(7)
∂ x * i
~iΓ^
ni
i2 Pri iPrj inj i2 Prj iPri ini
ixiixj ixj iI j ixj2 ixi iI
( ∂2 Prj ∂2 Pr. ∂2 Pr. ∂2 Prj 'ï
i- i
∂x 2 ∂x.2 ∂x i ∂x ∂x i ∂x
j . .j .j
∀.≠ j, .,j=L,H
Rewriting (7) together with (5), we obtain the fundamental equation that generates all
the comparative statics results:
(8)
∂ x*.
∂I
ɪ ∂2 Pri
B ∂xi ∂x
.j
ηjn.
1∂2Prj
B ∂xj 2
η. nj
∀.≠ j, .,j=L,H
where B =I n. n
i Pr
I dx j
∂2 Pr.
∂x.2
i Pr i Pr Ï
∂x i ∂x. ∂x i ∂x.
i j i j J
and all second-order partial
derivatives are computed at the Nash equilibrium (x *H , x*L) . The first term in (8)
represents the strategic rival’s-stake (“substitution”) effect. The sign of this term is
∂2 Pr.
equal to the sign of------η j. The second term represents the own-stake ( income )
∂x. ∂x
.j
effect. The sign of this term is equal to the sign of η . . By assumption,
∂2 Pr. (x. , x )
----i±÷j < 0 and, by (2),
∂ x. 2
i 2 Pr-(xi, xj )
∂ x. ∂ x j
12Prj(xj, xi)
∂ x. ∂ x j
< 0 . Hence, B>0.
A. . Public Policy, Efforts and Winning Probabilities
When a change inI only affects the stake of one of the contestants, as in reforms type
(ii) and (iii), η . or η j is equal to zero and (8) reduces to
(8’)
∂x*.
∂I
( ∂ 2Pri ï
--------η j ni
∂x i ∂x,
ij
or
∂x*i
∂I
∂2 Prj
∂xj2
ï
ηi nj ïï
J
∀i≠ j, i,j=L,H
In these cases the change in player i’s effort corresponding to the change in nj is
equal to the strategic rival’s-stake (“substitution”) effect, when i≠j, or to the strategic
11
More intriguing information
1. THE MEXICAN HOG INDUSTRY: MOVING BEYOND 20032. Optimal Private and Public Harvesting under Spatial and Temporal Interdependence
3. The name is absent
4. The name is absent
5. The name is absent
6. The name is absent
7. Survey of Literature on Covered and Uncovered Interest Parities
8. On Evolution of God-Seeking Mind
9. The name is absent
10. Testing Panel Data Regression Models with Spatial Error Correlation