An increase in education funding is a popular measure because it is thought that an increase in funding will beget an increase in graduation rates. Demonstrably, this is a logical inference. An increase in inputs will produce an increase in output. An increase in education funding will produce an increase in graduation rates. Granted, there are limits to this logic; according to the laws of diminishing marginal utility, an increase in inputs will eventually yield decreasing output. Eventually, maximum utility is reached. Therefore, education spending could eventually reach a level where increases would yield negligible increases in graduation rates. However, it is assumed that this level has not been reached. Moreover, the thought that an increase in education funding yields an increase in graduation rates presumes a strong linear relationship between education funding and graduation rates. Fortunately, if such a relationship exists, it can be discovered. If there is a strong relationship between education funding and graduation rates, then it would support the conventional wisdom that an increase in education funding is necessary to increase graduation rates.
Examining the linear relationship between education funding and graduation rates, if one exists, can be accomplished by applying a linear regression model to the data. To examine this relationship, a regression model was applied to the graduation rates and per pupil funding data from each state and the District of Columbia-South Carolina's graduation rate was absent from the data, so it was excluded from the model. The average freshman graduation rate data for the 2007-2008 school year was the latest available from the National Center for Education Statistics. And the per pupil funding data for the 2007-2008 school year was the latest available from the U.S. Census Bureau. However, preceding the evaluation of the regression model, an analysis of the funding and graduation rate data will be beneficial in constructing an overview of the issue; this will also allow Michigan residents to compare Michigan's graduation and funding rates to the national averages.
The mean freshman graduation rate for the 2007-2008 school year for the U.S. was 75.97%. Wisconsin had the highest graduation rate at 89.6%, and Nevada had the lowest at 51.3%. Michigan's graduation rate was 76.3%. It was marginally higher than the national mean. The mean K-12 per pupil funding for the 2007-2008 school year for the U.S. was $10,408.16. New York spent the most at $17,173 per pupil, and Utah spent the least at $5,765 per pupil. Michigan spent $10,069 per pupil. Similar to the graduation rate, Michigan only marginally deviated from the mean.
Now that the funding and graduation data have been reviewed, the regression model can be applied. First, if the conventional wisdom is accurate and there is a strong correlation between graduation rates and education funding, the model will display it; the model would yield a strong correlation coefficient. However, after applying the model to the data, it belies the conventional wisdom. The regression model yields a correlation coefficient of only .13. That is a weak correlation coefficient. This indicates that there is a weak relationship between education funding and graduation rates. This can be further demonstrated by plotting the data on a scatter plot-for convenience, Michigan's data plot is colored red. As is demonstrated on the graph, the slope of the regression line is minimal; minimal slope is consistent with a weak correlation coefficient which further evidences a weak correlation between education funding and graduation rates.
Consistent with its position to the national mean for education funding and graduation rates, Michigan is again average. Michigan adheres to the weak correlation and is placed nearly exactly on the regression line. For what it is worth, Michigan seems to be neither underachieving nor overachieving regarding its appropriations for education funding.
It is important to note the limitations of the data. The data is quite narrow and omits other variables that could have a positive correlation with graduation rates. Further refinement of the data and model is necessary to increase the accuracy and utility of the outcome. Nevertheless, the data is useful in determining if an increase in education spending will yield a linear increase in graduation rates. And as previously noted, the correlation is weak. Therefore, it can be inferred from the model that an increase in education funding will not yield a linear increase in graduation rates. Far from it, actually.
Now, this does not suggest that education funding is impertinent and should be ignored. However, it does suggest that other variables have a higher correlation with graduation rates and that a focus on increasing education funding may be wasteful and detract attention from more effective policy prescriptions. Therefore, as Michigan and the U.S. as a whole seek to increase the stagnant graduation rates, the focus on funding increases should be shifted to researching and implementing other measures that would have a stronger correlation with graduation rates.