year.
The survey instrument consists of three parts: demographic and farm type questions, crop
production and input values, and the value of various insurance alternatives. The farm-attribute
questions ask whether the grower contracts his or her production, whether the grower belongs to
a cooperative, the degree of vertical integration, crop insurance purchase history, percentage of
family income earned off-farm, the farm’s debt-to-assets ratio, the grower’s age, and his or her
level of education. The crop production section provides information on the number of crops
grown, acreage of the primary crop, the distance between parcels of that crop in miles, a five-year
history of irrigated and non-irrigated acreage and yields, the average price per pound of 1995
output, and value of fertilizer, chemicals, labor, water, seed, fuels, and other variable costs used in
1995. This section also asks growers to rank the importance of yield, output price, labor cost,
and input cost risk on a Likert scale (1 = high risk, 5 = low risk).
The third section of the survey asks growers to place a subjective value on several
insurance alternatives. This study concerns variants of one - the cost of production insurance
option. Growers are asked to submit what they would be willing to pay, on a per acre basis, for
insurance that guarantees them 100% of their variable costs of production if their yield falls below
65%, 75%, or 85% of their historical average yield. Approximately 30% of the sample growers,
however, would not consider insurance, so entered a “zero” willingness to pay for all options.
Because of the number of null responses, the distribution of the willingness to pay for
insurance is assumed to be truncated at zero. Therefore, the empirical insurance demand model
uses a Tobit estimation procedure (Maddala). A Tobit approach is required if growers reveal a
positive willingness to pay ( W > 0) only if the latent, or unobserved willingness to pay