Response: More on ;lies and damn lies'

Rich Braunstein and William Anderson

On July 28, distinguished professors Carole Goldberg and Kevin Washburn wrote an opinion piece in Indian Country Today [Vol. 28, Iss. 8] stating, ;'It would be a mistake to accept the conclusions of the South Dakota study,'' which is set to be published this winter in American Indian Culture and Research Journal.

We feel a response to that Perspective is needed to correct some of what Goldberg and Washburn wrote. Also, we, as two of the four co-authors of that study, would like to further challenge the community of interested academics, advocates and those generally concerned about American Indian criminal justice to continue to pursue reliable and context-sensitive research in this area.

First, some clarifications are in order. To professors Goldberg and Washburn's claim that ''the South Dakota report asserts that victimization surveys are unreliable,'' it is important to note that this is not what we wrote in our forthcoming AICRJ article. In our study, we wrote that we were concerned about the reliability of the Bureau of Justice Statistics methods, which rely almost exclusively on the National Crime Victimization Survey. We think it is important to note that this is not the same as asserting victimization surveys are unreliable as Goldberg and Washburn claimed we stated.

Further, our concern for the BJS reliance on NCVS comes from a long line of published research critical of this approach. In fact, a reviewer of our AICRJ article suggested that our work should not be published without articulating this concern. Upon further examination of that reviewer's position, we learned that many criminal justice publications, including well-established introductory textbooks and more detailed journal articles, share the concern that NCVS is subject to over reporting, under reporting and a wide array of sampling errors.

For our research, the greatest concern for sampling error is rooted in the fact that BJS does not seem to include any respondents from reservation communities in its two reports on American Indian crime. While this shortfall is being addressed by Julie Abril and others who are focusing attention on gathering reservation-specific data, analysis and academic research on Native crime victims have relied for years almost exclusively on the BJS' more limited sample of Native victims in urban areas, where the majority - nearly 75 percent - of Native respondents lives and works.

The second clarification is concerned with the Goldberg and Washburn claim that our research ''refuses to supplement the police data with data from victims of crime.'' This is simply not the case. Our argument in the AICRJ article is that BJS' reliance on NCVS data in all but homicide cases where there are no victims to survey is problematic. The fact that the BJS approach does not sample or report on reservation-based crime risks misleading researchers who use the data and policymakers who rely on it. Ultimately, a mixed-methods design would be the most compelling of all approaches to studying this issue, a point we will elaborate on below.

Our AICRJ piece took one approach - a case study of South Dakota, which has reservations with high incidences of crime, a high percentage of American Indian population, and a landscape that reflects, but does not represent, Indian country in the northern Plains. We brought to bear a different type of data than was used in the BJS research to more rigorously assess the notion that Native crime victims are disproportionately victimized by non-Native defendants. We found in South Dakota that this is not the case. In fact, we find that Native crime victims are most frequently targeted by Native defendants, a conclusion that contradicts much of what the BJS collection has told the academy and policymakers for nearly a decade.

Our research was not intended to reject the use of victim data or any other research designs. We have no qualm with the BJS research, or use of victim surveys, if it were made more clear that the BJS' data collection approach is but one of many ways to understand how Native Americans become crime victims. Unfortunately, such transparency is limited and, more importantly, public policy regarding crime in tribal territories is being crafted based on the findings of what we believe to be an incorrect assessment of crime in our home state and, perhaps, around the country.

Of course, we are not so bold as to suggest that our approach is completely right. In fact, we know that the federal data that our AICRJ piece uses probably under report crimes in general in South Dakota's tribal areas. We also are aware that this systematic under reporting of crime likely owes to a culture of longstanding mutual distrust between Native crime victims, their tribes, and state, local and federal authorities in the state. We believe, however, that our case-focused approach is useful and that it sheds light on a research question that remains unanswered by the BJS' existing data collection.

Further, we recognize that American Indian crime research requires complex and sensitive research designs, multiple approaches, and a reliance on both qualitative and quantitative data. In many ways, such a research agenda is exactly the opposite of a deterministic, federal government-driven quantitative approach produced by BJS. That approach seems to advance a ''one-size-fits-all'' research design on cultures that are diverse, geographically spread, and that have unique criminal justice concerns that cannot hope to be uncovered with an instrument as blunt as BJS' series of surveys from 2000 and after.

Our findings in the AICRJ article support the call for methodological pluralism, or a diversity of approaches understanding American Indian crime victimization. Our research approach argues for the importance of understanding how crimes happen in individual tribal territories. Taken to its logical conclusion, our research is challenging scholars in this area - and the policymakers who rely on that scholarship - to recognize that the research approaches that are taken to understanding crime in Native America should reflect the diversity of Native America. That recognition may encourage scholarship and, ultimately, public policies that reflect the needs of Native and non-Native communities in the United States.

Rich Braunstein is an associate professor of political science at that University of South Dakota. He is the principal investigator of an ongoing study regarding American Indians and criminal justice.

William Anderson is director of the Government Research Bureau and an assistant professor of political science at the University of South Dakota. He worked as an analyst in the U.S. intelligence community and continues to serve as a consultant with the federal government.