Office of Public Health Emergency Preparedness (OPHEP)

Organizing State and Local Health Departments for Public Health Preparedness

Jeffrey Wasserman, Peter Jacobson, Nicole Lurie, Christopher Nelson, Karen Ricci, Molly Shea, James Zazzali, Martha Nelson

TR-318-DHHS

December 2005

Prepared for the U.S. Department of Health and Human Services Office of the Assistant Secretary for Public Health Emergency Preparedness

Contents

Preface
                                       Figures
                                       Tables
                                       Summary
                                       Acknowledgments
                                       1. Introduction
                                       Bioterrorism Funds and Public Health
                                       Purpose of Study and Specific Objectives
                                       2. Analytic Approach
                                       Quantitative Analysis of Public Health Structure on Preparedness Levels
                                       Centralization
                                       Regionalization
                                       Freestanding Versus Umbrella Agencies
                                       Dependent Variables
                                       Data Sources
                                       Independent Variables
                                       Analyzing the Relationship Between Structure and Preparedness
                                       Case-Studies Approach
                                       Data Collection
                                       Data Analysis
                                       Review of Alternative Governance Structures and Strategies
                                       Selection Criteria
                                       Literature Review
                                       3. Results
                                       Quantitative Analysis of Effect of Structure on Preparedness
                                       Results from Case Studies of State and Local Public Health Departments
                                       Organizational Structure of Case-Study Sites
                                       Centralized States Included in Site Visits
                                       Decentralized States Included in Site Visits
                                       Mixed State Included in Site Visits
                                       Observations from Site Visits
                                       Public Health’s Changing Role in Emergency Preparedness and Its Effect on Organizational Structure
                                       Organizational Barriers to Improving Preparedness
                                       Barriers to Organizational Change
                                       Facilitators of Organizational Change
                                       The Relationship Between State and Local Health Departments
                                       The Relationship Between Health Departments and State, Regional, and Local Emergency Management Functions
                                       Roles Played by Other Local Stakeholders in Preparedness Activities
                                       Lessons Learned on Alternative Governance Structures and Strategies From Other Sectors
                                       Private-Sector Approaches
                                       Use of Information Technology
                                       Accountability
                                       Applicability of Formal Organizational Structures Used in Other Sectors to Public Health
                                       4. Summary And Conclusions
                                       Results of the Quantitative Analysis
                                       Case Studies
                                       Lessons from Alternative Governance Structures and Strategies in Other Sectors
                                       Regionalization
                                       Defining Public Entrepreneurs
                                       Privatization Criteria
                                       Limitations
                                       Final Thoughts
                                       Appendix A. CDC Progress Report Indicators (2004)
                                       Appendix B. Public Health Laboratories’ Survey Questions Used in Analysis
                                       Appendix C.  Robust Regression with Centralization-Regionalization Interactions
                                       References
                                       Copyright Information: Rand Corporation
                                    

Figures
Figure 3.1:  Preparedness and Degree of Centralization
Figure 3.2:  Preparedness and Structure of State Health Department
Figure 3.3:  Preparedness and Regionalization
Figure 3.4:  Effect of Centralization in Regionalized and Nonregionalized States

Tables
Table 2.1:  Focus Areas in 2004 CDC Progress Reports
Table 2.2:  Constructs and Items in Summary Preparedness Index
Table 2.3: Comparison of Systems by Selection Criteria
Table 3.1:  Robust Regression Results
Table 3.2:  Robust Regression Results by CDC Focus Area
Table 3.3: Organizational Structure of Case Study States
Table 3.4: Comparison of Services by Descriptive Features
Table 3.5: Comparison of Services by Organizational Features
Table A.1: CDC Progress Report Indicators (2004)
Table B.1: Survey Questions for Public Health Laboratories
Table C.1: Robust Regression with Centralization-Regionalization Interactions

Preface

Improving the ability to respond to bioterrorism and other emergencies is an important challenge facing the U.S. public health system.  Even with a knowledgeable workforce, practice and experience, capacity, and partnerships with other responders in the community, the system’s ability to respond may depend largely on its structure.  This study examines a key question:  Are state and local public health agencies related to one another in a way that facilitates emergency response?

Our specific objectives are to explain the factors influencing the particular ways in which state and local public health systems are organized, how state and local public health departments have arrived at the various types of relationships that exist between them, and, most important, the consequences of such structures and relationships for emergency preparedness.  We also examine alternative structures from several different types of service industries (public education, banking, the welfare system, and port authorities).  Finally, we recommend concrete strategies to improve public health preparedness.

This report will be of interest to policymakers and to public health professionals at the state and local levels who are involved in bioterrorism response and emergency preparedness, as well as to other agencies involved in emergency response.

This work was carried out from the beginning of October 2004 through October 2005.  This report was prepared specifically for the Office of Public Health Emergency Preparedness, but it should also be of interest to individuals working in public health preparedness at the federal, state, and local levels.

This work was prepared for the U.S. Department of Health and Human Services Office of Public Health Emergency Preparedness (OPHEP) and was carried out within the RAND Health Center for Domestic and International Health Security. RAND Health is a division of the RAND Corporation. A profile of the Center, abstracts of its publications, and ordering information can be found at http://www.rand.org/health/healthsecurity/.  Comments or inquiries should be sent to the RAND principal investigators, Jeffrey Wasserman (Jeffrey_Wasserman@rand.org) and Nicole Lurie (Nicole_Lurie@rand.org), at the RAND Corporation, 1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407.  More information about RAND is available on our Web site at http://www.rand.org.

Summary

The organization of the United States’ governmental public health system varies dramatically across federal, state, and local levels.  The federal government provides funding and support, but it delegates authority to the states to organize and deliver public health services (in accordance with the Constitution).  Every state has an agency that is responsible for public health, but the structures of those agencies vary.  In some states, the public health agency is a freestanding agency, functioning on its own; in other states, public health is part of a health and human services department, known as an umbrella organization.

Relationships between the state agency and local health departments also vary.  Some states have a centralized organizational structure, in which the state agency has direct control over local public health services. Other states are decentralized, giving local health departments considerable discretion over decisionmaking and service delivery.  Still others have mixed, or shared, authority.  A 1998 National Association of County and City Health Officials (NACCHO) survey found that 13 states are centralized, 26 states are decentralized, and 11 states are mixed (NACCHO, 1998).  Local health departments, again, vary widely in terms of their jurisdictional size and responsibilities.  Some local health departments have jurisdiction over a city or individual county; others are more regionalized, with jurisdiction over multicounty blocs (ASTHO, 2004; Wall, 1998).

The menus of services provided by public health are equally diverse.  Services are both population-based (e.g., environmental health services, screening, disease surveillance, health promotion, health education, and emergency response), and individualized (e.g., clinics for uninsured patients and maternal and child health care).

Since September 11, 2001 (9/11), bioterrorism preparedness has been a key focus of public health. Following the terrorist attacks and subsequent anthrax attacks, Congress passed an emergency supplemental appropriation that included $900 million for bioterrorism preparedness.  The funding for preparedness has continued from 2002 through fiscal year 2006, with approximately $1 billion per year allocated to state and local health departments-an unprecedented infusion of federal money into public health infrastructure development.

This study assesses whether there is a link between how state and local public health departments are organized and the level of their emergency preparedness.  In particular, we examine organizational variations (i.e., differences in structure and function) of state and local public health departments and the relationships between state and local agencies.  Additionally, we identify alternative governance structures and strategies from other sectors that could be applied to meeting public health preparedness goals and improving preparedness outcomes.

Work Included Statistical Analyses, Case Studies, and Cross-Industry Comparisons

Our study methodology consists of three integrated tasks.  First, we used multivariate statistical techniques to model the effects of various organizational structural characteristics on public health preparedness. Second, we conducted a series of comparative case studies to investigate how the structure of public health systems affects public health preparedness. Third, we conducted a comparison of current governance structures used in other sectors (such as port authorities) and strategies (such as regionalization) that may offer ways to improve the delivery of public health.

In addition to the three tasks outlined above, we interviewed key national, state, and local policymakers and stakeholders (both public and private) to ascertain participants’ views of how the public health system can best meet preparedness goals.

Findings

Statistical Analysis Found Relatively Little Relationship between Structure and Preparedness

Our analyses indicated that organizational structure, as measured by the degree of centralization, had little effect on states’ levels of public health preparedness.  In fact, analyses of variance models suggest that whether a state is centralized, decentralized, or mixed explains almost none of the observed state-to-state variation in the preparedness indices used in our analysis.

The analysis yielded similar findings for whether the state health department is freestanding or part of a larger umbrella agency.  States with umbrella agencies reported slightly higher task-completion rates.  However, once again, the differences are small relative to the amount of overall state-to-state variation.

Finally, regionalization does not explain the variations in self-reported preparedness among the states (and any differences across the groups were dwarfed by within-group variation).

We also considered the possibility that the effect of any one element of public structure might depend on other elements. For example, is centralization an effective organizational structure if it is accompanied by formal regionalization arrangements?  Tentatively, the answer is yes.  We found that centralized systems are associated with the lowest preparedness levels in nonregionalized states but the highest preparedness levels in regionalized states.

The Case Studies Revealed Some Common Themes

Public health’s changing role in emergency preparedness has required reorganization.  The national focus on bioterrorism and emergency preparedness since 9/11 and the subsequent influx of emergency preparedness funds have dramatically affected state and local public health departments.  Although the increased funds were welcome, many health departments found themselves without the organizational infrastructure or constrained by overarching government policies (such as hiring freezes and bureaucratic barriers, etc.) to adequately manage and disburse the funds to both “traditional” public health functions, such as responding to infectious disease outbreaks, enforcing sanitary codes, and monitoring public water supplies, as well as “new” public health functions, such as bioterrorism preparedness.  In all of the health departments we visited, some degree of restructuring was necessary, including creation of new positions, new programs, and/or new departments.  However, tension between the traditional and new public health functions is widespread.  Many feel that the focus on preparedness has diverted needed staff and other resources away from traditional public health activities.

Conflicts between state and local health departments are common.  Regardless of whether a state is centralized or decentralized, there are often conflicts over the respective roles of state and local departments.  While there is general agreement that “all health is local in terms of immediate response” and that local health officials must have the basic skills to screen and triage, many state officials feel that certain types of expertise (such as outbreak management) reside at the state level.  In some cases, these differences have impeded the ability to agree on an appropriate course of action.

Public health is a newcomer to emergency management.  The influx of federal funds and new public health preparedness mandates has thrust public health into a more central role in statewide emergency response efforts.  Respondents generally agreed that, since the events of 9/11 and the anthrax attacks, there is far greater involvement of public health at the local level as well.  Some reported that, in the past, public health typically did not interact with law enforcement, firefighters, or other first responders.  With the funding influx and its associated requirements for building community-level relationships, interagency cooperation is growing.  However, the quest for an equal seat at the table continues to be a difficult process.

Coordination with hospitals is improving.  Hospitals are major stakeholders in local emergency preparedness activities.  The relationship between hospitals and public health departments varies tremendously, but, generally, the relationships have been strengthened significantly by the Centers for Disease Control and Prevention (CDC) and the Health Resources and Services Administration (HRSA) requirements that public health departments and hospitals work with other members of the community on emergency preparedness activities.  These activities have taken various forms.  State and regional hospital associations and regional steering committees have proven successful in integrating hospital and public health preparedness activities (see Davis et al., 2006).

Barriers to improving preparedness.  Common themes across many sites included the following:

• Staff shortages.  Staffing for preparedness can “rob” needed staff of performing more-traditional public health activities.  In addition, some staffing issues were due to turnover:  Some state agencies’ inability to hire staff or fill vacancies is a result of state-level hiring freezes and overly burdensome bureaucratic processes that contribute to delays in disbursing federal grant monies for preparedness.
• Inadequate training.  For some, the concept of Incident Command Structure and the new reality of public health’s first-responder role are foreign.  Institutionalizing these concepts requires ongoing training.
• Poor communication within health departments, as well as between local and state health departments, is a barrier to organizational change.  In some cases, communication problems were primarily the result of poor leadership; in others, they were due to lack of adequate funding.
• Bureaucratic impediments, especially at the state level, and politics at both the local and state levels can retard organizational change.  In one decentralized state, limited financial resources at the local level were a major barrier to state-wide change.  Limited staff time and turf issues -“who’s in charge?”- can present additional barriers.

Facilitators of organizational change.

• Funding was most often identified as the main facilitator of organizational change at the state level.  Without the increases in funding to state health departments, organizational change necessary to meet the requirements of the CDC Cooperative Agreements would not be possible. One of these requirements is that states show that they have a regional response structure in place.  For some highly decentralized states, creating a regionalized structure is challenging.  However, control over the funding allows the state health department to require local health departments to implement changes as a condition of receiving state funds.
• Skilled leadership and good communication facilitate organizational change, especially within health departments.  Good leadership often takes the form of aggressive training programs for employees. 

• Regional alliances, community work groups, or coalitions that were originally formed to achieve other goals (such as economies of scale, coordination of activities within a region, and sharing of information) can facilitate organizational change by encouraging both formal and informal communication and coordination. 

Applicability to Public Health of Formal Organizational Structures Used in Other Sectors

Our review of organizational structures used in other sectors also yielded information on some of the formal structural characteristics that we examined in the quantitative analysis and in the case studies:  degree of centralization and the regionalization of services.

Centralization-Decentralization.  Our cross-sectoral analysis cannot speak specifically to the level of centralization or decentralization that is desirable for public health, but it is clear that the optimal level of centralization will be contingent upon the nature of the particular public health function or task.  Our review of the public health literature did not reveal any extant criteria about which public health preparedness tasks might be centralized as opposed to those that should be decentralized and, as indicated previously, our empirical analysis did not provide any guidance in this regard.

Consistent with the case-study interviews, the cross-sectoral analysis suggests that decentralized structures are feasible for those activities based on community-specific needs that might not benefit from economies of scale.  But centralization remains salient for certain functions.  In education, centralized authority is required for developing standards that all students must meet to fulfill education’s function of providing societal benefits.  In public health, laboratories or epidemiological services that are specialized and applicable across a state should have centralized control because of their wide benefits.  Even so, the risk-communication function in public health preparedness contains elements of both structures.

Regionalization.  Regionalization of public health services is an idea generating considerable attention.  Researchers have proposed that public health services might be more effectively and efficiently delivered on a regional basis, merging counties or states into geographic regions linked by similar health status, economic, or geographic characteristics.  For example, the Mississippi Delta region might be more effectively served through a regional public health authority, rather than by county-level or state-level agencies (Mays and Halverson, 2005).  Our cross-sectoral analysis suggests that regionalization of public health services might be incorporated, but that this approach must be directly linked to the public health objectives.  Both Michigan and Wisconsin have imposed regional bioterrorism preparedness structures on existing local functions.  Because the case studies were not designed to evaluate regionalization, our results are only suggestive that regionalization might be a useful structural mechanism.

Private-Sector Approaches.  Local health departments frequently contract out public health services to for-profit companies in efforts to meet public health goals (Keane, Marx, and Ricci, 2001).  Each of the four systems in our cross-sectoral analysis has been trying to determine the “right” degree of public versus private responsibility over ownership, service delivery, monitoring, and policy development.  This approach requires the governmental partner to conduct smart, aggressive management.

In education, contracting out services that do not have a direct relationship to public education’s mission or values (such as cafeteria, janitorial, and transportation services) is relatively uncontroversial.  But privatization of services related to education’s core function, such as vouchers and for-profit management, is much more controversial.  The analogous challenge for public health is to articulate what the basic functions are that the public sector must maintain, either for ethical or pragmatic reasons, and what services might be contracted to private organizations.

Limitations

Our analyses are subject to a number of important limitations, which should be kept in mind when interpreting our findings and considering our policy recommendations.  With respect to the multivariate analysis, the measures of preparedness employed are limited for a variety of reasons, not the least of which is that there is no agreed-upon definition of preparedness.  Moreover, there are serious questions about the extent to which the self-reported data used to construct the preparedness indices represent valid indicators of capacity to actually mount a successful response.

A second limitation is that our measures of public health structure include only formal and institutional manifestations of structure.  It is well known that the actual operation of institutions depends on a host of informal norms, relationships, and practices that are not represented in our analysis.  As noted above, we speculate that regionalization might be a proxy for informal networks, but this speculation is just that.

Another important limitation is that structure may be endogenous.  That is, it is possible that a given state may have adopted a more centralized organizational structure if it had reason to doubt the ability of local health departments to mount an effective response or because of the political history of the state (which may be unrelated to the historical development of public health within a given state).

The case-study analysis is also subject to a number of important limitations.  First, although we attempted to select sites for study that are broadly representative of all CDC public health preparedness grantees (or awardees), we have no way of knowing whether we were successful in this regard.  As a result, we cannot claim in any rigorous sense that our findings and policy implications are applicable to all grantees.  Second, in conducting the site visits, we attempted to obtain multiple points of views and to synthesize what we heard to arrive at a reasonable portrait of how events unfolded, strategies were considered, and solutions implemented.  But we have not subjected our syntheses to any formal validity and reliability testing, mainly for practical purposes.  Finally, our case-study results are driven largely by the degree to which interviewees were candid in their responses to our questions.

Our review of alternative governance structures and strategies contains many of the limitations found in other studies.  For instance, despite our efforts and intentions, we cannot claim that we have captured and analyzed all of the relevant literature.  Second, it is entirely possible that other sectors may have provided important insights for public health, but budget and other constraints precluded us from casting too wide a net.  Finally, we chose to organize our review by sector, as opposed to by practice (e.g., privatization), reasoning that doing so would provide a preferred context for analyzing our results.  This decision, however, compromised our ability to use more formal analytic techniques, such as meta-analysis, for presenting the lessons learned from the relevant literature.

Final Thoughts

Overall, we are optimistic about the prospects for meaningful organizational reform of the public health system.  The increased funding that public health has experienced as a result of 9/11 and the anthrax scares, coupled with a widespread recognition at the state and local levels of the importance of mounting an effective public health response to a wide range of emergencies, presents significant opportunities for change.

Our analyses, however, failed to reveal any panaceas or magic bullets.  But, taken together, they tell us that the time is right for an extended dialogue that includes federal, state, and local public health officials regarding, essentially, who should do what.  Our case-study results demonstrate convincingly that policymakers should refrain from imposing a one-size-fits-all approach and recognize that risks, infrastructure, and capabilities vary, and will continue to vary, among and within states.  At the same time, efforts must be made to determine how public health systems can provide equal levels of protection to their jurisdictions’ populations.  Finally, policymakers need to harmonize public health preparedness measures across federal agencies, including CDC, HRSA, and the Department of Homeland Security (DHS), and arrive at a consensus regarding which governmental entities will be held accountable for developing and maintaining relevant public health preparedness capabilities, and at what level.

Acknowledgments

The authors wish to thank the many state and local public health officials who contributed their time to helping us develop a deeper understanding of public health organizational structure and its relationship to preparedness.  We greatly appreciate the assistance of Amanda Spendlove for help in preparing this manuscript.  We also thank Pamela Russo and Yee-Wei Lim for their critical review of the report.  And we are especially grateful for the support, feedback, and insights provided by project officer Lara Lamprecht.

Chapter One.  Introduction

After years of neglect, the U.S. public health system is at a crossroads.  Accustomed to living in relative anonymity, public health officials have been thrust into the limelight as the nation collectively confronts a wide range of new and dangerous threats to its well-being, especially bioterrorism and, more recently, natural disasters.  The public health system provides multiple services, with the overarching goal of improving population health, including population-based services (i.e., environmental health services, screening, disease surveillance, and emergency preparedness) and individualized health services (including clinics for uninsured patients and maternal and child health care).  By public health system, we are referring to the federal, state, and local governmental apparatus designed to assess and respond to threats to the public’s health and to provide the core functions of surveillance, policy development, and assurances (Institute of Medicine [IOM], 1988).

The 2003 Institute of Medicine report The Future of the Public’s Health in the 21st Century describes a multisectoral arrangement to deliver public health, involving six collaborative partners: the governmental public health infrastructure, communities, the health care delivery system, employers and businesses, the media, and academia (IOM, 2003).  In reality, however, the public health infrastructure is largely governmental, and every state has a health agency responsible for public health.  Local public health departments (LHDs) function in cities, counties, and multicounty or regional blocs to deliver population health and personal health services to the local population.

True to the IOM’s vision of public health comprising multiple sectors, alternative organizational structures for public health have been suggested over the past decades.  Public-private partnerships, including partnerships between public health agencies and managed care organizations, have been implemented and promoted over the past ten years.  The Turning Point Program, an initiative of the Robert Wood Johnson Foundation and the Kellogg Foundation, has created a network of public-private collaborations in 23 states and 41 community-level organizations with the goal of transforming and strengthening the public health system (see http://www.turningpointprogram.org).  In addition, there is a national network - the National Network of Public Health Institutes -(http://www.nnphi.org/) that supplements and complements the governmental public health infrastructure in 25 states.

There is great variation in how public health systems are organized across states.  Although the federal government retains authority to protect the country’s health and welfare, primary responsibility for protecting the public’s health resides largely at the state level.  The federal government delegates authority to the states to organize and deliver public health services.  Every state has an agency designated as responsible for public health.  In some states, this is a freestanding agency, an agency that functions on its own; in other states, public health is part of a state’s health and human services department (known as an umbrella organization). States are evenly divided into umbrella and freestanding state organizations (ASTHO [Association of State and Territorial Health Officials], 2004).

Relationships between the state agency and local health departments (LHDs) also vary across states.  Some states have a more centralized organizational structure, with the state agency having direct control over the state’s local public health services; others are decentralized in their organization, giving local health departments more discretion over decisionmaking and service delivery.  A 1998 National Association of County and City Health Officials (NACCHO) survey of state-local health department liaisons characterized states by their level of centralization as centralized, decentralized, or mixed/shared authority (NACCHO, 1998).  By this typology, which is currently in the process of being modified, 13 states are centralized, 26 states are decentralized, and 11 states are mixed (the District of Columbia was not characterized).   LHDs also vary in their jurisdictional size and responsibilities.  Some LHDs have jurisdiction over cities or individual counties; others are more regionalized and have jurisdiction over multicounty


blocs within states (Association of State and Territorial Health Officials [ASTHO], 2004; Wall, 1998).

In several important respects, the public health system, generally speaking, resembles what the private health services delivery system looked like 20 to 30 years ago, when there were enormous variation in the quantity and quality of care delivered to patients with similar clinical conditions and few mechanisms to hold providers accountable for both the costs of care and patient outcomes.  Currently, there is considerable variation across state and local health departments in meeting preparedness objectives, and accountability challenges persist at all levels of government.  These challenges are exacerbated by the lack of evidence-based measures of public health system performance in general and of public health preparedness in particular.  Moreover, the lack of evidence-based research on such systems means that there is limited consensus on how state and local health departments might optimally be organized with respect to public health preparedness and on how the relationships between such departments should be structured.

Improving our understanding of how the public health system should be organized is important because the system’s organizational structure (i.e., whether freestanding or part of an umbrella organization, whether centralized, decentralized, or mixed) may explain differences in public health preparedness outcomes.  How the system is structured influences practitioners’ abilities to respond and the system’s capacity to adapt to changing circumstances.  In view of the scarce resources policymakers have allocated for public health, it is essential to have a structure in place that most appropriately and efficiently responds to preparedness needs - especially at a time when the public health system is expected to incorporate multiple mandates.

Bioterrorism Funds and Public Health

Following the September 11, 2001, attacks and subsequent anthrax attacks, Congress passed an emergency supplemental appropriation that included $900 million for bioterrorism preparedness, to be distributed by the Centers for Disease Control and Prevention (CDC). Beginning in 2002, a program of Cooperative Agreements was made available to state and four local health departments (the funding sent to the state is intended to be applied to local areas as well) to help strengthen capacity to respond to bioterrorism and other public health emergencies.  The funding for bioterrorism preparedness has continued from 2002 through fiscal year 2006, with approximately $1 billion per year allocated to state and local health departments.

  In May 2005, U.S. Department of Health and Human Services (HHS) Secretary Michael Leavitt announced that HHS disbursed another $1.3 billion to continue bolstering emergency preparedness levels around the United States. The 50 states and eligible territories received funding through two separate but interrelated cooperative agreements.  CDC provided $862.8 million for strengthening public health preparedness to address bioterrorism, outbreaks of infectious diseases, and public health emergencies.  The Health Resources and Services Administration (HRSA) provided $471 million for states to develop hospital surge capacity to deal with mass-casualty events.  These funding streams represent an unprecedented infusion of money into the development of public health infrastructure and hold the promise of making a significant contribution to transforming the system’s efficiency and effectiveness.

Purpose of study and specific objectives

In light of the new infusion of funds for bioterrorism preparedness and the considerable variation that exists across states and localities with respect to public health organization, we examined the relationship between public health structure and preparedness. The findings of our investigation might inform decisions about how states and localities structure their public health emergency preparedness activities.  Our specific objectives are to explain the factors influencing how and why states organize their public health systems in particular ways and how they have arrived at the various types of relationships that exist between state and local public health departments.  We also analyze alternatives to the status-quo structure of public health.  In particular, we address variation in the organization (i.e., the structure and function) of state and local public health departments and the governance relationships (i.e., authorities, funding, decisionmaking, among others) between these departments.  Finally, we identify strategies for improving public health preparedness outcomes.

As described in Chapter Two, we used a multipronged approach to address the study’s objectives.  We used the following research questions to guide our work:

• What are the key organizational factors in explaining public health preparedness outcomes?  What key trends have been observed since 2001?
• Why have states chosen to organize their public health preparedness efforts in particular ways?
• How can the public health system be restructured to achieve improvements in public health preparedness outcomes?
• What specific roles should federal, state, and local governments play?
• What lessons can be learned from other sectors?
• Do any states and/or localities offer promising models for organizing public health preparedness functions?
• Are there critical differences in the ways in which public health preparedness and other public health functions are (or should) be organized?

The remainder of this report is organized as follows.  Chapter Two describes our analytic approach, including a detailed discussion of our qualitative and quantitative research methods.  Chapter Three presents the results of our analyses, and Chapter Four summarizes key lessons learned and discusses the policy implications of our work.

Chapter Two.  Analytic Approach

We used qualitative and quantitative research techniques to address the questions posed in Chapter One.  Our study methodology consists of three integrated tasks.

First, we used multivariate statistical techniques to model the effects of various organizational structural characteristics on public health preparedness.

Second, we conducted a series of comparative case studies to determine how and why state and local health departments have chosen to organize their public health preparedness work - specifically:  how governmental officials at all levels have defined their roles in enhancing public health preparedness; how those roles have changed over time; the internal and external factors affecting decisions on how to organize the preparedness activities; gaps in preparedness activities; how those gaps are being filled - such as collaborations with the private sector; and how new models for state-local relationships are emerging.

Third, we conducted a synthesis of the literature on alternative governance structures (such as port authorities) and strategies (such as regionalization) that may offer ways of providing public health services, especially those related to public health preparedness.  We also reviewed documents describing health department organizational structures.

In addition to the project’s main analytic tasks, which are described in greater detail below, we interviewed key national, state, and local policymakers and stakeholders (both public and private) to ascertain their views on how the public health system can meet the preparedness challenges.  These interviews were designed to learn how stakeholders view the current public health preparedness efforts, what changes should be made, commonalities of organizational structures (such as the two typologies described in Chapter One), emerging state-local preparedness models, and the ways in which federal governmental initiatives contribute to improving integration of preparedness and response activities.  The interviews also helped identify case-study sites.

Quantitative Analysis of Public Health Structure on Preparedness Levels

Variation is to be expected in public health structures, and it creates an opportunity to examine empirically whether public health emergency preparedness (PHEP) varies with public health departments’ organizational structure and, in particular, the relationship between state and local health departments.   Public health emergency preparedness is defined by the U.S. Centers for Disease Control and Prevention as the capacity to prevent, detect, investigate, control, and recover from a large-scale outbreak of a natural or man-made pathogen (CDC, 2005).  The specific elements of structure included in this analysis are (1) degree of centralization, (2) regionalization, and (3) whether the state’s public health department is part of a larger umbrella agency or exists as a freestanding agency.  Given that the nation’s focus on bioterrorism and other aspects of PHEP is fairly recent, it might be particularly important to examine the institutional structures that guide the development of the capacity to respond to public health emergencies and the actual deployment of assets in response to an emergency.

Our analysis relies on data provided by CDC and a number of public health organizations that track, study, and seek to develop PHEP capacity among the nation’s state and local health departments. Before describing our data and analytic techniques, we consider how and why the elements of structure that we address might influence levels of PHEP.

Centralization

Centralization is the extent to which decisions are made by a relatively small number of individuals, groups, or organizational levels (Hendrick and Kleiner, 2001).  In most instances, centralization implies that decisions are concentrated near the top of the organizational hierarchy.  Yet, centralization might also be compatible with delegation of decisions to lower-level units, particularly where the “principals” who delegate such discretion maintain some measure of control over the “agents” to whom discretion is delegated.  The mechanisms by which principals maintain control might include providing performance incentives, specifying hiring criteria, or otherwise influencing selection of lower-level employees; oversight; training; and task standardization (see, e.g., Miller, 1992; Moe, 1984).  Thus, centralization is a multifaceted concept that can be embodied in many different ways.  In some cases, centralization might denote situations in which local officials are formally employed by a state agency.  In other cases, local officials might be formally independent but, for all intents and purposes, quite constrained by policies and decisions made at the state level.

A considerable body of organizational and institutional theory provides reason to expect a linkage between degree of centralization and level of preparedness.  First, centralization is often thought to be desirable when a comprehensive perspective is required to execute tasks (see, e.g., Hendrick and Kleiner, 2001; Fukuyama and Shulsky, 1997).  For instance, public health surveillance systems will be more effective if composed of large networks of laboratories and human sensors - especially when, as is often the case, detecting a bioterrorism or other threat depends on observing anomalous patterns that are evident only in large samples.¹  Similarly, centralization might minimize coordination problems.  The central direction of resources and staff, for instance, might speed the process of mobilizing staff for a large-scale response by avoiding the need to negotiate agreements between state and local authorities.  Finally, to the extent that tasks are routine (e.g., handling biological samples), certain economies of scale might be exploited by centralized structures.

Decentralized structures, on the other hand, offer greater flexibility in adapting to local circumstances, which may call for very different preparedness strategies and responses across locales.  However, the desirability of decentralization might depend on the level of skill, appropriate motivations, and resources possessed by local units.²  Thus, decentralization might be a viable strategy in states with well-staffed local health departments, but not in other states.  And even in the former states, the effect of centralization as opposed to decentralization might depend on the strength of formal and informal networks among the local health departments.

Regionalization

Regional approaches to public health preparedness can be implemented both within and across state boundaries.  However, in this study, we confined our analysis of regionalization to formal, within-state arrangements.  In some respects, regionalization might be viewed as a lesser degree of centralization-essentially the existence of multiple nodes of centralization in a system.  Thus, many of the same arguments should apply to this structural element.  However, regionalization might have the advantage of being perceived as more palatable in states with strong traditions of local autonomy.  In some instances, regional structures have been initiated at the grassroots level by local health departments.

Freestanding Versus Umbrella Agencies

A final structural element we consider is whether the state’s public health agency is part of a larger, umbrella agency or is a freestanding agency.  As with the arguments about centralization, umbrella agencies might, by virtue of being integrated with other agencies, be more capable than freestanding public health agencies of providing a comprehensive perspective on problems.  Additionally, economies of scale might accrue to umbrella agencies, because member agencies might be able to share the expenses associated with certain functions.  There might also be spillovers from one agency to another, in much the same way that studies of firm structure have found that innovations are more likely to transfer from units within a common firm than between other locations (Darr, Argote, and Epple, 1995).  On the other hand, umbrella agencies might have less autonomy to pursue public health-oriented missions and find their fiscal and staff resources diluted by membership in a larger agency.

In short, organizational theory provides reasons to expect a relationship between public health structure and public health emergency preparedness.  Yet, as the arguments above make clear, there is reason to think that the relationships might depend on the nature of the particular preparedness tasks in question and on the characteristics of the external environment.

Dependent Variables

The first requirement of a study of the relationship between public health structure and public health emergency preparedness is a measure of preparedness (the dependent variable).  Good measures of preparedness, however, remain elusive, owing largely to the fortunate rarity of significant public health emergencies.  While quality in clinical care and production and safety processes can often be measured by examining patterns in large numbers of events, researchers seeking to study public health emergency preparedness must base their measures on small numbers of actual events or on more-frequent-although less-serious-proxy events.  After action reports based on exercises, drills, and actual events probably provide the richest and most nuanced view of preparedness.  Yet, we need a relatively consistent measure that represents variation across most or all states.  Given that after action reports provide information on very different kinds of exercises, and that there is no agreed-upon standard format for preparing them, we did not attempt to extract information from a large number of after action reports.  Thus, we had to rely on other sources.

To measure or assess state-level preparedness, we used three sets of indicators, described in the Independent Variables subsection below, that provide information on most states:  (1) the full-range of indicators that the CDC required grantees to report in 2004 in exchange for funding under the cooperative agreement; (2) a preparedness index that we created from a subset of CDC indicators, and (3) data on lab capacity provided by the Association of Public Health Laboratories (APHLs).  These indicators were used to calculate task-completion rates, which, in turn, served as our dependent variables.

Data Sources

To our knowledge, the most comprehensive and complete attempt to collect data on preparedness in all states comes from progress reports submitted by recipients of funds provided through the CDC bioterrorism cooperative agreements.  Twice each year, the 62 states, metropolitan areas, and territories receiving these funds self-report their preparedness on a range of critical benchmarks that cover seven Focus Areas, shown in Table 2.1.

Our analysis focused on structured items (e.g., those calling for yes/no or percentage complete responses) from the 2004 reports.  In the parlance of educational testing, structured items are akin to multiple-choice questions, and unstructured items are akin to “blue book essays.”  The 2003 CDC reports included 11 critical benchmarks.  However, few are clearly linkable to items in the 2004 reports.  To keep the task manageable, we restricted our examination to the end-of-year CDC progress reports.³  Data were extracted from reports housed at HHS and entered into an electronic database.

Table 2.1:  Focus Areas in 2004 CDC Progress Reports

With 186 separate items, the CDC reports represent something of a “kitchen-sink” approach to measuring preparedness.  Thus, we also developed a more focused measure of preparedness for use in this analysis.  This measure is based on constructs identified by RAND researchers, who used a modified consensus process, as central to public health emergency preparedness. We then identified items from the 186 CDC indicators that map onto these constructs.  The constructs are shown in Table 2.2, along with the number of CDC items.  A detailed list of all CDC items, including those used in the index, is provided in Appendix A.

Table 2.2:  Constructs and Items in Summary Preparedness Index

We also obtained data from a survey of public health laboratories fielded by the APHL.  This survey included items designed to assess states’ readiness to deal with bioterrorism and other public health emergencies.  We extracted a number of relevant items from this survey, which we used to construct a simple additive index of preparedness.  The items used called for simple yes/no answers.  The index, therefore, is simply the proportion of 22 relevant items on which the respondent answered yes.  Appendix B provides a list of the APHL survey questions used in our analysis.

Independent Variables

As noted above, we examined three elements of public health structure:  degree of centralization, whether a state has a regional public health structure imposed by the state (as opposed to one that develops “organically”), and whether the state health department is freestanding or part of an umbrella agency.  Degree of centralization (according to NACCHO, 1998) was measured using the following coding scheme:  1 = decentralized, 2 = mixed, and 3 = centralized.  Centralized states are those in which local health departments are a formal part of the state health department.  Decentralized states, by contrast, have local health departments that are independent of the state health department.  Finally, mixed states include those in which public health services are provided through some combination of activities on the parts of the state agency, the local health department, and possibly even local boards of health and other agencies in the health jurisdictions.

Data on whether state departments of health are freestanding or are part of an umbrella agency were obtained from the Association of State and Territorial Health Officers (ASTHO, 2004), and data on whether states employ a formal regionalization approach to organizing public health services were obtained from the states’ Web sites.

Analyzing the Relationship Between Structure and Preparedness

Using these measures, we employed a variety of statistical techniques to assess the extent to which there is an empirical relationship between the measures of preparedness and public health structure described above.  Given the small number of cases (i.e., 46 states), we relied as heavily on inspection of graphs as on formal statistical methods.  Statistical analysis began with graphical examination of bivariate relationships, but also included regression models for estimating the effect of each structure variable on preparedness, holding the others constant.

The regression analyses modeled the task-completion rate (TCRi) - the proportion of all preparedness tasks reported as “complete”⁴ - for state i as a function of centralization, freestanding or umbrella, and whether a state has regional structures.  Degree of centralization was represented by two dummy variables.  The first variable (MIXEDi) takes the value 1 if state i has a mixed structure, 0 otherwise.  Similarly, the variable CENTRALIZEDi takes the value 1 if state i has a centralized structure, 0 if otherwise.  Another dummy variable (FREESTANDINGi) takes the value 1 if the state health department is freestanding, 0 if it is part of an umbrella agency.  Finally, the variable REGIONALi takes the value 1 if the state employs regional structures, 0 if otherwise.  These definitions yielded the following basic specification:

                                                            (2.1)

where b0 is a constant, b1 through b6 are slope coefficients, and ei is an independently and identically distributed Gaussian disturbance term.  We also included two measures of size as control variables, since providing effective coordination among state and local entities might depend on the overall scale of the intergovernmental system.  The variable LOGPOPi is the logarithm of total population of state i for 2004, and the variable LOGDENSITYi is the logarithm of the number of people per square mile for state i in 2004.⁵  Both size variables are centered at their means.  Thus, the intercept term b0 provides the mean task-completion rate for states of average size and density, and with decentralized, nonregionalized, umbrella structures.

Given the small number of cases, regression findings are likely to be highly sensitive to outliers.  Thus, we used robust regression and iteratively reweighted least squares (IRLS) estimation to reduce sensitivity to outliers and the consequences of violations of standard regression assumptions.  In light of the discussion above, it seems likely that any structure-preparedness relationships might be conditional on task and contextual conditions.  Thus, where possible, we sought to allow slope-coefficient estimates to vary by task or contextual contingencies.  Because of the small number of cases, we did not estimate interactions with tasks.

Given that structure varies by state, the unit of analysis in the regression models presented below is the state.  As noted above, to make the reports comparable and to simplify the analysis, we treated preparedness as a simple dichotomy:  “complete” or “not complete.”⁶

Case-Studies Approach

In consultation with the project officer and various organizations representing state and local public health departments, we selected ten state and six local health departments for in-depth case studies.  Criteria for site selection included geographic diversity, diversity of organizational structures, and states or localities that have a reputation for creating innovative public health policies and programs. The states we visited are Florida, Hawaii, Massachusetts, Nevada, Texas, Oregon, New Mexico, New York, Ohio, and Washington.  The local health departments included two rural locales and four urban locales (local sites are not identified, to protect confidentiality).  To achieve economies of scale and reduce the burden of site visits on health department officials, we coordinated selection of local health departments with other tasks that are part of the larger project.  Thus, the sample of local health departments was primarily a convenience sample with no specific selection criteria.

Data Collection

Within each state or local site, we conducted between six and ten in-depth interviews with key informants who have decisionmaking authority within the organizations.  Such informants included state officials responsible for organizing public health preparedness efforts, the chief health officer of a local health department, and state or local political leaders.  We developed and used a semi-structured interview protocol to conduct each interview.  The interviews were designed to obtain information on the following:

• Why and how state and local public health organizational structures evolved the way they did
• Perceived internal and external barriers to organizational change
• Perceived opportunities for organizational change
• Previous attempts to change the organizational structure and subsequent evaluations
• Strategic planning under way to devise alternative structures
• Perceived feasibility of the alternatives established through our literature review of alternative governance structures and strategies
• The relationship between statewide and local responsibilities for projects
• The relationship between health departments and state, regional, and local emergency management functions
• Roles played by other local stakeholders (particularly private-sector actors) in preparedness activities
• The legal and political relationships of local public health departments to the state public health department
• Whether local public health departments have autonomy and ability to promulgate and enact local ordinances to address preparedness issues or whether state law preempts local responses
• Suggested modifications to existing legislation to facilitate change.

In addition to the interview data, we collected relevant documentation from each organization we visited, including analyses of the public health system’s structure and organization, reports and brochures describing state and local public health initiatives, and descriptions of recently implemented programs.

Data Analysis

Our analysis of the site-visit data attempts to synthesize the documented evidence and interview data to portray the current organizational strategies being developed, along with options that public health systems might consider adopting.  The analysis focuses on identifying common themes across sites.  The primary form of analysis is descriptive, comparing and contrasting information across sites along several dimensions of interest, including the nature of the strategic adaptations considered and implemented; perceived successes and failures; organizational characteristics associated with these strategies; and financing implications.

Review of Alternative Governance Structures and Strategies

Although the case studies provided insight into ways of improving public health organizational structures and processes, we examined governance structures in other sectors that we believe may provide lessons for public health.  Moreover, we thought that it was important to examine these sectors because a significant barrier to considering new organizational forms for the public health system is simply the lack of empirically tested alternative models to consider.  Despite the limited existing organizational research comparing the U.S. public health system with other social service or economic systems, several other public and private systems in the United States share organizational features and difficult challenges with public health.  In the subsections below, we describe the criteria we used to select sectors for study, as well as our strategy for identifying and reviewing the relevant literature.

Selection Criteria

To select cases of alternative systems, we considered three criteria:  a public-private continuum; multiple missions; and regional strategies.  From an organizational perspective, these three dimensions capture some of the key issues a public health system would need to address in considering alternative ways of structuring public health delivery.  Because public health is a service, not a consumer good, we selected from other systems that deliver a service and not a specific good (such as a retail product) to enhance comparability.  Table 2.3 compares the four selected alternative systems - public education, welfare, banking, and port authorities - with public health across the three selection domains:

Table 2.3: Comparison of Systems by Selection Criteria

a In rural areas only.

• Public versus private.  Public health has been under pressure to adopt privatization or market competition as an approach to organizational reform.  By privatization, we are referring to the transfer of decisionmaking authority, delivery, or financing from a public to a private entity.  Public-private partnerships are growing increasingly common.  We thus chose alternative sectors characterized by a mix of public and private services.  For example, public education is predominantly a public service, delivered by a combination of federal, state, and local (county) governments.  Also as with public health, private reforms are often introduced into public education.  Such reforms include vouchers for individuals to attend private schools, creation of charter schools, and private management of public school systems.  The welfare system is predominantly public, whereas port authority systems are usually public-private hybrids, with the public regulatory mechanism monitoring the private sector’s operational responsibilities.  For contrast, we selected a sector, banking, that is delivered on a predominantly private basis, although there is considerable public regulation of banks. 
     
• Multiple missions.  The public health system is characterized by multiple, sometimes competing, missions.  In particular, the public health system is expected to assume a bioterrorism preparedness function while delivering the three core public health functions of assessment (surveillance), policy development, and assurance of services.  Traditionally, the public health system has also assumed a safety-net function, providing care of last resort to uninsured and underinsured populations.  Therefore, we selected other sectors that equally face multiple missions.  The public education system is expected both to convey knowledge and to contribute to future U.S. citizens’ socialization and citizenship processes.  The welfare system provides employment, self-sufficiency development, and aid to low-income individuals.  As with public health, welfare agencies monitor policy adherence, along with service quality and effectiveness.  The port-authority system simultaneously promotes the local or regional economies by facilitating trade and assures the
  safety of the port facilities.  And banking serves a range of financial services too numerous to list.
   
• Regional strategies.  One possible approach to reform is regionalization of public health services, either within states or among states (e.g., a Great Lakes Public Health System model).  Because of our interest in regionalization as an organizational reform to the public health system, we selected systems in which regionalization may be a viable method for service delivery.  Since port authority systems span state and local borders as a result of the naturally occurring distribution of ports and waterways, this system provides information on how services might be organized within and across states.  Similarly, banks span state borders, providing banking services in certain regions or nationally.  Here, it should be noted that both intra- and interstate regionalization strategies are potentially applicable to public health system reform.

Literature Review 

For each of the five systems, we conducted a thorough review of published studies in the scholarly literature about organizational structures and/or features and their relationship to performance indicators.  We also searched the Internet for reports (such as published reports about No Child Left Behind or Temporary Assistance for Needy Families) from national research and policy organizations.  For the port authorities’ case study, we combined a review of published literature with assessment of port structures and activities from Internet sites for individual ports.  (For this report, we use the terms port authorities and ports interchangeably.)

We focused the literature review around several key concepts, divided into descriptive features and organizational features of the system.  To organize our data, we constructed two matrices, following the methodology of Miles and Huberman (1984).  The first matrix (see Table 3.4 in Chapter Three) displays information about descriptive features of each system, including what services were provided, the missions, the consumers of the service, the public or governmental responsibility, the “visibility” of the system, whether the system has strong advocacy support, quality issues, and characterization of disparities and/or inequities within these systems.  The second matrix (see Table 3.5) displays information about the organizational features of the system, including the traditional organizational structure, geographic distribution, accountability for performance, professionalism, the relationship of the system with the community, notable reforms to the organizational structure, structural factors related to improved performance, and barriers to change. 

For each system studied, we analyzed the descriptive characteristics that were similar with or different from public health.  We then assessed which approaches to reform or organizational features work for each system and which do not seem to work, incorporating organizational theory into our analysis.  Finally, we looked for crosscutting themes and considered which ideas might be transported into public health.

Chapter Three.  Results

In this chapter, we present the findings from our multivariate analysis, case studies, and literature review on alternative governance structures and strategies, respectively.

Quantitative Analysis of Effect of Structure on Preparedness

As noted above, theories of organization suggest that centralization might be desirable when, among other things, there is a need for a comprehensive perspective on a task or problem, economies of scale can be achieved, and the task at hand is relatively routine.  However, given that preparedness involves a wide variety of tasks - each with potentially different organizational requirements - we might expect that centralization would not have much of an effect when all of these tasks are aggregated as a single indicator of preparedness.  This is exactly what we found.  Figure 3.1 provides box plots of state-to-state variations in the three preparedness indices according to whether the states are classified as decentralized, mixed, or centralized. Although there are differences in the median values across the three groups, they are small relative to the amount of within-group variation.  Indeed, analyses of variance models suggest that whether a state is centralized or not explains almost none of the observed state-to-state variation in aggregate preparedness indices.⁷  Adding the other structure variables - freestanding, regional - does not change this basic picture.  As is evident from the regression estimates reported in Table 3.1,⁸ the mixed and centralized states each reported task-completion rates of approximately 3 percentage points higher than decentralized states, but the differences were statistically indiscernible.

 Figure 3.1:  Preparedness and Degree of Centralization

Table 3.1:  Robust Regression Results (standard errors in parentheses)

NOTE: ^a Four states were missing data on the “regionalization” variable.

R2 estimates are based on analysis of variance models.

The analysis yielded similar findings for whether the state health system is regionalized.  As is evident in Figure 3.2, regionalized agencies reported slightly higher task-completion rates.  Once again, however, the differences are small relative to the amount of overall state-to-state variation.  The multivariate regression models present a picture similar to that for centralization, with small but statistically indiscernible differences between the two groups of states, even after holding other differences constant.

 Figure 3.2:  Preparedness and Structure of State Health Department

Finally, Figure 3.3 shows that whether a state health department is freestanding or part of a large umbrella agency it does no better in explaining variations in self-reported preparedness among the states: Any differences across the groups are dwarfed by within-group variation.  Once again, the multivariate analysis yields similar findings.

Figure 3.3:  Preparedness and Regionalization

As noted above, it is possible that the added value associated with various elements of structure might depend on the nature of the task at hand and on features of the broader context.  For instance, we might expect centralization or regionalization to explain more of the variance in task-completion rates when there are fairly clear economies of scale, as for education and training, for which states might produce standardized materials that could then be distributed to and adapted by local authorities.   Similarly, we might expect similar economies of scale in Health Alert Networks⁹ given that the utility of such networks is directly related to their scope and inclusivity.  Thus, we analyzed the data for each of the CDC Focus Areas shown in Table 2.1.  The more granular findings are presented in Table 3.2.  Structure explains more of the variation than before, especially with respect to education and training (Focus Area G; R2=0.22) and Health Alert Network (Focus Area E; R2=0.26).  However, the coefficient estimates do not suggest any clear advantage to centralization in these Focus Areas.  Thus, there are no clear interactions between structure and task characteristics.

Coefficients on the specific models do not provide a clear picture of what types of structures might be most effective.  States with mixed intergovernmental structures reported task-completion rates of approximately 13 percentage points higher for the Health Alert Network Focus Area.  But there was no such advantage for mixed states for education and training activities.  Overall, however, these findings are suggestive, but far from conclusive.¹⁰

Table 3.2:  Robust Regression Results by CDC Focus Area (standard errors in parentheses)

NOTE: ^a Four states were missing data on the “Regionalized” variable.

R2 estimates are based on analysis of variance models

We also considered the possibility that the effect of any one element of public structure might depend on other elements.  To operationalize this possibility, we created separate box plots for regionalized and nonregionalized states (the small number of cases [i.e., 46 states] made regression models with interactions impracticable).

The left panel in Figure 3.4 shows the preparedness-centralization relationship in regionalized and nonregionalized states using all CDC items as the measure of preparedness.  The panel on the right shows the same relationship using the APHL index as the preparedness measure.  In both panels, centralized systems are associated with the lowest preparedness levels in non-regionalized states but the highest preparedness levels in regionalized states.  These findings are confirmed by a regression model (see Appendix C), which included interactions between the regionalization variable and the mixed and centralization variables.  Using the CDC data as a dependent variable, we found that the performance advantage for states that are both regionalized and centralized is statistically significant (p = 0.01 and 0.07, respectively).¹¹  One possible explanation for this finding is that centralization might be more effective if accompanied by regional structures that help ensure the development of less-formal relationships among state and local officials.  More generally, formal centralization might have little effect without other mechanisms designed to foster more routine, face-to-face collaboration and communication.

 Figure 3.4:  Effect of Centralization in Regionalized and Nonregionalized States

KEY: 1 = Decentralized   2 = Mixed   3 = Centralized

Results from Case Studies of State and Local Public Health Departments

Organizational Structure of Case-Study Sites

The states and locales we selected for case studies represent a diverse range of geographic locations and public health organizational structures (Table 3.3).  We visited four centralized states, five decentralized states, and one mixed state.  Having visited only ten states total, the site visits should not be viewed as a comprehensive study of states and local health departments but, rather, an investigation of sample states for potential trends in health departments related to structural organization.  In this section, we present observations from visits to health-department sites and discuss trends found within centralized and decentralized state structures, as well as trends found across the various organizational structures. Finally, we discuss observed barriers to and facilitators of change within health departments.

Centralized States Included in Site Visits

Generally, in states with centralized public health structures, the state health agency or board of health has direct authority over local health departments (NACCHO, 1998).¹²  In all of the centralized states we visited, local health department staffs are state employees.  However, the similarity ends there.The four states, which are described below, are unique in size, population, geography, and future risks for types of public health emergencies.

New Mexico is divided into five health districts created by the state department of health, and health offices are located in each of the 33 counties. The state funds all public health employees at all levels (state, district, and county), and county health offices are facilitated through the larger health districts.

 Nevada has only three local health districts, all of which function with a considerable degree of autonomy and provide public health coverage for the more-populated portions of the state.  The remaining rural areas that are not included by a local health district receive services directly from the state health department.

  Florida and Hawaii represent different approaches to one of the challenges that a state with a centralized public health structure faces:  balancing local ownership with state-level support.  In Florida, the public health system is made up of 67 county health departments that report to the state as part of a state-county contract.  However, day-to-day public health activities are managed at the local level.  In contrast, Hawaii is highly centralized and has only four district health offices.  The state health department is located in Honolulu, on the island of Oahu, and District Health Offices are located on the other major islands--Kauai, Maui, and Hawaii.  Public health policy and budget decisions are made at the state level, although local public health officials are encouraged to implement policies in ways that are best suited for them. Local residents view these officials as the “local health department.&rdquo

Decentralized States Included in Site Visits

There is an equal amount of variation across the five states we visited with decentralized public health structures.  In decentralized states, local governments have direct authority over local health departments, with or without a board of health.  The local health department staff is generally hired at the local health department level; in some cases, they are hired at a regional level.  In New York and Oregon, local health departments are organized for the most part along county lines.  New York has 57 county health departments, 36 of which are considered “full-service,” meaning that they are capable of conducting all public health services, including environmental health.  Twenty-one of the 57 county health departments are considered “partial-service” and are supported by nine state district health offices.  The “partial service” health departments operate much like the rural areas in Nevada, both relying on the services and support from the state health department to provide the same services as the “full-service” departments.

Oregon has 34 local health departments, but these local health departments have levels of functionality that differ with the population of the county and therefore the amount of funding received from the state.  In the state of Washington, there are 35 local health jurisdictions, each governed by a local board of health.  In contrast, in Ohio there are 137 local health jurisdictions: Some are city-level, some are county-level and some comprise multiple counties.  Unlike the other decentralized states we visited, Massachusetts is structured uniquely.  There are local boards of health in each of 351 municipalities¹³ in the state, no county-level public health departments, and no county-level government structure.

Unlike centralized states, the decentralized states had a higher degree of autonomy to hire staff and could more easily allocate funding according to local needs. However, with a greater degree of local control, coordination within the state (i.e., redirecting resources or providing standardized training across a state) can become more challenging.

The extent to which local health departments rely on the state for technical assistance, templates, and personnel varies to a greater degree in decentralized states than in centralized states.  For example, in a centralized state, many directives on technical assistance and training would apply to the entire state, regardless of the size of the local health department; however, in decentralized states, large cities do not need to rely as heavily on state assistance because resources have been apportioned to address these issues, and local health departments in home-rule¹⁴ states are expected to be self-reliant.  Rural areas of decentralized states present yet another view: Local health departments may be as reliant or more reliant on state assistance than the local health departments in centralized states. Usually, local health departments in decentralized states receive a portion of their funding from local entities - county, city, or town government - and thus are accountable to those entities, as well.  Federal funds for public health activities are typically funneled through the state to local health departments, although some local health departments can apply directly for federal dollars for particular programs.  In some decentralized states, local health departments also receive limited per-capita funding from the state.   Although the local health departments in these states are autonomous, public health activities can be heavily influenced by state policies and, in some cases, regulated by state public health laws.  For example, New York is a home-rule decentralized state.  Although county health departments in New York have significant autonomy in day-to-day public health activities, the State Health Commissioner’s powers are broad.  In addition, the state determines public health regulations and policies, which local health departments are responsible for implementing.  Moreover, the state health department reviews the hiring of county-level health directors, and these local health directors are considered to be “deputies” of the State Health Commissioner.  This is one way in which the State Health Commissioner has oversight of local health departments, especially in emergency situations.

Mixed State Included in Site Visits

In those states in which the public health structure is considered “mixed,” local health services are provided by a combination of the state agency, local government, boards of health, or health departments in other jurisdictions.   The public health system in Texas, the only mixed state we visited, is a combination of local-, regional-, and state-level health departments. The local health departments are either county- or city-level departments that staff and govern themselves according to federal guidelines and grants.  State employees staff the regional health services offices, and they deliver public health services to those areas that are not served by a local health department.  Since we visited only one mixed state, there is not enough information to make generalizations regarding trends in mixed health departments across the country.

Table 3.3: Organizational Structure of Case Study States

aThe extent to which decisions are made by a relatively small number of individuals, groups, or organizational levels.

bWhether public health functions are administered by a freestanding agency or are part of a single umbrella health organization.

cTraditional: States in which local health departments have jurisdiction over individual counties or cities or, in some cases, multiple county blocs.  Regional: States that are divided
into regional units (health regions or health districts) with local health departments operating within them.

Observations from Site Visits

Similarly to the quantitative findings, we were unable to discern any strong relationship between formal organizational structures and respondents’ perceptions of their level of preparedness. The success of public health emergency preparedness programs seemed to be more conditional upon the presence of other factors, such as personal relationships, forms of coordination across levels of government, effective leadership, and norms of cooperation, or other ties borne of past experience or other arrangements.  For example, in one centralized state we visited, the Emergency Program Manager at the state health department responsible for coordinating preparedness efforts of hospitals across the state was mentioned in every interview we conducted.  Repeatedly, we heard that his strong communication and coordination skills and ability to reach consensus were critical factors in the success of that state’s preparedness efforts in the past few years, particularly with regard to coordination among hospitals.  By contrast, in a different centralized state, local health officials claimed that the state’s leadership style and lack of cooperation on key preparedness issues were major barriers to coordination and hindered the overall strength of the preparedness efforts at both the state and local levels.

There was also variability both across and within states in the degree to which the state and local governments integrated their public health preparedness-related activities with other public health-related work.  There was limited support for the notion that some local public health departments were better able to integrate their preparedness and nonpreparedness activities than the larger state bureaucracies.  Although we did not study the barriers and facilitators to integrating preparedness and nonpreparedness activities, the smaller size of local health departments may facilitate their being able to adopt an integrated approach because smaller organizations generally have less differentiated structures (i.e., less departmental specialization) and staff is frequently required to adjust for multiple functions.

  Large state bureaucracies, on the other hand, may be more likely to have more-specialized structures and may be more difficult to change because of their sheer size.  For example, in one state in which the state-level health department was moving toward an integrated approach, there was a parallel organizational structure outside of public health that was designated to deal with “disasters.”  While the functions that such parallel structures provide may indeed be required for states to plan and respond to disasters, their existence points to the challenges that states face in trying to integrate public health preparedness and non-public health preparedness activities.  Caution needs to be exercised in generalizing these results, because integration was not systematically assessed in our study.  Future research should be focused on the organizational barriers and facilitators that state and local governments face in moving toward an integrated approach.

Structure may not have a main effect or account for much of the variation in preparedness between states.  Centralized, decentralized, and mixed state organizations are all working to modify state and local health departments in order to effectively address issues of public health emergency preparedness.  While each type of structure requires attention to slightly different aspects of the process, all three structures seem capable of preparing effectively for public health emergencies. Within centralized and decentralized structures (we visited only one mixed state and thus cannot draw conclusions regarding the degree of variation in such states), the case studies revealed a wide range of preparedness levels within health departments, regardless of the type of structure, a finding that is quite consistent with our quantitative results.

Public Health’s Changing Role in Emergency Preparedness and Its Effect on Organizational Structure

In many of our site visits, we heard that the national focus on bioterrorism and emergency preparedness since 9/11, and the subsequent influx of emergency preparedness funds, dramatically affected state and local public health departments in various ways.  Although the increased funds were welcome, many health departments found themselves with neither the organizational infrastructure to adequately manage and disburse the funds nor the ability to perform the new functions that were expected.  In all of the health departments we visited, some degree of restructuring was necessary--creating new positions, new programs, new departments, etc.  One state health department made major organizational changes that created tension within the health department and, according to some respondents, has hindered their statewide preparedness activities.  Other health departments reported tension within their departments because some key staff members felt that both the financial and the programmatic focus on preparedness has negatively impacted traditional public health activities.

Most public health departments were not accustomed to serving as first responders, and they were not structured in a way that would facilitate a quick, coordinated response.  Nor was the traditional first-responder community accustomed to public health’s new role.  In some communities, this new role created tension, especially since large amounts of funds were involved.  In one state we visited, turf issues with a critical emergency response agency were hindering final strategic planning for national stockpile distribution.  As one respondent indicated, “There was lots of collaboration until the terrorism money started to come in; the money is coming from several sources; it caused rifts and questions about who’s in charge.  This was an unintended consequence.”

We also heard from two different states that public health’s new role thrust them into a new light with hospitals and others in the community - that state public health departments are now viewed more as partners than as simply a licensing agency or as “big brother.”  However, this new role in the community has also created pressure on health departments to develop better communications capabilities.  One health director indicated that, as a result of public health’s changing role in the community, his department now needs a group of employees that can handle “tumult and confusion.”  This recent addition to the role of public health practitioners is congruent with findings from our work on the integration of public health and hospital preparedness programs (see Davis et al., 2006).   In that report, hospitals appear to have been more extensively involved in interagency coordination on preparedness and for a longer period of time than have local health departments.

Organizational Barriers to Improving Preparedness

In our case study visits, we asked respondents about what they perceived to be organizational barriers to improving public health emergency preparedness.  Although many of the responses were specific to the individual state or health department, some common themes emerged.  However, these commonalities tend to be related more to problems with public health infrastructure in general than to specific types of public health organizational structures.

The most common organizational barrier to improving preparedness that we heard about during our site visits was the shortage of public health staff.  As one respondent indicated, there needs to be enough personnel to maintain the focus on public health preparedness without “robbing” the other areas of public health.  Some of the staffing issues were a result of turnover.  In two of our case studies, turnover among the senior state-level staff was a major issue.  In one of these states, local-level respondents reported that turnover contributed to a lack of coordination between the state and local agencies on preparedness issues.  In the same state, the inability of the state agency to hire staff or fill vacancies as a result of state-level hiring freezes and overly burdensome bureaucratic processes for obtaining exceptions contributed to delays in federal grant monies for preparedness issues being disbursed to local levels.

A related issue is staff training and how best to institutionalize the changing role of public health within an agency.  For some health department officials, Incident Command Structures (ICS)¹⁵ and the notion that public health staff must be prepared to serve as first responders are entirely foreign concepts.  As one respondent stated, “…it takes time, training, exercises, and discussion.”  States - whether centralized, decentralized, or mixed - are struggling to incorporate sufficient ICS practice and training so that the health department staff would all be able to smoothly function in an emergency situation, since (unlike law enforcement officers) they are not accustomed to participating in emergency situations on a regular basis.  Although staffing, turnover, and training challenges were common themes among respondents when asked about organizational barriers to preparedness, we did not find evidence to indicate that these issues are related to a particular public health organizational structure.

Other organizational barriers to improving preparedness included lack of technological capabilities, especially in rural areas; the continued presence of “silos”; and outdated public health laws and regulations.  Again, these issues may be related more to the big picture of public health infrastructure across the country than to the structure of public health in any particular state.  In some states, laws that govern public health interventions were developed years ago and were based on organizational structures that are not conducive to current public health needs, especially with regard to emergency preparedness.  For example, in Massachusetts, local boards of health have jurisdiction over isolation and quarantine policy, but they do not have the staff or the expertise to make isolation and quarantine decisions.

Barriers to Organizational Change

In our case-study interviews, we also addressed barriers to organizational change that both state and local health departments are struggling with as they attempt to alter their public health structures.  Such barriers include poor communication within and between health departments; large bureaucracies that need to quickly move resources; turf issues confronted when coordinating across agencies or local health departments; and time and personnel required to implement changes in the system. Again, many of the responses were specific to the individual state or health department, but there were some common themes.  Some respondents noted poor communication within health departments, as well as between local and state health departments, as a barrier to organizational change.  One local health department indicated its frustration with poor communication from the state by claiming that they could “count on a weekly surprise from the state health department.”  Some of the respondents felt that communication problems were primarily the result of poor leadership.

Bureaucratic impediments, especially at the state level, were also mentioned as barriers to organizational change in public health--particularly with shifting resources.  In one decentralized state, limited financial resources at the local level were a major barrier to organizational change across the state.  Specifically, the state health department tried to regionalize testing for sexually transmitted diseases and tuberculosis; however, none of the communities in the regions wanted to contribute money.  A turf issue of “who’s in charge?” was a barrier mentioned by respondents in many different locations. Lastly, time and personnel were an issue in many locations in which health departments struggled to keep up with traditional health department functions and received little funding, but were expected to adjust to multiple new responsibilities for preparedness in a short amount of time.

Facilitators of Organizational Change

Despite these barriers, respondents were quick to identify those factors that, from their experience, are facilitators of organizational change.  We observed several facilitators across the sites we visited, such as strong leadership and successful regional coordination within states. Interestingly, although increased funding caused some difficulties, as noted above, it also cut the other way, because many respondents identified funding as the main facilitator of organizational change on the state level.  An underlying tenet of the CDC Cooperative Agreements is to “…upgrade and integrate State and local public health jurisdictions’ preparedness for and response to terrorism and other public health emergencies with Federal, State, local, and tribal governments, the private sector, and Non-Governmental Organizations” (CDC, 2005).  For some highly decentralized states, improved integration has been and continues to be a difficult but necessary struggle.  Given that decentralized states do not have authority over local health departments, organizational change cannot be legislated.  Consequently, the state health department may persuade local health departments to implement changes--e.g., regional approaches--by requiring such changes as a condition of receiving state funds.  According to some respondents, this has been a very effective method for facilitating change that otherwise would be extraordinarily difficult.

The importance of skilled leadership and good communication in facilitating organizational change, especially within health departments, is another common theme apparent from our case studies.  This facilitation often took the form of aggressive training programs for employees.  In one health department, we were told about a program promoting public health leadership and cultural change in which employees are taught how to manage meetings, manage performance, etc.  As one interviewee commented, this type of training can be helpful in promoting emergency preparedness in an organization.

Another facilitator of organizational change, which we heard about in more than one state, is the existence of regional alliances, community work groups, or coalitions that were formed to achieve certain goals, such as economies of scale, coordination of activities within a region, and sharing of information.  Such coalitions are not formal structures that the states developed but, rather, their formation was more organic and from the bottom up.  For example, in one decentralized state, downsizing state government in the late 1980s spurred counties to pool their resources and create their own surrogate regionalization in the form of alliances.  Even as state government and, consequently, the state public health department, recovered from this downsizing, the regional alliances have remained active and continue to effect change in the structure and function of public health in particular areas.  In fact, the strength of these alliances and their desire for individualized information technology (IT) systems was the impetus for a compromise.  The state provided support to the alliance by creating space on the main statewide system, thus ensuring that both the state and alliance IT systems are integrated and compatible while meeting the alliance’s individualized preferences and needs.  The state also helps to maintain integration by setting data standards. This well-integrated system has increased connectivity between state and local health departments, among local health departments, and between the various players involved in emergency preparedness (e.g., hospitals), thus enhancing communications capabilities during emergency situations.

There were other facilitators of organizational change that we heard about from a single state or local health department but that may be common to other states with similar characteristics.  For example, many respondents in one state felt that the relatively small size of the state was a critical factor in facilitating organizational change even at the legislative level.  One interviewee commented, “…you can reason with the legislators and get it through; it’s the smallness of the place.”  Similarly, in our report on integration of public health and hospital preparedness programs (Davis, 2006), we found that, in general, smaller localities with fewer stakeholders find it easier to coordinate on preparedness.  Another facilitator of organizational change that may be common to other states is past experience with natural disasters or other events that bring people together to work for the common good.  As one interviewee stated, “Coordination is great because of past events, exercises; we have no choice but to work together.”  Finally, and not surprisingly, involving those people who will be affected by change in the planning and implementation was noted by many to be a strong facilitator of organizational change.

The Relationship Between State and Local Health Departments

Regardless of whether a state is centralized or decentralized, there are often conflicts between state and local public health departments, especially with regard to the role each one plays.  In centralized states, conflict may be related primarily to the degree of autonomy that local health officials should have.  We heard repeatedly that, “all health is local in terms of immediate response” and that local health officials must have the basic skills to screen, triage, and know when they need additional assistance.  However, one centralized state we visited reported struggling with this issue.  On the one hand, the state wants to empower the locals and increase their ability to function independently.  Yet, on the other hand, some state officials feel that the expertise resides at the state level especially with regard to control of outbreak management.  In some cases, there have been significant differences of opinion between local-level and state-level officials on the appropriate course of action; since each situation varies, it has been difficult to formalize a procedure.

In decentralized states, the tension between state and local health departments is related primarily to the degree of guidance state officials should give to locals, and when and under what circumstances appropriate hand-offs should occur.  In rural areas of a decentralized state, the locals are often looking for more assistance and guidance from the state; whereas, in the larger cities with potentially more funding than at the state level, the cities are waiting for the state to catch up with regard to preparedness activities.  In two instances, health officials in rural counties in decentralized states expressed a need for more specific guidance from the state, especially in the form of templates for plans and documents.  Local officials reported that templates typically received from the state have been focused on large urban areas and are not applicable to small rural locales.  Yet, in some larger, more advanced local health departments, officials felt that the state health department’s role should be more policy-oriented and less directive.

Some respondents reported that there was poor coordination between state and local levels of the public health infrastructure.  In states that have highly decentralized organizational structures, the need for some forms of coordination across levels of government may be particularly important.  In strong home-rule states, however, introducing such coordinating mechanisms will be a challenge.  In the course of our case studies, there were two examples of major cities in strong home-rule states that had a reputation for rebuffing state involvement.  However, tension between state officials and health departments in large cities was also found in some centralized states.

The Relationship Between Health Departments and State, Regional, and Local Emergency Management Functions

In all of our case-study sites, we heard about the changing role of public health in relationship to other agencies and organizations involved in emergency preparedness and response.  At the state level, the influx of federal funds has thrust public health into a more central role in statewide emergency response efforts, regardless of whether they were centralized or not.  Although it has been challenging, respondents generally agreed that, since the events of 9/11 and the anthrax outbreaks, there is far greater involvement of public health at the local emergency-response level, as well.  Some reported that, in the past, public health typically did not interact with law enforcement, firefighters, or EMS.  With the CDC preparedness funds and the associated requirements for building community-level relationships, public health and other first responders have started to work more cooperatively.  However, public health’s quest for an equal seat at the table has been a difficult process that many in public health had little experience with.  Some public health officials with whom we spoke reported that they have had to learn to deal with very savvy players on the public safety side who are well organized and part of strong unions.  One respondent stated, “Keep relationships open with the mayor’s office and figure out the key people in the fire department.  Know who to make phone calls to and get a constituency before you walk into meetings.&rdquo

It is important to note that all of the foregoing observations in this subsection pertain to both centralized and decentralized states.

Roles Played by Other Local Stakeholders in Preparedness Activities

In the course of our case-study visits, hospitals were noted as major stakeholders in local emergency preparedness activities (sometimes hospitals are as large a stakeholder in emergency preparedness as are police and fire).  The relationship of hospitals and public health departments varies tremendously, but generally the relationships have been strengthened significantly as a result of the federal preparedness funds to both public health departments (through the CDC cooperative agreements) and to hospitals (through the HRSA cooperative agreements).  Over the past few years, CDC and HRSA have expanded their requirements that public health departments and hospitals work together with other members of the community on emergency preparedness activities.  These activities have taken various forms.  However, state and regional hospital associations are a strong collaborating force in at least three of the states we visited - Hawaii, New York, and Texas.  In other states, regional steering committees have also proven successful in integrating preparedness activities between public health and hospitals.  But from our perspective, there was no discernible relationship between the roles played by these stakeholders and any formal organizational-structure variables.

Lessons Learned on Alternative Governance Structures and Strategies From Other Sectors

In addition to relying on the case studies for insight into organizational changes that might lead to improvements in public health preparedness outcomes, we examined governance structures in four other sectors - public education, welfare, banking, and port authorities, as described in Chapter Two - with an eye toward identifying promising strategies from those sectors that could be applied to public health.  After providing an overview of the descriptive and organizational features found in each of these sectors, as well as in public health, we discuss key themes that emerged from our comparative analysis.

Tables 3.4 and 3.5 display descriptive and organizational features of public health and the four alternative sectors that we used to form the basis of our comparative analysis.  Our analysis suggests several important organizational themes that cut across the sectors.  Public health might learn from the structural features with which the other systems have experimented, including private-sector approaches, use of information technology (IT), accountability mechanisms, adaptation to the community and environmental factors, collaborations and networks, and enhanced professionalism.

Table 3.4: Comparison of Services by Descriptive Features

Table 3.5: Comparison of Services by Organizational Features

NOTES: CBO = community-based organization; FDIC = Federal Deposit Insurance Corporation; NCLB = No Child Left Behind Act; TANF = Temporary Assistance for Needy Families.

Private-Sector Approaches

Our cross-sectoral analysis revealed that each of the sectors studied has had a long history of experimenting with private-sector approaches.  There are many possible models in which public services can incorporate approaches from the private sector, with the expectation that doing so will make service delivery more efficient.  Public-sector services can incorporate market forces through arrangements with private not-for-profit or for-profit organizations, such as by contracting out services, using single-bidder or multiple-bidder approaches, and establishing other partnerships (Kettl, 2003).  Each system in our cross-sectoral analysis has been trying to determine the appropriate mix of public and private responsibility for specific partnership arrangements for ownership, service delivery, monitoring, and policy development.  For any service, the optimal degree of private responsibility depends on economic considerations, and also administrative and employee concerns, practical challenges, and public acceptance.

Despite these challenges, our cross-sectoral analysis suggests that benefits can be related to the introduction of competitive forces in these services.  In the port system, for example, increased competition has improved port efficiency in smaller cities.  In the ideal type of port, the public sector would control regulation and land ownership, concentrating on core functions, while the private sector manages operations.  Research suggests that quasi-public ports are likely to be most effective when the ownership and regulatory functions remain public and the public sector focuses on a few core functions, such as surveillance and traffic control.  In banking, competitive forces have spurred greater entrepreneurship and the creation of new organizational forms (such as cross-functional teams).

Our analysis of wide-ranging market-based strategies used in the sectors we studied implied an important lesson for public health:  There need to be criteria to help determine the public sector’s role - what can be shifted to the private sector (through contracts or partnerships) and what must remain public.  In education, contracting out services that do not have a direct relationship to the mission or values of public education (such as cafeteria, janitorial, and transportation services) is relatively uncontroversial.  But when new, private structures are introduced, as with vouchers and for-profit management of schools, criteria for the public role must be linked to the mission and core competencies of the public service.  In the welfare system, there is much variability across states in which services are contracted, although criteria to make this determination are beginning to be developed.  Potential balancing criteria for contracting out in welfare include costs, savings, technology needs, skill-set needs, and the effect on clients (which includes access, quality, and equity) (Bandoh, 2003).

In addition to various privatization approaches, there are other ways to incorporate concepts from the private sector into public health while preserving the public role.  For instance, there may be a role for a “public entrepreneur” in public health:  individuals who are challenged to create innovative, efficient solutions to public health problems within the governmental structure.  While entrepreneurship generally refers to innovative behaviors that result in profits (in a business model), this concept might be applied to public-sector employees as well.

One model of public-sector entrepreneurship, used in urban redevelopment, is known as municipal capitalism (Chapin, 2002).  In this model, the public sector plays a more active entrepreneurial role than in a managerial service delivery approach.  According to Chapin, in municipal capitalism “the public sector is now the lead player during the entirety of the redevelopment process, an actor as focused on the return of investment . . . as the private sector.”  Given the nature of public health functions, it seems highly unlikely that public health entrepreneurial activity will generate revenue.

Nonetheless, the port-authority literature provides examples of entrepreneurial activity in generating revenues within a public-private organizational structure.  Likewise, the banking system provides an example of how organizational structures can influence entrepreneurial attitudes.  When decentralized decisionmaking is encouraged, entrepreneurial and innovative activities result (such as cross-functional teams).  Continuous education of employees also helps to stimulate innovation.  These concepts should be incorporated into public health.  Decentralized decisionmaking (combined with enhanced training of public health professionals) should be encouraged when there is a need for local decisions that are tailored to individual situations or communities, because such encouragement could spark ingenuity.

Use of Information Technology

Our cross-sectoral analysis demonstrates that information-technology capacity has proven to be an essential component for facilitating decisionmaking in collaborations (real and virtual) in the banking, ports, and welfare sectors.  Fukuyama and Shulsky (1997) argue that, over the past 50 years, the major change organizations have faced has been the change from organizations that produce a product to information-based organizations that deliver services.  In the future, they suggest, information technology will be integral to organizational change toward flatter organizational structures dependent on efficient information flow.  In banking, information technology is required to make decisionmaking less dependent on proximity.  In ports, information technology is necessary to coordinate various partners.  In welfare, the development of information technology is required to integrate programs effectively.  Public education could benefit from better information technology for information dissemination.

In public health, enhanced information technology is consistently cited as a requirement for better surveillance functions, but it could also be used to facilitate collaboration.  As noted in CDC’s Focus Areas (particularly for the Health Alert Network) and in our case-study interviews, the development of information technology is a key element in successful bioterrorism preparedness.  What emerges from our interviews is the need to understand better how to develop sophisticated but user-friendly IT systems for public health departments.  Enhancement of IT will likely be an essential component of any reforms to the public health system.

Accountability

Broadly speaking, public health departments at all levels of government are subject to political accountability; but there is a notable absence of specific financial and outcomes-related measures.  In other words, what is lacking are a measure of whether the returns generated by the public health systems are worth the money invested and a system for linking any performance shortfalls to consequences, financial or otherwise.

Our review of alternative structures suggests the need across all sectors for better empirical evidence to assess performance and measure accountability.  Yet the accountability challenges of public service contracting are manifest in the sectors selected for study.  As the welfare system demonstrates, increased contracting of services introduces the need to evaluate and manage the contracts, including adding full-time staff dedicated to oversight of external contacts.  In public health, most staff members have been trained to organize and deliver services and may not have been trained to manage contracts or external relationships.  The trend toward contracting out suggests that the public health system must, as in welfare, devote staff members and resources to contract oversight.

Another accountability-related challenge of private approaches to public services is exemplified by school choice programs (e.g., vouchers, and magnet and charter schools).  For such programs to be effective and accountable there must be well-publicized information about each school option and its performance in a given region.  One barrier to the implementation of private approaches in education is a lack of data about quality, which may remain elusive because of not just the methodological complexity of the research but also the ideological commitments of researchers.  The experience in public education should provide a lesson to public health about the necessity, yet difficulty, of good evaluations of such private-sector approaches.

Thus, data-driven approaches to accountability are essential.  To implement such approaches well, performance measures for the public health system must be developed and tested, and better informatics must be developed to track performance indicators (Roper and Mays, 2000; Mays et al., 2004; Barry, 2000).  One might imagine implementing jurisdiction-level accountability changes for health departments that underperform on key statistics, such as the No Child Left Behind Act does.  For example, in public education, the superintendent might be removed from a consistently poorly performing district.  One might consider similar punitive actions in public health, such as removing the health officer from a health jurisdiction or department.  Another option that might be imported from NCLB is implementing a “coach” at a health department to provide guidance and consultation for improvement, the strategy that underperforming public schools in Michigan employed.  Although such actions are not standard for public employees, the NCLB model suggests that these governance changes are possible even within a public system.

Accreditation.Of course, the use of any standards or testing measures is predicated upon knowing what to measure and how to measure it, which is not yet clear in the case of public health (Roper and Mays, 2000).  In response, public health organizations, such as the American Public Health Association and the National Association of County and City Health Officials are currently developing a mechanism to accredit public health professionals.  Doing so will establish criteria and performance measures to hold public health professionals accountable.

Valuation.Another related aspect of accountability is to examine how to value the outputs of public programs.  It is much easier to estimate value of private goods and services than to assess the value-added of public goods and services.  For instance, there are methodological limitations to examining the value of preventing a flu pandemic.  Yet the port-authority literature provides some elements that might be extended, particularly to provide insight into the value of public entrepreneurship.  Understanding how to value public health services is important for the resource-allocation decisions facing public health practitioners.  As our case-study interviews imply, practitioners need to make choices on how to integrate bioterrorism preparedness with responsibility for ongoing activities.  Inevitably, the necessary trade-offs will be easier to make with more information about which functions and services deliver the greatest return on the public’s investment.

Applicability of Formal Organizational Structures Used in Other Sectors to Public Health

Our review of organizational structures used in other sectors also yielded information on some of the formal structural characteristics that we examined in the quantitative analysis and in the case studies:  degree of centralization and the regionalization of services.

Centralization-Decentralization.  Our cross-sectoral analysis cannot speak specifically to the level of centralization or decentralization that is desirable for public health, but it is clear that the optimal level of centralization will be contingent on the nature of the particular public health function or task.  Our review of the literature did not reveal any extant criteria regarding those preparedness tasks that might be centralized as opposed to those that should be decentralized, and, as indicated above, our empirical analysis did not provide any guidance in this regard.

Consistent with the case-study interviews, the cross-sectoral analysis suggests that decentralized structures are feasible for those activities that are based on community-specific needs that might not benefit from economies of scale.  But centralization remains salient for certain functions.  In education, centralized authority is required for developing standards that all students must meet to fulfill education’s function of providing societal benefits.  In public health, labs or epidemiological services that are specialized and applicable across a state should have centralized control because of their wide benefits.  Even so, the risk-communication function in public health preparedness contains elements of both.

Regionalization.  Regionalization of public health services is an idea generating considerable attention.  Researchers have proposed that public health services might be more effectively and efficiently delivered on a regional basis, merging counties or states into geographic regions linked by similar health status, economic, or geographic characteristics.  For example, the Mississippi Delta region might be more effectively served through a regional public health authority, rather than through county-level or state-level agencies (Mays and Halverson, 2005).  Our cross-sectoral analysis suggests that regionalization of public health services might be incorporated, but that this approach must be directly linked to public health objectives.  Both Michigan and Wisconsin have imposed regional bioterrorism preparedness structures on existing local functions.  Because the case studies were not designed to evaluate regionalization, our results are only suggestive that regionalization might be a useful structural mechanism.

However, although there is no reason to believe a priori that regionalization of public health services will always be the best approach for meeting public health challenges, it might be a promising strategy.  Small, less-affluent communities lack the resources to meet all of their public health needs (Mays and Halverson, 2005).  The current structure places too many responsibilities on local units with inadequate resources to address them.  Thus, combining resources across local units has the potential to provide more services through economies of scale and scope currently lacking in certain communities.  Further, many of the risks to which public health must respond do not respect political boundaries.  And a regional approach helps attenuate the manpower shortage facing smaller units.

Yet, there are two major impediments to a regional approach.  The overriding difficulty is political.  Local political units will lose the considerable control over resource allocation that they now enjoy (as suggested in our case-study interviews).  Losing control over public health removes one mechanism for demonstrating to constituents what the county officials have done for them.  The other difficulty is logistical.  Which counties should be grouped within a regional structure?  Combining the proper local units into a larger, regional entity is a solvable, but complex, issue (Mirvis, 2005).  Encouraging the development of informal, or organic, regional approaches may provide an attractive alternative to “top-down” regionalization approaches, as well as a means for dealing with these two difficulties.

Cross-Sectoral Barriers to Reform.Our cross-sectoral analysis suggests some persistent barriers to organizational reform within the public health system.  The most-cited barriers include prohibitive costs and time needed to plan and implement change, lack of adequate IT capacity, and lack of quality empirical data on performance factors.

Chapter Four. Summary and Conclusions

This analysis employed various data sources and analytic methods with an eye toward identifying relationships between public health department organizational structures and preparedness outcomes.  Our hope was that we could use the data sources and methods to triangulate on a clear set of recommendations regarding how public health systems should be restructured to increase their effectiveness.   However, as is often the case, reality is more complicated than we would like, and judging from our analysis, we are not able to recommend any wholesale restructuring of the public health system to better meet preparedness goals and objectives.  However, a clear set of lessons emerged that can and should be applied to public health systems, regardless of the formal structural characteristics that may exist.  Below, we summarize our key findings and discuss their implications for improving public health preparedness in state and local health departments.

Results of the Quantitative Analysis

This research sought to assess whether there is a linkage between degree of centralization in states and the degree to which states are prepared to respond to public health emergencies.  We defined structure in terms of three elements:  (a) degree of centralization, (b) degree of regionalization, and (c) whether the state’s public health department is part of a larger umbrella agency or exists as a freestanding agency.  Such structural features might affect the extent to which state public health systems can effectively coordinate the activities of their own and local jurisdictions’ activities in response to a bioterrorism event or other public health emergency.  To the extent that there is such a relationship, it might inform decisions about restructuring state public health structures as the nation seeks to improve its capacity to respond to such emergencies.

Inasmuch as we found evidence of empirical relationships between structure and preparedness, they were weak and conditional on other variables.  For instance, we found that in some of our analyses mixed and centralized states generally reported somewhat higher task-completion rates than other states.  But we must emphasize that the findings were hardly consistent.  Similarly, we found that states whose health departments are part of umbrella agencies and employ formal, top-down regional structures perform slightly better than others.  But in all instances, the differences across groups of states based on structural characteristics were small relative to the total amount of variation within any particular group.

Finally, it is possible that structure matters, but in ways not measured by the data available to us.  In short, there might be better and worse ways of operating centralized and decentralized structures.  Our tentative finding that centralization provides greater benefits in states with regional structures provides some insight into the conditional nature of structure.

Case Studies

As we embarked upon the case studies, we were interested in understanding how public health organizational structure affects the ability of state and local public health agencies to conduct emergency preparedness activities.  Our findings revealed wide variability in the organizational structures for public health.  Even within states that were centralized or decentralized, we observed a high degree of variation in how state and local levels were coordinated, and variability in how public health was organized at both the state and local levels.  This finding supports our quantitative results, which revealed greater within-group variation than between-group variation with respect to formal structural characteristics.  Moreover, much of what we found may be related more to broader public-health-infrastructure issues than to specific state or local public health organizational structure.  That said, several themes emerged from the case-study interviews regarding how organizational issues are related to preparedness.

In all but the most centralized states, there needs to be greater coordination between the state and local public health departments, and a greater specification and shared understanding of roles and responsibilities for preparedness activities.  We observed the formation of informal coalitions and working groups of local public health officers to fill this gap in some states, and we found these groups to be facilitators for sharing information among geographic regions within a state.

The challenges of recruiting and retaining qualified staff to work on public health preparedness was a strong and consistent theme of our case studies.  On the positive side, interviewees recounted stories of individuals who were able to work effectively regardless of the formal organizational structures that were in place.  On the negative side, other individuals told us about the effects of turnover and the inability, in some states, to fill vacant positions as a result of hiring freezes or burdensome bureaucratic processes around hiring.

In short, the case studies demonstrated that there is no clear recipe that states can follow: No one organizational structure was equally effective across all states.  At the same time, the case studies provided evidence that effective leadership, good communications and IT systems, and an adequate number of staff as well as staff training were critical ingredients for a well-prepared state or locale, regardless of organizational structure.

Lessons from Alternative Governance Structures and Strategies in Other Sectors

The most important policy implication from the analysis of other organizational structures is that there is no obvious alternative organizational model that could be the basis for a redesigned public health system.  What emerges, instead, is that there are some innovative ways of thinking about public health delivery that deserve further analysis.  In particular, policymakers should consider the following.

Regionalization

Further examination of both inter- and intrastate regionalization as a means of organizing public health services is warranted.  Our review of the literature does not indicate that more-radical organizational alternatives are either available or feasible.  As an interim strategy that could well act as a catalyst for states to reconceptualize the delivery of public health services, therefore, we recommend additional exploration of this approach.  One of its advantages would be to stimulate states to reassess the issue of centralization versus decentralization.  In the process of devising a regional strategy, states would almost certainly examine which services would more appropriately be delegated to the regional structures and which would be best provided at the state or local level.

Defining Public Entrepreneurs

The concept of the public entrepreneur emerges most particularly in the literature on quasi-public organizational approaches across the sectors.  A “public entrepreneur,” in contrast to a traditional “public servant,” is somebody who develops innovative processes to achieve some public health goal or value - for example, improved public health preparedness.  One hypothesis with regard to organizational change is that a public health leader who is more “entrepreneurial” is more likely to embrace reforms.  Future research might assess the organizational capacity of the public health system to engage in entrepreneurial activities, perhaps considering the public entrepreneur leader as a variable facilitating change.  Research questions that might be addressed include the following:  What are the activities a public entrepreneur would pursue in public health?  Who are the public entrepreneurs of a health department?  What qualities distinguish somebody who considers innovative approaches within public health?  What value does a public entrepreneur work toward maximizing?

Privatization Criteria

Our analysis across sectors suggests the need for criteria to help public health professionals make important decisions about which public health services can be privatized and which should remain public.  Instead of individual health department leaders making these critical decisions in isolation, some broad criteria should be developed to help.  Risk represents one nascent criterion that might be explored.  Services that relate to risk (including bioterrorism and communicable disease) should be maintained by the public sector to ensure that there is adequate oversight of these services and to minimize transaction costs related to contracting these services out.  Services that involve routine and repeated events may well be shifted to the private sector and monitored by the local health department through computer-based systems (Kettl, 2003).

In any event, policymakers need to determine which functions must remain with the public health system and which might easily become private-sector responsibilities.  In so doing, they must be mindful of distinguishing between contracting out individual tasks or services and the wholesale privatization of critical functions or services, because each of those approaches has implications for oversight and accountability.  Future research might consider the process through which such criteria should be developed and consider some (preliminary) criteria through which public health professionals can determine what functions should be maintained under a public role.  Additional work is also needed to develop mechanisms to hold private entities accountable for their performance.

Limitations

Our analyses are subject to a number of important limitations, which should be kept in mind when interpreting our findings and considering our policy recommendations.  With respect to the quantitative analysis, and as mentioned previously, the measures of preparedness employed are limited for a variety of reasons, not the least of which is that there is no agreed-upon definition of preparedness.  Moreover, there are serious questions about the extent to which the self-reported data used to construct the preparedness indices represent valid indicators of capacity to actually mount a successful response. As noted in unpublished RAND research, rankings based on the CDC data are of questionable face validity given our experience in the field.  There, we speculated that well-developed understanding of the complexity of preparedness might bring with it a heightened awareness of one’s limitations and of what one does not know.  In other words, some of the best-prepared places could potentially rank themselves relatively low because their leaders realize that much more needs to be accomplished. Thus, opinions about the face validity of a given measure will inevitably broach disagreements about what counts as preparedness, as well as opinions about the adequacy of the measure as an accurate representation of those constructs.

Additionally, other aspects of the CDC grantee reports raise flags regarding the reliability of measures derived from them.  For instance, the reporting formats do not specify who is supposed to fill out each section of the reports.  As a result, at least some of the apparent variation in preparedness across states might be due to variation in the positions and perspectives of the individuals filling out the reports.  Similarly, grantees might have varying interpretations of the response categories.  For instance, one respondent’s view of the difference between whether an item is complete, partially complete, or incomplete might differ from that of another respondent.  Thus, variation in the reported data might reflect these differences, in addition to variation in true preparedness levels.

 A second limitation is that our measures of public health structure include only formal and institutional manifestations of structure.  It is well known that the actual operation of institutions depends on a host of informal norms, relationships, and practices that are not represented in our analysis.  As noted above, we speculate that regionalization might be a proxy for informal networks, but whether it truly is just that--speculation.

Another important limitation is that structure may be endogenous.  That is, it is possible, for instance, that a given state may have adopted a more centralized organizational structure if it had reason to doubt the ability of local health departments to mount an effective response.  This type of behavior, in turn, may call into question any attempt to draw causal conclusions from correlations between structure and preparedness.   Conceptually, the soundest method for assessing the causal impact of public health structure on public health emergency preparedness would involve random assignment of structures to states.  Under such circumstances, any observed differences among states with different structures could be attributed to variation in structures.  But it is very unlikely that structural elements are randomly distributed across the states.  Structure might change as a result of perceived deficiencies in performance.  If, for instance, reformers centralize states’ public health systems in response to perceived deficiencies, then the level of preparedness would determine the structure, rather than vice versa.  Moreover, states’ knowledge of or propensity to report performance deficiencies in surveys and CDC reports might be linked to structure.

The case-study analysis is also subject to a number of important limitations.  First, although we attempted to select sites for study that are broadly representative of all CDC public health preparedness grantees, we have no way of knowing whether we were successful in this regard.  As a result, we cannot claim in any rigorous sense that our findings and policy implications are applicable to all grantees.  Second, in conducting the site visits, we attempted to obtain multiple points of views and to synthesize what we heard to arrive at a reasonable portrait of how events unfolded, strategies were considered, and solutions implemented.  But we have not subjected our syntheses to any formal validity and reliability testing, mainly for practical purposes.  Finally, our case-study results are driven largely by the degree to which interviewees were candid in their responses to our questions.  Although we provided interviewees with strict confidentiality assurances prior to beginning our questioning, we have no way of determining the extent to which they provided truthful and accurate responses.

Our review of alternative governance structures and strategies contains many of the limitations found in other literature reviews.  For instance, despite our efforts and intentions, we cannot claim that we have captured and analyzed all of the relevant literature.  Second, it is entirely possible that other sectors may have provided important insights for public health, but budget and other constraints precluded us from casting too wide a net.  Finally, we chose to organize our review by sector, as opposed to practice (e.g., privatization), reasoning that doing so would provide a preferred context for analyzing our results.  This decision, however, compromised our ability to use more-formal analytic techniques, such as meta-analysis, for presenting the lessons learned from the relevant literature.

Final Thoughts

Overall, we are optimistic about the prospects for meaningful organizational reform of the public health system.  As indicated earlier, the increased funding that public health has experienced as a result of 9/11 and the anthrax attacks, coupled with a widespread recognition at the state and local levels of the importance in mounting an effective public health response to a wide range of emergencies, presents significant opportunities for change.

Our analyses, however, failed to reveal any panaceas or magic bullets for how best to structure governmental public health with regard to preparedness.  But together they tell us that the time is right for an extended dialogue that includes federal, state, and local public health officials about, essentially, who should do what.  Our case-study results demonstrate convincingly that policymakers should refrain from imposing a one-size-fits-all approach, and should recognize that risks, infrastructure, and capabilities vary, and will continue to vary, among and within states.  At the same time, efforts must be made to determine how public health systems can provide equal levels of protection to their jurisdictions’ populations.  Finally, policymakers need to harmonize public health preparedness measures across federal agencies, including CDC, HRSA, and Department of Homeland Security (DHS), and arrive at a consensus regarding which governmental entities will be held accountable for developing and maintaining relevant public health preparedness capabilities, and at what level.

Appendix A. CDC Progress Report Indicators (2004)

Table A.1: CDC Progress Report Indicators (2004)

Appendix B. Public Health Laboratories’ Survey questions used in analysis (fielded by Association of Public Health LABORATORIES [aphl])

Raw data sent by APHL included 45 items.  The index includes 22 yes/no items out of 45 in total.  We excluded items that sought to elicit subjective opinions (e.g., "Have delays in receiving reagents caused testing delays?"), redundant items (in some cases, there were separate items for "yes" and "no"), and items on which there was little or no variation (e.g., "Does your laboratory maintain a database of all the sentinel laboratories and pertinent laboratorians in your state?").

Table B.1: Survey Questions for Public Health Laboratories

Appendix C.  Robust Regression With Centralization-Regionalization Interactions

Table C.1: Robust Regression with Centralization-Regionalization Interactions (standard errors are in parentheses)

NOTE: ^a Four states were missing data on the “Regionalized” variable.

R2 estimates are based on analysis of variance models

References

Association of State and Territorial Health Officials (ASTHO), Resources for States Considering Health Agency Reorganization, Association of State and Territorial Health Officials, Issue Brief, January 2004. Available at: http://www.astho.org/
pubs/ReorganizationIssueBrief.pdf.

Bandoh, Evelyn,  “Outsourcing the Delivery of Human Services,” Welfare Information Network Issue Note, Vol. 7, No. 12, October 2003.  Available at:  http://www.financeprojectinfo.org/Publications/outsourcinghumanservicesIN.htm.

Barry, M. A., “How Can Performance Standards Enhance Accountability for Public Health?” Journal of Public Health Management and Practice, Vol. 6, No. 5, 2000, pp. 78-84.

Centers for Disease Control and Prevention, 1990 Profile of State and Territorial Public Health Systems, 1991. Atlanta GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Epidemiology Program Office, Division of Public Health Surveillance and Informatics. Available at: http://wonder.cdc.gov/
wonder/sci_data/misc/type_txt/stprof91.asp.

Centers for Disease Control and Prevention, Continuation Guidance for Cooperative Agreement on Public Health Preparedness and Response for Bioterrorism – Budget Year Five Program Announcement 99051, Atlanta, Ga.: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Emergency Preparedness and Response, June 14, 2004.

Centers for Disease Control and Prevention, Cooperative Agreement Guidance for Public Health Emergency Preparedness, July 1, 2005. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Emergency Preparedness and Response. Available at: http://www.bt.cdc.gov/planning/guidance05/pdf/annoucement.pdf

Centers for Disease Control and Prevention, Planning the Receipt and Distribution of the CDC National Pharmaceutical Stockpile, April 2002. Atlanta, Ga.: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention.

Centers for Disease Control and Prevention, Program Announcement AA-154, May 13, 2005.  Atlanta, Ga.: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Emergency Preparedness and Response. Available at: http://www.bt.cdc.gov/planning/guidance05/pdf/
annoucement.pdf.

Chapin, Tim,  “Beyond the Entrepreneurial City: Municipal Capitalism in San Diego,” Journal of Urban Affairs, Vol. 24, No. 5, 2002, pp. 565-581.

Cohen, H. W., R. M. Gould, and V. W. Sidel, “The Pitfalls of Bioterrorism Preparedness: The Anthrax and Smallpox Experiences,” American Journal of Public Health, Vol. 94, 2004, pp. 1667-1671.

Cuillinane, Kevin, and Dong-Wook Song, “Port Privatization Policy and Practice,” Transport Reviews, Vol. 22, No. 1, January 1, 2002.

Darr, D. E., L. Argote, and D. Epple, “The Acquisition, Transfer, and Depreciation of Knowledge in Service Organizations: Productivity in Franchises”. Management Science, Vol. 41, 1995, pp.1750–1762.

Davis, Lois M., Jeanne S. Ringel, Sarah E. Cotton, Belle Griffin, Elizabeth Malcolm, Louis T. Mariano, Jennifer E. Pace, Karen Ricci, Molly Shea, Jeffrey Wasserman, and James Zazzali, Public Health Preparedness: Integrating Public Health and Hospital Preparedness Programs, Santa Monica, Calif.: RAND Corporation, TR-317-DHHS, 2006.

De Langen, Peter W., and Evert-Jan Visser, “Collective Action Regimes in Seaport Clusters: The Case of the Lower Mississippi Port Cluster,” Journal of Transport Geography, April 29, 2004.

Fukuyama, F., and A. N. Shulsky, “Organizational Trends in the Commercial Sector,” in The “Virtual Corporation” and Army Organization, Santa Monica, Calif.: RAND Corporation, MR-863-A, 1997, Chapter 1.

Glaser, B. G. and A. Strauss, The Discovery of Grounded Theory: Strategies for Qualitative Research, Hawthorne, N. Y.: de Gruyter, Inc., 1967.

Hendrick, H. W. and B. M. Kleiner, Macroergonomics: An Introduction to Work System Design, Santa Monica, Calif.: Human Factors and Ergonomics Society, 2001.

Institute of Medicine, The Future of Public Health, Washington, D.C.: National Academy Press, 1988.

Institute of Medicine, The Future of the Public’s Health in the 21st Century, Washington, D.C.: National Academy Press, 2003.

Keane, C., J. Marx, and E. Ricci, “Privatization and the Scope of Public Health: A National Survey of Local Health Department Directors,” American Journal of Public Health,Vol.91, 2001, pp. 611-617.

Kettl, D. F.,  “Contingent Coordination:  Practical and Theoretical Puzzles for Homeland Security,” American Review of Public Administration, Vol. 33, 2003,  pp. 253-277.

Kissam, S., and P. D. Jacobson, “Integrating Health and Social Services in the Safety Net,” unpublished manuscript, University of Michigan School of Public Health, July 2005.

Kirshnitz, Eliot J., “Recent Development in New York Law,” St. John’s Law Review, Summer 2000. Available at: http://www.findarticles.com/p/articles/mi_qa3735/is_200007/ai_n8898153; accessed October 18, 2005.

Luberoff, David, and Jay Walder, “U.S. Ports and the Funding of Intermodal Facilities: An Overview of Key Issues,” Transportation Quarterly, March 2000.

Lurie, N., “The Public Health Infrastructure: Rebuild or Redesign?” Health Affairs, Vol. 21, 2002, pp. 28-30.

Masten, S., “The Organization of Production:  Evidence from the Aerospace Industry,”  Journal of Law and Economics, Vol. 27, 1984, pp.  402-417.

Mays, G. P., and P. K. Halverson, “Disparities in the Economic and Organizational Capacity to Deliver Public Health Services in the Delta,” Abstract from the 2005 Public Health Systems Research Meeting (Academy Health), Boston, Mass., June 25, 2005.

Mays, G. P., M. C. McHugh, K. Shim, N. Perry, P. K. Halverson, D. Lenaway, and R. Moonesinge,  “Identifying Dimensions of Performance in Local Public Health Systems:  Results from the National Public Health Performance Standards Program,” Journal of Public Health Management Practice,Vol. 10, 2004, pp. 193-203.

Miles, M. B. and A. M. Huberman,  “Matrix Displays,” in Qualitative Data Analysis, Thousand Oaks, Calif.:Sage Publications, 1984.

Miller, G., Managerial Dilemmas:  The Political Economy of Hierarchy, Cambridge:  Cambridge University Press, 1992.

Mirvis, D.,  “The Mississippi Delta Region:  Is It a Region?” Abstract from the 2005 Public Health Systems Research Meeting (Academy Health), Boston, Mass., June 25, 2005.

Moe, T.,  “The New Economics of Organization,” American Journal of Political Science, Vol. 28, 1984, pp. 739-777.

National Association of County and City Health Officials (NACCHO), NACCHO Survey Examines State/Local Health Department Relationships, NACCHO Research Brief, No. 2, 1998. Available at http://archive.naccho.org/documents/
Research_Brief_2.pdf.

National Association of County and City Health Officials, 1990 Profile of State and Territorial Public Health Systems, Washington, D.C.: NACCHO, 1991.

Rohrer, J., “Performance Contracting for Public Health:  The Potential and the Implications,” Journal of Public Health Management and Practice, Vol. 10, 2004, pp. 23-25.

Roper, W. L., and G. P. Mays,  “Performance Measurement in Public Health: Conceptual and Methodological Issues in Building the Science Base,” Journal of Public Health Management and Practice, Vol. 6, 2000, pp.  66-77.

Sclar, E. D., You Don’t Always Get What You Pay For: The Economics of Privatization, Ithaca, N. Y.: Cornell University Press, 2000.

Scott, W., M. Ruef, P. Mendel, and C. Caronna, Institutional Change and Healthcare Organizations: From Professional Dominance to Managed Care, Chicago: University of Chicago Press, 2000.

Stoto, M., David J. Dausey, Lois M. Davis, Kristin Leuschner, Nicole Lurie, Sarah Myers, Stuart Olmstead, Karen Ricci, M. Susan Ridgely, Elizabeth M. Sloss, and Jeffrey Wasserman, Learning from Experience: The Public Health Response to West Nile Virus, SARS, Monkeypox, and Hepatitis A Outbreaks in the United States, Santa Monica, Calif.: RAND Corporation, TR-285-DHHS, 2005.

U.S. Department of Education, National Commission on Excellence in Education, A Nation at Risk, Washington, D.C.: U. S. Department of Education, 1983.

Wall, S., “Transformations in Public Health Systems,” Health Affairs, Vol. 17, 1998, pp.  64-80.

Weick, K. and K. Sutcliffe, Managing the Unexpected:  Assuring High Performance in an Age of Complexity, San Francisco: Jossey-Bass, 2001.

Wohlsetter, P., C. L. Malloy, D. Chau, and J. L. Polhemus, “Improving Schools Through Networks: A New Approach to Urban School Reform,” Educational Policy, Vol. 17, 2003, pp. 399-430.

Wohlstetter, P., R. Smyer, and S. Mohrman, “New Boundaries for School-Based Management:  The High Involvement Model.”  Education Evaluation and Policy Analysis, Vol. 16, 1994, pp. 268–286.

 Copyright Information: Rand Corporation

The research described in this report was prepared for the U.S. Department of Health and Human Services. This research was produced within RAND Health’s Center for Domestic and International Health Security. RAND Health is a division of the RAND Corporation.

The RAND Corporation is a nonprofit research organization providing objective analysis and effective solutions that address the challenges facing the public and private sectors around the world. RAND’s publications do not necessarily reflect the opinions of its research clients and sponsors.

R® is a registered trademark.

A profile of RAND Health, abstracts of its publications, and ordering information can be found on the RAND Health home page at www.rand.org/health.

© Copyright 2006 RAND Corporation

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from RAND.

Published 2006 by the RAND Corporation
1776 Main Street, P.O. Box 2138, Santa Monica, CA 90407-2138
1200 South Hayes Street, Arlington, VA 22202-5050
4570 Fifth Avenue, Suite 600, Pittsburgh, PA 15213
RAND URL: http://www.rand.org/
To order RAND documents or to obtain additional information, contact
Distribution Services: Telephone: (310) 451-7002;
Fax: (310) 451-6915; Email: order@rand.org

-------------------------------------------------------