THE EFFECTS OF NEGATIVE AND POSITIVE FREEDOM ON ECONOMIC GROWTH, HUMAN DEVELOPMENT AND POVERTY by Jacqueline Lytle B.A., University o f Northern British Columbia 2004 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN DEVELOPMENT ECONOMICS UNIVERSITY OF NORTHERN BRITISH COLUMBIA April 2013 © Jacqueline Lytle, 2013 1+1 Library and Archives Canada Bibliotheque et Archives Canada Published Heritage Branch Direction du Patrimoine de I'edition 395 Wellington Street Ottawa ON K1A0N4 Canada 395, rue Wellington Ottawa ON K1A 0N4 Canada Your file Votre reference ISBN: 978-0-494-94144-7 Our file Notre reference ISBN: 978-0-494-94144-7 NOTICE: AVIS: The author has granted a non­ exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distrbute and sell theses worldwide, for commercial or non­ commercial purposes, in microform, paper, electronic and/or any other formats. L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'lnternet, preter, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conform em ent a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these. W hile these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada Abstract Freedom is argued by some to have instrumental as well as intrinsic value. Economic and political freedoms are argued to have effects on various economic outcomes in previous research. The definition of freedom is unclear and implied by the components of the economic or political freedom index used in the study. Although the concept of freedom continues to be debated and contested, this thesis employs Isaiah Berlin’s dichotomy of positive and negative freedom. Berlin defines negative freedom as freedom from government and coercion and positive freedom as the ability to participate in the governance of oneself. This study indicates that the effects of negative and positive freedom on economic growth are statistically non-significant. The human development model is sensitive to the inclusion of property rights and thus not stable. Both positive and negative freedoms have a positive, statistically significant and robust relationship with reductions in absolute poverty. Table of Contents Abstract ii Table of Contents iii List of Tables iv Chapter 1: Introduction 1 Chapter 2: Negative and Positive Freedom 9 Chapter 3: Econometric Research - Freedom and Economic Development 17 Chapter 4: Data & Methodology 34 Chapter 5: Results & Discussion 47 Chapter 6: Conclusion 62 Bibliography 68 Appendix 1 Summary o f Indices 77 Appendix 2 Summary o f All Variables & Sources 82 Appendix 3 Economic Growth Models - Countries Included in the Sample 83 Appendix 4 Human Development Models —Countries Included in the Sample 84 Appendix 5 Poverty Models - Countries Included in the Sample 85 Appendix 6 Fixed Effects for Economic Growth Models 86 Appendix 7 Fixed Effects for Economic Growth Models Including Property Rights 91 iii List of Tables Table 1 Components of the Index of Economic Freedom 39 Table 2 Components of the Freedom in the World Index 40 Table 3 Description of The Cingranelli-Richards (CIRI) Human Rights Dataset 41 Table 4 Independent Variables in the Economic Growth Model 44 Table 5 Independent Variables in the Human Development Index Model 45 Table 6 Independent Variables in the Poverty Head Count Model 46 Table 7 Economic Growth Models with Country & Period Fixed Effects 48 Table 8 Economic Growth Models Including Property Rights - Fixed Effects 50 Table 9 Human Development Model - OLS 53 Table 10 Human Development Model Including Property Rights - OLS 55 Table 11 Poverty Models —OLS 58 Table 12 Poverty Models Including Property Rights —OLS 60 Table 13 List of Freedom Indices and their Sources 81 iv Chapter 1: Introduction Using Isaiah Berlin’s dichotomy to clearly define freedom, this study examines the relationship between negative and positive freedom and economic growth, human development and poverty. The theories of Milton Friedman and Amartya Sen provide the mechanisms through which freedom may affect economic outcomes. The relationships between negative and positive freedom and economic growth, human development and poverty as suggested by these two economists are tested using regression analysis. Recent political rhetoric in the United States of America indicates that the debate regarding positive and negative freedom as articulated by Berlin (1992) in his famous essay endures. This is evidenced by the following quotes from Republican and Democratic presidential candidates: Opportunity expands ... when taxes are lowered... when constitutional freedoms are preserved.. .Liberals would replace opportunity with dependency on government largess. They grow government and raise taxes .. .Dependency is death to initiative, risk-taking and opportunity. (Mitt Romney, Republican Presidential Candidate, 2012) We have to love our freedom not just for the material benefits it provides, not just for the autonomy it guarantees us, but for the goodness it makes possible (John McCain, Republican Presidential Candidate, 2008). America cannot have a strong, growing economy without a strong, growing middle class, and the chance for everybody, no matter how humble their beginnings, to join that middle class ... a middle class built on the idea that if you work hard, if you live up to your responsibilities, then you can get ahead; that you can enjoy some basic guarantees in life. A good job that pays a good wage. Health care that will be there when you get sick. ... A secure retirement even if you’re not rich. ... An education that will give your children a better life than we had. ... These are simple ideas. These are American ideas. These are union ideas. That’s what we’re fighting for. (Barack Obama, Democratic Presidential Candidate 2008 and 2012) 1 Democracy in a nation of 300 million can be noisy and messy and complicated.. .These arguments we have are a mark of our liberty. We can never forget that as we speak people in distant nations are risking their lives right now just for a chance to argue about the issues that matter, the chance to cast their ballots like we did today (Barack Obama, Democratic Presidential Candidate 2008 and 2012). The Republican candidates advocate less government and lower taxes which, they argue yield increased opportunities. The Democratic candidate argues that a growing economy relies on basic guarantees for citizens, and a strong democracy. The proponents of small government call for more freedom in order to ensure prosperity and development, as proponents of self-government and empowerment of the citizenry also call for more freedom. Are the two opposing groups referring to the same freedom? Berlin clearly outlines in his essay “Two Concepts of Liberty” there are two very different kinds of freedom. Berlin defines the two freedoms as negative freedom - freedom from government interference - and positive freedom - freedom to govern oneself, to be self-directed, to determine goals and realize them (Berlin 1992, 303). Politicians calling for increases in freedom anticipate different outcomes from that increase, depending on to which freedom they are referring and which economic theory they rely upon. Republicans may follow Friedman’s theories, which state that increases in negative freedom lead to increases in economic growth and increases in positive freedom may lead to the opposite. Democrats may be more comfortable with Sen’s theories that indicate that increases in positive freedom lead to increases in human development. The economic theory subscribed to by the policy makers -Friedman’s or Sen’s —and thus the type of freedom deemed beneficial - negative or positive - have very important policy implications. If Friedman is deemed correct then increasing negative freedom by reducing taxes is more important than increasing positive freedom. In fact, increased positive freedom 2 may be seen as having a detrimental effect on economic growth. If Sen’s ideas are viewed as correct, then increasing positive freedom by empowering people to determine and realize their own goals and objectives would be viewed as more beneficial to improve human development. The tradeoff between the two freedoms is essential to many policy decisions. Increased positive freedom often comes at the expense of decreased negative freedom and vice versa. For example, an increase in positive freedom may lead to increased taxes that decrease negative freedom. An increase in negative freedom characterized by reduced taxes may be achieved only through the reduction of services that some see as necessary for all members of society to enjoy positive freedom, for example public education. The concept of freedom is not universally acknowledged to involve the notions of positive and negative freedom. It is a highly debated area of philosophy (Carter 2012). This econometric study has chosen to utilize the definitions of negative and positive freedom to define and categorize types of freedom. Recent research has examined the theoretical implications of viewing freedom as negative or positive as described by Berlin (Prendergast 2004; Stroup 2007) and clarifies the mechanisms through which freedom is hypothesized to affect economic outcomes. One empirical study investigates the effects of negative or positive freedoms on economic outcomes (Kaun 2002). Kaun’s empirical investigation into the effects of negative freedom on some well-being indicators in the United States of America (Kaun 2002) heavily informs my empirical investigation. The contributions of Prendergast, Stroup and Kaun’s work to this thesis will be more thoroughly examined in Chapter 2. However, most previous econometric research has relied on indices of economic and political freedom. The indices are often employed to measure freedom without questioning 3 the components of the index or the objectives of the publishers of the indices. The conclusions are then often framed by the authors or others as if the indices do represent freedom for individuals. According to Hanke and Walters, the economic and political freedom indices are aimed at policy makers and scholars, consisting mainly of indicators of government policy. These indices are designed to help determine what institutions are necessary for prosperity and what policies may be beneficial to economic growth (Hanke and Walters 1997,133-4). This study attempts to utilize a more precise definition of freedom by employing the dichotomous notions of negative and positive freedom of Berlin. A thorough examination of previous research investigating the effects of economic and political freedom assists in model specification and selection of other variables to include in the model. In addition, the results of previous research are compared to the results of this study. Analysis of previous research regarding the relationship of economic and political freedom to economic growth has returned two different conclusions: (1) there is a positive relationship between economic freedom and economic growth (See Doucouliagos & Ulubasoglu, 2006 for a meta-analysis) and (2) there is a positive relationship between some components of economic freedom, a negative relationship with some and no relationship with others (See de Haan, Lundstrom, & Sturm, 2006 for a critical survey). Empirical research investigating the joint effects of economic and political freedom on economic growth has also been inconclusive. Some conclude that political freedom and economic freedom enhance economic growth (Goldsmith 1995) and others conclude that given economic freedom there is no relationship between economic growth and political freedom (Wu and Davis 1999). Empirical research on the effects of freedom on poverty and human development is not as extensive but still contains contradictory conclusions. One study 4 concludes that some economic freedoms and civil liberties reduce poverty, but political liberties do not reduce poverty (Hasan, Quibria and Kim 2003). Another one finds that economic freedoms have a larger effect on quality of life than do political rights, but political rights are still positively related to quality of life (Stroup 2007). A consistent thread through the previous research is the importance of property rights in positive relationships between freedom and economic growth, human development and poverty. Property rights indicators are found in both economic freedom indices and political freedom indices. Previous research investigating the effects of components of either economic freedom, political freedom or both often find the property rights component is very important in the positive relationship (Carlsson and Lundstrom 2002,342; Dawson 2003, 493; Hasan et al 2003, 23; Norton 2003, 36; and Blume and Voigt 2007, 534). Property rights as a freedom is a debated and contested issue. Property rights have been claimed to be “a means of preserving liberty.. .an embodiment of liberty, or as a type of liberty” (Gaus 1994, 209). Although, the measurements of property rights available could be interpreted as an individual freedom (most likely a negative freedom) - they have also been interpreted as an institutional factor. Rodrik follows Lin and Nugent (1995, 2306-7) that “it is useful to think of institutions broadly as ‘a set of humanly devised behavioral rules that govern and shape the interactions of human being, in part by helping them form expectations of what other people will do’” (2000,4) Due to the contested role of property rights, they are not used in this study to represent either negative or positive freedom. However, a property rights component is added to the models to evaluate the appropriateness of the model and to assist in comparing the current results to previous research. 5 This study examines the relationship between positive and negative freedom and economic growth, human development and poverty. The nature of this relationship is important, as the assumed nature of the relationships is often used to justify various policy positions. The conclusions of previous econometric research that economic freedom, as defined by economic freedom indices, is positively related to economic growth, is a reasonable and valid conclusion. However, some have generalized from that research that freedom is positively related to economic growth and the type of freedom is left to be assumed. It is important to adequately account for the context and limitations of the existing econometric research keeping in mind the type of freedom and the method by which it is measured. The context and limitations of my study are explicitly stated as is the definition of freedom used. The theories of Friedman and Sen are prevalent in society, distilled down to slogans or political claims such as: “big government reduces economic growth” or “increased democracy and citizen rights increases well-being or quality of life.” These political claims are then used to justify policy directions which involve tradeoffs. Using regression analysis on panel data, this study examines such claims by estimating the relationship between freedom characterized as either negative or positive freedom and economic outcomes such as economic growth, human development and poverty. The above mentioned components are taken from the various published indices of freedom including the Index of Economic Freedom by The Heritage Foundation, Freedom in the World by Freedom House and indices from CIRI Human Rights Data Project (Cingranelli and Richards 2012). The main results of the study are as follows. There is no evidence of a relationship between positive or negative freedom and economic growth or the Human Development 6 Index. The addition of a property rights component to the economic growth model revealed a surprisingly negative and significant relationship between property rights and economic growth. The use of the property rights component in the Human Development Model, called into question the appropriateness of the model and its usefulness. The relationship between both freedoms and the reduction of absolute poverty is positive and significant. The original poverty model and the model including property rights returned similar results and provide strong evidence for a beneficial relationship between increases in both positive and negative freedom and the reduction of poverty. Although the relationship was significant for both positive freedom variables and for only one negative freedom variable, all freedom variables were positively related to the reduction of absolute poverty. The contributions of this study is in the use of Berlin’s dichotomy defining negative and positive freedom and the use of panel data to model the relationships between freedom and the economic outcomes. By explicitly defining freedom as either negative or positive using Berlin’s dichotomy and selecting proxies to represent that freedom in a transparent way, what is meant by freedom is more clearly understood. This clear definition underpins the regression analysis that follows. Through the careful definition of freedom and the separate analysis of the effect of property rights on economic outcomes significantly different conclusions are reached than indicated in previous econometric research. The use o f panel data from countries around the world over a fifteen year time period for the Economic Growth models is an addition to the previous research as most econometric analysis of the effects of freedom on economic growth utilizes cross-section data. The rest of the thesis is organized as follows. Chapter 2 situates this study within the existing discussion regarding freedom and economic development; the previous research provides the study a conceptual basis to identify the mechanisms through which negative and positive freedom can be understood to affect economic growth, human development and absolute poverty. Chapter 3 reviews the literature on the empirical relationship between economic and political freedom and economic growth, human development and poverty to inform the model specification, assist in the selection of non-freedom variables and identify previous research regarding positive and negative freedom. The literature review reveals a lack of empirical research utilizing the positive and negative freedom dichotomy to relate freedom to economic outcomes. Chapter 4 describes the data and methodology used for the regression analysis to test the relationship between positive and negative freedom and economic outcomes. Chapter 5 reports and discusses the results o f the regression analysis, and chapter 6 concludes the thesis by arguing that the approach employed in this empirical investigation has revealed a very different relationship between freedom for the individual and the outcomes of economic growth, human development and absolute poverty than has previously been understood. 8 Chapter 2: Negative and Positive Freedom Having chosen to define freedom as negative or positive using Berlin’s dichotomy and identified two economists ,Friedman and Sen, it is necessary to identify the mechanisms through which negative and positive freedom are theorized to affect economic outcomes. Both Friedman’s and Sen’s theories have been examined in light of Berlin’s dichotomy in previous empirical and theoretical economic investigations (Prendergast 2004; Stroup 2007). A recent empirical investigation has relied on Berln’s definitions of positive and negative freedom (Kaun 2002) to examine the empirical relationship between negative freedom and indicators of well-being for the United States. The contributions of Prendergast, Stroup and Kaun provide guidance in understanding the theoretical concepts of Berlin, Friedman and Sen regarding the relationship between negative and positive freedom and economic growth, human development and poverty Isaiah Berlin distinguishes between negative and positive freedoms or liberties. His conceptualization of negative and positive freedom is summarized by Renee Prendergast (2004) as follows: Isaiah Berlin distinguished between two concepts of liberty: a negative view in which freedom consists in “not being prevented from choosing as I do by other men” and a positive view in which freedom consists in being one’s own master (Berlin, 1969, p. 131). The negative view of liberty has a long tradition going back to Hobbes and Locke. In the twentieth century, it has been associated among others with Hayek, Nozick and to a lesser extent Rawls. The positive view of liberty as effective power to do specific things has links with Hegelian and Marxian traditions and with liberal political philosophy in the US. (Prendergast 2004,45). 9 The negative concept of freedom is entirely concerned with protecting individual freedom from restraint and/or coercion by others, including the government. Berlin argues that a line must be drawn between private life and that of public authority. Since justice demands that all individuals be entitled to a minimum of freedom, all other individuals were of necessity to be restrained, if need be by force, from depriving anyone of it. Indeed, the whole function of law was the prevention of just such collisions: the state was reduced to what Lassalle contemptuously described as the functions of a night-watchman or policeman (Berlin 1992, 3001). It is not necessary to have self-government in order to enjoy a level of negative freedom. The level of freedom is determined by the question “How far does the government interfere with me?” rather than the question “Who governs me?” (Berlin 1992,302). This is central to understanding the difference between the two notions of liberty - positive and negative. According to Berlin, [T]he “positive” sense of liberty comes to light if we try to answer the question, not “What am I free to do or be?”, but “By whom am I ruled?” or “Who is to say what I am, and what I am not, to be or do?” To be one’s own master then is to be “conscious of myself as a thinking, willing, active being, bearing responsibility for [my] choices and able to explain them by reference to [my] own ideas and purposes (Berlin 1992, 303). The difference between the two questions and the two conceptions of freedom stem from two different desires of the individual. He continues: The desire to be governed by myself, or at any rate to participate in the process by which my life is to be controlled, may be as deep a wish as that of a free area for action, and perhaps historically older. But it is not a desire for the same thing. So different is it, indeed, as to have led in the end to the great clash of ideologies that dominates our world (Berlin 1992, 303). As noted above, Berlin “distinguishes between two different concepts of freedom negative and positive - that run through the history of political thought” (Berlin 1992, 298). 10 Berlin’s identification of negative versus positive freedom is echoed in the distinctive views of freedom by two prominent economists -Milton Friedman and Amartya Sen. Friedman identifies his views with negative liberalism “in its original sense [as developed in the 18th and 19th centuries]”, which “emphasized freedom as the ultimate goal and the individual as the ultimate entity in society” (Friedman 1962, 5-6). Freedom for the individual is considered to be desirable and beneficial. Friedman’s position on individual freedom is embedded in his liberal philosophy. He argues: The heart of the liberal philosophy is a belief in the dignity of the individual, in his freedom to make the most of his capacities and opportunities according to his own lights, subject only to the proviso that he not interfere with the freedom of other individuals to do the same ... The liberal will therefore distinguish sharply between equality of rights and equality of opportunity, on the one hand, and material equality or equality of outcome on the other (Friedman 1962, 195). In contrast, Sen’s assertion is that increasing an individual’s freedom improves the individual’s life and enables the individual to determine the nature of his own freedom. He argues: Individual freedom is quintessentially a social product, and there is a two-way relation between (1) social arrangements to expand individual freedoms and (2) the use of individual freedoms not only to improve the respective lives but also to make the social arrangements more appropriate and effective (Sen 1999, 31). Sen describes his book Development as Freedom as “particularly concerned with the agency role of the individual as a member of die public and as a participant in economic, social and political actions” (Sen 1999,19). Friedman and Sen provide two very different concepts of freedom. It is essential to Friedman’s sense of freedom that the state not determine what is good or beneficial, but that be left to the individual - “according to his own lights.” Sen’s concept of freedom embraces 11 the state as a means to expand individual freedom and create a space for individuals and the state to “make ...social arrangements more appropriate and effective.” Their conceptualizations of freedom are part of the foundation of their respective arguments and with different assumptions they necessarily arrive at different conclusions. Thus, a negative freedom creates a space for action, but does not ensure results. This fits with Friedman’s sharp distinction between “equality of opportunity” and “equality of outcome” (Friedman 1962, 195). A positive economic or political freedom is a freedom enjoyed by an individual, which enables him to participate in his own governance and create his own choices and realize those choices. This is compatible with Sen’s “two-way relationship” between social arrangements and individual freedoms. Previous research has examined the mechanisms by which freedom may affect economic growth or development using Friedman and Sen’s theories. Stroup utilizes the framework proposed by Friedman regarding the effects of freedom on economic growth (Stroup 2007, 54). In an examination of Sen’s thinking on development and freedom Prendergast argues that Sen builds on Berlin’s conception of freedom (Prendergast 2004,45-48). The context for Kaun’s empirical investigation into the costs of negative freedom in the United States is Berlin’s concept of positive and negative freedom (Kaun 2002, 372). These prior contributions inform and contribute to the theoretical foundation of my empirical investigation. Stroup equates economic freedom with negative freedom and political freedom with positive freedom to test the tradeoff between democracy and economic freedom. He does not question the components of the indices. Stroup extrapolates Friedman’s theories regarding free markets to implications regarding economic freedom and economic growth. He argues 12 that increased economic freedom will lead to increased economic growth (Stroup 2007, 54). He also asserts that Friedman believes that democratic governments put individual’s economic freedoms in peril (Stroup 2007, 56). Stroup argues that development studies1 assume that democracy is better at improving non-monetary indicators of quality of life and he uses the results of his empirical study to deny that assumption, asserting that increased economic freedom will lead to better development results than will increased political freedom (Stroup 2007, 62). According to Stroup, Friedman’s theories indicate that increases in economic freedom leads to economic growth and increased political freedom could put economic freedom in jeopardy. In an analysis of Sen’s thinking on development and freedom, Prendergast employs Berlin’s definitions of negative and positive freedom. She contends that Sen views freedom as both a principal means to development and a development outcome in itself (Prendergast 2004, 39). According to Prendergast, “Sen makes extensive reference to positive and negative views of freedom” (Prendergast 2004,48). Sen is quoted by Prendergast as arguing that “the natural interpretation of the traditional view of positive freedoms is in terms of the capability to function” (Prendergast 2004, 45). According to her, “Sen’s proposal is that it is more useful to see positive freedom as the person’s ability to do things” (Prendergast 2004, 47). Prendergast agrees with some of Sen’s critics who have identified some ambiguities in Sen’s argument regarding freedom (Prendergast 2004, 50). Prendergast concludes that Sen’s 1 Stroup does not define “development studies” in the referenced paper. 13 capabilities approach2 encompasses broad conceptions of both positive and negative freedom although his concept of positive freedom evolved beyond Berlin’s clear definition (Prendergast 2004, 50). The empirical research into the effects of negative freedoms on human development by Kaun employs Berlin’s definitions of negative and positive freedom and questions the followers of Friedman that support negative freedom at the expense of positive freedom. Kaun equates taxes and regulations as negative freedoms as they are viewed as violations of freedom according to the conservative think tank that generates the index of economic freedom used by Kaun (Kaun 2002, 379). He contends that many if not all parties agree that there is a degree of tradeoff between positive and negative freedom (Kaun 2002, 376). To summarize, previous research has examined the relationship between freedom and economic outcomes in reference to Berlin, Friedman and / or Sen. Only Kaun has used the clear definitions provided by Berlin to evaluate the effects of negative freedom empirically. Others have analyzed the theories of Friedman and Sen with regards to freedom. It is argued that Friedman believed that negative freedom is important for economic growth and positive freedom could be detrimental to negative freedom. Sen’s detailed position on positive freedom is complicated by his view that freedom is both a means and an end of development. For the purposes of this empirical analysis I only examine the effect of positive freedom on human development and not the effect of human development on freedom. Nor, will I 2 “The capability approach is a theoretical framework that entails two core normative claims: first, the claim that the freedom to achieve well-being is o f primary moral importance, and second, that freedom to achieve well-being is to be understood in terms o f people's capabilities, that is, their real opportunities to do and be what they have reason to value” (Robeyns 2011). 14 extend the definition of positive freedom outside the bounds of Berlin’s definition to include entitlements or capabilities as does Sen, but will remain with the confines of Berlin’s original definition. To examine these relationships, proxies to represent positive and negative freedom are obtained from existing indices of economic and political freedom. Hanke and Walters (1997) provide a review and comparison of some of the indices of economic and political freedom. They argue that it is important to evaluate the indices with their intended purposes in mind. According to Hanke and Walters, the economic and freedom indices published by the Fraser Institute, the Heritage Foundation and Freedom House are aimed at policy makers and scholars, consisting mainly of indicators of government policy and are variables believed to be necessary for growth. These indices are designed to help determine what institutions are necessary for prosperity and what policies may be beneficial to economic growth (Hanke and Walters 1997, 133-4). It is not the purpose of this investigation to evaluate how well the indices are performing their intended purposes or what criteria or ideology may have contributed to the composition of the indices. Components of the existing indices are appropriated to act as proxies for positive and negative freedom indicators for this empirical work. The components used in this empirical inquiry are chosen based on Berlin’s definitions of positive and negative freedom and on their relevance to individual freedom - indicators that measure the level of freedom to which the individual is subject or where the individual is the agent. Components that measure the strength of institutional factors or freedom that is enjoyed by states, systems or corporations are not used in this study. 15 An important component of the existing indices is a measurement of property rights. Indicators regarding property rights are included in both economic and political freedom indices and it is often evident this component has a large influence in positive correlations between freedom indices and desirable economic outcomes. Therefore, without classifying it as either a negative or positive freedom or even identifying it as a freedom or an institution a property rights indicator is utilized in this study. The purpose of adding a measurement of property rights is twofold: to evaluate the appropriateness of the original model specifications and to enable comparisons with previous research. As noted, the role of property rights as a freedom is highly debated, however, its importance in the previous research dictates its inclusion in this empirical study. To examine the relationship between positive and negative freedom and economic growth, human development and poverty, this study builds on the previous research of Stroup, Prendergast and Kaun. Extracting components of existing freedom indices, the relationships between freedom and economic outcomes as identified by Friedman and Sen are examined through the lens of Berlins’ definitions of positive and negative freedom. In the following literature review, property rights emerge as an important driver in these relationships and are therefore included in this analysis. By carefully defining freedom as negative and positive, then identifying the mechanisms through which negative and positive freedom are hypothesized to affect economic outcomes this study provides a clearer understanding of the effects of freedom of the individual on economic growth, human development and poverty. 16 Chapter 3: Econometric Research - Freedom and Economic Development In order to examine the relationship between positive and negative freedom and economic outcomes, previous approaches are examined to assist in selecting appropriate model specifications and identify common areas. Previous research uses the available economic and political indices to identify freedom without defining freedom as positive or negative, with the exception of Kaun (2002) who does use Berlin’s dichotomy. The literature review reveals mixed conclusions regarding the relationship between freedom and economic outcomes —such as economic growth, poverty and human development. The dominance of one component of both economic and political freedom —property rights - emerges as a major driver in previously reported positive relationships. Differences in model specification, the use of aggregate or disaggregated indices, and the presence of direct or indirect, linear or non-linear relationships differentiate the empirical literature. The literature regarding the effect of economic freedom on economic growth is more extensive. The research examining the effects of political freedom or the joint effects of political and economic freedom on economic growth and other economic outcomes is not as extensive and relies heavily on the former research for much of its methodological and theoretical basis. The literature review informs the subsequent empirical analysis. Methodology choices such as the use of an augmented Solow growth model and, the selection of non-freedom variables have been influenced and informed by the research described below. Generally, previous research agrees that economic freedom (or some of its components) is positively correlated with economic growth. The investigations regarding the relationships between political freedom and economic growth and economic and political freedom and 17 human development and poverty are less conclusive. A relatively consistent finding throughout the research is the importance of property rights in relationships between freedom and economic outcomes. Positive correlations are often based on the strength of the property rights components effect on the dependent variable. The inconclusiveness of the literature regarding the roles of various freedoms may be related to the economic / political categorization of freedom indices employed in the literature. As noted earlier, this categorization is problematic for a number of reasons. First, there is overlap between the two with some components being included in both categories, for example property rights and rule of law. Second, the stated intent of the indices is not to measure individual freedom. Third, some of the components have been argued to be institutional factors rather than indicators of freedom (Hanke and Walters 1997, 133-4; Rodrik 2000, 6-8). Fourth, there are many components in the indices and although some may be argued to measure freedom of some definition, it is difficult to make the same argument for all the components included in a particular index and the definition of freedom is implied rather than explicitly stated. For a summary of indices used in the econometric research cited in the following literature review, refer to Appendix 1. Kaun’s work (2002) uses the alternate categorization which is rooted in the conceptual philosophical view of Isaiah Berlin who identifies freedom as either negative (freedom from) or positive (freedom to). As discussed, this study follows Kaun’s approach to investigate the potential role of various freedoms - categorized as negative or positive - on the economic outcomes of economic growth, human development and poverty. My thesis expands on Kaun’s work in several respects. First, it works with a large sample of countries from around the world rather than the states of the United States as in 18 Kaun’s work. Second, in the spirit of the dominant approach in the literature, it utilizes an augmented Solow growth model to control for the impact of conventional macroeconomic variables on economic growth, human development and poverty. Third, it attempts to identify the effects of negative and positive freedom on economic growth, human development and poverty, while Kaun identified the effects of negative freedom on well­ being indicators. Freedom and Econom ic Growth There appears to be a consensus that economic freedom or some components of it and economic growth are positively related. The general relationship between economic growth and economic freedom is confirmed in meta-analysis (Doucouliagos and Ulubasoglu 2006) and critical surveys of the existing literature (de Haan, Lundstom and Sturm 2006; Berggren 2003). For instance, Cole (2003) compares the effect of the Economic Freedom of the World (EFW) composite index on economic growth with two very different growth models. One is an augmented Solow growth model and the other is a model proposed by Gallup, Sachs, and Mellinger (1999), which explains per capita income growth in terms of the convergence effect and geographical variables. Cole concludes that the effect of the EFW index on economic growth is quite robust with respect to major changes in model specification (Cole 2003, 191,196). He found that the effects of the EFW index on economic growth were the same in both models indicating a strong statistical relationship. In their meta-analysis of 52 studies on the relationship between economic freedom and economic growth Doucouliagos and Ulubasoglu (2006, 60) find an overall positive association. In a critical survey of some of the studies using the EFW index de Haan, et al 19 (2006) identify some of the problems with using the aggregate index as well as the models utilizing the index. In addition, they review the evidence regarding the relationship (linear or non-linear; direct or indirect) between economic freedom and economic growth. They advise that the relationship between economic freedom and economic growth is a complex issue and it is important to consider the different types of economic freedom, as they seem to have different effects on growth (de Haan et al, 179). Similarly, Heckelman and Stroup (2005) investigate the aggregation of the EFW index and suggest that empirical research relating economic freedom to growth or other variables should keep the various elements of economic freedom separate in order to allow each element to speak freely (2005, 964). The following studies follow this advice and examine the relationship between specific components of economic freedom and economic growth. Studies examining the effect of components of the economic freedom indices on economic growth produce a variety of results. One trend that emerges is the importance of property rights as a driving force in the relationship between economic freedom and economic growth (Carlsson and Lundstrom 2002; Acemoglu and Johnson 2005; and Yishay and Betancourt 2008). There are a variety of conclusions regarding the other components of economic freedom, but property rights emerge relatively consistently as having a positive and significant relationship with economic growth. Using an extension of the Solow growth model, Dawson (2006) isolates the effect of government regulation on economic growth. He finds that regulation is negatively related to growth. However, his results are not significant when the level of aggregate economic freedom is included, which he attributes to high levels of correlation between the two variables. The effect of government regulation remains negative and statistically significant 20 when changes in the index are included (Dawson 2006, 508). Using variables that describe colonial origins rather than an augmented Solow growth model, Acemoglu and Johnson (2005), compare the impact of property rights institutions versus contracting institutions. The authors distinguish between property rights institutions that protect citizens against expropriation by government and elites and contracting institutions that enable citizens to enter into private contracts that are enforced in society. They find that property rights institutions have a major positive influence on long-run economic growth (Acemoglu and Johnson 2005,949). Carlsson and Lundstrom (2002) also find that some components of the EFW do not have a positive relationship with economic growth. In addition, they find that increased freedom to trade with foreigners decreases the growth rate. Increased freedom in terms of lower government consumption and transfers decreases the growth rate at index values lower than 8.86. Consequently, there is a hump-shaped relation between government size and growth (Carlsson and Lundstrom 2002, 342). The only variables they find to be positively and robustly related to GDP growth are legal structure and private ownership and freedom to use alternative currency (Carlsson and Lundstrom 2002, 343). A summary of the results of empirical studies examining the effects of different categories of economic freedom on economic growth is provided by Carlsson and Lundstrom (2002). It reveals that few categories have an unambiguously positive relationship with economic growth (Carlsson and Lundstrom 2002, 337). The category including protection of property rights and the rale of law does appear to show a clear positive relationship with economic growth by a number of studies. 21 The studies examining political freedom’s effects on economic growth are more limited in number and are less conclusive as a whole. A review of the studies that examine political freedom’s effects on economic growth has produced inconclusive results. Some empirical inquiries into the effect of political and civil freedom on economic growth conclude that the relationship is weakly negative (Helliwell 1994; Barro 1996). Others find a non-linear relationship (Goldsmith 1995). De Haan and Siennann (1996) show that although a positive relationship between political freedom and economic growth can be found, it is not robust when subjected to sensitivity analysis. The conclusions of Yishay and Betancourt (2008) are in conflict with the other studies. They find a positive, significant relationship between political freedom and economic growth. However, the indicator of political freedoms that shows a significant positive effect in the Yishay and Betancourt analysis includes property rights, which as noted earlier, some researchers consider this an economic freedom, not a political or civil liberty and others argue it is not a freedom at all but an institutional factor. Helliwell (1994) examines the empirical relationship between democracy and economic growth using an augmented Solow growth model and Gastil’s index of political and civil liberties as a measurement of democracy. He finds that the effect of economic growth on the political and civil liberty index is robust and positive, but the index’s effect on economic growth is negative and insignificant (Helliwell 1994, 225). De Haan and Siermann (1996) come to the same conclusion regarding political freedom’s effect on economic growth using an extension of an augmented Solow growth model, with a vector of explanatory variables suggested by other studies and a measure of how many years a country has been a 22 democracy. Their main conclusion is that the effect of democracy on economic growth is not robust when analyzed using extreme bound analysis (de Haan and Siermann 1996, 175). Barro (1996) confirms these results and extends the analysis to find that there is a suggestion of a nonlinear relationship in which more democracy enhances growth at low levels of political freedom but depresses growth when a moderate level of freedom has already been attained (Barro 1996,1). At the same time he finds that improvements in the standard of living (measured by GDP, health status and education) substantially increase the probability that political freedoms will grow. These results draw him to the conclusion that “political freedom emerges as a sort of luxury good ... Thus, in the long run, the propagation of Westem-style economic systems would also be the effective way to expand democracy in the world” (Barro 1996, 24). Yishay and Betancourt (2008) decompose the Freedom House Index of political and civil liberties and conclude that one subcategory - Personal Autonomy and Individual Rights outperform all available indicators of property rights institutions in explaining long-term economic growth. This subcategory is said to capture the level of second generation human rights - defined as economic and social freedoms - that affect the mobility of individuals with respect to housing, employment and higher education, as well as the level o f protection of property rights. This result is robust with respect to reverse causation. The authors test the robustness of their estimate by including geographic variables and other growth model explanatory variables (Yishay and Betancourt 2008,2). Some of the empirical studies mentioned above have included both economic freedom and political freedom in their models to avoid omitted variable bias. The following studies look at the relationship of both categories of freedom and economic growth more 23 purposefully. Goldsmith (1995) looks at three hypotheses, (1) democratic countries perform better economically; (2) countries with strong protections of economic rights perform better; and (3) democratic institutions and economic rights enhance economic performance. He utilizes the Freedom House index and the Heritage Foundation’s property rights index in an augmented Solow growth model to test his hypotheses for 59 developing and transitioning countries in the 1980s and 1990s. He concludes that the regression results do support the view that political rights and property rights enhance economic growth in those countries (Goldsmith 1995,167). Blume and Voigt (2007) explore the association between various human rights and economic performance in terms of investment, total factor productivity and economic growth. They distinguish four groups of human rights: (1) basic human rights which they identify as freedom from state interference or negative rights; (2) economic rights are associated mainly with property rights; (3) civil and political rights; and (4) social or emancipatory rights which the authors label as positive rights. In Blume and Voigt’s study positive rights refer to endowments from the state to individuals, as in rights to food or housing (Blume and Voigt 2007, 511). They attempt to shed empirical light on the nature of the relationship between these human rights and economic performance as posited by different economists including: Hayek’s position that basic rights have a positive effect on welfare and growth, while social rights are counterproductive; the Barro-Posner view that it is an issue of sequence - i.e. that property rights are necessary first and other rights will follow; and finally the position of Sen that freedom, fairness and reciprocity are important and these components of social capital have a positive effect on welfare and growth (Blume and Voigt 2007,513). 24 Blume and Voigt search many sources to obtain their explanatory variables and then use factor analysis to condense multiple variables into combined variables and minimize correlations. They then test the effects of the explanatory variables on growth, investment, and productivity. They summarize their results as follows: basic human rights (first generation rights) have a strong positive influence on investment; property rights have strong impacts on growth, investment and total factor productivity; civil rights impact investment and total factor productivity; and emancipator rights (second generation human rights) positively influence total factor productivity (Blume and Voigt 2007,534). In regards to the competing hypotheses, Blume and Voigt contend that the results indicate that the Hayek hypothesis is not supported by the data. The regressions did not reveal the significant negative impact that Hayek expected from strong emancipator rights with regard to economic development. Some support is evident for the Barro-Posner hypothesis. The Sen hypothesis is not fully confirmed as neither basic human rights nor civil and emancipatory rights have a significant impact on GDP growth (Blume and Voigt 2007, 534). As indicated above, the relationships between economic freedom, political freedom and economic growth have been studied in a variety of ways with a variety of results. The overall conclusion is that economic freedom has a positive correlation with economic growth. Investigations regarding the relationship between political freedom and economic growth return mixed results. An augmented Solow growth model is used in this study to examine economic growth, human development and poverty; this is a commonly used model for this type of empirical study (Cole 2003,191,196; Dawson 2006,492; Helliwell 1994,236). The non-freedom variables that are used in this empirical analysis are used in previous studies and include: 25 investment, government spending, openness to trade, population growth, and a measure of education (R. Barro 2003, 231; Blume and Voigt 2007, 528). It is clear from the previous discussion that measures of property rights are important in the relationship between freedom and economic growth (Blume and Voigt 2007, 534; Carlsson and Lundstrom 2002, 343; Yishay and Betancourt 2008,2). And finally, the general approach of Blume and Voigt informs this study. Blume and Voigt identify the theories of three economists and three different mechanisms to test the relationship between freedom and economic outcomes. The mechanisms identified in this investigtion are substantially different, although Blume and Voigt also rely on Sen’s work to some extent. Freedom and Human Development and Poverty The following studies are not in agreement regarding the relationship of economic freedom or political freedom on human development or poverty. Some conclude that economic freedom is more influential than political freedom and some claim that both are important. Norton’s research indicates that only overall economic freedom and property rights have a robust, positive and significant relationship with well-being indicators and poverty reduction (Norton 2003, 36). The study by Hasan et al (2003) finds civil liberties to be an important contributor to poverty reduction (Hasan et al 2003, 23). It should be noted that the civil liberties indicator from Freedom House includes a measure of property rights and is highly correlated with the political rights indicator. 26 Norton (1998) uses the property rights component of the EFW and EEF in a regression analysis with Human Development Index (HDI) and Human Poverty Index (HPI) and its components as dependent variables3. He converts the property rights measure to dummy variables for weak and strong rights to use ordinary least squares regression. He finds that better specified property rights are associated with higher levels of human development as represented by the HDI. In regards to the HPI, Norton finds that where property rights are strong, the HPI is reduced substantially. The relationship is not as strong when he tests the effects of property rights on the components of the HPI (Norton 1998,238-9). He concludes that these results are generally more robust in the cases where the EFW measures of property right are used rather than the IEF’s. (Norton 1998, 244). In a follow-up study Norton (2003) uses both the EFW and the International Country Risk Guide in a model with geographical explanatory variables to investigate the effects of economic freedom on measures of human development (HDI) and poverty (HPI). He argues there is a strong case for inclusion of geographic variables in estimates of the relationship between economic institutions and human well-being (Norton 2003,29). However, for the model with the HDI index as the dependent variable, the only robust results for the 3 “ The components are similar to those in the HDI including three basic dimensions o f well-beinglongevity, knowledge, and a decent living standard. However, using the deprivational approach, the HPI entails different measures. The first dimension is m easured by the number o f people in the population not expected to survive to age 40. The second dimension is m easured by the proportion o f adults who are illiterate and therefore excluded from the world o f reading and communication. The third dim ension is a composite o f three variables— the percentage o f people without access to health services, the percentage o f people without access to safe water, and the percentage o f malnourished (underweight) children under the age o f five” (Norton 1998,237). 27 geographic variables are for urbanization. The institutional [economic freedom] variables uniformly support the hypothesis that institutions favorably affect human development as measured by the HDI and HPI (Norton 2003, 30-31). Norton concedes that the institutional variables reveal a somewhat mixed pattern. Fourteen of the twenty estimates are statistically non-significant and the most robust results relate to simple property rights and economic freedom (Norton 2003, 31). Esposto and Zaleski (1999) examine the impact of levels and changes o f composite economic freedom index on levels and changes in literacy rates and life expectancy measures (two of the three components that comprise HDI). Their study uses literacy rates and life expectancy as proxies for quality of life. They conclude that although the effect of economic freedom on increased life expectancy is more significant than the effect on literacy rates, the evidence supports the hypothesis that greater economic freedom leads to an improvement in the quality of life (Esposto and Zaleski 1999,186). As the literature reviewed so far indicates, the research into the effects of economic freedom on growth, poverty and human development is varied in model specifications, sensitivity tests and data selection. Berggren (2003) provides a survey of this empirical research. He notes that one needs to be careful when interpreting empirical studies, especially when sensitivity analyses are lacking and panel data is not used. In addition, the causal relationship between variables can be unclear (Berggren 2003,200). Inquiries into the linkages between economic freedom, political freedom, and outcomes like poverty and human development have a variety of assumptions or hypothesis regarding the relationship. Some, like Goldsmith (1997), assume that political freedom impacts economic freedom which impacts well-being and growth, others, like, Hasan et al. (2003) 28 test the direct effects of political and civil liberties on poverty reduction. Stroup (2007), on the other hand, investigates the interaction of political and economic freedom concurrently on indicators of human development. Goldsmith (1997) investigates the relationship between aggregate economic freedom and the HDI; aggregate freedom and growth; and political freedom’s impact on economic freedom using a series of models. His regression results show a strong relationship between the economic freedom indices and the HDI (Goldsmith 1997, 36). Goldsmith appraises the determinants of economic freedom and finds that political freedom and per capita GDP explain about half the variance in the Economic Freedom of the World index (Goldsmith 1997,41). Hasan et al (2003) uses the Freedom House indices of political and civil freedoms as explanatory variables and compile explanatory variables to measure the effect of economic freedom on the incidence of absolute poverty. They advise that their findings are tentative. However, they conclude that openness to trade is robustly associated with poverty reduction. Labour market regulation does not have a direct significant impact on poverty. Total government expenditure is positively related to poverty, which lead the authors to speculate that increased government expenditure may indicate fiscal irresponsibility. However, high incidences of poverty may cause governments to increase expenditures rather than high government expenditures preceding high levels of poverty. Civil liberties, which in their study include property rights, contribute significantly to poverty reduction while political liberties have seemingly no impact on poverty reduction (Hasan et al 2003, 23). Stroup (2007) uses fixed effects specification but tries to capture the interaction of economic freedom (EF) and political freedom (PF) by testing the impacts of EF, EF and high 29 PF, PF, PF and high EF on six different dependent variables measuring aspects of quality of life. The latter include: years of life expectancy at birth; child mortality rate; adult literacy rates; percent of population with grade 5; percent of population with access to improved water and percent of two-year-old children with adequate vaccination. His results imply that economic freedoms in society have a larger positive influence on all the measures of quality of life examined than do political rights. The relationship is not as strong when a relatively robust democracy exists, but is still statistically significant and beneficial. Democracy has a relatively smaller, positive effect on five of the six measures of well-being (Stroup 2007, 53) The research regarding freedom, categorized as economic and political, and human development and poverty is less conclusive than the economic growth literature. The results are mixed. There are a variety of models used to estimate the relationships with no generally accepted model emerging. In light of this, I have chosen in this study to use the augmented Solow growth model as does Goldsmith (1997, 37) to examine human development and poverty as well as economic growth. Similar to the freedom and economic growth literature, property rights’ measures are a driver in positive correlations between freedom and human development and poverty, therefore a measure of property rights is added to the models used in this study to aid comparisons to previous research (Hasan, Quibria and Kim 2003, 23; Norton 1998, 238-9; Norton 2003, 31). Positive and Negative Freedom and Human Development Very little empirical research examining economic outcomes utilizes the negative versus positive rights dichotomy - one empirical paper examines the impact of different levels of negative freedom on economic outcomes within the United States (Kaun 2002). Kaun’s use 30 of Isaiah Berlin’s classification of freedom as positive or negative is only concerned with the states within the United States of America rather than a cross country evaluation. According to Kaun (2002), the empirical investigation into the probable effects of negative and positive freedoms on economic outcomes and human well-being indicators is testable, unlike the ideological debate regarding the nature of freedom and which type of freedom is more desirable for its own sake. Kaun (2002) investigates the impact of negative freedom across states using the index of freedom from the Center for Policy and Legal Studies, which he argues is comparable to the Heritage Foundation and Fraser Institute Index in subject and ideology. This ranking, he argues, “allows the debate regarding the value of one freedom over another to move from the theoretical to the empirical, that is, to begin to measure just what is so positive about negative freedom” (Kaun 2002,375). To test the effects of negative freedom, Kaun selects ten measures of human well-being4. In addition to the freedom variables, the model also includes geographic and ethnic dummy variables. His conclusion is that greater negative freedom does have positive effects on some measures of well-being and a negative relationship with others. He found a perverse relationship between negative freedom and levels of suicide, medical coverage, quality of education, voting behavior and labour union participation (Kaun 2002,385). In particular his analysis revealed that higher negative freedom is associated with increased levels of suicide, lower SAT scores and decreased voting and labour union participation. He did find that 4 The ten dependent variables are: overall level o f poverty, child well-being, child poverty, rates o f suicide, violent crime, no medical insurance, voter turnout, union labour and net-migration and SAT scores. The SAT Reasoning Test (originally the Scholastic Aptitude Test) is a college readiness assessment tool used in the United States by many colleges to screen applicants. 31 states with higher degrees of negative freedom had relatively low rates of poverty, child poverty, crime of all sorts and higher rates of net migration (Kaun 2002, 381). Kaun contends that the analysis suggests that the minimalist government aspects of negative freedom are often inversely related to some conditions of well-being within the states. He argues that such an adverse relationship may well stem from the behaviour patterns encouraged by an excess of negative freedom (Kaun 2002, 371). As indicated earlier, Kaun’s empirical study is the only empirical research to date testing the effects of freedom on economic outcomes where freedom is defined using Berlin’s dichotomy. There is extensive previous research that has examined the relationship between freedom, categorized as economic or political, and economic outcomes at the country level. In my empirical study, the question of the effect of different types of freedom on economic development is redefined using the positive / negative freedom dichotomy outlined by Berlin rather than the categories of economic or political freedom. This investigation of the effects of negative and positive freedom on economic growth, poverty and human development uses country level data and some of the components of the commonly utilized indices of freedom building on the previous research through a new lens. Components of the existing freedom indices are selected as indicators of negative and positive freedom and the relationship between positive and negative freedom and economic outcomes is examined using regression analysis. The econometric relationships between negative and positive freedom for the individual and the economic outcomes of economic growth, human development and poverty have been theorized by Friedman and Sen. The relationship between indices of economic and political freedom and these economic outcomes have been tested and the results have been framed has 32 representing the relationship between freedom and the economic outcomes without defining freedom independently of the index used. The assumption that economic and political indices measure freedom enjoyed by the individual is unsupported and unexamined in the previous research. This study defines freedom using Berlin’s dichotomy of negative and positive freedom, identifies proxies to represent negative and positive freedom of the individual from available measures and re-examines the relationships. The details o f data and methodology used in this study are explained in the following chapter. 33 Chapter 4: Data & Methodology The relationship between positive and negative freedom and economic outcomes is examined using data from between 1995 and 2009 for up to 148 countries. The economic outcomes are measurements of economic growth, human development and absolute poverty. Positive and negative freedoms are represented by certain components of existing indices of freedom. The relationships are examined using an augmented Solow growth model. The intent is to examine the relationship between negative and positive freedom experienced by individuals in a country and economic outcomes measured at the country level. To this end, an attempt is made to extract components of existing indices that are appropriate proxies for positive and negative freedom where the agent is the individual. The indices utilized in this study were originally intended to measure human rights, economic or political freedoms. The criteria for choosing the individual components to include in the model, is necessarily subjective. The goal of extracting measures of freedom enjoyed by the individual that are definitely either freedom from government restraint or involvement - negative freedom - or freedom to participate in the governance of oneself— positive freedom - is not clearly and inarguably achieved. The selection of the measures used is, however, transparent. There is a prevalence of the use of the augmented Solow growth model in previous studies that measure the effect of economic freedom on economic growth. The basic Solow growth model states that output per worker depends on: initial output per worker; level of technology; rate of technological progress; savings rate; growth rate of the workforce; depreciation rate; share of capital in output and the rate of convergence (Hoeffler 2000,3). This model was used to model convergence of gross domestic product per capita. The initial 34 levels of output per worker are necessary when modeling for convergence, but are not used in augmented Solow growth models which are being used to identify the determinants of long run economic growth. Augmented Solow growth models used in previous econometric research into this area typically are more general and include a measure of investment ratio to GDP as a proxy for physical capital, a measure of government spending, a trade openness indicator, school enrollment as a proxy for human capital and a measure of population growth. The measurement of investment, government spending and trade openness are also included in economic freedom indices, this is an overlap between the augmented Solow growth model and the indicators chosen by the indices publishers. The relationship between positive and negative freedom and economic growth is estimated in this study using fixed effects in a panel data model. The model with the Human Development Index as the dependent variable has only 2 time periods available for analysis. The panel data model is not suited to this type of data. Therefore, the data is modeled using ordinary least squares. Similarly, the data available for the poverty model is quite porous with limited variation over time due to few observations per country and is therefore modeled using ordinary least squares as well (Beck 2004, 4). The data and methodology used are discussed further below. Data Sources Economic Outcomes - Dependent Variables The three economic outcomes examined in this study are economic growth, human development and poverty, which are key indicators of economic development. Data 35 quantifying these economic outcomes is available for countries on an annual basis with varying regularity. The data for economic growth is obtained from the Penn World Tables 7.0 (Heston, Summers and Aten 2011). The gross domestic product (GDP) per capita is reported in 2005 constant prices and transformed to percent change. Data for 116 countries for the time period 1995 to 2009 is used for the economic growth model. See Appendix 3 for a list of the countries used in these models. The Human Development Index was introduced in 1990 as a more comprehensive measurement of development. It is a composite index combining indicators of income, life expectancy, and educational attainment. The income component is measured by gross national income (GNI) per capita instead of GDP per capita. The logarithm of income is used to indicate the diminishing importance of increases in income with increasing levels of income. The life expectancy at birth component is calculated with a minimum of 20 years and a maximum of 83.4 years. The education component is measured by combining the average years of schooling for adults aged 25 years and the expected years of schooling for children entering school (Human Development Index 2011). The Human Development Report 2011 cautions against using data from previous reports, since the sources of the data used are continually improving their data (Human Development Report 2011 - Readers Guide 2011). Therefore, this study uses the trend data as recommended and provided in the 2011 report for the years 2000 and 2005 for 148 countries. See Appendix 4 for a list of countries used in these models. Absolute poverty data is obtained from World Bank data (World Development Indicators (WDI) and Global Development Finance (GDF) 2012). Poverty headcount ratio at $2 a day 36 based on the Purchasing Power Parity (PPP) at 2005 international prices is the data series chosen to represent absolute poverty. The data is reported as the percentage of the population living at $2 or less per day. Data for this economic outcome is inconsistently reported from fewer countries than data for other economic outcomes. The data was obtained for 87 countries for 13 different time periods between 1996 and 2008. See Appendix 5 for a list of countries used in the poverty models. Positive and Negative Freedom Indicators In order to measure positive and negative freedom, indicators from three indices are utilized. The components are from The Heritage Foundation’s Index of Economic Freedom; Freedom House’s Freedom in the World and The Cingranelli-Richards Human Rights Dataset, CIRI Human Rights Data Project. These indices were designed to measure political and civil liberties, economic freedom and human rights, respectively. Isaiah Berlin’s distinction between negative and positive freedom is used as the criteria for selecting the indicators as proxies of negative and positive freedom enjoyed by individuals. As explained before, Berlin defines negative freedom as freedom from coercion from others including the government and positive freedom as the desire to be self-governed. The concepts of negative and positive freedom as outlined by Berlin are in relation to the individual. The theories of Friedman and Sen are being examined in this analysis. Both economists argue that it is the individual’s freedom that is important in achieving desirable economic outcomes such as economic growth (Friedman) and development (Sen). Thus, I am attempting to select indicators that measure to what extent the individual is free from coercion from the government and to what extent the individual is free to govern himself. 37 As the indices were designed for other purposes, many components do not meet the criteria as outlined above. Some of the components of the indices do not easily represent a freedom that is enjoyed by the individual, although they do represent important institutional factors. For instance, in the Index of Economic Freedom (IEF), the measurement of Financial Freedom is a measure of the level of regulation on banks imposed by the government. This is a freedom that is considered an economic freedom by The Heritage Foundation and may be considered a negative freedom, but it is difficult to categorize it as a freedom enjoyed by the individual. Rodrik labels economic freedoms such as this as market supporting institutions (2000, 6-8). The IEF also includes property rights. Many argue that property rights are a negative freedom, however, the role of property rights as a freedom is not uncontested (Gaus 1994, 209) and some argue it is an institutional factor (Rodrik 2000, 4). If it is understood as a freedom, it is not clearly understood if it is an economic or a political freedom as both economic and political freedom indices include measure of property rights. Due to the debated and contested nature of property rights, this study does not label property rights as a freedom of any type, but I do use property rights in alternate models to further evaluate the initial models and enable comparisons to previous research. There is a strong tendency in the previous empirical literature to use the EFW index from the Fraser Institute, although some do prefer the Heritage Foundation’s IEF, arguing that the Fraser Institute measures are dominated by outcome variables whereas the Heritage measures are primarily policy variables the government can actually control (Heckelman 2000, 73). The Index of Economic Freedom is used here partially because of its methodology and partially due to data availability. 38 From the Index of Economic Freedom, one component most closely met the criteria outlined for representing negative freedom - Fiscal Freedom. Fiscal Freedom is the measurement of taxes in all forms; it is calculated using individual and corporate tax rates and total tax revenue as a percentage of GDP. This measure is a reasonably close fit to the concept of negative freedom although the corporate tax rate is weighted equally with the other two components. A lower taxation burden is indicated by a higher score (The Heritage Foundation 2008,45). This is the only negative freedom indicator in the Index of Economic Freedom which reflects freedom of the individual from state intervention. The indicator is not ideal as it includes the highest corporate tax rate as well as individual tax burdens, but it is the best fit of the ten indicators. In addition, taxes are often equated with a violation of negative freedom in the debate (Kaun 2002, 379). The complete list of the components of the Index of Economic Freedom is given in Table 1. Therefore, the only component from the IEF used in this study to represent negative freedom is Fiscal Freedom, as it meets the criteria of being a negative freedom enjoyed by the individual. Table 1. Components of the Index of Economic Freedom Comnonent Business Freedom Trade Freedom Fiscal Freedom Government Size Monetary Freedom Investment Freedom Financial Freedom Property Rights Freedom From Corruption Labour Freedom Descrintion Freedom of entrepreneurs to start businesses, amount of regulation and impediments Level of tariffs and other obstacles to free trade between nations Level of taxes of individuals and corporations Measurement of government expenditures Measure of inflation, price stability Restrictions on foreign investment Amount that banks are controlled or regulated by governments Security of property rights Measurement of amount of corruption in government Ease of companies to hire and fire employees 39 Freedom in the World measures Political Rights and Civil Liberties in two indices. The Political Rights index includes three subcategories: electoral process; political pluralism and participation; and function of government (Freedom House 2012, 34). This index is a good approximation of some aspects of positive freedom as defined by Berlin. The Civil Liberties index has a measure that could represent positive freedom, Associational and Organizational Rights, and one that could represent negative freedom, Freedom of Expression and Belief. However, the Civil Liberties index also has a measure for the rule of law and includes a component regarding property rights which causes it to be even less suitable as a measure of positive freedom (Freedom House 2012, 35). Therefore, only the Political Rights index is included in this study as a positive freedom indicator. A lower score indicates a higher level of positive freedom, however, for the purposes of this study the index is reordered so that a higher score indicates a higher level of positive freedom. Therefore, a positive effect on the economic outcomes would be indicated with a positive coefficient in this study. This reordering simplifies interpretation of the regression results. All the components of political freedom as measured by Freedom House are listed in Table 2. Only the Political Rights index is used to represent positive freedom in this study as the Civil Liberties index is difficult to classify as measuring negative or positive freedom. ComDonent Political Rights Civil Liberties Table 2. Components of the Freedom in the World Index Descriotion Electoral Process Political Process and Participation Functioning of Government Freedom of Expression and Belief Associational and Organizational Rights Rule of Law Personal Autonomy and Individual Rights (including Property Rights) 40 Table 3. Description of The Cingranelli-Richards (CIRI) Human Rights Dataset Index Physical Integrity Rights Index Disappearance Extrajudicial Killing Political Imprisonment Torture Modified Empowerment Rights Index Freedom of Assembly & Association Electoral SelfDetermination Worker’s Rights Descrintion adds values from Torture, Extrajudicial Killing Political Imprisonment and Disappearance Indicators “cases in which people have disappeared, political motivation appears likely, and die victims have not been found” “killings by government officials without due process of law” Indicator of how “many people are imprisoned because of their religious, political or other beliefs” “the purposeful inflicting of extreme pain ...by government officials” Adds values from Freedom of Assembly and Association, Electoral Self-Determination, Workers Rights Freedom to “assemble ... and associate with other persons in political parties, trade unions, cultural organizations or other special-interest groups” “freedom of political choice and the legal right and ability to practice to change the laws and officials that govern them through free and fair elections” Freedom of association at workplaces and the right to bargain collectively. The data from the CIRI Human Rights Data Project is more easily adopted as proxies for negative and positive freedom as each right is individually reported. Two sub-groups from the CIRI Human Rights Data Project’s many rights measurements are used in this study the Physical Integrity Rights and the Modified Empowerment Rights. The Physical Integrity Rights index is an additive index of four individual indices of clearly negative freedoms. The four individual indices measure freedom from coercion: the incidence of disappearances, extrajudicial killings, political imprisonment and torture. The Modified Empowerment Rights Index is an additive index constructed from three individual rights indices which are reasonably argued to be positive freedoms. These are freedom of assembly and association, electoral self-determination, and workers’ rights. The Modified Empowerment index will 41 thus be used to represent positive freedom. The individual indices are measured on a scale of 0 to 2 with 0 indicating lack of freedom and 2 indicating virtually unrestricted freedom. The additive indices range from 0 to 8 for the Physical Integrity Rights Index and from 0 to 6 for the Modified Empowerment Rights Index. Table 3 describes the various CIRI indices employed in this study (Cingranelli and Richards 2012). Macroeconomic Determinants of Growth Variables Other independent variables were chosen as macroeconomic determinants of growth and are commonly used in the augmented Solow growth model. The additional variables are investment as a percentage of GDP, used as a proxy for growth of physical capital; government consumption as a percentage of GDP, used as a proxy for government spending; openness as a percentage of GDP, used as a proxy for trade impacts; the rate of growth of population; and the percentage of the population over 15 with secondary education, used as a proxy for human capital. The selected variables are theoretically important and have been shown in previous work to be important determinants of economic growth (Barro 2003,231; Blume and Voigt 2007, 528). According to theory, investment as a percentage of GDP is expected to be positively related to economic growth as is human capital. Population growth is expected to be negatively correlated with economic growth (Mankiw, Romer and Weil 1992,410). The inclusion of government spending and openness is common in the relevant literature. Government spending is expected to have a negative relationship with economic growth (R. J. Barro 1991,407; Mitchell 2005; Carlsson and Lundstrom 2002, 343; Dawson 2003,487). Openness is expected to have a positive relationship with economic growth (Blume and Voigt 2007, 528). Similar relationships are expected to hold between these variables and the 42 Human Development Index. However, opposite relationships are expected to hold between those independent variables and poverty head count as the dependent variable. Investment, government consumption and openness are from Penn World Tables 7.0 (Heston, Summers and Aten 2011). These variables are in terms of 2005 constant prices. Population growth data is from World Bank (World Development Indicators (WDI) and Global Development Finance (GDF) 2012). Education data is obtained from the Barro-Lee Educational Attainment Data Set (Barro-Lee Educational Attainment Dataset 2012). A list of all variables and sources is available in Appendix 2. Methodology To investigate the role of positive and negative freedoms on economic growth, human development and poverty, three separate models are considered each considering one of the three economic outcomes. These models are named Economic Growth Model, Human Development Model and Poverty Model respectively. Each model is estimated using two alternative specifications: one using the Fiscal Freedom (negative freedom) and Political Rights (positive freedom) measurements, which is labeled Model A; the other using the Physical Integrity (negative freedom) and the Modified Empowerment Index (positive freedom), which is labeled Model B. This specification choice was made mainly because the Modified Empowerment Index and the Political Rights measurement have a strong positive correlation (0.81). In addition, the variables in Model A are taken from economic and political freedom indices and the variables in Model B are taken from a human rights dataset. 43 Economic Growth Models In these models the economic growth rate is the dependent variable in an augmented Solow growth model. As mentioned, Model A utilizes the Fiscal Freedom and Political Rights and Model B utilizes the CIRI Rights data. The panel data that is available for the Economic Growth Model combines time and country dimensions. The availability of panel data enables the use of fixed effects modeling which isolates the effects of each country or time period or both on the dependent variable separate from the independent variables. This model is estimated using cross-section (country) and period (time) fixed effects. The independent variables used in the two specifications of the Economic Growth Model are listed in Table 4. Table 4. Independent Variables in the Economic Growth Model Model A Investment (% of GDP) Government Spending (% of GDP) Openness to Trade (trade as % of GDP) Population Growth (% change) % of population over 15 that have completed Secondary Education Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Model B Investment (% of GDP) Government Spending (% of GDP) Openness to Trade (trade as % of GDP) Population Growth (% change) % of population over 15 that have completed Secondary Education Physical Integrity Rights Index (negative freedom) Modified Empowerment Rights Index (positive freedom) Human Development Models To study the relationship between positive and negative freedoms and the human development index, I utilize a similar base model as was used to model the relationship with economic growth. However, the education indicator is omitted from this model as there is an education component in the HDI. Due to the limited amount of trend data available for the 44 Human Development Index that overlaps with available freedom indicators, there are an insufficient number of time periods to conduct meaningful panel data analysis (Beck 2004, 4). Therefore, the models are estimated using ordinary least squares. Similar to the Economic Growth models, two versions of the model are specified: Model A that includes indicators from the Index of Economic Freedom and Freedom in the World indices and Model B that includes indicators from the CIRI dataset. The independent variables for both models of Human Development are described in Table 5. Table 5. Independent Variables in the Human Development Index Model Model B Investment (% of GDP) Government Spending (% of GDP) Openness to Trade (trade as % of GDP) Population Growth (% change) Physical Integrity Rights Index (negative freedom) Modified Empowerment Rights Index (positive freedom) Model A Investment (% of GDP) Government Spending (% of GDP) Openness to Trade (trade as % of GDP) Population Growth (% change) Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Poverty Models Following the same methodology, the relationship between positive and negative freedom and absolute poverty levels is examined in the Poverty models. The other variables for the poverty model are investment, government spending, openness, population growth and an education indicator. As before, Model A and Model B include two separate sets of freedom indicators. The limited and sporadic data on poverty for most countries in the sample prevented the use of panel data. Therefore, these models are estimated using ordinary least squares (Beck 2004, 4). Table 6 describes the independent variables used in the poverty models. 45 Table 6. Independent Variables in the Poverty Head Count Model Model A Investment (% of GDP) Government Spending (% of GDP) Openness to Trade (trade as % of GDP) Population Growth (% change) % of population over 15 that have completed Secondary Education Fiscal Freedom Index (negative freedom) Political Freedom (Political Rights Index) Model B Investment (% of GDP) Government Spending (% of GDP) Openness to Trade (trade as % of GDP) Population Growth (% change) % of population over 15 that have completed Secondary Education Physical Integrity Rights Index (negative freedom) Modified Empowerment Rights Index (positive freedom) 46 Chapter 5: Results & Discussion The regression results for the Economic Growth, Human Development and Poverty models are reported and discussed separately in the following analysis. Economic Growth Models The results from the Economic Growth models estimation indicate no statistically significant relationships between the freedom indicators and economic growth. The regression results for the two versions of the Economic Growth models are outlined in Table 7. Recall that Model A includes freedom indicators from The Heritage Foundation and Freedom House and Model B includes rights indicators from the CIRI dataset. The “t” ratios are provided in brackets under each parameter estimate5. The relationship between economic growth and positive and negative freedom is estimated using cross-section and period fixed effects on unbalanced panel data. Testing indicated that the fixed effects method was more suitable than random effects for this model6. The adjusted R-squared for both models is relatively low. The panel data used spans 14 time periods from 1996 to 2009 and 116 countries which are listed in Appendix 3. The estimated coefficients for the period and cross-section fixed effects are reported in Appendix 6. Both period and cross-section fixed effects are appropriate and jointly significant. 3Due to the variation in the scales of the independent and dependent variables it would be desirable to report standardized coefficients, however, standardized coefficients are not defined for panel data models. 6 The Likelihood Ratio test that is built into EViews 7 indicated that the presence o f fixed effects were likely. In addition, the Hausman test indicated that fixed effects is a better specification than random effects. 47 Table 7. Economic Growth Models with Country & Period Fixed Effects Economic Growth Models Model A IEF& FW Variables -0.033622 (-1.289597) 0.002547 (7.257156)*** -0.001342 (-1.006361) -0.0000495 (-0.370842) -1.154384 (-10.53712)*** 0.000793 (1.911767)* -0.0000285 (-0.157415) 0.003317 (1.613410) Intercept Investment Government Spending Openness Population Growth % of population over 15 that have completed Secondary Education Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Physical Integrity Index (negative freedom) Modified Empowerment Index (positive freedom) Adjusted R-squared Number of Observations *** Denotes a 1% level of statistical significance ** Denotes a 5% level of statistical significance * Denotes a 10% level of statistical significance 0.294288 1618 Model B CIRI Variables -0.029030 (-1.275212) 0.002555 (7.358040)*** -0.000900 (-0.669808) -0.0000531 (-0.401546) -0.982219 (-5.834102)*** 0.000826 (1.983661)** 0.001072 (0.939392) -0.000593 (-0.446302) 0.288789 1620 The results for the other included variables have mixed statistical significance. Investment exhibits the expected positive sign and is statistically significant at the 1% level. If Investment increases by 1 %, one can expect a 0.00255 % increase in per capita economic growth in Model A and a 0.00256 % increase in Model B. The education proxy is also positively related to economic growth. It is statistically significant at the 5% level in Model A and only at the 10% level in Model B. In Model A, a 1 % increase in the population over 15 that have completed Secondary Education will lead to a 0.000793 % increase in per capita 48 economic growth and in Model B, a 0.000826 % increase. Population growth is negatively related to economic growth, this is consistent with the findings of other empirical investigations (Mankiw, Romer and Weil 1992,410; Dawson 2006, 503). For a 1% change in population growth, economic growth will change by 1.15% in the opposite direction. The other two variables - government spending and openness - are not statistically significant in either model. In these models, there is no relationship between economic growth and the proxies for freedom. The freedom indicators are not statistically significant in either Model A or Model B. In addition, the freedom indicators show mixed results with both negative and positive correlations with the dependent variable. Individual components of indices have proved to be statistically non-significant in previous studies (Ayal and Karras 1998, 7; Carlsson and Lundstrom 2002, 342). As discussed earlier, property rights has been identified as a driver in some positive relationships between freedom and growth (Carlsson and Lundstrom 2001, 337; Yishay and Betancourt 2008, 1; Dawson 2003, 493-4). To evaluate the suitability of the model and to enable comparisons with previous models, the models are re-estimated adding a measure of Property Rights. The results are reported in Table 8. The coefficients for period and crosssection effects are reported in Appendix 7. The period and country fixed effects are very significant for all of the Economic Growth models and account for much of the explanatory power of the models. 49 Table 8. Economic Growth Models Including Property Rights with Country & Period Fixed Effects Economic Growth Models Model A IEF& FW Variables -0.016990 (-0.623097) 0.002492 (7.162774)*** -0.001398 (-1.038090) -0.00000216 (-0.016200) -1.115821 (-10.75218)*** 0.000932 (2.218261)** -0.000577 (-3.599967)*** -0.0000227 (-0.126619) 0.004205 (2.041421)** Intercept Investment Government Spending Openness Population Growth % of population over 15 that have completed Secondary Education Property Rights Index Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Physical Integrity Index (negative freedom) Modified Empowerment Index (positive freedom) Adjusted R-squared Number of Observations *** Denotes a 1% level of statistical significance ** Denotes a 5% level of statistical significance * Denotes a 10% level of statistical significance 0.301748 1618 Model B CIRI Variables -0.004053 (-0.164936) 0.002549 (7.379451)*** -0.001288 (-0.951516) -0.0000135 (-0.101390) -1.112940 (-10.21859)*** 0.000954 (2.248178)** -0.000536 (-3.368433)*** 0.001272 (1.118973) -0.000250 (-0.186787) 0.300215 1615 The results for the Economic Growth models including the Property Rights index from The Heritage Foundation are surprising - the coefficient for Property Rights indicates it is negatively and significantly related to economic growth. However, the magnitude of the coefficient is very small. The coefficient indicates that a 1 unit increase in the Property Rights measure is associated with a 0.000577 % decrease in economic growth. This 50 departure from previous research may be due to the use of a different time period or the use of panel data instead of cross section data. The data set includes more recent years, up to and including 2009. The latter part of this time period has exhibited weak economic growth due to the financial crisis of 2008 and the subsequent recession in areas with strong property rights - North America and Western Europe - and strong growth in areas with weak property rights. It must be noted that the direction and significance of the relationship remain substantively the same when China is omitted from the sample. The inclusion of the property rights indicator also causes the Political Rights index to become statistically significant at the 5% level. Testing indicates that the Fiscal Freedom, Physical Integrity and Modified Empowerment variables are redundant in this model and in the original model7. The only freedom indicator that exhibits a relationship with economic growth is the Political Rights Index and this is only true in the model that includes Property Rights. Aside from this tenuous relationship, the other negative and positive freedoms have no statistically significant relationship with economic growth. In particular, there is no evidence that Friedman’s assertion that if individuals have increased negative freedom, there will be increased economic growth. The adjusted R-squared for the original Model A is 0.294288 and for Model B is 0.288789. The addition of Property Rights increases the adjusted R-squared to 0.301748 for Model A and 0.300215 for Model B. This increase is quite small and indicates that the 7 The built in feature o f EViews 7 was used to test the likelihood ratio that the variables were redundant to the model. This is done by performing an F-test o f joint significance. 51 addition of Property Rights did not add significantly to the explanatory power of the models. However, testing indicates that the Property Rights measure is not a redundant variable with an F-statistic of 16.83265 for Model A and an F-statistic of 14.62448 for Model B. In addition, much of the explanatory power of the Economic Growth models is evident in the fixed effects coefficients for country (cross-section) and year (period) rather than the independent variables. This is due to the large variation that exists between countries. Human Development Models The estimates of the Human Development models reveal some statistically significant relationships between the freedom indicators and human development. The regression results are outlined in the Table 9 with t-statistics in parentheses. The human development model is estimated using ordinary least squares because of insufficient data for panel estimation. The sample consists of only 2 time periods 2000 and 2005 due to availability and comparability of data as noted in the methodology section. Two time periods are not sufficient for fixed effects modeling as there is inadequate variation. Panel data usually has at least 5 time periods and fixed effects is not generally recommended for fewer time periods (Beck 2004,4). Therefore, the model is estimated using ordinary least squares (Blumenstock 2011, 7). The countries utilized in this model are listed in Appendix 4. 52 Table 9. Human Development Model - OLS Human Development Models Intercept Investment Government Spending Openness Population Growth Model A EEF & FW Variables 0.524433 (8.849109)*** 0.002061 (2.036106)** -0.010859 (-5.896481)*** 0.000582 (2.475745)** -3.898803 (-5.117226)*** -0.000105 (-0.167228) 0.032796 (8.182039)*** Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Physical Integrity Index (negative freedom) Modified Empowerment Index (positive freedom) Adjusted R-squared 0.525545 Number of Observations 270 *** Denotes a 1% level of statistical significance ** Denotes a 5% level of statistical significance * Denotes a 10% level of statistical significance Model B CIRI Variables 0.542708 (15.34141)*** 0.002271 (2.332083)** -0.012852 (-6.607379)*** 0.000421 (2.665904)*** -4.391908 (-5.849870)*** 0.023327 (4.979925)*** 0.015051 (2.636267)*** 0.578008 274 The non-freedom variables are investment, government spending, openness and population growth. These variables are statistically significant and have the expected signs with investment and openness positively correlated with human development and government spending and population growth negatively correlated with human development. These relationships are anticipated based on the economic growth literature and theory (Barro 2003,241). In addition, government spending’s negative relationship to human development is consistent with one previous study which speculates that higher government expenditure may indicate fiscal irresponsibility (Hasan et al 2003, 23), although as noted 53 previously, the correlation may be due to higher levels of poverty leading to higher government expenditure. The positive freedom indicator, the Political Rights index, is positively related to the Human Development Index and is statistically significant at the 1% level. Both, the CERI freedom variables, the Physical Integrity index as negative freedom indicator and Modified Empowerment index as a positive freedom indicator are positively and significantly correlated to the HDI. This indicates that increases in positive freedom as represented by political rights and empowerment lead to increases in human development. If the Political Rights index improved by one unit, the HDI would improve by 0.0328 of a unit, if the Modified Empowerment index improved by one unit, the HDI would inprove by 0.0151 of a unit. The same is true for increases in negative freedom as represented by Physical Integrity Rights index. An improvement in the Physical Integrity Rights index of one unit is associated with a 0.0233 improvement in the HDI. The negative freedom indicator, the Fiscal Freedom index, is negatively related, but not statistically significant. There is a strong relationship between the proxies for positive freedom used and the Human Development index in the model as specified. The Physical Integrity index has a statistically strong, positive relationship with the development indicator, while the other negative freedom indicator, Fiscal Freedom, has no statistical significance in this model. The two indicators measure two different aspects of negative freedom - freedom from state violence and freedom from state taxation and are not significantly correlated. The mixed result leads to ambiguity regarding the relationship between negative freedom and human development. 54 Again, previous studies have found that property rights have positive, statistically significant effects on human development (Norton 1998, 243; Norton 2003, 36). In order to further test the relationship between the positive and negative freedom variables and human development and to enable comparisons to previous literature, the model is estimated with a measure of Property Rights included. The results of the models with the Property Rights index included are reported in Table 10 below. Table 10. Human Development Model Including Property Rights - OLS Human Development Models Intercept Investment Government Spending Openness Population Growth Property Rights Index Model A IEF & FW Variables 0.399005 (6.292050)*** 0.001251 (1.555154) -0.007322 (-4.820327)*** 0.000325 (1.843292)* -4.133617 (-5.850321)*** 0.003593 (7.369871)*** 0.000701 (1.178554) 0.011122 (2.276446)** Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Physical Integrity Index (negative freedom) Modified Empowerment Index (positive freedom) Adjusted R-squared 0.621709 Number of Observations 270 *** Denotes a 1% level of statistical significance ** Denotes a 5% level of statistical significance * Denotes a 10% level of statistical significance 55 Model B CIRI Variables 0.471005 (13.88295)*** 0.001168 (1.500741) -0.008289 (-5.290187)*** 0.000224 (1.611233) -4.360717 (-6.139402)*** 0.003513 (8.366418)*** 0.014840 (3.631824)*** -0.003649 (-0.606836) 0.652372 268 The addition of the Property Rights index changed the some o f the results. Investment became statistically non-significant in both model specifications. Openness became statistically non-significant in Model B. The Modified Empowerment index changed from a positive, statistically significant relationship to a negative, statistically non-significant relationship with human development. The coefficient for the Physical Integrity index maintained a positive and significant relationship with human development as it had in the original model. Although, the addition of Property Rights creates some havoc in the model, it does not increase the explanatory power of the model significantly. The models without a measure of property rights reported an adjusted R-squared of 0.525545 (Model A) and 0.5578008 (Model B). Adding Property Rights increased the adjusted R-squared to 0.621709 and 0.652372 respectively. Testing indicates that the Property Rights measure is not a redundant variable with an F-statistic of 67.85615 for Model A and an F-statistic of 71.02520 for Model B. The changes in the coefficients indicate a lack of reliability in the original model. The Political Rights index and the Modified Empowerment index lose some statistical significance in the model, when the Property Rights measure is included. The original model may be subject to omitted variable bias and the Property Rights index may be useful in explaining human development. Nonetheless, there is some evidence of a statistically significant relationship between a negative and a positive freedom indicator and human development. Keeping in mind the limitations of the model, such evidence does not provide strong support for Sen’s hypothesis that increased positive freedom leads to increased human development. 56 Poverty Models The results of the regressions for the Poverty models exhibit a strong relationship between freedom and poverty. The relationship between positive freedom and reduced absolute poverty is more evident than the relationship between negative freedom and poverty levels. The regression results are outlined in Table 11 with t-statistics in parentheses as before. As with the human development models, ordinary least squares is used to estimate the poverty models. There is inadequate data on poverty over time to use fixed effects methods (Beck 2004,4). The data is from 13 different time periods between 1996 and 2008. The countries used in the estimation are listed in Appendix 5. The non-freedom variables for the poverty models are the same as for the economic growth models: investment, government spending, openness, population growth and education. The education and openness indicators are not statistically significant in the poverty models and are likely redundant variables; however, they exhibit the appropriate sign with respect to poverty. Investment has a statistically significant negative relationship with poverty, indicating that increases in investment may lead to decreases in absolute poverty levels. A 1 % change in investment is associated with a 0.463 % change in the poverty ratio in the opposite direction for Model A and a 0.623 % change for Model B. The population growth coefficient is large, statistically significant and positively related to poverty, which is consistent with expectations that higher levels of population growth accompany higher levels of poverty. The government spending indicator is also statistically significant and positively related to poverty. As indicated previously, some have speculated that higher government spending and higher poverty levels exhibiting a positive correlation may indicate government 57 irresponsibility (Hasan et al 2003, 23) or it may indicate that higher incidences of poverty may lead to higher government expenditure as a policy response. Table 11. Poverty Models - OLS Poverty Models Intercept Investment Government Spending Openness Population Growth % of population over 15 that have completed Secondary Education Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Model A IEF& FW Variables 69.33695 (7.085748)*** -0.462916 (-3.012545)*** 0.984775 (3.875622)*** -0.046537 (-1.336040) 936.8117 (4.741623)*** -0.138548 (-0.999874) -0.280433 (-2.769485)*** -4.725556 (-7.482176)*** Physical Integrity Index (negative freedom) Modified Empowerment Index (positive freedom) Adjusted R-squared Number of Observations *** Denotes a 1% level of statistical significance ** Denotes a 5% level of statistical significance * Denotes a 10% level of statistical significance 0.502213 309 Model B CIRI Variables 44.98814 (5.740612)*** -0.623097 (-3.824484)*** 1.230007 (4.621508)*** -0.046568 (-1.220040) 1012.067 (5.073083)*** -0.130399 (-0.897216) -0.646512 (-0.963984) -4.435774 (-6.137983)*** 0.468030 311 As with the human development models, the poverty models indicate a statistically significant correlation between the positive freedom variables and absolute poverty measures, while the negative freedom variables reveal mixed results. Positive freedoms as represented by the Political Rights index and the Modified Empowerment index are statistically significant to the 1 % level, as is the negative freedom variable, the Fiscal 58 Freedom index. This means that increased political rights, empowerment and lower taxes are associated with lower poverty rates. An increase of one unit in the Political Rights index is associated with a 4.73 % decrease in the poverty ratio; an increase of one unit in the Modified Empowerment index is associated with a 4.44 % decrease in the poverty ratio; and a one unit increase in Fiscal Freedom index is associated with a 0.28 % decrease in the poverty ratio. However, the Physical Integrity index is not statistically significant at any generally accepted level although it is negatively correlated with poverty. Overall, the results indicate a strong relationship between increases in the indicators used to represent positive freedom and decreases in absolute poverty as measured by the poverty headcount ratio. There is a relationship between higher scores on Fiscal Freedom (negative freedom) and decreases in the poverty head count ratio and no relationship between the Physical Integrity index (negative freedom) and the poverty measure. As with the human development model, the two negative freedom indicators have delivered mixed results. These results are contrary to previous research regarding the relationship between freedom and poverty. Hasan et al conclude that political rights have no impact on poverty reduction; however, they do find that civil liberties do contribute to poverty reduction (Hasan et al 2003, 23). Recall, that the civil liberties index was not utilized in this study as it includes a property rights component. As with previous models, property rights in the form of the Property Rights index from The Heritage Foundation is included in the poverty models to evaluate the original model specification and for comparisons with previous research. The results of this regression are reported in Table 12. 59 Table 12. Poverty Models Including Property Rights - OLS Model A IEF&FW Variables 70.78444 (6.779413)*** -0.454633 (-2.912160)*** 0.951019 (3.496748)*** -0.046151 (-1.318607) 936.2030 (4.750960)*** -0.139076 (-1.001059) -0.044731 (-0.555312) -0.284038 (-2.786914)*** -4.529888 (-5.850369)*** Poverty Models Intercept Investment Government Spending Openness Population Growth % of population over 15 that have completed Secondary Education Property Rights Index Fiscal Freedom Index (negative freedom) Political Rights Index (positive freedom) Physical Integrity Index (negative freedom) Modified Empowerment Index (positive freedom) Adjusted R-squared Number of Observations *** Denotes a 1% level of statistical significance ** Denotes a 5% level of statistical significance * Denotes a 10% level of statistical significance 0.501095 309 Model B CIRI Variables 46.66065 (5.471913)*** -0.588191 (-3.512280)*** 1.151810 (4.119599)*** -0.048993 (-1.271943) 1009.236 (4.970581)*** -0.133217 (-0.908582) -0.074522 (-0.957401) -0.508448 (-0.762839) -4.105795 (-4.957342)*** 0.462526 309 When property rights are included in the poverty models, the signs and statistical significance of the coefficients of the non-freedom and freedom variables remain the same as in the original model specification. The Property Rights index exhibits a negative correlation with the poverty head count, but the coefficient is statistically non-significant. This result is inconsistent with Norton who found a significant correlation between property rights and the components of the Human Poverty Index (1998, 240-1). This may be because his study did 60 not include similar indicators of freedom and only looked at the effects of property rights. As indicated by the adjusted R-squared values the addition of a measure of Property Rights did not increase the explanatory power of the model, but actually decreased it. For Model A the adjusted R-squared decreased from 0.502213 to 0.501095 and for Model B it decreased from 0.468030 to 0.462526. Testing indicates that the property rights measure is a redundant variable in both Model A (F-statistic of 0.324959) and Model B (F-statistic of 0.832007). The results of this regression indicate that increases in positive freedom decrease poverty levels. Increases in the proxies for negative freedom also decrease poverty levels, but only the fiscal freedom variable is statistically significant. The addition of a measure of Property Rights does not significantly impact the model. 61 Chapter 6: Conclusion To examine the effects of freedom on economic growth, human development and poverty, Berlin’s definitions of negative and positive freedom are employed. Berlin’s dichotomy is used to extract measurements of freedom as to what degree the individual is free from coercion (negative freedom) or to what degree the individual is free to govern himself (positive freedom). To identify the mechanisms through which negative and positive freedom might affect economic development the theories of two economists, Friedman and Sen, are consulted. Friedman’s theories and followers predict that increases in negative freedom will lead to increases in economic growth. Whereas, Sen’s theories anticipate that increases in positive freedom will correlate with increases in human development. It is also hypothesized that increases in either freedom will lead to decreases in poverty. Using an augmented Solow growth model these predictions are tested using regression analysis. Friedman’s argument that individual negative freedom is necessary for economic growth and that too much positive freedom could put economic growth into jeopardy is not supported by the regression results of this model. The regression results do not reveal any relationship between either negative or positive freedom and economic growth in an augmented Solow growth model using country and period fixed effects on panel data. This holds true when property rights are included in the model. Surprisingly, property rights exhibit a statistically significant negative relationship with economic growth. Most previous research indicates a positive relationship between property rights and economic growth (Carlsson and Lundstrom 2002, 337). This contrary result may be a result of the use o f a later time period than previous research had available. The period fixed effects reported in 62 Appendix 7 reveal negative coefficients for 1998, 1999, 2001, 2002,2008 and 2009. The negative coefficients coincide with global economic slowdowns. Alternatively, it is possible that the type of data used influenced the results as previous research generally uses crosssection data rather than panel data analysis (Blume and Voigt 2007, 527; Dawson 2006,497). As noted earlier, the country and period fixed effects coefficients are jointly significant and much of the variation in economic growth is explained by the country and the period coefficients. The ideas of Sen that increased positive freedom would have a positive effect on human development were also not supported by this empirical study. Initially, a positive relationship between freedom and human development seemed evident, especially between positive freedom and development. However, the inclusion of property rights in the model changed the significance and / or sign of three of the four proxies for freedom, calling into question the reliability and appropriateness of the model. This could be due to a complex relationship between freedom and human development as proposed by Sen (1999, 10). As noted previously, I attempt to identify the effects of freedom on human development, not the effect of development on freedom or the linkages between freedoms. This approach has proven inadequate to obtain valuable results. There does appear to be a favourable relationship between freedom and poverty with increasing freedom associated with reductions in poverty. Only one of the proxies for negative freedom is statistically significant, but both proxies for positive freedom are statistically significant, and all the freedom variables exhibit a negative relationship with poverty levels. The inclusion of property rights did not change the size or direction of the coefficients of any of the freedom or other variables. In fact, the property rights variable 63 proved to be statistically non-significant. Increased freedom, especially positive freedom, exerts a negative effect on poverty. This study, as do other studies of similar nature, is subject to substantial limitations. The first limitation is data availability, specifically, the limited overlap between available measures of freedom and economic outcomes over time. The second is the necessarily contestable process of proxy selection. Although the criteria for choosing the proxies for negative and positive freedom, is transparent and defendable, the choices are not beyond debate or question. The degree to which the variables can be said to measure what they were chosen to measure determines the significance of the results for theory and policy. Third, the mechanisms through which freedom is theorized to affect economic outcomes have varying degrees of theoretical support. In light of the results of this study and the limitations of it, future research may be able to access more extensive panel data to test the models chosen, especially for the relationship with human development. The inclusion of more time periods may lead to increased support for the model specification. Alternatively, it is possible that the model specification is not appropriate and another model would be a better fit to the data. For instance, an indirect relationship between positive and negative freedom and economic growth may be revealed through investment or total factor productivity. Perhaps an investigation into the effects o f human development on levels of freedom would be more fruitful than the approach attempted here. The choice of proxies for negative and positive freedom is necessarily constrained by available indices. A thorough review of available indices was performed to select the variables used in this study; however, variables that are more clearly representative of freedom of the individual would be very valuable in testing these relationships. The classification of freedom as negative or positive rather than economic or political and the restriction of the indicators used to only those components of the indices that meet the definition of positive and negative freedom as outlined by Berlin could be one reason why the freedom indicators and the property rights measurement do not have the same effects on growth in this study as they do in previous studies. In addition, the use of fixed effects modeling on panel data may have impacted the results. Finally the use of a later time period may also contribute to the contrary results - as the characteristics of the global economy shift so too might the types of freedoms and institutions that drive economic growth change. The promotion of freedom for individuals as a necessary condition for economic growth is not justified by the empirical evidence. Of course, freedom has intrinsic value and is usually pursued by people for its intrinsic value rather than its instrumental value. It is the policy makers that imply that more negative freedom leads to growth and more positive freedom may jeopardize that growth thus putting greater emphasis on the instrumental value of freedom. Those supposed consequences of increased or decreased freedom are not supported by the economic growth or human development models estimated in this thesis. Freedom, especially positive freedom, is associated with a reduced incidence of poverty. This is not unexpected, as it has been hypothesized that empowered people will not support regimes that do not address issues of poverty. In the development process, reduction in the incidence of absolute poverty appears to come hand in hand with increased positive freedom and to some extent negative freedom. Overall, there is no winner between positive and negative freedom when evaluated according to their effects on economic outcomes. In regards to economic growth, neither type of freedom appears to be an effective driver or determinant. The relationship is not 65 clear with human development. The strong relationship between increased positive freedom and reduced poverty levels is unambiguous, but there is also evidence of a relationship between increased negative freedom and reduced poverty, indicating both are desirable in the development process when the goal is poverty reduction. The pursuit of freedom of both types - positive and negative - continues regardless of the instrumental value of freedom. However, it is encouraging that both positive and negative freedom are positively related to the reduction of poverty. Previous research categorizing freedom as economic or political concludes that increased economic freedom is a more effective method of improving economic growth or well-being than increasing political freedom or promoting democracy (Barro 1996, 24; Stroup 2007, 63). Although this study categorizes freedom in a more explicit and precise way than these previous studies, it does generally show that freedom is not likely to have any effect on economic growth or human development directly but is likely to be beneficial in the reduction of absolute poverty. The alternate categorization of freedoms as either negative or positive assists in understanding the effects of freedoms enjoyed by the individual on the economic outcomes of economic growth, human development and poverty. This study attempts to explicitly outline the assumptions underlying what is meant by freedom by clearly defining what is negative and positive freedom. This definition of freedom is not a unversally accepted concept and the concept of freedom continues to be debated. This approach could have an advantage over the previous approach of using indices of economic and political freedom and allowing the components of the indices to implicitly define what is meant by freedom. 66 The debate regarding the intrinsic value of freedom and the instrumental value of freedom continues. Hopefully, it will continue to explicitly define freedom using stated frames of reference and assumptions, although the frames of reference and assumptions may differ from the ones employed in this study. That is our generation’s task - to make these words, these rights, these values — of Life, and Liberty, and the Pursuit of Happiness —real for every American. Being true to our founding documents does not require us to agree on every contour of life; it does not mean we will all define liberty in exactly the same way, or follow the same precise path to happiness. Progress does not compel us to settle centuries-long debates about the role of government for all time —but it does require us to act in our time (Obama, Inauguration 2013 2013). 67 Bibliography Acemoglu, Daron, and Simon Johnson. "Unbundling Institutions." Journal o f Political Economy 113, no. 5 (2005): 949-995. Ayal, Eliezer B., and Georgios Karras. "Components of Economic Freedom and Growth: An Empirical Study." Journal o f Developing Areas (Western Illinois University) 32, no. 3 (Spring 1998): 327-338. Barro, Robert J. "Determinants of Economic Growth in a Panel of Countries." Anals o f Economics and Finance 4 (2003): 231-274. —. "Democracy and growth." Journal o f Economic Growth 1 (March 1996): 1-27. —. "Economic growth in a cross section of countries." The Quarterly Journal o Economics 106, no. 2 (May 1991): 407-443. Barro-Lee Educational Attainment Dataset. 2012. http://www.barrolee.com/ (accessed November 15, 2011). Beck, Nathaniel. Panel Data (Especially Modeling "Effects"). New York, NY, Spring 2004. Berggren, Niclas. "The Benefits of Economic Freedom: A Survey." The Independent Review VIII (Fall 2003): 193-210. Berlin, Isaiah. "Two Concepts of Liberty." In Social and Political Philosophy, by John Arthur and William H. Shaw, edited by John Arthur and William H. Shaw, 298-305. Englewood Cliffs, New Jersey: Prentice-Hall Inc., 1992. Blume, Lorenz, and Stefan Voigt. "The Economic Effects of Human Rights." KYKLOS 60, no. 4 (2007): 509-538. Blumenstock, Joshua. "Fixed Effects Models.” Econ 174/274: Global Poverty andImpack Evaluation (UC Berkeley). Spring 2011. http://www.jblumenstock.com/courses/econl74/FEModels.pdf (accessed November 24, 2012). 68 Boesveld, Sarah. Canada rises to Top Five in world economic freedom rankings as U.S. plummets to 18th. National Post, September 18, 2012. Business Environmental Risk Intelligence. n.d. http://www.beri.com/ (accessed January 27, 2013). Byars, John, Robert McCormick, and Bruce Yandle. Economic Freedom in the America's 50 States. Economics, Clemson University, Clemson: The Center for Policy and Legal Studies, 1999. Carlsson, Fredrik, and Susannna Lundstrom. "Economic freedom and growth: Decomposing the effects." Public Choice 112 (2002): 335-344. Carter, Ian. "Positive and Negative Liberty." Vers. Spring 2012. The Stanford Encyclopedia o f Philosophy. March 5,2012. http://plato.stanford.edu/archives/spr2012/entries/liberty-positve-negative (accessed May 4, 2012). Chong, Alberto, and Cesar Calderon. "Causality and Feedback between Institutional Measures and Economic Growth." Economics and Politics 12, no. 1 (March 2000): 69-81. Cingranelli, David L., and David L. Richards. The Cingranelli-Richards (CIRI) Human Rights Dataset. Vers. 2011.12.09. 2012. http://www.humanrightsdata.org (accessed January 25,2012). Cole, Julio H. "The Contribution of Economic Freedom to World Economic Growth, 198099." Cato Journal 23, no. 2 (Fall 2003): 189-198. Dawson, John W. "Causality in the freedom - growth relationship." European Journal o f Political Economy 19 (2003): 479-495. —. "Regulation, Investment and Growth Across Countries." Cato Journal 26, no. 3 (Fall 2006): 489-509. 69 de Haan, Jakob, and Clemens L.J. Siermann. "Further Evidence on the Relationship between Economic Freedom and Economic Growth." Public Choice 95, no. 3/4 (June 1998): 363-380. —. "New Evidence on the Relationship between Democracy and Economic Growth." Public Choice 86, no. 1/2 (January 1996): 175-198. de Haan, Jakob, and Jan-Egbert Sturm. "On the Relationship between Economic Freedom and Economic Growth." March 1999. de Haan, Jakob, Susanna Lundstom, and Jan-Egbert Sturm. "Market-Orientated Institutions and Policies and Economic Growth: A Critical Survey." Journal o f Economic Surverys 20, no. 2 (2006): 151-191. Doucouliagos, Chris, and Mehmet Ali Ulubasoglu. "Economic freedom and economic growth: Does specification make a difference?" European Journal o f Political Economy 22 (2006): 60-81. Esposto, Alfredo G., and Peter A. Zaleski. "Economic Freedom and the Quality of Life: An Empirical Analysis." Constitutional Political Economy 10, no. 2 (1999): 185-197. Farr, W. Ken, Richard A. Lord, and J. Larry Wolfenbarger. "Economic Freedom, Political Freedom, and Economic Well-Being: A Causality Analysis." Cato Jouranl 18, no. 2 (Fall 1998): 247-62. Freedom House. Freedom in the World. 2012. http://www.freedomhouse.org/reporttypes/freedom-world (accessed December 1,2012). Freedom House Index - Methodology. 2006. http://www.freedomhouse.org/template.cfrn?page=35&year=2006 (accessed October 11, 2010 ). "Freedom in the World." Freedom House. 2012. http://www.freedomhouse.org/sites/default/files/FIW%202012%20Booklet_0.pdf (accessed November 26,2012). 70 Friedman, Milton. Capitalism and Freedom. 1982. Chicago, EL: The University of Chicago Press, 1962. Garrett, Thomas A., and Russell M. Rhine. "Economic Freedom and Employment Growth in U.S. States." Federal Reserve Bank o f St. Louis Review 93, no. 1 (January / February 2011): 1-18. Gastil, R.D., and L.M. Wright. "The state of the world: Political and economic freedom." In Freedom, democracy and economic welfare, edited by M. Walker. Vancouver, BC: The Fraser Institute, 1988. Gastil, R.D. Freedom in the World. New York, NY: Freedom House, 1973-1989. Gaus, Gerald F. "Property, Rights and Freedom." Social Philosophy and Policy 11, no. 02 (1994): 209-240. Goldsmith, Arthur A. "Democracy, Property Rights and Economic Growth." The Journal o f Development Studies 32, no. 2 (December 1995): 157-174. —. "Economic Rights and Government in Developing Countries: Cross-National Evidence on Growth and Development." Studies in Comparative International Development 32, no. 2 (Summer 1997): 29-44. Grubel, Herbert G. "Economic Freedom and Human Welfare." Cato Journal 18, no. 2 (Fall 1998): 287-304. Gwartney, J.D., W.E. Block, and R.A. Lawson. "Economic Freedom of the World: 19751995." 1996. http://oldfraser.lexi.net/publications/books/econ_free95/(accessed January 20,2013). Gwartney, James, and Robert Lawson. Economic Freedom o f the World. The Fraser Institute, 1997. —. "The concept and measurement of economic freedom." European Journal o f Political Economy 19 (2003): 405-430. 71 Gwartney, James, Robert Lawson, and Joshua Hall. 2012 Economic Freedom Dataset. Fraser Institute, 2012. Hanke, Steve H., and Stephen J.K. Walters. "Economic Freedom, Prosperity, and Equality: A Survey." Cato Journal 17, no. 2 (Fall 1997): 117-146. Hasan, Rana, M. G. Quibria, and Yangseon Kim. "Poverty and Economic Freedom: Evidence from Cross-Country Data." East-West Center Working Papers 60 (September 2003). Heckelman, Jac C. "Economic Freedom and Economic Growth: A Short-Run Causal Investigation." Journal o f Applied Economics III, no. 1 (May 2000): 71-91. Heckelman, Jac C., and Micahel D. Stroup. "A comparison of aggregation methods for measures of economic freedom." European Journal o f Political Economy 21 (2005): 953-966. —. "Which Economic Freedoms Contribute to Growth?" KYKLOS 53, no. 4 (2000): 527-544. Helliwell, John F. "Empirical Linkages between Democracy and Economic Growth." British Journal ofPolitcal Science 24, no. 2 (April 1994): 225-248. Heritage Foundation. 2010. http://www.heritage.org/index/FAQ.aspx (accessed October 27, 2010). Heston, Alan, Robert Summers, and Bettina Aten. PWT 7.0. Vers. 7.0. May 2011. www.pwt.econ.upenn.edu/php_site/pwt_index.plq) (accessed November 30, 2011). Hoeffler, Anke E. "The Augmented Solow Growth Model and the African Growth Debate." Explaining African Economic Growth Performance Conference Series. Center for International Development at Harvard University, 2000. Human Development Index. 2011. http://hdr.undp.org/en/statistics/hdi/ (accessed November 16,2012). 72 Human Development Report 2011 - Readers Guide. 2011. http://hdr.imdp.org/en/media/HDR_201 l_EN_Readers.pdf (accessed September 25, 2012). Judson, R. A., and A. L. Owen. "Estimating Dynamic Panel Data Models: A Guide for Macroeconomists." Economic Letters 65 (1999): 9-15. Justesen, Mogens K. "The effect of economic freedom on growth revisited: New evidence on causality from a panel of countries 1970-1999." European Journal o f Political Economy 24 (2008): 642-660. Kaufinann, Daniel, Aart Kraay, and Massimo Mastruzzi. Governance Matters VII: Aggregate and Individual Governance Indicators 1996-2007. Policy Research Working Paper, The World Devleopment Research Group Macroeconomics and Growth Team & World Bank Institute Global Governance Program, 2008. . Wolrdwide Governance Indicators. n.d. http://info.worldbank.org/govemance/wgi/index.asp (accessed November 15, 2011). Kaun, David E. "What does Economic Freedom Buy?" Journal o f Socio-Economics 31 (2002): 371-390. Kiviet, J.F. "On Bias, Inconsistency, and Efficiency of Various Estimators in Dynamic Panel Data Models." Journal o f Econometrics 68 (1995): 53-78. Lin, Justin Yifu and Jeffrey B. Nugent. Institutions and Economic Development. Vol. 3A, in Handbook of Economic Development edited by J.Behrman and T.N. Srinivasan. Amsterdam: North Holland, 1995. Lundstrom, Susanna. "The Effect of Democracy on Different Categories of Economic Freedom." Edited by 967-980. European Journal o f Political Economy 21 (2005). Mankiw, N. Gregory, David Romer, and David N. Weil. "A Contribution to the Empirics of Economic Growth." The Quarterly Journal o f Economics, May 1992: 408-437. 73 McCain, John. "Remarks by Senator McCain to the Republican National Convention." The Washington Post. August 30,2004. http://www.washingtonpost.com/wpdyn/articles/A47237-2004Aug30.html (accessed November 25,2012). Mello, Luiz de, and Randa Sab. "Government Spending, Rights and Civil Liberties." International Review o f Law and Economics 22 (2002): 257-276. Mitchell, Daniel J. The Impact o f Government Spending on Economic Growth. March 15, 2005. http://www.heritage.org/research/reports/2005/03/the-impact-of-govemmentspending-on-economic-growth (accessed January 19, 2013). Norton, Seth W. "Economic Institutions and Human Well-Being: A Cross-National Analysis." Eastern Economic Journal 29, no. 1 (Winter 2003): 23-40. Norton, Seth W. "Poverty, Property Rights and Human Well-Being: A Cross-National Study." Cato Journal 18, no. 2 (Fall 1998): 233-245. Obama, Barack. "Barak Obama's 2012 victory speech." National Post. November 7, 2012. http://news.nationalpost.com/2012/11/07/barack-obama-victory-speech-201-full-textand-video/ (accessed November 25,2012). —. "Inauguration 2013." The Wall Street Journal. January 21, 2013. http://stream.wsj.com/story/inauguration-2013/SS-2-137898/SS-2149774/?mod=wsj_streaming_inauguration-2013 (accessed January 28, 2013). —. The White House Blog. September 6,2010. http://www.whitehouse.gOv/blog/2010/09/06/president-obama-labor-day-fightamericas-workers-continues (accessed January 13, 2013). Pitlik, Hans. "The Path of Liberalization." KYKLOS 55 (2002): 57-80. Prendergast, Renee. "Development and freedom." Journal o f Economic Studies 31, no. 1 (2004): 39-56. Puddington, Arch. "Freedom House." Freedom in the World. 2012. http://www.freedomhouse.org (accessed August 16, 2012). 74 Robeyns, Ingrid. ""The Capability Approach"." The Stanford Encyclopedia o f Philosophy. Edited by Edward N. Zalta. Summer 2011. http://plato.stanford.edu/cgibin/encyclopedia/archinfo.cgi? entry=capability-approach (accessed January 20, 2013). Rodrik, Dani. "Institutions for High-Quality Growth: What They Are and How to Acquire Them." Studies in Comparative International Development 35, no. 3 (Fall 2000): 331. Roll, Richard, and John R. Talbott. "Political Freedom, Economic Liberty and Prosperity." Journal o f Democracy 14, no. 3 (July 2003): 75-89. Romney, Mitt. "Republican Convention 2008: Mitt Romney's Speech." The Patriot Post: Voice o f Essential Liberty. The Republican National Convention. 2008. http://patriotpost.us/pages/183 (accessed November 25,2012). Safaei, Jalil. "Is Democracy Good For Health." International Journal of Health Services 36, no. 4 (2006): 767-786. Scully, Gerald W. Constitutional Environments and Economic Growth. Princeton, N.J.: Princeton University Press, 1992. —. "The Institutional Framework and Economic Development." Journal o f Political Economy 96, no. 3 (1988): 652-662. Scully, Gerald W., and Daniel J. Slottje. "Ranking economic liberty across countries." Public Choice 69(1991): 121-152. Sen, Amartya. Development as Freedom. Toronto, Ontario: Random House of Canada Limited, 1999. Spindler, Zane A. "Liberty and development: A further empirical perspective." Public Choice 69, no. 2 (February 1991): 197-210. 75 Stroup, Michael D. "Economic Freedom, Democracy, and the Quality of Life." World Development 35, no. 1 (2007): 52-66. The Heritage Foundation. Index o f Economic Freedom. 2008. http://www.heritage.org/Index/download (accessed November 15, 2011). Total Economy Database. January 2012. www.conference-board.org/data/economydatabase (accessed April 08, 2012). Vanssay, Xavier De, and Z. A. Spindler. "Freedom and Growth: Do Constitutions Matter?" Public Choice 78 (1994): 359-372. Vega-Gordillo, Manuel, and Jose L. Alvarez-Arce. "Economic Growth and Freedom: A Causality Study." Cato Journal 23, no. 2 (Fall 2003): 199-215. Weede, Erich. "Economic Freedom and Development: New Calculations and Interpretations." Cato Journal 26, no. 3 (2006): 511-524. Weede, Erich. "Human Rights, Limited Government and Capitalism." Cato Journal 28, no. 1 (Winter 2008). World Development Indicators (WDI) & Global Development Finance (GDF'). 09 18, 2012. http://databank.worldbank.org/ddp/home.do (accessed November 1, 2012). Wu, Wenbo, and Otto A. Davis. "The two freedoms, economic growth and development: An empirical study." Public Choice 100 (1999): 39-64. Yishay, Ben Ariel, and Roger R. Betancourt. "Civil Liberties and Economic Development." econ-server.umd.edu. August 2008. http://econserver.umd.edu/~betancourt/development/BenYishay%20and%20Betancourt%20200 8%20-%20Civil%20Liberties%20Paper%20-%20Draft%2016.pdf (accessed December 15,2009). 76 Appendix 1 Summary of Indices The empirical literature on the effects of political and economic freedoms utilizes a variety of indices or indicators that have evolved over time. The earliest index cited by studies in this literature review was originally conceived as a measurement of political freedom but was soon combined with measures of economic freedom. This combined GastilWright index was a forerunner to some of the later economic and political freedom indices (Gastil and Wright 1988). Raymond Gastil’s index of political freedoms and civil liberties was first published in 1978 (Gastil, 1978). His index assigned a number between 1 and 7 to denote the level of political freedom where 1 indicated a higher level of freedom (Freedom House Index Methodology 2006). To measure economic freedom, Lindsay Wright assigned a composite rating (high, medium high, medium, medium low or low) to 165 countries based on four economic freedoms: (1) freedom of property, (2) freedom of association, (3) freedom of movement and (4) freedom of information (Spindler 1991,198). It should be noted that, freedom indices published currently do not generally categorize the last 3 indicators as economic freedoms. Wright’s economic freedom index supplemented Gastil’s index and continued to be published by Freedom House. In 1989, the compilation of Gastil’s index was taken over by a team of analysts and continues to be published today by Freedom House. The Fraser Institute also published a joint effort by Gastil and Wright - an early effort at quantifying economic and political freedom by the organization that now publishes Economic Freedom of the World annually (Gastil & Wright, 1988). Another early index was compiled explicitly to measure the effects of economic liberty on various outcomes by Scully 77 and Slottje (Scully and Slottje 1991). These early indices did not continue to be published over time. In 1996 Economic Freedom in the World (EFW) published its first data set of seventeen measures of economic freedom in four categories for 103 countries (Gwartney, Block and Lawson 1996). This provided researchers a new data set to test their hypothesis regarding the relationship between freedom and growth. After the publication of the EFW, investigation into the relationship between economic freedom and economic outcomes, such as growth, increased. The 2012 Economic Freedom of the World covers 42 measures in 5 broad categories: size of government; legal structure and property rights; access to sound money; freedom to trade internationally; and regulation of credit, labour and business (Gwartney, Lawson and Hall 2012, v). The Heritage Foundation’s Index of Economic Freedom (IEF) was first released in 1995 and is jointly published with The Wall Street Journal. They measure ten components of economic freedom, using a scale from 0 to 100, where 100 indicates the maximum freedom. The ten components of economic freedom used by the Heritage Foundation are: business, trade, fiscal freedom, government spending, monetary, investment, financial freedom, property rights, freedom from corruption and labor freedom (Heritage Foundation 2010, 2). Worldwide Governance Indicators (WGI) is another index that measures perceptions of governance in 215 countries. The indicators measure perceptions regarding: voice and accountability; political stability and absence of violence or terrorism, government effectiveness, regulatory quality, rule of law, and control of corruption (Kaufmann et al, 2008,1). The index reports on perceptions, rather than more objective measures. The authors cite three reasons for this methodology: (1) people base their decisions on their 78 perceptions; (2) there are few alternatives to perception data and (3) even if objective or factbased data is available, the data may reflect what is ‘on the books’ rather than the that reality that exists ‘on the ground’ (Kaufinann et al, 2008, 3). Freedom House publishes Freedom in the World. The survey first appeared in book form under this title in 1978 and was originally produced by Raymond Gastil and is now produced by a team of analysts (Freedom House 2012). Countries are evaluated on political rights and civil liberties. The political rights indicator includes an evaluation of the following categories: electoral process, political pluralism and participation and functioning of government. The civil liberties indicator evaluates the categories of: freedom of expression and belief, associational and organizational rights, rule of law and personal autonomy and individual rights. The individual rights component includes property rights which as noted is also included in economic freedom indices and its role as a freedom is contested.. (Freedom in the World 2012, 33-5). The CIRI Human Rights Dataset reports on government respect for human rights. The dataset includes information on physical integrity rights, civil liberties, workers’ rights and rights of women. The authors indicate that the rights that are included in the dataset are chosen based on whether there is reliable and systematically available information across time and space. (Cingranelli and Richards 2012). The Business Environmental Risk Intelligence is a private source of data for country level business risk analysis. The data is available on a subscription basis. Information on each country includes measures of political risk, operations risk and remittance and repatriation factor ratings. This information is based on “qualitative judgments and candid assessments” (Business Environmental Risk Intelligence n.d.). 79 An index measuring economic freedom in the United States is used by Kaun (Kaun 2002, 379). This index was published by The Center for Policy and Legal Studies at Clemson University. The index utilizes over 100 measures of economic freedom “spanning government spending, regulation, welfare, school choice, taxation and the judicial system” (Byars, McCormick and Yandle 1999). Table 13 summarizes various indices of freedom as identified above8. 8 Each index reports a varying number o f countries year to year depending on data availability. In addition, the components of the indices have been subject to change. 80 Table 13. List of Freedom Indices and their Sources Index Gastil’s Index Wright’s Index Gastil-Wright Index Freedom in the World Scully - Slottje Economic Freedom Indicators Economic Freedom of the World Index of Economic Freedom World Governance Indicators The CingranelliRichards (CIRI) Human Rights Dataset Business Environmental Risk Intelligence Economic Freedom in America’s SO States Title of Publication Freedom in the World: Political Rights and Civil Liberties “A Comparative Survey of Economic Freedoms.” In R.Gastil, Freedom in the World: Political Rights and Civil Liberties, 1982. “The State o f the World: Political and Economic Freedom.” In M.A. Walker (ed.) Freedom, Democracy and Economic Welfare Freedom in the World Publisher Freedom House Author Gastil, R.D. Years Published 1972-1988 Freedom House Wright, L.M. 1982 The Fraser Institute Gastil, R.D. and Wright, L.M. 1988 Constitutional Environments and Economic Growth World Survey o f Economic Freedom: 1995-1996 Princeton University Press Freedom House Economic Freedom of the World (EFW) The Fraser Institute Index of Economic Freedom (IEF) Heritage Foundation World Governance Indicators (WGI) The International Bank for Reconstruction and Development / The World Bank CIRI Human Rights Data Project Business Risk Reports and Historical Ratings Research Package (HRRP) Economic Freedom in America’s 50 States Freedom House Scully, G.W. Messick, Kaku Kimura Gwamey, Lawson & Block Johnson and Sheehy 81 1996 1996 1995-Present The World Bank Group 1996 —Present David L. Cingranelli and David L. Richards 1981-2009 Business Environmental Risk Intelligence The Center for Policy and Legal Studies, Clemson University 1989-Present: Team of Analysts 1992 1980-2010 Byars, McCormick & Yandle 1999 Appendix 2 Summary of All Variables & Sources Indeoendent Variable Economic Growth Human Development Index Poverty Head Count living on less than $2 / day (% of population) Investment (% of GDP) Government Spending (% of GDP) Openness (% of GDP) Population (% change) % of population over 15 that have completed Secondary Education Property Rights Fiscal Freedom Index Political Rights Index Physical Integrity Index Modified Empowerment Index Source Penn World Tables 7.0 Human Development Report 2011 World Development Indicators Penn World Tables 7.0 Penn World Tables 7.0 Penn World Tables 7.0 World Development Indicators Barro-Lee Educational Attainment Data Set The Heritage Foundation The Heritage Foundation Freedom House CIRI Human Rights Data Project CIRI Human Rights Data Project 82 Appendix 3 Economic Growth Models - Countries Included in the Sample Albania Algeria Argentina Armenia Australia Austria Bahrain Bangladesh Belgium Belize Benin Bolivia Botswana Brazil Bulgaria Cambodia Cameroon Canada Chile China Version 1 Colombia Congo, Republic of Costa Rica Cote d'Ivoire Croatia Cuba Cyprus Czech Republic Denmark Dominican Republic Ecuador Egypt El Salvador Estonia Fiji Finland France Gabon Gambia, The Germany Ghana Greece Guatemala Guyana Haiti Honduras Hungary Iceland India Indonesia Iran Ireland Israel Italy Jamaica Japan Jordan Kenya Korea, Republic of Kuwait Laos Latvia Lesotho Libya Lithuania Malawi Malaysia Mali Mexico Moldova Mongolia Morocco Mozambique Namibia Nepal Netherlands New Zealand Nicaragua Niger Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Russia 83 Rwanda Saudi Arabia Senegal Slovak Republic Slovenia South Africa Spain Sri Lanka Swaziland Sweden Switzerland Syria Tanzania Thailand Trinidad &Tobago Tunisia Turkey Uganda Ukraine United Kingdom United States Uruguay Venezuela Vietnam Yemen Zambia Zimbabwe Appendix 4 Human Development Models - Countries Included in the Sample Albania Algeria Argentina Armenia Australia Austria Bahrain Bangladesh Belgium Belize Benin Bolivia Botswana Brazil Bulgaria Cambodia Cameroon Canada Chile China Version 1 Colombia Congo, Republic of Costa Rica Cote d'Ivoire Croatia Cuba Cyprus Czech Republic Denmark Korea, Republic of Kuwait Laos Latvia Lesotho Libya Lithuania Malawi Malaysia Mali Mexico Moldova Mongolia Morocco Mozambique Namibia Nepal Netherlands New Zealand Nicaragua Niger Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Dominican Republic Ecuador Egypt El Salvador Estonia Fiji Finland France Gabon Gambia, The Germany Ghana Greece Guatemala Guyana Haiti Honduras Hungary Iceland India Indonesia Iran Ireland Israel Italy Jamaica Japan Jordan Kenya 84 Romania Russia Rwanda Saudi Arabia Senegal Slovak Republic Slovenia South Africa Spain Sri Lanka Swaziland Sweden Switzerland Syria Tanzania Thailand Trinidad &Tobago Tunisia Turkey Uganda Ukraine United Kingdom United States Uruguay Venezuela Vietnam Yemen Zambia Zimbabwe Appendix 5 Poverty Models —Countries Included in the Sample Albania Algeria Angola Argentina Armenia Azerbaijan Bangladesh Belarus Belize Benin Bolivia Bosnia and Herzegovina Brazil Bulgaria Burkina Faso Burundi Cambodia Cameroon Cape Verde Central African Republic Chad Chile China Version 1 Colombia Congo, Dem. Rep. Congo, Republic of Kenya Kyrgyzstan Laos Latvia Lesotho Lithuania Macedonia Madagascar Malawi Malaysia Mali Mauritania Mexico Moldova Mongolia Morocco Mozambique Nepal Nicaragua Niger Nigeria Pakistan Panama Papua New Guinea Paraguay Peru Philippines Poland Costa Rica Cote d'Ivoire Croatia Czech Republic Djibouti Dominican Republic Ecuador Egypt El Salvador Estonia Ethiopia Gabon Gambia, The Georgia Ghana Guatemala Guinea Guinea-Bissau Guyana Haiti Honduras Hungary India Indonesia Iran Jamaica Jordan Kazakhstan 85 Romania Russia Rwanda Senegal Sierra Leone Slovak Republic Slovenia South Africa Sri Lanka Suriname Swaziland Tajikistan Tanzania Thailand Togo Tunisia Turkey Turkmenistan Uganda Ukraine Uruguay Uzbekistan Venezuela Vietnam Yemen Zambia Appendix 6 Fixed Effects for Economic Growth Models Period Coefficients Model A 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Model B 0.008684 0.009882 -0.006439 -0.006348 0.006557 -0.008537 -0.003612 0.006531 0.017223 0.015664 0.011827 0.007637 -0.009682 -0.049387 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 86 0.008974 0.009142 -0.006162 -0.006429 0.006816 -0.008546 -0.003880 0.006767 0.017912 0.015743 0.012027 0.007662 -0.009896 -0.050131 Cross-Section Coefficients Model A Albania Algeria Argentina Armenia Australia Austria Bahrain Bangladesh Belgium Belize Benin Bolivia Botswana Brazil Bulgaria Cambodia Cameroon Canada Chile China Version 1 Colombia Congo, Republic of Costa Rica Cote d'Ivoire Croatia Cuba Cyprus Czech Republic Denmark Dominican Republic Ecuador Egypt El Salvador Estonia Fiji Finland France 0.004162 -0.008835 -0.010354 0.013446 -0.034709 -0.044430 -0.009886 0.000925 -0.044092 0.002962 0.007005 0.009731 -0.038602 0.002215 0.000883 0.070108 0.028191 -0.027018 -0.020977 0.015482 -0.000627 Gabon Gambia, The Germany Ghana Greece Guatemala Guyana Haiti Honduras Hungary Iceland India Indonesia Iran Ireland Israel Italy Jamaica Japan Jordan Kenya Korea, Republic of Kuwait Laos Latvia Lesotho Libya Lithuania Malawi Malaysia Mali Mexico Moldova Mongolia Morocco Mozambique Namibia Nepal 0.028560 0.011079 0.042270 -0.009192 0.069832 -0.016240 -0.045720 -0.037740 0.032889 -0.014869 0.038135 0.012879 -0.027174 -0.019644 -0.018379 -0.032220 87 -0.024097 0.064128 -0.048647 -0.000981 -0.024046 0.025655 0.005994 0.024101 0.013234 -0.031239 -0.031698 0.038492 -0.006454 0.000296 -0.003518 -0.012679 -0.052619 -0.028668 -0.063258 -0.046409 0.031668 -0.063395 -0.002889 0.061616 -0.009503 -0.028605 0.061633 -0.004562 0.005448 -0.016150 0.046415 -0.007809 -0.011819 -0.004966 -0.016316 0.072251 -0.017056 0.006411 Netherlands New Zealand Nicaragua Niger Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Russia Rwanda Saudi Arabia Senegal Slovak Republic Slovenia South Africa Spain Sri Lanka Swaziland Sweden Switzerland Syria Tanzania Thailand Trinidad &Tobago Tunisia Turkey Uganda Ukraine United Kingdom United States Uruguay Venezuela Vietnam Yemen Zambia Zimbabwe -0.019124 -0.012064 -0.000543 0.049078 -0.037694 0.019186 0.018246 -0.006634 -0.012851 0.012225 0.008305 -0.040686 -0.019010 0.013721 0.115246 -0.005698 0.010751 -0.001505 -0.059704 -0.012407 -0.027639 88 -0.016504 0.021943 -0.036512 -0.055353 0.051979 0.065761 -0.021823 0.050137 -0.003893 0.011080 0.078914 -0.008111 0.000327 -0.033943 -0.006726 -0.000965 0.029131 0.027677 0.060328 -0.016162 Model B Albania Algeria Argentina Armenia Australia Austria Bahrain Bangladesh Belgium Belize Benin Bolivia Botswana Brazil Bulgaria Cambodia Cameroon Canada Chile China Version 1 Colombia Congo, Republic of Costa Rica Cote d'Ivoire Croatia Cuba Cyprus Czech Republic Denmark Dominican Republic Ecuador Egypt El Salvador Estonia Fiji Finland France Gabon Gambia, The Germany Ghana Greece Guatemala Guyana Haiti Honduras Hungary Iceland India Indonesia Iran Ireland Israel Italy Jamaica Japan Jordan Kenya Korea, Republic of Kuwait Laos Latvia Lesotho Libya Lithuania Malawi Malaysia Mali Mexico Moldova Mongolia Morocco Mozambique Namibia Nepal Netherlands New Zealand Nicaragua Niger 0.004905 -0.020354 -0.004341 0.009331 -0.028353 -0.035710 -0.026475 0.003792 -0.034955 0.005051 0.005780 0.010467 -0.039620 0.005358 0.007685 0.058438 0.017750 -0.019332 -0.014154 0.000695 0.003308 0.019112 0.013042 0.034164 -0.007371 0.050324 -0.012045 -0.039464 -0.029077 0.036890 -0.013101 0.027824 0.015320 -0.018573 -0.027361 -0.009976 -0.023528 -0.033975 0.050620 89 -0.040634 0.001363 -0.016330 0.022678 0.008493 0.014983 0.008566 -0.023315 -0.023335 0.045305 -0.005285 -0.005465 0.003084 -0.005427 -0.043680 -0.027186 -0.054809 -0.055797 0.027331 -0.056274 -0.012154 0.043232 -0.003580 -0.022841 0.045879 0.001364 -0.000053 -0.022047 0.043049 -0.000072 -0.008802 -0.005405 -0.021216 0.072189 -0.010321 0.004419 -0.012786 -0.005381 -0.006446 0.037955 Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Russia Rwanda Saudi Arabia Senegal Slovak Republic Slovenia South Africa Spain Sri Lanka Swaziland -0.030341 0.012038 0.020008 -0.007686 -0.009252 0.018540 0.017175 -0.031021 -0.015232 0.012020 0.100805 -0.020755 0.011704 0.005023 -0.051506 -0.003091 -0.019437 -0.013826 0.007974 Sweden Switzerland Syria Tanzania Thailand Trinidad &Tobago Tunisia Turkey Uganda Ukraine United Kingdom United States Uruguay Venezuela Vietnam Yemen Zambia Zimbabwe 90 -0.028314 -0.047160 0.037639 0.056307 -0.020068 0.054654 -0.011673 0.013635 0.068453 -0.006167 0.008914 -0.026216 0.001911 0.001925 0.014857 0.019567 0.054187 -0.029605 Appendix 7 Fixed Effects for Economic Growth Models Including Property Rights Period Coefficients Model A 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Model B 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 0.012529 0.013916 -0.003206 -0.003084 0.008860 -0.007511 -0.004101 0.005766 0.015653 0.013281 0.009057 0.004378 -0.012931 -0.052609 91 0.011641 0.013001 -0.003294 -0.003563 0.008857 -0.007671 -0.004358 0.006268 0.016532 0.013632 0.009546 0.004711 -0.012696 -0.052606 Cross-Section Coefficients Model A Albania Algeria Argentina Armenia Australia Austria Bahrain Bangladesh Belgium Belize Benin Bolivia Botswana Brazil Bulgaria Cambodia Cameroon Canada Chile China Version 1 Colombia Congo, Republic of Costa Rica Cote d'Ivoire Croatia Cuba Cyprus Czech Republic Denmark Dominican Republic Ecuador Egypt El Salvador Estonia Fiji Finland France -0.006207 -0.014764 -0.013565 0.006545 -0.014304 -0.025876 -0.010134 -0.010152 -0.027759 0.001557 -0.000130 -0.001237 -0.029546 0.002876 -0.007194 0.059164 0.021229 -0.007429 -0.001444 0.006115 -0.005535 Gabon Gambia, The Germany Ghana Greece Guatemala Guyana Haiti Honduras Hungary Iceland India Indonesia Iran Ireland Israel Italy Jamaica Japan Jordan Kenya Korea, Republic of Kuwait Laos Latvia Lesotho Libya Lithuania Malawi Malaysia Mali Mexico Moldova Mongolia Morocco Mozambique Namibia Nepal 0.013102 0.007412 0.034623 -0.021842 0.051496 -0.005594 -0.043034 -0.015767 0.020803 -0.023348 0.038997 0.013941 -0.021392 -0.030689 0.003571 -0.022065 92 -0.023309 0.066246 -0.029436 -0.003834 -0.020937 0.019655 0.002093 0.002674 0.005322 -0.026776 -0.010438 0.041239 -0.013384 -0.020021 0.012321 -0.003639 -0.045706 -0.027660 -0.047661 -0.043820 0.032433 -0.049655 0.006242 0.044495 -0.014179 -0.031833 0.041171 -0.012751 0.007047 -0.019858 0.044113 -0.008451 -0.016940 -0.012199 -0.017016 0.061949 -0.015571 0.000835 Netherlands New Zealand Nicaragua Niger Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Russia Rwanda Saudi Arabia Senegal Slovak Republic Slovenia South Africa Spain -0.001400 0.009994 -0.011759 0.043935 -0.018314 0.015067 0.004908 -0.020283 -0.018546 0.007896 0.014475 -0.029160 -0.032898 0.007338 0.105460 0.001022 0.013638 -0.007537 -0.064578 -0.013894 -0.016489 Sri Lanka Swaziland Sweden Switzerland Syria Tanzania Thailand Trinidad &Tobago Tunisia Turkey Uganda Ukraine United Kingdom United States Uruguay Venezuela Vietnam Yemen Zambia Zimbabwe 93 -0.020282 0.025148 -0.022561 -0.038653 0.044922 0.062008 -0.015535 0.062422 -0.001914 0.017647 0.081787 -0.022618 0.024292 -0.013021 0.002090 -0.008731 0.009061 0.018352 0.060820 -0.029213 Model B Albania Algeria Argentina Armenia Australia Austria Bahrain Bangladesh Belgium Belize Benin Bolivia Botswana Brazil Bulgaria Cambodia Cameroon Canada Chile China Version 1 Colombia Congo, Republic of Costa Rica Cote d'Ivoire Croatia Cuba Cyprus Czech Republic Denmark Dominican Republic Ecuador Egypt El Salvador Estonia Fiji Finland France Gabon Gambia, The Germany Ghana Greece Guatemala Guyana Haiti Honduras Hungary Iceland India Indonesia Iran Ireland Israel Italy Jamaica Japan Jordan Kenya Korea, Republic of Kuwait Laos Latvia Lesotho Libya Lithuania Malawi Malaysia Mali Mexico Moldova Mongolia Morocco Mozambique Namibia Nepal Netherlands New Zealand Nicaragua Niger -0.007586 -0.026009 -0.008381 -0.000279 -0.010380 -0.021307 -0.022718 -0.006554 -0.023363 0.008862 0.002187 0.002772 -0.029551 0.007852 -0.000381 0.050761 0.010320 -0.002500 0.003113 -0.008523 -0.001461 0.005533 0.014025 0.026435 -0.021414 0.035095 -0.000158 -0.038315 -0.012236 0.027495 -0.020267 0.028937 0.018012 -0.015594 -0.038355 0.007767 -0.016476 -0.034866 0.055838 94 -0.025804 0.000645 -0.014438 0.019721 0.005177 -0.002803 0.006362 -0.021005 -0.006986 0.051403 -0.010552 -0.025070 0.017760 0.006455 -0.039842 -0.024395 -0.043109 -0.051857 0.029570 -0.045538 0.000217 0.028442 -0.008830 -0.027764 0.028268 -0.006511 0.005372 -0.024788 0.045527 -0.001312 -0.015917 -0.009482 -0.024512 0.064744 -0.011750 0.001509 0.002176 0.014238 -0.012197 0.037461 Norway Pakistan Panama Paraguay Peru Philippines Poland Portugal Romania Russia Rwanda Saudi Arabia Senegal Slovak Republic Slovenia South Africa Spain Sri Lanka Swaziland -0.014924 0.010047 0.011499 -0.019992 -0.015335 0.015903 0.020697 -0.023254 -0.029510 0.003261 0.092956 -0.015231 0.015543 -0.001330 -0.059511 -0.004644 -0.010310 -0.017778 0.010400 Sweden Switzerland Syria Tanzania Thailand Trinidad &Tobago Tunisia Turkey Uganda Ukraine United Kingdom United States Uruguay Venezuela Vietnam Yemen Zambia Zimbabwe 95 -0.018577 -0.033824 0.031772 0.057543 -0.015430 0.064580 -0.013832 0.019569 0.076292 -0.022897 0.029952 -0.007227 0.008899 -0.005521 -0.005563 0.013173 0.056931 -0.039876