اثر نهادها بر رشد اقتصادی: یک تجزیه و تحلیل جهانی بر اساس برآورد پانل پویا GMM

This study examines how institutional indicators influence economic growth in a theoretical framework proposed by North (1981). Thirty-one indicators each covering 84 countries over a span of 5 years have been used to extract factors based on principal component analysis. Factors based on these indicators are classified as institutional and policy rents, political rents and risk-reducing technologies. These institutional factors are then used in a formal growth model employing panel OLS and GMM-based estimation methodologies. The findings suggest that favorable institutions positively affect economic growth. This study also shows that for a developing country the institutional and policy rent is more important than other two indices that curb political rents and those that reduce transaction risks. This study also highlights the positive complementarities between index of political rents and index of risk-reducing technologies.

The theoretical and applied issues of the relationship between institutions and economic growth have thoroughly been examined both in the developed and in developing countries. This study revisits the issue and tests the role of institutions in economic growth using new methodology and three newly constructed sets of institutions. North and Thomas (1973) argued that institutions affect economic growth by influencing property rights, incentive structures and transaction costs. Rodrik (2000) explained the pivotal role of various non-market institutions in creating complete and contingent markets. Institutions contribute to growth and to development by reducing the risk of doing business, thus directing resources toward innovation rather than protecting property rights or earning predatory rents. Empirical literature has identified numerous institutions that influence economic growth, including governance, law enforcement, justice, regulations, tax administration, and institutions that manage monetary and fiscal policies.2Moers (1999) found that a broader measure of institutions has the strongest effect on growth. Acemoglu et al., 2001 and Acemoglu et al., 2002, and Acemoglu and Johnson (2005) show that qualities of institutions have a stronger effect on long term growth than in the short run. Méon and Weill (2006) and Olson et al. (1998) found evidence that institutional factors influence total factor productivity and that those countries with better institutions exhibit higher productivity. A few studies have used instrumental variable techniques to indicate evidence of causality running from institutions to economic performance (Olson et al., 1998, Acemoglu et al., 2001 and Rodrik et al., 2004). Institutions are multi-dimensional. Commercial organizations including Political Risk Service (PRS) and Business Environment Risk Intelligence (BERI), and non-commercial organizations including the World Economic Forum (WEF), Global Integrity, Freedom House, Fraser Institute (FI) (Gwartney and Lawson, 2008), Heritage Foundation (Miller et al., 2009), and Bertelsmann, Marshall et al. (POLITY) have been developing diverse indicators to measure institutional quality across countries. Some of these indicators have been used in the empirical literature.3 Their aim was to select those indicators that explore the dimensionalities as proposed by North (1981), Dawson (1998), and Rodrik (2000). However, each of them was deficient in this area (Kormendi and Meguire, 1985, Knack, 2002 and Kaufmann and Kraay, 2008; and Van de Walle, 2005). There are statistical limitations to the use of diverse indicators in a single regression framework as the strong correlation among indicators creates a risk of multicollinearity (Moers, 1999). This might be the reason why studies such as Méon and Sekkat (2004) and Acemoglu et al. (2001) have used different indicators separately in different equations. Alternatively, various studies have attempted to aggregate different indicators by means of simple averages. Francisco and Antonio (2004), Hall and Jones (1999), and Knack and Keefer (1995) standardized and averaged five indicators of PRS and four similar indicators from BERI in two indices, while Mauro (1995) averaged eight indicators from BI (now EIU) in two indices. An advantage of aggregating indicators is that it cancels out source specific measurement error4 (Laura and Knack 2010; Knack and Nick 2000; Knack, 2002 and Mauro, 1995). The literature acknowledges that as compared to individuals, the aggregate indicators are more reliable as these organize and summarize very large and disparate information in very concise way (Catrinescu et al., 2009, Kaufmann and Kraay 2008, and Van de Walle, 2005]. However, Knack and Nick (2000) and Kaufmann and Kraay (2008) emphasized that the gain in reliability from aggregation comes at the expense of a loss in specificity and conceptual precision. World Governance Indicators (WGI) (Kaufmann et al., 2008) is a widely used indicator to measure the quality of institutions across countries.5 It attempted to cover a wide range of institutions by categorizing six “governance indicators.” Its validity, however, is being questioned in different studies on conceptual as well as empirical grounds. First, the governance indicators are poorly identified, multi-faceted and lack proper theoretical basis as pointed out by Martin and Petra (2011), Laura and Knack (2010), Arndt and Oman (2006), Johnston (2007), and Thomas (2007). These studies show evidence of overlap between six different categories of WGI and seem to be tautological, which makes it difficult to delineate them separately. Secondly, these studies also show that the WGI encounters difficulties on the empirical front, as most of its categories have an extremely high degree of inter-correlation, exceeding 95 per cent in some cases. Thirdly, the WGI indices are affected by lack of dimensionality. Martin and Petra (2011), and Laura and Knack, 2008 and Laura and Knack, 2010 indicate that there is only a single dominant factor in the WGI indicators, which shows that the indices in fact measure the same basic concept. That could be reason why various empirical studies such as Al-Marhubi (2004), Bjornksov (2006), Easterly (2002), and Easterly and Levine (2002) averaged all six WGI indices in their analysis. Finally, the WGI aggregated indicators failed to depict political institutions that are crucial to this analysis. This point is highlighted in a recent empirical growth study by Glaeser et al. (2004), who conclude that ICRG and WGI both fail to include political constraints whereas POLITY indicators are narrowly political; in sum each of these indicators is an incomplete and imperfect proxy for institutions. The present study addresses all of the above-mentioned shortcomings by constructing three indices to measure institutional qualities. This study performed exploratory factor analysis on 31 diverse institutional indicators from different data sources. Three dimensions of institutions were identified, namely the factor of Institutional and Policy Rents (RiiF1), the Factor of Political Rents (RpiF2) and the factor of Risk-reducing Technologies (SiiF3). The RiiF1 assesses institutions’ ability to limit rent-seeking opportunities that divert innovation and resources from productive avenues, whereas RpiF2 focuses on political competition and participation. The SiiF3 measures the quality of institutions that reduce transactional risk through proper enforcement of property rights. Weak risk-reducing institutions increase transaction cost as people divert their resources from productive activities to private arrangements. RiiF1 and RpiF2 represent two dimensions of institutions—risk-reducing and anti-rent-seeking, which are theoretically motivated by Douglass North's two theories of the state—a “contract theory” and a “predatory theory.” Three factors are then aggregated into the Index of Institutionalized Social Technologies (IIST) which measures the overall quality of institutions. The paper is organized as follows. Section 1 introduces the paper and the relevant literature. Section 2 explains the methodology and rationale for the indices. Section 3 describes the details of indices and their analysis. Section 4 provides regression estimates. Section 5 gives conclusions and policy recommendations.

Institutions undoubtedly are important for economic growth but earlier studies have not exposed which institutions matter the most and how they affect growth. This study investigated the same issue by using larger datasets and more robust techniques. It presented an Index of Institutionalized Social Technologies for the period 2002–2006 along with its sub-indices (Institutional and policy rents, political rents and risk reducing institutions), as measures of institutions. These indices captured institutions’ effects on growth in a formal model along with variables such as human capital, savings, and trade openness. It also factored in initial conditions to measure signs of convergence. Overall results of dynamic panel difference estimation and panel OLS estimations suggest that institutions exert a large and positive influence on economic growth. We found evidence that institutions of Institutional and policy rents (RiiF1) substantially influence long-run economic growth, whereas institutions that reduce risk and inhibit political rent (SiiF3 and RpiF2) were not statistically significant in some cases. A similar conclusion is reached by Acemoglu and Johnson (2005). They argue that in the absence of formal risk-reducing institutions, informal arrangements develop to provide protection in their place. As in earlier times, when formal institutions like courts and police did not exist or were ineffective, people resorted to living in groups where contracts are honored through informal pressure and risk of expulsion from the group. Hence, their rights are secured in other ways. Otherwise, protection from rent-seeking behavior would depend only upon the behavior of the State toward its people. If government's behavior promotes corruption, inefficiency, or predatory rents for privileged classes then all those people who do not have a level playing field cannot enter into similar contracts, and this would largely diminish long-run growth prospects. When all institutional measures were combined in one index, the effects on growth were more pronounced, which revealed a high degree of complementarity among institutions generally and between the institutions that protect property rights and those that inhibit political rent seeking, particularly. As political institutions improve, the measures such as improvement in law and order conditions would yield immediate results in the form of economic development. Other control variables showed that human capital, savings, and international trade have significant influence on growth as theory predicts. The results also confirm the conditional convergence predicted in modern theories of growth. From the policy perspective the emphasis should be given to the strengthening of all forms of institutions. Elimination of corruption and bureaucratic inefficiencies and provision of competitive market systems may promote economic prosperity but relying only on risk-reducing institutions for that purpose may not be very fruitful. Nonetheless, enforcing law and order and strengthening justice systems would be effective only if supported by strengthening political rights, competition and civil liberties.