dC1
∂θ1
< 0 and
⅛- > 0
∂θ1 .
The intuition is that when country 1’s costs increases, this leads to smaller re-
ductions by country 1, smaller reductions increase the marginal benefit for
country 2 of the same level of reductions, and, hence, it will reduce more in re-
sponse.
Appendix 2
Proof of result 1: We will show that a qi' < q(H,0) exists that satisfies both 4.a
and 4.b. Define qi' =min arg{NBi(L, qi,0) = NBi(L,q(L,1),1)}, i.e. qi' is the smallest
qi satisfying 4.b. If at the same time, qi' satisfies 4.a with strict inequality, then
existence is guaranteed. Given that NBi(H,q'i,0) > NBi(H,q(H,0),1), then also
NBi(H,qi',0) > NBi(H,q(L,1),1) , since q(H,0) is the unique maximizer of
NBi(H,ρ,0).
From this we have that
NBi(H,q'i,0)-NBi(H,q(L,1),1)>NBi(L,qi',0)-NBi(L,q(L,1),1). Hence, in order
to get separation, there must exist a qi' < qi (H,0), such that
NBi(H,qi',0)-NBi(H,q(L,1),1)>NBi(L,qi',0)-NBi(L,q(L,1),1).
Proof of result 2: Weak domination for the low cost type in this set-up corre-
sponds to the following inequality. If NBi (L,q(L,1),1) ≥ NBi (L, qi ,0), then all qi
are weakly dominated by q(L,1). But this expression is equal to 3.2, the condi-
tion satisfied by all separating equilibria. Hence, all qi are weakly dominated by
c(L,1) for the low cost type.
Define qH = minarg{NBi (H, q 1H ,0) = NBi (H, q(H,0),1)}. We now show that no
q 1H ≠ qH is weakly dominated. This amounts to showing that
43