|
Article Excerpt
I. INTRODUCTION
The contention that "institutions rule," as Rodrick, Subramanian, and Trebbi (2004) put it, has become a core belief among both scholars and practitioners of economic development. This consensus results from a wide body of evidence suggesting that a country's overall economic performance is affected by its institutional framework. (1) It has been observed that ailing institutions are associated with slower growth, for instance by Mauro (1995) or Knack and Keefer (1995), lower total factor productivity (TFP), by Hall and Jones (1999), lower TFP growth, by Olson, Sarna, and Svamy (2000), or lower per capita income, by Acemoglu, Johnson, and Robinson (2001). Thanks to different instrumental variables, Hall and Jones (1999) and Acemoglu, Johnson, and Robinson (2001) convincingly argued that the causality ran from institutions to economic performance.
The debate now focuses on the channels through which institutions affect overall economic performance. Capital accumulation first appeared as the most likely culprit, and this is why the institutional determinants of investment have been closely scrutinized. Accordingly, the quality of institutions has been found to affect total investment, among others by Mauro (1995) or Knack and Keefer (1995), public investment, by Mauro (1998), and foreign direct investment, by Wei (2000). These results explain the impact of governance on capital accumulation but only incidentally address its impact on productivity. This is where integration in world trade comes to the fore due to its observed relationship to TFP, for instance documented by Edwards (1998).
Strikingly, however, the impact of institutions on the trade of goods has received little attention so far. This is puzzling in view of the key role that integration in world trade plays in development and the fact that many developing countries either remain on the periphery of world trade or are stuck with exporting primary products. This observation is allegedly one of the main puzzles of international economics, since relative factor endowments should result in substantial North-South trade. To date, the main explanation of that puzzle has been to blame developing countries' restrictive trade policies. Thus, Sachs and Warner (1995a) found that economies that adjust more slowly from primary-intensive to manufactured-intensive exports were those whose trade was less liberalized.
However, recent research suggests that for countries to fully benefit from openness strategies, institutions might be crucial. Anderson (2005) thus suggested that the risk of predation and imperfect enforcement of contracts impairs foreign trade because it increases both the costs and the risks of trading abroad. Anderson and Marcouiller (2002) accordingly observed that bad opaque public policies and an ineffective legal system reduce the volume of trade. Dollar and Kraay (2003) also reported a positive correlation between openness and the rule of law, although they argued that the causality between the two variables may be bidirectional. In a similar vein, Giavazzi and Tabellini (2005) observed that political liberalizations were associated with trade liberalizations. The former moreover tended to lead rather than follow the latter, which suggests that political liberalizations at least partly cause trade liberalizations.
However, trade is not homogenous, and the volume of trade is only one dimension of a country's integration in the world economy. Moreover, all exports are not equivalent in terms of development and growth. The development economics literature suggests that manufactured exports are more likely to lead to development than nonmanufactured exports. The empirical support for this is based on the observation that countries with more diversified exports seem to do better and that growth tends to be positively correlated with growth in manufactured production and exports. To highlight the reason for such a relationship, some authors stress the importance of "forward and backward linkages" in creating higher positive externalities coming from manufacturing than from natural resource sectors (see, e.g., Hirschman, 1958; Seers, 1964). Matsuyama (1992) underlined the importance of learning by doing in manufacturing and its implication for the rate of human capital accumulation. Other arguments hinge on the Dutch disease. Natural resources have more volatile world prices than manufacturing, and this induces greater uncertainty for primary commodity producers, which extends to other sectors. Uncertainty is known to be detrimental to factor accumulation and hence to growth (see, e.g., Sachs and Warner, 2001).
There is another reason to distinguish manufactured and nonmanufactured exports. The impact of institutional quality on exports of primary products is likely to differ from its effect on manufactured exports. Endowments of natural resources create natural rents that are usually controlled by the administration and generate corrupt competition over their distribution as Ades and Di Tella (1999) suggested. In such a context, exports of nonmanufactured goods may be positively rather than negatively associated with the lack of quality of institutions. Distinguishing exports of manufactured goods from exports of nonmanufactured goods may therefore lead to a better specification of exports regressions, hence more reliable estimates of the impact of institutions on trade.
Moreover, the current debate over the relationship between institutions and trade remains fairly vague on what is meant by institutions. Governance is indeed a multifaceted concept. It ranges from the rule of law to the degree of democracy of the country. Unsurprisingly, there exists a wide choice of indicators, each aiming to assess one dimension of the wider phenomenon. Unfortunately, most indicators have been developed separately, thus producing a lacunal picture of countries' overall institutional environment. As a result, if each dimension of governance has been independently studied, attempts to compare the impact of different facets of the institutional framework remain scarce. The lack of comparability of the various dimensions of the institutional framework was remedied by Kaufmann, Kraay, and Zoido-Lobaton (1999a) who synthesized existing indicators to provide a distinct and consistent assessment of the main dimensions of governance. They thus constructed six governance indexes devoted to six dimensions of the institutional framework. A second contribution of this paper is to take advantage of the set of indicators of Kaufmann, Kraay, and Zoido-Lobaton (1999a) to compare the effects of several dimensions of institutional quality on trade flows. The results should allow identifying the institutional reforms that may be the most promising in economic terms.
Using a panel of around 60 countries over 1990-2000, this paper examines separately the impact of each dimension of the institutional framework on total exports of goods, manufactured goods exports and nonmanufactured goods exports. In a first stage, we use a simple least squares approach relating a given category of exports to a given dimension of the institutional framework plus a set of control variables. In a second stage, we use a two-stage least squares (2SLS) approach to take account of potential endogeneity of the institutional variables and get a clearer idea of their causal effect.
In the first stage, we find a positive impact of the quality of institutions on the manufactured exports ratio but no significant impact on the ratio of nonmanufactured exports. When using the 2SLS approach, we find a positive impact of the exogenous component of institutions on exports of manufactures. In particular, an improvement in the quality of the regulatory framework is significantly and positively associated with exports of manufactures. In contrast, we observe a negative impact of the exogenous component of institutions on the exports of nonmanufactured goods. In particular, nonmanufactured exports appear negatively affected by a better accountability of political leaders and higher respect of the rule of law. This opposite correlation of the two kinds of exports with institutions results in the total exports ratio being not significantly related to institutions.
To reach those results, the rest of the paper is organized as follows: the next section discusses why different categories of exports may be differently affected by the...
|
