Tricky numbers and school construction costs
You can find the data behind Sunday's construction costs story with the N.C. Department of Public Instruction (click on "construction costs" to the left). Note: I had to revise the 2005 averages in the spreadsheet because the Northern numbers listed are inaccurate.
When I started working on the series about school construction in Guilford County, one of the questions I had was "why did Northern High cost so much more than Reagan?" The reason? From board meetings, to blogs, to phone calls, I have heard people compare the two. I found in my research that there really is no "apples to apples" comparison, but you can come close by looking at inflation-adjusted cost per square foot and cost per student (a bit more difficult to come by).
When you look at the state comparison data, you can find reasons to criticize many school districts, not just Guilford. It just depends on the angle you take. When you take out projects in Wake, Mecklenburg and Guilford -- the state's largest school systems, cost averages come down. (but then you have to wonder, how valuable is data that omit school systems with the largest student populations and the most schools built?)
At any rate, advantages and disadvantages exist to comparing total building costs, per square foot costs and costs per student:
Total building costs: Tells you up front how much it cost to build a school, but doesn't include pieces that are harder to compare, such as furniture, equipment and soft costs. Some reasons to not compare total costs: One school system may benefit from buying all its furniture in bulk through one contractor as opposed to buying it per project. Another district may decrease its soft costs by working with one architect on several different schools or using a prototype design. The downside to total building costs is you can't really compare a 200,000 square foot school to a 80,000 square foot school.
Building cost per square foot: Allows comparisons in that you can better control for disparate building sizes and student populations (for example, a school may be built to house a disproportionate number of students with disabilities, which requires smaller class sizes). The downside to cost per square foot is that you can have schools with the same cost average, giving the appearance of equity, but one building can be more wasteful in space than the other.
Cost per student: Helps make up for the weakness in cost per square foot by taking into account wasteful space (i.e. two schools may have the same cost per square foot, but one devotes more square footage to students than needed). The downside to cost per student is that school districts use different formulas to allocate space and this may give a false representation (i.e. 30 students per classroom versus 25). Another issue is some students by law require more space than others (i.e. classroom wing for students with autism or the state's move to lower class sizes in K-3). The other downside is that student enrollments fluctuate, so a school may have been built to originally house 1,000 students but because of enrollment or programmatic changes, the population could shift to say, 900 students, or 1,100 students. That would change the cost per student, while the cost per square foot would stay the same.
Comments (8)
To report abuse of the comment feature on this site, please use the feedback form at the bottom of any page.
Morgan:
This was a great article that showed that you really need to look carefully at two projects before deciding that one is unfairly priced. You've done a great service.
That said, the construction cost price index that you used seems awfully high. I can't find anything that suggests that the deflator should have been 50 percent.
For instance, the producer price index for "intermediate materials, supplies and components" for construction rose from 151 in 2000 to 182 in 2005 (see http://www.gpoaccess.gov/eop/tables06.html, table B-65), a 20 percent increase over 5 years.
Without the 50 percent deflator, you end up with much different costs between the two schools. What was the source of the deflator?
Posted on October 1, 2007 2:10 PM
Mr. Ribar,
I adjusted based on the average per square foot costs for high schools bid in 2003 through 2005 (average per square foot costs were $105.12 for 2003 and $157.08 for 2005). To make sure that this method was not way off, I asked Bill Powell, facilities manager with Forsyth schools if he agreed with Joe Hill that Reagan would have cost closer to $150 per square foot if the school was bid in 2005 instead of 2003. Powell said yes he did, and that the system had actually documented per square foot costs of $131 in early 2004.
Interesting to note: According to DPI, general construction costs alone (not including HVAC, electrical and plumbing) were $67.96 per square foot in 2003 and represented 73 percent of construction costs. In 2005, those costs were $104.71 and represented 76 percent. That is a 54 percent increase. You can find the table on www.schoolclearinghouse.org under prime construction costs. Regarding the table you referenced, I'm not sure how North Carolina stacks up in those numbers.
Also interesting: According to School Planning and Management, the national MEDIAN high school now costs $26,111 per student, $151.52 per square foot, with 166.7 sq ft provided per student. Around 2003, those numbers were $19,438, $122.16, and 154 sq ft, respectively. Unfortunately, because SP&M reports in medians and the state reports in averages, I could not compare the two.
Bottom line is: You can only get so far comparing school construction costs. I think both taxpayers and school officials will be better served by getting on board about what they want to see in future schools. GCS facilities staff and a construction advisory committee are working on developing comprehensive design guidelines now. They need to communicate those guidelines effectively to the overall community.
The other question is this: who will be best and least served if a) the proposed school bond passes and b) the bond fails.
Posted on October 1, 2007 3:47 PM
Morgan:
A concern with that approach is that it leads to the same kinds of apples-to-oranges comparisons that you discuss in the article. There were only a few high schools built in each of those years (5 per year?). A few outliers could lead to very different schools being compared. Moreover, wouldn't those cost averages include the very schools that the article is comparing?
When I looked at the average state figures, it appeared that costs actually decreased from 2002 to 2003 and from 2005 to 2006, which doesn't seem right.
The price index that I referenced was a national price index. As you said, that might not be an accurate indicator for NC. However, in the article, you pinned much of the construction cost increase on Katrina; so, if there was a big jump, we should see it in the national figures.
Your bottom line is correct. It's hard to compare costs across locations and years.
Posted on October 2, 2007 8:07 AM
this is off-topic, but does anyone else just hate this new format? Only 1/3 of the page is the blog, 1/3 advertisements and then 1/3 other blogs! Does that have to be so large? It seems I have to scroll forever to get to the blog itself. I understand that they have to have advertisements, but does the other blog info have to be located where it is and as large as it is....
Posted on October 2, 2007 9:20 AM
Mr. Ribar,
I admit the adjustment is not perfect, but I would say most, if not all, inflation figures need to be taken with a grain of salt (the core inflation index comes to mind, as it does not include food or gasoline). It would be great if we were literally comparing apples. An apple is an apple is an apple, for the most part. Not so with schools. But I think the state figures paint a decent picture of how costs are now.
Posted on October 2, 2007 11:20 AM
Morgan:
Your last statement seems to undercut the entire point of your article. The article, correctly and carefully, made the point that you can't compare the costs of one high school to another without considering all of the features of each school. However, to adjust for inflation the article then uses average costs from 4 schools in 2003 to 5 schools in 2005 adjusting only for square footage.
In 2003, the school districts that built high schools were Forsyth (2), Lee, and Union. The largest district was Forsyth, and they might have had economies of scale from building two schools at the same time. They were also the low-cost builder in that year.
In 2005, the school districts that built high schools were Carbarrus, Chapel Hill-Carrboro, Guilford, Mecklenburg, and Union. Chapel Hill is a relatively affluent district (not surprisingly its high school had the highest costs per square foot). Mecklenburg and Guilford are relatively large districts.
Your post states that, "When you take out projects in Wake, Mecklenburg and Guilford -- the state's largest school systems, cost averages come down." Yet you end up comparing data from one year that includes two of these school districts to another year that includes none of these.
The growth in the national median figures that you cited seems more in line with the producer price index numbers (growth of 25 percent). As I said initially, 50 percent just seems very, very high.
Dave
Posted on October 2, 2007 2:55 PM
Debora, I agree this format is horrible.
As many of you know I've ranted and raved on this blog about Bush's NCLB legislation which might be losing support on the Hill. Anyway, did anyone chuckle when the President recently stated while commenting about education,
"...this shows that children's do learn......."
And this is the guy who wants to reform education in the USA.
Lord hear our prayers
Posted on October 4, 2007 1:43 PM
The main problem with the article is its enabling premise: that cost per sq ft is the best critierion for comparison. That suppresses or ignores entirely a crucial question: why does a school for 10% more students need 50% more space? The "downside" you identify above is almost scrupulously avoided in the article. There's a passing reference to "community standards," but smaller classrooms and no atrium are fine by me, and I'm a member of the community.
Posted on October 12, 2007 11:52 AM