dear sirs
The disturbance is composed of two influences: Vit are random
variables assumed to be iid N(0, Γ蓋子
v2) and independent of Uit, which are non-negative random variables
assumed to account for the cost of inefficiency in production. The
later are assumed to be independently distributed as truncations at
zero of the N(mit, Γ蓋子U2) distribution; where mit = zitΓ蓋残 is a
(p×1)
vector of variables Z which influence the efficiency of a firm; and
Γ蓋残
is an (1×p) vector of parameters to be estimated.
The error term is decomposed using the conditional distribution
approach proposed by Jondrow et
al. (1982) for a truncated normal distribution; providing an unbiased,
though inconsistent, estimate
of the cost of inefficiency. The measure of cost inefficiency relative
to the cost frontier is defined
as: E[U|Γ蓋仕] =
[Γ蓋子Γ蓋史/(1+Γ蓋史2)][Γ蓋師(Γ蓋嗣*/Γ蓋子)/Γ蓋撒(Γ蓋嗣*/Γ蓋子)
+ Γ蓋嗣*/Γ蓋子] (3)
where Γ蓋子 = Γ蓋子V2 + Γ蓋子U2, Γ蓋史 =
Γ蓋子U2/(Γ蓋子V2+ Γ蓋子U2), Γ蓋嗣* = Γ蓋仕Γ蓋史/Γ蓋子 +
Γ蓋嗣/(Γ蓋子Γ蓋史), Γ蓋師 is the
standard normal density
function, and Γ蓋撒 the ***ulative normal density function, and all
other
terms are as previously
defined.
THE QUESTION IS HOW CAN WE DECOMPOSE E IF WE WANT TO DO IT MANUALLY ?
I CAN DO IT USING FRONTIER 4.1 BUT I WANT UNDERSTAND HOW CAN IT BE
DONE
THANKS
|
