Estimating the Value of Precipitation Forecast Information in Alfalfa Dry Hay Production in Ontario

Abstract

The value of weather forecast information to farmers is an important component of program evaluation of weather information systems. No empirical studies have attempted to provide this kind of information for Canadian producers. In this study, a farm level production economics model was used to characterize the value of precipitation forecast information to alfalfa (Medicago sativa L.) dry hay producers in the province of Ontario. This model was applied to precipitation forecast data from the Windsor and the London Environment Canada weather offices for the crop years of 1994 and 1995. Four forecast methods are compared. Precipitation damage relationships during harvest are estimated using an agronomic simulation model combined with expert opinion. The value of weather forecast information was found to vary considerably between 1994 and 1995. For the region served by the London Ontario weather station, our results indicate that the Environment Canada daily precipitation forecast was worth $30.23 (Cdn)/acre in 1994 and $26.39 (Cdn)/acre in 1995. For the region served by the Waterloo weather station, the 1994 forecast was worth $36.08/acre. In contrast, producers who followed the 1995 forecast would have been worse off by $5.72 (Cdn)/acre relative to what they would have been able to earn by naively assuming that precipitation over the next 4 d would be the same as the last 4 d. The level of risk aversion of the producer was not found to be an important determinant of the value of weather forecast information. Averages of the range of values of weather forecast information obtained in this study indicate that precipitation forecast information is of considerable value to producers at critical times in the production process for alfalfa dry hay. Research Question Weather forecast information has been made available to Canadian farmers at no cost, or for a nominal charge, for many years. Recent budget problems have prompted the federal government as well as most of the provincial governments to review the package of services that they provide. In this context, the value of weather forecast information has come under scrutiny. No estimates of this value are available. This paper reports the findings from an application of a farm level production economics model for characterizing the value of weather forecast information for alfalfa dry hay producers in southern Ontario. Literature Summary The available empirical studies of the value of weather forecast information to farmers have generally used an ex ante or an ex post approach. The ex ante approach generally employs farm surveys or interviews. It attempts to measure what farmers indicate they would be willing to pay for weather forecast information services. The ex post approach generally uses farm decision models based on production economics. These models are used to assess how much better off a representative producer would have been if he had known what the nature of a future weather event, say the amount of precipitation that would fall during the harvest period, was going to be before the decision to begin harvesting was made. Study Description This study uses the ex post approach. A farm decision model that incorporated various levels of aversion to risk was developed. This model was applied to alfalfa dry hay harvest decisions in the face of uncertainty about the timing and amount of precipitation during harvest. Data for two locations in southwestern Ontario for 2 yr were used. Was precipitation forecast information valuable to alfalfa dry hay producers in Ontario? The findings of this study indicated that precipitation forecast information was of considerable value to alfalfa dry hay producers in southwestern Ontario in most of the situations that were considered. For the region served by the London weather station, the estimated value of Environment Canada's precipitation forecast was $30.23 (Cdn)/acre in 1994 and $26.39 (Cdn)/acre in 1995. For the region served by the Waterloo weather station, the estimated value of the precipitation forecast was $35.08 (Cdn)/acre in 1994, but following the forecast would have lost the producer $5.72 (Cdn)/acre in 1995. Does intensity of risk preference matter in deriving a value of weather forecast information? It is often thought that information is more valuable to more risk averse people. It has been demonstrated theoretically, however, that this commonly held intuition is not generally true. So it is an empirical question in any given situation if the value of information varies inversely, directly, or not at all with the level of aversion to risk. We used sensitivity analysis to study the effect of variations in the level of risk aversion, from low to moderate risk aversion, on the value of weather forecast information. We found that this range of risk aversion did not change the estimated value of precipitation forecast information in alfalfa dry hay production in southwestern Ontario.