Crop Yield Distributions: A Reconciliation of Previous Research and Statistical Tests for Normality

Abstract

This study revisits the large but inconclusive body of research on crop yield distributions. Using competing techniques across 3,852 crop/county combinations we can reconcile some inconsistencies in previous studies. We examine linear, polynomial, and ARIMA trend models. Normality tests are undertaken, with an implementable R-test and multivariate testing to account for spatial correlation. Empirical results show limited support for stochastic trends in yields. Results also show that normality rejection rates depend on the trend specification. Corn Belt corn and soybeans yields are negatively skewed while they tend to become more normal as one moves away from the Corn Belt.A gricultural economists have long investigated yield uncertainty because this fundamental attribute of agricultural production profoundly affects economic decisions such as crop acreage allocations, input use, and crop insurance demand and rate making. While this vein of literature is extensive, it has failed to produce a consensus regarding the appropriate approach to crop yield distribution specification.Several studies have rejected the assumption of normally distributed crop yields. In addition, some studies conclude that crop yields are positively skewed while other studies conclude that they are negatively skewed. Because these studies use a wide variety of data, normality tests, and detrending techniques, one may conjecture that the inconsistency of results is caused by the techniques used.

Crop Insurance Premium Ratemaking Based on Survey Data: A Case Study from Dingxing County, China

“…7, the three probabilities' average values of all farmers are 0.1672 (low yield), 0.7059 (average yield), and 0.1269 (high yield), which indicates that the probability of getting below average yields is greater than that of getting above average yields. It is consistent with the research results of Gallagher (1987) and Harri et al (2009) that the yield distribution is negatively skewed. This is also consistent with the survey result.…”

Yield Responses to Planting Density for US Modern Corn Hybrids: A Synthesis‐Analysis

“…Comparable yield distribution for the entire meta-database herein analyzed was previously documented ( Just and Weninger, 1999;Claassen and Just, 2011;Assefa et al, 2014). However, there are studies which argued that crop yields (at county scale) were not normally distributed, but instead were negatively or positively skewed (Harri et al, 2009;Hennessay, 2009). A small skewness to the negative on both the 2000-2010 ('training data') and the 2011-2014 ('validation data') datasets were evident but not highly deviated from the normal distribution (Fig.…”

Spatial Pattern of Yield Distributions: Implications for Crop Insurance

“…Day 1965;Gallagher 1987;Nelson and Preckel 1989;Taylor 1990;Moss and Shonkwiler 1993;Tirupattur, Hauser, and Chaherli, 1996;Ramirez 1997;Wang et al 1998;Goodwin and Ker 1998;Just and Weninger 1999;Stokes 2000;Chen and Miranda 2004;Ker and Coble 2003;Chen and Miranda 2006;Harri et al 2008;Zhu et al 2008;Zhu et al 2009). A variety of yield modeling approaches have been proposed, including parametric, semiparametric (Ker and Coble 2003), and nonparametric (Goodwin and Ker 1998) methods.…”