Do farmers waste fertilizer? A comparison of ex post optimal nitrogen rates and ex ante recommendations by model, site and year

While there appear to be costs to farmers of over-applying nitrogen, there is evidence in many regions that farmers are applying nitrogen at levels that exceed those suggested by government extension services. A major reason why farmers would apparently waste money by applying more fertilizer than a crop can use is a perception that the general recommendations are not appropriate for their individual situations. In this paper we estimate the economically optimal rates for 7 Ontario sites over multiple years using four yield response functional forms to examine whether the profit-maximizing rates as determined at the end of the growing season (ex post optimal rates) are generally higher than recommended rates due to site, year, or yield response functional form differences. There was no statistical or economic basis for selecting one response model over another suggesting functional form choice or perception is not a reason for over-application. However, there was a great deal of variability found between the actual optimal rate for the season and the ex ante recommended rate, which is constant across seasons for a given site. While the recommended rate is higher than the maximum economic rate of nitrogen (MERN) on the majority of sites examined, the distribution is skewed due to a few large differences. When the recommended rate is lower than the ex post MERN in a given year, it tends to occur on less productive sites and is much lower. The pay-off function to alternative nitrogen rates is generally flat as the difference between the MERN and recommended was less than $10/ha on approximately one-third of the trials but there are large differences in the good years on less productive sites. Thus, the decision to apply more than average to take advantage of the good years is appropriate since the cost of over-application is low compared to the cost of under-application. While the pay off to soil testing for nitrogen and consequently variable rate application technology is brought into question, there appears to be significant value to information on the growing conditions for the upcoming season particularly on less productive sites. Another implication of the study is the need to have a sufficient range of application rates in nitrogen field trials for accurate estimation of the underlying response function.