Goldberg and Washburn: Lies, damn lies, and crime statistics


By Carole Goldberg and Kevin Washburn

Are American Indians more often victims of crime than members of other ethnic and racial groups? Are most of the offenses committed against them committed by non-Indians, as opposed to members of their own group? Ever since the federal Bureau of Justice Statistics began issuing reports on this subject in 2000, the clear answer to both of those questions has seemed to be ''yes.''

Now the South Dakota attorney general and researchers at the University of South Dakota have challenged that conclusion, issuing a report that focuses on only one state but questions the Indian data nationally. Their challenge to the federal data is much too quick to dismiss the BJS findings.

Over the past eight years, the BJS, which is a component of the U.S. Department of Justice, has released some startling figures. Although American Indians are .9 percent of the total population, they represent 1.4 percent of all crime victims, a very significant overrepresentation. At least two-thirds of all crimes against Indians, and 80 percent of all sexual assaults, are committed by non-Indians.

Indian women, according to BJS data, are 2.5 times more likely than non-Indian women to be raped or sexually assaulted during their lifetimes.

These statistics have been difficult to ignore. Tribes and Native women's groups have raised them before the Senate Committee on Indian Affairs in order to secure greater support for Indian country criminal justice initiatives. Amnesty International included some of those statistics in a much broader analysis of sexual assault of Indian women in the United States, and used case studies from Indian country to make their point.

The BJS statistics do have some limitations. They are based primarily on nationwide victimization surveys, administered through interviews of individuals at a nationally representative number of households. These data fail to distinguish between Indian reservations and other rural and urban settings. The BJS victimization data encompasses the more than 60 percent of all Indians who live outside reservations, including large urban Indian populations in places such as Los Angeles and Minneapolis.

Indians represent a tiny percentage of the population in such areas, so high percentages of victimization by non-Indians would be expected. Furthermore, 2000 Census data indicates that nearly 55 percent of all Native women marry non-Indians, a greater rate of out-marriage than other racial groups. Under such circumstances, domestic violence would predictably be an interracial event.

But the South Dakota study goes much further in contesting the BJS figures. Using data from crime reports, arrests and prosecutions in the federal and state systems in South Dakota, it claims that Indian country crime is mainly an Indian-against-Indian phenomenon. The report confirms the BJS point that Indians are disproportionately victimized by crime compared with other groups. For example, Indians were 64 percent of the intentional homicide victims according to the South Dakota study, yet only 8.3 percent of the state population. But the South Dakota researchers also argue that Indians are the main crime perpetrators, making crime against Indians in that state intra-racial rather than interracial.

The big difference between the BJS crime figures and the South Dakota numbers is the sources of crime data. The BJS relied on direct interviews with individual crime victims, on the assumption that some crimes will go unreported. The South Dakota report asserts that victimization surveys are unreliable, and therefore the researchers there rely only on crime reports to official agencies. Remarkably, the South Dakota researchers claim that reports by victims are not ''actual crime data'' and their research refuses to supplement the police data with data from victims of crime.

In considering reports by victims of crime, BJS has the more reliable approach. For Indian country generally, and for South Dakota especially, exclusive reliance on official crime reports is a serious problem.

Rape, which is one of the crimes the South Dakota report highlights, is a notoriously underreported crime everywhere and for all groups. We would expect even greater underreporting of rape and sexual assault against Indian women in South Dakota, particularly where the perpetrator is non-Indian. Due to the U.S. Supreme Court's decision in Oliphant v. Suquamish Indian Tribe, Indian nations have lost control over arrests and prosecutions of non-Indian assailants. Other decisions by the Supreme Court have disestablished considerable reservation territory in South Dakota, eliminating Indian country status. As a result, Native victims in South Dakota must depend on the state and federal systems to bring non-Indian perpetrators to justice.

Yet Indian women may wonder whether they should bother to report the crimes committed against them. Reservation residents in that state regularly complain that the BIA, FBI and federal prosecutors decline to prosecute crimes committed in Indian country. Even if tribal police are specially commissioned to carry out federal law enforcement responsibilities, there are too few of them, spread too thin around vast reservations. AI has shown how deficient the physical examinations and evidence collection have been for rape victims who must rely on the IHS.

Outside Indian country, Native women victimized by non-Indians confront a state system they perceive to be biased against them. And these women can point to considerable support for their view. Indians in South Dakota have recently prevailed in lawsuits brought under the Voting Rights Act of 1965, proving discrimination against them in the design of county voting districts. In 2000, the U.S. Civil Rights Commission issued a report documenting charges of discrimination against Indians in that state. The bottom line is this: If Indian women do not believe that their reports will be investigated, they are much less likely to report them.

The SCIA has recently released a draft bill designed to enhance the administration of criminal justice in Indian country. One of the sections of that bill focuses on improved crime data collection, a real need. Until we have better data, it would be a mistake to accept the conclusions of the South Dakota study which ignores data from crime victims.

Carole Goldberg is a Distinguished Professor of Law at the University of California - Los Angeles. Kevin Washburn, Chickasaw Nation, is a Rosenstiel Distinguished Professor of Law at the University of Arizona. They are co-principal investigators on a $1.45 million grant to examine the administration of criminal justice in Indian country, and both serve as co-editors and co-authors of Felix Cohen's ''Handbook of Federal Indian Law.''