A binding constraint is a constraint used in linear programming equations whose value satisfies the optimal solution; any changes in its value changes the optimal solution. Once an optimal solution is obtained, managers can relax the binding constraint to improve the solution by improving the objective function value. Managers should not tighten the binding constraints as this worsens the value of the objective function.
Constraints whose changes do not affect the optimal solution are called nonbinding. The shadow price is the amount associated with a unit change of a particular constraint. Nonbinding constraints have a shadow price of zero, while binding constraints typically have other shadow prices than zero.
Each variable within the objective function must be represented in the constraints, including those that are not explicitly specified. Linear programming equations typically use deterministic objective functions, but they apply sensitivity analysis in their real world application. Sensitivity analysis examines the sensitivity of the optimal solution to changes in its parameters as reflected in the constraints report and the changing cells report within Excel. The 100 percent rule states that the values of variable coefficients of an objective function may change without affecting its solution if the deviation is less than 100 percent.