Coordination Failures and Government Policy: Evidence From Emerging Countries

Rodrik (JIE 1996) argues for the use of government policy in cases where emerging countries are stuck in a low wage equilibrium because of a coordination failure. His coordination failure arises between downstream producers and their upstream intermediate suppliers when the production of high tech final goods requires a special set of differentiated intermediate goods. To qualify, intermediate inputs need to be produced under increasing returns to scale and nontradeable, the latter necessitating domestic production. In addition, a threshold number of varieties of the intermediate goods must be used before high tech production becomes viable. Then because the marginal returns associated with changes in individual intermediate output are markedly below the return that can be realized when a threshold number of intermediate goods are produced in concert, the expectation that a sub-optimal number of varieties will be forthcoming is sufficient to discourage entry into intermediate production and so prevent a high tech sector from becoming viable. The result is that a country’s high tech sector may not take off even when the country has the requisite skills and levels of human capital. In such cases, skilled labor will earn abnormally low returns and net social returns can be realized by affecting coordinated action (through government). Rodrik argues that this scenario plagues many transition countries, especially those characterized by high levels of human capital relative to their endowed level of physical capital. We test two hypotheses from Rodrik's paper: Hypothesis 1: A coordination failure among intermediate producers of inputs into the high tech sector leads to a low wage equilibrium. Hypothesis 2: Government policy is needed to solve the coordination failure among intermediate producers of inputs into the high tech sector and can move a country from a low to high wage equilibrium. Tests of these hypotheses are performed on data from the transition experiences of two sets of Eastern European, African and Asian countries. The tests form part of the growing empirical literature underlying the modeling of long run growth and shorter term transition processes. Such empirical knowledge is needed to better inform policy debate and to suggest directions of promise for future theoretical research. Our empirical tests take a more immediate step by beginning to test the general theory of threshold externalities (e.g. Azariades and Drazen, QJE 1990). 4 COORDINATION FAILURES AND GOVERNMENT POLICY: EVIDENCE FROM EMERGING COUNTRIES