In their article, "A Model of the Distribution of Branded Personal Prod- ucts in Jamaica,''* John U. Farley and Harold J. Levitt developed the following model (the personal products considered were shaving cream, skin cream, sanitary napkins, and toothpaste):
Y1i = α1 + β1 Y2i + β2 Y3i + β3 Y4i + u1i
Y2i = α2 + β4 Y1i + β5 Y5i + γ1 X1i + γ2 X2i + u2i Y3i = α3 + β6 Y2i + γ3 X3i + u3i
Y4i = α4 + β7 Y2i + γ4 X4i + u4i
Y5i = α5 + β8 Y2i + β9 Y3i + β10 Y4i + u5i
where Y1 = percent of stores stocking the product
Y2 = sales in units per month
Y3 = index of direct contact with importer and manufacturer for the product
Y4 = index of wholesale activity in the area
Y5 = index of depth of brand stocking for the product (i.e., average number of brands of the product stocked by stores carrying
the product)
X1 = target population for the product
X2 = income per capita in the parish where the area is
X3 = distance from the population center of gravity to Kingston
X4 = distance from population center to nearest wholesale town
a. Can you identify the endogenous and exogenous variables in the pre- ceding model?
b. Can one or more equations in the model be estimated by the method of least squares? Why or why not?