Title page for ETD etd-111597-193238


Type of Document Dissertation
Author Zaloshnja, Eduard X.
Author's Email Address ezaloshn@vt.edu
URN etd-111597-193238
Title Analysis of Agricultural Production in Albania: Prospects for Policy Improvement
Degree PhD
Department Agricultural and Applied Economics
Advisory Committee
Advisor Name Title
Taylor, Daniel B. Committee Chair
Littlefield, James E. Committee Member
McDowell, George R. Committee Member
McGuirk, Anya M. Committee Member
Norton, George W. Committee Member
Keywords
  • Agricultural Supply Response
  • Econometric estimation
  • Production economics
Date of Defense 1997-09-11
Availability unrestricted
Abstract
The overall objective of this study is to develop a framework to predict the impacts of government policies on agricultural production in Albania. The specific goal of this study is to provide some empirical estimates of the farmers' short-run supply response to government policies that effect output and input prices.

Different theoretical approaches to integrating the questions this study purports to answer were considered. Two models were deemed as most appropriate for Albanian agriculture. The first is a semi-commercial farm household model and the second is the well-known indirect profit function model. The first model was preferred. However, the second was used instead, due to the lack of information necessary for an empirical application of the semi-commercial farm household model.

A quadratic functional form was selected to approximate the profit function. It satisfied the Taylor series approximation convergence test. Two approaches were used to estimate the empirical model. In the first, the traditional approach, the symmetry and homogeneity conditions were imposed beforehand and then the system of equations was estimated using the ITSUR procedure in SAS. Following common practice, a joint Rao test of these conditions was conducted, implicitly assuming that the test statistic has a Fisher distribution or, stated differently, assuming that parameter estimators are normally distributed. The test results indicate that the conditions are met.

A second approach, proposed by McGuirk, et al., was also used in this study. The approach proposed by McGuirk, et al., requires that, before imposing and/or testing any theoretical assumption, the unrestricted model is estimated and tested to see if all the underlying statistical assumptions of the linear regression are met.

The misspecification tests suggested that the model is not statistically adequate. This finding indicated that the theoretical test conducted in the traditional approach was invalid. An alternative estimation procedure is proposed in the study for cases when a statistically adequate model cannot be specified. Named the sub-sample or the bootstrapping method, this procedure consists of randomly selecting a large number of sub-samples from the cross-sectional sample and running a regression for each of them. The large number of estimates for each of the coefficients serves as a basis for estimating 95-percent confidence intervals.

An inspection of the supply and input demand elasticities calculated based on coefficients estimated through the sub-sample method revealed that half of them have wide 95 percent confidence intervals. Therefore, predicting policy impacts across all output and input equations is not possible. However, elasticities that have narrow confidence intervals and make economic sense can be used to predict isolated policy impacts, if Albania returns to the conditions that prevailed before the political turmoil of 1997.

Files
  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  EDDIE.PDF 396.31 Kb 00:01:50 00:00:56 00:00:49 00:00:24 00:00:02

Browse All Available ETDs by ( Author | Department )

dla home
etds imagebase journals news ereserve special collections
virgnia tech home contact dla university libraries

If you have questions or technical problems, please Contact DLA.