Saleh Amirkhalkhali

A varying-coefficients model of export expansion, factor accumulation and economic growth: Evidence from cross-country, time series data

Abstract In this study we re-examine the role of export expansion in developing countries in a production function framework. However, our study goes beyond previous studies by not only using more recent data, but more importantly, by adopting a random coefficients model that can be viewed as a refinement of laws as stated by Pratt and Schlaifer. The random coefficients model is clearly more general than those adopted in previous studies, not only because it more correctly describes the law relating growth to its determinants, but also permits the impact for those determinants to be country specific. This is particularly important when considering a diverse group of developing countries. We estimate the model using Swamy-Mehta methods. Our results show that economic growth bears a statistically significant relationship to export growth for all except the strongly inward oriented group of countries. However, our findings suggest that there is no major difference between moderately inward oriented countries and moderately and strongly outward-oriented countries in-so-far as the impact of export expansion on economic growth is concerned.

Testing for Capital Mobility: A Random Coefficients Approach

This paper extends the Feldstein-Horioka (1980), Feldstein (1983) and subsequent studies on the degree of capital mobility, by adopting a random coefficients model. This approach is more general in that it permits inter-country variations in the degree of capital mobility to arise due to the differences in size as well as in other institutional or structural characteristics. In addition, it is a refinement of stochastic laws as defined by Pratt and Schlaifer (1984, 1988). Our results point to significant inter-country differences in the degree of capital mobility, thereby lending support to the random coefficients approach. In particular, our results indicate that, on average, the degree of capital mobility is much higher than implied by fixed coefficients approach. Finally, country size itself does not appear to bear a systematic relationship with the degree of capital mobility as suggested by Murphy (1984).

Saving-investment correlations, capital mobility and crowding out: some further results

Abstract This study attempts to reassess the evidence on the degree of capital mobility and crowding out by applying a varying coefficients model to data on 19 OECD countries over the 1971–1999 period. Our period-specific results strongly support the crowding-out effect as well as the low capital mobility argument for this group of countries as a whole. However, the strength of the crowding-out effect appears to weaken and the degree of capital mobility to increase in the 1990s as compared to the 1970s and 1980s. We also classify countries into five groups according to the relative size of the government sector. Our group-specific results indicate that the degree of capital mobility is generally lower and the crowding-out effect generally stronger, in country groups with smaller governments. The differences are especially evident when we compare the group with largest government size with all other groups, those differences between the latter being much more modest. However, significant differences in the country-specific results suggest that it is prudent to be cautious when we draw conclusions about crowding-out and capital mobility for specific countries from the period-wise or group-wise results. This is particularly important in drawing policy implications for specific countries.

On the Forecasting Performance of Estimators for a Structural Equation in a Large System: Some Monte Carlo Results

This paper examines the relative forecasting performance of published estimators proposed for a structural equation in a large system using Monte Carlo experiments with antithetic variates. The performances of the estimators are compared in terms of the accuracy of the within-sample as well as post-sample predictions for 10 structural equations by using the mean absolute percentage error of forecasts. It is concluded that the ridge-type estimator developed by Kadiyala and Nunns and the estimator proposed by Swamy and Holmes performed consistently better than other estimators in both the within-sample predictions and ex post forecasts.

Prediction in regression with autocorrelated normal and non-normal errors

This simulation study focuses on the relative small sample properties of some widely applied predictors in regression with AR(1) errors where there errors are allowed to follow normal and non-normal distributions. The conclusions are: all predictors considered are significantly unbiased; the relative performances of predictors, from the efficiency point of view, seemed insensitive to the nature of the error distribution; and the standard errors of predictors computed from the asymptotic formulas are very useful for purposes of inference in small sample and under all assumed distributions.

A simulation study of estimators of sur models with unequal numbers of observationsand with non-normal disturbances

This paper reports the results of a Monte Carlo experiment conducted to compare the small sample properties of selected feasible estimators of the seemingly unrelated regressions model with unequal numbers of observations and with non-normal disturbances. Our results indicate that the choice of estimator depends mainly on the size of the sample, the number of extra observations and the degree of correlation among explanatory variables across equations. For instance, methods which use the extra observations in estimating all elements of the disturbance covariance matrix are relatively more efficient when data on explanatory variables are highly correlated. However, the usual estimator which is based on only the matched observations in estimating the disturbance covariance matrix performs well and competes closely with the other feasible estimators when the correlation of explanatory variables across equations is low and the sample is small or the extra number of observations is large relative to the sample...

On the Dynamics of the Interrelationships between Fiscal Deficits and Some Important Macroeconomic Aggregates in Canada

Broadly speaking, three views about the macroeconomic implications of fiscal deficits can be identified. The traditional Keynesian argument, assuming chronic unemployment, states that fiscal deficits stabilize aggregate demand, increase private saving and foster investment and growth. The neoclassical view, assuming self-equilibration of the economy, states that deficits crowd out private investment, and are, therefore, damaging to a countrys growth prospects. Between these two diametrically opposing views, is the Ricardian Equivalence Theorem. This theory argues that increasing fiscal deficits are matched by an equivalent increase in private saving with no real effects. In this paper, we investigate the dynamics of the interrelationships among fiscal deficits, private saving, investment and other macroeconomic aggregates in Canada over the period 1961-90 using vector autoregressions, cointegration tests and error-correction models.

On the reliability of quasi-t-statistics: Some Monte Carlo results

This study intends to shed light on the relative reliability of computed t-statistics (quasi-t-statistics) of some commonly used estimators for the regression model with autoregressive non-normal errors. Our Monte Carlo results indicate the overall reliability of such t-statistics for making inferences about the true values of the coefficient parameters where the coefficient of autocorrelation (@r) is negative. However, the computed t- statistics are unreliable when @r assumed high positive values and the data are trended.

The P-version of the finite element method: Theory and applications: On the use of estimated structural equations for prediction: Some empirical results

The purpose of this simulation study is to address the question of choice between the reduced-form and the estimated structural equations for purposes of prediction in simultaneous-equation models.