Search this site:

3.2 Land Use Forecasts

To forecast land-use change to 2020, we employed a county-level model developed by Hardie and others (2000). This econometric model assumes that:

The allocation of land between urban and rural uses is driven by population density, personal income, and housing values.

The allocation of rural land to agricultural and forest uses is driven by returns to local crops, returns to grazing, agriculture costs, land quality, and timber prices. All of these variables except timer prices are defined at the county level of resolution. Timber prices are defined for two or three subregions per State defined by the Timber Mart South price reporting service.

The model was estimated based on land-use patterns recorded in 1982, 1987, and 1992 by the National Resource Inventories. See Hardie and others (2000) for modeling details. Detailed land-use categories were lumped into four classes: urban/residential, cropland/pasture, forest, and other. The urban/residential class includes areas in transportation, and other corridors. The "other" class can be considered a transitional zone where land use is unclear due to changing conditions.

Before land uses could be projected, we had to forecast the factors that drive changes. Accordingly, we acquired county-level forecasts of population density and personal income and developed forecasts of housing values.

Two core projections were developed to (1) isolate the influence of general economic and population growth on the region and (2) Completely assess land-use changes that account for market responses to increased scarcity of timber as rural land is developed. The two core projections were defined for the following scenarios:

Base scenario. An initial scenario was developed assuming that the population, income, and housing value forecasts are correct and that the relative positions of timber and agricultural markets do not change in the future. Effects of population growth and economic growth on urban land uses are estimated.

Market scenario. A scenario was also constructed to evaluate how rural land uses might be influenced by a relative shift in returns to agricultural and timber management. This scenario assumes that the population and economic-change forecasts in the base scenario hold and that the real price of softwood timber will increase by 35 percent by 2020, consistent with timber market forecasts developed in TMBR-1. Agricultural returns are held at their 1992 levels. This scenario was built by imbedding the land-use model described here within the timber market model as described in TMBR-1. This procedure allowed land use, timber management, timber harvesting, and timber prices to be jointly and consistently determined. (See TMBR-1 for a detailed description of modeling assumptions with respect to timber productivity, timber demand, and other factors. See Murray and others (2001) for a description of how these models are linked together.)

A sensitivity analysis was conducted to see how land uses would be affected by different forecasts for timber prices. Results show where rural land use may be most sensitive to timber market changes in the South.

The histories of key driving variables were analyzed. Population changes in counties were plotted. Changes in timber and agricultural prices over time were also analyzed.