EDITOR’S PROOF
236 G. Serra
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
Lemma 5The expected utility ofR’s leadership for each value ofmRis
EURE(mR=elite)=−(XRE−XL)πL( 1 −πRI)
−(XRE− 0 )
[
πLπRI+( 1 −πRI)
]
−(XRE−XRE)( 1 −πL)πRI
EURE(mR=primary)=−(XRE−XL)πL
(
1 −(πRI+S)
)
−(XRE− 0 )
[
πL(πRI+S)+( 1 −πL)
(
1 −(πRI+S)
)]
−|XRE−XRM|( 1 −πL)(πRI+S)
Armed with these results, the leadership in partyRcan measure the conse-
quences of choosing one CSM over the other.
7 The Optimal Selection of a CSM
The leadership in partyR will choose the optimal rule mR by comparing
EURE(mR=elite)andEURE(mR=primary). It will choose the CSM that yields
the highest expected utility, and if it is indifferent, I assume that it will choose
an elite selection. A primary will be adopted if and only ifEURE(mR=elite)<
EURE(mR=primary). That condition leads to the following result, recalling that
dR≡|XRM−XRE|.
Theorem 3The leadership of partyRwill adopt a primary election if and only if
dR<T
withT≡S[X(RE 1 −(π^1 −L)(ππL)RI−+XS)LπL].
The intuition behind this result is thatR’s leadership will delegate the nomination
if and only if the RAF’s ideology is close enough to its own. In other words, inter-
nal party democratization will only ensue from enough elite-mass congruence. How
close do primary voters need to be to the party elite? It depends on a certain thresh-
old,T, introduced in the theorem. If the preferences of the elite and the mass of party
Rare so incongruent thatT≤dRthen the leadership will not adopt a primary elec-
tion. This could happen for two reasons. On one hand, the RAF could be so far on
the right of the leadership thatXRE+T≤XRM. In that case the leadership will not
adopt a primary election because the primary voters are tooextremist. On the other
hand, the RAF could be so far on the left of the leadership thatXRM≤XRE−T.In
that case the leadership will not adopt a primary election because the primary voters
are toocentrist.
As it turns out, the first reason (that primary voters might be too extreme) is fre-
quently found in some way or another in scholarly comments about primary elec-
tions. Yet the second reason (that primary voters might be too moderate) is equally