This preview shows pages 1–3. Sign up to view the full content.
Econ 513, USC, Fall 2005
Lecture 15. Discrete Response Models:
Multinomial, Conditional and Nested Logit Models
Here we focus again on models for discrete choice with more than two outcomes. We
assume that the outcome of interest, the choice
y
takes on nonnegative integer values
between zero and
J
;
y
∈ {
0
,
1
, . . . , J
}
. Unlike the ordered case there is no particular
meaning to the ordering. Examples are travel modes (bus/train/car), employment status
(employed/unemployed/outofthelaborforce), marital status (single/married/divorced/widowed)
and many others.
We wish to model the distribution of
y
in terms of covariates. In some cases we will
distinguish between covariates
x
i
that vary by units (individuals or ﬁrms), and covariates that
vary by choice (and possibly individual),
x
ij
. Examples of the ﬁrst type include individual
characteristics such as age, or education. An example of the second type is the cost associated
with the choice, for example the cost of commuting by bus/train/car. This distinction
only arises from the economics (or general scientiﬁc) substance of the problem. McFadden
developed the interpretation of these models through utility maximizing choice behavior. In
that case we may be willing to put restrictions on the way covariates aﬀect choices: costs of
a particular choice aﬀect the utility of that choice, but not the utilities of other choices.
The strategy is to develop a model for the conditional probability of choice
j
given the
covariates. Suppose the model is Pr(
y
=
j

x
) =
P
j
(
x
;
θ
). Then the log likelihood function is
L
(
θ
) =
N
X
i
=1
J
X
j
=0
1
{
y
i
=
j
} ·
ln
P
j
(
x
i
;
θ
)
.
I. Multinomial Logit
Suppose we only have individual speciﬁc covariates. Then we can model the response
probability as
Pr(
y
=
j

x
) =
exp(
x
0
β
j
)
1 +
∑
J
l
=1
exp(
x
0
β
l
)
,
for choices
j
= 1
, . . . , J
and
Pr(
y
= 0

x
) =
1
1 +
∑
J
l
=1
exp(
x
0
β
l
)
,
for the ﬁrst choice. This is a direct extension of the binary response logit model. It leads to
a very wellbehaved likelihood function and is easy to estimate. More interestingly it can be
viewed as a special case of the following conditional logit.
1
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentII. Conditional Logit
Suppose all covariates vary by choice (and possibly also by individual, but that is not
essential here). Then McFadden proposed the conditional logit model:
Pr(
y
i
=
j

x
i
0
, . . . , x
iJ
) =
exp(
x
0
ij
β
)
∑
J
l
=0
exp(
x
0
il
β
)
,
for
j
= 0
, . . . , J
.
This is the end of the preview. Sign up
to
access the rest of the document.
 Fall '07
 Rashidian
 Econometrics

Click to edit the document details