Addressing endogeneity in aggregate logit models with time-varying parameters for optimal retail-pricing
| Autor: | Daniel Guhl |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
050210 logistics & transportation
Discrete choice 021103 operations research Information Systems and Management General Computer Science 05 social sciences Aggregate (data warehouse) Logit 0211 other engineering and technologies 02 engineering and technology Management Science and Operations Research Industrial and Manufacturing Engineering Control function Modeling and Simulation Normative model of decision-making 0502 economics and business Econometrics Economics Relevance (information retrieval) Endogeneity Generalized method of moments |
| Zdroj: | European Journal of Operational Research. 277:684-698 |
| ISSN: | 0377-2217 |
| DOI: | 10.1016/j.ejor.2019.02.058 |
| Popis: | It is well known that price endogeneity is a severe problem in demand models for market-level data (e.g., aggregate logit models) because it leads to biased estimates and therefore incorrect managerial implications. If the price parameter varies over time, as is usually the case, the relevance of the issue increases because standard methods to correct endogeneity biases (e.g., generalized method of moments) fail. This paper presents a control function approach as a remedy. A comprehensive simulation study demonstrates this method’s suitability, such that addressing endogeneity with the control function approach is the best choice. Moreover, addressing the endogeneity problem incorrectly may be even more harmful than simply ignoring it. To further illustrate the control function approach, we analyze the demand for canned tuna using aggregate retailer-level data. Here, all utility parameters vary over time and price endogeneity is indeed an issue. Effectively addressing price endogeneity correct has positive economic consequences: a normative model analysis reveals that implementing the control function approach yields a 3 % increase in retailer profits. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect