The figures indeed suggest that more land was developed between 1992 and 1997 than in the previous five years. But a close scrutiny of the data also show that some of the fastest land development was taking place in states that supposedly have controlled their sprawl through smart-growth planning.
Data for urban and built-up areas for 1992 and 1997 are in appendix one of the report. According to these data, 97 percent of the increase in developed land between 1992 and 1997 was in the urban-and-built-up category rather than the rural-transportation category.
However, Gore's press release says that "the loss of farmland is no longer centered predominantly around major metropolitan areas, but is affecting growing numbers of small- and mid-sized cities." This suggests that the 1997 inventory counted rural developments as "built-up areas" when previous inventories would have counted them as "rural" or "rural transportation."
One check on the validity of the urban-and-built-up numbers can be found in the Federal Highway Administration's annual Highway Statistics books. Table HM-72 lists the area of land in every U.S. urbanized area. In 1997, the total area of all metropolitan areas (outside of Alaska) was about 54.2 million acres. This is two-thirds of the 80.8 million urban -and-built-up acres counted by the Natural Resources Inventory.
Unfortunately, the 1992 edition of Highway Statistics only includes urbanized areas larger than 200,000 people, while the 1997 edition goes down to 50,000 people. The 200,000-and-larger areas hold 84 percent of the people and occupy 75 percent of the land of the 50,000-and-larger areas.
Between 1992 and 1997, the land area of urbanized areas larger than 200,000 people grew by just 13 percent (including several areas that grew to that size in the meantime). By comparison, the urban-and-built-up land in the Natural Resources Inventory grew by 24 percent. This confirms that much of the land newly counted as developed is around urban areas of less than 50,000 people.
It seems hard to believe that the administration would direct the Natural Resources Conservation Service to tinker with the data just to support its war on sprawl. But there are two indications that this is precisely what it did.
First, prior to releasing the final report, the Natural Resources Conservation Service's web site presented preliminary results from the 1997 inventory ("Natural Resources Inventory: 1997 State of the Land Update"). According to this paper, the inventory found less than a 0.5 percent loss of croplands, most of which went into pasture, range, or forest, not urban areas.
Yet the final inventory data reported a 1.9 percent decline in croplands. The final data also showed declines in pasture land, range land, and conservation reserves and only slight increases in forest land and "other rural" land. This leads to the conclusion that most of the crop, pasture, and range lands were urbanized. This completely contradicts the preliminary report. Naturally, the preliminary report has been deleted from the NRCS website but hard copies are still available from the agency.
Second are some other peculiarities with the data pointed out by David Riggs of the Competitive Enterprise Institute. For the comparison of 1992 with 1997 data, the 1997 NRI reduced the amount of 1992 developed land by 3 million acres nationwide. This downward revision accounts for nearly 20 percent of the "growth" in developed lands in the 1997 inventory. Meanwhile, the 1997 report increased the amount of developed land in 1987 by 5 million acres.
Original Reports 1997 Report
1987 77 82
1992 92 89
1997 105 105
The
original data suggest that developed lands grew by 13 million acres from 1992
to 1997, less than the 15-million acre growth from 1987 to 1992. But the
revised data show developed lands growing by 16-million acres from 1992 to
1997, twice as fast as the 7 million acres from 1987 to 1992. This change in
the data allowed Gore to claim that "the conversion of farmland and other open
space to development more than doubled in recent years" and that therefore
federal action was needed to "preserve farms threatened by sprawl."These apparent political manipulations of the data reduce the credibility of the Natural Resources Conservation Service. A better idea of the exent of recent urbanization will be gained when the Census Bureau completes the 2000 census. We have to hope that those data are not politically manipulated as well.
The results do not look good for smart growth. If smart-growth policies such as urban-growth boundaries, comprehensive land-use planning, and open-space initiatives worked, then urban areas in the states that have adopted such policies -- primarily Oregon and Florida, but also California -- should have increased their population densities, or at least not decreased them as much as comparable states without such laws. But that's not what the numbers show.
Nevada, Arizona, and Colorado are all reputed to contain sprawling metropolises. No wonder: Unlike Oregon, to which they are roughly similar in size and population, they don't have any growth control or smart-growth laws. Yet these are the only three states whose urban densities actually increased between 1992 and 1997.
Meanwhile, Oregon is famous for its strong land-use planning. It effectively adopted smart growth in the early 1990s, when the land within its urban-growth boundaries was getting full. Rather than expand the boundaries, Oregon cities decided to rezone land within the boundaries to higher densities. Increasing land prices also encouraged people to build on smaller lots. Yet Oregon's urban densities decreased by 1.7 percent between 1992 and 1997.
In the South, Florida is a fast-growing state that has also adopted many smart-growth policies. In comparison, Texas is well known for its sprawling urban areas such as Houston and Dallas-Ft. Worth. Yet Florida urban densities fell by 2.5 percent whereas Texas' fell by just 1.6 percent -- less even than Oregon's.
Unlike Florida and Oregon, California has no statewide growth-control laws. But many of its cities have local adopted such laws, including urban-growth boundaries and other growth limits. California urban densities fell by 2.1 percent, nearly a third more than Texas', which it resembles in population and size.
This isn't hard proof that smart growth doesn't work. But it confirms calculations done with other available data. It also suggests that Vice-President Gore should not be so glib about urging cities to adopt smart-growth policies as a way of curbing sprawl.
The roughly 800,000 sample points are used to represent the United States as a whole. The USDA estimates that the results are statistically valid at the state level but may be questionable at the county level. The errors can also be high when dealing with smaller land classifications, such as the urbanized category.
The USDA says that the national estimate of urbanized land is within plus or minus 0.6 percent of the true total (standard error at the 95 percent confidence interval). For the individual states the average standard error is plus or minus 5.5 percent. For a few states, such as Arizona and Montana, the error is well over 10 percent. Of course, this says nothing about possible changes in methodology between the 1992 and 1997 inventories.
In the Rust Belt cities of the East and Midwest a different process is taking place. Many of the central cities are actually declining in population. The suburbs may be growing, but more slowly than the suburbs of the West and South. Thus, slow population growth is accompanied by a more rapid growth in developed land.
It is likely that these older inner cities must decrease their populations before they can attract gentrifiers. Such depopulation followed by gentrification is already taking place in Cleveland and a few other Rust Belt cities.
The decline in croplands between 1987 and 1992 was 24 million acres, while the decline from 1992 to 1997 was only 7 million acres. Between 1992 and 1997, the number of acres of forest lands actually increased.
Between 1987 and 1992, 20 million acres was taken up by the USDA conservation reserve program, in which the federal government effectively pays farmers not to farm. Between 1992 and 1997, the decline in the total number of acres in farms, including pastures, croplands, and rangelands, was less than the shift from croplands to the conservation reserve program between 1987 and 1992. So which causes the greater erosion in farming -- land development or federal subsidies to stop farming?
The USDA admits that "conversion of agricultural land does not threaten America's food supply." But Gore worries about "the elimination of vital open spaces." How fast is open space disappearing and how much is left?
According to the inventory data, 95.4 percent of the U.S. (not counting Alaska) was open space in 1992, falling to 94.6 percent in 1997. At this rate -- and the rate is probably exaggerated -- more than 75 percent of the U.S. will still be open space in 2100. This is hardly cause for alarm.
Of the total land area, 3.4 percent was "urban and built-up" in 1992, increasing to 4.2 percent in 1997. According to the 1997 Highway Statistics, only 2.8 percent of the nation's land is in metropolitan areas of 50,000 people or more, so the remaining 1.4 percent must be in small towns. Again, these figures are hardly a cause for alarm.
Suburbanization represents peoples' desires to leave the high-density central cities, especially those in the Rust Belt, and move to lower density suburbs, especially those in the Sun Belt and the West. Suburbanization is no threat to the nation's farms, forests, or open space. No federal intervention is needed except to eliminate federal subsidies that promote both high- and low-density.