Growth change in the E-470 corridor

Material Information

Growth change in the E-470 corridor
McGuire, David Joseph
Publication Date:
Physical Description:
viii, 70, [20] leaves : charts, maps ; 28 cm


Subjects / Keywords:
Cities and towns -- Growth -- Colorado -- Denver Metropolitan Area ( lcsh )
Regional planning -- Colorado -- Denver Metropolitan Area ( lcsh )
Cities and towns -- Growth ( fast )
Regional planning ( fast )
Colorado -- Denver Metropolitan Area ( fast )
theses ( marcgt )
non-fiction ( marcgt )


General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Urban and Regional Planning, College of Architecture and Planning.
Statement of Responsibility:
by David Joseph McGuire.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
17464641 ( OCLC )
LD1190.A78 1987 .M3137 ( lcc )

Full Text
David Joseph McGuire
B.A., University of Wyoming, 1974 M.S., University of Wyoming, 1980
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Urban and Regional Planning School of Architecture and Planning

This thesis for Master of Urban and Regional Planning degree
by David Joseph McGuire has been approved for the School of Architecture and Planning

I would like to express my gratitude to Dr. Thomas Clark who acted as Chairman of my Committee and provided a great deal of expertise during this study. I would also like to thank Dr. Yuk Lee who served as Second Reader and helped organize my research. Additional thanks goes to ray third Committee member, Lawrence Mugler, for expert input.
Technical support and assistance was provided by Joe Tempel (Colorado Department of Highways), Ted Coffelt (Boulder County, Kathy Cushman (environmental consultant) and Kathryn Joyner (cultural resource consultant). Permission to use E-470 mapped information was provided by Greg Henk of the E-470 Authority. Joseph Joyner (University of Northern Colorado and Rang King) assisted in the production of slides for project presentation. Wayne Clark (Boulder County) provided recommendations for graphic presentation.
I would like to thank Dr. Daniel Schler for encouraging me to attend the Planning program at the University of Colorado.
I would also like to thank my classmates who encouraged me during my studies. Special thanks go to Martin Landers who empathized with my non-existent summer as he was also writing his thesis.
Finally, I would like to thank Kathy Joyner who typed the manuscript and shared me with the Planning program for what seems
like forever

McGuire, David Joseph (MURP)
Growth Change in the E-470 Corridor
Thesis directed by Associate Professor Thomas Clark
An outer beltway or circumferential highway is currently being planned for the Denver Metro Region. The system is divided into three parts including C-470 (in the southwest quadrant of the metro area), W-470 (in the northwest quadrant), and E-470 (eastern half of the region).
The proposed E-470 corridor will "open" a large area for development which is currently under cultivation or vacant. This study models population growth and environmental constraints to 1996 and 1998.
Research results indicate a spatially unbalanced growth pattern in the study area and a degree of environmental sensitivity across all cells in the study area. Determinants of growth location in the region appear to be primarily dependent on three of eight factors selected to represent study area growth. They are employment, interchange scores, and Public Service meter sets.
The form and content of this abstract are approved. I recommend its publication.
Signed ________________________________________
Dr. Thomas Clark

Scope of the Study...........................
Study Area...................................
Natural Environment..........................
Topography/Slope Failure...................
Soil and Rock Hazards......................
Aggregate Resources........................
Wildlife Habitat...........................
Human Environment............................
Existing Development (Urban/Residential)...
Health Hazards.............................
LITERATURE REVIEW..............................
History of Land Use and Allocation Models....
Components of Urban Planning Models..........
NBER Urban Simulation Model..................
Clark-Hansen Model...........................
Impact Assessment............................
Model Calibration

Results of Calibration........................ 39
Clark-Hansen Model Components................. 43
District Characteristic Values.............. 45
IV. RESULTS........................................... 49
Introduction.................................. 49
Projected Study Area configuration: 1996...... 49
Environmental Factors: 1996................... 55
Projected Study Area Configuration: 1998...... 57
Summary....................................... 62
Recommendations............................... 64
BIBLIOGRAPHY............................................ 67
A. Variable Measurements
B. Results of Simple Linear Regression
C. Selected Variable Measurements and Multiple Regression Results
D. Correlation Results (Pearson Product Moment Correlation)
E. Step-Wise Regression Results
F. Clark-Hansen Model Components (First Run: 1986-1996)
G. Clark-Hansen Model Components (Second Run: 1986-1998)

2.1 Summary of Beltway Influences......................... 20
4.1 Environmental Factors By Census Tract................. 56

1.1 Proposed Study Area, E-470 Corridor................ 2
1.2 Proposed E-470 Corridor (study area)............... 4
1.3 Floodplain Hazard.................................. 7
1.4 Slope Hazard....................................... 9
1.5 Rock Hazard....................................... 11
1.6 Aggregate Resources............................... 12
1.7 Wildlife Habitat.................................. 13
1.8 Existing Development.............................. 15
1.9 Transportation.................................... 17
1.10 Health Hazards.................................... 18
2.1 Urban Subsystems as Components of EMPIRIC
(after Wilson 1975)............................... 29
3.1 C-470 Calibration Zone............................ 37
4.1 Relative Spatial Distribution of Growth by
Census Tract for Proposed E-470 Corridor: 1996... 50
4.2 Study Area Environmental Sensitivity: 1996........ 58
4.3 Relative Spatial Distribution of Growth by
Census Tract for Proposed E-470 Corridor: 1998... 59
4.4 Study Area Growth Pattern with Environmental
Sensitivity: 1998................................. 60

INTRODUCTION Background/Purpose
An outer beltway or circumferential highway is currently being planned for the Denver metro region (Figure 1.1). The system is divided into three parts including C-470 (in the southwest quadrant of the metro area), W-470 (in the northwest quadrant) and E-470 (eastern half of the region). Portions of C-470 have been completed and currently serve traffic while W-470 is in planning with routing alternatives being considered. E-470 is also being planned and evaluated with a target completion date of 1996. In addition, E-470 will serve as a major access to the proposed regional airport scheduled for completion in the mid-1990's (Denver Regional Council of Governments [DRCOG] 1987)-
The E-470 corridor will "open" a large area to development east of Denver which is currently mostly undeveloped or agricultural. There is considerable speculation about the effect that the facility will have on land use, growth and development in a regional context.
This study models potential impacts in a study area containing the E-470 corridor by estimating locations of future growth and potential environmental effects associated with such growth. To accomplish this task the study area was defined, its environmental setting described, and future growth forecasted. After growth was forecasted it was distributed

Proposed Denver Internatlonel Airport Site
A Point to Serve Access to The
Jewell/lliff Avenue Extension
A Point Between Smoky Hill Road and Quincy Avenue moky Hill Road
A Point Between Parker Road and Smoky Hill Road Parker Road Jorda\i Road
A Point Between Peoria Street and Jordan Road
FIGURE 1.1 Proposed Study Area, E-470 Corridor

growth within sub areas of the study area. More specific siting of growth change can then be accomplished by applying known historical patterns of highway/beltway impacts to the study area.
With this information in hand alternatives may be explored to minimize the potential adverse effects of facility construction.
Scope of the Study
Because of the complexities involved with regional urban and transportation systems modeling and impact analysis it is necessary to set the scope of the project to produce meaningful results given time and cost constraints. Limitations (aside from time and cost) which affect this study include the following:
data quality and availability
E-470 corridor changes (current location assumed to be accurate)
interchange/corridor shifts
Further, delimitations were established to constrain the research tc a manageable project and are listed in the following:
concentrate on impacts at and near interchanges
select variables from existing data sources when available
limit study area to a zone of beltway influence supported in the literature
The impact study area, referred to throughout this document as the E-470 corridor study area (CSA), is a two mile wide strip of land running the length of the roadway. The data collection study area is composed of all of the census tracts intersected by or immediately adjacent to the E-470 CSA (Figure 1.2).


fire, police, health and educational facilities will be provided by development and won't constrain growth
water and sewerage not a constraint; provided by developer and assisted by jurisdiction to encourage expanded tax base.
Highway will be built under current private financing
All land not currently developed will be available for development
each jurisdiction will encourage growth in the corridor.
Study Area
The E-470 CSA is located in the eastern half of the Denver Metro area. The general boundaries which enclose the region containing the E-470 CSA include 1-25 on the west, 160th on the north, County Line Road on the south, and Powhaton Mile Road on the east. Primary jurisdictions which control lands in the E-470 CSA include Douglas County (1986), Arapahoe County (1985), Aurora and Adams County (1984). The corridor also intersects planned growth areas of Commerce City and Thornton (1985) (see Figure 1.2).
Existing land use was investigated through the use of comprehensive plans prepared by primary jurisdictions. In addition, 1"=200' scale aerial maps of the corridor were studied. Portions of the E-470 right-of-way were field checked in areas where comprehensive plan information was dated and aerial maps were ambiguous.

Natural Environment
The E-470 CSA intersects a variety of ecologically diverse areas. The southern and southeastern portion exhibit broken topography and stands of ponderosa pine with active springs. The central portion is more shortgrass prairie (although currently cultivated) incised by intermittent drainages. The northeast extreme crosses the South Plate River and ascends the steep northwest terrace where the topography turns to rolling grasslands (currently cultivated). Certain aspects of the environmental setting will constrain development in the E-470 CSA. Those aspects include the following:
Topography/slope failure
Soil and rock
mining/commercial aggregate areas
wildlife habitat
The following is a discussion of the above aspects as they relate to the E-470 CSA.
The 100 year floodplain is used as a standard area by most jurisdictions to delineate a zone which is unsuitable for development (HUD 1978). All of the 100 year floodplain areas intersected by the E-470 corridor are located along active and intermittent drainages (Figure 1.3). The South Platte River floodplain represents the most extensive area; particularly at the confluences of Second and Third Creeks. The next two drainages to the south (First Creek and Coal Creek) exhibit


substantial floodplains up to half of a mile wide in some areas. Further to the south Cherry Creek and its tributaries produce broad floodplains. Most jurisdictions regulate land use within floodplains to protect life and property and to preserve water quality (HUD 1978).
Topography/Slope Failure.
The topography intersected by the E-470 CSA is highly variable ranging from near flat on the east side of the South Platte River near Barr Lake/Brighton to greater than 10% slope on the west side of the River in the same location and south near Cherry Creek.
Slope failure is in part due to topography, soil characteristics, and ground water. The usual manifestation of slope failure is landslide or rock slide. These areas are unsuitable for development because of instability and historical recurrences of such phenomena (HUD 1978). The E-470 corridor crosses such a zone on the east bank of Cherry Creek in the southern portion of the study area (Figure 1.4).
Soil and Rock Hazards.
Areas which contain soils with a high clay content are not recommended for development, or, at the least, should require site by site inspection to determine the hazard of soil swell. Some soils can swell one and a half times their normal volume depending on moisture content (HUD 1978). Likewise, drying can cause the same soil-type to shrink considerably in volume. Such fluctuations are hazardous to building and roadway foundations.


The E-470 CSA intersects a vast area of such soils on the north end of the study area and another significant area on the south and southeast portions (Figure 1.5). Development in those areas must be carefully planned to either alter the soil properties or design buildings to mitigate potential hazards.
Aggregate Resources.
Commercial sand, gravel and aggregate resources which are considered statutory commercial deposits are regulated by the State of Colorado and many local jurisdictions (H.B. 1529-1973). Each jurisdiction controlling such deposits is required to formulate a master plan for their extraction prior to developing or allowing other land use which might restrict their future extraction. Figure 1.6 displays these locations in relation to the E-470 CSA.
Wildlife Habitat.
Wildlife habitat represents a constraint for development particularly in areas that support threatened and endangered species. Wildlife habitat may be described as a natural location which supplies food, water and cover for seasonal sustenance of wildlife populations. Significant habitats are those which support breeding populations or large numbers of species (HUD 1978). Figure 1.7 shows wildlife habitats in or near the E-470 CSA.
Human Environment
Human development which will affect growth in portions of the E-470 CSA include the following:



Figure 1.7

Existing urban development (residential and employment)
Location of commercial development
Existing transportation facilities
Landfills (Lowry, Arapahoe, Tower Road)
Proposed new international airport
In addition, local growth policies and the availability of water, sewerage, and public utilities all affect the location and intensity of development (see;Scope of Project).
Existing Development (urban/residential).
Urban development in the path of E-470 may constrain/affeet future growth in two ways. First, it will remove acreage from development consideration simply because space is not available. Second, existing development can encourage certain types of future activity. For example, low density residential could select for neighborhood retail development while high density residential might attract a regional commercial center. Figure 1.8 displays current development in and near the E-470 corridor.
In areas of vacant/agricultural land it is assumed that development can take place, therefore growth will be allocated to those areas. It is assumed that "normal" beltway-patterned development will occur in those areas (Introduction, Chapter II). In addition, existing employment centers provided the increased access from E-470 will influence the pattern of growth in certain areas; particularly in the southern and southeastern portions of
the study area.


The existing transportation system will have a considerable affect on growth patterns within the E-470 CSA. Of particular importance will be the intersections of major and minor radial arterials with the beltway. It is anticipated, based on the literature, that a significant portion of beltway growth will occur at these sites (Figure 1.9)(see Introduction, Chapter II).
Health Hazards
The final considerations in the human built environment are hazardous zones which could preclude growth. Two significant areas are present near the E-470 CSA including the Lowry Landfill and the Rocky Mountain Arsenal (Figure 1.10). The Rocky Mountain Arsenal has been allocated E.P.A. Superfunds and clean-up is expected to take 40-50 years due to high levels of contamination. The clean-up of the Lowry Landfill is also being planned, however, it is doubtful that the area will be suitable for human habitation during the initial stages of development in the E-470



During the 1950's/1960's residential flight to the suburbs occurred which in turn drew certain employment opportunities to the same locations. Manufacturing and distribution industries were particularly free to move since their dependence on railroads located in the centers of many of the large cities shifted to highways. In addition, shifts from traditional industries to services and high tech industry also facilitated employment mobility and a move to the suburbs. Suburban areas "pulled" employment out of large cities by offering virtually unlimited development potential in the form of inexpensive land and a resident work force (DRCOG 1987).
The development of suburbs as satellite cities changed travel patterns from a radial system (central business district to suburb) to a circumferential system (suburb to suburb)(Wachs 1980). In short, the construction of beltways created certain impacts on some metropolitan fabrics including location and density of residential areas, the composition of the central business district, the location of coramercial/industrial facilities, retail function/location, and the location and density of office space (Table 2.1). All of these impacts have common thread: they all involve the distribution/redistribution of inter-regional population (Economic and Demographic Forecasting Team 1972). It should follow that in order to

TABLE 2.1 Summary
Single Family
Central City
of Beltway Influences
suburbanization taking place regardless of beltways;
non-beltway cities tend toward a higher suburb to central city housing ratio than beltway cities;
beltways have some impact on multi-family rental units due to accessibility and visibility; may draw some units from central city;
suburbanization of population and employment is evolutionary process rather than than as a result of beltways;
manufacturing, wholesale and some service employment losses to beltway corridors may be attributed to land cost and accessibility at beltways;
manufacturing, warehousing, some service employment growth in suburbs may be attributed to beltways (lower land cost, and better access; industry no longer rail-dependent );
some beltway cities have 30-80% of industrial development within beltway corridors
retail is following population regardless of beltways;
beltways provide accessibility and visibility for large retail centers;
some indication that retail is now moving beyond beltways, following population;
suburbanization of office employment taking place regardless of beltways, following population;
beltways provide accessible, visible sites for suburban office parks which might otherwise locate elsewhere in suburbs.
(after DRCOG 1987)

understand changes in land use it is necessary to understand shifts in population.
The literature supports the notion that beltways have a significant effect on the distribution of development within a region rather that as creators of development pulling new opportunities in from outside the region (Dyett et al. 1980, Baerwald 1979). In addition, the pattern of development related to a beltway is believed to depend on the physical components of the facility such as interchange spacing as well as from competition from other arterials in the region. The resulting development associated with the actual beltway generally takes two forms: either clustered at nodes around interchanges or spread out along the entire corridor (Dyett et al. 1930). Conditions for nodal development include interchanges spaced more than one mile apart with no frontage access and equal traffic volumes on the intersecting roadways. Greater accessibility along the beltway generally encourages linear development throughout the corridor. It is also believed that beltway construction in undeveloped areas tends to encourage cluster development at interchanges (DRCOG 1987). Beltways are believed to encourage in-fill between major arterials intersecting the beltway. It is believed, however, that the initial attraction to these areas diminishes over time as land prices escalate in response to increased demand and decreased availability.
When forecasting growth change and examining impacts for an area such as that associated with a circumferential beltway, it is desirable to examine the location of certain land uses by

first assigning growth to vicinities or cells with a region. Allocation models are widely used for this purpose in the field of land use planning (Putman 1979, Garin 1966, Clark n.d.).
The effects of the intensity and location of land use on the location and intensity of transportation facilities has been the object of many research endeavors (Wilson et al. 1966., Harris et al. 197^, Lounsbury 1981, Putman 1979). Brown (1972) elaborates on this "interrelationship" and describes the feedback which must occur between the two processes (land use and transportation).
As one side of the equation is changed, the other responds in an iterative process which tends toward equilibrium. As applied to the E-470 corridor, current and forecasted land uses indicated the need for a beltway facility, and the forecast of the facility in turn alters land use demand on and near the corridor. While many of these traditional studies rely on land use data to forecast demand for transportation facilities, this study conversely focuses on the impact a transportation facility has on current and future growth change.
History of Land Use and Allocation Model Development
A model is a representation of real-world phenomena using mathematical equations (Foot 1981). Land use models might best be described as mechanisms which facilitate the identification, description, and forecasting of land developments for defined zones (Krueckeberg and Silvers 1974). These models are the products of interdisciplinary works including the fields of planning, geography, mathematics, economics, business administration, regional science, and others (Papageorgio 1976).

Models which are used for predictive purposes carry the potential to propagate error over time. This may occur when explanatory errors occur early in the model process and forecasts are made from unreliable assumptions. Some researchers believe that foundation errors can be lessened by incorporating more variables into the model thus diluting negative effect. An example would be to disaggregate a study region into the smallest sub regions which can still yield a usable data set (Harris 1975).
Even though urban systems are difficult to describe in terms of modeled variables, models still provide a quantitative and empirical means for formulating and evaluating plans. Such research, if conscientiously pursued, helps eliminate, or at least minimize, qualitative subjectivity (Echenique 1975).
Model complexity is not necessarily a measure of productivity or advancement of model research. For example, Echenique (1975) describes an urban systems model developed in the late 1960s which was extremely complex with hundreds of equations, feedback loops, and non-linear relationships (Forrester 1969). The model proved that low-income people are attracted by job programs and low-cost housing.
From an historical perspective, the complex transportation and land use models originated from origin and destination surveys used to describe existing transportation facilities and project future travel demand (Voorhees 1959). These surveys were often inaccurate because they were missing critical land use components which affected travel demand (e.g., employment

location and population density factors to name two).
Gradually, iterative models evolved which described the state of equilibrium existing between transportation facilities and land uses over time (Detroit Transportation Study 1961, Chicago Transportation Study I960). These models were successful in replicating and describing existing travel patterns. It naturally followed that in order to project future travel patterns (given an equilibrium/relationship between transportation factors and land use factors) certain land use data must be included.
Two concepts were identified which moved model research forward from the descriptive stage:
land use inputs to transportation models should be provided by planners (intuitive selection)
land use inputs could be modeled and thus predicted because they behaved as natural laws
These two mutually exclusive concepts split United States and
European planners due to differences in who controlled land use
planning. In addition, in the United States transportation was
viewed as a function which served land and shaped its use while
in Europe transportation was viewed as a land use (Holmes 1974).
American models continued to describe the development of land use patterns as natural laws (density, location, accessibility) while critics warned that description of existing land uses was quite different from predicting future land use patterns (Webber 1961). More considerations were entered into predictive models such as accessibility factors and distance/time traveled (Hansen 1959, Stewart 1948) signaling a recognition of the correlative link between land use and transportation.

Model complexity continued to grow in the 1960s as Lowry (1964) developed one of the first urban models which included multiple land uses and accessibility factors. Many of the models which followed were modifications of the complex and comprehensive Lowry model (Steger 1965, Crecine 1968). Many of the models in use today are adaptation of the Lowry model.
While Lowry was praised for his furthering of model research, certain components and results of the model were critiqued. Fleisher (1965) noted that the model was guided by "allocation rules" rather than human behaviors which were certainly a plus. His criticisms were linked to the results of three experimental applications which exposed the following weaknesses:
£d_ hoc constraints must be added depending on each
model not sensitive to retail location when distributing households
model assigns residential population symmetrically
model assigns inner zonal densities (at too high a rate)
retail allocations too sensitive to location of central business district
Fleisher critiques model foundations in general by stating that time varying systems such as transportation and land use cannot be simulated by recording a series of observations at equilibrium. He further asserts that allocation rules must be able to capture the sense of dynamics in urban systems. In this regard, Lowry's model is described as a weak predictor due to the fact that forecasts are made with current estimates which mirror extant conditions. Value is attached to the models success in

exploring policy alternatives and their consequences. The purpose of applying any simulation or predictive model to a problem is to derive the following benefits:
project probable future growth location and its intensity
identify growth impacts
identify potential alternatives
influence development policies and public actions
generate plans
test/evaluate plan effectiveness
identify future plan modifications (Harris 1965)
There are currently hundreds of locational models available which are outgrowths of early, prolific-model building endeavors and the advancement in the field of computers. Despite criticisms, the frequent application of models to urban planning problems remains the most promising research tool available to urban and regional planners. An awareness of previous research can act to improve past failures and help to tailor future research so that, while not all encompassing, it may provide a sound data base and encourage future investigation.
Components of Urban Planning Models
As described by Kilbridge et al. (1970), urban land use/planning models usually address four subjects including land use, transportation, economic activity, and population. The complexity of models focusing on land use is variable although the goal of such models is fixed, and that is the allocation of land uses to cells within a region. Transportation models may

include land use allocation as a component but also require the allocation of all other activities which affect travel in a region (i.e., economic activity, recreation use, residential concentrations, etc.). Economic activity as a model focus is generally expressed in such quantities as labor/employment, wholesale and retail sales, and income. These values are often derived from existing data sources or produced from model/submodel runs. Finally, population allocation is a common subject of urban planning models. Population forecasts often require a variety of data inputs including social factors, biological factors, and policy considerations due to the complex nature of human settlement.
In addition to subject matters, urban planning models usually perform one or a combination of three functions including projection, allocation, and/or derivation (Kilbridge et al. 1970). An urban planning model should describe the future condition of a subject (projection), describe the location of a subject (allocation), and/or describe the effect one subject may have upon another regarding generation or change (derivation).
The following models are discussed in light of this urban planning model framework.
EMPIRIC (Hill 1965).
A widely used growth allocation model, EMPIRIC, perhaps represents one of the best examples of operational comprehensive urban and transportation models which have grown from the Lowry model process (see Hill 1965). The subject of Hill's use of the model is the distribution of exogenous growth (population and

land use) represented in all urban sub systems to sub areas in a
given study region. The model assumes a correlation among
development patterns of urban activities which can in turn allow
for their prediction (ibid). The components of the model
represent urban sub systems (Figure 2.1). The method of the
model is explained in the following (Hill 1965:113)!
The change in the subregional share of a located variable in each subregion is proportional to: one, the change in the subregional share of all other located variables in the subregion; two, the change in the subregional share of a number of locator variables in the subregion; and three, the value of the subregional shares of other locator variables. This concept is expressed by the following equation system:
ARi = Z:aijARij + Zbik(ZkorAZk)
where: i or j = 1,2,. i 1 j N: number of the
located variables (a total of N equations).
= 1,2,... ,k,.. M : number of the
locator variables.
or j = change in the level of
the ith or jth located variable over the calibration or forecast time interval.
= level of locator variable k at the beginning of the calibration or forecast time interval.
= change in the level of the kth locator variable over the calibration or forecast time interval.
aiibik = coefficients expressing the
J interrelationships among variables

Figure 2.1 Urban Subsystems as components of EMPIRIC (After Wilson 1975).

The model is calibrated using multiple regression analysis which produces activity coefficients determined from historical information. The model was touted to produce "an efficient and desirable pattern of residential and industrial development"
(Hill 1965: 1 19).
Critics of Hill's claims suggest that the model must be operated iteratively to produce the desired results and that submodel applications of model outputs would perhaps satisfy this perceived deficiency (Brand 1967; Peat, Marwick, Mitchell and Co. 1972; Boyce et al. 1970).
NBER Urban Simulation Model (Ingram et al. 1972).
The purpose of the NBER urban simulation model is to model change in the urban spatial structure through computer simulation. The model relies on the interaction between the economy and the consumer. Consumers (demand sector) choose the location of residence based on the location and configuration of available housing (supply sector). Cost is the final element included in the model. Submodels are recursively operated to simulate the interaction of the above elements. The subjects of the submodels include employment location, movers, vacancy, housing demand, allocation, filtering, supply, and market-clearing (Lee n.d.). Submodels generate model and submodel inputs. The end result is the simulation of residential growth for a chosen time period. Several periods may be run. Subsequent runs require additional inputs to simulate evolutionary market changes.

The complexity of the NBER model contributes to its shortcomings, not the least of which is the tremendous data requirement. Tied to this is the difficulty in calibrating the model. It does offer the opportunity to glimpse urban spatial structure and the simulated effects of market and policy changes.
Clark-Hansen Model (Clark n d .).
The Clark-Hansen Model provides a process for allocating population growth to sub areas within a region while controlling for interregional consistency. The model can be applied to a situation when it is desired to project growth and its distribution within sub areas of a given region using a reliable projection of growth for that region. Model assumptions are listed below:
Growth (change) within a district is derivative of growth (change) within the larger region in which it resides.
Since regional growth is finite, growth within districts is a competitive process in which not all districts have equal ability to compete.
The development potential of any district within a region is proportionate to its attractiveness.
Attractiveness is a linear function of one or more objectively measurable characteristics of districts.
Over time the parameters of the attractiveness scoring function remain constantat least in the period of future time for which projections are sought.
Districts are internally homogeneous.
Households have common preferences regarding residential environment and location (Clark n.d.).
The author of the model describes the components of population growth and change in regional population in the following:

There are two dimensions of growth (change). One dimension is the growth within the population which currently reside within an area, and which can be expected to continue to reside within the area over the period of future time for which a projection is sought. This dimension is natural increase. The second dimension is the net effect of population...migrating to and from the area over the projection interval. This is the competitive dimension of growth. The more competitive districts within the region will be likely to realize a net competitive increase due to migration from less competitive districts as well as new growth attracted to the region as a whole (Clark n. d.)
The competitive edge one sub area may have over another can change through time due primarily to the following reasons:
Criteria for "competitiveness" is dynamic and changes due to shifts in population preferences and physical sub area modifications.
When competition criteria are held constant, growth impact itself may change the attractiveness of a sub area.
In order to assess the ability of a sub area to attract growth it is necessary to observe prior population changes in the area of interest and assess which criteria most affected that change.
This process produces a quantifiable factor or weight which can be used to describe the importance of given criteria in predicting changing. Assuming that these weights would remain relatively constant over time allows the use of these scores in forecasting future change. The model is as follows:
APit =ARt Ait
t = i t j
, ,
m (time interval) .. ,n (sub area)
= change in population in the ith_ sub area

during the tth future time interval.
^Rt = projected change in the total population
aggregate (exogenous) region as a whole, during the tth future time interval.
= "attractiveness" score of the ith sub area at
the beginning of the tth future time interval.
n = number of sub areas in region/study area.
m = number of future time intervals for which district allocations are sought (after Clark n.d.).
The attractiveness score is used to indicate the relative appeal
of a sub area and the "power it will have in competing for
shares of regional growth. The score may be composed of measures
and observations of growth among analogous sub areas in previous
years. The technical character of the score is presented in the
f ollowing:
Ait = a 1 Cut + a2c2it + ceceit where:
= the weighting "coefficient" for the k characteristic, C^it
Ckit = the kMl characteristic of the iMl sub area at the beginning of the time interval t.
e = the number of characteristics by which we identify each sub area, hence k = 1,...,e (after Clark n.d.).
The model must be calibrated to tailor it to the study area. The
net product of calibration is the set of weighting coefficients
used to determine sub area attraction scores (See Methods,
Chapter III).
Impact Assessment
Most current forms of impact analysis were initiated with the passage of the National Environmental Policy Act of 1969 (PL91-190, 31 STAT. 852; 42 U.S.C. 4321-4347). Impact is defined by this Act as any modification or alteration of the human

environment. This study concentrates on one aspect of impact assessment; population growth and distribution within the E-470 corridor study area.
Modeling growth change has become an accepted method of predicting effects of transportation facility construction and development impacts in general (National Technical Information Service 1975). Impacts caused by the construction of transportation facilities may be categorized as primary and secondary (Council of Environmental Quality 1975). Primary impacts are those effects which result from environmental disturbance due to actual construction. Secondary effects are not as apparent as construction disturbance and are associated with the physical human environment including social, economic, and psychological factors (Canter 1977). Many of these effects may not be observed until well after the facility is built and in operation.
Before the effects of impacts can be assessed it is necessary to measure current environmental conditions in the area of interest (see Introduction--Study Area). The process of impact assessment then takes into account the placement of the transportation facility (E-470) based on the analysis of critical factors collected regarding the present condition of the study area (Jain 1978).
Decisions on where to direct secondary effects of highway construction follow the process applied to primary effects. For example, characteristics of highway interchanges are known to influence development patterns around and near them. As mentioned in Chapter II (Introduction), concentrations of nodal

development will often occur when interchanges are spaced more than a mile and no frontage access is provided. This pattern is further encouraged by highways built in undeveloped locations (Pennsylvania Department of Transportation 1975). In addition, development tends to occur in greater density along the crossroad intersecting a limited access facility such as E-470 (ibid). Knowing these historical effects of highway/interchange construction and accepting the validity of these trends (as this study does), one can make some preliminary observations regarding growth location and density in the E-470 corridor study area.
For example, it is apparent that E-470 will have widely spaced interchanges intersecting mostly major and minor arterials. There is no frontage access along most of the corridor and the corridor itself is located in undeveloped areas. According to these factors E-470 will probably experience nodal concentrations of development near interchanges. There may certainly be exceptions to this general observation due to other development influences.
Accepting these facts as valid it is possible to plot the area of influence around an interchange and examine the impacts such development will cause. Since growth change due to construction is also a known (forecasted) factor relative densities can also be determined. In short, by modeling impact through the use of impact assessment and forecasted estimates of growth change, it is possible to project impacts and direct growth to minimize adverse effects. This approach also allows for the development of alternatives to help direct growth to lessen projected impacts on the human environment.

One of the purposes of this research was to allocate the forecasted population growth to sub areas of a study region.
The Clark-Hansen Model was chosen to accomplish this task because of the excellent fit between model data requirements and the E-470 configuration and data availability.
The E-470 corridor was defined by the E-470 Authority on a variety of maps including U.S.G.S. Quadrangles (1:24,000), blueline project maps (1:200) and on blue line aerials of the study area. Individual interchange maps were provided which displayed configuration, area of disturbance, etc. All of these sources were used to plot the corridor on project base maps.
As mentioned before, the segment of C-470 completed between U.S. 85 and 1-25 served as an analogous transportation corridor to the E-470 study area. After the facility was chosen, a study area was identified along the corridor which included a two mile impact zone on either side of the C-470 centerline. Next, census tracts which were intersected by the impact zone were plotted and used as the primary data gathering grid (Figure 3.1). Portions of census tracts were included if it was known that the majority of tract development activity occurred in or near the highway corridor.
Both multiple variable and simple linear regression were used to select those factors which best explained total aggregate


population change in the study area between 1980 and 1985. Those dates represent the time between the completion of the Environmental Impact Statement (final stages of planning) to the time automobile traffic began to use the facility.
Model Calibration. The purpose of model calibration is to assure it is tailored to the study area in question. The process of calibration yields the weighting coefficients (a^) through multiple regression/linear regression analysis. Regression analysis is an effective technique to better understand change phenomena and the interrelationships between descriptive variables and a dependent variable (Lee 1973). The formula expressing this relationship is as follows:
Y = a + b -j f x 1.. .bnxn
T-] = dependent variable
a = point where regression line intersects Y axis (constant)
t>n = slope of the line
xn = independent or descriptive variable.
The descriptive variables represent characteristics of the sub area which best describe growth change in a region. The dependent variable represents population change in each sub area during a given time period.
To calibrate the model it was necessary to prepare a matrix of variable measurements as seen in Appendix A. As mentioned before, the results obtained from this matrix are the weighting coefficients used to prepare the attractiveness score for the sub areas of the selected study area.
As previously mentioned, in order to calibrate the model for the E-470 study a local analogous transportation facility was

used. Centennial Parkway or C-470 displayed considerable similarities to the E-470 study area including the following attributes:
Both are in the Denver metropolitan region
Both have shared data sources available
Both are physically accessible facilitating field visits
Both occupy similar environmental zones
Both are subject to the same local market trends
C-470 is being built in segments as will be E-470
Most of the same jurisdictions are involved in both projects (Haase 1980).
On the other side of the issue are the dissimilarities enumerated in the following:
The portion of C-470 used for calibration has a parallel frontage facility (County Line Road)
The completed portion of C-470 is a much shorter segment than the entire E-470 roadway
Dissimilar land ownership patterns exist
Project financing differs between projects
Despite the dissimilarities, the C-470 facility represents the best analog available for the E-470 study.
Results of Calibration
After identifying the C-470 facility as the appropriate parallel" roadway to calibrate the Clark-Hansen Model for the E-470 research, the following steps were accomplished:
mark an area of influence around the chosen section of roadway (calibration zone)
grid the area in the smallest increment which would still allow for data measurements (see Figure 3-1)

identify descriptive variables for the calibration zone
run regression analysis on selected variables
A two mile buffer on either side of the C-470 roadway was used to mark an area of influence for the calibration zone. The buffer area was established based upon a variety of research results on the matter of transportation facility influence areas (U.S. Department of Transportation 1976, Engelen 1982, Gamble et al. 1978). The buffer area also allowed for the inclusion of the maximum number of fixed grid units (census tracts) in the calibration zone without compromising the unique impact characteristics of the C-470 roadway (e.g., by excluding areas too near other parallel major arterials).
The grid units selected to represent sub areas within the calibration zone were census tracts identified by the U.S.
Bureau of the Census. These units were selected because they represent the smallest discrete boundaries which are represented by reliable data sources (DRCOG 1985). Additional census tract characteristics include variability in size (Gilpin County, Colorado is a single census tract 138.00), and they are based on population size (4,000-6,000 minimum). As of the 1980 census there were 407 census tracts in the Denver Region (ibid). The variable selection process produced one dependent variable (total sub area population change between 1980-1985) and eight independent variables listed in the following:
Households (X^)
Household size (X2)
Population Density (X^)
Total Employment (X4)

Land Use (X5)
Interchange (Score) (5)
Meter Sets (Xy)
Population in Households (Xg) (See Appendix A)
Regression results showed that all were indicators for
explaining overall population growth in the study area although a certain degree of collinearity is evident. Regression graphs were generated to test goodness of fit in each individual variable (Harvey 1987). Each variable was entered independently with base year data, end year data, and absolute change data by census tract where appropriate (See Appendix A).
Households represents the total number of households in the study area in 1980. This factor includes all occupied dwelling units. Household size reflects the overall household size by census tract in 1980. The next factor measures the total Population Density (per acre) within each census tract in 1980. Employment measures total employment opportunities by census tract in 1976 (1980 baseline score not available data in study format). Land Use scores were derived from mapped data. Each census tract was measured as to the amount of acreage developed in 1980.
The Interchange score assigns a value to a census tract containing all or part of an interchange as a function of the total acres of the tract within the study zone in 1980. Meter Sets measures total electrical hook-ups in the calibration zone in 1980 thus measuring development. Finally, Population in Households measures the total number of residents in all
households in 1980. It should be noted that Meter Sets was not

included in the E-470 portion of the research because of the difficulty in assessing what portion of the E-470 CSA relies on Public Service for electricity as opposed to rural electric associations. Due to data availability, most of these variables are site measurements rather than situation estimators with the exception of the interchange score.
Other variables were identified and examined as to their potential for explaining population change in the C-470 calibration zone. For example, research has shown that distance to/availability of water and sewerage is an important indicator. The C-470 calibration zone is already located in urban service areas which will provide such utility services and it was therefore not considered a critical factor or potential constraint to development (Haase 1980).
Another variable which has been shown to be a good predictor of population change is change in travel time from a given location to a major arterial such as 1-25. Travel time was not included because the existence of County Line Road as frontage access would not present an analogous comparison to the E-470 study area regarding travel time, service level, and general access. The variables that were chosen for this study appear to have a general applicability to both the C-470 calibration zone and the E-470 CSA.
As mentioned before, it was believed that all of the identified variables had a relationship to total population change in the C-470 calibration zone. A simple linear regression procedure was performed to determine to what extent the

independent variables predict change in the dependent variable
Each variable was run independently to determine isolated affect
(e.g., Y-| a + bx-|) (Appendix B). It was determined that all eight variables had a relationship to the dependent variable.
The variables were then run in a single multiple regression equation to obtain a combined interdependent set of coefficients (Appendix C).
A Pearson Product Moment Correlation matrix was generated to verify variable relationships (Appendix D). All of the descriptor variables showed a very strong relationship to the independent variable. In order to assess the primary driving variables in the model a step-wise regression was also run (Appendix E). The variables accounting for most of the model prediction include, in order, meter sets (X7), interchange score (X6), and employment (X4).
With the first phase (model calibration) completed, the second phase of the study was initiated. In order to apply the Clark-Hansen Model to the E-470 study area it was necessary to, as accurately as possible, locate the roadway on 15* series U.S.G.S. quadrangle maps (county series) and define the corridor study area. That scale allowed for reasonably accurate representation of development constraints in the form of topographic features, geographical disconformities, geologic hazards, etc.
Clark-Hansen Model Components
The Clark-Hansen Model was applied to the E-470 CSA in two time sequences. The first was for a period between 1986

(present) and 1996 (one probable completion date for the E-470 beltway). The second period between 1986 and 1998 was chosen to demonstrate post-construction development. In essence, all of the development of ancillary facilities associated with the beltway are represented in the latter model run.
Five data components were established to operate the model including the following:
District characteristic values (C^it)
Regression Coefficients (a^)
Estimate of regional growth (Rj.)
Attractiveness scores by census tract (A^t^
Growth allocation factors by census tract ( P^t)
(see: Clark-Hansen Model)
These five data components were used in both model runs (1986 1996; 1986-1998), each with estimated district characteristic values projected for the target baseline year (i.e., 1996 and
1998)(Appendices F and G).
To tailor the data for the most accurate model results several adjustments to direct linear and direct exponential forecasts were required (Ottensman 1985). District characteristic values had to be modified to include events which will presumably occur in the future in the study area related to roadway construction (e.g., new airport, interchange construction, local policy changes, growth constraints, etc.). Some variable values were more affected than others and are discussed in the following section.

District Characteristic Values.
Population forecasts to 1996 and 1998 were adjusted for census tracts 68.06, 70.29, 70.31, 71.00 and 85.11 to bring them into a more proportional relationship with DRC0G Regional Statistical Area (RSA) projections to 1990, 2000, and 2010 (DRC0G 1986). This was necessary since the baseline population estimates for this study were based on DRCOG figures for 1980 and 1986. In addition, the 1990, 2000 and 2010 DRCOG population projections included a "bonus for beltway RSAs thus incorporating the facility effect which was a requirement of this research (DRCOG 1986).
Total number of households (X-j) remained a product of the projected population and the average household size thus figures for the aforementioned census tracts were affected by population forecast adjustments. No buildout limit was established since no population capacity was used in either model run. Such a constraint was not considered necessary due to the considerable amount of undeveloped, open land in the E-470 CSA. Vacancy rates were not applied to this variable.
Average household size (X2) was assumed to continue a gentle downward trend although some tracts in the study area exhibit projected increases (census tracts 83-53 83.03> 71.00 and
70.30). Average household size was forecasted from 1980 census figures with the aid of DRCOG's 1986 estimates (DRCOG 1936).
Population density (X3} was affected by the adjustment in the population forecasts. Density remained the relationship between total census tract population and square miles in each

Total employment (X^) was adjusted for the 1998 run of the model to reflect airport operation. The two tracts which will absorb the majority of the estimated 35,000 jobs associated with the airport include 85.14 and 83.53* Other census tract figures were adjusted to closely correspond to DRCOG projections for 1990, 2000, and 2010 (DRCOG 1986). No negative values were considered to be possible/realistic in light of the proposed development in the study area. Census tracts containing proposed interchanges were also given employment bonuses since this study assumes that development will concentrate at those nodes. It was felt that employment bonuses tied directly to the roadway were adequately represented in DRCOG figures used as baseline data in this study.
Land use (X^) was the most affected variable value; particularly for the second model run (1998). The baseline scores for the study area (1996) were held relatively constant from extant conditions. This was done for a number of reasons. The first is the difficulty involved in accurately predicting a level of fragmented growth tied to a segmented and phased roadway construction project. Another reason involved the difficulty in accurately predicting which areas will attract development without attraction developments already in place like the airport, beltway interchanges, and the roadway itself.
Land use values were adjusted for the 1 998 model run by calculating bonus factors for impacted census tracts in the form of acres per development. The following values were used:
interchange640 acres were allotted if a facility was wholly contained in a census tract. Appropriate proportions were allocated where an interchange fell on a

boundary line (cf., Pennsylvania Department of Transportation 1975).
Roadwaytotal acreage in the E-470 CSA in a given census tract was calculated to account for potential frontage development.
Airportthe total acreage of the proposed airport was calculated and added to census tracts 85.14 and 83.53.
That area included both the facility area and the 60 LDN buffer area.
Ground unsuitable for development was subtracted from census tracts which contained development constraints like Lowry Landfill and the West Arapahoe Soil Conservation District in census tract 70.29 (Arapahoe County 1985).
Interchange scores (Xg) were assigned by point totals for all or portions of an interchange located in or bordering a census tract. All potential interchanges have not been located definitely in the central portion of the study area to date due to uncertainty regarding the airport site. Interchanges were assumed for the central study area where they were considered most likely to occur. Interchange effect has also been considered as a component of land use (see above).
Population in households (Xy) represents the non-transient portion of the population. This figure closely approximates the resident population of the study area and was thus used to parallel population figures. The values for this variable were forecasted like population with similar checks against baseline value criteria (DRCOG 1986).
As mentioned before, meter sets were not included in either of the study area model runs due to questionable validity. Of concern was the uncertainty of the number of Public Service Company customers versus rural electric association customers.

The meter sets were therefore not reliable estimators of total new growth in the study area.
Finally, growth allocated to census tracts 85.14 and 83*53 was not redistributed to neighboring census tracts due to the proposed location of the new airport. It was determined that both tracts contained sufficient undeveloped area to absorb the allocated growth and the new airport.

The first run of the model established the "baseline characteristics of the study area, i.e., in 1996. For the purpose of this study it was assumed that growth to 1996 would follow historic trends since the construction will move by segments from the central portion (if the airport is built) from 1-70 to the proposed airport, from South 1-25 to Parker Road, from Parker Road to 1-70, then from the airport to North 1-25. This is one scenario which demonstrates the difficulty in forecasting growth for non-contiguous roadway construction.
With "ideal" conditions assumed for 1996 growth, the 1998 allocation was run with the predicted influences in place which could then function to draw growth according to competitive attraction. The following results were achieved in the modeling process.
Projected Study Area Configuration: 1996
Figure 4.1 displays the relative spatial distribution of growth across all study area census tracts for aggregate population change (1986-1996). Two positive growth peaks are pronounced at census tracts 70.29 and 141.00.
Although there are certain growth restrictions attached to tract 70.29 it is believed that the model is accurately representing possible buildout in the area as a result of infill. The tract contains a portion of Aurora urban development, the

Census Tracts
Relative Spatial Distribution for Proposed E-470 Corridor:
of Growth by 1996
Census Tract

western extreme of the Lowry Landfill, and a Soil Conservation Service preserve in between.
Census tract 141.00 likewise is expected to continue growth for the following reasons:
Location of the I-25/E-470 interchange
Availability of developable land (the entire tract was included in this study
Booming construction taking place at Highland Ranch fronting C-470.
Tracts exhibiting moderate positive growth to 1996 include 70.31, 85.13, 85.14 and 140.00. Census tract 70.31, like 70.29 appears to be a likely growth area due to infill. Access within the tract will most certainly be increased off of Quincy Avenue (Airline Road) on the north, Parker Road on the west/southwest, and E-470 on the south/southeast. Smoky Hill Road and the Smoky Hill/E-470 interchange are also located in the southeast portion of the tract. The area currently exhibits stratified development with intense urbanization in the northern extreme and scattered large lot residential to the south. There is considerable room for development.
Tract 85.13 contains all of three major interchanges and a portion of a fourth. Major arterials crossing the area include from west to east 1-25, Washington Street (to be relocated), Colorado Boulevard, and Quebec Street. The area is currently mostly agricultural although scattered large lot developments occur as do numerous oil and gas wells. Much of the area is in unincorporated Adams County and Thornton Planned Urbanization Area (PUA).

Tract 85.14 contains all of two major intersections, a considerable area of undeveloped land, and will contain most of the new airport as proposed (not a factor in this model run).
Major accesses in and bounding the tract from north to south include U.S. 85, 1-76, 120th Avenue, 104th Avenue,, and 1-70.
The area is mostly agricultural with some large lot residential. The eastern part of the tract is in the Brighton PUA while the southern extent falls in the growth area of both Denver and Aurora.
Tract 140.00, located in the southern extent of the study area, is bound on the west by 1-25 and contains all or part of four interchanges including 1-25, Peoria Street, Potoraac/Chambers Road, and Jordan Road. The northern part of the tract contains residential development and borders Centennial Airport and associated industrial development. The remainder of the tract is primarily agricultural with some scattered large lot residential. All of the tract is located in unincorporated Douglas County.
Slow positive growth is noted in two tracts including 85.09 and 139.00. Census tract 85.09 is an area characterized by residential urbanization, large lot residential, and cultivated agricultural ground. The western boundary is Sheridan Boulevard, 1-25 is located on the eastern boundary, Dillon Road is located on the north, and 120th Avenue is the southern boundary. The tract includes or borders the jurisdictions of Adams County, Westminister, Broomfield, and Northglenn.
Stable/positive tracts include 85.10, 85.11, and 85.12, all on the south side of proposed E-470 and adjacent to one another (see Figure 1.2). Tract 85.10 is bordered on the west by 1-25, on

the north by Dillon Road (144th Avenue) and on the east by Colorado Boulevard (except for the extreme southeastern corner). Thornton, Northglenn and Adams County are the primary jurisdictions involved and most of the land is large lot residential or cultivated. The southern-most portion is urbanized with development proceeding from south to north toward the E-470 corridor.
Tract 85.11 primarily is located between tract 85.10 (to the northwest) and the South Platte River (eastern boundary). The land is in agriculture with some large lot residential development. Adams County is the controlling jurisdiction although the ground is in Thornton's growth area.
Census Tract 85.12 is bordered on the west by the South Platte River, on the north and west by tract 85.14, and on the south by 104th Avenue. South of 85.12 is 87.01 (not in the study area) which contains Rocky Mountain Arsenal. Land use is characterized by large lot residential, strip development (U.S.
85 and 1-76), some urban development, and irrigated and dry farm lands. Jurisdictions with primary interests in the tract include Adams County and Commerce City. The proposed E-470 corridor parallel the Second Creek floodplain in the northern portion of the tract. The floodplain zone on Second Creek and the South Platte River is listed as poor for development potential in the Adams County Comprehensive Plan (1984).
Three census tracts closely parallel, on the negative side, the previously discussed stable/positive tracts. They include 71.00, 83.03 and 83.53. Census tract 71.00 is the largest in the study area and the farthest removed from urbanized area. Despite the fact that the tract contains several interchanges it is

apparently far enough removed from the primary infill and development area that growth will occur slowly to 1996. Arapahoe County is the primary jurisdiction. The E-470 right-of-way jogs east around the Lowry Landfill which is located in 71.00 and 70.29. Primary land use is agricultural/vacant.
Tract 83.03 (combined with 83.50) represents a fairly isolated land area as far as access is concerned; however, it is in Denvers urbanization area and will likely contain an interchange. These factors apparently combined to produce the stable/negative projection for growth to 1996. First Creeks floodplain runs diagonally through the tract precluding development from that corridor.
Tract 83*53 will contain a portion of the proposed new airport; however, up to 1996 it will be a large isolated tract incised by several drainages. Most of the land is in dry farming and is considered to be suitable for development. The land is almost entirely under the jurisdiction of Adams County with the southern portion included in the Aurora PUA.
The model indicates that two tracts (68.06 and 67.03) will contribute some of their growth to other beltway census tracts. This may be due to the fact that the model measures and allocates population growth and both 67.03 and 68.00 are currently and will probably continue to experience regional commercial and light industrial development related to the E-470/1-25 interchange. It is possible that this continued development will displace residential growth to other census tracts. Both tracts are in Arapahoe County and are urbanized. Both, especially 68.06, maintain population densities which may

act as growth reservoirs to nearby census tracts. Tract 67.03
is bordered on the south by County Line Road/C-470.
Environmental Factors: 1996
In order to assess the relative position of each census tract in the study area with regard to environmental constraints, values were calculated for the following factors:
flood hazard
slope failure
expansive soil/rock
aggregate resources
wildlife habitats
urban development
man-made hazards
Tracts were assigned a numerical score for each environmental factor which occurred in the E-470 CSA (Table 4.1). Scores equaled the amount of effect or concern certain environmental factors had in each tract. Scores were based on the proportion of acres each environmental factor occupied in and adjacent to the E-470 CSA in each census tract. Possible scores included the following:
0 = not applicable/not present
1 = of slight concern
2 = or moderate concern
3 = of significant concern

Table 4.1
Environmental Factors By Census Tract
67-03 II 1 I 0 I 3 I 0 I 1 I 2 I 0 I 7 I
68.06 II 2 I 0 I 1 I 1 I 0 I 2 I 0 I 6 I
70.29 II 2 I 0 I 3 I 0 I 0 I 2 I 2 I 9 I
70.31 II 2 I 3 I 3 I 1 I 2 I 1 I 1 I 13 I
71.00 II 3 I 0 I 3 I 1 I 3 I 2 I 2 I 14 I
83-03 II 3 I 0 I 0 I 0 I 2 I 1 I 0 I 6 I
83.53 II 2 I 0 I 1 I 0 I 2 I 1 I 1 I 7 I
85.09 II 1 I 0 I 3 I 0 I 0 I 1 I 0 I 5 I
85.10 II 1 I 0 I 3 I 0 I 1 I 1 I 0 I 6 I
85.11 II 3 I 0 I 3 I 3 I 0 I 1 I 0 I 10 I
85.12 II 2 I 0 I 1 I 1 I 1 I 1 I 1 I 7 I
85.13 II 2 I 0 I 3 I 1 I 1 I 1 I 0 I 8 I
85.14 II 2 I 0 I 0 I 1 I 3 I 1 I 1 I 8 I
139.00 II 1 I 1 I 2 I 1 I 1 I 1 I 0 I 7 I
140.00 II 1 I 1 I 2 I 1 I 1 I 1 I 0 I 7 I
141.00 II 1 I 1 I 2 I 1 I 0 I 2 I 1 I 8 I
0 = NA (not present)
1 = of slight concern
2 = of moderate concern
3 = of significant concern
NOTE: Scores based on proportion of acres of environmental factor occurring in E-4-70 CSA in each census tract.

The assignment of values was a somewhat subjective process relying not only on acreage calculation but also on intuition based on aerial photograph analysis and field assessments.
Results were plotted for 1996 assuming no environmental change in the values of those selected factors (Figure 4.2). The plot of those values allowed for comparison with the projected growth allocation for 1995 (see Figure 4.1). All tracts display some level of environmental sensitivity to development in the E-470 CSA, particularly tracts 70.31 and 71.00. Tract 70.31 displays moderate growth potential according to the allocation model while 71.00 shows a stable/negative value. The two peak growth tracts (70.29 and 141.00) display raoderate/high environmental sensitivity to growth.
Projected Study Area Configuration: 1998
When growth variables including interchange growth, the proposed airport, etc., are added to the model equation for 1998 tract growth allocations shift from the 1996 model run (Figure 4.3). Again, 1987 environmental conditions were held constant. Again, all tracts display the same environmental sensitivity to development as seen in the 1996 comparison. Now the most sensitive tracts exhibit more moderate growth allocations because of growth being assigned to more attractive tracts. There are more tracts in the moderate/positive growth range affecting more environmentally sensitive areas (Figure 4.4). Of particular note is the balance' between 1998 growth distribution and environmental sensitivity where the tracts with the most growth

Census Tracts
Figure 4.2 Study Area Environmental Sensitivity: 1996

Census Tracts
Relative Spatial Distribution of Growth by Census Tract for Proposed E-470 Corridor: 1998

Figure 4.4 Study Area Growth Pattern with Environmental Sensitivity: 1998

occur in areas of least total environmental sensitivity. The most environmentally sensitive tract, 71.00, will not compete extremely well for growth according to the 1998 model run.
Although specific site analyses are beyond the scope of this research some areas where attention should be focused are evident. For example, the roadway itself directly intersects floodplains, soil/rock hazard areas, wildlife habitats, slope failure zones, and aggregate deposits. In addition, the E-470 CSA as defined in this study takes in a portion of the Lowry Landfill, an extremely toxic facility. These locales warrant specific attention by controlling jurisdictions to assure that ancillary development is tailored to specific situations to avoid conflict between E-470, the host environment, and ancillary highway facilities.

It is difficult at this point to predict certain growth factors associated with the E-470 CSA. For example, total interchange influences is not completely predictable since all of the interchanges have not been finally located. Another major uncertainty is the ultimate location of the proposed international airport. These question marks affect the accuracy of model results since they contribute so much in determining sub area competition/attractiveness values in 1988.
It is of particular interest which variables account for the majority of the models prediction/allocation powers. The most surprising link may be the relationship between interchange score and population growth change. The potential to control or shape development based on numbers and location of interchange facilities may provide a tool to help shape residential patterns. This possibility deserves future research attention.
It is impossible for a research project of this scale to consider all of the growth components included in macro models such as EMPIRIC. Instead, the research method was designed to match the research goal: to describe growth change by allocating growth to sub areas of a study region and to assess the potential environmental effects of those changes. In achieving this goal the research produced several products including the following:

a model of growth to 1996 including roadway impacts but precluding total ancillary facility influences
a model of growth to 1998 including major growth influences assumed in 1987
a determination of 1987 environmental sensitivity in the E-470 CSA and vicinity which could or should influence planning and development decisions
a comparison between the 1998 modeled growth and possible study area environmental conditions
The model of 1996 growth supports observations found in the literature regarding beltway projects. For example, certain sub areas will contribute portions of their growth to more attractive zones. Cited research refers to this as the redistribution of local resources (population and employment) within a beltway rather than a phenomenon tied to the influx of new population and employment opportunities. The 1996 model also supports conclusions in the literature that infill will probably occur along some of the arterials intersecting the beltway.
The 1998 model goes further in describing the E-470 CSA and vicinity by adding key facility influences to the forecasted growth pattern. For instance, the addition of completed interchanges and related employment opportunities de-emphasizes many of the sharp growth peaks observed in the 1996 model run.
The proposed airport also serves to pull more of the growth toward the northern end of the study area as would be expected. Fewer negative values are seen with the addition of the effect of ancillary facilities and the remaining negatively valued sectors show a marked decrease in their competitive losses.
The trend toward more positive and "even" growth can have

positive and negative connotations. For example, the better showing in growth could indicate additional growth from outside the region, or that growth is coming to beltway sub areas from zones not included in the study but from within the immediate region.
Gross environmental sensitivity scores for the E-470 CSA and vicinity indicate that there is a need to closely monitor development along the corridor; particularly in areas of forecasted substantial growth. This portion of the study also indicates that there may be critical engineering concerns not only for construction of the roadway itself but also for ancillary facilities which will gravitate to the beltway.
The graphic representations of projected growth with the environmental sensitivity scores allow for identification of specific sub areas which may be targeted for expansive future growth yet which may not be environmentally capable of supporting such levels of growth or which will require considerable planning and engineering to accommodate certain inevitable development. Certainly the products of this research can be used to justify careful land use decisions in areas where development is economically desirable but potentially hazardous or environmentally unbalanced.
Questions raised by this project may be answered by continued research. As a baseline study this research produced the ground work to analyze growth change and development considerations in the E-470 CSA. Future research could take

several forms in an effort to assess the study area on a more "microscopic" level. In addition, the study area model variables could be interchanged with other district descriptors possibly more suited to a specific sub area or set of sub areas.
The next logical steps in the E-470 corridor research will depend on specific areas of interest and might include the following:
model a broader region to more accurately determine which metro sub areas will contribute portions of their future growth to the beltway
model more specific data to determine which employment sectors are most likely to relocate to a beltway corridor (e.g. distributors, regional commercial, light industry, etc).
map specific environmental constraints and their actual boundaries to allow for site specific analysis
It would undoubtedly be of interest to a jurisdiction to be able to predict which areas will experience future growth and tax base losses and which areas will gain from those losses. Such a scenario could be effectively modeled using the processes established in this research.
The use of environmental factors to aid in the allocation of growth could be modified to obtain more precise information on specific location and degree of sensitivity. These types of data are most often used to select a roadway corridor from a set of alternatives; however, in the case of E-470 the corridor has already been selected and concern must be shifted to the development which will occur as a result of the beltway.
A method to achieve greater insight into the environmental composition of sub areas would include mapping zones of each

environmental factor and overlaying them on a base map much like the process described by McHarg (1969). This approach would allow for more accuracy in identifying potential environmental hazards or environment/development conflicts in zones predicted to have increased growth pressures. Advance knowledge of these conditions would allow jurisdictions to prepare policy recommendations to mitigate potential impacts and conflicts.
The need to use current environmental conditions for some future-time forecast limits the accuracy of the approach. Certainly, the farther into the future one attempts to predict growth the less reliable the environmental baseline information will be. It would be most desirable if current and complete environmental assessments accompanied all development which have been determined to be growth altering facilities.
It is believed that this research, used as a baseline, can greatly improve our ability to locate project-specific areas of growth and isolate/determine potential locations of environmental conflict. Local jurisdictional follow-up could aid in producing balanced development which does not impact the host environment and which yields the maximum tax base benefits for which it is sought.

Adams County, Comprehensive Plan (Commerce City, CO, 1984).
Arapahoe County, Comprehensive Plan (Littleton, CO, 1985).
Baerwald, T.J., "Land Use Change in Suburban Clusters and Corridors (Transportation Research Board, Transportation Research Record 861, 1979).
Boyce, David E. et al., Metropolitan Plan Making: An Analysis of Experience with the Preparation and Evaluation of Alternative Land use and Transportation Plans (Philadelphia, PA, Regional Science Research Institsute, 1970).
Brand, Daniel et al., "Techniques for Relating Transportation Improvements and Urban Development Patterns" (Highway Research Board, No. 207, 1967) 53-67.
Brown, H. James, et al., Empirical Models of Urban Land Use: Suggestions on Research Objectives and Organization (New York,
NY, National Bureau of Economic Research, Exploratory Report 6, 1972).
Canter, Larry W., Environmental Impact Assessment (New York,
NY, McGraw-Hill Book Col, 1977) 153-162.
Chicago Transportation Study, Final Report (Chicago, IL, Chicago: The Study, I960).
Clark, Thomas, Clark-Hansen Model for Population Allocation (Denver, CO, University of Colorado, Article in Possession of Author, n.d.).
Crecine, J.P. A Dynamic Model of Urban Structure (Santa Monica, CA, Rand Corporation, 1968).
Denver Regional Council of Governments (DRCOG), 1985 Population and Household Estimates (Denver, CO, 1985.
Denver Regional Council of Governments (DRCOG), Population and Employment Forecast Distributions for 1990, 2000 and 2010 (Denver, CO, DRCOG, Public Affairs Office, 1986).
Denver Regional Council of Governments (DRCOG), Beltway Impacts:
A Review of the Literature and Implications for the Denver Region (Denver, CO, Discussion Paper, 1987).
Detroit Transportation Study, Final Report (Detroit, MI, Detroit: The Study, 1961).
Douglas County, Master Plan (Castle Rock, CO, 1986).

Dyett, M. et al., The Land Use and Urban Development Impacts of Beltways. (U.S. Department of Transportation, vol. 1, 1980).
Echenique, Marcial, "Urban Development Models: Fifteen Years of Experience" in Urban development Models, Baxter, et al. eds.(Lancaster, G.B., Construction Press Ltd., LUBFS Conference Proceedings No.3, 1975).
Economic and Demographic Forecasting Team, Suburbanization and Beltways (Federal Highway Administration, Interim Report on Beltway Impacts, 1972).
Engelon, Rodney E., Coordination of Transportation System Management and Land Use Management (Washington, D".C. National Cooperative Highway Research Program, No. 93, 1982).
Foot, D., Operational Urban Models: An Introduction (New York, NY, Methren and Company Limited, 198177
Fleisher, Aaron, "Review: A Model Metropolis" (Journal of the American Institute of Planners, May, 1965), p. 175.
Forrester, J., Urban Dynamics (Cambridge, MA, M.I.T. Press,
Gamble, Hays B. and Thomas B. Davinroy, Beneficial Effects Associated with Freeway Construction: Environmental, Social, and Economic (Washington, D.C., National Cooperative Highway Research Program, No.193, 1978).
Garin, R.A., "A Matrix Formulation of the Lowery Model for Inter-Metropolitan Activity Location" (JAIP, 31, 1966) 361-364.
Haase, E.N., Centennial Parkway Project IXFCU470-1(1): Final Environmental Impact Statement (Denver, CO, Colorado Department of Highways, 1980).
Hansen, W.G., "How Accessibility Shapes Land Use" (JAIP May, 1959), 73-76.
Harris, Britton, "New Tools for Planning" (Journal of the American Institute of Planners, Special Issue Vol. 31, No. 2,
1965), 90^97:
Harris, Britton, "Model Building and Rationality" (Journal of the American Institute of Planners, Special Issue Vol. 31, No. 2, 1965), 90-94.
Harris, C.C., Jr. et al. Regional Economic Effects of Alternative Highway Systems (Cambridge, MA, Ballinger Publishing Company, 1974).
Harvey, Greg, LOTUS 1-2-3 Desktop Companion (San Fransisco, CA, Sybex, Inc., 1987)

Hill, Donald M., "A Growth Allocation Model for the Boston Region (Journal of the American Institute of Planners, 31, No.
2, 1965) 111-120.
Holmes, Edward H., Coordination of Urban Development and the Planning and Development of Transportation Facilities (Washingtoin, D.C., Department of Transportation, Federal Highway Administration, Office of Planning, 1974).
HUD, Denver Metropolitan Area-Wide Environmental Impact Statement (Denver, CO, Department of Housing and Urban Development, 1978).
Ingram, G.K., J.F. Kain, and J.R. Ginn, The Detroit Prototype of the NBER Urban Simulation Model (Washington, D.C., National Bureau of Economic Research, Urban and Regional Studies No.1,
1972) .
Jain, Ravinder K. and Bruce L. Hutchings, Environmental Impact Analysis: Emerging Issues in Planning (Chicago, IL, University
of Illinois Press, 1978).
Kilbridge, Maurice D., R.P. O'Block, and P.V. Teplitz, A Conceptual Framework for Urban Planning Models (Boston, MA, Harvard University, Division of Research, Graduate School of Business Administration, 1970).
Krueckeberg, D.A. and Silvers, Arther L., Urban Planning Analysis: Methods and Models (New York, NY, John Wiley and
Sons, 1974).
Lee, Colin, Models in Planning, An Introduction to the Use of Quantitative Models in Planning TNew York, NY, Pergaraon Press,
Lee, Yuk, An Introduction to Urban Land-Use and Development Models (Denver, CO, University of Colorado, Article in Possession of Author, n.d.).
Lounsbury, J.F., et al., Land Use: A Spatial Approach (Dubuque, IW, Kendall/Hunt Publishing Company, 1981).
Lowry, Ira S., A Model of Metropolis (Santa Monica, CA, Rand Corporation #RM-4035-RC, 1964).
McHarg, Ian, Design with Nature (Garden City, NY, Doubleday and Company, Inc."i 1969)
National Transportation Information Service, Secondary Effects of Public Investments in Highways and Sewers (Newton, MA, Council on Environmental Quality Report No. EQC3172deff1, 1975).
Ottensman, John R., BASIC Microcomputer Programs for Urban Analysis and Planning (New York, NY, Chapman and Hall, 1985).

Papageorgio, G.J., ed., Mathematical Land Use Theory (Lexington, MA, D.C. Heath and Company, 1979).
Peat, Marwick, Mitchell, and Company, Study Design for Development and Application of "Empiric Activity Allocation Model (Atlanta, GA, Atlanta Regional Commission Report, 1972).
Pennsylvania Department of Transportation, Highway Interchange Area Planning and Control Study (Philadelphia, PA, Raymond, Parish and Pine, Inc., 1975).
Putman, S. H., Urban Residential Location Models (Boston, MA, Martinus Nijhoff Publishing Company, 1979).
Steger, W.A., "The Pittsburg Urban Renewal Simulation Model" (Journal of the American Institute of Planners, May, 1965), 144-150.
Steward, John Quincy, "Demographic Gravitation: Evidence and Applications" (Sociometry, Vol II, 1948), 31-58.
Thornton, City of, Comprehensive Plan (Thornton, CO, 1985).
U.S. Department of Transportation, Secondary Impacts and Consequences of Highway Projects (Washington, D.C., U.S. Department of Transportation Report No. TST-77-24, 1976).
Voorhees, A.M.,"The Nature and Uses of Models in City Planning"
(Journal of the American Institute of Planners, May, 1959).
Wachs, M., "Social Trends and Their Implications for Transportation Planning Models" (Proceedings, Conference on Urban Transportation Planning in the 1980s, Special Report 196, 1981).
Webber, M.J., Information Theory and Urban Spatial Structure (London, G.B., Croom Helm Limited, 1979).
Webber, M.M., "Transportation Planning Models" (Traffic Quarterly, 1951).
Wilson, Alan, "Urban and Regional Models: Some Organizing Concepts to Describe the State of the Art", in Urban Development Models, Baxter et al. eds (Lancaster, G.B., Construction Press Ltd., LUBFS Conference Proceedings No. 3, 1975).
Wilson, G., et al., The Impact of Highway Investment on Development (Washington, DC, Brookings Institute, 196FT.

Appendix A (Harvey 1985)
Variable Measurements

C470 data -file:VARl Y1
56.02 5910.00 8749.00 2839.00
56. OB 3597.00 3708.00 111.00
56.09 3173.00 6260.00 3087.00
56. 10 7590.00 7406.00 -103.00
56. 15 7492.00 9876.00 2384.00
56. 16 6073.00 7342.00 1269.00
66.03 3773.00 3302.00 -471.00
67.03 13963.00 21266.00 7330.00
68.06 10.00 4.00 -6.00
120.12 9644.00 9463.00 -181.00
120.13 6506.00 6300.00 -206.00
120.19 5841.00 16522.00 10681.00
141.00 4763.00 9505.00 4742.00
C470 data -file:VAR2 HOUSEHOLDS XI 1980 1985 AB.CHANGE
56.02 1836.00 2952.00 1116.00
56.08 1154.00 1204.00 50.00
56.09 888.00 1763.00 875.00
56. 10 2452.00 2472.00 20.00
56. 15 2370.00 3556.00 1186.00
56. 16 1756.00 2200.00 444.00
66.03 1286.00 1141.00 -145.00
67.03 4101.00 6434.00 2333.00
68.06 5.00 2.00 -3.00
120.12 2884.00 2881.00 -3.00
120.13 1818.00 1B66.00 48.00
120.19 1741.00 5155.00 3414.00
141.00 1507.00 3294.00 1787.00
C470 data file:VAR3 X2
56.02 3.21 2.96 -0.25
56.08 3. 12 3.08 -0.04
56.09 3.57 3.55 -0.02
56. 10 3.06 3.00 -0.06
56. 15 3. 16 2.87 -0.38
56. 16 3.46 3.34 -0. 12
66.03 2.93 2.89 -0.04
67.03 3.40 3.31 -0.09
68.06 2.00 2.00 0.00
120.12 3.34 3.28 -0.06
120.13 3.57 3.38 -0. 19
120.19 3.35 3.21 -0. 14
141.00 3. 16 2.89 -0.27

C470 data file:VAR4 X3
56.02 1168.00 1729.00 561.00
56. OS 932.00 961.00 29.00
56.09 2046.00 4037.00 1991.00
56. 10 4836.00 4770.00 -66.00
56. 15 3652.00 4814.00 1162.00
56. 16 2976.00 3598.00 622.00
66.03 5180.00 4533.00 -647.00
67.03 2748.00 4193.00 1445.00
68.06 1.00 0.00 -1.00
120.12 4869.00 4778.00 -91.00
120.13 4058.00 3930.00 -128.00
120.19 690.00 1952.00 1262.00
141.00 66.00 131.00 65.00
C470 data -filesVARS EMPLOYMENT X4 1976 1985 AB.CHANGE
56.02 479.00 999.00 520.00
56.08 385.00 903.00 518.00
56.09 197.00 662.00 465.00
56. 10 1974.00 4175.00 2201.00
56. 15 462.00 1669.00 1207.00
56. 16 1804.00 3191.00 1387.OO
66.03 617.00 1230.00 613.00
67.03 1814.00 6931.00 5057.00
68.06 2461.00 12030.00 9569.00
120.12 719.00 1445.00 726.00
120.13 505.00 641.00 136.00
120.19 637.00 358.00 -279.00
141.00 664.00 1383.00 719.00
470 data file:VAR6 X5
56.02 888.00 1920.00 1032.00
56.08 704.00 704.00 0.00
56.09 432.00 960.00 528.00
56. 10 528.00 960.00 432.00
56. 15 848.00 960.00 112.00
56. 16 400.00 520.00 120.00
66.03 256.00 386.00 130.00
67.03 592.00 3080.00 2488.00
68.06 1280.00 1640.00 360.00
120.12 1280.00 1280.00 0.00
120.13 1120.00 1120.00 0.00
120.19 80.00 100.00 20.00
141.00 120.00 12800.00 12680.00

56.02 5.00 27.50 5.50 13.75
56.08 2.00 15.00 7.50 7.50
56.09 1.50 20.00 13.33 10.00
56. 10 1.50 10.00 6.67 5.00
56. 15 2.00 15.00 7.50 7.50
56. 16 2.00 10.00 5.00 5.00
66.03 0.30 5.00 16.67 2.50
67.03 5.00 30.00 6.00 15.00
68.06 4.30 20.00 4.65 10.00
120.12 1.80 5.00 3.33 2.50
120.13 1.75 15.00 8.57 7.50
120.19 1.00 5.00 5.00 2.50
141.00 21.50 135.00 6.28 67.50
C-470 data -file-.VARB METER SETS X7 1979 1980 1981 1982 1983
56.02 128.00 113.00 90.00 297.00 368.00
56.08 0.00 49.00 29.00 16.00 13.00
56.09 0.00 186.00 264.00 125.00 139.00
56. 10 0.00 2.00 0.00 4.00 2.00
56. 15 0.00 178.00 158.00 110.00 411.00
56. 16 0.00 258.00 120.00 66.00 32.00
66.03 0.00 5.00 5.00 7.00 11.00
67.03 369.00 626.00 614.00 337.00 254.00
68.06 1.00 0.00 0.00 0.00 0.00
120.12 0.00 34.00 15.00 17.00 12.00
120.13 0.00 22.00 46.00 32.00 29.00
120.19 0.00 746.00 635.00 505.00 920.00
141.00 0.00 49.00 30.00 17.00 42.00

19B4 1985 1986 TOTAL AB.CHANGE
229.00 215.00 185.00 1625.00 1497.00
13.00 15.00 16.00 151.00 151.00
131.00 112.00 107.00 1064.00 1064.00
2.00 2.00 1.00 13.00 13.00
366.00 105.00 30.00 1358.00 1358.00
58.00 63.00 27.00 624.00 624.00
0.00 0.00 0.00 28.00 28.00
516.00 194.00 133.00 3043.00 2674.00
0.00 0.00 0.00 1.00 0.00
42.00 36.00 20.00 176.00 176.00
57.00 1.00 2.00 189.00 189.00
553.00 304.00 289.00 3952.00 3952.00
733.00 474.00 1155.00 2500.00 2500.00
C470 data -file:VAR9 X8
56.02 5898.00 8238.00 2340.00
56.08 3597.00 3488.00 -109.00
56.09 3173.00 5663.00 2490.00
56. 10 7590.00 7324.00 -185.00
56. 15 7491.00 9789.00 2298.00
56. 16 6073.00 7002.00 929.00
66.03 3773.00 3433.00 -340.00
67.03 13936.00 20430.00 6494.00
68.06 10.00 4.00 -6.00
120.12 9627.00 9543.00 -84.00
120.13 6494.00 6455.00 -39.00
120.19 5841.00 15712.00 9871.00
141.00 4763.00 8475.00 3712.00

Results of Simple Linear Regression

Regressi on
Std Err of Y Est R Squared
No. o-f Observations Degrees o-f Freedom
X Coe-f-f icient (s)
Std Err o-f Coe-f.
Regressi on
Std Err o-f Y Est R Squared
No. of Observations Degrees of Freedom
X Coefficient(s)
Std Err of Coef.
Output: Y/Xl
1.31 0. 15
Output: Y/X2
-15.19 3327.17 0.86
764.15 90.44
Regressi on
Std Err of Y Est R Squared
No. of Observations Degrees of Freedom
X Coefficient (s)
Std Err of Coef.
Regressi on
Std Err of Y Est R Squared
No. of Observations Degrees of Freedom
X Coef fi ci ent (s)
Std Err of Coef.
Output: Y/X3
0.87 0. 15
Output: Y/X4
262.17 3997.08 0.79

Regression Output:
Std Err of Y Est R Squared
No. of Observations Degrees of Freedom
X Coefficient(s) 3.41
Std Err of Coef. 0.57
Regression Output:
Std Err of Y Est R Squared
No. of Observations Degrees of Freedom
X Coefficient(s) 177.75
Std Err of Coef. 27.10
Regression Output:
Std Err of Y Est R Squared
No. of Observations Degrees of Freedom
X Coefficient Std Err of Coef. 0.68
Regression Output:
Std Err of Y Est
R Squared
No. of Observations Degrees of Freedom
X Coefficient(s) Std Err of Coef.

Selected Variable Measurements and Multiple Regression Results

Y1 XI X2 X3 X4
2839.00 1836.00 3.21 1168.00 479.00
111.00 1154.00 3. 12 932.00 385.00
3087.00 BBB.00 3.57 2046.00 197.00
-103.00 2452.00 3.06 4836.00 1974.00
2384.00 2370.00 3. 16 3652.00 462.00
1269.00 1756.00 3.46 2976.00 1804.00
-471.00 1286.00 2.93 5180.00 617.00
7330.00 4101.00 3.40 274B.00 1814.00
-6.00 5.00 2.00 1.00 2461.00
-181.00 2BB4.00 3.34 4869.00 719.00
-206.00 1818.00 3.57 4058.00 505.00
10681.00 1741.00 3.35 690.00 637.00
4742.00 1507.00 3. 16 66.00 664.00
31476.00 23798.00 41.33 33222.00 12718.00
Regressi on Output: Y/X1...XB
Constant -0.95
Std Err of Y Est 843.08
R Squared 1.00
No. of Observations 14.00
Degrees of Freedom 5.00
X Coefficient(s) 4.74 -243.52 0.07
Std Err of Coef. 2.93 403.34 0.32
X5 X6
888.00 13.75
704.00 7.50
432.00 10.00
528.00 5.00
848.00 7.50
400.00 5.00
256.00 2.50
592.00 15.00
12BO.OO 10.00
1280.00 2.50
1120.00 7.50
80.00 2.50
120.00 67.50
8528.00 156.25
113.00 5898.00
49.00 3597.00
1B6.00 3173.00
2.00 7590.00
178.00 7491.00
258.00 6073.00
5.00 3773.00
626.00 13936.00
0.00 10.00
34.00 9627.00
22.00 6494.00
746.00 5841.00
49.00 4763.00
2268.00 78266.00
-0.65 1.00 79.39 16.11
0.34 0.9B 27.50 2.69

Correlation Results (Pearson Product Moment Correlation)

Y1 XI X2 X3
Y1 1.000
XI 0.928 1.000
X2 0.925 0.989 1.000
X3 0.862 0.980 0.978 1.000
X4 0.890 0.965 0.970 0.951
X5 0.867 0.970 0.980 0.966
X6 0.884 0.903 0.918 0.857
X7 0.985 0.938 0.928 0.882
XB 0.929 1.000 0.989 0.979
X5 X6 X7 XB
1.000 0.965 O. 889 0.905 0.964

Step-Wise Regression Results

STEPs 1 ENTER X7 Rs .985 RSQUAREs .971
STEP = 2 ENTER X6 R = .992 RSQUARE= .985
STEP = 3 ENTER X4 Rs .996 RSQUAREs .993

Clark-Hansen Model Components (first run: 1986-1996)
A. District characteristic values (forecasted to 1996)()
B. Regression coefficients (weights)(A^)
C. ARt (forecasted growth change from 1986-1996)
D. Attractiveness scores by census tract (A^t)
E. Growth allocation factors by census tract (APit)

43830 13070 3.16 8641 13909 3033 5 43330
3440 4220 2 0 25501 1640 5 8440
3 0 0 C 2 11797 2.55 2507 227 3200 0 25671
43250 15670 2.76 5413 4130 3520 15 43250
20550 6607 3.11 18 5190 0 50 20350
5560 1544 3.6 1500 0 0 10 5560
9629 2767 3.48 146 3377 1152 40 9 523
14349 5459 2.72 1512 2275 1920 5 14649
16909 5975 2.33 1470 5942 1792 15 16905
9203 3121 2.95 1537 800 1230 0 9206
1706 674 2.53 30 2938 512 25 17 08
2654 945 2.62 69 -259 1600 45 2654
5776 2320 2.49 74 2059 1656 40 4366
29451 10674 2.76 425 2752 0 15 2 9 4 0 1
9376 3606 2.6 164 1042 300 50 8621
31553 12674 2.49 431 1754 12300 10 3155o

REG COhF 4.74 -243.52 0.07 -0.65 1
DELTA Rt 1986-1996 141136
-8795 67.03 -23145
-3274 66.Go -2 1776,
1790? 70.23 47 134
8105 70.31 21337
-1106 71.00 -2512
-1365 O': PC 3 J j - j o ^ 4
-706 vJ j b z> 1 j 0 O'
2344 35.03 7 4 j 7
309 o J 10 312
490 65.11 1255
450 - 5 12 1212
4856 35.13 13044
6263 63.14 1653/
31 j 3 i025 135.00 140.00 6161 16455
22406 141.00 5697o
5 0 u 2 0 TOTAL 141136

Clark-Hansen Model Components (second run: 1986-1998)
A. District characteristic values (forecasted to 1996)(C^t^
B. Regression coefficients (weights)(Ak^
C. AR^ (forecasted growth change from 1986-1996)
D. Attractiveness scores by census tract (A-j^)
E. Growth allocation factors by census tract (APit)

45520 15770 3.14 9762 15119 3259 5 49520
12250 6130 n C. 1532 27305 1732 5 12260
31459 12736 2.47 2522 0 3271 0 27055
62419 23032 2.71 6911 4 532 5320 5 62419
12558 4025 3.12 20 7331 9230 50 10196
6255 1700 3.63 1578 750 2240 15 6255
11238 3095 3.63 173 15192 2883 2 40 6836
16169 5939 2.7 1646 2457 2159 5 13 f4 0
18601 6739 2.76 4610 6568 3392 15 18501
9666 3333 2.9 1666 856 1422 0 6136
1765 712 2.43 83 4219 8032 25 1710
2795 1009 2.77 72 1875 13280 45 2707
8630 3555 2.43 79 28220 35616 40 7263
36416 13537 2.59 488 3423 2400 15 24559
12450 4940 2.52 196 2840 9120 50 521 5
37853 15542 2.42 516 2263 14115 10 13939

REG. COEF. 4.74 -243.52 0.07 -0.65 1
DELTA Rt 1935-1995 133350
-9249 67.03
-6465 63.05
20745 70.29
14671 70.31
10764 71.00
402 33.03
25200 33*53
7243 35.03
2704 35.10
1731 35.11
7357 65.12
15497 35.13
25279 35.14
2o So ^ 135.00
24207 140.00
57213 141.00
GROWTH DIST. -7739 -5445 17471
12j55 S0C5 j35 21222 6095 2277
14 j o 6195 13051 21235 22193 20536 43134
TOTAL 13^350