Estimating individual response coefficients in varying coefficients regression models

The objective of this paper is to suggest a method of estimating the individual response coefficients in varying coefficients regression models. An empirical

application of the method is demonstrated, using farm-level micro data from the Philip-

pines.     

Simultaneous testing of regression disturbances for heteroscedasticity and non-normality

Drawing on the procedures developed for estimating stochastic frontier production functions, the objective of this paper is to derive a practical two-directional test for both heteroscedasticity and non-normality, when the structure of heteroscedastic-ity is not known to the researchers.     

Stochastic varying coefficients gravity model: an application in trade analysis

In international trade analysis between countries, the central theme is the examination of whether or not there are any significant diff erences between the actual trade and potential trade, given the determinants of trade flows. Thus, estimating the potential trade is an important component in trade analysis. The objective of this paper is to suggest a methodology to estimate the potential trade flows between countries, using the gravity model, which has been established in the literature as the most successful empirical trade flow equation, usually producing a good fit. The application of the method has been demonstrated using trade flows between Australia and its trading partners in the Indian Ocean Rim.     

Bias free measurement of technical efficiency

Technical efficiency, which is a measure of production performance of a firm, has been estimated generally using a primal production frontier. Since the estimation is carried out for a given level of inputs, the efficiency measure includes the effect of 'input-mix' or 'input-allocation' and consequently, the technical efficiency estimate is biased. The objectives of this paper are to gauge the magnitude of 'input-mix' bias in technical efficiency estimate and to suggest a method to measure technical efficiency eliminating the bias. The workability of the suggested method is demonstrated through an empirical analysis using agricultural data from India covering the period 1970-1993.     

Estimating technical efficiency with non-neutral shift of the frontier

In the literature, technical efficiency is measured as the ratio of the observed output to potential output. Although there is no a priori theoretical reasoning, in the stochastic framework of measuring technical efficiency, potential output has been conventionally assumed as a neutral shift from observed output, owing solely to a larger intercept term in the frontier production function and without change in the input response coefficients. The objective of this paper is to propose and apply a method to measure technical efficiency without the above assumption. Furthermore, this methodology does not require the restrictive assumption of a particular distribution for the efficiency-related error term, as has been the case until now in the stochastic production function literature. A random sample of farmers from Madurai district in Tamil Nadu, India, was used. The analysis revealed substantial variation in the farm-specific input response coefficients between farms, which means that the contributions of individual inputs to the output differ from farm to farm, because the methods of application of the individual inputs vary. The frontier production function, which defines the potential of a technology, is determined by the highest values of the coefficients of each individual input which may come from one or more farms. Farm-specific frontier functions generally showed a considerable potential for improving the technical.performance of each input.     

On the estimation of a regression model with fixed and random coefficients

With the recent accent on varying parameter models, theoretical studies have concentrated on estimating equations where all the coefficients are stochastic. This paper analyses an estimation of a single equation in which the coefficients of certain exogenous variables are fixed and those of other exogenous variables are random.