Spending data for the 2005-2006 school year can be found at the National Center for Education Statistics. Data for the total number of students, the total number of FARMS students (free and reduced meals, an indicator of poverty), and the number passing the Maryland math assessment in 2008 can be found at the School Improvement in Maryland web site.

Here are the basic data for Montgomery County and for an average of all counties in Maryland (in each instance, I list Montgomery County first, followed by the average and standard deviation for all counties.)

proportion of non-FARMS students: .725, .697, .169

proportion passing math MSA: .819, .779, .080

expenditure per student: $14,373, $10,768, $1,283

In terms of rank, Montgomery County is number 11 out of 24 counties both in percent non-FARMS and in math. It is number 1 in terms of spending.

Montgomery County is 0.4 standard deviation above the state average for non-FARMS students, 0.5 standard deviations above average for passing the math assessment, and 2.8 standard deviations above average for spending.

The Maryland county that performs the best is Worcester County, which has a below-average proportion of non-FARMS students (.640) and the highest passing rate in the state (.882). Worcester has the second-highest spending rate in the state.

Cecil County has an above-average non-FARMS rates and a below-average passing rate. It is a low-spending school district, at $9,309 per student. On the other hand, Caroline County has a below-average non-FARMS rate and an above-average passing rate, and it spends only $9614.

Overall, the spending data are skewed. Montomery spends $14,373, Worcester spends $14,015, Howard County spends $12,960, and Kent County spends $11,731. All 21 other counties spend between $9309 and $11082.

The percent of non-FARMS is .206 in Baltimore city. Otherwise, there are no outliers, although the range goes from .399 in Somerset to .883 in Carroll.

The regression line for predicting the percentage of students passing the MSA based on the percentage non-FARMS has a slope of .376 and an intercept of .532, with a standard error of the regression of .049.

Montgomery County lies slightly above that line (meaning that the percent passing is slightly above what would be predicted based on the the percentage of non-FARMS students). The positive residual is .015, which is less than one-third of one standard error in the regression. Essentially, Montgomery County is right on the regression line.

Relative to the regression line with percent passing as the dependent variable and percent non-FARMS as the explanatory variable, 12 counties are above the line. Of these 6 are high-spending counties and 6 are low-spending counties. However, given that Worcester is way above the line and Montgomery is slightly above the line, my guess is that the statistical relationship between spending and school performance is positive, although not strongly so.

We want charts!

I suggest looking at the pass rates for only the FARM students to ascertain if the correlation still holds.

Does this include spending on athletics? How many pro athletes has Montgomery county manufactured over the years compared to the state average?

Isn't spending per pupil in large part a factor of the land and thus the labor costs in MontCo?

Q: Assuming that spending is the input and exam results are the output, then a _lack_ of correlation between spending and results tells us _what_ about the education system?

Here is the data on the MCPS budget:

http://www.montgomeryschoolsmd.org/departments/budget/FY2009/pdf/FY2009BudgetSummary.pdf

2007 Actual total spending was $1.8 billion. $1.25 billion of that was salaries and wages (20,739 positions, avg. $60k/position). $795 million went to "professionals", $291 million to "supporting services" and $80 million to "administrative". I'm still not sure how much went to schools and how much went to central administration though...

"Insurance & fixed charges" were the next largest expenditure ($374 million), followed by supplies & materials ($71 million), "grants & other" ($56 million), and utilities ($40 million).

You will also learn that MCPS used Value Stream Mapping (VSM), Process Design and Improvement Process (PDIP), and has also adopted Six Sigma!

Arnold, isn't what you're trying to show the same as doing a regression of spending vs test scores controlling for FARMS? If after controlling for FARMS, the effect of spending on test scores is statistically insignificant, then doesn't that make your point?