Working Paper: CEPR ID: DP13382
Authors: Matthew Gentry; Martin Pesendorfer
Abstract: We develop a structural model of demand with expectations-based reference effects following Koszegi and Rabin (2006). We apply this model to panel data on ketchup purchases and a repeated cross section on automobile purchases, finding significant reference effects in both cases. Estimated reference effects imply substantial differences between short- and long-run demand responses, with magnitudes comparable to a dynamic stockpiling model. This attractive model feature allows us to explore price policy alternatives such as high-low versus every-day-low-pricing at low computational cost. Finally, we embed the model within a fully dynamic framework additionally accommodating limited attention and forward-looking search.
Keywords: consumer behavior; reference effects; demand estimation
JEL Codes: D12; D90; D91
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| expectations-based reference effects (D84) | consumer demand for ketchup (D12) |
| lower price than expected (D41) | willingness to pay for ketchup (D10) |
| perceived high price (D41) | willingness to pay for automobiles (R48) |
| short-run demand responsiveness (J23) | price changes (P22) |
| high-low pricing strategies (D49) | long-run profit levels (L21) |
| everyday-low pricing strategies (D40) | long-run profit levels (L21) |