lect14_06mar15

lect14_06mar15 - Imbens Lecture Notes 14 ARE213 Spring...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Imbens, Lecture Notes 14, ARE213 Spring ’06 1 ARE213 Econometrics Spring 2006 UC Berkeley Department of Agricultural and Resource Economics Discrete Response Models IV: Nested Logit Models Here we focus again on models for discrete choice with more than two choices. We assume that the outcome of interest, the choice Y takes on non-negative integer values between zero and J ; Y ∈ { , 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/out-of-the-laborforce), housing choice: apartment, townhouse, sin- gle family house, and many others. One way to induce correlation between the choices is through nesting them. We do this in three steps. I. Nested Logit with a Three-valued Outcome The utility associated with choice j is U ij = μ ij + ij = X ij β + ij . Before in the multinomial logit model we assumed that three residuals were independent with extreme value distributions, implying that the joint distribution function of the three residuals has the form F ( , 1 , 2 ) = exp(- exp(- )) exp(- exp(- 1 )) exp(- exp(- 2 )) . Now suppose the set of choices { , 1 , 2 } can be partitioned into 2 sets, the first consisting of only the first choice 0, and the second consisting of the pair of choices { 1 , 2 } . We assume the joint distribution function has the form F ( , 1 , 2 ) = exp(- exp(- )) exp(- (exp(- 1 /ρ + exp(- 2 /ρ )) ρ ) . Imbens, Lecture Notes 14, ARE213 Spring ’06 2 So, is still independent of ( 1 , 2 ), but ( 1 , 2 ) are no longer independent, with joint distri- bution function F ( 1 , 2 ) = exp(- (exp(- 1 /ρ + exp(- 2 /ρ )) ρ ) . The correlation coefficient for ( 1 , 2 ) is 1- ρ 2 . If ρ = 1, we are back in the multinomial logit case. The marginal distribution function for 1 is F ( 1 ) = exp(- exp(- 1 )) , so that is just an extreme value distribution, with marginal probability density function...
View Full Document

This note was uploaded on 08/01/2008 for the course ARE 213 taught by Professor Imbens during the Spring '06 term at Berkeley.

Page1 / 6

lect14_06mar15 - Imbens Lecture Notes 14 ARE213 Spring...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online