Reflections on a Proposed Theory of Reservation-Dwelling American Indian Alcohol Use: Comment on

Abstract

In their recent article, N. Spillane and G. Smith (2007) suggested that reservation-dwelling American Indians have higher rates of problem drinking than do either non–American Indians or those American Indians living in nonreservation settings. These authors further argued that problematic alcohol use patterns in reservation communities are due to the lack of contingencies between drinking and “standard life reinforcers” (SLRs), such as employment, housing, education, and health care. This comment presents evidence that these arguments were based on a partial review of the literature. Weaknesses in the application of SLR constructs to American Indian reservation communities are identified as is the need for culturally contextualized empirical evidence supporting this theory and its application. Cautionary notes are offered about the development of literature reviews, theoretical frameworks, and policy recommendations for American Indian communities.

There is much to recommend the article, “A Theory of Reservation-Dwelling American Indian Alcohol Use Risk” (Spillane & Smith, 2007). The article underscores for psychologists the need to augment a focus on individual behaviors with a consideration of the broader contexts in which drinking occurs (e.g., culture, peers, families, communities). Moreover, the proposed theory generates testable hypotheses, which are always an important step in the conduct of good science. At the same time, the central arguments made by Spillane and Smith appear problematic in ways worthy of discussion. Furthermore, their article provides an opportunity to broach a broader discussion of the responsibilities of those reviewing literatures that are complex, are laden with stereotypes, and have major policy implications.

Central Arguments of Spillane and Smith (2007)

Two central arguments undergirded Spillane and Smith’s (2007) theory: (a) that reservation-dwelling American Indians have higher rates of problem drinking than do either non–American Indians or those American Indians living in nonreservation settings and (b) that this pattern arises from the lack of contingencies between drinking and “standard life reinforcers” (SLRs) in these reservation communities. A careful review of the literature shows that neither of these arguments is fully supportable.

Reservation-Dwelling American Indians at Higher Risk for Alcohol Problems

As Spillane and Smith (2007) pointed out, the epidemiological data on American Indian alcohol use have lagged behind those of other populations. Indeed, Spillane and Smith cited only two articles to support their assertion of greater risk for alcohol problems for reservation-dwelling American Indians in comparison with other populations. The first (Beauvais, 1992) compared reservation-dwelling American Indian youths with those surveyed in nonreservation settings—both American Indians and non-Hispanic Whites. The American Indian youths living on reservations were about as likely to have used alcohol but appeared more likely to have reported having been drunk than did their nonreservation counterparts. Unfortunately, statistical tests were not reported in this particular piece, so it is impossible to discern whether true differences existed. In a study not cited by Spillane and Smith, a group of predominantly reservation-based American Indian high school seniors did not differ from a nationwide sample of seniors in lifetime alcohol use (Plunkett & Mitchell, 2000). These American Indian youths did have significantly higher levels of 30-day use of alcohol, but those differences disappeared once the authors controlled for geographic region.

In the second article cited by Spillane and Smith (2007), O’Connell, Novins, Beals, and Spicer (2005) conducted a joint analysis of a large reservation-based American Indian data set, which included participants from the Northern Plains and the Southwest, with the National Longitudinal Alcohol Epidemiological Study (NLAES), which provided sufficient numbers of geographically dispersed (and largely nonreservation) American Indians for analysis. Here, Spillane and Smith highlighted the result that reservation residents were likely to report greater quantities of alcohol consumed per drinking occasion than were other Americans. Yet, the authors did not describe two other central findings. First, the Northern Plains reservation participants were most likely to have consumed alcohol in the past year, the Southwest reservation sample was the least likely, and the NLAES sample fell between these two samples. This finding underscores the tremendous variation in reservation-based samples. Second, among drinkers, the NLAES American Indian sample—much more geographically dispersed than the reservation-based samples—was arguably at highest risk for alcohol-related problems: They drank as frequently as other Americans (with the reservation-based samples drinking much less frequently) while also consuming greater quantities of alcohol than did others in the NLAES sample.

These are cases in point that reflect the more robust conclusion in a wide range of epidemiological research: Alcohol use rates across age groups and tribes are immensely variable (Beals et al., 2003; May, 1996; Spicer et al., 2003; Welty, 2002). Moreover, many reservation-dwelling American Indian adults—usually a higher percentage than among most populations—consume little or no alcohol; among drinkers, a subset exhibit a pattern of episodic heavy drinking at levels exceeding those of others in the United States (Beals et al., 2003; May & Gossage, 2001; Welty, 2002).

In sum, Spillane and Smith’s (2007) assertion that reservation-dwelling American Indians are at greater risk for alcohol-related problems than are nonreservation Indians was based on two articles from a small but growing literature. This assertion was the foundation of the authors’ theory, yet the literature fails to support it fully. Further, the authors overlooked the diversity and complexity of alcohol consumption patterns found in Native America today.

Lack of Contingencies Between Alcohol Use and Standard Life Reinforcers

The second claim—that the lack of contingencies between drinking and common reinforcers is responsible for the higher rates of problem drinking in reservation communities—was also tenuous. Key to this argument were the assertions that some reinforcers (employment and education) are almost universally unavailable on reservations, whereas others (housing and health care) are, for all intents and purposes, universally available. This portrayal of American Indians’ access to these standard life reinforcers both oversimplifies the matter and conflicts with numerous published accounts. For the sake of brevity, we focus on data from the 2000 U.S. Census in the next paragraphs.¹

Employment was listed by Spillane and Smith (2007) among the factors asserted to be almost universally unavailable. The 2000 U.S. Census estimated that 14% of those residing on reservations were unemployed compared with 6% of the U.S. total population (U.S. Census Bureau, 2000; Table GCT-P12). Although this level is more than twice the national level, it hardly reflects a universal lack of employment. Similarly, the authors described educational opportunities as being universally unavailable, yet, according to the 2000 U.S. Census, 73% of the reservation-based population had at least a high school degree, as compared to with 80% of the U.S. population as a whole (Table GCT-P11). In the general U.S. population, 24% had a bachelor’s degree, in comparison with 13% of the reservation-based population. These figures do not reflect the myriad of other postsecondary educational opportunities many American Indians have pursued that are likely more appropriate for the economic conditions of reservation communities; most notable is the increasing availability of degree programs available within the country’s 32 tribal colleges. Although the educational opportunities available to American Indians residing on reservations differ from those available to the general population, they are far from universally unavailable.

Spillane and Smith (2007) also described housing and health care in reservation communities as “nearly universally available” (p. 405). Using the 2000 U.S. Census, Spillane and Smith selected a sample of 159 reservations and found no homelessness on 139 reservations and less than 0.1% on the remaining 20. However, because the 2000 U.S. Census used specialized methods to include homeless persons, users of these data are cautioned that “any attempt to use the results from these specialized operations as a measure of … homelessness would be inaccurate” (Smith & Smith, 2001, p. 1). Moreover, according to the U.S. Census Bureau (2000), persons living on tribal lands were more than twice as likely to live in overcrowded housing than were other Americans (15% vs. 6%; Table GTC-H8) and over 10 times more likely to live in substandard housing lacking complete plumbing or complete kitchen facilities (Table GCT-H7). Relative to other Americans those living on reservations paid about the same percentage of their monthly income in mortgage (32% vs. 31%) and only slightly less if renting (11% vs. 15%; Tables GCT-H6, GCT-H9, GCT-P14). In sum, a true picture of homelessness in reservation communities is not available, and extant data show that American Indian reservation housing is overcrowded and substandard and that it requires similar levels of monthly financial commitment as for other Americans.

Finally, the assertion that reservation-dwelling Indians enjoy universally available health care is problematic. Although it is true that those meeting certain criteria are eligible for Indian Health Service coverage, agreement exists that the Indian Health Service is “woefully underfunded” (Gone, 2004, p. 12). American Indians receive the fewest health care dollars per capita ($1,914 in 2003), as compared with other groups with federal health care funding, including veterans ($5,214), Medicaid recipients ($3,879), and federal prisoners ($3,803; Indian Health Service, 2002; U.S. Commission on Civil Rights, 2003). Furthermore, access to health care is often limited, not only by available resources, but also by the extensive distances many American Indian patients must travel as well as other barriers to care they face when they seek to avail themselves of services from the Indian Health Service (Dixon & Roubideaux, 2001).

Spillane and Smith’s (2007) assertions that a lack of contingency exists between drinking and alternative reinforcers were also inaccurate. Contingencies do exist. As Spillane and Smith themselves noted, American Indians are more likely than middle-class Whites to experience serious alcohol-related health problems and premature death (May, 1996; Nixon, Kayo, Jones-Saumty, Phillips, & Tivis, 2007). Spillane and Smith also cited the high incarceration rates for alcohol offenses. Furthermore, the laws of some tribes are comparatively severe; for example, possession of alcohol or drinking while pregnant can lead to incarceration (Berman, Hull, & May, 2000). American Indians experience higher alcohol-related arrest rates than all other races (Perry, 2004), and ethnographic work has consistently demonstrated how deeply troubled American Indian individuals are by the consequences of their drinking (Mohatt et al., 2004; Spicer, 1997).

In sum, existing research does not support the central arguments made by Spillane and Smith (2007)—that most reservation-dwelling American Indians are at increased risk for alcohol problems and that a lack of contingency exists between problematic drinking and other aspects of their lives.

Responsibilities of Reviewing Alcohol Use Among American Indian Populations

Spillane and Smith (2007) do advance the literature by commencing their theory generation within the reservation context rather than the more typical approach of seeking to adapt theories developed in other cultures to American Indians. Also important, their article has provided an opportunity for a discussion of the responsibilities of scholars committed to the use of theory and research to address issues such as alcohol use in American Indian reservation communities. Indeed, those familiar with American Indian health research know that many community members and leaders feel that the extent and consequences of alcohol use have received disproportionate attention among the health issues they face. In comparison with other Americans, American Indians are at higher risk for a number of physical ailments (e.g., diabetes, oral health diseases, unintentional injuries and trauma, tobacco-related illnesses, certain infectious diseases; Rhoades, 2000) and mental health problems (e.g., depression, posttraumatic stress disorder, suicide; Manson, 2001). Further, high levels of comorbidity among all of these have been noted (as is true generally). Although few community members or researchers would dispute that alcohol-related health disparities exist, many argue that these problems receive undue attention because they fit the stereotype that many Americans hold of American Indian drinking. To be guided by such stereotypes and focus on only one aspect of the interrelated health problems facing American Indian communities severely limits our understanding of the etiology of these problems.

The Complex Problems of American Indian Alcohol Use

Problems surrounding alcohol use in American Indian communities are multifaceted. Spillane and Smith’s (2007) goal of an integrative theory is an admirable one, yet inclusion of interdisciplinary perspectives would strengthen it further (Stokols, 2006). Although we appreciate that a full summary of such multidisciplinary perspectives is impossible in a journal article, reference to these extensive literatures is important, and any theory development should be amenable to their inclusion. For instance, how might the proposed theory integrate the relatively high levels of recovery from alcohol use disorder found in at least some studies of American Indians (Spicer et al., 2003)? Further, how does this theory incorporate common comorbidities, such as posttraumatic stress disorder among those with alcohol problems, particularly given that these problems often predate significant alcohol use (Beals et al., 2002; Kunitz et al., 1999)?

A major—and important—component of Spillane and Smith’s (2007) argument focused on the centrality of the contexts in which American Indians live; we would encourage the adoption of multilevel approaches to better capture contextual relationships. Alcohol-related findings in demography, sociology, family studies, and other fields have advanced our understanding of the importance of placing individual behavior in the broader context (Duncan, Duncan, & Strycker, 2002). In addition, it is critical to incorporate ethnographic literatures. Anthropology has the longest history of grappling with the meaning and context of alcohol use in American Indian communities, and anthropologists have explicitly grounded their theories in the lived experiences of those with whom they work. The inclusion of such perspectives could help to contextualize the rationalist approach in Spillane and Smith’s model, as well as to underscore the important cultural variations that shape alcohol use (Levy & Kunitz, 1974; Spicer et al., 2003).

In fact, researchers have a responsibility to explicitly include the perspectives of those most impacted by their work. Several remarkable incidents in American Indian communities where researchers acted with perceived negligence have fueled explicit expectations in this regard (Chapman & Khabbaz, 1994; Foulks, 1989). Tribes increasingly require review and approval of projects before their implementation and of publications before submission; tribal institutional review boards are becoming common-place. These actions emanate from a need to ensure that research conducted in American Indian communities is culturally grounded and that both individuals and communities are protected from harm (National Congress of American Indians, 2005).

Certain statements in Spillane and Smith’s (2007) article are unrepresentative of many reservation residents’ experiences, for example, “In addition, on reservations, there is little contact with Caucasians, so one is unlikely to experience the advantages biculturalism brings” (p. 401) and “A comparison of reservation-dwelling American Indians to middle-class U.S. Caucasians may essentially involve comparing two groups that are on opposite ends of the continuum of SLR access and drinking–SLR contingency” (p. 406). Had community advisory groups been consulted, such overgeneralizations likely would have been challenged.

In summary, it is incumbent upon those developing literature reviews or generating theories to ensure that their work considers these varying perspectives and accounts for available empirical evidence from different disciplines. Most importantly, researchers have a responsibility to accurately reflect the experiences of those about whom they write.

The Stereotypes Surrounding American Indian Alcohol Use

Negative stereotypes about American Indian alcohol use are pervasive and enduring. It is impossible to work in the field of American Indian health and not be confronted by astounding stereotypical assumptions. Yet, alcohol problems are not universal in American Indian communities, and researchers should be careful not to suggest that they are. Further, the diversity within American Indian communities is immense and should be underscored. In this vein, those summarizing literatures focused on ethnically defined groups are urged to attend carefully to the problems inherent in “ethnic gloss”—the assumption, often implicit, that those in a given group are inexorably the same (Trimble, 2007).

Responsibilities Around Public Policy Recommendations

Finally, in areas as complex and replete with misunderstanding and gross generalizations as American Indian alcohol use, policy recommendations must always follow, not precede, the evidence. Spillane and Smith (2007) concluded,

A comprehensive model might consider both the evolution of institutionalized discrimination against American Indians and the emergence of apparently well-meaning, perhaps compensatory efforts to provide housing and health care on reservations. Both of these factors may actually facilitate American Indian drinking, if our analysis is correct. (p. 412)

Such recommendations are not only suspect, as argued above, but also extremely dangerous, because they could offer the impetus to further decrease the already minimal funding for housing, health care, education, and employment in reservation communities. Unlike Medicare and Medicaid, for instance, Indian Health Service monies are not entitlements but rather derive from discretionary Congressional budgets (Dixon & Roubideaux, 2001). Although the treaties that the U.S. government forced upon tribes promised these very supports, the fiscal pressures that the federal government faces continue to diminish the available resources. Without solid empirical evidence to support their assertions, those invested in American Indian health must take extreme care not to precipitously provoke further cuts.

Possible Adaptations to Spillane and Smith’s (2007) Model

This commentary was prepared by a diverse group of scholars, both multidisciplinary (e.g., psychology, sociology, anthropology, medicine, social work) and multicultural (non–American Indian and American Indian, the latter representing many tribes and urban American Indian communities). Of particular salience are the perspectives of scholars and community members who have spent substantial periods of their lives on reservations, because they are especially qualified to identify inaccuracies with Spillane and Smith’s (2007) depictions of American Indian reservation life. The theoretical model presented in Figure 1 of Spillane and Smith’s article was discussed. Many expressed concerns about the assumed “universals,” suggesting instead that it would be critical to measure—within specific communities—resources such as education, health care, employment, and housing. Some of these assessments could be derived from sources such as the U.S. Census, but many would optimally require specifically designed, culturally contextualized data collection efforts. For instance, it is not uncommon for more traditional tribal members to earn a subsistence living; goods obtained in this manner may not easily be captured by Census questions. Examples exist for each of the four domains (education, employment, housing, and health care), and use of such data would help to prevent inaccurate generalizations and misrepresentations. In addition to individual-level measures, community-level constructs, such as trauma exposure, social disorganization, alienation, and hopelessness, would also provide important context. Indeed, a comprehensive test of Spillane and Smith’s theory would require a very large study involving multiple reservation communities representing not only the cultural diversity of American Indian populations but also other populations where levels of SLRs vary. All of the authors involved in this commentary were in agreement that if such a study were to be conducted, a critical first step would be a meaningful dialogue with community members—including elders, tribal leaders, and youths. At the same time, all were impressed by the practical hurdles to such a study, even if funding for such an effort could be obtained.

Conclusions

As we close this commentary, we again acknowledge the opportunity that Spillane and Smith’s (2007) article has provided, bringing attention to the importance of theory development concerning problematic alcohol use in American Indian communities. Yet, a full understanding of the scope of alcohol use and—most importantly, viable remedies—requires a very careful and culturally contextualized consideration of the facts and their implications.