The I-93 project includes widening the existing highway from four lanes to eight lanes for a 29-km (18-mile) -long segment near the New Hampshire/Massachusetts border. Several organizations, including U. S. Environmental Protection Agency (EPA), Region 1, Smart Growth group and other New Hampshire-based natural resource agencies and interest groups expressed concern that the project would increase sprawl and result in the loss of natural resources in its path. PB was hired under subcontract to convene and manage an expert panel to develop growth allocations and to determine secondary land use impacts—those that would result from the project but later in time and removed in distance.

Overview of PB’s Procedure

The procedure we developed to calculate the amount of residential, commercial and industrial land needed to accommodate allocated future growth included the following elements:

• Conversion of the project’s blended average allocations for population and employment growth by 2020 into land area needed (in acres). The growth allocations were prepared by the expert panel using the Delphi Technique¹
• Identification of the amount of available developable land in each community based on a review of community master plans and available build-out analyses. Twelve of the 29 communities in the study area had already identified and calculated the remaining developable land in acres via buildout analyses, and some had indicated this amount in their master plans.
• Identification of the general location where secondary growth was wanted according to future land use plans (if included as part of the adopted master plan) and zoning district maps.
• Identification of the general locations of environmental resources in the study area using available geographic information system (GIS) data to prepare a series of individual and composite maps.
• Qualitative observation of whether the secondary land use development would affect these resources as shown on the maps.

The methodology we developed to convert the panel’s allocations to acres of consequent residential development was for unique application to this project; however, it could be applied to other projects. We consulted our in-house land use and environmental law experts throughout the U.S., the Urban Land Institute (Washington, DC), the American Planning Association (Chicago) and Bonz and Company, a Boston-based real estate economic consultant to identify a national or regional index that could be used. Because no reliable index was available, these parties agreed that applying PB’s methodology based on zoning was reasonable rather than applying a per capita index. They cautioned that the use of an index such as acres per capita absorbed over time should not be used because inputs to such type of indices vary depending on the source and the purpose, and could lead to these indices being too gross an indicator in a specific area.

Residential Land Conversion Assumptions

Average Lot Size. We obtained zoning data for the towns in New Hampshire through the regional planning commissions, and determined the average single-family and multi-family lot sizes. Zoning was used because it is the primary legal land use and density control officially delegated to local governments. Existing density of development allowed by present zoning was assumed to be the same in 2020.

The six municipalities in the Massachusetts portion of the study area had completed build-out analyses, so the resulting data were used to compute average lot size in these communities rather than zoning data. It was assumed that future development will be consistent with the percentage of each type of housing calculated in the build-out scenarios. It should be pointed out that the panel was not given an upper limit for their allocations and, in some cases, may have allocated more people than build out scenarios indicated.

Number of Housing Units Required. The number of existing housing units in each town in the study area in 1998 was determined using housing permit data. It was assumed that the proportional mix of single-family, multi-family, and manufactured housing (mobile homes or trailers) in each town would be the same in 2020. In the case of one town, a review of its 2001 master plan update indicated no future growth in multi-family housing so our calculations took this into consideration.

In determining the number of housing unites required, we assumed that the average household size in each municipality would be approximately 0.05 person less in 2020 than that recorded in the 2000 census. This was done to reflect the trend of decreasing household size in the study area since at least the 1980 census, though the rate of decrease has been slowing over time. The panel’s blended average allocation for population in 2020 was divided by the reduced household size to estimate the number of housing units needed to accommodate the allocated growth for the No Build and Build Alternatives.

Acreage Required. For each municipality, the average lot size for single family, multifamily and manufactured housing was multiplied by the number of units needed and then weighted by the proportion of that type of housing in the municipalities’ housing stock. The three numbers were then added together to determine the estimated number of acres required to accommodate secondary residential growth in the study area by the year 2020.

Impacts and Conclusions. The study area’s developable land areas were compared with areas designated by the towns for new growth according to their future land use plans and with locations of high value resources. We concluded that most of the communities had enough upland areas to accommodate the allocated growth without affecting the significant natural resources. We recommended that more technical and financial assistance be provided from the state to localities to improve local and regional land use planning. EPA and the natural resource agencies recommended that NHDOT acquire major wetland areas as “mitigation” of secondary impacts.

EPA’s Region 1 Smart Growth manager said that our approach to addressing secondary land use impacts was “the best they had seen nationwide”.