Testing a model of county government influence on health care safety-nets.
Abstract: In the United States, health care is not equitably distributed. As indicated in the literature, age, income, and other socio-economic indicators contribute to substantial differences in the variety and scope of health services. The 2010 Affordable Care Act illustrates the United States' effort to bring balance and equity to the health care system. In the meantime, county governments are struggling with rising health care costs on their budgets (Eaton, 2009; Phaup, 2009; Clark, 2003), particularly health care for low-income residents (Benton, Byers, Cigler, Klase, Menzel, Salant, Streib, Svara, & Waugh, 2008). However, as learned in this study, county governments across the country continue to address the health care needs of uninsured and underinsured citizens through participation in health care safety nets. This research identifies possible county government influences on health care safety-nets. This study analyzed 123 responses from county government administrators and elected officials along with secondary data from the U.S. Census Bureau and the International City/County Manager Association using a variety of statistical techniques, culminating in structural equation modeling. These analyses provided reasonable explanation for the variation among the variables leading to network performance improvement in meeting the health care needs of uninsured and underinsured people as well as the significant influence of county government involvement.
Article Type: Report
Subject: Community service (Management)
County services (Management)
Public health administration (Research)
Author: Knepper, Hillary
Pub Date: 06/22/2012
Publication: Name: Journal of Health and Human Services Administration Publisher: Southern Public Administration Education Foundation, Inc. Audience: Academic Format: Magazine/Journal Subject: Government; Health Copyright: COPYRIGHT 2012 Southern Public Administration Education Foundation, Inc. ISSN: 1079-3739
Issue: Date: Summer, 2012 Source Volume: 35 Source Issue: 1
Topic: Event Code: 200 Management dynamics; 310 Science & research Computer Subject: Company business management
Product: Product Code: 8340000 Community Services NAICS Code: 6242 Community Food and Housing, and Emergency and Other Relief Services
Geographic: Geographic Scope: United States Geographic Code: 1USA United States
Accession Number: 304051557
Full Text: INTRODUCTION

The local government burden of funding health care for the uninsured and underinsured has certainly been well documented as significantly burdensome (Eaton, 2009; Phaup, 2009, Benton, Byers, Cigler, Klase, Menzel, Salant, Streib, Svara, & Waugh, 2008, Clark, 2003). Notably, a study conducted by leading county government researchers examined county governance issues and identified the extreme resource demands that low-income health care needs place on county governments (Benton et al., 2008). Further, the continuing turbulent economic climate exacerbates the struggle of county governments to balance declining revenues with rising service demands (Eaton, 2009; Phaup, 2009; NACO, 2000b). The current climate illustrates the importance of studying county government involvement and participation in health care safety-nets.

This study is the first of its kind to develop and examine a model of county government influences, environmental pressures, and community resources in the context of health care safety-nets. While several research questions were identified in the overall study, this article examines one: Does the model accurately portray county government influence on health care safety-nets? This analysis suggests the model does provide evidence to address the research question: County governments appear to be important to the maintenance of health care safety-nets in their communities.

For this study, health care safety-net is defined as a loosely structured network of public and private health care organizations that provide health care access for uninsured and underinsured individuals (Hoffman & Sered, 2005). Examples of safety-net providers include, not exhaustively, federally qualified community health centers, public health departments, hospitals, and indigent care clinics.

The article is presented in the following format: the background of the topic is discussed in terms of health care spending, county health care collaborations, and county governments and health care. The conceptual framework is briefly discussed. Next, the research methodology and limitations are followed by the analysis and discussion sections. The discussion concludes with implications for public administration management and policy, suggestions for future research, and a summary of lessons learned.

BACKGROUND

Conservative estimates identify approximately 47 million Americans under the age of 65 who are uninsured (Kaiser, 2009; U.S. Census Bureau, 2009) and about a fifth of all Americans who are underinsured (Rovner, 2009; Kaiser, 2002). This substantive market failure has led the public sector to become involved in funding health care for certain populations such as the very poor, pregnant women, children, and seniors. Community health care expenditures vary considerably in the U.S., with evidence suggesting this funding variation contributes to differences in health outcomes around the U.S. (Mays & Smith, 2011). Indeed, while health care spending is measured in trillions of dollars in the U.S., the public sector accounts for more than half of these dollars in a complicated web of federal, state, and local funding streams (Selden & Sing, 2008; TFAH, 2008). The recently enacted Affordable Care Act is expected to inject another $15 billion into the health care industry (Mays & Smith, 2011). The private sector, both organizational and individual, accounts for the remaining 50% of health care spending.

In recent years, however, federal and state budget shortfalls have resulted in reductions and limitations in public health funding (Posner, 2003; Holahan, Buettgens, Chen, Carroll, & Lawton, 2011). Local governments are increasingly being called upon to make up the difference. These federal and state budget reductions and shortfalls suggest that for local governments, health care is becoming an increasingly unmanageable burden given a paradox of rising need and dwindling fiscal resources (NACO, 2002, 2011a; Smith, Gifford, Ellis, Rudowitz, O'Malley, & Marks, 2008). A little more than twenty years ago Agranoff and Pattakos (1989) suggested that county governments should leverage resources and contract for service to navigate the turbulent tide of service demands placed on counties. Those suggestions remain relevant in today's health care safety-net environment. Recognizing this, in 2002 the National Association of Counties (NACO) and the National Association of Community Health Centers (NACHC) partnered on a study that examined county health care funding. Of the 700 counties responding, 89% reported funding for public health departments, 39% contracted with other health care providers, and 26% funded community health centers (NACO, 2002). The results of the study presented here found a similar distribution of county government funding related to health care.

Addressing health care services for the poor is a substantial social challenge facing contemporary county governments (Benton et al, 2008; NACO 2011a). For county government policy-makers, it is important to recognize the changing world of county government service delivery. One example is how county governments are participating in health care safety-nets (West, 2004; Benton et al., 2008). Further, capacity building for local public health organizations is predicated in part upon the role of the public sector in collaboration with community-based resources (Campbell & Conway, 2005). At the same time, a funding crisis is looming over health care safety-nets (Kaiser Foundation, 2002; Holahan et al., 2011). A combination of dwindling fiscal resources and an increasingly dispersed service delivery system for health care has created unique challenges for both safety-net providers and local governments.

Theoretical Underpinnings

Conceptually, this study is grounded in two theories: resource dependency theory and complex adaptive systems theory. Both complexity theory and resource dependency theory link environmental constraints and resources to the functioning of organizations (Anderson, 1999; Pfeffer & Salancik, 2003). These two theories focus on how resource constraints and complex relationships affect the composition, and impact of, the health care network model constructed for this study.

Resource dependency theory links organizations through the control, acquisition, and maintenance of resources (Evan, 1965, Pfeffer & Salancik, 2003) within a dynamic environment which affects the continual distribution and redistribution of resources (Kiel, 1994). Further, complexity suggests that partnerships and collaborations are natural adaptations to complex social and public policy problems just as organizations must adjust to their environment (Meier & O'Toole, 2003; Davis, Eisenhardt, & Bingham, 2007). The formation of networks of organizations is securely linked to the complexity literature (Kapucu, 2006; Davis et al., 2007; Kilduff & Tsai, 2003; Nambisan, 2008). These secure theoretical linkages provide evidence to support the conceptual model developed for this study

Figure 1 provides an illustration of the conceptual model developed here after consideration of contemporary county government, health care, and network literature. The model provides a visual understanding of the relationships among the latent constructs of environmental pressures, community resourcefulness, pervasiveness of county influence and health care network performance.

Structural equation modeling was selected for this study for several reasons, among them the ability to measure latent constructs. Wan (2002) notes the relevance of structural equation modeling in studying health care systems through the utilization of theoretically informed latent constructs. The results of some of the structural equation modeling and path analysis will be discussed later in this article.

[FIGURE 1 OMITTED]

Table 1 identifies the measurable (or indicator) variables associated with this study. For the purposes of this article, pervasiveness of county influence and network performance will be discussed. The initials below each measurable variable are the data names for that variable as it appears in Figure 2, the final covariance structure model.

Framework

The purpose of this research was to identify and test a model of county government influence on health care safety-nets. The model was developed in part to test the hypothesis: pervasiveness of county influence (types of relationships, intensity of county relationships, and number of community oriented health organizations) has a direct effect on network performance (access to care, health care coordination, and health information exchange). This was accomplished by examining the county environment in which these safety-nets are providing services and the relationships the health care providers have established with county government. This hypothesis was developed in part on the Kaiser Foundation's research on uninsured and low-income individuals, which determined that health care is provided within an informal network of health care providers (2007), and in part from research suggesting that county governments are financially supporting health care for low-income residents (Benton et al., 2008; Eaton, 2009; Phaup, 2009). Understanding county government influence on health care safety-nets may inform local government policy decisions and improve the health care practitioner-county government relationship.

METHODS & LIMITATIONS

The unit of analysis in this study is the county. As illustrated earlier, county governments are investing heavily in health care despite fiscal stress. Leading county researchers like Benton et al., (2008) have called for county based health care and network research. The relevant variables of the model are pervasiveness of county influence and network performance. In this model, pervasiveness of county influence functions as an endogenous and exogenous variable, being both acted upon by, and acting upon, variables.

This is the first study to examine a macro model of health care networks based upon county administrator perception. In large part, this study observes the county government factors and conditions extant in a community and their impact on health care safety-nets. This is a cross-sectional study that uses multiple quantitative analyses to examine the data and to understand the relationships among the constructs. This analysis provides for goodness of fit regarding the study model. Goodness of fit analyses identify how well the measurable indicators generate the latent constructs (Byrne, 2001; Wan, 2002). County data were obtained from the U.S. Census Bureau. Five hundred counties were randomly selected to receive surveys. A total of 123 counties fully responded to the survey for a response rate of 25%. Univariate, correlation, structural equation modeling and path analyses were run. Figure 2 presents the final covariance structure model used in this study.

There are multiple limitations in this research. First, a disproportionate regional response rate may provide an inaccurate representation of variation among counties using network structures to deliver low income health care services. For this study, 23% of responding counties were located in the South Atlantic states and another 31% in the North Central states. If response rates are not even, it is possible that overrepresentation from one region or under-representation from another may obscure the data. Further, state policies differ in how state Medicaid contributions are generated with some states (like Hawaii) not requiring counties to provide funding, while other states (like Florida) do (Nakama, 2006; Murphy, 2010).

Delimitations include the potential bias linked to the use of technology in gathering the data for this study via the internet. However, because the survey was first sent via U.S. Mail, it is assumed this delimitation was constrained as those county managers who preferred the use of web-based survey were able to do so and those who preferred written responses had equal access.

This study involved the use of survey tools that collected and analyzed self-reported data, some of which is the perception of the public manager interpreting the question. While every attempt was made to make the variables clearly understood in the data collection instrument, validity of self-reported data may be a concern due to knowledge discrepancies, collection methods and interpretation.

Timing of the surveys could also be considered to be a delimitation. A total of nearly six months separates the mail-out of the first survey and receipt of the final completed survey. However, this concern may be unfounded given the high degree of skewness that was apparent among some of the fiscal variables. Even given the time lapse, many counties were experiencing similar environmental pressures.

Sample size is another limitation. The sample size of 123 counties is considered to be fairly small, especially in terms of structural equation modeling. This means that in some cases, statistically significant relationships may appear to be falsely insignificant, resulting in inaccurately rejecting hypotheses. Finally, the goodness of fit statistics suggest there is room for adjustment among the variables. Perhaps some of the variables should be reconsidered, even though the literature and theory support their inclusion. Exchanging some of the variables with less predictive influence may improve model fit.

ANALYSIS

The goodness of fit statistics are varied; some of the indicators suggest a reasonable fit with the model, and some of them do not. Byrne (2001) notes two items of importance in evaluating the fit of a model to its data; first, due to a small sample size, fit may not be accurately indicated by the RMSEA and second, the goodness of fit statistical indicators are only part of assessing the model's adequacy because theory and practicality must be considered along with statistics. Given Byrne's observation coupled with the statistical indicators, the match between the literature and these constructs suggests a reasonable fit among the data and the model. Figure 2 provides visual evidence of the final model that was identified through this research. An analysis is briefly provided following Figure 2 and Table 2. Figure 2 provides the analytical version of the model identified in Figure 1. The indicator variables are identified in Table 1 along with the data names that appear in Figure 2.

[FIGURE 2 OMITTED]

In examining the parameter estimates, the critical ratio should generally exceed 1.96. For this article, only one of the two paths in this model that meet this qualification for statistical significance is discussed. The path discussed here is between pervasiveness of county influence (PCI) and network performance (NP) with a critical ratio of 5.012, which is statistically significant at the .05 level. This illustrates the importance of county government influence on network performance. Subsequently, breaking these constructs down into their indicator variables suggests that types of relationships, the intensity of county relationships, and the number of community oriented health care organizations contribute to access to care, health care coordination, and health information exchange.

The goodness of fit statistics vary for this model. While the high [X.sup.2] statistic should indicate a poor model fit, Byrne (2001) notes the trend toward using the Likelihood Ratio ([X.sup.2]/df) statistic, particularly for small samples and its growing acceptance as an acceptable alternative to a low [X.sup.2]. Given this study only analyzed a small response group of 123 counties, it is the likelihood ratio considered here. In the case of the likelihood ratio, the 2.598 is lower than 4 and, therefore, considered evidence of a reasonably good fit between the data and the model. While the p value is not statistically significant for this model (.000), the CFI value of .915 indicates reasonable fit. While the GFI and AGFI statistics should be .9 or greater, and as close to 1 as possible respectively, the GFI (.840) and AGFI (.763) for this model do not quite meet the test for goodness of fit. However, these numbers do not indicate an enormous gap between a good fit and a poor fit. Finally, the RMSEA (.114) and the Hoelter (66) statistical values fail to suggest a reasonable fit between the model and the data. This illustrates that future research with these variables is required in order to improve the relationships among the model and the data.

Results

Using path analysis the relationship between pervasiveness of county influence (PCI) and network performance (NP) is discussed. The hypothesis is: Pervasiveness of county influence (types of relationships, intensity of county relationships, and number of community oriented health organizations) has a direct effect on network performance (access to care, health care coordination, and health information exchange). As evidenced by the path analysis, a direct association indicates the positive effects of pervasiveness of county influence (PCI) on network performance (NP). Pervasiveness of county influence (PCI) has a large positive effect (.550) on network performance (NP). These associations indicate the covariance structure model reasonably accounts for the change in the constructs, pervasiveness of county influence and network performance.

DISCUSSION

One research question is discussed in this article: What impact does pervasiveness of county influence (types of relationships, intensity of county relationships, and number of community oriented health organizations) have on network performance (access to care, health care coordination, and health information exchange)? This research question specifically examined the impact of three variables contained within the construct "Pervasiveness of county influence" on health care network performance. When asked whether or not health care access for the underinsured & uninsured improved as a result of network activities, 46% of respondents believed there had been significant or substantial improvement. In considering the improvement in the degree health care coordination for the underinsured and uninsured improved as a result of network activities, 40% believed there had been significant or substantial improvement. It is important to note this improvement occurred during a period that county administrators identify in the study as one of tremendous fiscal stress, fiscal year 2009-2010.

As noted previously in the hypotheses testing section, pervasiveness of county influence (PCI) did exert direct effects on network performance (NP). As identified in Figure 2, the results of the path analysis specify that pervasiveness of county influence (PCI), which is indicated by the types of relationships the county maintains with the health care providers (TR), the intensity of the relationships the county maintains with the health care providers (ICR), and the number of community oriented health organizations in the community (COHO) have a large predictive value (P =.55) on network performance (NP), which is indicated by access to care (AI), health care coordination (HCCI), and health information exchange (HIEI).

So, what impact does pervasiveness of county influence have on network performance? This study indicates that there is a significant impact. The path analysis indicates that as county influence increases, network performance will experience increasing improvement. This influence has implications for county administrators and policy-makers which are presented here.

Public Administration Management and Policy Implications

County governments are engaged in community health care (NACO, 2009a, 2011b). This study provides further evidence of such engagement and supports the impact county government influence has on health care network performance. Health care safety-nets are comprised of dispersed, independent health care organizations working to serve uninsured and underinsured populations. The levels of support county governments provide these safety-nets was identified in this study as ranging from relationships that were diffuse (information sharing/policy advice) to intense (shared outcomes). Nonetheless, this study identified that county influence has more than twice the predictive value of network performance than the other two variables that affect health care network performance. For county administrators, this is important to understand. The relationships county governments maintain with health care organizations affect health care service delivery for vulnerable populations. Noting that not all of this support is financial, this study suggests substantial variation exists in how county managers influence health care services within their communities. There are many routes to meeting the health needs of vulnerable populations, and county governments appear to be invested in several of these routes. As the 2010 Affordable Care Act (ACA) rolls out implementation, care should be taken by counties to continue to assist with the maintenance of a stable safety-net until such time as their support is no longer necessary.

As county government continues its participation in service networks, some public administration problems may be resolved (e.g. bureaucratic inefficiency, fiscal burdens). However, these resolutions may give rise to concerns about the hollow state including inadequate management and poor accountability. As counties have expanded beyond their historical service provision, it has become important to analyze these new service arenas of which health care is rapidly becoming significant (Agranoff & Pattakos, 1989; NACO, 2002, 2009b).

As this study substantiates, county governments are delving deeper into services such as health care. The funding commitment alone substantiates the influence of county governments in community health issues. As health care in the United States is in flux with ACA and litigation is pending in the Supreme Court, county governments are also at a crossroads. Economic conditions significantly constrain county governments. However, the health of a community is considered to be a form of capital that supports economic development (Mirvis & Bloom, 2011). Counties must continue to move their economies forward, and investing in health care may bring appropriate avenues for such activities.

As counties struggle to maintain the services to which residents have grown accustomed and tax revenues decline with a deepening economic downturn, county governments were still continuing to fund health care as of this study in 2009. While the level of involvement varied considerably, nearly all counties responding in this study participated to some degree in health care services, and all counties indicated some level of involvement with some health care provider. It is important to note that some county governments participate voluntarily in health care, while others are mandated by their state governments.

Wholey (1999) noted the importance of developing appropriate oversight among public managers in order to facilitate more effective performance measurement. The necessity of this issue is more clearly seen in the complex network of providers. Given the influence county governments have in health care, management capacity among the county administrator and the network service providers must adapt. County administrators who function within networks must engage those skills necessary for effective collaboration. At a minimum, these skills include interpersonal and interagency communications, building relationships, interagency planning, and the maximization of administrative resources (Austin, 2003). Further, Martin (2001) suggests that oversight of such service delivery systems requires management skills that facilitate, coordinate, and evaluate both the services and the organization providing them. These skills are required to engage network partners, bringing multiple stakeholders together for a common public good (Salamon, 2002). This study identifies this link with county government's involvement in health care networks beyond the basic funder-contractor relationship, further substantiating how county managers must apply new knowledge and skill sets.

Societal benefits may be obfuscated by marketplace and government failures, which leads to skepticism among citizens regarding the benefits of social service programs (Thayer & Fine, 2001; Considine 2003; Hayes, 2010). However, this study has provided evidence that county governments can respond successfully to these health care failures. Public-private collaborations for resolving health care disparities can be effective. The study data evidence the impact of county government and private health care organizations working together in a safety-net capacity to meet community health needs. For example, 47% of responding counties indicated there were between five and nine community oriented health care providers in their communities. While no distinction was made in this analysis between county owned and county supported organizations, the modest number of providers offers opportunities for substantive relationship building. Notably, more than a decade ago county government emerged as a fundamental participant in collaborative partnerships (Cigler, 1999), thus paving the way for its contemporary role in health care networks. Perhaps this is due in part to how county roots in social welfare date to the inception of county government itself (Fairle, 1904). The historic role of county governments in actively addressing societal ills is well founded. In terms of this health care study, more than 50% of responding counties provided substantial or significant financial support to public health departments, more than 30% provided financial support to federally qualified community health centers, and more than 50% supported emergency medical services. These financial contributions position county governments as significant partners in health care safety-nets.

This study has taken the first step in considering the influences of county governments on health care safety-net performance. County government policy-makers may learn from this research to facilitate the development and maintenance of safety-nets across a variety of service disciplines. This study provides county administrators with information necessary to identify areas within a health care network that may need stabilizing or shoring up in order to improve network performance. As learned in this study, resource-poor counties require neither a strong economy nor many health care providers to improving health care access. The study asked respondents to rate the degree to which health care access for the underinsured and uninsured improved as a result of network activities. For 46% of the respondents, there had been significant or substantial improvement. Further, 40% indicated significant or substantial improvement in the degree of health care coordination for the underinsured and uninsured as a result of network activities. Finally, 40% believed there had been significant or substantial improvement in health information exchange. Ultimately, this knowledge may contribute to improved health resource utilization.

In the 21st century community, county governments operate within the paradigm of contemporary governance. This governance approach embraces partnerships determined by service needs and economies of scale. It involves integrating stakeholders and communities to resolve problems. As such, this governance embraces new tools, necessary for achieving success in addressing complex social problems. As this study has identified, health care safety-nets are an example of how these tools are being wielded to resolve problems, and county governments do exert influences on health care safety-nets.

Recommendations for Future Research

This study examined the role of county government in health care and opens opportunities for new research. First, the high [X.sup.2] statistic suggests the conceptual model would benefit from additional revisions to improve goodness of fit. Variations among the exogenous variables may strengthen the model fit. Similarly, Hoelter's critical N was below the recommended number of 200. This relates to the need for a larger study sample. Replicating the study on a larger scale may provide a larger sample sufficient to strengthen Hoelter's critical N.

Second, this study could benefit from network analysis of the data. Health care network analysis could prove fruitful in testing the centrality of county governments within networks. Because this study was conducted among county government officials, it would be interesting and useful to compare these findings with a study of the network participants to gauge their perceptions of county government influence on health care network performance.

Finally, several questions emerge that were not addressed in this study. First, how can the inherent complexity in public problem solving be turned to an advantage? To understand this, comparative studies could be undertaken that first examine the size and scope of networks and then identify relative policy advantages and disadvantages of these arrangements. To gain a richer understanding, case studies could inform this research to identify evidenced based decision-making among policy makers. Therefore, a logical outgrowth of this study is how best can policy implementation within these network arrangements be measured? Researchers debate the best mechanisms for measuring policy success. Should efficiency be measured? What determines effective service deliverables? What constitutes health care network success? Is it comprehensive access for all? Is it a decrease in emergency room utilization? Could it be somewhere in the middle? The United States is currently wrestling with an unwieldy, expensive, inequitable health care system. The Affordable Care Act of 2010 is yet another factor that may affect not only county government fiscal involvement given the proposed Medicaid expansion, but also the very roles of the current safety-net providers. These changes will need to be explored. The role of government and its involvement in health care is changing. County governments are active participants in this change. As health care evolves in the United States, it will be important to continue to investigate the role of local government in health care.

CONCLUSION

Turbulent economic conditions force county governments to balance declining revenues with rising citizen service demands (Eaton, 2009; Phaup, 2009). Yet within this turbulence, county governments continue their support for health care safety-nets. As learned in this study, this commitment has occurred even though counties are experiencing significant fiscal distress. Consequently, this study explored the relationships among county governments and health care safety-nets using exogenous and endogenous constructs with indicator variables that examined multiple community factors.

This study provided reasonable explanations regarding the relationships among the pervasiveness of county influence and network performance variables to support this assertion. The pervasiveness of county influence emerged as having more than twice the predictive value as it relates to health care network performance. This is a critical lesson learned in this study: county involvement has a direct impact on health care networks. As this research has identified, health care networks can improve health care access for vulnerable populations and county governments affect network performance.

As the theoretical framework of complex adaptive systems and resource dependency suggested, networks rely upon the successful raising of resources (e.g. capital, financial, and personnel) within a complex web of stakeholders. Within this web, county governments may provide the necessary leadership to make services happen by validating the problem and leveraging resources.

Three lessons emerge from this study that may be useful for both county government decision-makers and health care providers. First, county influence over health care safety-nets can improve access to health care for uninsured and underinsured individuals. Second, the model developed for this study was reasonably founded, suggesting county administrators have the ability to understand some of the necessary variables of a health care safety-net. Third, fiscal stress is ubiquitous for county governments, yet they have remained involved in health care safety-nets. Reasons for this could be mandatory or voluntary, depending upon state policy. County government participation was the largest predictor of network performance in this study. Therefore, it is important for county administrators to recognize and understand their influence on health care safety-nets to inform policy decisions and to make improvements in the health care practitioner-county government relationship.

REFERENCES

Agranoff, R., Pattakos, A.N. (1989). Management of human services in local governments: A national survey. State and Local Government Review, 21 (Spring) 74-83.

Anderson, P. (1999). Complexity theory and organization science. Organization Science, 10(3), 216-232. Special Issue: Application of Complexity Theory to Organization Science.

Austin, M.J. (2003). The Changing relationship between nonprofit organizations and public social service agencies in the era of welfare reform. Nonprofit and Voluntary Sector Quarterly, 32(1) 97-114.

Benton, J. E, Byers, J., Cigler, B. A., Klase, K., Menzel, D. C., Salant, T., Streib, G., Svara, J. H, Waugh, J. W. L. (2008). Service challenges and governance issues confronting American counties in the 21st century: An overview. State and Local Government Review, 40 (1) 54-68.

Byrne, B.M. (2001). Structural equation modeling with AMOS- Basic concepts, applications, and programming. Mahwah, New Jersey: Lawrence Erlbaum Associates, Publishers

Campbell, P., & Conway, A. (2005). Developing a local public health infrastructure: The Maine turning point experience. Journal of Public Health Management and Practice. 11(2): 158-64.

Cigler, B.A. (1999). Pre-conditions for the emergence of multi-community collaborative organizations. Policy Studies Review, 16(1), 86-102.

Clark, R. L. (2003). Counties in crisis: A survey of current budgetary situations facing America's counties. Carl Vinson Institute of Government, prepared for the National Association of Counties.

Considine, M. (2003). Governance and competition: The role of nonprofit organisations in the delivery of public services. Australian Journal of Public Administration, 35(1), 63-77.

Davis, J. P., Eisenhardt, K. M., and Bingham, C.B. (2007). Complexity theory, market dynamism, and the strategy of simple rules. Stanford Technology Ventures Program Working Paper, April, 2007.

Eaton, L. (2009, September). Cities brace for a prolonged bout of declining tax revenues. Wall Street Journal Eastern Edition, 254(53), A5. Retrieved from Academic Search Premier database.

Evan, W.M. (1965). Toward a theory of interorganizational relations. Management Science, Vol.11, (10), Series B, B217-B230. Retrieved from: http://www.jstor.org/stable/2627811

Fairlie, J. (1920). Local government in counties, towns, and villages. New York: The Century Co. Retrieved from: http://books.google.com/books?id=5W8WAAAAYAAJ&printsec=frontcover&dq= Local+Government+in+Counties,+Towns,+and+Villages& source=bl&ots=K5nEl15Aae&sig=8AmrcxnBH6fiWaYV1ROlTlzp7bk&hl=en&ei= CGwETOj9HcP38AaUsdGxDA&sa=X&oi=book_result&ct=result&resnum=1&ved= 0CBYQ6AEwAA#v=onepage&q&f=false

Hayes, C. (2010). System Failure. Nation. 290(4). 22-24.

Hoffman, C., Sered, S.S. (2005). Threadbare: Holes in America's health care safety-net. The Kaiser Commission on Medicaid & the Uninsured. The Henry J. Kaiser Family Foundation.

Holahan, J., Buettgens, M., Chen, V., Carroll, C., Lawton, E. (2011). House Republican budget plan: State-by-State impact of changes in Medicaid financing. Kaiser Commission on Medicaid and the Uninsured.

Kaiser Foundation, H.J. (2002). Underinsured in America: Is health coverage adequate? The Kaiser Commission on key facts. Retrieved 11/13/09 from: http://www.kff.org/uninsured/upload/Underinsuredin-America-Is-Health- Coverage-Adequate-Fact-Sheet.pdf

Kaiser Foundation, H.J. (2007). Covering the uninsured: Growing need, strained resources. Kaiser Commission on Medicaid and the Uninsured. Retrieved from: http://www.kff.org/uninsured/upload/7429-02.pdf

Kaiser Foundation, H.J. (2009). The uninsured: A Primer. Key facts about Americans without health insurance. Kaiser Commission on Medicaid and the Uninsured.

Kapucu, N. (2006). Public-nonprofit partnerships for collective action in dynamic contexts of emergencies. Public Administration. 84 (1), 205-220.

Kiel, L. D. (1994). Managing chaos & complexity in government: A new paradigm for managing change, innovation, and organizational renewal. San Francisco: Jossey-Bass Publishers.

Kilduff, M., Tsai, W. (2003). Social networks and organizations. London: Sage.

Martin, L. L. (2001, May 31-June 1, 2001). Designing Alternative Delivery Options. Paper presented at the Guarantor Government Workshop on Alternative Service Delivery Options for Local Government, Taby, Sweden.

Mays, G.P., Smith, S. (2011). Evidence links increases in public health spending to declines in preventable deaths. Health Affairs; 30(8).

Meier, K. J., & O'Toole, Jr. L. J. (2003). Public management and educational performance: The impact of managerial networking. Public Administration Review, 63(6), 689-699.

Mirvis, D.M., Bloom, D. E. (2011). Population health and economic development in the United States. JAMA, 300(1).

Murphy, K. (2010). Counties and Medicaid: A snap shot. National Association of Counties. A Publication of the Research Division of NACo's County Services Department. Retrieved on 11/18/11 from: http://www.naco.org/research/pubs/Documents/Health,%20Human%20Services%20and %20Justice/Research%20Health%20Human%20Serv ices%20and%20Justice/Counties%20and%20Medicaid.pdf

Nakama, S.K. (2006). Medicaid and quest provider payment and Reimbursement rates Report No. 6, 2006. Legislative Reference Bureau. State Capitol Honolulu, Hawaii 96813. Retrieved on 11/18/11 from: http://hawaii.gov/lrb/rpts06/reim.pdf

Nambisan, S. (2008). Transforming government through collaborative innovation. IBM Center for The Business of Government. Retrieved from: http://www.businessofgovernment.org/sites/default/files/ NambisanReport.pdf

National Association of Counties (NACO). (2002). Community health centers & public health services: Where are counties devoting their resources for health care? National Association of Counties.

National Association of Counties (NAC0) (2009a). Annual Report, 2008-2009, 1-24. Retrieved from: http://www.naco.org/Content/NavigationMenu/ About_NACo/Annual_Report/Default274.htm

National Association of Counties (NAC0) (2009b). How are counties doing? An economic status survey. Retrieved from: http://www.naco.org/Template.cfm?Section=Surveys&template= /ContentManagement/ContentDisplay.cfm&ContentID=32326

National Association of Counties (NACO). (2011a). Statement to House on Cost-Shifting Proposals to Medicaid. National Association of Counties.

National Association of Counties (NACO). (2011b). 2011 Key legislative priorities. Retrieved from: http://www.naco.org/legislation/Pages/2011LegislativePriorities.aspx.

Pfeffer, J. Salancik, G.R. (2003). The external control of organizations: A resource dependence perspective. Stanford Business Books: Stanford, CA.

Phaup, M. (2009). Federal use of implied guarantees: Some preliminary lessons from the current financial distress. Public Administration Review, 69(4), 651659.

Posner, P. L. (2003). Highlights of a GAO symposium: Addressing key challenges in an intergovernmental setting. (P. L. Posner, Ed.) (GAO- No. 03-365SP). Washington, D.C.: Government Accountability Office. Retrieved from: http://www.gao.gov/new.items/d03365sp.pdf

Rovner, J. (2009). The Richard and Hinda Rosenthal Lecture 2008: Prospects for health reform in 2009 and beyond. Washington D.C.: Institute of Medicine National Academies Press.

Salamon, L.M. (2002). The tools of government: A guide to the new governance. Oxford University Press: New York.

Selden, T.M., Sing, M. (2008, July). The distribution of public spending for health care in the United States, 2002. Health Affairs, 27(5) 349-359. doi: 10.1377/hlthaff.27.5.w349.

Smith, V., Gifford, K., Ellis, E., Rudowitz, R., O'Malley, M., Marks, C. (2008) An Update on Medicaid Spending, Coverage and Policy. Headed for a Crunch: Heading into an Economic Downturn. Kaiser Commission on Medicaid and the Uninsured. Kaiser Family Foundation.

Thayer, C., E. and A. H. Fine (2001). Evaluation and outcome measurement in the nonprofit sector: Stakeholder participation. Evaluation and Program Planning, 24, 103-108.

Trust for America's Health (TFAH). (2008). Shortchanging America's health 2008: A state by state look at how federal public health dollars are spent. Retrieved from: http://healthyamericans.org/reports/shortchanging08/Shortchanging08.pdf.

U.S. Census Bureau (2009). Income, poverty, and health insurance coverage in the United States: 2008 . Retrieved from http://www.census.gov/prod/2009pubs/p60-236.pdf

Wan, T.T. H. (2002). Evidenced-Based health care management: Multivariate modeling approaches. Boston: Kluwer Academic Publishers.

West, K. (2004). Return on community investment: Opening doors to improve health outcomes. National Association of Counties Special Report.

Wholey, J. S. (1999). Performance-based management: Responding to the challenges. Public Productivity & Management Review. 22(3), 288-307.

HILLARY KNEPPER

Pace University

Hillary Knepper is an assistant professor in the Department of Public Administration at Pace University. With over 20 years as a public and nonprofit practitioner, her main research interests are health care, local government, complexity in service delivery, and social media in governance.
Table 1

Indicator Variables for Conceptual Model Constructs

Construct & Relevent Indicator Variables:

Environmen-       Community          Pervasive-       Network
tal Pressures     Resourceful-       ness of          Performance
(EP)              ness (CR)          County           (NP)
                                     Influence
                                     (PCI)

Indicator         Indicator          Indicator        Indicator
Variables:        Variables:         Variables:       Variables:
Population        County financial   Types of         Access to care
growth (PG)       support for        relation-ships   (AI)
                  health care        (TR)
                  (CinFinSup)

Geographic        County general     Intensity of     Health care
region (Greg)     revenue/           county           coordination
                  Number of          relation-ships   (HCCI)
                  county             (ICR)
                  employees
                  (CTYCOMB)

Population size   Structure of       Number of        Health
(PS)              county             community        information
                  government         oriented         exchange
                  (Struc)            health care      (HIEI)
                                     organiza-tions
                                     (COHO)

Political         Number of
leadership        health care
pressure (PLP)    organizations
                  (NOHO)
                  Indirect public
                  health (IPH)

Table 2

Covariance Structure Model- Parameter Est.

                                 Std. Err   Std.
                    Reg.         of the     Reg
Path Parameter      Coeff        Est.       Weights   C.R.      P

pci [right arrow]   .516         .103       .550      5.012 *   ***

* Significant at the p.05 level.

PCI= Pervasiveness of County Government Influence; NP= Network
Performance

Table 3

Goodness of Fit Statistics for the Covariance Structure Model

Statistic                  210.406

Chi-Square ([X.sup.2])

Degrees of Freedom (df)      81

P value                     .000

Likelihood Ratio            2.598
([X.sup.2/]df)

Normed Fit Index (NFI)      .871

Goodness of Fit Index       .840
(GFI)

Adjusted Goodness of Fit    .763
Index (AGFI)

Comparative Fit Index       .915
(CFI)

Root Mean Square Error      .114
(RMSEA)

HOELTER (.01)                66
Gale Copyright: Copyright 2012 Gale, Cengage Learning. All rights reserved.