New Institutional Economics has revived the important role of institutions
 on economic growth. North (1990) was a pioneering work. Institutions are defined as ‘the humanly devised constraints that structure human interaction. They are made up of formal constraints (for example, rules,
 laws, constitutions), informal constraints (for example, norms of behaviour, conventions, self-imposed codes of conduct), and their enforcement characteristics’ (North 1994, p. 360). Formal institutions are constraints
 sanctioned by state power if individuals violate them, while informal institutions are self-imposing constraints. According to North, of primary importance to economic performance is the economic institutions that
 determine transaction costs and influence the incentive structure in society such as the structure of property rights and the presence and perfection of markets. 
 There are now various empirical studies on the effect of institutions on economic growth. Most studies used crosscountry regressions to determine the effect of institutional quality on economic growth. Knack and Keefer (1995) was a pioneering work. Four important institutional variables were proposed by Knack and Keefer (1995): protection of property rights, rule of law, corruption and bureaucratic quality. Such
 data were compiled from International Country Risk Guide (ICRG) data, published by the U.S.-based Political Risk Services Group, and from Business Environment Risk Intelligence (BERI), based in Switzerland. The ICRG index includes protection of property rights (expropriation risk and repudiation of contracts by government), rule of law, corruption, and bureaucratic quality. The BERI index includes contract enforceability, nationalisation potential, bureaucratic delays and infrastructure quality. Knack and Keefer run a regression vfor 97 countries in the period 1974-89. The explanatory variables include institutional quality (ICRG or BERI), initial per capita GDP, initial human capital, average annual government consumption share/GDP, distortion index (absolute value of deviation of investment price level), the number of revolutions and coups per year and the number of political assassinations per year per million population in the period 1974-89. To avoid possible simultaneity between growth and institutional quality, the authors chose the initial value of the institutional
 indices rather than the average for the whole period. The earliest release of BERI was 1974 and that of ICRG 1982. The scale for BERI was from 0 to 4 and for ICRG from 0 to 10 (the higher the better). The findings indicated that the ICRG index was positive and highly significant across the specifications. The BERI index was positive and significant for most specifications. Mauro (1995) used a different dataset of institutions from Business International (later incorporated into the Economist Intelligence Unit). His institutional variables included corruption and bureaucratic efficiency (including corruption, efficiency of the judiciary system, and bureaucratic red tape). The data were collected for the period 1980-83. The dependent variable
 was average per capita GDP growth during 1960-85. The explanatory variables included initial per capita income in 1960, population
 growth, primary education in 1960, government expenditure share, revolutions and coups, assassinations, political instability, two distortion indices (absolute value of deviation of investment price level and its standard deviation), dummies for regions, and Mauro’s corruption index or
 bureaucratic efficiency index. The finding was that both low bureaucratic efficiency and high corruption exerted strong and negative effects on growth. Their effects were statistically significant. Other significant studies include Sachs and Warner (1997a, 1997b), Barro (1998), Brunetti et al. (1997), Kaufman et al. (1999b), Aron (2000). Their findings in general indicate
 positive effects of institutional quality on economic growth.
 This paper is aimed to explore a different but relevant relationship, i.e.,
 the question is how institutions affect on efficiency of investment. The efficiency of investment is defined as the incremental capital-output ratio (ICOR). The ICOR measures the additional amount of capital required to produce an additional unit of output. The reciprocal of ICOR measures
 the productivity of additional capital (Gillis et al. 1992). The efficiency of investment is vital to growth because the level of investment alone cannot fully explain growth performance across countries. It is noteworthy that some countries can achieve a fairly high investment rate, but only slow growth. For example, during the period 1961-85, Argentina, Jamaica and
 Zambia achieved an investment/GDP rate as high as that of Taiwan, Malaysia and Thailand, but could only achieve a growth rate less than a third of the latter group. The main hypothesis of this paper is that quality of institutions has positive effect on investment efficiency.
Read full abstract