Kenkel, Norris
Real-Time Weather Information 367
with higher sales are likely to pay slightly more for the mesoscale weather data. In addition,
those producers with irrigated acres could be expected to pay more for the weather data
(about $0.76 per month more, all else constant) than producers who do not irrigate. Irrigated
crops are generally more intensively managed, and irrigators could be expected to value
information which aids them in monitoring soil moisture and scheduling irrigation. Finally,
as expected, those producers who have suffered larger weather-related crop losses expressed
a higher willingness to pay for the weather data; an increase of one point in the percentage
of crop sales lost due to weather would result in a $1.16 per month higher bid. It is not
surprising that producers with higher weather-related losses would be interested in ways to
reduce weather-related risks.
For the raw data/value-added combination model, sales, irrigation, and weather-related
losses were found to significantly impact willingness to pay. The relationships of sales,
irrigation, and weather-related losses to willingness to pay were again positive, as expected.
The impact of sales is again slight. Irrigators bid about $0.76 per month higher than
nonirrigators, and an increase of one point in the percentage of crop sales lost due to weather
would mean a bid of $ 1.40 more per month for the value-added information. The coefficients
for crop acres and number of crops were significant at the 0.15 and 0.16 levels. While the
negative crop acres coefficient suggests that the diversification effect of larger acreages
reduces the perceived value of weather data as a risk management tool, a positive coefficient
on number of crops suggests the opposite.
For the raw data∕value-added model, the production of peanuts, cotton, or alfalfa did not
significantly impact willingness to pay. This was somewhat surprising since much of the
proposed value-added weather information addresses problems and needs specific to those
producers.
Results from the maximum likelihood models were used to calculate mean and median
willingness to pay for the raw weather data and the value-added information. Mean
willingness to pay for the raw weather data was $5.83 per month (with a standard error of
0.58); the median was $4.05. Respondents indicated that they would pay only slightly more
for the value-added weather information; the mean willingness to pay for both raw data and
value-added information was $6.55 per month (with a standard error of 0.84). The median
willingness to pay estimate for the raw data∕value-added model ($3.85) was slightly lower
than for the raw data model; this occurred because of the number of observations included
in the raw data∕value-added regression with a zero bid which were excluded, as protest bids,
from the raw weather data model.
The CV method can also be used to estimate the aggregate value of the system. Mesonet’s
developers are interested in the aggregate willingness-to-pay estimate because it represents
the value of the system to agricultural producers. If public funds were to be used instead of
user fees, officials would require information on the value of the system to justify the
expenditure of public funds. The aggregate willingness-to-pay estimate also provides an
upper limit on the proportion of annual operating and development costs which can be
recovered from agricultural user fees. The actual revenue which could be collected would
be less than the calculated aggregate value, unless the Mesonet developers could implement
a system of perfect price discrimination which captured all consumer surplus.
Based on alternatives for aggregating mean willingness to pay which have been applied
in the literature (Loomis; Mitchell and Carlson), a range within which the aggregate value