Eric Bohard

Determining Developable Vacant Lands Through A Growth Management Process

As part of Washington State's Growth Management legislation, municipalities needed to inventory and forecast the amount of lands available for future urbanized growth, specifically residential, commercial and industrial buildable vacant lands. In a very high growth area as exists in Clark County, Washington, fears of not enough future land to develop erupted from the building and real estate community, whereas fears that massive urban sprawl will occur greatly concerned the environmental advocates. A method needed to be devised to spatially show where developable lands currently exist and, based on future land use and zoning scenarios, where and how much land would exist in the future. Through ArcInfo GIS, Clark County, Washington developed a comprehensive model which answered all these questions. Based on numerous land use and critical land criteria, this model mapped and provided land inventories. This model also greatly assisted in defining the urban growth boundaries of seven cities based on projected buildable land needs and population projections for the next 20 years. This paper will describe the process used to determine buildable vacant lands and how the model was developed and tested.

INTRODUCTION

In 1990, the State of Washington legislated a Growth Management Act which was intended to place controls on urban sprawl and conserve rural and resource lands. An element of the Act was to determine boundaries in which urbanized growth, particularly urban services, would be contained. This urban growth boundary was delineated to contain all potential urbanized growth for the next twenty years.

A critical part of the determination of urban growth boundaries is the inventory and identification of vacant lands suitable for development. These lands were divided into three development type categories:

residential buildable vacant lands
commercial buildable vacant lands, and
industrial buildable vacant lands.

Each division had its own criteria for inclusion as buildable vacant, however, there were similarities that all shared.

This analysis requires a number of assumptions about the choices property owners would make. For example, "under-utilized land" is land that is not developed to the maximum that zoning would allow. The model assumes that only 70 percent of the under-utilized land will redevelop to higher densities or intensity uses. Not all land owners will be willing to convert their land to the maximum allowed under zoning during the twenty year period of this model. Another assumption is that the data provides a reasonable picture of the vacant land but does not show areas that are available for urban land uses. In other words, whether property owners are willing to release the vacant property for development or if there are other problems that would prevent development.

BUILDABLE VACANT LAND CRITERIA

General Criteria

Several issues regarding the definition of buildable vacant land were common for all three categories. Much of the data used to make these determinations came from the Clark County, Washington Assessors land record database, assessors parcel maps, FEMA floodplain maps, USDA- SCS soil survey and data from Clark County's Water Quality Department in assessing wetlands. The exclusion factors were:

All publicly owned and tax exempt land (federal, state, and local holdings, parks, open space, public and private school properties, wildlife refuges, etc.) are eliminated from the land inventory.
Major right-of-ways, water features, and major utility easements were excluded.
Parcels less than 5,000 square feet with a land value of zero were excluded because they are considered too small to develop readily or are slivers of land with irregular shapes.
All mobile home parks, personal property, mining resource lands, and private commons areas were excluded.
Parcels that are covered by 50 percent or more by critical areas.

These critical areas were divided into two categories, critical 1 and critical 2 lands. These will be defined later in this paper.

Pending plats were identified and removed from the vacant land count as to not over estimate the vacant land available for development.

Critical Lands

Critical lands were determined on the ability to develop them with established and conventional building methods and on environmental regulations and laws. Critical lands were divided into two categories, critical 1 which were exclusion lands in that no development could occur on these lands. Elements of critical 1 lands were as follows:

100 year flood way
High quality wetlands (Class I and II) with a 75 foot buffer
Slopes over 40%

If a parcel contained over 50 percent critical 1 lands, the parcel is undevelopable.

Critical 2 lands consisted of elements less restrictive to development but still can prevent certain uses on the land. The elements for critical 2 were:

100 year flood fringe
Slopes 25% to 40%
Hydric soils with a 50 foot buffer
USGS hydrography with a 50 foot buffer
Priority Habitat and Species designations from the Washington State Department of Wildlife
Unstable slopes and landslide risk areas

Depending on the proposed land use, the percentage of critical 2 cover to limit development would vary. For example, residential uses could occur if the parcel had a large percentage of critical 2 lands. However, industrial uses could be severely limited. For the model, critical 2 lands were noted as existing for residential analysis, and used as exclusions for commercial and industrial.

Residential Criteria

In addition to the general criteria used to exclude lands from being buildable, the residential analysis had its own specific criteria. Those were:

Parcels with structures less than $10,000 in assessed value were classified as vacant and available for development.
Critical 2 lands did not exclude lands from development.
Under-utilized parcels were defined as parcels that are three times the minimum lot size (based on allowable density) and at least 2.5 acres in size. This standard recognizes that many small parcels are actually accessory to adjacent parcels and will never be utilized for infill or redevelopment.
Parcels 2.5 acres or more in size having an assessed value of $250,000 or more were excluded.
All parcels with structures having an assessed value of $750,000 or greater were excluded.

Commercial Criteria

Commercial buildable vacant lands had additional criteria also. Those were:

Parcels with structures less than $50,000 in assessed value were classified as vacant.

If critical 2 lands exceeded 50 percent, a special category was developed identifying those instances. Depending on what type of commercial use would dictate whether critical 2 lands would be significant or not.

Under-utilized parcels were defined as parcels where the land value exceeded the building value. The Assessor values property that is zoned commercial for its highest and best value. So, if a parcel has a commercial structure, but the real value is its location, all the tax value will be placed with the land and often times, the building will have no value. However, the land is not currently vacant.

Industrial Criteria

Industrial lands where divided into three categories according to economic viability and the speed at which the parcel could be developed. These categories are primary, secondary and tertiary and described below:

Primary -has immediate access to arterial streets, availability to sewer within 500 feet, location within an urban service area, little or no critical 1 or 2 lands, parcels larger than 10 acres, and slopes less than 8 percent.

Secondary -has sewer available but no more than 500 feet from the parcel, slopes between 8 and 15 percent, parcels smaller than 10 acres, and has critical 2 lands on site.

Tertiary -has critical 1 lands on site, slopes greater than 15 percent, located outside the urban growth area, and parcel size is less than 5 acres.

The use of these categories will be explained in the methodology section. In addition to dividing up industrial lands into economic viability, the following criteria was used for determining buildable vacant industrial lands:

Parcels with structures less than $50,000 in assessed value were considered vacant
If critical 2 lands were over 50 percent, the parcel was excluded.

Land Utilization Assumptions

Once a count of vacant acres and parcels was made, assumptions about land utilization were developed to allow the consideration of factors that limit the availability of the land such as utilization preferences, already platted property, density, and infrastructure. These considerations are as follows:

Ninety percent of all vacant land will convert over the 20 year time frame to its designated use.
Seventy percent of all underutilized land will be converted to its designated use with the 20 year planning period.
Vacant parcels between 5,000 and 20,000 square feet in size will be counted as one potential lot each, and assumed to develop as planned.

Deductions for infrastructure were established based on the size of the parcel. Development capacity was reduced for infrastructure including roads, drainage, parks, and other miscellaneous facilities such as pumping stations, power transformers, etc. The infrastructure deduction assumptions are:

Vacant Land

No deduction for parcels between 5,000 and 20,000 s.f.
25 percent deduction for parcels from 20,000 s.f. to 1 acre
30 percent deduction for parcels one acre to 2.5 acres
40 percent deduction for parcels 2.5 acres and larger

Underutilized Land

No deduction for parcels one to 2.5 acres
40 percent deduction for parcels 2.5 acres or larger

Density

6 dwelling units/acre for single family
16 dwelling units/acre for multi family
4.5 dwelling units/acre for pending plats
4 dwelling units/acre for some critical areas, where development could occur with mitigation

An error factor of five percent was deducted because of data acquisition limitations. For example, the assessor updates new commercial development once a year at one particular time. Thus, data is always behind actual counts.

Development Standards

In order to determine how much employment or population could be accommodated on the vacant land, the following development standards were assumed:

Housing composition was a 60/40 split for new single family/multi family.
2.33 persons per single family household or 1.8 persons per multi family household based on internal demographic analysis specific to this county.
Commercial/Industrial projections were 12 jobs per acres for commercial and 9 jobs per acres for industrial, both based on regional trends.
123,000 new people and 58,000 new jobs in the next 20 years based on demographic forecasts.

In recognition that land owners act in their own interest, not for the long-term benefit of the community, an extra supply of vacant land was included to prevent artificial inflation of land prices which might make it impossible to achieve affordable housing and employment goals. The market factors were 25 percent for residential and commercial zones and 50 percent for industrial zones.

METHODOLOGY

Prior to any vacant lands analysis, a comprehensive plan needed to be in place in which to test. Four alternative comprehensive plans were developed with varying urban growth areas and percentage of available land for development. The four alternatives are described briefly as:

No Action: No change from the comprehensive plan developed in 1979
Alternative A: Defined new resource land designations
Alternative B: Urban Growth Boundaries reflective of cities preferences
Alternative C: Most conservative urban growth option.

The model was run against each of these alternatives for evaluation as to available land to support future population needs.

Steps to the model are as follows:

Step 1: Starting with all parcels within the county, eliminate all public and tax exempt lands. These would be government holdings, parks, greenways, municipal facilities, schools, and most institutional facilities. Major electricity and gas line utility easements were removed. Hydrography and road rights of way were eliminated also. All parcels that contained more than 50 percent of critical 1 lands were eliminated (see discussion earlier in this paper on critical lands). Parcels less than 5,000 square feet (unbuildable), parcels with no assessed value are excluded.

Sources for this information were found within the county assessor's data base for ownership, parcel size and tax status. Other governmental agencies provided their ownership data.

From step 1, all buildable vacant land was found. No determination was made to categorizing the lands into residential, commercial, or industrial. This gross buildable vacant land product provided acreage of land excluded by critical lands within each city and urban growth boundary, and acres of land by parcel sizes. These products were used to check the components of critical lands to insure sufficient lands still existed once critical lands were removed. The parcel size analysis allowed for initial forecasting of residential, commercial, and industrial vacant lands since each requires minimum acreage for development.

Step 2: This step is the residential lands analysis. In addition to the above, parcels with structures less than $10,000 were eliminated (this amount was used since a building with this assessed value is thought to be uninhabitable). To account for the large land holdings and the existing mansion class residential lands, parcels of 5 acres or less with assessed values of $250,000 and all parcels assessed at $750,000 or more were eliminated (these parcels most likely would not be subdivided further). Next, the gross vacant buildable land coverage created in step 1 was overlaid with each of the comprehensive plan options. This categorized the land into residential based on future land use scenarios. From that, all residential parcels that were three times the minimum allowable density where the resultant polygon is 2.5 acres or more have the potential to be further subdivided. These are the underutilized lands.

Sources for this step came directly out of the alternative comprehensive plans and the assessor's data base.

The products derived from this step provided the number of acres of residential buildable vacant land and underutilized lands within each city and urban growth area. To satisfy the various interests in the building and development industry, acres of buildable vacant land by acreage range was also provided.

Step 3: After step 2, the lands that were left were either designated as commercial or industrial for future land uses. For step 3, commercial land was evaluated. Commercial lands include rural, neighborhood, convenience, community, limited, regional, highway, freeway, downtown, central business district, tourist, general, downtown business district. All factors in step 1 were used to eliminate parcels with commercial uses. However, two distinctions were made. The building value for commercial is $50,000 instead of $10,000 and critical 2 lands of 50 percent or more were classified and noted. For future zoning, critical 2 lands would be available for development for limited types of commercial use. Underutilized lands, as described earlier, were used for commercial analysis.

Sources for commercial analysis were from the comprehensive plan alternatives and the assessors data base.

The products of step 3 allowed planners to adjust the urban growth boundaries so that sufficient land was available for future commercial development to satisfy the future population and employment needs for the next 20 years.

Step 4: All that is left is industrial future land uses. Those consisted of office park, light and heavy industry, industrial park, general industrial, rural industry. Again, all factors to identify buildable vacant lands in step 1 were used, except for two major differences. First, structures assessed less than $50,000 were considered vacant. Second, critical 2 lands which exceeded 50 percent were eliminated as vacant. Also, unless parcels were contiguous, all parcels less than 5 acres were dropped from vacant and are not developable as industrial.

Sources for this analysis came from the comprehensive plan alternatives and the assessor's data base.

The products contained data based on the earlier mentioned levels of industrial land; primary, secondary, and tertiary industrial lands. Based on future employment needs, certain acreage of primary and secondary industrial land are needed to maintain balanced growth. This analysis provided acreage to again adjust urban growth boundaries to accommodate future needs.

CONCLUSION

The determination of buildable vacant lands is a quantitative analysis as well as a political exercise. Granted, the physical characteristics of the landscape are easily identifiable, however, the political realism of meeting population and employment needs are often much more fuzzy. Organizations such as the homebuilders, chamber of commerce, and the economic development committee were brought into the process early to assist with the political component of this project.

This model offers only a snapshot of what the inventory of vacant buildable land is at a given moment. In a dynamic world of development, this is a handicap of the model. To compensate, the model was written in AML to easily accept updated data and run automated with minimal operator intervention.

As the model is refined, additional data inputs are desired. Market factors and concurrency issues will be the greatest enhancements.

ACKNOWLEDGMENTS

I wish to express my appreciation to Jim Seely, now affiliated with David Evans and Associates for his extensive input into the construction of this model. Jim acted as the planner involved with vacant buildable lands.

Eric Bohard
GIS Coordinator
Clark County, Washington
P.O. Box 5000
Vancouver, WA. 98666-5000
Telephone: (360) 699-2391