Based on the Data

• By Paul Abramson
• 05/01/15

About 25 years ago, I received a call from a friend working in the U.S. Department of Education. His colleagues were in something of a panic because a new Secretary of Education had been appointed and among the first questions he had asked was, “how many new schools do we build in the United States in a year.” No one in the department had an answer that was more than a guess. My friend wondered if I could provide an answer based on the information I had been gathering in my annual national survey of school construction.

I started compiling and publishing annual data on school construction in 1974 (all the while suggesting that this was the type of information that the Department of Education ought to be collecting and disseminating) so, although the number of schools being built was not my focus, I had data that I thought might allow me to make a reasonable estimate.

The data that I had was the total estimated dollars spent on constructing new schools in a given year. To get an estimate of what a new school cost, we asked architects around the nation to estimate the cost of a new school in their area. We took these estimates to establish a base price for an average new elementary school or high school and then did some educated guessing about the percentage of new schools that were elementary or secondary. It was hardly an exact science, but based on our data, information and best guesses, we provided the department with a good “guestimate” of the number of new schools being constructed that year.

I never received any acknowledgement, but noted a few months later that the then-Secretary of Education was able to tell Congress how many new schools had been built the previous year and, remarkably, it was the number I had worked out with friends in the department.

Since that time, I have continued to estimate the number of new schools constructed each year, but now with a little more structural backbone. That backbone comes from dividing the nation into twelve regions and, within each region, estimating the dollars spent and the median cost of new elementary schools, new middle schools and new high schools. The accompanying table shows my estimate of the number of new schools constructed each year since 2000.

Thus, last year, we had estimated that $112 million had been spent on new elementary schools in New England, and that the median cost per school in that region was almost $37 million. Based on that, I estimated that three new elementary schools were opened in New England last year. I did the same for middle and high schools and for 12 regions of the nation then added up the regional figures by building type to get a national figure. Because we had also estimated spending by school type on a national basis — and because the samples in some regions were small — I also did a national estimate of schools constructed without reference to regions or regional cost factors.

The bottom line for 2014 is that, based on a region by region count, I estimate that 279 new schools were opened last year, including 145 elementary, 62 middle or junior high and 72 high schools. Looking at national figures without reference to regional differences, it appears as many as 296 new schools were opened, including 156 elementary, 67 middle and 73 high schools. For simplicity, each year I average the national and regional totals to get a cohesive estimate. In this case, that 288 new schools were opened in 2014.

As the accompanying table shows, we are building far fewer new schools today than we were building at the beginning of the century, but the cost per project has gone up significantly. That will make interesting fodder for a future column. Meanwhile, however, since the federal government still is not producing information on school construction, let’s make these the official estimates of new school construction since the beginning of the century.

This article originally appeared in the issue of .