Delay Of Inventory Leaves Sprawls Pace In Question
By John Nagy, Staff Writer
Federal data released nearly a year ago that shows an alarming state-by-state disappearance of open space remains unverified after the discovery of a computer error in March cast doubt on its validity.
The Natural Resources Conservation Service ( NRCS ), a division of the U.S. Department of Agriculture that oversees the land use survey, initially projected that it could take weeks to deliver corrected numbers. But federal officials now say state and local planners and others waiting for the new data will have to keep waiting at least until the end of the year.
"Mid-December is not a guarantee, but I feel fairly comfortable that we will reach it," said Warren M. Lee, who directs the resources inventory program for NRCS.
NRCS touts the Natural Resources Inventory as "the most comprehensive database of its kind ever attempted anywhere in the world."
Launched under Congressional mandate in 1982, it gathers data from over 800,000 statistical sample areas on the growth and decline of cropland, pasture and rangeland, forests, wetlands and new development. No one disputes that road building and construction are quickening, but the glitch leaves open the question of the rate at which it is happening.
Last December, Vice President Al Gore and Agriculture Secretary Dan Glickman used the data to announce the loss of nearly 16 million acres nationwide between 1992 and 1997, an area slightly larger than West Virginia.
At 3.2 million acres per year, developers appeared to have clear-cut and steamrolled at more than double the pace of the previous ten years, bringing the total amount of U.S. land covered by cities, highways, parking lots, strip malls, office parks, residential subdivisions and other signs of humanity to 7.1 percent. mobilized anti-sprawl forces and provided the grist for debates in community meetings and legislatures over policies intended to shape development more efficiently.
The figures confirmed Texas' standing as the nation's top land-eater since 1982, converting an average of 381 square miles of open space into developed land each year. Pennsylvania, Georgia, Florida and North Carolina rounded out the top five.
Only four states Hawaii, Rhode Island, Vermont and Colorado successfully slowed the speed of growth. depicted the spread of development with blood-red hexagons, each representing the development of over 50,000 acres.
The largest of the ruddy smears appeared around Atlanta, Philadelphia, California's San Joaquin Valley and along Florida's Atlantic coast.
Intense development has also taken place in and around Dallas, Houston and Brownsville, Tex., Albuquerque, N.M, Minnesota's Twin Cities, the western suburbs of Chicago and Pittsburgh, southwestern Michigan and northern Indiana, Detroit, Indianapolis, central Ohio, portions of Alabama, Kentucky and Tennessee, South Carolina's Greenville-Columbia corridor, Charlotte, N.C., and the Research Triangle, Long Island, Boston and New Hampshire.
But all of these findings are subject to change. NRCS continues to post the "preliminary" 1997 data on its web site with a prominent disclaimer and occasional updates on the status of the corrections being performed by statisticians at Iowa State University.
According to Warren M. Lee, a statistical computer program "double counted information in certain situations."
As soon as the error was identified, NRCS strengthened its quality assurance system. Lee says the long wait for final figures is due to the difficulty of verifying detailed information for all fifty states. While he declined to predict the size of the anticipated adjustments, he rejected the idea that the corrections would seriously alter the inventory's original findings.
"There'll be some changes, and we know that the original estimate was greater than what the new numbers will show." But changes will differ from one state to the next, he said.
No data is yet available for Alaska.
While the flawed data has fallen short of halting land use research in its tracks, it has forced NRI customers to get used to qualifying their work.
"We have not used (the data) in an official publication, because we want the new revised estimates to rely on. But I certainly use them in presentations as evidence of the upper bound," said Dr. Samuel Staley, an economist and self-described smart-growth skeptic who directs the Urban Futures program of the Los Angeles-based Reason Public Policy Institute.
"We are concerned about it. We would like the data," said Betsy Garside of the American Farmland Trust , which uses the inventory as one piece of its comprehensive resources on land trends in agriculture. The NRCS error has stalled updates to AFT's 1997 "Farming on the Edge" map , which "showed areas where farmland was at risk of development."
The American Planning Association is using the NRI data "to look at trends from 1987 to 1997 in terms of development and the number of people per acre," said spokesman Denny Johnson. That information will set the backdrop for an APA-commissioned survey of sprawl's impact on voter's behavior in 2000 that APA plans to release in December.
Staley, author of a study entitled The Vanishing Farmland Myth and the Smart-Growth Agenda said errors and delays of this kind are commonplace for statisticians. But he contended that continued postponement of a downward revision in the numbers works in Gore's favor as the Democratic presidential nominee. "If it wasn't for the election, very few people would question whether this was a politically-motivated delay," he said.