加拿大essay：公共机构的质量与经济发展研究

Quality Of Public Institutions And Economic Growth Economics Essay
 

The role of institutions both public and private, formal and informal has become a subject of detailed examination relatively recently. Once proposed by D. North (North, 1990) the idea of a strong correlation between economic growth and institutional indicators has been tested and developed in numerous heterogeneous studies. The primary goal of this part of the present paper is to make a profound research of what has already been done in the this area in order to have an idea of the estimates obtained by other authors, study the various problems inherent to institution identification, data collection and use, as well as research methodology and results interpretation.

 

How does the institutional framework affect economic growth? According to North, "Third World countries are poor because the institutional constraints define a set of payoffs to political/economic activity that do not encourage productive activity”. Those payoffs (transaction costs induced by a high level of corruption, for example) affect negatively all economic agents, including individuals, political and economic organizations, educational and social bodies, who, in response, reduce their activity (investment, production, consumption) thus provoking diminishing growth. Integrated into the basic Solow’s growth theory (Solow, 1956), North’s approach helps to determine and evaluate the possible channels though which the economic growth can be influenced by institutional quality and social capital.

 

Basing on the numerous papers reviewed, we can assert that researchers often use one of two types of models: structural models or reduced-form models. Both types have their pros and cons. Thus, structural models are easier to interpret and they are more helpful in determining direct and indirect effects of institutional quality and its changes but one can face a problem of endogeneity while using them. Reduced-form models are, vice versa, endogeneity free since all the variables in the right hand side of the equation are exogeneous, but the estimated coefficients are really hard to interpret. The additional problem, connected to the model-type choice lies in the fact that reduced-form models are not always transformable to structural models. A profound discussion on this topic can be found in (Deaton and Miller, 1996).

 

There are different ways to broadly introduce institutions into the growth model, and the appropriate one is chosen on the basis of the available data and purposes of the study. While the purpose of the study varies from paper to paper, all the data on the public institutions and their quality can be classified into 5 groups, listed below (Aron, 2000).

 

Institutional quality measures (subjective indicators concerning corruption, quality of bureaucracy, contract enforceability etc, based on the rankings of local experts, observers, entrepreneurs and political risk services).

 

Social capital measures (objective and subjective indicators concerning civil rights and freedoms)

 

Social characteristics (objective measures of ethnic tension, social development and capability etc)

 

Political characteristics (objective measures of democracy, the and duration of political regime, etc)

 

Political instability ( objective measures of propensity of government change, sociopolitical instability, duration of civil wars etc)

 

The first two of them are generally used as separate regressors representing the quality of institutions itself, like in (Knack and Keefer, 1995), (Knack and Keefer, 1997), (Knack and Keefer, 1977) (Barro, 1996), (Clague and others, 1996), (Hassan and Sarna, 1996), (Knack, 1996), (Lane and Tornell, 1996), (Sachs and Warner, 1997), (Mauro, 1995), (Helliwell, 1996). The last three are often included into models as proxies for institutions, property rights in particular (see (Clague and others, 1996), (Barro, 1996a), (Barro, 1996b).

 

While modeling institutional influence on economic growth one must keep in mind several problems intrinsic to the qualitative data. The first and the most evident of them is the problem of endogeneity. It’s not obvious whether institutional quality influences economic growth, or, on the contrary, growth rates determine the quality of public institutions. There are different ways to resolve this problem in different ways, for example, using one indicator describing the whole period (instead of one for one point of time), like in (Knack and Keefer, 1995). The second serious problem with the institutional data is ordinality. In other words, the regression estimates suggest that a change in institutional measure from 3 to 2 is equal to a change from 9 to 8, which is not undoubtedly. In order to include the diminishing returns of institutional quality (or any other kind of returns if supposed) the indexes can be turned from ordinal into cardinal either though linear or nonlinear transformation (see (Barro, 1996a)).

 

The third difficulty that most of the researcher face while constructing a single quality index out of multiple components is right weighting. One should bring certain evidence to add or average the components of the aggregate index. The most common technique that is used in this case is factor analysis (see (Temple and Johnson, 1998)). One more concern that is to be kept in mind is the high correlation of different measures of institutional quality that takes place quite often. The problem of multicollinearity which can follow must be taken into consideration.

 

To conclude the literature review, we would like to list some of the findings from the papers examining the effect of institutional quality on the economic growth.

 

Early studies on the quality of institutions and economic growth ((Kormendi and Meguire, 1985) and (Scully, 1988) used the cross-section regression and indices of political and civil rights. Having solved the ordinality problem, the authors of these papers figured out an indirect effect on growth through investment, but due to the endogeneity problem no causal relation was stated.

 

Knack and Keefer (Knack and Keefer, 1997) found that trust and civic norms really cause economic growth, and this conclusion is robust to changes in specification. But when the investment is added to the equation those variables become insignificant, most probably because the investment variable captures those two effects in full.

 

Mauro (Mauro, 1995) examined the problem with the help of 3 equations describing the same process from different points of view: investment equation, a reduced-form growth equation and an augmented Solow growth model. The estimation results showed that efficient bureaucracy is significant in the investment regression (thus, for economic growth as well); in the second equation bureaucracy appeared to be robustly significant, while in Solow regression it is barely significant. Surprisingly, corruption turned out to be insignificant for all three equations. This unusual finding, as well as insignificance of efficient bureaucracy in the Solow model, may be the result of the problems inherent to qualitative data, mentioned above.

 

In his work (Helliwell, 1996b) the author managed to resolve the endogeneity problem. The variable describing institutional quality was insignificant, though investment was not included as a separate regressor, i.e. couldn’t be the factor that explains most of growth rate variance as it usually does.

 

Knack and Keefer ((Knack and Keefer, 1995), (Knack, 1996)) analyzed two institutional indices (property rights and security of contract) and found out that endogeneous institutional indices (taken from the previous periods) are significant in the investment equations during 1974-1989 and 1960-1989, though the absence of any statistical proof makes one hesitate about the reliability of this result. Nevertheless, it supports the general idea of a strong relation between institutions and growth.#p#分页标题#e#

 

Lane and Tornell (Lane and Tornell, 1996) focused their analysis on the variables that are usually omitted in growth regressions – natural resources and institutions – in order to find an explanation to the higher growth rates in resource-poor countries. The reduced-form regression is run in the paper, and on the basis of the estimation the authors asserted that a coexistence of weak institutions and manufacture concentration affect the growth rates negatively and significantly.

 

Barro (Barro, 1996a) based his research on a 1960-1990 panel, using the Solow growth model, endogeneous rule of law index and the ordinal institutional index, lately transformed into three categories (low, middle and high measures of institutional quality). He finds a substantial positive effect on the growth rates.

 

Coupet (Coupet, 2003) analyzed the total productivity function of the economy and showed that there is a negative interrelation between the level of corruption and investment and thus economic growth for non OECD countries and oil importing countries.

 

Gupta, Davoodi, Alonso-Terme (Gupta et al., 1998) showed that high and growing corruption reduces economic growth, efficiency of government spending and human capital development and finally leads to the expansion of the gap between the richest and the poorest.

 

The rest of the studies mentioned above are all based on the data of risk-rating agencies and prove that there is a strong positive correlation between efficient bureaucracy and economic growth, while corruption has a negative effect on economic growth.

 

Thereby in most of the studies examined the null hypothesis that there is a positive correlation between institutional quality and economic growth was nit rejected. This fact let us expect to obtain similar results in our own research.