Events

Economics Seminar

Author:

• Andrew Ching
  University of Toronto

• Matthew Osborne
  University of Toronto

We develop a new empirical strategy for identifying the parameters of dynamic structural models in markets for storable goods, with a focus on identification of the discount factor.  The identification strategy rests on an exclusion restriction generated by discontinuities in package sizes:  In storable goods product categories where consumption rates are exogenous and package sizes  are  discrete,  current  utility  does  not  depend  on  inventory  unless  a  package  gets  used up.  Most of the time, inventory only enters a consumer's expected future value.  We develop conditions  for  identification  in  two  situations:  when  inventory  is  observed  to  the  researcher, and when it is unobserved.  When inventory is unobserved, we show that the shape and slope of the purchase hazard can identify the discount factor.  We demonstrate the feasibility of our identification strategy with an empirical exercise, where we estimate a stockpiling model using scanner data on laundry detergents.  Our estimates suggest that consumers are not as forward-looking as most papers in the literature assumes; our estimates of weekly discount factors average at about 0.69, which is significantly lower than the value used in previous research (it typically is set at 0.99, using the market interest rate).  We also find evidence of significant unobserved heterogeneity  in  discount  factors,  estimating  the  25th  and  75th  percentiles  of  the  population distribution of discount factors to be 0.64 and 0.75.  We find evidence that discount factors are correlated with demographic characteristics: higher income, older, and larger households appear to behave in a more forward-looking way.