barley productivity decomposition in iran: comparison of tt, gi, mgi, and gtti approaches

In this paper, the authors present new indices for estimating technical change, return to scale, and tfp growth, as well as its decomposition. these indices, modified general index (mgi), generalized modified general index (gmgi), and general time trend index (gtti), are generalization of general index approaches. these approaches were used for productivity decomposition of iran's barely production across the period 20002012 in 20 provinces. to select the best approaches, estimated tfp growth of tt, gi, mgi and gtti are compared with divisia index. results show that differences between barley tfp growth of tt, gi, mgi, and gtti approaches with divisia index are 39.12, 17.94, 9.71, and 1.61 percent, respectively. the findings revealed that mgi method is appropriate when time series data or panel data with limited crosssection data are used. in addition, when only need to compare periods of time that are not regular (for evaluation plan or policies), it is suggested gmgi method. when time series data or panel data with limited cross section data are used, and there is a trend in every period, the gtti method is recommended for estimating technical change, return to scale, and tfp growth.