The poster gets the proportions all wrong. You could argue that legislators indeed should be held accountable for the economic conditions which their legislations create. "Job growth," and "Economic activity" are splendid criteria, relevant criteria, germane criteria by which to evaluate the effectiveness of politicians today. More should be.
But they are of a diametrically different order than using students' test scores to assess teachers.
To accurately match the incongruity of assessing teachers' effectiveness by students' test scores, legislators should be evaluated for their effectiveness by the numbers of bushels of wheat that were grown in their tenure, or the number of clogged drains their constituents experienced, or the number of auto accidents occurring, or arrests made in their districts. Stuff over which they have NO control, influence or impact.
The SINGLE largest factor influencing the performance of students on standardized tests is the relative wealth of poverty of the students. Kids from wealthier homes perform better than kids from poorer ones. A kid's ZIP-code is a better, more accurate predictor of school success than grades.