration
In fact, at- risk populations are all around us. We as researchers might not initially recognize how vulnerable the population is or at least feels itself to be. I faced this situation when I was part of an NSF ADVANCE grant to conduct interviews with faculty women on gender equality within academia. As I began the interviews, I came to realize how sensitive the interviews were, how difficult it was—even for top tenured professors or college presidents—to speak openly about gender discrimination in their lives, and how vulnerable many of them still felt. Several would not meet me on campus, insisting we meet where no one would recognize them and guess they were talking with me about gender discrimination at UCI. Initially, I was skeptical of the need for caution, but I respected their wishes.
I soon learned first-hand, and painfully, how sensitive these data truly were. The ADVANCE grant to UCI was one of the first funded by the NSF, and right-wing groups had threatened to sue for the data. The initial protocol thus required me to tape over interviews (destroy data, effectively) and to keep all transcripts in one computer in a locked office, with access restricted to the Monroe research team. There was no problem with this, but trouble came elsewhere. As the interviews proceeded, the higher-ups on the NSF ADVANCE program on UCI’s campus realized my findings were not depicting the UCI administration in a favorable light. The woman who served as co-PI for the grant and who had herself moved into a high administrative post—cynically, I believe, not unrelated to her work on the ADVANCE—asked me to provide the description of my work for the NSF by breaking down the responses by school and rank. I told her I could not do that without revealing the identities of some of the speakers since, in some cases, there was only one female full professor in a School, such as computer science or engineering. She sent me an e-mail, chastising me for refusing to report data in a “full and scientific” manner, telling me I was jeopardizing all the NSF funding on campus by what she dubbed my obstinate refusal to comply with accepted scientific norms.
I opened the particular e-mail as I was sitting on the floor in the basement of the hospital where my mother had just been wheeled into exploratory surgery, to detect whether she had serious late-stage stomach cancer. As I opened the e-mail it is fair to say I was not at my best, and felt extremely vulnerable, away from any support groups, and fearing I was losing my mother. Despite this vulnerability, there was no question in my mind, no hesitation at all. I was not going to compromise the anonymity of my sources. Fortunately, the Institutional Review Board (IRB) protocol I had submitted, on which my antagonist had signed off, stated that no one but members of the Monroe research team would ever have access to these data. The IRB thus protected me from this particular undue pressure. This unpleasant personal experience, however, and the fact that my last year of funding for this grant was eventually cut, reinforced my sense that we have to protect scholars from anything so threatening. A journal editor’s request for full data before publishing a manuscript can easily feel coercive. It is wrong to pressure researchers to share data in order to be published when it might endanger research subjects in any way. While this does not seem to be DA-RT’s intention, and the initiative is careful to stress the importance of IRB protocols, strong measures to protect authors from the pressures of the passion for transparency when other important values are a stake should be a prerequisite before we adopt DA-RT discipline-wide.
Our Ability to Get at Important Political Questions
I worry that DA-RT may discourage young scholars in particular from tackling big topics. As a practitioner, I may disagree with some scholars on the value of qualitative work and what its place should be in the discipline. But even if colleagues do not share my belief that qualitative data better capture nuances in human behavior than do quantitative data in many instances, I think we all can agree that qualitative data and work do add different insights that improve our understanding of political life. Thus we all should be concerned about how DA-RT may impede the collection of qualitative data, its analysis and its presentation.
Gathering first-hand information on the experience of illegal refugees, political dissidents or victims of human trafficking, or conducting interviews with people in the inner cities about police violence or prejudice in the workplace—to cite just a few examples—all help to study big, important political questions. Footnote 18 Research of this nature is sensitive because it involves speaking with vulnerable populations who need to be sure they can trust the researcher. Individuals, activists, and government officials need to know they can speak off the record in confidence to researchers. Understanding the specific context and respecting the particular demands of certain types of research is key in establishing the trust necessary to obtain honest answers about highly charged questions. To establish such trust may sometimes require taking a stance that feels unscientific and political, for example by sympathizing with activists fighting state repression or by talking to citizens and government opponents when an authoritarian government discourages it, or refraining from taking notes or taping interviews. This stance may put researchers in situations where morality and human decency put them at odds with DA-RT’s concern for maximal transparency, but it is these very acts that help us obtain information that is not readily available by any other means. Footnote 19 Transparency cannot trump the search for answers about important questions.
Scholars sometimes need to go to extraordinary lengths to protect privacy. Interviews often cannot be taped since simply turning on a recorder for many people will invoke fears about coercive security. Footnote 20 This produces interviews that often are unrecorded and unverifiable but still legitimate research data. Further difficulties on transparency and replicability arise when researchers try to get interview material out of countries with repressive regimes. Notes often need to be scribbled in the margins of magazines to elude detection if bags are searched when a scholar leaves a country. Female researchers sometimes hide flash drives in tampon boxes. Many scholars wipe their hard drives clean and protect photos, tape recordings and other interview notes and material with passwords. In this day and age, however, storing data on a cloud outside the country being investigated offers no iron-clad guarantee of security. Qualitative data gathered under authoritarian or anarchic conditions do not easily lend themselves to the simple principles DA-RT endorses. If we are to make use of such data, the best practices to minimize risk while allowing for sharing need to be developed.
Conclusion
I have tried to suggest my support for the essential values behind the DA-RT even as I conclude that DA-RT must be implemented with greater care to avoid crowding out qualitative research, imposing unacceptable costs on its practitioners, or creating conflicts between getting published and protecting human subjects in precarious or sensitive political situations. DA-RT should do more than promulgate abstract principles; it also should develop protocols to assure that pursuit of transparency does not impinge upon the collection and reasonable use of qualitative data and its important contributions to political science. Among the issues that need to be addressed are: (1) the protection of human subjects; (2) provision of adequate journal space to publish complete analysis of qualitative data; (3) attention to ethical issues of cleaning data before sharing it; and (4) protecting the anonymity of interviewees, especially those from vulnerable populations.
A piece I found online, ironically—given the concerns for replicability—unsigned by any person and identified only by Twitter: @PolSciReplicate), dated November 7, 2015, dismisses the impact of the chilling effect of DA-RT by journal editors, noting: “The guidelines have inbuilt exemptions so that no journal editor will have reason to reject a paper that uses sensitive data.” To test this, I wrote to a few top journals, describing a manuscript I wanted to submit but saying I could not make all my data available because some of the subjects had asked to keep their interviews private. The AJPS was the most extreme in refusing flat-out to even consider the manuscript. The APSR response was ambiguous but seemed to indicate they would not consider the manuscript, a position I assume has changed now that new guidelines are in place, as of fall 2017. Footnote 21 Perpectives on Politics and Political Behavior were the most open and willing to be flexible in response to the ethical considerations involved.
So DA-RT’s impact is critical. DA-RT potentially matters for what is deemed legitimate “scientific” work, and for the career trajectories of scholars. Can young scholars, given the increasing pressures to produce for tenure, take risks by answering big questions drawn from interviews they conduct themselves, or should they work only with institutionalized data infrastructures? Data with human subjects who can be identified in any way should not be required to be shared or made public unless and until the privacy of the subject can be fully protected. Even a request from a journal editor to a scholar—untenured or not—can have a chilling effect, and I fear such a policy at top political science journals will deter graduate students from pursuing qualitative work on critical topics.
All the issues discussed here relate directly to our ability to do good political science. Paradoxically, the notion of science that underlies DA-RT’s more rigid interpretations seems at odds with what is a fundamental premise of careful scientific work: the notion that “it could be otherwise.” Footnote 22 The creative uncertainty of good science thrives on more than routine, more than elegant models systematically applied in a pre-designed manner. It must observe and capture the unexpected. Generating large amounts of data—transparently or not—still requires interpretation by human beings who are sensitive and open to allowing for the unanticipated, the unforeseen. I do not want political science to adopt a rigid and—I would argue—a narrow conceptualization of the scientific method in the search for reproducibility and replication. Doing so will lessen the richness of our discipline’s analyses. It will inhibit the kind of innovative scholarship whose value arises, as Frank Knight noted, “out of the inherent, absolute unpredictability of things, out of the sheer, brute fact that the results of human activity cannot be anticipated.” Footnote 23
So while I believe strongly in sharing data and have always made mine available, and while I applaud the general sentiment of making clear the foundations of research, I nonetheless strongly oppose the rush to adopt and institutionalize DA-RT without more meticulous consideration of its impact on different forms of work and without the working out and adopting of standards to protect the diversity of the discipline. The DA-RT initiative, while well intended, was hastily adopted without fully considering the impact on data-intensive qualitative work. Before it can be embraced by the discipline as a whole, DA-RT is in need of serious reconsideration because of (1) ethical concerns about human subjects and their privacy, (2) the needs of scholars to have first right to data they have collected at great cost in time, and (3) because an implementation of DA-RT that is based on a narrow conceptualization of the proper scientific approach in political science can cast a chilling effect on our discipline’s ability to tackle important substantive issues. I encourage editors to rethink and modify their position on DA-RT, adopting the more flexible policies such as those recently put into place at the APSR.
