The policy side of the answer is that they don't need to be. We have standardized course outcomes, so the state has specified which outcomes you should assess, but not how. Alice comments:
The assessment needs to measure the level of learning about the subject matter as stated in the course competencies.This could allow for differing assessments from course section to course section, as long as they are measuring the same outcome and the same level as indicated by the verb in the competency. (For example, if the verb is "demonstrate," a strict concept based multiple-choice task wouldn't cut it.)
We commonly use pooled questions on multiple choice exams, so even students within sections taking those exams are not evaluated based on exactly identical assessments, although they all face the same assessment method.
My instantaneous reaction to the question was that we would not want to require all faculty teaching the same class to use the same assessments, however I think we can make a case for some sharing of assessments (that isn't based on students complaining to me that their best friend in the other section has different/easier homework assignments). Good assessments are very difficult to design and most of us don't have a background in assessment design. If we share the development time we can all have better assessments without killing ourselves. We can also update the assessments or consider developing a few pools of assessments in order to avoid students sharing with each other in ways we don't feel are appropriate.
As usual in these discussions we end up trying to walk the line between sharing the workload and sharing the costs of development. :^)