1. Demand systems 4178
1.1. Characteristic space: The issues 4181
1.2. Characteristic space: Details of a simple model 4185
1.3. Steps in estimation: Product level data 4188
1.4. Additional sources of information on demand parameters 4190
1.4.1. Adding the pricing equation 4190
1.4.2. Adding micro data 4192
1.4.3. Identifying restrictions 4196
1.5. Problems with the framework 4198
1.6. Econometric details 4202
1.7. Concluding remark 4204
2. Production functions 4205
2.1. Basic econometric endogeneity issues 4205
2.2. Traditional solutions 4207
2.2.1. Instrumental variables 4207
2.2.2. Fixed effects 4209
2.3. The Olley and Pakes (1996) approach 4210
2.3.1. The model 4211
2.3.2. Controlling for endogeneity of input choice 4214
2.3.3. Controlling for endogenous selection 4217
2.3.4. Zero investment levels 4220
2.4. Extensions and discussion of OP 4222
2.4.1. A test of Olley and Pakes’ assumptions 4222
2.4.2. Relaxing assumptions on inputs 4226
2.4.3. Relaxing the scalar unobservable assumption 4228
2.5. Concluding remark 4232
3. Dynamic estimation 4233
3.1. Why are we interested? 4234
3.2. Framework 4235
3.2.1. Some preliminaries 4237
3.2.2. Examples 4238
3.3. Alternative estimation approaches 4241
3.3.1. The nested fixed point approach 4242
3.3.2. Two-step approaches 4244
3.4. A starting point: Hotz and Miller 4246
3.5. Dynamic discrete games: Entry and exit 4249
3.5.1. Step 1: Estimating continuation values 4250
3.5.2. Step 2: Estimating the structural parameters 4253
3.5.3. Multiple entry locations 4255
3.6. Models with discrete and continuous controls: Investment games 4257
3.6.1. Step 1: Estimating continuation values 4258
3.6.2. Step 2: Estimating the structural parameters 4262
3.6.3. An alternative approach 4264
3.7. A dynamic auction game 4266
3.7.1. Estimating continuation values 4268
3.7.2. Estimating the cost distribution 4269
3.8. Outstanding issues 4269
3.8.1. Serially correlated unobserved state variables 4270
References 4271