The Consumption Choice of Organics: Store Formats, Prices, and Quality Perception –A Case of Dairy Products in the United States - Organic Food and Agriculture: New Trends and Developments in the Social Sciences - page 10

 

 

x β ' x β '

 

Prob Yx  1|  e / 1  e     x '   , (3)

where Λ denotes the logistic cumulative distribution function.
The elements of exogenous variable set ( x ) include the price premium of organic versus
conventional products in the store at trip t (price premium), the discount use rate (%
discount), days between trips, distance between consumer's home and store (distance),
number of stores within the shopping range by format, total spending and organic
percentage in other items purchased in the same trip, income and household demographics.
The set of demographic characteristics includes household size and dummy variables
identifying the percentages of 1) college educated householder, 2) married householder, 3)
family with preschool children, 4) family with school-aged children, and 5) the elderly. In
addition, we include household's format loyalty to three formats alone and their interaction
terms with price premium variable to allow for differential price sensitivity and fixed
component of utility to differ among households with differentiated preferences over
formats. In addition, we employ the same set of demographics described above for
preference heterogeneity.
In sum, the complete empirical specification of organic choice model is as follows:

 

Prob Yx  1|    (  loyalty    ( price premium * loyalty )

f 1 f f f 2 f f

  % discount   days between trips

3 4

  distance   total spending of the trip

5 6







% organic in other items of the trip

income household size
college educated

7
8

(4)

 

9

    married

10 11

  preschool children   school  age children

12 13

  elderly )  . 

14

Definitions and measures of variables are summarized in table 4 and further discussed in
next subsection.

3.3 Estimation procedures and measures
Within each set of two years, we use the first 26 weeks as our “initialization” period to
identify shopper types and format-specific indexes to avoid potential endogeneity between
quality and store format choices. The remaining 78 weeks were used as the “estimation”
sample. The estimation is based on every shopping trip of households with shopping
duration being no longer than 30 days during the estimation period at seven major retail
chains in the market.
Format Loyalty . The format-specific loyalty for a household is represented by the percentage
of trips that the household made to the format during the initialization period. Specifically,
we use the following standard measure of loyalty ( FL ) used also by (Bell, Ho, and Tang
1998; Briesch, Chintagunta, and Matzkin 2010):

   

FL  NV  0.5 /  NV  1 for f  AC , (5a)

hf , hf , f hf ,