Cost Per Student As An Analysis Tool

The State of New Jersey, I and most analyzers of school district expenses use the cost per student number as guides when looking at the cost of education.

I have been part of many schol budget presentations. There is one comment that inevitably would arise. “Why are you using this cost per student stuff? It hides too much and I pay taxes on the total expenses anyway.” 

Many people believe the cost per student hides information. When used properly, it provides more information than most other measures. No one measure is best in all cases. My goal is to show how the cost per student gives context that no other method provides.

I use five charts and two data sets to make my case. I use the actual Barnegat enrollment for the 2003-04 to the 2010-11 school years. I also provide fictional total administrative costs for the same period. Administrative costs are made up to illustrate a situation where the cost per student is most useful. I will walk you through each chart and summarize. This methodology (charts and cost per student along with associated data) was used in my first blog. Charts and cost per student will be used in future posts, so I hope to provide a better foundation for and confidence in its use.

CHART 1: ASSUMED TOTAL ADMINISTRATIVE COSTS

This chart is a bar graph of costs. There appears to be a large increase between the first two years with smaller increases there after. The information is good, as far as it goes, but it is hard to get a sense of or context for the changes. We have no idea what caused the increases or how they compare to other costs.

CHART 2: ASSUMED TOTAL ADMINISTRATIVE COSTS, shown as percents 

This is the same information with the first year shown as 100 percent. The rest of the years are shown in reference to the first year as a percent. Notice the scales are different. Chart 1 starts at $3,000. This is done so the viewer may more easily see the changes.

The reader my not notice this and if their mind sees the starting point as zero, the changes look larger than they actually are. Percents give the chart perspective.  Year 2004-05 grew by 25 percent. The second through the seventh year increased by about 17 percent, or on average about 3 percent per year. 

Total administrative costs rose 42.4 percent in six years. The percent chart allows the reader to make judgements by making a few simple calculations in their mind.  Chart 1 shows only large numbers and does not allow for this. This is why you may see me use percentages in this way.  I will use it for my own analysis even if I do not show a chart.

CHART 3: ACTUAL BARNEGAT ENROLLMENT

This is the actual Barnegat enrollment in two sections. The section on the left is four years.

2003-04 is the last year where we had no high school students. In 2004-05 we had our first freshman class. In 2005-06 we had a sophomore and freshman class, and in 2006-07 we have a sophomore, freshman, and junior class. 

The four years to the right are the first four years Barnegat had all four high school classes. The enrollment growth is significant the first four years as we increased the population each year for the last three years. The enrollment stabilized and shrunk slightly once we had all four high school classes. 

This makes some analysis very difficult. Some expenses may rise to meet the needs of more students. You will see one time increases in things like books and other support costs to accommodate higher enrollment. Some ethically diminished politicians will say some normal expense increases due to enrollment growth represent bad management techniques when they reflect required costs due to enrollment increases.  We need tools to help us understand the truth.

CHART 4: ASSUMED ADMIN COSTS VS. ACTUAL BARNEGAT ENROLLMENT 

This chart combines chart 1 and chart 3. We now see the administrative cost in context. 

The reader still has to make many interpretations. Are administrative costs rising faster than enrollment? The two scales are not proportional so there is distortion in the increase in the line. I tried to make the scales proportional, so it is pretty close, but not perfect. I could have used the default scales given me by the software.  I would have grossly distorted the comparison between enrollment and administrative costs.  This would not be a purposeful distortion but just my trust in the software. The software may have a bias towards aesthetics that sacrifices scale legitimacy. This is a perfect example of unintentional distortion.

CHART 5: ADMINISTRATIVE COST PER STUDENT 

This chart shows the cost per student only. I may look at this chart and without looking at the other charts see the information in context. 

There is a significant increase between 2003-04 and 2004-05. This is because the administrative costs increased faster than enrollment. The chart does not tell you why (no chart can). 

My fictitious numbers accounted for setting up the infrastructure for the next three years of increased students. This is a wise thing to do. The unethical politician will call this a waste of money, but they should know better. There may be room for criticism, but the sole reason should not be that the number went up. We would want the number to go up, so we have the right number of staff to meet the growing enrollment needs we know are coming.

The cost per student drops for three years in a row (2005-06 to 2007-08). This is because the cost are paid before the new students arrive. Enrollment is increasing faster than the increase in administrative cost. Good management would anticipate this. The first year increase was getting the district ready for the new enrollment. The new enrollment brings the cost per student down, and the world is happy. 

The last three years show an increase in the cost per student. This is because enrollment has stabilized and normal year to year increases in costs are faster than increases in enrollment. Notice the cost per student in 2003-04 is the same as 2010-11. This was by design. The cost stayed the same over seven years, but I guarantee there there would be claims of incompetence by challengers during phases of increases and claims of divine intelligence by incumbents during years of reduction. The arguments are about nothing. 

SUMMARY 

I hope we learned two things: 1) When it comes to analyzing spending, cost per student is our friend, and 2) we may be given facts by politicians, but it takes some work to determine the truth.  

Our elections are too often fought over things that do not matter. We would be better off if we selected politicians who did the right thing as they see it, even when we do not agree with them. We just went through a school board elections where both sides gave us facts, but did they give us information? We soon will know who told us the truth.