Continuous Growth in the Delaware Valley Region
-- Forecasting Demand- and Supply-Driven Developments in 2020

To: Sean Thompson, Chair of the Delaware Valley Regional Planning Commission (DVRPC)
From: Leila Bahrami and Elizabeth Wang, Policy Analysts at Penn Urbanistics Initiative (PUSI)
Re: Continuous Growth in the Delaware Valley Region: Forecasting Demand- and Supply-Driven Developments in 2020
————————————————————————–

1. Executive Summary

Comprehensive regional development rooted in multi-nodal growth engines has been the new norm in the context of American urban planning. Cross-boundary collaboration brings benefits towards resource distribution, a balanced growth and overall sustainable development. Conjoining Pennsylvania, New Jersey, and Delaware and in proximity to New York City and Washington D.C., the Delaware Valley Region (DVR) serves as an example of regional collaboration in growth management. With the forecasts of next ten-to-thirty years development, the region witnesses the potential of increased population joining the workforce, along with the expansion and enhancement of regional infrastructure networks.

This report will summarize our recent study predicting future development demand based on the demand-driven factor of population change and the supply-driven factor of infrastructure network enhancement. It is predicted that development would intensify in Chester and Gloucester County as the northern parts of Montgomery, Bucks and Burlington County. The development would be further intensified along the proposed highway, concentrating in north Montgomery and Burlington County. The findings provide insights for planners to formulate better policies to manage urban sprawl, leverage smart growth and balance regional development.

2. Background

The DVR as core urban clusters of the Greater Philadelphia Region connects five adjacent states. While population grew from 5.1 million to 5.6 million, the developed area expanded by 1.5 times, signaling significant urban sprawl. The DVR region is forecasted to receive over 658,000 residents and more than 372,000 job opportunities from 2015 to 2045. (DVRPC, 2017) The region will face severe environmental damage if urban clusters continue sprawling at the preceding speed. Therefore, a projection study of future development demand will help observe development patterns and strategically alleviate sprawl through formulation of policy instruments.

3. Data and Study Methods

The study adopted a quantitative approach using spatial and demographic data, along with regression analysis to predict future development demands.

3.1 Data

Data of land cover in 2000 to 2010 were obtained from USGS. 2010 and 2011 population census and Esri Data and Maps have been referred to for demographic and transportation features. Population and employment projections came from DVRPC’s studies.

3.1.1 Land Cover Change as Development Indicators

The change of land cover from 2001 to 2011 indicates major developments in the decade. (Figures 1 to 3)

3.1.2 Major Development Engines

Population is the foremost driving factor in development since an increase in population will bring about growth of density and employment. Intensive population growth has been observed in Montgomery County and Bucks County in the north of Philadelphia, and northwest Gloucester County and Burlington County. (Figures 4 to 6)

Road network is another critical development engine to consider. In view of the high level of auto-dependency and consequently highway-oriented development patterns, our assumption is that development demand is likely to occur near highways to access better connection to destinations. (Figure 7)

3.2 Methods

To achieve a level of accuracy, the study divided the roughly 8 thousand square-kilometers land into over 12 million grids and used the roughly one-acre grids as units of analysis. By joining spatial data including land cover change, population distribution and distance to highway to the grids of both 2000 and 2010, we used logit regression analysis to study the development patterns <1>. Based on the hypothesis that future development demand shares continuity of current development, the patterns learned from the previous decade would serve as the model of development demands in the upcoming decades. Both demand- and supply-side development demand are projected based on future population projection and a proposed highway respectively.

4. Machine Learning of Development Patterns

4.1 Exploratory Analysis

The spatial lag (i.e. the calculation of spatial weighted average) to 2001 development illustrates development probabilities. Higher chance of development occurs in the periphery of the region and close to certain sections of the highway networks. (Figures 8 and 9) Land cover change also indicated positive correlation with population change from 2000 to 2001. (Figure 10) Among different types of land cover, farmland has the highest conversion rate, follow by developed land and forests, at 1.16%, 1.07%, and 0.67% respectively. (Figure 11)

4.2 2010 Prediction and evaluation

2010 prediction was based on analysis of six logit regression models. (Figure 12) As observed, most developments have been predicted with probabilities of lower than 20%. (Figure 13) The model failed to provide robust results. For example, 5% threshold correctly predicts a lower number of new development areas (Sensitivity) whereas incorrectly predicts a higher number of not changed areas (Specificity). (Figure 14 to 16) Similarly, all counties demonstrate higher numbers of incorrect predictions of not changed areas. (Figure 17) Thus this model is not representative enough (low generalizability) in 2010 development prediction.

5. Predicted Demand of Development

According to DVRPC (2016), population will continue increasing in the coming ten to thirty years.<2> Chester County and Gloucester County will witness close to one-third of population growth – the largest increase in density in the region.

Similar trends of population change per county are demonstrated by the model. (Figure 18) Based on the population projection in 2020, it is observed that development demand concentrates in Chester and Gloucester Counties and the northern periphery of Bucks, Montgomery, Burlington and Mercer Counties. (Figure 19)

5.2 Supply-driven Development

Both Bucks County and Burlington County are projected among the lowest population growth based on the current topography. However, both counties demonstrate potential connectors to edge cities of the Greater Philadelphia Region – in particular, north Bucks County is projected to hold roughly 30% to 50% increase of employment. (DVRPC, 2017) To leverage growth, a new highway is proposed crossing North Philly, reaching the north crossing the boundary of Montgomery and Bucks County and Burlington County, and connecting Allentown East and Atlantic City in the north and south respectively. (Figure 20) The 2020 predicted development demands, therefore further intensify in Bucks and Burlington Counties where the highway is proposed. (Figure 21)

5.3 Environmental Sensitivity

An environmental sensitivity analysis cross-examines the effects of the proposed transportation infrastructure. The predicted development, however, will evade forests and wetlands, as concentrated in southern New Jersey and other counties. (Figure 22 and 23)

6. Discussions and Suggestions

The following figures summarize this study by showing the development potential of each county. Development demand implies growth opportunities including jobs, real estate investment, small business and retail supporting bedroom communities. On the other hand, however, the projected growth in density unveiled a possible pattern of urban sprawl to north Bucks County. The proposed highway further connects north Bucks County and South Jersey where most land is currently Conservation Focused Areas (CFA). The environmental sensitivity analysis further revealed vulnerable preserved land that may be invaded under urban sprawl. While the consequent highway-oriented development is inevitable, your office is suggested to observe closely residential and commercial development along the new highway. In particular, CFA should continue as a tool in the growth control mechanism. Your office should work with relevant parties to study the land value and release land for development. Legislation and regulations should be formulated as precaution of over-development.

Footnotes:
<1> Spatial lag of development will be used to identify current developed locations from 2000 to 2010 and logit regression will be used to predict future developments.
<2> The projected population in 2020 are 640495, 840934, 543702, 1594787, 568337, 377328 , 514006, 307766, 459344 for Bucks County, Montgomery County, Chester County, Philadelphia County, Delaware County, Mercer County, Camden County, Gloucester County, Burlington County respectively, according to DVRPC(2016)’s population forecast. The projected population in 2040 are 691111, 918918, 645562, 1683402, 584329, 398669, 525101, 366383, 488026 respectively.

Reference

DVRPC (2016). “County and Municipal-Level Population Forecasts, 2015-2045”, website. Retrieved, 5/13/2020.
DVRPC (2017). “Connections 2045 Plan for Greater Philadelphia”, website. Retrieved, 5/13/2020.