Cross-Jurisdictional Sharing for Emergency Management-Related Public Health: Exploring the Experiences of Tribes and Counties in California

Each American Indian tribe is unique in several ways, including in its relationships with local governments and risk for emergencies. Cross-jurisdictional sharing (CJS) arrangements are encouraged between tribes and counties for emergency management-related population health, but researchers have not yet explored CJS experiences of tribes and counties for emergency management. This investigation used collaboration theory and a CJS spectrum framework to assess the scope and prevalence of tribe–county CJS arrangements for emergency management in California as well as preconditions to CJS. Mixed-methods survey results indicate that tribes and counties have varied CJS arrangements, but many are informal or customary. Preconditions to CJS include tribe–county agreement about having CJS, views of the CJS relationship, barriers to CJS, and jurisdictional strengths and weaknesses in developing CJS arrangements. Areas for public health intervention include funding programs that build tribal capacity in emergency management, reduce cross-jurisdictional disagreement, and promote ongoing tribe–county relationships as a precursor to formal CJS arrangements. Study strengths, limitations, and future directions are also discussed.

governmental, and societal process and encouraged public health officials to engage in preparedness with outside entities, including local governments (8,9). Other researchers have examined the associations between emergencies, social contexts, and behavio ral outcomes (10). As reviewed by Ejeta and colleagues, scholars have applied the health belief, extended parallel process, and social cognitive models to studies about emergency health pre paredness (11), while O'Sullivan and colleagues use complexity theory to suggest that coping with uncertainty, providing surge capacity, being a response organization, and communicating with the public are crosssectoral, structural emergency management challenges that impact the dynamic system(s) in which emergen cies occur (12).
An emerging area of public health research is the study of crossjurisdictional sharing (CJS) for emergency management. Health and social scientists promote CJS as one collaborative mechanism for sharing resources to address the population health of two or more jurisdictions, particularly in fiscally limited areas of the country and world (13). CJS is important because emergencies do not always occur within one geographic setting or sector, and CJSguided emergency management across mul tiple boundaries can increase the effectiveness of preparedness, mitigation, response, and recovery efforts (14,15). Collaboration theory provides a lens for exploring CJS for emergency manage ment because it emphasizes that there are preconditions which make CJS possible, processes by which CJS occurs, and outcomes of CJS (16). Collaboration theory has guided research about strategic alliances and the need for independent entities to share resources to achieve progress toward mutual goals (17) and can be used as a framework for understanding CJS models like the one proposed by the Center for Sharing Public Health Services (CSPHS), which describes CJS as a deliberate exercise whereby two or more jurisdictions explore, prepare and plan, and imple ment and improve CJS to deliver essential population health via a spectrum of CJS arrangements. CSPHS suggests that CJS occurs through one or all of the following types of arrangements: infor mal or customary (sometimes known as asneeded assistance), servicerelated, shared functions with joint oversight, and/or regionalization (18).
Literature reviews using PubMed, National Institutes of Health Library, and books and journals specific to public health, public administration, community interventions, and emergency management topics (search terms: tribal, governments, American Indian/Alaska Native, emergency management, emergency pre paredness, crossjurisdictional sharing) appear to yield no studies about CJS for emergency management population health between tribal and nontribal governments. This gap in the empirical literature is problematic because American Indians and Alaska Natives (AIAN) experience disproportionately poorer health out comes than nonAIAN, all before taking into account the impact of natural and nonnatural emergencies on tribal populations (19,20). The forced relocation of AIAN to rural areas of the United States, far from major hospitals and other emergency resources, contributes to tribes' vulnerability to emergencies. Social scholars also argue that societal forces place vulnerable populations like tribes at greater risk for harm by emergencies by preventing access to needed resources before, during, and after emergencies (3,6). For example, in the United States, tribal governments are classified as sovereign nations with the right to engage in federaltofederal decisionmaking with the United States government (21). Yet, tribes are not guaranteed access to disaster funds awarded to counties by the federal government and could not directly declare emergencies to the President of the United States until 2013 [i.e., Robert T. Stafford Disaster Relief and Emergency Assistance Act (22)].
Tribal governments may benefit from establishing CJS arrange ments with counties to pool resources or formalize assurances that counties will share federal emergency funding, but only a tribe as a selfdetermined, sovereign governing body can choose to enter into a CJS arrangement with a county (21,23). Currently, there are tribal governments engaging in CJS to exercise sover eignty regarding the health of their people, including tribes using public health authority, an extension of tribal sovereignty, to meet specific CJS accreditation benchmarks associated with popula tion health (24-28). Some tribes could be hesitant to participate in CJS with county governments due to the historical mistrust between tribal and nontribal societies, and others may avoid CJS due to limited infrastructure and capacity. However, little is known about the scope, prevalence, and existing conditions for tribe-county CJS for emergency management.
The purpose of this study is to explore the scope and prevalence of tribe-county CJS for emergency management in California, a state with large AIAN and nonAIAN populations susceptible to natural and nonnatural emergencies each year (29). This inves tigation also explores associations between scope and prevalence of CJS and preconditions for CJS, including the size of tribal and county jurisdictions and whether tribes and counties are working toward accreditation in emergency management. Due to histori cal mistrust between tribal and nontribal societies, associations are also explored between the scope and prevalence of CJS and whether tribes and counties agree about having CJS arrange ments. Qualitative data from a second wave of the study provide insight into three additional preconditions to CJS: Views of the CJS relationship, barriers to tribe-county CJS for emergency management, and jurisdictional strengths and weaknesses in developing CJS arrangements for emergency management.

Recruitment
Research staff sent a project recruitment letter to the elected Tribal Chairman/woman of 111 federally recognized tribes in California, including the 109 tribes with federal recognition status in 2015 and 2 additional federally recognized tribes from neighboring states with tribal lands in California (30). Followup telephone calls and emails were conducted by research staff to finalize a list of participating tribes. A recruitment letter was sent to the Clerk of the Board of Supervisors of any county cor responding to a participating tribe, and research staff undertook methods similar to those for tribes to increase county participa tion. Tribes and counties were not told which respective county or tribal jurisdiction(s) participated in the study.

Participants
Initial CJS survey participants were tribal and corresponding county representatives who received recruitment letters or who were designated to participate. Participants were from 83 of 111 tribal jurisdictions (participation rate = 74.7%) and all 29 counties corresponding to the 83 tribes (participation rate = 100.0%). The 83 participating tribes differed from the 28 nonparticipating tribes in geographic location [F(1,110) = 4.45, p < 0.05], with more partici pants being from northern and central California than southern California (31). Geographic region was not statistically associated with scope and prevalence of CJS arrangements; therefore, the sample of 83 tribal jurisdictions was considered representative.

Procedure
After providing informed consent, tribal and county participants completed an Institutional Review Board (IRB)approved, mixed methods survey adapted from CSPHS "Existing CJS Arrangement" instrument (32). Questions were tested and refined with an advisory group of tribal and county emergency managers (32,33), and per mission was obtained from CSPHS before using the adapted survey. As is appropriate in studies with AIAN, research staff respected local protocols for data collection, including asking tribal representatives whether they preferred data collection via webbased survey, paper survey, telephone discussion, or inperson discussion (33)(34)(35). All tribal representatives chose to complete the survey via telephone or during inperson meetings, with research staff writing down participant responses and reviewing answers with participants to guarantee accuracy. County representatives opted to complete the survey via secure internet link. Participants received a small gift after the initial CJS survey and were invited to discuss findings at roundtable meetings after the project ended (36).

Measures
Data about participant demographics, jurisdictionrelated infor mation, scope and prevalence of CJS arrangements, tribe-county agreement about having no or any CJS, and accreditationrelated CJS arrangements were obtained using the initial CJS survey.

Demographic Information
Participant role in jurisdiction was an openended item about job title/department. Research staff categorized tribal and county responses by department type.

Jurisdiction-Related Information
Jurisdiction population and geographic sizes were selfreported by tribal and county representatives. For data analysis, research staff used a California tribal lands directory to derive the total number of tribes in a county (37). Research staff then derived two proportional items: Proportion of total number of tribes per county to county population size and proportion of total tribes per county to county geographic size.

Scope and Prevalence of CJS Arrangements for Emergency Management
In adapting the CSPHS spectrum of CJS arrangements framework (38), initial CJS survey responses were coded into nonmutually exclusive dichotomous variables (0 = no, 1 = yes) to represent jurisdictions that have: (1) Formal CJS arrangements (1 item; for malized in a written document); (2) informal or customary CJS arrangements (1 item; "handshake agreement, " not formalized in a written document); (3) servicerelated CJS arrangements (1 item; formalized in a contract for asneeded work/consulta tion); (4) shared functions with joint oversight CJS arrangements (1 item; successful in establishing people from both jurisdic tions who make decisions before, during, after emergencies); and/or (5) regionalization arrangements (1 item and response to openended comments at end of survey; successful in establish ing a group of people from both jurisdictions who make decisions before, during, after emergencies and comments that jurisdic tions have merged into one department). This coding adaptation is inclusive of CJS arrangements in tribal settings, which can be formal agreements for partnership (e.g., Tribal Resolutions) without specific emergency management tasks and/or informal or customary arrangements unrelated to specific emergency events (see Figure 1).
Research staff reconciled all coding with comments made by participants at the end of the survey. If discrepancies arose between coding and comments, 4person agreement was reached about how to resolve the inconsistency. For example, if a representative reported no formal CJS arrangements but later stated that the jurisdiction has a tribe-county Memorandum of Understanding for emergency management, the jurisdiction was coded as having formal CJS arrangements. Finally, a sum CJS arrangement score 0-5 was calculated for each jurisdiction.

Tribe-County Agreement about CJS Arrangements
Tribe-county agreement about having no or any CJS arrangements for emergency management was crosstabulated by research project staff. There were four total possible combinations of agree ment and disagreement. Agreement was coded when both juris dictions reported having no (0) or any (1-5) CJS arrangements. Disagreement was coded when one jurisdiction reported having no (0) CJS arrangements, and the other jurisdiction reported hav ing any (1-5) CJS arrangements.

Accreditation-Related CJS Arrangements
Jurisdictions had accreditationrelated CJS for emergency man agement when participant reported one reason for establishing the CJS arrangement was to meet national accreditation stand ards (1 item).

Recruitment, Participants, and Procedure
Research staff narrowed the original sample to 24 representa tive tribes and 13 corresponding counties (i.e., 6 tribe-county pairs from each CJS agreement and disagreement category in the initial CJS survey). The IRBapproved CJS followup survey questions were generated by the advisory group to broadly assess CJS preconditions using openended questions (33). After taking the initial CJS survey, tribal and county representatives were recontacted, provided additional informed consent, and answered questions over the telephone. Research staff wrote down participant responses and reviewed written information with participants for accuracy. County representatives were prompted to answer questions for each tribe in their jurisdiction. Thus, 24 tribe and county responses to each item were possible.

Measures
Openended items on the followup CJS survey were about views of tribe-county CJS relationship for emergency management, historical or cultural barriers to the CJS relationship, and strengths and weaknesses of the jurisdiction in developing tribe-county CJS arrangements. Research staff developed an emergent qualitative coding scheme based on respondents' answers to followup CJS survey items. Emergent themes were added to a preliminary code list using NVivo for Windows Version 11 (39). Independent raters applied the codes, discussed coding discrepancies, and developed new subcodes if necessary. Raters continued this iterative pro cess until 2coder agreement was reached for all qualitative data and a final master code list could be applied. For views of the CJS relationship and strengths and weaknesses of the jurisdic tion in developing CJS arrangements, codes were applied once. For historical or cultural barriers to the CJS relationship, several participants reported two or more types of barriers. These juris dictions were coded as having "multiple barriers, " with subcodes applied for specific themes.

resUlTs initial cJs survey
Descriptive statistics were used to assess demographic infor mation, jurisdictionrelated information, scope and prevalence of CJS arrangements for emergency management, tribe-county agreement about having no or any CJS arrangements, and accreditationrelated CJS arrangements. Correlational and inferential statistics were used to test associations between jurisdictionrelated information and scope and prevalence of CJS arrangements; tribe-county agreement about having no or any CJS arrangements and scope and prevalence of CJS arrangements; and accreditationrelated CJS arrangements and scope and prevalence of CJS arrangements.

Demographic Information
Frequency scores indicated that tribal participants were elected leaders or other tribal administration staff [

Scope and Prevalence of CJS Arrangements for Emergency Management
The most frequently reported CJS arrangement for both tribes and counties was informal or customary CJS arrangements, while approximately 15% of tribes and 28% of counties reported having formal CJS arrangements (e.g., Memorandum of Under standing) for emergency management. See Table 1

Tribe-County Agreement about CJS Arrangements
Approximately 55% of tribe-county pairs were in agreement about having no or any CJS arrangements, (see Table 2).

Associations between Accreditation-Related CJS Arrangements and Scope and Prevalence of CJS Arrangements
Given the small number of tribes and counties that reported having CJS arrangements to meet national accreditation stand ards, it was not possible to statistically test associations between accreditationrelated CJS arrangements and scope and prevalence of CJS arrangements.

Follow-up cJs survey
Qualitative analyses were conducted to determine emergent themes in response to three followup CJS survey items. Descriptive sta tistics were used to tabulate how frequently themes were reported for each item.

Views of the Tribe-County CJS Relationship
When asked to describe views of the tribe-county CJS relationship for emergency management, representatives from 24 tribes and 13 corresponding counties provided emergent themes of positive, negative, neutral, and non-existent. Positive views were limited to comments depicting the tribe and county as having a strong or collaborative relationship. Neutral views were comments without a discernable positive or negative sentiment.

Associations between Jurisdiction-Related Information and Scope and Prevalence of CJS Arrangements
There were significant positive associations between three juris dictionrelated variables and specific types of CJS arrangement. For tribes, there were significant positive associations between population size and having formal CJS arrangements (r = 0.24, p < 0.05); population size and having shared functions with joint oversight CJS arrangements (r = 0.25, p < 0.05); and geographic size and having shared functions with joint oversight CJS arrange ments (r = 0.24, p < 0.05). For counties, there was a significant positive association between total number of tribes per county and having informal or customary CJS arrangements (r = 0.43, p < 0.05).
Multiple linear regression analyses revealed one significant negative association between proportion of tribes per county to county population size and tribereported sum of CJS arrange ments (t = −2.22, p < 0.05): a greater total number of tribes to county population size were associated with fewer total CJS arrangements reported by tribes. There were no significant associations between proportional variables and countyreported prevalence of CJS arrangements.

Associations between Tribe-County Agreement about CJS and Scope and Prevalence of CJS Arrangements
Chisquare and simple linear regression analyses determined that tribe-county agreement about having no or any CJS arrange ments was positively associated with tribereported formal CJS arrangements [χ 2 (1) = 4.42, p < 0.05], informal or customary CJS arrangements [χ 2 (1) = 7.64, p < 0.01], and shared functions with joint oversight CJS arrangements [χ 2 (1) = 7.42, p < 0.01]. Tribe-county agreement was also positively associated with tribereported prevalence of CJS arrangements (β = 0.79, Compact, but if there is a level of conversation at the county level about emergency services, the tribe is not at the table or made aware of the meetings. The tribal council hasn't made it a priority to ask the county about these meetings. (Tribe, neutral) There is no ongoing relationship between the tribe and county. (Tribe, nonexistent) Any time the [surrounding] tribes start casinos, the county tries to benefit. The tribe is skeptical about outreach [to attend emergency management and pre paredness meetings] because the tribe feels the county is trying to meet a requirement. If there were a natural disaster, the tribe would feel uncomfortable and would be skeptical about whether or not state or county ser vices would be provided to the tribe. (Tribe, negative) Because only tribes reported having a negative view of the CJS relationship, tribal and county views were sometimes disparate: The relationship is nonexistent. The county as a whole hasn't really heard from the tribe since the flood [omit ted] when the tribe lost property. (County, nonexistent) The tribe had experienced high waters and nobody from the county came to check on our wellbeing or alert tribal members. (Tribe, negative)

Historical or Cultural Barriers to the CJS Relationship
Emergent themes regarding historical or cultural barriers to the CJS relationship for emergency management were legal/ jurisdictional restrictions, distrust, limited knowledge about tribal systems, multiple, other, and no/unknown. Legal/jurisdictional barriers were present when a respondent noted laws (e.g., Public Law 83280 known as "PL280") and/or governmental limitations to the tribe-county CJS relationship. Distrust was comments referencing current or historical mistrust between the tribe and county, including nontribal governments participating in the massacre and displacement of tribal peoples. Limited knowledge of tribal systems was comments about limited or lacking knowl edge about tribal sovereignty, tribal government and infrastruc ture (e.g., funding and tribal jobs/roles), or cultural protocols. Multiple barriers were present for jurisdictions with two or more of the aforementioned barriers, and other barriers were beyond an emergent theme. No/unknown barriers were coded for tribal and county jurisdictions with no known restrictions. Tribes most frequently reported multiple barriers to the CJS relationship for emergency management, while counties most frequently reported no/unknown barriers. See Tables 3 and 4 for frequency scores and examples by theme.

Strengths and Weaknesses of the Jurisdiction in Developing Tribe-County CJS Arrangements
A total of 14 tribal representatives and 13 county representatives identified strengths of their own jurisdiction in developing CJS arrangements for emergency management. Relationships and jurisdictional preparedness were tribal strengths, and relationships and resources were county strengths. Relationship strengths were coded for 11 tribal and 12 county representatives who said that their jurisdiction's strength was networking, engaging in ongoing CJS communications, or collaborating on tribe-county projects to develop CJS arrangements: A total of three tribal participants said their jurisdiction's strength was being ready for an emergency (i.e., jurisdictional preparedness), and one county representative said their jurisdic tion's strength was its resources: The tribe has a good standalone emergency manage ment program. (Tribe, jurisdictional preparedness) The county has resources that tribes can access. (County, resources) A total of 10 tribal and 7 county representatives identified weaknesses of their own jurisdiction in developing CJS arrange ments. For tribes, limited or no capacity for emergency management was a weakness reflected in statements about limited tribal training, funding, or manpower for emergency management work.
There are time restrictions on tribal staff. (Tribe, limited or no capacity) Sometimes the capacity of the tribe isn't large enough to support the full structure that may be required [for emergency management]. (Tribe, limited or no capacity) The tribe needs to convince their elected officials that money needs to be spent in this area. (Tribe, limited or no capacity) For counties, jurisdictional weaknesses were laws that restrict the development of CJS arrangements, having limited staff for emergency management ("the county is scrambling"), and lack of information, including not knowing who to contact within a One unexpected set of findings emerged during qualitative analyses because some tribal and county jurisdictions reported weaknesses of the other jurisdiction in addition to weaknesses of their own jurisdiction. Tribal representatives said that county representatives exclude tribes from emergency management train ings and events. County representatives reported that tribes have frequent staff turnover and do not have designated tribal emergency managers who can help develop tribe-county CJS arrangements ("tribal staff wear many hats"). One county representative cited lack ing formal arrangements as a weakness of tribes. Examples include: The tribe is never sent an official invitation to the emer gency management development meetings hosted by the county, even though the tribe would like to attend. (Tribe, discussing county weakness) It's a challenge that most tribes do not have an established emergency manager, often because of [not having] funding. If there is an emergency manager within the tribe, sometimes it is someone who has multiple responsibilities and won't have time to engage in the advanced planning for emergency management. (County, discussing tribe weakness) The lack of formal written agreements [is a weak ness] because relying on handshakes [with tribes] can be iffy. (County, discussing tribe weakness).

DiscUssiOn
This study provides novel information about tribe-county CJS experiences for emergency management and population health. This is the first empirical investigation to use collaboration theory and the adapted CSPHS CJS spectrum to explore the scope, prevalence, and preconditions of CJS between tribes and counties for emergency management. Results indicate that CJS arrange ments vary in scope, with all five types of arrangements reported by California jurisdictions with existing tribe-county CJS for emergency management. Less integrated CJS arrangements were predominant in this sample, with both tribes and counties most frequently reporting informal or customary CJS arrangements (18). This finding supports literature about tribal leaders not wanting to establish written arrangements so they can maintain flexibility in crosssectoral responsibilities and/or allow time for the development of trust between tribal and nontribal partners (27). It is noteworthy, however, that qualitative data from this study highlight limited tribal capacity and funding for emergency management as preconditions which weaken tribe-county CJS. Additional findings reveal that only tribes with larger population sizes are associated with having formal CJS arrangements. Thus, although CJS is a mechanism by which resourcelimited jurisdic tions can pool manpower and funding to improve population health (13,18), results of this study suggest that smaller tribal jurisdictions may seek and benefit from informal or customary CJS arrangements because they do not possess the minimal emergency management assets required to enter into written tribe-county agreements.
Because nearly 20% of the total number of United States fede rally recognized tribes are located in California, results of this study offer preliminary understanding about the nationwide prevalence of tribe-county CJS for emergency management (30). Findings suggest that more than half of tribes and threefourths of coun ties report having at least one CJS arrangement for emergency management. Yet, results also highlight a gap among tribes and counties with no CJS arrangements. Statistical analyses further indicate that tribes report fewer total CJS arrangements when located in counties with a greater total number of tribes to county population size. In these areas, the tribe-county CJS relationship could be diminished because county representatives assume there are enough other people to help tribes before, during, or after emergencies (40). However, due to the unpredictability of emergencies, county representatives do not know whether tribal resources (e.g., shelter, food, and staff) will be the only means by which to protect nontribal citizens (41). It would benefit tribes and counties to formulate CJS arrangements to protect both populations.
Another important finding is the discordance between tribal and county perspectives about having CJS for emergency management. Nearly half of tribes and counties disagree about working together, with disagreement most frequently occur ring when counties report CJS but tribes do not. Different governmental priorities and an inability or limited desire to communicate contribute to disagreement between tribal and nontribal representatives, but increased meaningful interactions and mutual understanding of respective jurisdictional values can mitigate conflicting perspectives (42). Results of this study also suggest that tribe-county CJS agreement is associated with tribe reported scope and prevalence of CJS, signifying the importance of concordance as a precondition for tribal CJS experiences. In turn, population health efforts should be aimed at facilitating longterm, tribe-county CJS collaborations that increase mutual awareness, decrease disagreement, and establish successful CJS arrangements (43).
The mixed methodology of this study allows for both systematic and emergent identification of preconditions to CJS. As previously mentioned, jurisdictionrelated indicators (i.e., jurisdiction size and relation to total number of tribes in county) and tribe-county CJS agreement are statistically associated with scope and prevalence of CJS arrangements for emergency management. Qualitative data offer nuanced infor mation about three additional preconditions to CJS. Specifically, views of the CJS relationship, barriers to CJS, and jurisdictional strengths and weaknesses in developing CJS arrangements seem to impact jurisdictions' willingness and ability to participate in tribe-county CJS. Emergent themes reveal varying sentiments about the CJS relationship (i.e., positive, negative, neutral); jurisdictionspecific strengths and weaknesses in developing CJS arrangements (e.g., relationship building and limited capacity); and weaknesses of the other jurisdiction in tribe-county CJS (e.g., tribal staff turnover, preference for informal arrangements, exclusion by counties). Tribes and counties report barriers to CJS that range from systemslevel legal issues like Public Law 280, which contributes to tensions between tribal and county law enforcement in at least six states (44,45), to distrust, limited knowledge of tribal systems by county representatives, and a combination of these barriers. In the future, research ers can utilize emergent themes to further explore direct and indirect relations between preconditions to CJS and scope and prevalence of tribe-county CJS arrangements for emergency management. This will be profitable when examining themes with subcategorizations that vary by jurisdiction type, including negative views of the CJS relationship which were only reported by tribes and no/unknown barriers to the CJS relationship which were frequently reported by counties but not tribes. Additional mixedmethods research with regional and national samples of tribes and counties will improve understanding about tribe, county, and tribe-county (i.e., actor-partner) models of CJS preconditions and arrangements in the United States.
There is interest in traditional, practicebased accreditation for population health throughout Indian Country, and some tribes report developing CJS arrangements to meet national accreditation standards (27,28,46). In this study, however, accreditationrelated CJS arrangements were not common or relatable to scope and prevalence of tribe-county CJS. One explanation for this finding is that tribes and counties value CJS for emergency management above and beyond meeting accredi tation requirements. Another possibility is that tribes have dif ficulty aligning emergency management, population health, and accreditation efforts with national benchmarks. In California, for example, tribal governments can be involved in emergency management and also designate health functions to tribal health programs (23). Unfortunately, neither tribal government nor health systems are typically compatible with the organizational structures required to achieve national accreditation (47). In this study, lacking accreditationrelated CJS arrangements highlight an opportunity for population health programs to advance tribal capacity for accreditation and tribe-county CJS for emergency management.
Strengths of this study include using culturally responsive research methodology and social theory to examine tribe-county CJS for emergency management. Tribal Epidemiology Center staff with experience navigating tribal protocols used an iterative process for study improvement which involved seeking guidance from advisory group members and tribal leaders during project recruitment, data collection, and dissemination (48). Adapting data collection methods can contribute to higher participation rates by tribes because protocols respect and adhere to sovereign preferences for research within tribal nations (33). By grounding this study in collaboration theory and the CJS framework, results also contribute to the literature about emergency management as a faction of societal health and wellbeing and expand our knowledge about tribal and county representatives from varied disciplines who participate in emergency management to protect the health of their people (7).
One limitation of the present research is the assumption that CJS is a linear process in which preconditions are antecedents to the scope and prevalence of tribe-county CJS arrangements for emergency management. Instead this process may be inter dependent in nature, and future investigations should explore this possibility. The purpose of this study was to examine CJS for emergency management between federally recognized tribes and corresponding counties, but staterecognized tribal governments may have disparate experiences with CJS given their status as nonfederal entities (49). Thus, future research should address the second limitation of this study, which is the exclusion of staterecognized tribes. Finally, results and generalizability of this investigation, including qualitative data, would be enhanced by mixedmethods explorations of tribe-county CJS for emergency management across the United States. At the time of this publica tion, there were 567 federally recognized tribes (50), and it may be that region and communityspecific best practices and experi ences, including information unpublished in the formal empirical literature, could vary in places like frontier Alaska Native villages, which represent approximately 40% of tribes nationwide. Future studies about tribe-county CJS for emergency management would provide a clearer picture of the generalizability of this study.
cOnclUsiOn This study used a public health framework to examine tribe-county CJS experiences for emergency management and preconditions to CJS arrangements. Quantitative and qualitative findings sug gest that tribes and counties have varied arrangements for CJS, many of which are informal or customary. Preconditions to tribecounty CJS are jurisdictionrelated indicators, tribe-county CJS agreement, views of the tribe-county CJS relationship, barriers to tribe-county CJS, and strengths and weaknesses of jurisdictions in developing CJS arrangements. Areas for public health inter vention include programs that build tribal capacity in emergency management and accreditation, reduce crossjurisdictional dis agreement, and promote ongoing relationship building between tribes and counties as a precursor to formal CJS arrangements. Future public health practice and research initiatives about tribe-county CJS for emergency management can help reduce emergencyrelated health disparities among AIAN people and protect both tribal and county jurisdictions before, during, and after emergencies.

eThics sTaTeMenT
This study was carried out in accordance with the recommenda tions of the California Rural Indian Health Board, Inc. (CRIHB) Institutional Review Board (IRB) with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the CRIHB IRB.
aUThOr cOnTribUTiOns MW served as the principal investigator of this project and par ticipated in all aspects of research, including developing surveys; facilitating California Rural Indian Health Board, Inc. (CRIHB) IRB approvals; planning and conducting project advisory group meetings; collecting onsite, telephone, and webbased data from tribes and counties; supervising staff in data collection; conduct ing quantitative and qualitative analyses to test research questions; and writing this manuscript. MW circulated the document for review within CRIHB and received approval of the manuscript from the CRIHB IRB.

acKnOWleDgMenTs
The author would like to thank Michael Mudgett, Christine Smith, Cassandra Call, and Elizabeth Benton who assisted with data collection and/or preliminary coding. The author would also like to thank CRIHB staff in leadership at the time of this publica tion, including Dr. Mark LeBeau. Appreciation is also extended to members of the project advisory group.
FUnDing Support for this manuscript was provided by the Robert Wood Johnson Foundation (72458), and the views expressed herein do not necessarily reflect the views of the Foundation.