Parameter prediction error:
This is another aspect of faulty planning. As Hongren says, ‘planning decisions are based on predictions of future costs, future selling price, future demands and so on. In many cases there will be a difference between the actual value and the predicted value’. Such differences are not only due to uncertainty about the future: the predictions may not have taken proper accounts of conditions existing at the time when it was made, like a recently agreed pay rise, or an agreement to increase wages in three months time.
Inappropriate decision models:
Variance can arise when chosen decision model fails to capture important aspects affecting the decision. The solution to a linear programming model can be used when setting standards for direct material purchase prices. These standards, however, may be inappropriate if the LP solution is not feasible because the LP models fail to recognize a constraint on labor availability or storage capacity’. (It is the relationship between the variables that causes the problem here, not the failure to predict accurately.)
Randomness of operating processes:
A standard is an average figure: really it represents the mid-point of a range of possible values and therefore individual measurements taken at specific times will deviate unpredictably within this predicable range.