Value-added measures of teacher effectiveness have become very popular in educational reform efforts. Using 2010 data from elementary schools in the Los Angeles Unified School District (LAUSD), we engage in a new and innovative statistical approach to contribute to the discussion concerning such measures. We hypothesize that quality teachers are not independent of students and schools. Through two-stage least squares, we find that in the aggregate, high-quality teachers do not improve aggregate student outcomes in urban districts controlling for student and school factors. While good teachers can have individual, positive, and dramatic effects on individual students, their impacts on larger populations are diluted. Therefore, although high-quality teachers can improve individual student outcomes, other factors are more likely to create a more equitable, effective system of education.