function V (π1, π2, Y, L). We assume V is nondecreasing and strictly concave in each of its
arguments.
Usually the LG signs a managerial contract with the TVE manager that ties manager
compensation to output and profit levels (Whiting, 1996), suggesting the LG likely influences
TVE production choices. Also, a survey conducted by Raiser (1997) suggests LGs induced
over 16 per cent of the TVEs to meet minimum labor hiring goals. In what follows we assume
the LG has some but different influence over the output choice or hiring decisions of the PE
and TVE. Also, we assume the local government exerts some but different effort in securing
loans for the PE and TVE. Then, under such assumptions the LG optimization problem is
given by:
^,... , . . ,,,
V(l*,Y *,I ) = max {V [π1 (E1 (I ) ,Y1,l1) , ∏2(E2 (I ) ,Y2,l2),l1 + l2,Y1 + Y2] (3)
e1 ,e2,l1,l2,Y1,Y2
-C1 (Ei (I) ,Yi - Y*,l1 - l*) - C2 (E2 (I) ,Y2 - Y2*,l2 - l*)
subject to Y1 - Y1* ≥ 0, l1 - l* ≥ 0,Y2 - Y2* ≥ 0, l2 - l* ≥ 0} .
Here, Cj (Ej,Y∙ - Y*,lj∙ - l*) is the effort cost to the LG of raising Ej∙ in loans for firm j,
while inducing the firm to produce Yj units of output and hire lj employees. We assume Cj
is non-decreasing and convex in each of its arguments, and for all (e, l, Y) ,C1 (E, l, Y) <
C2 (E, l, Y) and Ci1 (E, l, Y) <Ci2 (E, l, Y) ,i=2, 3. The assumptions on C1 and C2 imply
that compared with TVEs, inducing PEs to choose output targets or minimum employment
levels is more difficult/costly for the LG.
The solution to (3), denoted EË,, lɪ, Y1, E2, l2, Y2 ) , might involve output-targets and min-
imum employment goals that are inconsistent with profit maximization. Also, depending on
the institutional regime I, LG efforts might result in loan amounts where ∕T'∣ = E2. The prior
discussion would suggest that before 1994 E1 might be larger than E^2, but post-1994 this
might not be the case.
Since no well-defined theoretical framework is used to specify the structural and be-