THE GROWTH OF PRIVATE VOLUNTARY ORGANIZATIONS 1968-2004 Except where reference is made to the work of others, the work described in this dissertation is my own or was done in collaboration with my advisory committee. This dissertation does not include proprietary or classified information. _______________________ Felita Nanette Boldin Certificate of Approval: _____________________ _____________________ Cynthia Bowling Cal Clark, Chair Professor Professor Political Science Political Science _____________________ _____________________ D?Linell Finley Johnny Green Professor Assistant Professor Political Science & Political Science Public Administration _______________________ Joe F. Pittman Interim Dean Graduate School THE GROWTH OF PRIVATE VOLUNTARY ORGANIZATIONS 1968-2004 Felita Nanette Boldin A Dissertation Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Auburn, Alabama December 15, 2006 iii THE GROWTH OF PRIVATE VOLUNTARY ORGANIZATIONS 1968-2004 Felita Nanette Boldin Permission is granted to Auburn University to make copies of this dissertation at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ________________________ Signature of Author ________________________ Date of Graduation iv VITA Felita Nanette Boldin was born December 11, 1972, in Manhattan, New York. She spent the majority of her formative years in Europe where she graduated from Ramstein American High School in Germany. She was an Air Force ROTC student at the University of Alabama where she graduated and was commissioned in August, 1996. She served in the Air Force as a Contracting Officer and Flight Training Officer for seven years. Upon separating from the Air Force in July, 2003, she became a full-time Ph.D. student at Auburn University. v DISSERTATION ABSTRACT THE GROWTH OF PRIVATE VOLUNTARY ORGANIZATIONS 1968-2004 Felita Nanette Boldin Doctor of Philosophy, December 15, 2006 (M.A. Webster University, 1999) (B.S. University of Alabama, 1996) 170 Typed Pages Directed by Cal Clark The increase in the number of private voluntary organizations (PVOs) registered with the United States Agency for International Development was analyzed for the period 1968-2004. The factors that were used to explain the growth pattern of PVOs included: partisanship of the President and of Congress, public funding of PVOs, and the passage of Public Law 97-113. Four hypotheses were used to evaluate growth in private voluntary organizations for the study period. The population size for this research was 31 to reflect cases for the years 1968-2004. Data for each year were collected from the Voluntary Agency Reports produced by USAID annually, creating 31 data points. The vi entire private voluntary organization population was evaluated in aggregate by year. The central finding was that the partisanship of the President, but not Congress, had a strong impact on PVO growth. In addition, the both of the public policy measures considered exercised moderate influences in the predicted direction. vii ACKNOWLEDGEMENTS The author would like sincerely thank Dr. Cal Clark for his insight and knowledge, Dr. Bowling for her continued encouragement throughout this process, and Dr. Johnny Green and Dr. D?Linell Finley for their guidance during this project. She is especially grateful to God, her son, and her family for their continued inspiration, which made the publication of this document possible. viii Style manual or journal used Publication Manual of the American Psychological Association, 5 th Edn. Computer software used Microsoft Office, Student Teacher Edn., 2003; SPSS v.14.0 ix TABLE OF CONTENTS LIST OF TABLES........................................xiii LIST OF FIGURES........................................xiv CHAPTER 1. INTRODUCTION.................................1 Significance........................................2 Discussion of Literature............................7 Purpose.............................................9 Research Hypotheses................................10 Methodology........................................10 Population....................................10 Cross-Sectional Multiple Regression...........11 Models........................................12 Organization of the Study..........................12 Chapter 1 Appendix A...............................16 CHAPTER 2. LITERATURE REVIEW...........................19 Introduction.......................................19 Historical Perspective.............................20 Theoretical Perspective............................27 Nonprofit Administration......................27 Government Failure............................33 Outsourcing and Privatization.................41 x Research Variables.................................52 Partisanship of the President.................52 Congress......................................54 Public Law 97-113.............................57 Private Voluntary Organization Funding........59 Public Funding................................62 Research Model.....................................65 Conclusion.........................................67 CHAPTER 3. METHODOLOGY.................................68 Introduction.......................................68 Explanation of Hypotheses..........................69 Description of Population..........................71 Data Collection....................................73 Instrumentation...............................73 Dependent Variable............................73 Independent Variables.........................74 Data Measurement..............................76 Data Limitations..............................78 Stages and Types of Analysis.......................78 Regression Analysis...........................79 Multiple Regression...........................85 Conclusion.........................................91 xi CHAPTER 4. EVALUATION OF THE VARIABLES OVER TIME.......92 Introduction.......................................92 Evaluation of Variables............................92 Private Voluntary Organizations...............92 Executive Branch Partisanship.................95 Partisanship of Congress......................97 Public Funding...............................101 Summary of Graphical Analysis.....................106 CHAPTER 5. TESTING THE THEORY OF PVO GROWTH...........108 Introduction......................................108 Modeling the Direct Relationships.................109 Partisanship of the President................114 Congress.....................................115 Public Law 97-113............................116 Public Funding...............................117 Summary...........................................118 Multiple Regression...............................118 Stage One Multiple Regression Analysis.......119 Stage Two Multiple Regression Analysis.......121 Path Diagram......................................123 Analysis of Hypotheses............................125 Partisanship of the Executive Branch.........125 Congress.....................................126 xii Public Law 97-113............................127 Public Funding Increase......................127 Chapter 5 Appendices..............................129 CHAPTER 6. CONCLUSION AND RECOMMENDATIONS.............133 Introduction......................................133 Significance of Study.............................133 Research Question.................................134 Instrumentation...................................135 Summary of Findings...............................135 Implication for Policy and Administration.........137 Limitations of Study..............................138 Recommendations for Future Research...............139 Future Policy Changes.............................141 Conclusion........................................142 REFERENCES.............................................143 xiii LIST OF TABLES Table Page 2.1 Primary Legislation.................................26 2.2 PVO Funding Laws....................................56 3.1 Data Collection.....................................76 3.2 Data Measurement....................................77 3.3 Approximate Strength of Correlation Coefficient.....83 3.4 Correlation and Regression Statistics for the Relationship Between PVO Increase and Government Spending Increase...................................85 3.5 Multiple Regression Explaining PVO Growth by Republican Presidencies and Growth in Government Spending for PVOs...................................87 5.1 Partisanship of the President?s Correlation with PVO Increase...........................................111 5.2 Partisanship of the Congress?s Correlation with PVO Increase...........................................111 5.3 Public Law 97-113?s Correlation with PVO Increase..112 5.4 Public Funding?s Correlation with PVO Increase.....113 5.5 Variables Selected for Statistical Analysis........113 5.6 Multiple Regression Explaining PVO Growth by Partisanship of the President and Congress.........120 5.7 Multiple Regression Explaining PVO Growth by Presidential Partisanship, Public Law 97-113, and Government Spending................................122 xiv LIST OF FIGURES Figure Page 1.1 Number of PVOs registered with USAID 1967-2005.......3 2.1 Variable path diagram...............................66 3.1 Increase in number of PVOs vs. increase in public spending............................................79 4.1 Number of PVOs from 1968-2006.......................94 4.2 PVO changes 1968-2004...............................95 4.3 Republican executives and PVO growth................97 4.4 Republican Senate/Congress 1968-2004................99 4.5 Public Laws 97-113 and 99-83 1968-2004.............101 4.6 PVO public support 1968-2004.......................103 4.7 Changes in public support of PVOs 1968-2004........105 4.8 Percentage change in PVOs and public funding 1968-2004..........................................106 5.1 Variable path diagram..............................124 1 CHAPTER 1. INTRODUCTION Countries around the world have organizations that are neither government agencies nor a part of the business sector that contribute to meeting the needs of the local community. Various researchers have classified this body of organizations as the third sector within nation-states. In developed countries, third sector organizations are commonly called non-profit organizations, while in developing countries they are referred to as non- governmental organizations. In the last fifteen years, the numbers of third sector organizations have grown rapidly at both the national and international levels. The international organizations operating in the third sector are called international non-governmental organizations or just referred to as non-governmental organizations (NGOs). ?Nongovernmental organizations are governed by boards of directors that tend to reflect the particular culture, history, and mandates of the organizations concerned? (Natsios, 2001, p. 392). 2 Private Voluntary Organizations (PVOs) are a subset within the non-governmental organization (NGO) community. The Government Accountability Office (GAO) states that ?the term NGO includes for-profit firms, educational institutions, cooperative development organizations, and PVOs. PVOs are tax-exempt, nonprofit organizations that receive voluntary contributions of money, staff time, or in-kind support from the general public and are engaged in voluntary, charitable, or development assistance activities? (United States General Accounting Office, 2002, p. 8; see also Gorman 1984). The United States government uses USAID as the medium for dispersing aid to other governments by employing the services of PVOs through contracts, grants, and cooperative agreements. The private sector also supports the PVO community in performing relief efforts and providing development assistance to lesser developed countries. According to Thompson (2003), ?because PVOs are neither fully in the public sector nor the private sector, they remain under-studied and under- analyzed? (p. 43). Significance The number of PVOs engaged in overseas programs in developing countries has grown significantly (Smith, 1984). Wallace and Lewis (2000) further explain that NGOs (PVOs) have a relatively long history, but their numbers and profile did not grow significantly until the 1980s. They cite increased violence and neoliberal economic ideas as the reason for the growth. However, the reasons for this growth are more anecdotal than empirical in nature. The question of the cause of the growth in the number of PVOs registered with USAID is one of significance since the number of registered PVOs has increased from 73 organizations in 1967 to 539 organizations as of 2005. 0 100 200 300 400 500 600 1967 1969 1971 1973 1979 1981 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Years # of NGOs Number of NGOs Figure 1.1. Number of PVOs registered with USAID 1967-2005 3 4 The growth trend is obvious in PVOs from an initial glance at Figure 1.1. In 1981, there were 156 US PVOs registered with USAID with the number rising to 301 in 1991, and today there are 539 registered PVOs. PVO numbers started to rise at a faster rate during the Reagan administration. This may have been a result of the Private Enterprise Initiative in 1981, which fostered an environment favoring the growth of PVOs. The number of PVOs receiving funding from USAID has tripled since 1982. Further, the gross revenue of the organizations registered with USAID has grown from $433 million in 1967 to $19.9 billion in 2004. According to the GAO, ?in fiscal year 2000, USAID directed about $4 billion of its $7.2 billion assistance funding to nongovernmental organizations, including at least $1 billion to private voluntary organizations working overseas? (United States General Accounting Office, 2002, p. 2). Given this information, it is important to understand the cause of the growth in this sector. This study will attempt to examine four variables which may be contributing factors to the increase in the number of PVOs. The variables selected are not a comprehensive list of factors due to the limited data set 5 available, which covers the growth in PVOs by year over a thirty two year period. The variables that will be examined in relation to the rise in the number of PVOs are the following: Republican partisanship of the Executive Branch and Congress, the implementation of PVO funding minimums through Public Law 97-113, and the amount of public funding going to PVOs. Most fundamentally, partisanship is a likely factor due to the pro-privatization position of conservative Republicans. For example, in 1981 at the start of the Reagan administration there were 153 PVOs; at the end of his second term as President in 1988 the number of PVOs had increased by 52, or approximately one-third, to 205 organizations. Further, during the administration of Bush I the number of PVOs increased by 142 organizations in a four-year period. Conversely, the rate of growth slowed during the Clinton administration, with the number of PVOs increasing by 92 organizations in an eight-year period. The trend in the number of PVOs depicted in Figure 1.1 shows that short growth spurts commenced during the Reagan and George W. Bush presidencies. Thus, the partisan environment will be indicated by two major independent or explanatory variables: 1) whether or not a Republican was 6 President; and 2) whether or not Republicans held a majority in at least one of the two Houses of Congress. The partisanship is significant because if either Congress or the President or both are Republican, they are likely to support privatization. Specific policy initiatives by Republicans should also promote or stimulate the formation of PVOs. Thus, two indicators of policy are included in the model as intervening variables that explain why the Republican control of government leads to an increase in the number of PVOs. One explanatory factor is Public Law 97-113 which required USAID to give at least 12% of the Development Assistance budget to Private Voluntary Organizations. The other policy indicator is government funding of PVOs, which has risen at a slow but steady rate from about $16 million in 1968 to $3.1 billion in 2004. The importance of the PVO growth rate is twofold. First, they have become an available resource to be used by USAID to meet some foreign policy objectives in the areas of humanitarian and development assistance; second, with the amount of private funds they receive, they are representatives of the American public. Whether they are publicly or privately funded, the rise in the number of PVOs has made them second only to the 7 US military in representing American culture to the world. They are able to ?advance U.S. national and political interests? (Zimmerman & Hook, 1996, p. 65). PVOs have become an instrument used by the American government to promote foreign policy objectives and an instrument used by the American public to promote good will. USAID?s 1982 policy paper, Partnership in the International Development with Private and Voluntary Organizations refers to PVOs as ?a heterogeneous group of agencies reflecting the diverse nature of American society, PVOs active in developing countries embody the basic American values of pluralism, voluntary action and concern for others? (USAID, 1982, p. 4). There are times when government objectives may be met by letting the PVO market, with the support of private donations, respond to international situations. Further, though USAID does fund PVOs through grants, contracts, and cooperative agreements, the majority of PVO funding comes from the private sector. Discussion of the Literature The literature covering Private Voluntary Organizations provides the basis for the historical perspective, theories, and selected variables. These three aspects provide a three dimensional look into the world of 8 Private Voluntary Organizations. First, the historical perspective gives a starting point for the growth of these organizations within the United States as a result of foreign policy initiatives to assist nations around the world. Second, the three major theories which support the growth phenomena of the Private Voluntary Organizations in the United States are nonprofit administration, government failure, and outsourcing and privatization. Lastly, government funding, partisanship of the president and congress, and Public Law 97-113 are examined in light of the literature and their relationship to Private Voluntary Organizations. These three components are discussed in detail in Chapter 2. A thorough review of the literature does not provide adequate information concerning the impact of the explanatory variables considered here on the growth of the PVO sector. Further, while funding patterns are addressed largely by case, aggregate information still remains largely unstudied. Thus, there is a need for further research to provide clarity about the PVO sector. The results of this study will serve to further define and describe with a quantitative method the PVO community in the United States. Understanding this community is 9 integral to keeping this sector as a viable medium for providing humanitarian, relief, and developmental assistance to other countries in times of need. Purpose The purpose of the research is to add to the body of knowledge that covers Private Voluntary Organizations. According to Chabbott (1999), the degree that the proliferation of PVOs has been examined as a result of government policy initiatives rather than popular interest is an area that remains to be explored. This research analyzes the impact of the partisanship of the President and Congress, Public Law 97-113, and government funding on the number of Private Voluntary Organizations registered with USAID from 1968 to 2004. No prior research has focused specifically on the growth of the number of Private Voluntary Organizations during this time period. American Private Voluntary Organizations have increased in importance as the American private and public sectors rely on them to fill the relief, humanitarian, and development needs of lesser developed countries and/or disaster areas around the world. 10 Research Hypotheses This research examines the following hypotheses: H1. In terms of comparing the annual increase in PVOs, Republican control of the Executive Branch will have a positive impact on the annual increase in PVOs. H2. In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs. H3. In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase in PVOs. H4. In terms of comparing the increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. Therefore, the annual increase of PVOs registered with USAID is expected to be determined by the partisanship of the executive branch and legislative branch, Public Law 97- 113, and public funds going to PVOs. Methodology Population The population size for this project is 31. The population includes longitudinal data from 1968-2004 on the 11 number of PVOs registered, the partisanship of the President and Congress, whether or not Public Law 97-113 had been passed, and the amount of government funding devoted to PVOs. Though the data are longitudinal in nature, for the purpose of this project, they will be treated as cross-sectional to analyze the causal relationship between the variables and PVO growth. Data are available from 1968-2004 on the number of PVOs and government support. There are holes in the data from 1975- 1978, and 1982, as there were no USAID annual reports for those years. The case for 1979 was eliminated from the study to eliminate outliers created from transposing the data. The data in dollar amounts will be analyzed in terms of constant 1990 dollars. The USAID American Voluntary Agency Reports also give specific information for each PVO registered every year for 1968-2004. However, the data for this study are used in the aggregated form presented in the reports. Totals are given for the number of PVOs, the total amount of funding, and the amount of public funding for each year. Cross-Sectional Multiple Regression This project is designed as a cross-sectional study using multiple regression as the instrument to analyze the 12 relationship between the variables. In essence, the increase (or decrease) in the number of PVOs for a given year is statistically explained through multiple regression analysis by the values of four independent or predictor variables: 1) the partisan affiliation of the President; 2) the partisanship of Congress; 3) whether or not Public Law 97-113 was in existence; and 4) the amount of change of government funding used to support PVOs. Models For these regressions, the data will be transformed by converting the number of PVOS to the increase or decrease in the number of PVOs for each year. Likewise, the data on government funding will also be converted to annual increases or decreases. Organization of the Study Chapter 1 provides general information about Private Voluntary Organizations and their significance as an area of research. Private Voluntary Organizations are organizations that fall within the various types Nongovernmental Organizations involved in development assistance around the world. The growth trend of these organizations is highlighted for the years 1968-2004. Also, in this chapter the four independent variables are 13 introduced and the dependent variable is defined for the research project. Four hypotheses are outlined for further examination in subsequent chapters. Chapter 2 provides an in depth look of the body of literature that surrounds Private Voluntary Organization. This chapter is organized into three sections. First, the history of Private Voluntary Organizations is discussed in detail using the conceptual framework developed by Porter (1990). Second, the three main theories that are explored are: 1) nonprofit administration; 2) government failure; and 3) outsourcing and privatization. Lastly, the four explanatory variables are clearly defined and analyzed through the literature; and their predicted relationship to the increase in the number of Private Voluntary Organizations is explained. Chapter 3 details the research methods used for this study of the increase of Private Voluntary Organizations. The hypotheses for the four variables are explained, with support from the literature. Detailed information is provided about the population of the cases examined. Further, the measures and the method for data collection are explained for each of the dependent and independent variables. In addition, this chapter introduces two 14 methods of analysis of the relationship of the increase in PVOs to the explanatory variables: 1) graphical analysis of the temporal relationships between the two variables and 2) more sophisticated regression models on the precise statistical relationships among these variables. Lastly, the expected results from study are presented. Chapter 4 presents the graphical analysis. Trends and patterns in the data are noted in relation to time and to the increase in PVOs. The hypotheses are evaluated in light of the data and the supporting literature. Graphs are presented for each variable to give a sense of the dynamics in the PVO environment and provide a visual display of the data and of how PVO growth is related to the values of the independent variables. Chapter 5 is the presentation of the statistical findings for each of the hypotheses based upon simple and multiple regression. The nature of the relationship between each of the four independent variables and the annual increase or decrease in the number of PVOs is first established through simple regressions of the direct association between them. More complex multiple regression models are applied to describe how the independent variables act together to shape PVO growth. Based upon 15 these results, each of the research hypotheses is evaluated and related to the theoretical literature. The concluding chapter provides a discussion of the findings and a conclusion to this study. The limitations of the study are noted, along with recommendations for further study in the area of Private Voluntary Organizations. In addition, some implications of the research for policy and administration are offered in the area of PVO autonomy and independence. 16 CHAPTER 1 APPENDIX A Definitions Contract-an agreement with a contractor, usually awarded competitively, to provide goods and services with substantial direction and technical direction. Contracting Out/Outsourcing-the shifting of selected government functions to a nongovernmental organization for performance to fulfill agency requirements. Cooperative Agreement-an award made competitively to an organization to implement programs with limited involvement from the grantor. Faith Based-defined by GAO as a term used for a PVO if its website, mission statement, objectives, or priorities directly mentioned an affiliation with a religious organization or God, Allah, another deity, prayer, faith or other overtly religious terms. General Accountability Office (GAO)-an independent, nonpartisan agency that works for Congress. The GAO gathers information to help Congress determine how well executive branch agencies are doing their jobs. 17 Government Failure-a state that occurs when the public sector is not able to provide goods and services to the public efficiently. According to Charles Wolf Jr. (1979), the four types of government failure are internalities, redundant and rising costs, derived externalities, and distributional inequity. Grant-an award made competitively to an organization to implement programs independent from the grantor. Gross Domestic Product (GDP)-is a measure of the value of the economic production of a particular territory in financial capital terms during a specified period. It is one of the measures of national income and output. Lesser Developed Country (LDC)-a country with a low income average, a relatively undeveloped infrastructure and a poor human development index when compared to the global norm. Non-Governmental Organizations (NGOs)-an umbrella term as defined by the GAO which includes for-profit firms, educational institutions, cooperative development organizations, and private voluntary organizations. Private Voluntary Organizations (PVOs)-according to the GAO; tax-exempt, nonprofit organizations that receive voluntary contributions of money, staff time, or in-kind 18 support from the general public and are engaged in voluntary, charitable, or development assistance activities. United States Agency for International Development (USAID)- the U.S. agency responsible for distributing foreign aid around the world. 19 CHAPTER 2 LITERATURE REVIEW Introduction The purpose for this literature review is to provide a foundation in history, theory, and the proposed variables encompassing the growth in the number of Private Voluntary Organizations. ?There has been an explosion of nongovernmental organizations dedicated to some aspect of humanitarian action? (Barnett, 2005, p. 723); and according to Aall (1996), ?the rapid rise in the number of these organizations has resulted in more institutions that are able to mobilize resources for humanitarian crises in hitherto inaccessible and neglected parts of the world? (p. 6). The idea that private voluntary organizations have experienced tremendous growth and are performing critical functions worldwide is commonly stated. However, what is not specified is the cause of this growth through empirical means. The rise of these organizations is chronicled as a sequencing of events rather than a cause and effect relationship. Fitzduff and Church (2004) agree that the 20 rapid expansion of PVOs in number has been attributed to the end of the Cold War, an increase in outsourcing, a growth in humanitarian norms, and the perception of PVOs as more trustworthy than governments. This survey of the literature explores the growth in number of Private Voluntary Organizations in the United States first, through the historical framework of Porter (1990). Next, three theories which provide a conceptual basis for the growth of Private Voluntary Organizations in society are examined. Lastly, the proposed variables of government funding, partisanship of the executive branch, partisanship of congress, and Pubic Law 97-113 are presented. Historical Perspective The history and background of NGOs in the US are tied to the evolution of US foreign aid policy. According to Porter (1990), foreign aid policy can be broken down into three periods. ?The first period began with the close of World War II and ends with the Mutual Security Act of 1951? (p. 6). This is significant because the US adopted foreign aid policy to assist in the reconstruction of Europe in peacetime. It was not unusual to provide foreign aid during times of war, which had been the case for WWI and WWII. However, the idea that it was in the best interest 21 of the world to continue aid in a time of peace was a new concept that marked this first period. US aid was provided in the form of economic grants and loans to European countries with a small portion of total aid expenditure going to military assistance (Porter, 1990). The Marshall Plan, which was established in 1948, embodied the US commitment to Europe and is considered a great success in development assistance. Europe received bilateral support from the United States through the Marshall Plan and through US-based International NGOs (PVOs) during the post- war period instead of multilateral support through the United Nations (Chabbot, 1990). This is a critical point in history because it establishes PVOs as credible organizations capable of meeting development and humanitarian efforts in devastating conditions. Meyer (1999) recalls that the PVO-government partnership was solidified when the War Relief Control Board was established to coordinate the relief effort of U.S. PVOs after WWII. This board was dissolved in 1946 and replaced with the Advisory Committee on Voluntary Foreign Aid which consisted of representatives from the government and the PVO community. This historic step of funding PVOs to aid 22 in development assistance is the precursor for later policies that stimulated an expanded role for PVOs. The second period discussed by Porter is characterized by the growing emphasis on national security in the United States during the rise of communism and the onset of the cold war. It begins with the Mutual Security Act of 1951, which was designed to ?contain Communism the United States began to allocate foreign assistance to our allies on the rim of the Soviet Union? (Porter, 1990, p. 7). This period of aid is significant because foreign aid policy was used by the US to secure positions around the world in countries considered to be in the national interest for the containment of communism. According to Porter, this period of economic aid became overshadowed by military aid to countries as seen by the reversal of aid allocation ratios ?for every dollar allocated for economic aid the United States was allocating four dollars for military aid?(p. 7). However, during this period Public Law 480 was passed in 1954 as part of the Food for Peace initiative. This authorized government surplus commodities to be donated to PVOs registered with the Advisory Committee on Voluntary Foreign Aid and distributed in developing countries (Meyer, 1999; Ruttan, 1996). The use of PVOs as official 23 instruments to provide for the world?s hungry served further to strengthen the formal relationship between PVOs and the government. The last period outlined by Porter reflects the notion of foreign aid for the purpose of developmental assistance. The idea of developmental assistance was originally proposed by President Truman in his 1949 inaugural address, but did not come to fruition until 1961 in the inaugural address of President Kennedy and the passage of the Foreign Assistance Act of 1961. It was in this period of foreign aid policy that the groundwork was established for PVOs with the creation of the Agency for International Development (USAID). Initially, foreign aid was coordinated through USAID as bilateral transfers from the US to LDCs. ?USAID focused primarily on large-scale improvement projects; however focus changed in 1973 due to administrative and program reforms aimed at providing the basic human needs (BHN) to LDCs? (Berrios, 2000, p. 10). Since PVOs were already providing food to LDCs under PL 480, it was a natural progression to also include them in the provision of BHN to the same countries receiving food aid. Thus, in 1973 Congress declared, in an amendment to the Foreign Assistance Act, that PVOs should be encouraged 24 to expand overseas development efforts with USAID support (Meyer, 1999; Ruttan, 1996; Smith, 1990). However, substantial funding for the PVOs from USAID took time; and the focus remained of government-to- government aid transfers. But, by the 1980s there was growing skepticism about the ability of lesser-developed governments to utilize developmental aid and to provide effective service delivery to their citizens. There was an emphasis in the media on corrupt governments that had taken aid and built palaces and neglected the people the aid was intended to help. According to Chabbott (1999), Donor organizations became less sanguine about the administrative capacities of low-income states to manage structural adjustment programs and, at the same time, deliver services to disadvantaged groups. By the late 1980s international NGOs, as well as national development NGOs emerging in the low-income countries, became the social delivery agents of choice for many donors. (p. 242) The view expressed by Chabbott began to permeate America in the late 1970s and early 1980s and gave rise to an expanded role for PVOs. The choice of the United States government to move towards endorsing PVOs as development 25 agents is further described by Minear, who states that the ?widespread disenchantment with US government foreign aid has made them a preferred channel for government resources? (1987, 206). Though Porter does not recognize another phase in foreign policy, there was also a definite policy shift at the onset of the Reagan administration. The 1980s saw the dawning of a new era in foreign developmental aid under President Reagan, who wanted foreign aid for development dispersed to NGOs to become much more driven by initiatives from the private sector. Further, the first comprehensive USAID policy including PVOs in development and legitimizing their role was adopted in the 1982 USAID policy paper, Aid Partnership in International Development with Private and Voluntary Organizations (Gorman, 1984). This is a pivotal point from which NGOs mushroomed under US foreign policy. Duffield states, ?the transfer of official aid away from government and towards NGOs represents a unilateral decision on the part of donors based upon a wish to improve effectiveness and accountability? (1991, p. 28). NGOs, and specifically PVOs, have risen from half a century of US foreign policy and changing views on foreign aid. 26 Table 2.1 represents the policy legislation that prompted changes in foreign aid. This table builds on the information provided by Porter by adding initiatives from the Reagan Administration. Porter?s work does not provide a current framework for understanding the growth in expansion of the PVO sector after 1980. Therefore, his model has been updated to reflect the changes. Table 2.1 Primary Legislation Year Primary Enabling Legislation Primary Distribution Pattern 1945-1951 Marshall Plan 1948 Western Europe 1951-1961 Mutual Security Act 1951 Food for Peace 1954 Countries on the Rim of the USSR 1961- Present Food For Peace Foreign Assistance Act 1961 LDCs 1973-1981 Basic Human Needs Amendments 1973 LDCs 1981- Present Private Enterprise Initiative 1981 Partnership for Business Dev. Initiative 1990 LDCs Table 2.1 is significant because NGOs responded to the changes in US foreign aid policy that curtailed direct transfers to governments in favor of transfers to NGOs. The GAO reports that in FY 2000 NGOs received $4 billion, while only $1.9 billion was given out as direct transfers to foreign governments (United States General Accounting 27 Office, 2002). This transition occurred in response to the idea that LDCs receiving direct assistance were not effectively utilizing the resources. Further, Radelet (2003), Senior Fellow for the Center of Global Development advocates ?in nations with weaker, more corrupt governments that show no interest in development, USAID should direct funds carefully, with most activities carried out by NGOs rather than by the government? (p. 159). Theoretical Perspectives The research on the growth in the number of private voluntary organizations registered with USAID is embedded in three major discussions within Public Administration: nonprofit administration, theories of government failure, and outsourcing and privatization. Nonprofit Administration Nonprofit organizations are created when people voluntarily come together to provide a service to the public or to assemble around common causes and interests, either from the need to be social and connected with others or as rational individuals motivated by extra incentives, as presented in the works of Truman (1951) and Walker (1983) on interest groups in America. Further, ?in a civil society, citizens freely engage one another in dialogue and 28 debate; and if they so desire, they form groups?the voluntary associations? (Garvey, 1997, p. 307). In other words, insomuch as at the basic level PVOs are groups of people with common interests serving common causes, their formation is natural. According to Truman (1951), people have a need to be social which manifests in varying degrees of organization; that is, certain of the relationships among a collection of men regularly occur in consistent patterns and sequences. At one end of the spectrum, nonprofit organizations are grassroots community associations of people that provide local goods and services, while at the other end of the spectrum they are classified as Private Voluntary Organizations by USAID with associations and service provision extending around the world. Moreover, like local nonprofits they generally have fewer tiers of authority and use consensus to make decisions (Weiss & Collins, 2000). Private voluntary organizations like nonprofit organizations fill a niche within the market or society that is underserved or not served at all by the government (Berrios, 2000). People organize and form private voluntary organizations to do what they feel the market cannot do or will not do. These organizations are not 29 perfect but they are better than the alternative, which in many cases in LDCs is no provision at all. ?Given expanding market economies and shrinking states, PVOs fill a growing void by responding to the needs and demands of the poor and marginalized sections of society? (Kamat, 2003, p. 65). Not only do PVOs provide services to the underserved and underprivileged peoples of the world, they have a reputation for being efficient and effective lifesaving organizations. The presumed effectiveness of private voluntary organizations in program delivery and meeting the needs of the poor is what has attracted interest and support for private voluntary organizations (Brodhead, 1987). Private Voluntary Organizations are nonprofit organizations that provide goods and services to the international community with a primary focus on Lesser Developed Countries. They are tax exempt nonprofit organizations, as classified by their tax status of 501(c)3, that are registered with USAID and provide humanitarian, development, and relief services to foreign countries. Because PVOs are simply nonprofit organizations providing services internationally, most of the same concepts apply to them that apply to local nonprofits. 30 As nonprofits, PVOs receive funding from both public and private sources to carry out their objectives and the objectives of their donors. PVOs, because of their diverse private funding sources and government funding, have been increasingly challenged to be better stewards of their donated resources and public funds. The idea that the greater the public funding, the more the nonprofit is subject to government control is a belief that has been widely discussed in the nonprofit community. It is here that the impact of government funding on nonprofit organizations is seen most clearly. Nonprofit organizations dependent on government funding over time become very responsive to the values of the government. Like local nonprofits, international nonprofits are also subject to government dependency and neutrality concerns. PVOs, regardless of their funding source, espouse independence of influence from political interest. According to the Code of Conduct for the International Red Cross Movement and NGOs in Disaster Relief, Humanitarian aid is not a partisan or political act and should not be viewed as such. Aid will not be used to further a particular political or religious standpoint ? Humanitarian NGOs shall endeavor not to 31 act as instruments of government foreign policy. Humanitarian organizations are agencies which act independently from governments. (www.icrc.org/web/eng/ siteeeng0/) The idea of independence is not limited to NGOs performing humanitarian work, but also extends to organizations doing development work and providing other services. GAO Report NSAID-96-34 cites PVOs as ?valuing their independence of action, and some believe that close associations with governments could limit their freedom to pursue their mission? (United States General Accounting Office, 1995, p. 4). Moreover, the United Nations (UN) also endorses NGO independence when it solicits support from this community to provide humanitarian and development assistance to lesser developed countries (LDCs) around the world. UN Resolution 46/182 states, ?humanitarian assistance must be provided in accordance with the principles of humanity, neutrality and impartiality? (un.org/documents/ga/res/46). The resolution calls for NGOs to be politically neutral in their provision of services to people in need. However, in the view of Barnett (2005), humanitarian principles were completely shattered in places like Kosovo, Afghanistan, and Iraq, 32 where many agencies were funded by the very governments that were combatants and thus partly responsible for the emergency. This is a clear example of how ?government funding of nonprofit organizations tends to transform nonprofit organizations from representatives of minority interest to representatives of broader constituents? (Lipsky & Smith, 1993, p. 97). As nonprofit agencies, whether working at the local, national or international level as a PVO, grow in number and their government funding increases, must find the proper balance between public and private funding in order to maintain their independence. Lipsky and Smith (1993) summarize many of the contemporary issues about nonprofits in five imperatives that may be impacted when services are provided by the third sector: equity, responsiveness, accountability, efficiency and fiscal integrity. These imperatives are significant in the nonprofit discourse, whether the organization provides services domestically or internationally, because of the relationships between the nonprofit, public, and private sectors under this decentralized system of development aid provision by the third sector. A buffer is created between the government and the client or the donor and the client when services 33 are provided by PVOs, which may call into question the five imperatives. Smith (1984) maintains that PVOs are indeed accountable organizations with mandatory auditing requirements and published annual reports. In spite of the decentralized system, the government and private donors have demonstrated that they trust PVOs to provide services in an efficient manner that reflect American policy and sentiment to foreign citizens. Government Failure In public administration, Wolf (1979) lays the foundation for the theory of government failure in his classic work A Theory of Non-Market Failures. Three of the four types of non-market failure outlined by Wolf, which he attributed to the greater role of private voluntary organizations in the area of development assistance (DA), are internalities, redundant and rising costs, and derived externalities. According to Grand (1991) internalities, redundant and rising costs, and derived externalities are reasons for government failure based on inefficiency. This idea of efficiency versus inefficiency is central to the idea of government failure. If government is inefficient, then what organization is efficient? Internalities are the standards that agencies apply to guide, regulate, and 34 evaluate agency performance and the performance of agency personnel. They become problematic when they ?bear no clear connection with the public purpose the agency is intended to serve? (Wolf, 1979, pp. 121-122). If government is disconnected from the public, then what is the remedy and what organization is more connected to the public? The USAID development goals of the 1970s emphasized program growth instead of program results. Redundancy and rising costs reflected the agency?s focus on program expansion and the failure to streamline efforts, and were a hallmark of USAID before the Reagan era of outsourcing and privatization. It is widely documented that USAID?s decision to contract out was made largely due to concerns regarding issues such as the inability to contain costs and poor economic efficiency (Berrios, 2000; Ruttan, 1996; Rodinelli, 1989; Zimmerman & Hook, 1996). Derived externalities occur when ?government intervention to correct market failure may generate unanticipated side-effects, often in an area remote from that in which the public policy was intended to operate? (Wolf, 1979,p. 127). The bilateral government-to- government aid provided by USAID had the unintended consequence of bolstering corrupt regimes in lesser 35 developed countries. Kasich (2004) called on Congress to ?cut ineffective foreign aid programs that put money into the hands of corrupt and inept governments? (p. 11). Wolf?s theory of government failure provides a framework for the sentiments expressed by the GAO, legislators, and academics that USAID, the government agency charged with international development, has failed along with the public sector in lesser developed countries. Ebrahim (2003) notes that the increase in NGO funding and attention grew from ?evidence of state failure in service provision and an attendant neo-liberal economic climate of state retrenchment? (p. 1). Governments? inability to provide service to those in need and reflect a positive outcome to the United States Congress and the American public further perpetuated the image of widespread government failure in LDCs. According to Korten (1990), governments and international agencies had acknowledged that they were unable to ?get a range of goods and services to the poor? (p. 6). The late Senator Jesse Helms spoke out strongly against USAID and its shortcomings, and it has been suggested that all U.S. development assistance should be channeled through nongovernmental organizations, including private voluntary organizations. 36 This negative view of the role of the government in development assistance has caused donors to look for alternate sources to carry out development efforts around the world. The shift to PVOs by official aid agencies in the early 1980s was a result of the disappointments of government-to-government aid projects by Congress and the public (Summer, 1977; OEDC, 1988). Meyer (1992) reiterates this point by stating, ?disenchanted with the public sectors in developing countries, donors have turned to NGOs and PVOs as vessels to further their program objectives? (p. 1118). The perception of government failure may have artificially bolstered trust and confidence in the abilities of PVOs to augment the shortcomings of government in development assistance. Further, donors have revealed that they trust NGOs and PVOs more than the governments of developing countries (Al-Saqqaf, 2004). It is clear that PVOs have directly and indirectly benefited from the loss of confidence in the public sector?s ability to help the poor and decrease poverty. ?Support for a greater PVO role in delivering assistance seems to stem from (1) general disappointment with the results of over 40 years of government to government assistance and (2) a perception 37 that private organizations are better able to identify development needs and deliver help? (United States General Accounting Office, 1995, p.4). The rising number of PVOs could be related to failures of state and market to provide all of the requirements of society (Shah, 2003). When the market does not provide a needed service because of profitability, it is normally the role of government to step in. However when government fails, private voluntary organizations are now stepping in to fill the gap. PVOs can take on the function of the state where it has failed; but they are not elected and not indebted to a constituency (Kovach, Neligan, & Burall, 2003). However, though most of the focus of the failure of development assistance is placed on USAID, the problem is actually wider in scope than a single agency. Gradstein (1993) notes, One of the causes of intervention failure has to do with the observation that for reasons such as pressure by interest groups, wasteful rent seeking, legislative failure and others, the government is unlikely to act purely in the interest of the public. (p. 1236) Donor and recipient governments have acquiesced under political pressure and may have rendered bilateral aid 38 ineffective and inadvertently promoted the growth of PVOs who appear to have the purest interest of the poor people. Though government failed to adequately provide for the world?s poor, the market had completely left this segment of the world population out. However, in essence, the new prevailing thought is that ?imperfect markets are better than imperfect states? (Colclough & Manor, 1991, p. 7). The growth in popularity and support of PVOs may also have been a contributing factor in the further decline in state responsiveness to the poor. According to Farrington and Lewis (1993), the widening gap between government and NGO resources makes state inefficiency a ?self-perpetuating reality? (p. 333). The perceived failings of government to provide efficient assistance to the world?s poor is central to the adoption of policies to utilize PVOs as service providers in the areas of humanitarian and development aid. These organizations are viewed by both government agencies and the public as more efficient and cost effective service providers than governments, giving better value for money, especially in reaching poor people (Edwards & Hulme, 1996; Gorman, 1984; Meyer, 1992; Smith, 1984; Sollis, 1992; Vivian, 1994). The push to improve the failings of 39 bilateral aid through official government organizations created a positive environment for PVOs to expand operations into development assistance. Gorman (1984) analyzes six models for assisting Lesser Developed Nations. In the models, he shows the funding flows to LDCs starting from (1) multilateral agencies to governments, (2) government to government, and (3) donor agencies to PVOs. He concludes that PVOs ?constitute an especially suitable vehicle for transmission of BHN assistance, and they could be used in a more expanded and systematic way? (p. 69). ?PVOs have often been characterized as efficient vehicles through which resources can be channeled across national borders directly to the poor, with a minimum of the diversion and dilution that often accompanies aid flows through multilateral and national bureaucracies? (Gorman, 1984, p. 1). PVOs present a solution to the apparent government failure problem and serve as a medium between poor people and the global market economy, which is dominated by capitalistic ideals. Further, the optimism of the proponents of PVOs derives from a general sense of PVOs ?doing good,? unencumbered and untainted by the politics of government or the greed of the market (Fisher, 1993; Zivetz, 1991). Faced with the past of government provision 40 and the current neglect of market interests, PVOs are the logical choice to support development initiatives aimed at those in poverty. The research clearly supports the positive image of PVOs in the area of development which has been essential to their growth in numbers and support in this sector. According to Leonard (2002), PVOs have the resources of credibility, expertise, and appropriate networks readily available, while foreign governments do not. Private Voluntary Organizations are seen as a means to mitigate some of the weaknesses in the development process (Fisher, 1997). The picture of PVOs as willing and able vessels capable of functioning in areas that have thwarted government efforts almost gives them super hero characteristics. Edwards and Hulme (1996) characterized them as the ?magic bullet? able to penetrate the problems plaguing the world?s poor. PVOs have an easy reach behind states? borders (Matthews, 1997), and when properly funded, can out perform government in the delivery of many public services. For example, they are quicker than government to respond to new demands and opportunities (Matthews, 1997). With such laudatory comments about PVOs from the public and academics alike, it is logical that when placed next to the 41 government, which is saddled with the baggage of past failures, they shine like superstars. Outsourcing and Privatization The theoretical basis for outsourcing and privatization comes from public choice theory, which is grounded in laissez-faire individualism and free-market economics supported by public choice theorists. These theorists, such as Buchanan and Tullock (1965), advocate minimal state intervention in the market except in the case of market failures, arguing that private provision of goods and services is more efficient than public provision because of competition in the market place (Berrios, 2000). The idea of government inefficiency played into the push to shrink the role of the state in social areas both in the domestic and international environment. One of the neoliberal goals of the 1980s was to reduce the state?s role in the delivery of public services and rely more on voluntary organizations, which were viewed as being more efficient (Barnett, 2005). The United States government, under the leadership of President Ronald Reagan, embraced a policy of outsourcing and privatization. Contracting out or outsourcing can be defined as the process by which private individuals or 42 companies, working under government contracts, undertake functions that are generally recognized as public responsibilities (Garvey, 1997). The matter of what belongs in the public domain and what responsibilities can belong in the private domain generally depends on the political ideology of politicians, administrators and the public, though it is touted as a decision based on improved efficiency because of increased competition in the private sector. ?The theory of contracting stresses that competition between potential bidders creates incentives to minimize costs, which provides a competitive edge in the market? (Berrios, 2000, 28). Contracting out may be the answer to a sluggish government that lacks the incentive of competition to contain cost and improve service quality. According to King (1994), the potential benefits from contracting out include lower costs, improved service delivery and quality, and greater flexibility. The idea that PVOs? provision of services to LDCs is more efficient than public provision is supported continuously by different sources. Outsourcing or contracting out to PVOs is used by governments for activities where they believe these organizations have a comparative advantage and relevant experience (OECD, 1988). 43 Certain functions and roles are flexible and are able to be transferred from the public to the private sector. The governments? decision to outsource development assistance funds to PVOs is a logical conclusion, given that these organizations typically perform development functions on a regular basis with large amounts of private support. Privatization of bilateral aid to Private Voluntary Organizations entails ?the shifting of a function either in whole or in part, from the public sector to the private sector? (Butler, 1993, p. 17). ?It involves the increased reliance on private actors and market forces? (Feigenbaum & Henig, 1994, p. 185) to pursue humanitarian and development goals. In the outsourcing of development and humanitarian assistance by USAID to Private Voluntary Organizations, the activities of financing and decision making are retained by USAID, while implementation becomes the responsibility of the PVO. According to Henig (1989-1990), ?the Reagan administration began to target programs and assets for privatization early in its first term? (p. 649). The policy initiative that best captures the new approach to foreign aid is reflected in Component One of the Private Enterprise Initiative. Berrios (2000) records Component 44 One as the utilization of the private sector as a delivery mechanism, that is, a greater reliance on private rather than on the public sector. This policy change added a new dimension to foreign aid distribution by USAID and enlarged the role of Private Voluntary Organizations (PVOs) in humanitarian and development assistance. USAID no longer focused solely on advancing private enterprise abroad, but actively began to nurture it at home by encouraging U.S. private firms to become USAID contractors (Berrios, 2000; United States General Accounting Office, 1992a, 1992b). In essence, the outsourcing policy gave PVOs the endorsement of government to take the lead in development initiatives abroad. Increased U.S. government support to PVOs was based on the supposed advantages of PVOs over large aid organizations in seven areas: 1. focus on the poorest of the poor, 2. ability to encourage local participation, 3. ability to innovate quickly in response to needs, 4. small scale community approach, 5. exceptional commitment of staff, 6. low cost provision of services, and 7. focus on strengthening local capacity (Clark, 1991; Hoksenbergen, 1999; Marcussen 1996). There are academics and politicians on both sides of outsourcing theory. Zimmerman and Hook (1996) cite PVOs as 45 able to provide quick and flexible responses for people who need help with less cumbersome reporting requirements and bureaucratic structures. However, they have the weakness of being more costly and/or requiring more management attention than is initially apparent. There is no doubt that PVOs should be involved in development assistance, because PVOs are groups of people that have organized for the purpose of helping other people. Organizations such as CARE, Concern, the International Red Cross, and World Vision would help those impoverished and in need without the proverbial nod of the government. Supporters of PVOs view them as efficacious and economical organizations that are able to provide a people-to-people response to development challenges instead of a government-to- government approach. However, more research is needed on PVO activities to determine how accurate the claims about the advantages of PVOs are (Roberts, 1984). The human and personal touch, alongside apparent know-how, continually gives PVOs the advantage over governments characterized as bureaucratic and bumbling in the area of development assistance. The Office of Management and Budget issued a report in the late 1970s in support of government subsidies to PVOs 46 because they have both lower overhead costs than government aid programs and greater innovation (Smith, 2000). In the theory of outsourcing, the sector that is best able to perform the function efficiently should be assigned with its responsibility. PVOs should be rewarded with a greater role in development assistance if they have the comparative advantage relative to government provision. In addition, Smith (1984) states that PVOs use donor money efficiently and generally keep administrative and management expenses around 5 to 6% of their total budget, as compared to 20 to 30% overhead costs with government to government AID programs. Moreover, Gorman (1984) reemphasizes this point by asserting that, ?PVOs can typically make an aid dollar stretch much further than bilateral and multilateral agencies can, while focusing those dollars directly upon the needs of the poor? (p. 59). The ability of PVOs to present themselves as efficient carriers of help abroad directly to those in need makes them the choice of the public to carry out development assistance. A questionnaire given to aid agencies of Development Assistance Committee countries found that the most frequently quoted reason to give financial support to NGOs (PVOs) was their attempt to ensure that the poorest groups 47 benefit directly from developmental activities (OECD, 1988). The expansion of the PVO environment, if not caused by government outsourcing, was certainly done with the blessing of government because of their image as having the virtuous qualities of efficiency, accountability, and connectedness to people. Over time there has been some evidence that NGOs commonly perform better than government or commercial institutions in terms of their capacity to meet the needs of the poor (Farrington, Bebbington, Wellard, & Lewis, 1993; Smillie & Helmich, 1993). Proponents of outsourcing tout cost savings, flexibility, greater efficiencies, and competition; while those against outsourcing argue accountability issues, PVO dependency on government funding, and lack of empirical evidence (Berrios, 2000; United States General Accounting Office, 1996). In the absence of evidence and program evaluations, it is hard for PVOs to defend themselves against claims that they are no better than government. Hira and Parfitt (2004) criticize NGOs (PVOs) as suffering from the same problems of producing results, transparency, accountability, and learning as government agencies. PVOs need empirical evidence to support their success stories. Edwards and Hulme (1996) assert, ?there is no empirical 48 study that demonstrates a general case that PVO provision is cheaper than public provision? (p. 963). The topic of PVOs and empirical evidence continues to pervade the literature and cast some doubt on to the reason for the growth in the number of PVOs if not for reasons of efficiency. Ebrahim (2003) contends that NGOs (PVOs) are believed to be more efficient service providers than public agencies, more democratic, and effective in reaching the poor despite a dearth of supportive empirical evidence. It may be that PVOs are a less costly alternative to donor governments in lieu of them being saddled with program administrative details. Holloway (1998) asserts that PVOs are believed better by those controlling the development assistance purse strings. It is plausible that the conjecture of PVO ability was better than the known reality of government failure. Further, Brinkerhoff (2002) argues, The depth and universality of NGO (PVO) comparative advantages have been greatly exaggerated. Donors and practitioners are increasingly questioning assumptions about NGO (PVO) cost effectiveness?effectiveness, legitimacy, and credibility, as they often emerge or change agendas in response to available funding. (p. 49). 49 Evaluations of PVOs have not shown them to be spectacularly better than government. An evaluation performed by Development Alternative Inc. for USAID of 17 PVO projects in Kenya and Niger reported that 8 of the 17 projects failed to return benefits equivalent to cost (Ellis, 1984). Further, a study by Robert R. Nathan Associates, Inc., of 19 PVO projects in 1981 found that 40% of the projects were borderline to unsatisfactory in cost effectiveness (Ellis, 1984). It appears that when PVOs are evaluated they are weak in the area of evaluating the success or failure of their programs. The evaluation Tendler carried out for USAID of NGO activities revealed after a review of 75 cases that many NGOs did not reach the poor majority or the lowest 40%, were not participatory, and were not innovative (OECD, 1988). This study concluded that further research was needed to determine if PVOs really have a comparative advantage in the area of cost (OEDC, 1988). According to Cooley and Ron (2002), contracting out to private voluntary organizations generates incentives that produce dysfunctional outcomes. The contracts to PVOs are often performance based, renewable, and short term, thus creating counterproductive incentives and acute principal- 50 agent problems. This goes against the accepted thought that market based institutions increase PVO efficiency and effectiveness. If PVOs? provision of services is not superior to the government based upon the neoliberal ideals of efficiency and competition, then perhaps the evaluation of government inefficiency is also misguided. Moreover, competition for government dollars and official support among PVOs works against the formation of collaborative efforts and relationships in the PVO community (Covey, 1995; Miller, 1994). If this is the case then PVOs were not selected to provide services based on their comparative advantages. Edwards and Hulme (1996) suggest that the policy shift to outsourcing was based on ideological grounds rather than empirical evidence. There is some concern about whether or not PVOs? delivery of bilateral assistance instead of government-to- government aid is better or worse for the receiving country. Gorman (1984) makes the point that lesser developed countries would prefer PVO services to government-to-government aid because PVOs typically do not tie their aid to the purchase of donor country products or deflationary policies to acquire loans as the World Bank does. However, the most profound argument against the 51 advantages of PVOs over government provision portrays PVO provision of services as a band aid which covers the problem without addressing the causes of poverty and structural injustice throughout the third world. According to Manji and O?Coill (2002), The NGOs today contribute marginally to the relief of poverty in Africa, but significantly undermine the struggle of the African people to emancipate themselves from economic, social, and political oppression. Development NGOs have, unwittingly or otherwise, become a part of the neo-liberal system that has resulted in widespread impoverishment and the loss of the authority of the African states to determine their own agenda. (p. 567) Further research was done by Meyer (1999) to gain a local perspective on the PVO dynamic in place of public sector activities. She found that in the Dominican Republic and Ecuador, professionals in both countries unanimously considered NGOs more advantageous because of their position outside of the political process. However, after researching the impact of the sharp policy shift in the 1980s of USAID in Latin America, Meyer maintains that 52 ?externally funded NGOs [PVOs] cannot replace the public sector? (1999, p. 61). Woodrow Wilson (1887) contended that it was necessary to determine the things which government can do properly, successfully, and efficiently. If it is determined that government does not, or no longer has the advantage in the production of a desired good or service, then that function should be shifted to the sector that has the advantage. In the case of the growth of PVOs as service providers, political theorists and economists deemed them to be more efficient than government provision and shifted government funding to this sector accordingly. The theory of outsourcing prompted government to make public funds available to private organizations to carry out social support functions. Research Variables Partisanship of the President The partisanship of the president is extremely important for the policies that have encouraged the increase in PVOs. The policies which changed the funding requirements of PVOs saw the greatest expansion under Republican administrations. According to Berrios (2000), the Reagan administration wanted to streamline USAID; and 53 central to this was the Private Enterprise Initiative in 1981, which in effect shifted aid from basic human needs (BHN) in LDCs to the promotion of private sector activities (11). Further, Lebovic?s (1988) study analyzing foreign aid in the Carter and Reagan years indicates that economic interests determined assistance allocated to the least- developed countries in both Administrations; ?Human needs are found to play a secondary role in both Administrations? (p. 115). In effect, the policies adopted by the Reagan administration resulted in the outsourcing of service delivery by USAID to contractors. The popularization of the free-market policies of President Reagan convinced much of the aid community that the private rather than the public sector must lead development efforts (Meyer, 1992). The Bush I administration followed the same pattern as that of the Reagan administration. However, the focus of development assistance shifted during the Clinton Administration from reform to protecting it from further cuts pushed by a conservative Congress (Berrios, 2000). This research and the literature on the growth of private voluntary organizations presented earlier lead to the following hypothesis: 54 H1. In terms of comparing the annual increase in PVOs, Republican control of the Executive Branch will have a positive impact on the annual increase in PVOs. Congress Congress is responsible for the legislation passed either in support of PVOs or against them. The initial laws to include PVOs in development assistance were passed while both the Executive Branch and the Legislative Branch were Democratic. However, the laws to encourage the funding of these organizations with public funds were generated during a Republican Congress and Presidency. Thus, though the Democratic Congress passed legislation to include PVOs, they were not the catalyst for outsourcing development efforts. It is presumed that if the legislative branches are Republican, then they will pass laws that support outsourcing and privatization. Over the last 30 years U.S. laws have expanded to encourage the funding of PVOs by the federal government. On October 6, 1978 Public Law 95-424 was passed by Congress and specified that bilateral development assistance should be carried out by the private sector (icreport.loc.gov). This law on inclusion made it clear that PVOs had a place in development. Further, Public Law 96-533 passed on 55 December 16, 1980 directed the Administrator of USAID to encourage PVOs to work with world hunger problems abroad (icreport.loc.gov). By directing the USAID agency head to include PVOs in development more, Congress was expanding the PVO role. During Republican control of Congress and the Presidency, laws mandating funding minimums for PVOs were passed. Public Laws 97-113 and 99-83 required USAID to provide Private Voluntary Organizations with at least 12% and 15% of the Development Assistance fund, respectively. This laid the framework for an alternative means to provide development assistance. Instead of aid being provided government-to-government, the structure allowed aid to be outsourced to PVOs and services to be provided directly to those in need. This change in public funding has been a tool driving the growth in the PVO market. It is clear that over time the Congress has intervened favorably for the PVOs and encouraged and financed their participation in development assistance around the world. The requirement for PVOs to have 20% private funding was the only law that was a significant barrier to PVOs receiving government funds; and this was rescinded in 2005 during a Republican controlled 56 Executive Branch and Congress. A telephone interview with Ms. Harriet Powers of USAID revealed that the privateness rule was removed to encourage more organizations to register with USAID and receive funding. Undoubtedly, this will increase the competition among the PVOs for government funds. Table 2.2 charts these changes in policy. Table 2.2 PVO Funding Laws Legislation Date Purpose 95-424 October 6, 1978 Urge use of PVOs for Bilateral Development Assistance 96-533 December 16, 1980 Direct Administrator of USAID to use PVOs 97-113 December 29, 1981 12% minimum to PVOs 20% private funds required 99-83 8 October 1985 15% minimum to PVOs 22USC2151(u)g 2005 Rescind 20% private funds requirement It is undeniable that the number of registered PVOs increased amid all of the legislative incentives introduced to the PVO sector. The number of PVOs registered with USAID almost tripled from 144 in 1982 to 419 registered organizations in 1994 (United States General Accounting Office, 1995). Further, Dichter (2003) shows the increase in the number of organizations from 205 in 1988 to 439 in 57 1999 (73). This expansion is still continuing; and there are currently 534 PVOs registered with USAID. This research and the literature on the growth of private voluntary organizations presented earlier lead to the following hypothesis: H2. In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs. Public Law 97-113 Public Law 97-113 established the first funding minimum in support of PVOs. This legislation was introduced to the Senate Foreign Relations Committee by Sen. Charles Percy (R-Il). ?The Congress has continued to be confident about PVOs and in the early 1980s required that 12-16 percent of AID?s annual budget be channeled through PVOs working overseas? (Smith, 1990, p. 5). The push to encourage PVO funding was expanded by Congress in 1981 through Public Law 97-113, passed on December 29, 1981, which ?requires that at least twelve percent of the funds authorized for development assistance and disaster preparedness for fiscal years 1982-1984 be made available to private voluntary organizations? (Thomas.loc.gov; see 58 also OECD, 1988). Though the minimum was set at 12%, the target amount was 13.5%. From the mid-1960s until this legislation, USAID funding of PVOs doubled (Ruttan, 1996). Congress continued to be supportive of PVOs and increased the minimum amount authorized to PVOs from 12% to 15%, with a target amount of 16%. Public Law 99-83 was introduced to the Senate Foreign Relations committee by Sen. Richard Lugar (R-In) and passed on August 8, 1985; and ?it earmarked for private and voluntary organizations at least fifteen percent of the funds appropriated for any fiscal year for various development assistance programs? (icreport.loc.gov). The 15% amount is still in effect and has not been changed since the 1985 legislation. The impact of the funding minimums established in PL 97-113 and continued in PL 99-83 will be studied to see if the number of PVOs increased as result of the minimum. This research and literature on the growth of private voluntary organizations presented earlier lead to the hypothesis: H3. In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase in PVOs. 59 Private Voluntary Organization Funding According to Hira and Parfitt (2004), ?the NGO expansion happened as a result of them being embraced by the mainstream as development agencies, and huge resources followed to fund their activities? (p. 21). The funding of a PVO is the vital substance that keeps it alive and healthy. Programs cannot function and missions organized without financial support. PVOs derive their support from both public and private sources (Natsios, 2001). A report from the Development Assistance Committee within the Organization for Economic Development referred to the NGO environment as the growth sector given the amount of funding they receive from both private sources and other official aid agencies (Brodhead, 1987). PVO funding is significant because as nonprofit organizations their revenues are typically generated from donations through fundraising. An organization may provide excellent service, but if it fails to garner adequate support it will cease to operate. PVOs in recent years have started trying to diversify their funding streams. Reliance on fundraising has always been a part of the PVO environment; however, as the sector has evolved organizations have begun to seek both public and private funding (Dichter, 2003). 60 Fundraising is not easy and requires time, money, and connections. The biggest pressure facing PVOs is fundraising (Holloway, 1998). Over the past 30 years PVO funding has increased dramatically from both the public and private sector. According to Manji and O?Coill (2002), the increase in funding in western countries of development organizations caused them to mushroom and experience spectacular growth. However, empirical studies have not been done to substantiate the relationship between PVO funding and the proliferation of organizations that has taken place over the last 30 years. The majority of the funds in the PVO sector come from private contributions. Given the amount of money that is given to PVOs each year from private sources, most organizations fundraise year round. According to Natsios (2001), the funding is primarily generated from mass media appeals, most notably television, direct mail, and major donor contributions. PVOs compete for support, promoting their causes to the public; those which are the most compelling receive funds. Further, PVOs with a religious affiliation or sponsorship receive support from congregations and faithful parishioners. For example, Samaritan?s Purse is a PVO registered with USAID that is 61 specifically supported by Assemblies of God churches worldwide. There were ?439 organizations registered with USAID in 1999 and seventy-six percent of their overall support came from private sources. The total income of these organizations was $12.32 billion of which $9.38 billion was private support? (Dichter, 2003, p. 106). The laws and operational environment made it advantageous for PVOs to seek out private funding aggressively. Since 1981, private funding was a requirement in order to be eligible for government funding. Public Law 97-113 enacted on December 29, 1981 required that 20% of PVO income originate from the private sector in order for the organizations to be eligible for public funds. The law ?prohibits such funding for a program of a private voluntary organization if the program, after a specified date, does not obtain at least twenty percent of its total annual support for its international activities from non-Federal sources? (icreport.loc.gov; see also Natsios, 2001, p. 391). The law was passed in 1981 and went into effect on January 1, 1985. The purpose of the law was to ensure the autonomy and private nature of PVOs, who may also receive funding from the government and has become known as the privateness rule. However, according 62 to Minear (1987), a study showed that though this legislation was passed to protect PVOs from government control, there is no evidence of an increase in PVO dependency by comparing PVOs from 1979-1980 versus those in 1983-1984 given that private funding outpaces government funding in growth. PVOs received one dollar of government money for every two dollars from other contributions. The privateness rule was rescinded by Congress in 2005, and private voluntary organizations are no longer required to have 20% of their revenues from private sources to qualify for government funds. This change is reflected in U.S. Code Title 22, 2151u(g) and the 2006 VolAg Report of Voluntary Agencies, which state that, ?while PVOs are no longer required to meet the privateness rule, they must maintain a private nature? (VolAg, 2006, p. 5). Public Funding The U.S. government is an attractive source of income for PVOs to augment their support from the private sector (Roberts, 1984). According to Dichter (2003), PVOs prefer public funds even though they are harder to get because they typically come in larger amounts, can be used for overhead costs, and are less expensive to get than private funds. It is irrefutable that private funding has 63 consistently been significantly higher than public funding, but the rate of growth in public funding has outpaced private funding. ?Between 1973 and 1986, USAID assistance to U.S. NGOs [PVOs] increased from $39 million to $450 million, almost a twelvefold increase, while NGOs [PVOs] private source income went from $673 million to $1.86 billion [a threefold increase], an increase in volume of huge proportions but, in percentages, less dramatic than that of the public source of income? (Dichter, 2003, p. 73). The availability of government funds has fueled the expanding PVO environment. In 1981 there were 156 agencies registered with USAID, and a third of the organizations received 50% or more of their funding from federal funds (Roberts, 1984). Government funding as a cause of the PVO growth is plausible because according to Meyer (1999), PVO funding from 1982-1992 increased by 50%. No industry can absorb such a dramatic increase in funding from a previously undeveloped source without growing. The increase in funding had to have an affect on the PVO environment. Robbins (1999) suggests there is a good deal of evidence to indicate that PVOs are growing in size and number because of public funding. The evidence that he provides is based 64 on the theories outlined above of government failure and NGO comparative advantages in service delivery and on the qualitative research of Salamon and Anheir (1998), who focus on indigenous nonprofit growth in lesser developed countries, as well as Edwards and Hulme (1996), who cite economic and political ideology shifts as the reason for increased support from governments and official aid agencies. However, there remains no empirical evidence tying the increase in funding to PVOs for a period in time to the increase in their numbers during the same period of time. According to Smith (2000), overall government assistance as a percentage of PVO resources expanded, rising from 27.6% in 1973 to 38.9% in 1982; and the portion of the AID budget channeled through PVOs rose from 1.5% in 1972 to 14.3% by 1982. The shift to include PVOs in development is best understood when looking at the increase in government funding for various time periods. Whether the increase is analyzed in dollar or percentage terms, it is still dramatic. Clearly the scale of US official funding of PVOs has increased from the 1970s, when less than 2% of PVO income came from official donors, to the mid-1990s, when this figure rose to 30% (Manji & 65 O?Coill, 2002). The increase in funding clearly creates an incentive for new organizations to form. Holloway (1998) notes that the increase in government funding has not only gone to existing PVOs, but also to new organizations that have been created to avail themselves of the financial boom. However, regardless of the financial support, PVOs would still exist because people by nature organize themselves in groups, as confirmed by Truman (1951). The OECD report of 1988 makes it clear that PVOs would exist even if there were no public funding. These organizations were around before the cooperative relationships were formalized. This research and the literature on the growth of private voluntary organization presented earlier lead to the following hypothesis: H4. In terms of comparing the annual increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. Research Model The various factors identified in the hypotheses culminate in the model presented below in Figure 2.1. This model depicts four factors influencing PVO growth during 1968-2004 and the direction of the influence. In the model design, there are two independent variables, two intervening variables, and one dependent variable. The independent variables are the Partisanship of the President and the Partisanship of Congress. The independent variables will be used as predictor variables for the dependent variable in the regression model. The two intervening variables are Public Law 97-113 and the change in public funding to PVOs. These two variables are placed in the model as intervening variables because they are a result of the actions of both the President and the Congress. These relationships will be explored in greater detail in Chapter 3. Independent Variables Dependent Variable Partisanship of the President Partisanship of Congress + Change in Number of PVOs 66 + + Intervening Variables Public Law 97-113 Change in Public Funding of PVOs Figure 2.1 Variable Path Diagram 67 Conclusion This chapter explored the theories surrounding private voluntary organizations and the proposed variables linked to their growth. The literature is expansive, but what is missing is empirical evidence. It is the objective of this study to add quantitative data to the body of literature presented. None of the research presented in the literature addresses the causes of PVO growth through empirical research. 68 CHAPTER 3 METHODOLOGY Introduction This chapter describes the methods used in this research to determine the factors related to the growth in the number of Private Voluntary Organizations. The analysis in the following chapter seeks to answer the question: Is the number of Private Voluntary Organizations registered with USAID related to the partisanship of the President, the partisanship of Congress, Public Law 97-113, and/or government funding for PVOs? The numbers for changes in public funding were adjusted for inflation and converted to 1990 dollars. The conversion was performed by Trey Hamner from the Auburn MBA program. The motivation behind this research is to add empirical data to the body of literature explaining the PVO environment. The method for this research is presented in five sections: explanation of hypotheses, population, data collection, regression analysis, and expected findings. 69 Explanation of Hypotheses H1. In terms of comparing the annual increase in PVOs, Republican control of the executive branch will have a positive impact on the annual increase in PVOs. The partisanship of the executive branch is included as a variable in this research to determine if there is a relationship between the partisanship of the President and the increase in the number of PVOs. Lebovic (1988) claims that in both the Carter and Reagan administrations aid amounts were determined by economic interests. However, it is the Reagan administration that popularized free-market policies and private development efforts instead of public efforts (Meyer, 1992). In contrast according to Berrios (2000), the Clinton administration attempted to protect development assistance from further cuts. H2. In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs. The partisanship of the House of Representatives and the Senate is incorporated into this study to determine if the political ideology of Congress is a factor that causes increases of PVOs. Congress has consistently passed laws 70 that have been favorable to PVOs. In 1978 Congress encouraged the use of PVOs by USAID in development assistance. This was followed up with a law in 1981 that required 12% of development assistance funding to be channeled through PVOs (icreport.loc.gov). This was raised to 15% in 1985. Smith asserts that the support of the PVOs from Congress is reflective of their continued confidence in them in the early 1980s (Smith, 1990). According to Ruttan (1996), Congressional passage of legislation on PVO funding led USAID to increase the level of support quite considerably. H3. In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase in PVOs. Public Law 97-113 is included because it established a set public funding percentage for PVOs from the USAID development assistance budget. The law states that ?at least 12 percent of the funds appropriated for any fiscal year for various development assistance programs is earmarked for private voluntary organizations? (icreport.loc.gov). 71 H4. In terms of comparing the annual increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. Public funding is included in this study as a variable because of the very significant increase in the amount of public funding to PVOs over the last 30 years. According to Dichter (2003), there has been a twelvefold increase in the amount of government funds going to private voluntary organizations during 1973-1986. Meyers (1999) reports that government funds to PVOs increased by 50% between 1982- 1992. Further, the increase in public funding to PVOs did not just go to established organizations, but also to newly created PVOs (Holloway, 1998). Description of Population The USAID Voluntary Agency reports (?VolAg?) were chosen as the source for the population because of the annual requirement for PVOs to provide information to USAID, which they then compile and report to the public. Further, the reports are signed by the USAID Administrator, who reports to the Secretary of State, and are subject to review by the United States General Accounting Office. The number of PVOs for each year is given in the narrative section of each report and can be ascertained by counting 72 the PVOs listed in the report. The PVOs are presented individually in the report as line items. The extraction of the population from the VolAg reports provided a relatively quick and inexpensive means to collect data about PVOs without the need to conduct a survey. The VolAg reports generate a population size of 32 periods. Each period measures the number of PVOs registered with USAID. The population for this study was taken from the USAID Voluntary Organization Agency reports for the years 1968-2006. The population size is limited because USAID did not consolidate PVO information before 1968 and information for fiscal year 2005 has not yet been published. Though the population is collected from longitudinal data, it will be analyzed as cross-sectional in order to show a relationship between the dependent and independent variables outlined in this study. The VolAg reports provide statements on the amount of income that each organization receives from private and public sources, both individually and aggregately, and the number of organizations registered. An illustration of this is the income reported in the 2006 report for World Vision, Inc.: their total private support was $319,287,000 not including in-kind contributions, with government 73 financial support equaling $94,386,000 for 2004. The aggregate amount of total private support for all PVOs was $9.65 billion and $2 billion for government support in 2004, as presented in the 2006 VolAg report. For 1988 onwards, the information reflected in the yearly report presents information from two years prior. For example, the 2006 report records the number of PVOs and is a compilation of the financial information of PVOs from 2004. Hence, the 2006 report covers the fiscal year 2004. Data Collection Instrumentation The instruments used to collect the data were USAID VolAg reports and U.S. Government websites. These instruments are accepted as valid given that they originate from official sources. Dependent Variable The dependent variable for this research is the number of PVOs registered with USAID. These data were collected from the annual U.S. Private and Voluntary Organizations reports generated from USAID. The reports were downloaded from the Development Experience Clearinghouse site posted on the USAID webpage under the Policy heading. The 74 Clearinghouse is a library of all USAID documents, reports, and publications. Independent Variables The independent variables for this study are presidential and congressional partisanship, Public Law 97- 113, and the annual increase in government funding to PVOs. The data for the independent variables for this study were all collected from U.S. government sources. The data used to determine the partisanship of the executive branch were taken from the White House webpage. This site accurately records historical information concerning the presidents for each of the periods and provides their party affiliation. The data used to determine the partisanship of the Senate and the House of Representatives were taken from the Library of Congress webpage. The reliability and validity of these data on executive and congressional partisanship are without question. The passage of Public Law 97-113 was used as a variable in this study. This law was passed in 1981 and required USAID to allocate 12% of the annual Development Assistance Fund to PVOs. The information for this variable was gathered from the Library of Congress website. 75 The data for government funding were extracted from the U.S. Private Voluntary Organizations reports published by USAID yearly. The validity of government funding amounts provided in the VolAg reports was tested by reviewing the annual financial reports of individual PVOs and comparing it to the data in the VolAg reports. This was important because in order for the aggregate amounts to be accepted as accurately reported it was necessary to ensure the individual amounts for the PVOs were valid and reliable as verified by the annual audited statements. The financial reports of World Vision, Women for Women, Catholic Relief Services, Convoy of Hope, Samaritan?s Purse, and the Near East Foundation were reviewed and showed that the data in the VOLAG reports are excerpts from PVO tax reports and audited financial reports. Once the data for public funding was collected it was deflated using the GDP deflator index and converted into 1990 dollars. All of the five variables presented in this study are believed to be collected from credible sources. The variables and their source are presented below in Table 3.1. 76 Table 3.1 Data Collection Hypothesis Variable Source DV Number of PVOs USAID VOLAG Reports H1 Executive Partisanship Whitehouse.gov website H2 Congressional Partisanship Library of Congress website H3 Public Law 97-113 Library of Congress website H4 PVO Public Funding USAID VolAg Reports Data Measurement The data collected for the dependent and independent variables presented in the hypotheses statements are recorded using three different measurements, as summarized in Table 3.2. First, the dependent variable, which is the annual change in the number of PVOs, is computed by subtracting the number of PVOs in the previous year from the number in the year in question. For example, in 1967 there were 73 PVOs, while in 1968 there were 77. Thus, the annual change in 1968 is: 77-73=+4. Second, the change in government spending is also measured by the difference between spending in a particular year and the level of funding in the preceding year. For this variable, in addition, the dollar amounts collected for each observation 77 are deflated using the GDP deflator and converted into 1990 dollars to make the amounts for each year comparable. Third, the other three indicators used in this model are ?dummy variables? that are coded ?1? for the presence of an item and ?0? for its absence. Republican control of the executive branch is coded as ?1? if the President of the United States is a Republican and ?0? if the President is a Democrat. Republican power in Congress is coded as ?1? if the Republicans control either the House of Representatives or the Senate and ?0? if Democrats control both houses of Congress. Finally, Public Law 97-113 is coded ?1? for the years after the law was enacted and ?0? for the preceding years. Table 3.2 Data Measurement Hypothesis Variable Measurement DV Annual increase in PVOs Numerical units H1 Executive Partisanship 0=Democrat, 1=Republican H2 Congressional Partisanship 0=Democrats control both Houses; 1=Republicans control either House or Senate H3 Public Law 97-113 0=Before Law; 1=After Law enacted H4 PVO Public Funding GDP (Deflator adjusted $) 78 Data Limitations The data collected for government funding to PVOs from the VolAG reports were not reported by USAID for the years of 1975-1978 and 1982. Thus, there are several years of missing data. Further, the population size is a limiting factor for this study. There are four independent variables, with a population size of 31. Stages and Types of Analysis The statistical analysis testing the model that has been developed to explain the growth of PVOs will be presented in three stages. First, Chapter 4 uses graphs to show how each of the variables change over time and how closely the changes in the four independent variables appear to be associated with the growth pattern of PVOs. Chapter 5 then applies regression analysis to derive precise statistical estimates of these relationships. Two major types of regression results are reported. First, simple regression analysis is performed between the dependent variable and each of the predictor variables. Second, multiple regression analysis is performed with the relevant predictor variables used to explain the annual change in the number of PVOs. Regression Analysis Regression is the principle statistical technique used to model the relationship among variables at the interval level of measurement, which all the variables in this analysis are (this section is based on Clark, 2007, Chapter 12). The basic data for a regression can be expressed as a ?scatterplot? graph such as the one in Figure 3.1 below for the relationship between the change in the number of PVOs each year and the corresponding increase or decrease in the public spending devoted to PVOs in that year, measured in millions of dollars. Figure 3.1 Increase in number of PVOs vs. increase in public spending 79 80 By convention, the cause or independent variable is plotted along the horizontal or x axis, across the bottom of the graph, and the dependent variable is plotted along the vertical or y axis. In Figure 3.1, PVO increase is the dependent variable whose value for a specific year is explained by the amount of government spending for PVOs that year. Each of the 31 dots in Figure 3.1, therefore, represents one of the years in the analysis and is placed in the scatterplot according to its value for the two variables. For example, the lowest dot in the graph is for 1996, when the number of PVOs fell by 17, while government spending increased by $29 million. The fourth hypothesis predicted that there would be a positive relationship between these two variables. That is, the greater the increase in government spending, the larger the number of new PVOs will be. A cursory examination of Figure 3.1 suggests that this is in fact the case. As the rise in government spending increases across the graph, the increase in the number of PVOs also appears to go up. This impression is confirmed by the ?regression line? that is superimposed on the scatterplot. This regression line provides the best linear fit for the pattern of dots in the scatterplot. Technically, it is the 81 line that minimizes the squared deviations on the data points from the line. As for any straight line, it is defined by the equation Y=a+bX. Here, a represents the ?intercept? or the value of Y when x=0. In turn, b represents the slope or the amount that Y changes for every 1 unit increase in X. For the regression line in Figure 3.1, this equation is: Y=6.256 + 0.086X. Although often the intercept has little substantive meaning, here provides some relevant information by telling us that if government funding remains constant from one year to the next, the number of PVOs would still be expected to increase by just over 6. The slope coefficient is generally much more important. Here, the b value of .085 shows that every one million dollars of government spending is associated with only an increase of .086 PVOs. More meaningfully perhaps, this indicates that one additional PVO will be formed for each $12 million increase in government expenditures. Even the most cursory glance at Figure 3.1 shows that the regression equation does not provide a perfect prediction of PVO increase based on the change in government spending because many, if not most, of the data points lie at an appreciable distance from the regression 82 line. This brings us to probably the most important set of results in regression, which measure the strength of the association or relationship between the two variables. Two measures of strength are commonly reported in regression analysis. The first is the correlation coefficient, Pearson?s r, and the second is the statistical significance of the relationship. A correlation coefficient, such as Pearson?s r, provides a standardized measure of how strongly two variables are related. It varies between 0 when there is no association at all to either plus or minus 1 when there is a perfect relationship. Technically, a perfect relationship occurs when the values of the dependent variable can be exactly predicted from the values of the independent variable or, in graphical terms, when all the data points fall precisely on the regression line. Note that the positive or negative sign of the correlation denotes the direction of the relationship, not the strength of the association. For example, an r of -.65 is much stronger than one of .35. Table 3.3 provides a rough set of estimates to interpret the strength of a correlation. Based on this scale, r values of over .5 are considered strong, those between .25 and .49 are moderate, and so 83 forth. In addition, the square of r (r 2 ) represents the percentage of the ?variance? in the dependent variable that is statistically explained by the independent variable. Table 3.3 Approximate Strength of Correlation Coefficient (Cronk, 2006) Correlation Strength .00-.14 No Association .15-.24 Weak Association .25-.49 Moderate Association .5-1.0 Strong Association The other measure of strength is the statistical significance of r and b, which is the same for both. Significance is the probability that the observed relationship between two variables in a scatterplot reflects ?random chance? rather than actual association. When the probability is low, conventionally 5% or less, the relationship is said to be statistically significant. Significance is calculated from ?test statistics,? such as F or t, which have known distributions that can be used to calculate the probability of an F or t of a given size. The size of these statistics and, hence, the significance of a relationship is based on two factors. First, the 84 higher the correlation between two variables is, the more likely it is to be statistically significant. Second, the larger the number of cases in an analysis, the higher is the confidence that can be put in the results and, consequently, the more highly significant a correlation of a given size will be. While significance technically can only be applied to representative samples drawn from larger populations, it is often used in other analyses, such as the one here that uses data from all available years, as an estimate of how important a relationship is. Table 3.4 presents all the correlations and regression statistics for the relationship between PVO increase and government spending increase. The relationship between the two variables is a moderate positive one, as denoted by the correlation coefficient of +.41, which is in the direction predicted by H4 and which also corresponds to the upward sloping regression line in Figure 3.3. The r 2 of .17 shows that the increase in government spending explains 17% of the variance in the growth of PVOs; that is, it has a 17% predictive power. Finally, despite the fairly small number of cases (31), the r of .41 is strong enough to produce an F of 5.75, which is significant at the .025 or 2.5% level. Since this is well under the 5% standard that is often used 85 to denote statistical significance, we can conclude the relationship is highly significant. Table 3.4 Correlation and Regression Statistics for the Relationship Between PVO Increase and Government Spending Increase Statistic Value R +.41 r 2 .17 F 5.75 Significance .025 A 6.256 B .086 Multiple Regression Multiple regression is an extension of simple regression that uses two or more independent variables in a linear equation to predict the dependent variable (this section is based on Clark, 2007, Chapter 13). The linear equation for multiple regression is represented as Y=a+b 1 X 1 +b 2 X 2 +?b z X z . As with simple regression, in multiple regression the assumption is made that all the variables can be measured as either ratio or interval scales. As before, a reflects the value of Y when all the X values for the independent variables in the linear equation are zero. 86 As an example of multiple regression, let us examine how the increase in government spending and Republican control of the executive influence PVO growth together. As will be seen in Chapter 5, the partisanship of the President affects PVO growth at a two-year time lag, while the effects of changes in government spending occur in the same year that they are made. Thus, in this multiple regression the PVO growth for a given year is correlated with whether or not the President was a Republican two years previously. The results of this regression are presented in Table 3.5, which is actually composed of two separate tables. The top part of the table reports the statistics for the combined impact of the two independent variables on PVO growth. These results are exactly the same as in the simple regression summarized in Table 3.4. The strength of the association is measured by a correlation coefficient. Here, it is termed Multiple R, rather than simple r, to denote that it is estimating the combined impact of two or more independent variables. The Multiple R of .6 and R 2 of .36 in Table 3.5 show that the two independent variables have a strong impact on PVO growth that accounts for 36% of the variance in the dependent variable. Also, as in simple 87 regression, the statistical significance of Multiple R is measured by an F statistic. In this case, F is 7.72, which is highly significant at the .002 or .2% level. Finally, the value of a in this multiple regression equation is 0.72. That is, if both independent variables are zero in a specific year (there was no change in government spending and the president was a Democrat two years previously), the number of PVOs would be expected to grow by a little less than one. Table 3.5 Multiple Regression Explaining PVO Growth by Republican Presidencies and Growth in Government Spending for PVOs Overall Equation Dependent Variable: PVO growth Multiple R .60 Multiple R 2 .36 F 7.72 Significance .002 A .72 Separate Effects of Independent Variables Beta B t Sigma Republican President, (2 years lag) +.45 13.74 2.87 .008 Government Spending Growth +.31 0.064 1.97 .059 88 The bottom part of Table 3.5 contains the statistics summarizing the separate effect of each independent variable upon PVO growth, controlling for the impact of the other variable. One is the slope coefficient b, as with simple regression. Here, the b of 0.06 for government spending shows that each $1 million in public spending for PVOs brings an increase of only 0.06 PVOs, while the b of 13.74 for presidential partisanship shows that having a Republican, rather than a Democratic president, is associated with an increase of nearly 14 PVOs two years later. In simple regression, there was only one measure of statistical significance for both r and b. In multiple regression, in contrast, the value of b measures the separate effects of the independent variables and Multiple R measures their combined impact. Thus, each b needs its own measure of significance. Here, they are provided by the test statistic t. For presidential partisanship t is 2.87, which is highly significant at the .008 or .8% level. T is 1.97 for government spending, just missing the conventional level of 5% for statistical significance by being at the .059 level. 89 The b values for presidential partisanship and change in government spending indicate how much PVO growth will be affected in numerical terms by a one unit increase in one of the independent variables. They do not, however, give a measure of the relative strength of the influences exerted by these two factors because of the considerable difference in the scale of measurement for presidential partisanship (which varies between 0 and 1) and government spending (which varies between $58 and $216). Thus, a one-unit increase in the former would almost inevitably cause a much bigger change in the dependent variable than a one-unit change in the latter, as is certainly the case in Table 3.5. Consequently, statisticians have developed a statistic referred to as beta, or the ?Standardized Regression Coefficient?, that is the b that would result if all the variables had the same units of measurement. The larger beta is, therefore, the greater the relative impact of an independent variable. Betas are thus analogous to the r values in simple regression, and their strength can be estimated using the scale in Table 3.3 above. Indeed, in simple regression with only one independent variable, beta is equal to the value of r. For the results in Table 3.5, 90 for example, the beta for having a Republican President is .45 and that for government spending is .31. This indicates that both of these variables have moderate effects in the predicted direction (H1 and H4) on PVO growth, and that presidential partisanship exerts a somewhat stronger influence. According to Cronk (2006), ?multiple regression is generally much more powerful than simple linear regression? (p. 50). The increase in the predictive power of a multiple regression analysis as compared to the simple one in this illustrative example can easily be seen by comparing Tables 3.4 and 3.5. Adding presidential partisanship to government spending in the explanatory equation results in a substantial jump in the correlation coefficient, from an r of .41 to a Multiple R of .60. Moreover, the predictive power of the model, as denoted by r 2 and R 2 , more than doubles from .17 to .36. The relationship in the multiple regression, correspondingly, is much more significant at the .002 level than is the one in the simple regression at the .025 level. It is also instructive to compare the b values for government spending in the two analyses. In the simple regression, b is 0.086, but it falls to 0.064 in the 91 multiple regression. This indicates that some of the initially observed impact of the change in government spending on PVO growth is actually caused by the partisanship of the President: spending is higher in Republican administrations. Consequently, statistically controlling for the effects of the second independent variable in the multiple regression reduces the amount of change produced by a given increase or decrease in government spending. Similarly, the beta for government spending of .31 in the multiple regression model is markedly lower than its r of .41 in the simple regression. Conclusion This chapter provided a detailed overview of the methods used in this cross-sectional study of Private Voluntary Organizations. The aim of this chapter was to connect the data with the method of analysis in Chapters 4 and 5. In order for the data to have any meaning after it is collected, it must be properly analyzed using the correct method. This information is important because it aids in the interpreting of statistical figures about the relationship of the independent variables against the dependent variable in this cross-sectional study. 92 CHAPTER 4 EVALUATION OF THE VARIABLES OVER TIME Introduction The findings in this chapter serve to introduce the more sophisticated regression analysis of Private Voluntary Organizations and how the four explanatory factors included in our model influence the growth of PVOs. This chapter details how each of the variables included in the analysis evolved over time. In essence, this is a presentation and analysis of the raw data. Some trends were evident in the PVO environment by simply examining the collected data before statistical analysis. By reviewing the cases collected during 1968-2004, it is possible to gain a clearer understanding of the PVO environment. First the dependent variable is analyzed and then the independent variables are evaluated against the dependent variable. Evaluation of Variables Private Voluntary Organizations Descriptive analysis of the number of PVOs reveals that they have substantially increased in number since 93 1968. Figure 4.1 shows the general trend in the growth of PVOs. The largest amount of growth occurred between 1984- 1995 where 323 organizations were added. Further, the number of PVOs only increased by 27 organizations from the period of 1995-2002. However since 2002, there has been a renewed growth spurt in the increase of PVOs during the administration of Bush II. From 2002 until the present the number of PVOs has grown from 444 organizations to 533 organizations, an increase of 89 organizations. What also can be gleaned from this simple exploration of the data is that the growth in this sector did not really gain upward momentum until the late 1980s. The changes in the number of PVOs are a better reflection of the volatility in the PVO environment. Figure 4.2 reveals that the growth in PVOs has not been as smooth as Figure 4.1 would suggest. During the first term of the Clinton Administration the number of new PVOs dropped from 43 to 29 to 15 for the years 1993, 1994, and 1995, respectively, and eventually saw a loss of 17 organizations for 1996. Conversely, PVOs during the Bush II administration experienced a loss of 3 organizations in the first year and then increased by 8 to 45 to 20 for the years 2002, 2003, and 2004, respectively. 0 100 200 300 400 500 600 1968 1970 1972 1974 1981 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 Years PVO # Number of PVOs Figure 4.1. Number of PVOs from 1968-2006 There appears to be a stark inverse relationship between the increase of PVOs during the Clinton administration and the administration of Bush II. Figure 4.2 provides great insight into the movement of the PVO sector. However, more analysis is needed to determine the causes of the changes in the PVO environment. 94 -30 -20 -10 0 10 20 30 40 50 1968196919701971197219731974198019811983198419851986198719881989199019911992199319941995199619971998199920002001200220032004 Year # PVOs Figure 4.2. PVO changes 1968-2004 Executive Branch Partisanship H1. In terms of comparing the annual increase in PVOs, Republican control of the Executive Branch will have a positive impact on the annual increase in PVOs. Preliminary findings of the validity of this hypothesis that a Republican executive branch will have a positive effect on the increase in the number of PVOs indicate that this is likely to be true. There are 19 cases in this study where the executive branch is Republican: the periods of 1971-1974, 1981-1992 (excluding 95 96 1982), and 2001-2004, as reflected in Figure 4.3. In the first period there was marginal growth in the number of PVOs. However, the subsequent periods are marked with significant growth. The final period, for example, starts from a seven year PVO growth plateau and yields an increase of 45 organizations in 2003 and 20 organizations in 2004. The figure reflects an obvious stagnation in the growth of PVOs during the Clinton administration and though the number of PVOs does not start to increase immediately at the start of the Bush II administration, it is clear that the number of PVOs begins to rise again during the latter part of his first term. Based on the preliminary observations of the coincidence of PVO increases with a Republican executive, there may be a connection between PVO growth and a Republican executive. In cases of the largest PVO increase, the executive branch has typically been Republican. In addition, the effects of partisanship seem to take several years to develop. In particular, the spurts of PVO growth associated with the Reagan and George W. Bush presidencies and the stagnation during the Clinton era clearly commenced only several years after those three presidents took office. It also must be noted that though for the purposes of this study former President Clinton is used as the prototype for the Democratic party, his administration was moderately conservative. The United States has not had a "liberal" president in office in the last forty years. Rep. Exec. Rep. Exec. Rep. Exec. 0 100 200 300 400 500 600 1968 1970 1972 1974 1980 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Years PVO # Number of PVOs Figure 4.3. Republican executives and PVO growth 1968-2004 Partisanship of Congress H2: In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs. 97 98 The trend data in Figure 4.4 do not suggest the hypothesized positive impact of Republican control of Congress upon the growth in PVOs. Congress was controlled by the Democratic Party from 1968-1980 and was split with the Republican Party gaining control of the Senate from 1981-1986. Yet, there was minimal growth in the number of PVOs during both periods. However, both Public Law 97-113 and 99-83, which promoted PVOs (see Chapter 2), were passed during this time, in 1981 and 1985 respectively. Both pieces of legislation were introduced by Republican senators who were also members of the Senate Foreign Relations Committee. The period between 1987-1994 once again saw Congress controlled by the Democrats, followed by the Republican take over of Congress in 1995 until now. Based on Figure 4.4 the greatest amount of PVO growth occurred between the years of 1988 and 1995. This coincides with the timeframe of the Democratic control of Congress. The number of PVOs reached a plateau after 1995 and did not start to regain momentum until 2003. Thus, there appears to be little direct correlation between partisan control of Congress and the growth of PVOs. As stated earlier, the Congress is overshadowed by the agenda of the President in matters of foreign policy. This could very well explain the lack of association between the partisanship of the Congress and the increase in the number of PVOs. Rep. Congress Rep. Senate 0 100 200 300 400 500 600 1968 1970 1972 1974 1980 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Years PVO # Number of PVOs Figure 4.4. Republican Senate/Congress 1968-2004 Public Law 97-113 H3: In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase in PVOs. Preliminary analysis supports the impact of the passage of Public Law 97-113 on the increase in Private 99 100 Voluntary Organizations. The law was passed in 1981 and required that 12% of USAID?s Development Assistance Funds go to Private Voluntary Organizations. The 12% was subsequently raised to 15% with the passage of Public Law 99-83. The preliminary assessment to determine if there is a cause and effect relationship with the passage of the funding minimum to PVOs is a simple observation of the cases before the funding minimum was imposed and those after its implementation. Figure 4.5 depicts the growth of PVOs as stagnant after the passage of PL 97-113 until the funding level was increased to 15%. While there was a lag of several years, the number of PVOs began a major growth spurt in the mid 1980s which lasted until 1995. Even though there have been lulls since the passage of the initial legislation to establish a funding minimum, it is undeniable that the number of PVOs have grown dramatically. However, whether PL97-113 is the cause of this growth requires additional analysis through statistical methods. PL 99-83 PL 97-113 0 100 200 300 400 500 600 1968 1970 1972 1974 1980 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Years PVO # Number of PVOs Figure 4.5. Public Laws 97-113 and 99-83 1968-2004 Public Funding H4: In terms of comparing the annual increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. A preliminary analysis of the trend in public funding to Private Voluntary Organization would lead to the conclusion that PVO increases are a result of changes in public funding. However, more rigorous analysis is needed to conclude this definitively. Public funding for PVOs has 101 102 increased steadily over the past 36 years. Regardless of the reason, it is obvious that the government has shown its support of these organizations by backing them financially. Figure 4.6 below reflects the government?s support of these organizations. It is apparent that funding to PVOs had a slow start with commitment from the government to fund these organizations taking hold only gradually. Further, Figure 4.6 is also a powerful view of the government funding pattern of PVOs during the Reagan, Bush I, and Clinton administrations, and term 1 of the Bush II administration. The growth in PVO funding from the years of 1981-1989 represents a 114% increase in funding; funding from 1993-2001 increased by 84% in real dollars. During the four year period of the Bush I administration and the first term of the Bush II administration funding increased by 78% and 30%, respectively. Additionally, it is also apparent from a review of the trend in public funding during the years of 1994-1997 that changes in public funding of PVOs was marginal. This time period coincides with much of the Clinton administration even though the Republicans had gained control of Congress, thus supporting the earlier finding in this chapter that partisan control of the executive seems much more important than partisan control of Congress for influencing the growth of PVOs. $- $500,000,000.00 $1,000,000,000.00 $1,500,000,000.00 $2,000,000,000.00 $2,500,000,000.00 1968 1970 1972 1974 1981 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 years Public Support Figure 4.6. PVO public support 1968-2004 103 Figure 4.6 illuminates the funding pattern of the U.S. Government for PVOs. What is gained from this information is the certainty that the government has indeed increased support to PVOs over time at a rigorous pace. However, what is not clear is the reason for the increase in monetary support or if the increase definitively produced more PVOs. Even so, the growth of funding shown in Figure 4.6 is strikingly similar to the growth in the number of 104 PVOs in Figure 4.1. Both figures show a high growth period starting in the 1980s and eclipsing in the mid-1990s. There is then a period where both public support and the number of PVOs plateau, followed by another growth period starting in 1998 for public support and 2002 for PVOs. Figure 4.7 plots the actual change in the number of PVOs compared to the preceding year. The increases and decreases in public funding from year to year are not as stable as the growth in the total number of PVOs in Figure 4.6 would appear to suggest. Actually, the PVO funding environment from 1968-2004 has been very dynamic. The public money supply to Private Voluntary Organizations is continuously expanding and contracting from year to year. For example, the annual increase in funding to PVOs from the government declined from 1993-1997. PVOs experienced an increase in funding of $209 million in 1993 and only an increase of $9.1 million in 1997 in real dollars. Further, Figure 4.7 shows that public funding for Private Voluntary Organizations has grown substantially under the Bush II administration, increasing by $363 million dollars in 2004. $(100,000,000.00) $(50,000,000.00) $- $50,000,000.00 $100,000,000.00 $150,000,000.00 $200,000,000.00 $250,000,000.00 1968 1970 1972 1974 1980 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Years PVO Public Support Adjusted Dollars Figure 4.7. Changes in public support of PVOs 1968-2004 The increases and decreases in public funding displayed in Figure 4.7 are undoubtedly connected to the increases and decreases in the number of PVOs displayed in Figure 4.2. In order to link the gains in public funding and the increases in PVOs both should be analyzed together. A convenient way to do this is by looking at both in the light of the percentage changes in both variables during the period of 1968-2004 to measure them on the same scale. 105 Figure 4.8 presents these two time series in the same graph. While public spending has been considerably more volatile, the changes in the two variables do appear to move together in fairly close tandem. Figure 4.8 Percentage Change in PVOs, and Public Funding 1968-2004 -20 0 20 40 60 80 100 120 140 160 196 8 19 70 197 2 197 4 19 81 198 4 198 6 1 98 8 199 0 199 2 199 4 199 6 19 98 200 0 200 2 20 04 Year Per cen tag e PVO Percentage Change PVO Public Funds Percentage Change Figure 4.8. Percentage change in PVOs and public funding 1968-2004 Summary of Graphical Analysis Based on the preliminary graphical evidence in this chapter, some but not all of the hypotheses concerning the causes of PVO growth appear to hold. The expected 106 107 associations between PVO growth and Republican control of the executive, supportive public laws, and changes in government spending are evident in the graphs. However, partisanship in Congress does not appear to be associated with changes in the number of PVOs. These graphs, though, can only suggest associations. The actual relationships can only be confirmed by the more advanced statistical analysis in the next chapter. 108 CHAPTER 5 TESTING THE THEORY OF PVO GROWTH Introduction This chapter focuses on the presentation of statistical data to support or reject the suggested predictors for determining the increase in PVOs from 1968 to 2004. Each independent variable was analyzed statistically through simple and multiple regression analysis and found to be either significant or not significant with respect to the dependent variable. Using regression analysis builds upon the evaluation of the variables over time in the previous chapter and produces quantifiable findings. The data utilized for the analyses are included in the appendix section of this chapter. The original data set was collected primarily from USAID Voluntary Agency Reports. This chapter provides an overall test of the theoretical model that was developed in Chapter 2 to explain the growth of PVOs over the last three decades. The first section establishes the nature of the direct 109 relationships between PVO growth and each of the independent variables by examining the nature of the time lag, if any, that it takes for the causal factor to affect PVO growth. The second section then applies multiple regression to provide a full test of the theoretical model. Modeling the Direct Relationships The impact of some predictor variables on the dependent variable does not always occur immediately. In fact, there are many instances when there is a time delay between the cause and effect relationship of the two variables. For example, the passage of a law to give PVOs public funding may not result in the instantaneous creation of new PVOs. It is generally accepted that it will take a certain amount of time for the interested parties to respond to the mandated funding and form PVOs. Further, the effects may not be felt until a month, a year, or even a decade later for some dependent variables. To determine if the dependent variables? response to the changes in the dependent variables were more significant in subsequent years, the Pearson?s r and significance were calculated for the same year, a lag of 1 year, a lag of 2 years, and a lag of 3 years. This information was obtained by constructing a correlation 110 matrix for each of the independent variables versus the PVO growth. This was an important step in order to assess which variables would be best for the simple and multiple regression models presented later in this chapter. Examining the nature of the relationship between the annual increase in PVOs and the partisanship of the President demonstrates the importance of considering the possibility of time lagged relationships. If a Republican administration and PVO growth are correlated in the same year, there is almost no relationship between them (r=.08), suggesting that the first hypothesis should be rejected. Yet, the graphical presentation in Chapter 4 showed that there seems to be a lag of a year or two between a change in the partisanship of the President and the resulting impact on the dynamics of PVO formation. Indeed, when presidential partisanship is lagged so that it is correlated with the change in PVOs by one, two, or three years, it is clear that a moderate relationship exists since these r values range from .36 to .46. For further analysis, a Republican administration with a two-year lag period was selected as the independent variable because it had the most significant association with PVO growth, as shown in Table 5.1. 111 Table 5.1 Partisanship of the President?s Correlations with PVO Increase Independent Variable Explaining PVO Growth Simple r Significance President Partisanship .08 .68 PresParty, Lag 1 yr .36 .048 PresParty, Lag 2 yrs .46 .010 PresParty, Lag 3 yrs .45 .012 The partisanship of Congress clearly lacks the expected relationship with the annual increase in PVOs; none of the correlations in Table 5.2 are statistically significant. Moreover, except for a very low r of .07 with a three-year lag, all of these correlations are negative, the opposite direction from that predicted by Hypothesis 2. Somewhat arbitrarily, then, the partisanship of Congress will be lagged by two years to make it consistent with the partisanship of the President. Table 5.2 Partisanship of the Congress?s Correlations with PVO Increase Independent Variable Explaining PVO Growth Simple r Significance Congressional Partisanship -.31 .091 ConParty, Lag 1yr -.25 .179 ConParty, Lag 2yrs -.14 .448 ConParty, Lag 3yrs .07 .711 112 Unlike the partisanship of the President and Congress, the correlations between Public Law 97-113 and PVO growth are fairly consistent across the four possible time lags. As reported in Table 5.3, the correlations only range between .39 and .44. A lag of two years was selected for further analysis because the r with PVO growth was somewhat higher than for either the zero or one year lags, and almost exactly the same as the longer three-year lag. Table 5.3 Public Law 97-113?s Correlation with PVO Increase Independent Variable Explaining PVO Growth Simple r Significance Public Law 97-113 .37 .044 PL97-113, Lag 1 yr .36 .048 PL97-113, Lag 2 yrs .44 .015 PL97-113 Lag 3 yrs .44 .014 The increase in government support has an immediate impact upon PVO growth, as indicated by moderate r?s of .41 for the same year and a one year lag. The stimulus provided by government spending is quite short term, though, as the r?s drop to under .15 for longer time lags. Since using a time lag does not affect the strength of the association, the changes in government spending and PVO growth will be correlated with each other in the same year. 113 Table 5.4 Public Funding?s Correlation with PVO Increase Independent Variable Explaining PVO Growth Simple r Significance Public Funding Increase .41 .023 PubFun, Lag 1 yr .41 .024 PubFun, Lag 2 yrs .14 .493 PubFun, Lag 3 yrs .11 .597 The impact of the variables that will be used in further statistical analysis on PVO growth are summarized in Table 5.5, which contains the r, r 2 , b, and significance for these relationships as determined by the simple regression of PVO growth on each. Based on the information presented in Table 5.5, it is possible to accept or reject the hypotheses presented in this research. Table 5.5 Variables Selected for Statistical Analysis Independent Variable Explaining PVO Growth Simple r r 2 b Significance PresParty, Lag 2 yrs .46 .21 14.69 .010 ConParty, Lag 2 yrs -.19 .04 6.0 .339 PL97-113, Lag 2 yrs .44 .19 14.5 .015 PubFun .41 .17 .086 .025 114 Partisanship of the President H1: In terms of comparing the annual increase in PVOs, Republican control of the Executive Branch will have a positive impact on the annual increase in PVOs. Determination: The Pearson?s r of .46 revealed a moderate, almost strong, relationship between the Republican Executive branch and the increase in PVOs, with a predictive power of 21% as measured by r 2 . Moreover, the correlation was found to be highly significant, with a p- value of .010. The b of 14.69 indicates that having a Republican President results in an additional 14-15 PVOs two years later. Thus, the hypothesis that a Republican Executive branch is related to the increase in PVOs is supported. The finding that when the President is Republican the number of PVOs will increase is also supported in the literature. Berrios (2000), Meyers (1992), Barnett (2005), and Henig (1989-1990) all support the findings presented in this research. They assert that the neoliberal resurgence in the executive branch led to an increase in the number of PVOs. The Reagan administration is given credit for pushing outsourcing initiatives which laid the foundation for other administrations to come. According to Meyers 115 (1992), Reagan popularized free-market policies and convinced the aid community that the private sector must lead development efforts. Congress H2: In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs. Determination: The Pearson?s r of -.19 signifies a weak negative relationship between a Republican controlled Senate that is far from statistically significant. Thus, this hypothesis is rejected based on the statistical analysis. The fact that a Congress controlled by the Republican Party is not a predictor for the increase in PVOs is somewhat surprising since it is inconsistent with the literature that Congress passed laws during periods of Republican control which stimulated the growth of these organizations. According to Smith (1990), the confidence of Congress in the PVOs in the early 1980s led to the requirement for a 12% funding minimum of the development assistance fund in 1981 which was later increased to 15%. 116 Public Law 97-113 H3: In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase in PVOs. Determination: The Pearson? r of .44 and r 2 of .19 reveals a moderate relationship between the passage of Public Law 97-113 and the increase in PVOs. The relationship was found to be significant with a p-value of .015. The value of the slope coefficient b shows that beginning with two years after the law was enacted, the number of PVOs grew by an extra 14.5 per year. Thus, the hypothesis is accepted that PL 97-113 is related to the increase in PVOs. From the mid-1960s until this legislation, USAID funding of PVOs had doubled (Ruttan, 1996); before this legislation there were only 156 organizations. After its passage the number of PVOs remained stagnant until the late 1980s, when the number of PVOs began to increase by two- digit numbers annually. The establishment of a funding minimum for government support to PVOs through PL 97-113 was a reflection of the confidence and support of Congress (Smith, 1990). 117 Public Funding H4: In terms of comparing the increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. Determination: The Pearson?s r of .41 and r 2 of .17 reflects a moderate relationship between public funding and PVOs. The relationship between public funding and the increase in PVOs is found to be significant with a p-value of .023. The b shows that each $1 million increase in government spending was associated with the creation of 0.086 new PVOs, or that it took $12 million in additional public funding to stimulate one new PVO. The finding that public funding is a cause of the increase of PVOs from 1968 to 2004 is consistent with the literature. The increase of funding to PVOs through USAID was in part because of their perceived cost containment and economic efficiency (Berrios, 2006; Ruttan, 1996; Rodinelli, 1989; Zimmerman & Hook, 1996). Robbins (1999), for example, concludes that the growth if PVOs in number and size is a result of public funding. Therefore, in terms of comparing the increase in the number of PVOs, the greater the amount of public funding, the greater the annual increase in PVOs. 118 Summary Three of the four hypotheses are accepted based on the information presented in the correlation tables. The partisanship of the President, enactment of Public Law 97- 113, and public funding are all shown to be significantly related to the increase in PVOs when no other factors are controlled. The partisanship of Congress was determined to be unrelated to the increase in PVOs. Multiple Regression The analysis in this section builds upon the research presented in Chapter 4 and earlier in this chapter and applies the methodology laid out in Chapter 3. Of the methods utilized to analyze the relationships of the independent and dependent variables, multiple regression is the most sophisticated technique utilized. Multiple regression analysis was performed in two stages to test the theory of the causes of the increase in PVOs during 1968- 2004. The first stage analyzes the effects of the two independent variables, partisanship of the President and Congress on the increase in PVOs to determine if their influences are significant when their effects are considered together. Next, the second stage is an analysis 119 of the impact of the significant independent variables and the two intervening variables on the growth PVOs. Stage One Multiple Regression Analysis The first stage of the multiple regression analysis is an examination of the impact of the partisanship of the political actors the President and Congress on the growth in the number of PVOs. Both of the predictor variables are lagged by two years, as discussed previously. The regression results are presented in Table 5.6. The top part displays the combined impact of the independent variables, while the bottom part reports statistics describing the individual impact of each predictor controlling for the effects of the other. The regression analysis revealed a moderate relationship as depicted by the Multiple R of .47. The R 2 of .22 indicates that 22 percent variance in the increase in PVOs is explained by the two partisanship factors. Further, the F statistic of 3.84 is statistically significant at the .034 level. The combined impact of the two independent variables, therefore, is clearly significant. However, adding the partisanship of Congress to Republican Executive produces almost no increase in explanatory power because the Multiple r of .47 is almost 120 exactly the same as the simple r of .46 between presidential partisanship and PVO growth. Table 5.6 Multiple Regression Explaining PVO Growth By Partisanship of the President and Congress Overall Equation Dependent Variable: PVO Growth Multiple R .47 Multiple R 2 .22 F 3.84 Significance .034 A 5.23 Separate Effects of Independent Variables Beta b t Significance PresPart, lag 2 yrs .45 14.37 2.640 .014 ConPart, lag 2 yrs -.09 -2.70 -.552 .585 The statistics regarding the specific effects attributable to each independent variable shown in the bottom half of Table 5.6 confirm that partisan control of the executive influences the formation of PVOs, while partisan control of Congress does not. The beta of .45 for presidential partisanship denotes a moderate-to?strong impact on PVO growth that is significant at the .014 level. In contrast, controlling the President?s party affiliation 121 makes almost no difference in the weak and statistically insignificant association between PVO growth and party control of Congress. These results, hence, strongly confirm the finding of the simple regressions in the last section that Hypothesis One holds true but Hypothesis Two does not. Stage 2 Multiple Regression In stage two of the regression analysis, the independent variable presidential partisanship and the two intervening variables Public Law 97-113 and public funding are analyzed in relation to the increase in PVOs. The findings of the pooled and individual effects of the predictor variables on the increase in PVOs are recorded in Table 5.7. All three independent variables have moderate direct correlations of between .41 and .45 with PVO growth (see Table 5.5 above). Thus, the Multiple R of .64 which is very highly significant at the .003 level in Table 5.7 clearly indicates the two intervening variables, PL97-113 and government spending, affected PVO growth well beyond the impact of presidential partisanship. Indeed, the Multiple R 2 of .42 shows that three independent variables in combination have double the explanatory power exercised by presidential partisanship by itself (21%). 122 Table 5.7 Multiple Regression Explaining PVO Growth by Presidential Partisanship, Public Law 97-113, and Government Spending Overall Equation Dependent variable: PVO Growth Multiple R .64 Multiple R 2 .42 F 6.138 Significance .003 a -5.879 Separate Effects of Independent Variables Beta b t Significance PresPar, Lag 2 yrs .44 14.020 2.869 .008 PubLaw, Lag 2 yrs .37 12.179 2.005 .055 PubFun .12 .026 .664 .512 The statistics on the individual effects of the three predictor variables show that even though they have fairly equal simple correlations with PVO growth, they differ radically in their relative influence on the dependence variable. Presidential partisanship and Public Law 97-113 have moderate positive effects that were predicted by Hypotheses 1 and 3, as indicated by their respective betas of .44 and .37. In contrast, once the effects of these factors are statistically controlled, government spending 123 has no independent influence, as indicated by its very low and insignificant beta of .12. This makes sense both substantively and theoretically. Changes in government spending were certainly associated with changes in the number of PVOs, as demonstrated by their simple correlation of .41. Yet, the multiple regression shows that this is largely spurious, reflecting the common impact of Republican Presidency and Public Law 97-113 upon both. Hence, a combination of Republican administrations and the passage of PL97-113 brought significantly increased public spending for PVOs which, in turn, stimulated the creation of new PVOs. Path Diagram It was originally surmised that the relationship between the dependent variable and the two independent and two intervening variables would all be positive relationships. However, the study?s findings have revealed that only 3 of the predictor variables have a positive relationship to PVO growth. The relationships are summarized in the path diagram in Figure 5.1. It was also thought that all of the variables would be significant predictors in determining PVO growth. However, the data analysis showed that congressional partisanship is not a significant factor in PVO growth, and thus it was dropped from the final model. Presidential partisanship and Public Law 97-113, in contrast, exerted moderate impacts on PVO growth in both the simple and multivariate analyses. Public spending on PVOs had a significant direct association with PVO growth, but this turned out to be a spurious correlation caused by the other two factors in the predictive equation. Independent Variables Dependent Variable Partisanship of the President +.44 Change in # of PVOs 124 +.37 +.12 Intervening Variables Public Law 97-113 Change in Public Funding of PVOs Figure 5.1. Variable Path Diagram 125 Analysis of Hypotheses Each hypothesis was analyzed with a multiple regression model to determine whether it should be accepted or rejected. The model that was accepted to describe the increase in PVOs included the partisanship of the Executive Branch, Public Law 97-113, and public funding. This reflects the statistical analysis in Table 5.7 and the path diagram represented by Figure 5.1. The influence of presidential and congressional partisanship on the growth in PVOs was assessed and resulted in the finding that congressional partisanship is not an indicator for PVO growth, as reflected in Table 5.6. Partisan of the Executive Branch H1: In terms of comparing the annual increase in PVOs, Republican control of the Executive Branch will have a positive impact on the annual increase of PVOs. Determination: As in the simple regression model, the partisanship of the Executive Branch was found to be the most important indicator in PVO growth. Therefore, this hypothesis is accepted. The results match the outsourcing theory, which asserts that Republican administrations are more predisposed to outsourcing. One of the neoliberal goals of 126 the 1980s was to reduce the state?s role in the delivery of public service and rely more on private organizations (Barnett, 2005). It is widely documented that the Reagan administration promoted privatization initiatives, convincing much of the aid community that the private rather than the public sector must lead development efforts (Meyer, 1992). Further, President George W. Bush signaled that he would continue with the privatization efforts of President Reagan (Henig, 1989). The literature and the statistical findings thus provide support for each other. Congress H2: In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs. Determination: Republican control of Congress is not positively related to the increase in PVOs; there is a negative correlation between the two variables. This variable was not included in the final regression model. This hypothesis is rejected Republican control of Congress as a predictor of the increase in PVOs was supported in the theory of outsourcing. The premise was that since Republicans 127 advocated outsourcing foreign assistance when they were in control, the number of organizations would increase. Further, the initial law to establish a funding minimum to PVOs and the subsequent law to increase funding were introduced in the Senate Foreign Relations Committee by Republican members of Congress. However, this hypothesis was not supported by any of the statistical analyses. Public Law 97-113 H3. In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase of PVOs. Determination: Public Law 97-113 was found to be positively correlated to the increase in PVOs. Within the regression model Public Law 97-113 had a p-value of .055. Thus, this variable?s effect is marginally significant at best. However, the moderate beta of .37 indicates that PL97-113 did stimulate the formation of more PVOs. Public Funding Increase H4. In terms of comparing the increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. Determination: The increase in public funding is not a significant predictor of the increases in PVOs by itself. 128 Rather, government spending simply transmits the effects of the other explanatory factors to PVO growth. Consequently, while H4 was supported by the simple regression, its influence becomes insignificant in the multivariate model. 129 CHAPTER 5: APPENDICES APPENDIX A: Original Data Years # PVO PVO Incr Original Pub Sup Orig Pub Sup Incr Pres Con PL97 -113 1967 73 $16,760,274.00 0 0 0 1968 77 4 $15,930,429.00 $(829,845.00) 0 0 0 1969 80 3 $20,650,390.00 $4,719,961.00 0 0 0 1970 82 2 $21,107,781.00 $457,391.00 0 0 0 1971 82 0 $25,121,155.00 $4,013,374.00 1 0 0 1972 89 7 $39,373,050.00 $14,251,895.00 1 0 0 1973 91 2 $91,708,839.00 $52,335,789.00 1 0 0 1974 94 3 $101,951,805.00 $10,242,966.00 1 0 0 1979 138 44 $126,998,149.00 $25,046,344.00 0 0 0 1980 153 12 $231,466,245.00 $104,468,096.00 0 0 0 1981 156 3 $306,482,047.00 $75,015,802.00 1 0 1 1983 167 11 $292,519,483.00 $(13,962,564.00) 1 0 1 1984 161 -6 $255,569,622.00 $(36,949,861.00) 1 0 1 1985 170 9 $331,316,428.00 $75,746,806.00 1 0 1 1986 178 8 $481,962,070.00 $150,645,642.00 1 0 1 1987 198 20 $497,068,182.00 $15,106,112.00 1 0 1 1988 205 7 $563,043,058.00 $65,974,876.00 1 0 1 1989 243 38 $656,526,757.00 $93,483,699.00 1 0 1 1990 277 34 $743,065,331.00 $86,538,574.00 1 0 1 1991 301 24 $832,035,825.00 $88,970,494.00 1 0 1 1992 347 46 $1,041,366,774.00 $209,330,949.00 1 0 1 1993 390 43 $1,249,062,388.00 $207,695,614.00 0 0 1 1994 419 29 $1,403,509,231.00 $154,446,843.00 0 0 1 1995 434 15 $1,483,038,634.00 $79,529,403.00 0 1 1 1996 417 -17 $1,545,016,527.00 $61,977,893.00 0 1 1 1997 424 7 $1,554,144,496.00 $9,127,969.00 0 1 1 1998 425 1 $1,657,605,612.00 $103,461,116.00 0 1 1 1999 437 12 $1,819,112,682.00 $161,507,070.00 0 1 1 2000 439 2 $2,084,249,402.00 $265,136,720.00 0 1 1 2001 436 -3 $2,296,003,596.00 $211,754,194.00 1 1 1 2002 444 8 $2,574,744,414.00 $278,740,818.00 1 1 1 2003 489 45 $2,810,364,214.00 $235,619,800.00 1 1 1 2004 509 20 $3,173,592,418.00 $363,228,204.00 1 1 1 130 APPENDIX B: GDP Deflator and Multiplier YEAR GDP Deflator Index- Multiplier 1967 23.893 3.414807684 1968 24.913 3.27499699 1969 26.149 3.120195801 1970 27.534 2.963245442 1971 28.911 2.822109232 1972 30.166 2.704700656 1973 31.849 2.56177588 1974 34.725 2.349604032 1979 49.548 1.646686042 1980 54.043 1.509723739 1981 59.119 1.380097769 1983 65.207 1.251246032 1984 67.655 1.205971473 1985 69.713 1.170369945 1986 71.25 1.145122807 1987 73.196 1.114678398 1988 75.694 1.077892567 1989 78.556 1.038622129 1990 81.59 1 1991 84.444 0.966202454 1992 86.385 0.944492678 1993 88.381 0.923162218 1994 90.259 0.903954176 1995 92.106 0.885827199 1996 93.852 0.869347483 1997 95.414 0.855115601 1998 96.472 0.845737623 1999 97.868 0.833673928 2000 100 0.8159 2001 102.399 0.796785125 2002 104.187 0.783111137 2003 106.305 0.767508584 2004 109.099 0.747852868 Table 1.1.9. Implicit Price Deflators for Gross Domestic Product [Index numbers, 2000=100] Bureau of Economic Analysis Downloaded on 2/24/2006 At 2:52:16 PM Last Revised January 27, 2006 131 APPENDIX C: Adjusted Public Funding Data Years Public Support Public Support Increase 1967 $57,233,112.45 1968 $52,172,107.02 $(5,061,005.43) 1969 $64,433,260.17 $12,261,153.15 1970 $62,547,535.84 $(1,885,724.33) 1971 $70,894,643.44 $8,347,107.60 1972 $106,492,314.18 $35,597,670.74 1973 $234,937,491.73 $128,445,177.55 1974 $239,546,372.06 $4,608,880.34 1979 $209,126,079.30 $(30,420,292.77) 1980 $349,450,084.74 $140,324,005.44 1981 $422,975,189.27 $73,525,104.54 1983 $366,013,842.35 $(56,961,346.93) 1984 $308,209,673.48 $(57,804,168.87) 1985 $387,762,789.73 $79,553,116.26 1986 $551,905,758.47 $164,142,968.74 1987 $554,071,164.67 $2,165,406.20 1988 $606,899,927.37 $52,828,762.70 1989 $681,883,218.39 $74,983,291.02 1990 $743,065,331.00 $61,182,112.61 1991 $803,915,055.68 $60,849,724.68 1992 $983,563,293.29 $179,648,237.61 1993 $1,153,087,204.68 $169,523,911.39 1994 $1,268,708,030.86 $115,620,826.18 1995 $1,313,715,959.31 $45,007,928.45 1996 $1,343,156,229.36 $29,440,270.05 1997 $1,328,973,205.49 $(14,183,023.87) 1998 $1,401,899,430.75 $72,926,225.26 1999 $1,516,546,815.35 $114,647,384.60 2000 $1,700,539,087.09 $183,992,271.74 2001 $1,829,421,511.91 $128,882,424.81 2002 $2,016,311,024.78 $186,889,512.87 2003 $2,156,978,657.83 $140,667,633.05 2004 $2,373,380,190.33 $216,401,532.50 132 APPENDIX D: Data Used in Regression Analyses Year PVO Inc PubFun Inc in Mil Pres Lag 2 Congress Lag 2 PL97-113 Lag 2 1968 4 -5.06 0 0 0.00 1969 3 12.26 0 0 0.00 1970 2 -1.89 0 0 0.00 1971 0 8.35 1 0 0.00 1972 7 35.6 1 0 0.00 1973 2 128.45 1 0 0.00 1974 3 4.61 1 0 0.00 1980 . 140.32 1 0 0.00 1981 3 73.53 1 0 0.00 1983 11 -56.96 0 0 0.00 1984 -6 -57.8 1 1 0.00 1985 9 79.55 1 1 1.00 1986 8 164.14 1 1 1.00 1987 20 2.17 1 1 1.00 1988 7 52.83 1 1 1.00 1989 38 74.98 1 0 1.00 1990 34 61.18 1 0 1.00 1991 24 60.85 1 0 1.00 1992 46 179.65 1 0 1.00 1993 43 169.52 1 0 1.00 1994 29 115.62 1 0 1.00 1995 15 45.01 0 0 1.00 1996 -17 29.44 0 0 1.00 1997 7 -14.18 0 1 1.00 1998 1 72.93 0 1 1.00 1999 12 114.65 0 1 1.00 2000 2 183.99 0 1 1.00 2001 -3 128.88 0 1 1.00 2002 8 186.89 0 1 1.00 2003 45 140.67 1 1 1.00 2004 20 216.4 1 1 1.00 133 CHAPTER 6 CONCLUSION AND RECOMMENDATIONS Introduction As discussed in the preceding chapters, the purpose of this study was to determine the causes of the increase in Private Voluntary Organizations from 1968-2004. Further, this research builds on the largely anecdotal literature about PVOs and seeks to add empirical data to specify the nature of the relationships included in the model. The implications of this study of Private Voluntary Organizations for public policy and administration are not novel but rather verify that reality often proves theory. This chapter concludes this research by summarizing the findings for each hypothesis, discussing the implications of this study for public policy and administration in the PVO environment, and highlighting the limitations. Significance of Study This study provides a partial explanation of the increase in Private Voluntary Organizations during 1968- 2004. This is critical, as humanitarian and development 134 efforts have also increased in number and exposure. The world is now acutely aware when disastrous events occur, from genocide to hurricanes or tsunamis, and expect the assistance of Private Voluntary Organizations. ?There has been an explosion of nongovernmental organizations dedicated to some aspect of humanitarian action? (Barnett, 2005, p. 723). According to Aall (1996), ?the rapid rise in the number of these organizations has resulted in more institutions that are able to mobilize resources for humanitarian crises in hitherto inaccessible and neglected parts of the world? (p. 6). These organizations have grown in number and size with little resistance because of their purported efficiency and cost effectiveness in humanitarian and development assistance. According to Brodhead (1987), the presumed effectiveness of private voluntary organizations in program delivery and meeting the needs of the poor is what has attracted interest and support for them. Research Question This research explored the following question: What specific factors caused the increase in Private Voluntary Organizations from 1968-2004? 135 Instrumentation The data sets collected to answer the research question for this study are utilized as cross-sectional though they are in fact longitudinal in nature. The data specific to Private Voluntary Organizations, such as the number of them and their public funding, were taken from the annual United State Agency for International Development Voluntary Agency reports (?VolAg?) from the years 1967-2004. The information on the partisanship of the Executive Branch, the partisanship of Congress, and the passage of Public Law 97-113 was taken from official government websites. Each year recorded by USAID is taken as a case, for a total of 31 cases. USAID did not compile reports on Private Voluntary Organizations from 1975-1978 and the report for 1982 contains combined data for 1981 and 1982. The case for 1979 was deleted from the population because the data was converted to differences from year to year and there was no data on the number of PVOs for 1978. The case for 1980 was deleted due to data inconsistencies. Summary of Findings Four hypotheses were tested for this study over time and through statistical analysis to answer the research 136 question. What factors have caused the increase in PVOs from 1968-2004? The following hypotheses were developed: H1. In terms of comparing the annual increase in PVOs, Republican control of the Executive Branch will have a positive impact on the annual increase in PVOs. Both the bivariate and multivariate analyses supported this hypothesis. H2. In terms of comparing the annual increase in PVOs, Republican control of one or both Houses of Congress will have a positive impact upon the annual increase in PVOs Both the bivariate and the multivariate analyses showed that this hypothesis should be rejected. H3. In terms of comparing the annual increase in PVOs, the enactment of Public Law 97-113 will have a positive impact upon the annual increase in PVOs. Both the bivariate and multivariate analyses supported this hypothesis. H4. In terms of comparing the increase in PVOs, the greater the growth in the amount of public funding, the greater the annual increase in PVOs will be. 137 This hypothesis was supported by bivariate analysis. However, the relationship became insignificant in the multivariate analysis after controlling for the effects of the other explanatory variables statistically. Implications for Policy and Administration This research developed a model to explain the increase in PVOs from 1968 to 2004. It stepped beyond the literature and added empirical data to explain the increase PVOs. It is commonly agreed by researchers that the number of PVOs has increased dramatically over the 20 to 30 years (Smith, 1984; Lewis & Wallace, 2000; United States General Accounting Office, 2002). However, little empirical data are available on the causes of PVO growth. There is a plethora of literature that attributes the rise of PVOs to the advancement of neoliberal ideas (Berrios, 2005; Henig, 1989; Meyers, 1992). Others attribute the increase in the number of PVOs to their better efficiency than government at providing humanitarian and development assistance to the poor (Clark 1991; Hoksenbergen, 1999; Marcussen, 1996; Meyer, 1992; OECD, 1988; Summer, 1977). The findings in this study reveal that the increase in PVOs has largely been the result of the partisanship of the President. This information is significant to foreign 138 policy and the principal agent theory and begs the question to what extent are PVOs an extension of the United States Government if the PVO sector is being largely influenced by the Executive Branch? This is critical in places like Iraq, Afghanistan, and Kosovo, where PVOs need to be viewed as autonomous organizations assisting people in humanitarian and development efforts and not merely as instruments of U.S. foreign policy. This research is an indicator to the PVO sector of the causes of the expansion of the sector with empirical evidence. PVOs, though claiming autonomy and independence, must understand that they are the result of political forces. The extent to which this makes them political actors is an area for more research and intense debate. Limitations of Study All research has some limitations based on the availability of resources, information, and/or project design and this research is no exception. However, steps were taken to mitigate any limitations that would jeopardize the integrity of the research. Though founded in a solid research design and method, there are two significant limitations to this study. 139 The population size of this project is a major limitation. The population for this study was only 30 cases. In regression analysis, the larger the population size the better and more reliable the results. However, USAID did not start recording the number of registered PVOs until 1967. USAID also failed to compile annual reports for PVOs for the years of 1975-1978. Therefore, the population size for this research was limited by the available information. The longitudinal data are also subject to limitations. It was necessary to convert the data for public funding and the number of PVOs into increases and decreases to compensate for the natural trends upward during 1968-2004 in these variables. Further, the data were not analyzed as a time series study but as if collected and used in a cross-sectional study. Recommendations for Further Research Research leads to more research, as some questions are answered but new questions are raised. This study was largely conducted at the macro level and used data compiled over the aggregate to perform analysis. However, more research needs to be done at the micro level, analyzing specific PVOs such as the faith based organizations versus 140 the secular organizations and the special interest organizations such as those that focus on women?s issues. This study has exposed a number of areas for additional research. The major areas that require more analysis are funding, principle agent theory, and PVO efficiency. First, the funding of PVOs is an area of research that needs to be examined in greater detail. A study should be done to show the public and private funding streams to PVOs. It is important to know if the money that funds PVOs is being evenly distributed or whether a few organizations are getting the majority of the funds. GAO reports (United States General Accounting Office, 1995) indicate that a select few organizations were getting the majority of the funding, such as World Vision, Care, and Catholic Relief Services. Second, given that Presidential partisanship is a driving force in the increase in PVOs, research needs to be done to adequately define the relationship between the PVOs and the government. Is the relationship reflective of the principle agent theory or are these organizations able to maintain their independence and autonomy in spite being a part of the executive?s foreign policy strategy? To what extent do PVOs performing humanitarian and development 141 function in places like Iraq view themselves as instruments of foreign policy? Lastly, the literature continues to emphasize the efficiency of Private Voluntary Organizations over the government to provide humanitarian and development assistance to the world?s poor. However, this conclusion is not validated by research to the same degree. A substantial amount of Private Voluntary Organization program evaluations need to be conducted to determine empirically if PVOs are actually more cost effective and efficient service providers than traditional government agencies. Future Policy Changes In 2005 Congress rescinded the law which required PVOs to raise at least 20% of their funding from the private sources before they could receive public funds. This is a critical change and must be carefully watched in the immediate years to come. This change makes PVOs more susceptible to control by public funding and it may be determined in future research that public funding is a significant factor in the increase of PVOs. Again the same question is asked in this case as with presidential partisanship: if an organization receives 90% of its 142 funding from the government, to what degree is this organization independent and autonomous? With the onset of this change, it is necessary to warn Private Voluntary Organizations of the potential for being co-opted by the government to carry out foreign policy objectives. Conclusion This research provided information that is useful to Private Voluntary Organizations, USAID, the Senate Foreign Relations Committee, and other PVO researchers. The empirical evidence presented here suggests that the increase in PVOs is a partisan issue in the Executive Branch. 143 REFERENCES Aall, P. (1996). NGOs and conflict management. Washington, DC: United States Institute for Peace. Abdullah, A. (2004, August 26). NGOs: Another wing of US foreign policy. Yemen Times, 13, 767. Al-Saqqah, W. (2004, August 2). The new revolution. Yemen Times, 13, 760 Barnett, M. (2005). Humanitarianism transformed. Perspectives on Politics, 3(4), 723-739. Becker, E. (2004, December 22). U.S. cutting food aid that is aimed at self-sufficiency. New York Times, A3. Berk, K. & Carey, P. (1995). Data analysis with Microsoft Excel. Pacific Grove, CA: Brook/Cole. Berrios, R. (2000). Contracting for development: The role of for-profit contractors in U.S. foreign development assistance. Westport, CT: Praeger. Brinkerhoff, J. (2002). Partnership for international development. London: Lynne Rienner. Brodhead, T. (1987). NGOs: In one year, out the other? World Development, 15, 1-6. 144 Brownstein, V. (1994). Budget deficit reality check. Fortune, 130(13). 32. Buchanan, J.M., & Tullock, G. (1965). The calculus of consent: Logical foundations of constitutional democracy. Ann Arbor, MI: University of Michigan Press. Bureau of Economic Analysis. (n.d.) Implicit price deflator for gross domestic product. Retrieved February, 24, 2006, from www.bea.doc.gov/bea/dn/home/gdp Bureau of Economic Analysis. (n.d). Gross domestic product 1968-2004, Retrieved March 20, 2006, from www.bea.doc.gov/bea/dn/home/gdp Bureau of Public Debt. (n.d.). Federal budget deficits and surplus 1968-2004. Retrieved March 20, 2006, from http://www.publicdebt.treas.gov/opd/opdhisto4.htm Chabbott, C. (1990). Developing INGOs. In J. Boli & G. Thomas (Eds.), Constructing world culture: International nongovernmental organizations since 1875. (pp 223-248). Stanford, CA: Stanford University Press. Clark, C. (2007). Political science research methods: Exploring ?America at a crossroad?. New York: Longman. 145 Clark, J. (1991). Democratizing development: The role of voluntary organizations. West Hartford, CT: Kumarian. Colclough, C., & Manor, J. (1991). States or markets? Neo- Liberalism and the development policy debate. Oxford, UK: Clarendon Press. Conteh-Morgan, E. (1990). American foreign aid and global power projection: The geopolitics of resource allocation. Aldershot, UK: Dartmouth. Cooley, A., & Ron, J. (2002). The NGO scramble: Organizational insecurity and the political economy of transnational action. International Security, 27(1), 5-39. Covey, J. (1995). Accountability and effectiveness in NGO policy alliances. In M. Edwards & D. Hulme (Eds.), Beyond the magic bullet: NGO performance and accountability in the post Cold-War world. London: Earthscan. Cronk, B. (2006). How to use SPSS. CA: Pyrczak. Dichter, T. (2003). Despite good intentions: Why development assistance to the Third World has failed. Boston, MA: University of Massachusetts Press. Dielman, T.E. (1989). Pooled cross-sectional and time series data analysis, New York: Marcel Dekker. 146 Drury, C.A., Olson, R.S., & Van Bell, D.A. (2005). The politics of humanitarian aid: U.S. foreign disaster assistance, 1964-1995. Journal of Politics, 67(2), 454-474. Dudley, L., & Montmarquette, C. (1976). A model of the supply of bilateral foreign aid. The American Economic Review, 66(1),132-142. Duffield, M. (1991). Where famine is functional: Actual adjustment and the politics of relief in Sudan. (Middle East Rep. No. 172). Middle East Research and Information Project. Ebrahim, A. (2003). NGOs and organizational change: Discourse, reporting, and learning. Cambridge, UK: Cambridge University Press. Edwards, M. & Hulme, D. (1996). Too close for comfort? The impact of official aid on nongovernmental organizations. World Development, 24(6), 961-973. Ellis, G. (1984). Making PVOs count more: A proposal. In R.F. Gorman (Ed.) Private voluntary organizations as agents of development (pp. 201-214). Boulder, CO: Westview. Farrington, J., Bebbington, A., Wellard, K., & Lewis, D. (1993). Reluctant partners? NGOs, the state and 147 sustainable agriculture development. London: Routledge. Farrington, J. & Lewis, D. (Eds). (1993). With S. Satish, & A. Miclat-Teves, NGOs and the state in Asia: Rethinking roles in sustainable agriculture development. London: Routledge. Feigenbaum, H. & Henig, J. (1994). The political underpinnings of privatization: A typology. World Politics, 46(2), 185-208. Fisher, J. (1993). The road from Rio: Sustainable development and the nongovernmental movement of the Third World. Westport, CT: Praeger. Fisher, W. (1997). Doing good? The politics and antipolitics of NGO practices. Annual Review of Anthropology, 26, 439-64. Fitzduff, M. & Church, C. (Eds.)(2004). NGOs at the table: Strategies for influencing policies in areas of conflict. New York: Rowman & Littlefield. Garvey, G. (1997). Public administration: The profession and the practice. New York: St Martins Press. Gorman, R.F. (Ed.) (1984). Private voluntary organizations as agents of development. Boulder, CO: Westview. 148 Gradstein, M. (1993). Rent seeking and the provision of public goods. The Economic Journal, 103, 1236-1243. Healey, J.M., & Coverdale, A.G. (1981, January). Foreign policy and British bilateral aid: A comment on McKinlay and Little. British Journal of Political Science, 11(1), 123-127. Henig, J.R. (1989-1990, winter) Privatization in the United States: Theory and practice. Political Science Quarterly, 104(4), 649-670. Hira, A., & Parfitt, T. (2004). Development projects for a new millennium. London: Praeger. Hoksbergen, R. (1999). Give them a fish, teach them to fish, or organize a fishing club? NGOs, civil society and economic development. Faith & Economics, 34, 11- 18. Holloway, R. (1998). NGOs: Losing the high ground? UN Chronicle, 35(1), 93-95. Hook, S.W. (1995). National interest and foreign aid. Boulder, CO: Lynne Rienner. International Committee of the Red Cross. (1995, September 15). Code of conduct for the international Red Cross movement and NGOs in disaster relief. Retrieved 149 November 5, 2005, from www.icrc.org/web/eng/siteeeng0/ nsf/htmlall/57JMNB Kamat, S. (2003). NGOs and the new democracy: The false saviors of international development. Harvard International Review, 25, 65-70. Kasich, J. (2004, February 1). Deficits, fruit flies and the Beltway. New York Times, p. 11. King, S. (1994). Competitive tendering and contracting out: An introduction. Australian Economic Review, 107, 75- 78. Korten, D. (1990). Getting to the 21 st century: Voluntary action and the global agenda. Bloomfield, CT: Kumarian. Kovach, H., Neligan, C., & Burall, S. (2003, January, 1). Power without accountability? Global accountability report. Retrieved November, 5, 2005, from the One World Trust Web site: http://www.oneworldtrust.org/documents/GAP2003.pdf Lebovic, J. (1988) National interests and US foreign aid: The Carter and Reagan years. Journal of Peace, 25(2), 115-135. Leonard, M. (2002). Diplomacy by other means. Foreign Policy, 132, 48-53. 150 Lewis, D. & Wallace, T. (Eds.) (2000). New roles and relevance: Developing NGOs and the challenge of change. Bloomfield, CT: Kumarian. Library of Congress. (n.d.). Public Law 99-83.Retrieved February 15, 2006, from http://icreport.loc.gov. Library of Congress. (n.d.). Public Law 97-113. Retrieved February 15, 2006, from http://icreport.loc.gov. Lipsky, M. & Smith, S. (1993). Nonprofits for hire. London: Harvard University Press. Manji, F., & O?Coill, C. (2002). The missionary position: NGOs and development in Africa. International Affairs, 78(3), 567-584. Marcussen, H.S. (1996). NGOs, the state and civil society. Review of African Political Economy, 69, 506-423. Matthews, J. (1997). Power shift. Foreign Affairs, 76(1), 50-67. McHugh, G., & Gostelow, L. (2004). Provincial reconstruction teams and humanitarian-military relations in Afghanistan. London: Save the Children. McKinlay, R.D., & Little, R. (1977). A foreign policy model of US bilateral aid allocation. World Politics, 30(1), 58-86. 151 Meyer, C. (1992). A step back as donors shift institution building from the public to the ?private? sector. World Development, 20(8), 1115-1126. Meyer, C. (1999). The economics and politics of NGOs in Latin America. London: Praeger. Miller, V. (1994). NGOs and grassroots policy influence: What is success? Boston, MA: Institute for Development Research. Millikan, M.F. (1957, August). Future of American foreign aid. The Scientific Monthly, 85(2), 86-88. Minear, L. (1987). The other missions of NGOs: Education advocacy. World Development, 15, 201-211. Natsios, A. (2001) NGOs and the UN in complex humanitarian emergencies. In P. Diehl (Ed.), The politics of global governance (pp. 388-405). London: Lynne Rienner. Organization for Economic Co-operation and Development. (1988). Voluntary aid for development: The role of non-governmental organizations. Paris: OECD. Porter, D. (1990). Economic foreign aid: A case study of the United States Agency for International Development. New York: Garland. Radelet, S. (2003). Challenging foreign aid: A policymaker?s guide to the Millennium Challenge 152 Account. Washington, D.C.: Center for Global Development. Robbins, R. (1999). Global problems and the culture of capitalism. MA: Allyn & Bacon. Roberts, H. (1984). The domestic environment of AID- registered PVOs: Characteristics and impact. In R.F. Gorman (Ed.), Private voluntary organizations as agents of development (pp. 99-114). Boulder, CO: Westview. Rodinelli, D. (1989). Foreign aid policy: Constraints on development assistance. Policy Studies Journal, 18, 67-85. Ruttan, V. (1996). United States development assistance policy. Baltimore, MD.: Johns Hopkins University Press. Smillie, I., & Helmich, H. (1993). Nongovernmental organizations and government: Stakeholders for development. Paris: OECD. Smith, B. (1984). U.S. and Canadian PVOs as transnational development institutions. In R.F. Gorman (Ed.), Private voluntary organizations as agents of development (pp. 99-114). Boulder, CO: Westview. 153 Smith, B. (2000). More than altruism: The politics of private foreign aid. Princeton, NJ.: Princeton University Press. Sollis, P. (1992). Multilateral agencies, NGOs and policy reform. Development in Practice, 2(3), 163-178. Summer, J. (1977). Beyond charity: US voluntary aid for a changing Third World. Washington, D.C.: Overseas Development Council. Thompson, E. (2004). The role of NGOs in challenging the conservative agenda: Empirical studies. Social Alternatives, 23(1), 43-49. Truman, D. (1951). The governmental process: Political interests and public opinion. New York: Knoft. United Nations. (1991, December, 19). United Nations Resolution 46/182. Retrieved July 15, 2005, from http:// www.un.org/documents/ga/res/46/a46r182.htm United States Agency for International Development. (1982, September). AID partnership in international development with private and voluntary organizations. Washington D.C.: USAID. United States Agency for International Development. (n.d.). About USAID. Retrieved May, 19, 2005 from http://www.usaid.gov/about_usaid. 154 United States Agency for International Development. (n.d.). Voluntary agencies reports 1967-2006. Retrieved July 30, 2005 from http://dec.usaid.gov. United States General Accounting Office. (1992a). Aid management: Strategic management can help AID face current and future challenges. Report No. NSIAD-92- 100. Washington, DC: Author United States General Accounting Office. (1992b). Foreign assistance: AID?s private sector assistance program at a crossroads. Report No. NSIAD-93-55. Washington, DC: Author. United States General Accounting Office. (1995). Foreign assistance: Private voluntary organizations? contributions and limitations. Report No. GAO/NSIAD- 96-34. Washington, D.C.: Author. United States General Accounting Office. (2002). Foreign assistance: USAID relies heavily on nongovernmental organizations, but better data needed to evaluate approaches. Report No. GAO-02-471. Washington, DC: Author. Vivian, J. (1994). NGOs and sustainable development in Zimbabwe: No magic bullet. Development & Change, 25, 181-209. 155 Walker, J. (1983). The origins and maintenance of interest groups in America. The American Political Science Review, 77, 390-405. Weiss, T. & Collins, C. (2000). Humanitarian challenges & Intervention. Boulder, CO: Westview. Wikipedia. (n.d.). Budget deficit. Retrieved May 10, 2006 from http://En.wikipedia.org/wiki/budget-deficit. Wikipedia. (n.d.). Autocorrelations. Retrieved May 20, 2006, from http://En.wikipedia.org/wiki/ Autocorrelation#Applications Wikipedia. (n.d.). Multicollinearity. Retrieved May 20, 2006 from http://en.wikipedia.org/wiki/ Multicollinearity Wilson, W. (1887). The study of administration. Political Science Quarterly, 2(2), 197-222. Wolf, C. (1979). A theory of non-market failures. Journal of Law and Economics, 22(1), 107-39. Zimmerman, R.F. (1993). Dollars, diplomacy & dependency: Dilemmas of U.S. economic aid. Boulder, CO: Lynne Rienner. Zimmerman, R., & Hook, S. (1996). The assault on US foreign aid. In S. Hook. (Ed.), Foreign aid toward the new millennium. (57-73). Boulder, CO: Lynne Rienner. 156 Zivetz L. (1991). Doing good: The Australian NGO community. North Sidney, NSW: Allen & Unwin.