Equation (2.12a) shows that when both bid shading and industry level imperfect
competition are considered, the conjectural variation obtained with the NEIO model is a
“mixed” conjectural variation (Θ~) given by the sum of the “true” conjectural variation
(Θ) and the average bid shading δj, weighted by the ratio of elasticity of cattle supply to
the price of cattle times the Herfindahl index (εsf /( p f HHI )).
The encompassing model considering both bid shading and industry level
imperfect competition is obtained by substituting equations (2.12a) and (2.12) back into
the industry supply equation (2.9a) to yield:
(2.13)
pr
pf
- c(Y ) = [ p f
(1+θ)HHI ]+δ-f-,
εsf j pfHHI
The model represented by equation (2.13), is more general than the models represented
by equations (2.4a) and (2.9a) since it nests both (2.4a) and (2.9a). Industry-level
imperfect competition, which the NEIO model seeks to explain, is captured by Θ, the
industry conjectural variation. The price markdown considered by the auction model is
represented by the bid shading factor δj.
Notice that if δj = 0, there is no bid shading, and all perceived price markdown is
due to industry level imperfect competition. In this case, equation (2.13) becomes
equation (2.9a). If Θ = 0 and δj ≠ 0, equation (2.13) becomes equation (2.4a), and all
perceived price markdown is due to bid shading. If Θ ≠ 0 and δj ≠ 0, then perceived price
markdown are due to both bid shading and industry level imperfect competition.
Data and Empirical Application
This section uses data from a cattle procurement experiment to test the encompassing
model proposed in the previous section. The cattle experiment is described first,
16