Loading Events
  • This event has passed.

Ao Wang (CREST) – "Demand for Bundles: A Statistical Test", joint work with Alessandro Iaria (U. Bristol) and Julien Monardo (CREST)

October 24, 2017 @ 12:15 pm - 1:15 pm
The Microeconometrics Seminar: Every Tuesday at 12:15 pm.
Time: 12:15 pm – 1:15 pm
Date: 24th of October 2017
Place: Room 3001.
Ao Wang (CREST) – “Demand for Bundles: A Statistical Test”, joint work with Alessandro Iaria (U. Bristol) and Julien Monardo (CREST)

Abstract: On any given shopping trip, consumers are typically observed to purchase baskets of products rather than individual products. However, the econometric models commonly used to estimate demand elasticities to product characteristics – such as price – rely on the assumption that each of the products observed to be purchased in a basket was chosen independently from the others. This restrictive assumption is motivated by the complexity of the alternative, a demand model that accounts for the correlations across the products often observed to be jointly purchased (Gentzkow (2007)). We propose an easy-to-implement statistical test that allows one to verify whether the purchase data at hand reject the product independence assumption. The test can be implemented both with individual and market-level data, and it is unaffected by price endogeneity. The test can inform and formally guide the decision of estimating the more complex Gentzkow (2007) model, or conversely the simpler and more readily available standard demand model.

Organizers:
Laurent Davezies (CREST – ENSAE Paristech),  Arne Uhlendorff (CREST – ENSAE Paristech) & Yannick Guyonvarch (CREST – ENSAE ParisTech)
Sponsors:
CREST
Lunch registration:
food provided, no registration
The Microeconometrics Seminar: Every Tuesday at 12:15 pm.
Time: 12:15 pm – 1:15 pm
Date: 24th of October 2017
Place: Room 3001.
Ao Wang (CREST) – “Demand for Bundles: A Statistical Test”, joint work with Alessandro Iaria (U. Bristol) and Julien Monardo (CREST)

Abstract: On any given shopping trip, consumers are typically observed to purchase baskets of products rather than individual products. However, the econometric models commonly used to estimate demand elasticities to product characteristics – such as price – rely on the assumption that each of the products observed to be purchased in a basket was chosen independently from the others. This restrictive assumption is motivated by the complexity of the alternative, a demand model that accounts for the correlations across the products often observed to be jointly purchased (Gentzkow (2007)). We propose an easy-to-implement statistical test that allows one to verify whether the purchase data at hand reject the product independence assumption. The test can be implemented both with individual and market-level data, and it is unaffected by price endogeneity. The test can inform and formally guide the decision of estimating the more complex Gentzkow (2007) model, or conversely the simpler and more readily available standard demand model.

Organizers:
Laurent Davezies (CREST – ENSAE Paristech),  Arne Uhlendorff (CREST – ENSAE Paristech) & Yannick Guyonvarch (CREST – ENSAE ParisTech)
Sponsors:
CREST
Lunch registration:
food provided, no registration