| Summary: | Introduction: Industry reports and anecdotal evidence indicate that the death loss rate
in cattle feedlots has increased over time. Such increases in death loss rates impact
feedlot cost and thus profitability.
Objectives: The primary objective of this study is to examine whether feedlot death
loss rates in cattle have changed over time, to analyze the nature of any identified
structural change, and to identify possible catalysts for that change.
Methods: Data from the Kansas Feedlot Performance and Feed Cost Summary from
1992 through 2017 is used to model feedlot death loss rate as a function of feeder
cattle placement weight, days on feed, time, and seasonality in the form of monthly
dummy variables. Commonly used tests of structural change, including the CUSUM,
CUSUMSQ, and Bai and Perron methods, are implemented to examine the existence
and nature of any structural changes in the proposed model. All tests indicate the
presence of structural breaks in the model, including both systematic change and
abrupt change. Following a synthesis of structural test results, the final model is
modified to include a structural shift parameter for the period from December 2000
to September 2010.
Results: Models indicate that days on feed has a significant positive influence on death
loss rate. Trend variables indicate that death loss rates have increased systematically
over the period studied. However, the structural shift parameter in the modified model
is positive and significant for December 2000 to September 2010, indicating that death
loss is higher on average during this period. Variance of death loss percentage is
also higher during this period. Parallels between evidence of structural change and
possible industry and environmental catalysts are also discussed.
Conclusions: Statistical evidence does indicate changes in the structure of death loss
rates. Ongoing factors such as changes in feeding rations prompted by market forces
and feeding technologies may have contributed to systematic change. Other events,
such as weather events and beta agonist use could result in abrupt changes. No clear
evidence directly connects these factors to death loss rates and disaggregated data
would be required to facilitate such a study.
|
|---|