State of Maine's Environment 2005
Header image  
An Environmental Assessment  
     
   

Sprawl and the Future of Maine’s North Woods

Nathaniel H. Dick

 

Introduction

A growing concern that challenges the character of the American landscape is sprawl. Sprawl, or dispersed, auto-dependent development outside of compact urban and village centers, along highways, and in rural countryside, has the potential to permanently change the landscape of the United States.1 Between 1994 and 2002 real estate developers constructed approximately 1.5 million new units of housing each year, a majority of them being suburban single family homes.2  The expansion and migration of the public from urban centers to rural communities has focused attention on the challenge of managing and controlling this sprawl. Out of the entire 225 year history of the United States, in the last 15 years a quarter of the nation’s available land has been developed.1  Communities once densely assembled around the amenities of a typical town or city have transformed into communities that now stretch far from the centers of commerce.

The problem of sprawl is an increasing concern for Maine also. Over the last thirty years the fastest growing Maine towns have been “new suburbs” situated 10-25 miles from metropolitan areas.3  In 1990 to 1991 local governments in Maine spent $800 million more from1980-1981 in order to service widely dispersed households, an increase of 60% or $1,700 per household.3  On a per unit basis, costs are much higher to service homes spread across the land than costs to service traditional, densely settled neighborhoods.

During the 1980’s the population of Maine increased by less than 10%, but the vehicle miles driven increased by 57% or 40 million miles per year.3  The stress placed on the Maine Department of Transportation (MEDOT) to maintain new and old roads has increased to a level that requires higher taxes on the public. From 1987-1994 Maine municipalities were developing new roads at a rate of 100 miles per year to accommodate the spread of communities across the state.3 

            Similarly, there is a greater need for protection and police coverage for the larger more dispersed communities. From 1980-1993 the crime rate dropped by 17%, yet during the same time period the number of police officers increased by 10%.3  This expansion correlates to the spread of people to rural areas. The price of one patrol, a combination of a cruiser and four officers, costs taxpayers on average $175,000 per year.3

            The Maine taxpayer is charged for redundant infrastructure costs for the construction and renovations of schools associated with sprawl. Despite a decrease of 27,000 in enrollment in Maine’s primary and secondary schools enrollment from 1970-1995, $727 million were collected from taxpayers for new school construction during 1975-1995.3  Almost half the money was used to build new school capacity in fast-growth regions, while schools in denser populated regions are under-used and still in need of money for renovations. Busing costs have risen from $8.7 million in 1970 to a staggering $54 million in equivalent dollars even though the enrollment has decreased.3 

The implications for sprawl are not easily quantifiable, but most of the effects of sprawl are easily observed. Sprawl creates segregated communities, degrades landscapes, pollutes air and water, increases taxes and diminishes economic productivity.4  Sprawl requires lengthening of service routes for police, fire, emergency, road maintenance, and plowing. Also, sprawl neglects older cities and town centers, leaving them with declining populations and unused infrastructure.

Sprawl also drives the need for car use, which increases air pollution and obesity. People in sprawling communities are more likely to be trapped in traffic and suffer diminished productivity. Real estate analysts in the United States have noted that denser cities that boast car alternatives, primarily public transportation infrastructure, have seen better returns on investments in recent years than sprawling suburban agglomeration.4

Not only are the effects of sprawl anthropocentric, or seen or felt by the human population, they also stretch into the ecological realm. Over the last decade, according to the United States Department of Agriculture (USDA), Americans have converted open space to developed land at a rate of 2.2 million acres per year or 252 acres per hour.5  Increased population creates the need for more infrastructure, which in turn detrimentally affects ecosystems. The increase of impervious surfaces such as roads, parking lots, and buildings, enhances the rate of surface runoff and erosion, which compromises the integrity of habitats and ecosystems.5 

My research focuses on the issue of sprawl in Maine and the potential implications it has for Maine’s North Woods. To determine the potential effects of sprawl on undeveloped portions of Maine, I carried out an analysis comparing fast-growth and slow-growth communities. The analysis focused on the transitions the communities have made over a 20 year period from 1980-2000. I compared five regions to explore similarities, differences, drivers, and effects of different growth patterns. The five primary areas used in the study include Acadia, Baxter, Belgrade, Greenville, and Sebago (Figure 1). The primary indicators of sprawl I used in my analysis measure compactness (density) and connectivity (accessibility of one place to another) for each region. I then use these criteria to assess the vulnerability of the study regions, in particular, and the Maine North Woods, in general, to sprawl.

 

studyregions_trans_overview2

Figure 6:  Study regions and class 2-4 roads in Maine based on data from the Maine Office of Geographic Information Systems (MEGIS) from the years 1984-19866

 

Context

            The search for more natural amenities such as pure air and pure water, along with employment opportunities has brought many people from clustered, noisy urban centers to Maine.7  The promise of lower prices, cheaper lands, lower taxes, and privacy along with powerful government subsidies on mortgages entices young couples and retired workers alike to come to Maine.

            The search for the “triple dream” of house, land, and community, is a realization that drives the public to keep expanding out onto the fringes of towns and villages across the state.2  The truth behind the “triple dream” is that, over time, expenses increase with distance from community centers rather than decrease.

            The Augusta metropolitan area is representative of sprawl in Maine. A couple deciding between purchasing a home in Augusta or Windsor must consider real estate expenses. In 1997 property taxes were $600 more in Augusta versus a comparable home in Windsor.3  Most likely the choice of a private and secluded home in Windsor, which is situated in a rural setting, will be favored by the couple. The trouble with purchasing a home in Windsor is that costs to service a home in a rural setting increase property taxes, similarly raising transportation costs for the family. Eventually the rising property taxes and transportation costs more than offset the initial savings the couple received.

 

Methods

Case Selections  

I selected five regions to represent areas with and without sprawl, areas near protected areas, and regions with lakes. I defined the Acadia region as the towns of Bar Harbor, Gouldsboro, Sorrento, and Winter Harbor, which are adjacent to Acadia National Park. I selected this area to provide perspective into the formation and development of towns around Maine’s only national park. The Acadia region is a very popular area in Maine because of the attractions provided by the park and the implications for growth in this region are substantial.

I define the second region, Baxter as the two towns in Penobscot county, East Millinocket and Millinocket. I selected this area because these communities are situated in close proximity to Baxter State Park, Maine’s largest and most popular state park.

The Belgrade region in Kennebec County includes the towns of Belgrade, Rome, Oakland, and Sidney. I selected this region to incorporate perspective from a fast-growth community. The potential for growth in the Belgrade region is high because it is situated near a regional service center, Waterville, and provides recreational opportunities and real estate for people in search of summer or year round homes.

I defined the Greenville region of northern Maine to include Beaver Cove, Greenville, Munson, and Shirley all in Piscataquis County. The Greenville region, much like the Belgrade region, has implications for growth because of the recreational attractions, including Moosehead Lake and Lily Bay State Park. The growth of this community has been relatively calm in past years, but interest has blossomed recently with the proposal by Plum Creek Timber Company to develop over nine-hundred lots, along with a number of sporting camps around Moosehead Lake.8 

 

The final region I selected was the Sebago Lake region, including the towns of Raymond, Sebago, Standish and Windham. I chose this region to examine community that is situated around a recreational haven, much like Belgrade and Greenville that has experienced a large amount of growth, in the past and is today heavily developed.

Indicators

            The indicators of sprawl that I use in this analysis include compactness (density), represented by total population, population density, number of housing units, and housing density; and indicators of connectivity (accessibility of one place to another), represented by road length, road density, kilometers of road per person, and average commute time (Table 1). I analyzed each indicator based on available data for the years 1980 to 2000 and assess the growth trends for the period in each region. The indicators are based on population and housing data from the U. S. Census Bureau for the years 1980, 1990, and 2000.9, 10  Similarly, transportation data was analyzed from the Maine Office of Geographic Information Systems (MEGIS) from the years 1984-1986.6

 

                                                                                                                    

Table 3. Definition of indicators used in analysis of five study regions

Indicator

Definition

Population

Sum of people

Population Density

Sum of people divided by square kilometers of land area

Housing Units

Sum of housing units

Housing Density

Sum of housing units divided by square kilometers of land area

Road Length

Sum of roads in kilometers

Road Density

Sum of roads divided by square kilometers of land area

Km.Roads/Person

Sum of roads in kilometers divided by sum of people

Average Commute Time

Average commute time to work in minutes

 

Data Analysis

To tabulate road data I downloaded transportation data from MEGIS 1984-1986 based on Maine roads.6  In ArcGIS I calculated the total road lengths and road types for each of the five regions. For calculating road densities I divided the total sum of roads in the region and the total sum of class 2-4 roads in each of the five regions by total land area to produce two road densities.

 

Results

Compactness

            In three of the five regions population density increased, while two regions, Baxter and Greenville decreased. The largest change in population density occurred in Sebago, which increased 47.7% from 1980-2000, while Belgrade increased 31.5% and Acadia increased 6.9%. Baxter decreased 27.8% and Greenville decreased 12.5%.        

Figure 7:  Average population density for each of the five study regions. Population data and area are derived from the US Census Bureau9, 10 

 

            In terms of sprawl in urban areas, higher densities are positive changes because they encourage smaller lot sizes and efficiently utilize current infrastructure. In rural areas, however, population density can reflect expanding suburbanization. For the sake of this study high population densities and housing densities will be considered indicators of growth and potential signalers of sprawl. The Sebago, Acadia, and Belgrade regions all exhibit sprawling tendencies, while the Baxter and Greenville regions do not.

            The number of housing units in each town for the five regions was compiled as well. The housing trends, excluding Baxter, show trends similar to changes in population density, but show an overall increase in housing density in each region from 1980-2000. The region with the largest percent increase in housing density is Sebago (104.2%), while both Belgrade (97.2%) and Acadia (86%) made substantial increases. The Greenville and Baxter regions increased only slightly, 28.8% and 2.3% respectively.

 

Figure 8:  Average number of housing units per square kilometer or each of the five study regions. Population data and are from the US Census Bureau9, 10

Connectivity

To analyze the connectivity of the different regions, I examine road length, road density, types of roads, kilometers of roads per person, and average commute time to work in each of the five regions. The road types are categorized by class in Table 2. I first look at the road length in each region per class. To account for the different size of each region, I converted road length to road density (Figure 4). Consequently, I analyzed the road densities for all roads and for only roads class 2-4, the primary road types (Figure 5). There were no roads classified as “interstate” or “footbridge” in my study area.

 

Table 4. Road classes as categorized by MEGIS from 1984-19866

Class

Definition

1

Interstate

2

Primary Road

3

Secondary Road

4

Improved Road

5

Unimproved Road

6

Trail

7

Footbridge

 

Road densities are good indicators of sprawl because connectivity is a primary issue related to sprawl. The more populations migrate to rural areas the more infrastructure is required to facilitate transportation between homes and services. Higher road densities reflect a well-organized, accessible community. For the sake of this study, I consider higher road densities to reflect greater sprawl.

The results of the road density analysis suggests that the Baxter region has both the highest total road density, 4.8 kilometers of road per square kilometer, and class 2-4 road density, 8.5 kilometers of road per square kilometer. The Sebago region is the second highest, barely higher than the Acadia region, while the Belgrade region has the fourth highest density followed by the Greenville region.

Figure 4:  Length of roads by all classes for each of the five study regions based upon transportation data collected from MEGIS from 1984-19866

 

To devise a means of road length per person, I divided the total length of roads by the total population for each region. The result was a measurement of kilometers per person for each region. Population data was found from the census surveys of 1980, 1990, and 2000, while the road length data was taken from the period of 1984-1986.6, 9, 10  Therefore the kilometers of road per person analysis uses only road length data from one period versus population data from three different periods (Figure 6). The accuracy of this comparison is therefore limited. Nevertheless, this measure does provide a reasonable comparison across regions.

Similarly, data providing road length does not account for all infrastructure related to roads. Particularly, other impervious surfaces that are associated with roads include parking lots, driveways, and other parking spaces, are not accounted for in this data set.

 

Figure 5:  Road densities in the five regions for all roads and class 2-4 roads based on data from MEGIS 1984-19866

 

Figure 6:  Kilometers of road per person in the five regions of study based upon transportation and population data collected from both the MEGIS from 1984-1986 and the US Census Bureau from 1980, 1990, and 2000, respectively6, 9, 10

The final compilation of data for comparing connectivity of each region was the average commute time to work (Figure 7). Data for this comparison was taken from the 2000 US Census from each town within the five regions of study.10

National Average

 
The results of the average commute data reveal that the Sebago region (31.2 minutes) and the Belgrade region (27.2 minutes) have the longest average commutes to work. This suggests that both communities are not in close proximity to their centers of commerce and the primary employment opportunities for their population. Furthermore, both average commute times in these regions are above the national average of 25.5 minutes. The Acadia region has an average of 20.5 minutes to commute to work, while the Greenville region has an average time of 19.9 minutes. The Baxter region has the shortest average commute to work at only 13.4 minutes.

Figure 7:  Average commute time to work in the five regions based on US Census Bureau data from Census 200010

 

Commute time to work was a category in the United States census that was recently added. The 1980 and 1990 Census Surveys did not include this data as part of their survey. No comparison can be made between the three time periods regarding this category.

 

Relative Sprawl Comparisons

            To produce a relative sprawl index, I devised a five point scale to rank the categories where one represents the best or least sprawling and five the worst or most sprawling. I was unable to determine the percent change in road length, kilometers of road per person, road density of all roads, and road density of class 2-4 roads because of a lack of data available to allow comparison of these categories. The index is a sum of the rank of five different indicators: population density, housing density, road density of all roads, road density of class 2-4 roads, and average commute time (Table 3).

 

Table 5. Five variables used to calculate the sprawl indicators and rankings. Rankings for each variable, indicated in parentheses, are based on a scale from 1-5 with 5 the best or least sprawling and 1 the worst or most sprawling.

Region

Population Density (%Δ)

Housing Density (%Δ)

Road Density (total)

Road Density (class 2-4)

Avg. Commute Time

Sprawl Indicator

Sprawl  Rank

Sebago

47.7 (5)

104.2 (5)

3.3 (4)

1.45 (3)

31.2 (5)

.88

#1

Acadia

6.9 (3)

86 (3)

3.2 (3)

1.49 (4)

20.5 (3)

.68

#2

Belgrade

31.5 (4)

97.2 (4)

2.7 (2)

0.15 (1)

27.2 (4)

.60

#3

Baxter

27.8 (1)

2.3 (1)

4.8 (5)

8.5 (5)

13.4 (1)

.56

#4

Greenville

12.5 (2)

28.8 (2)

2 (1)

0.52 (2)

19.9 (2)

.36

#5

 

Sebago was clearly the most sprawling region. Population and housing densities had the largest percent changes over the span of twenty years than any other region studied in Maine. The average commute time to work also suggests that more families in this region are traveling to work from rural homes or areas distant from regional service centers or centers of commerce. Road density was second behind the Baxter region, while road density of class 2-4 roads was third behind the Baxter and Acadia regions.

Surprisingly, the Acadia region was the second most sprawling region studied. Despite the Belgrade region’s higher percent change in both population density and housing density, the Acadia region had higher road densities, 3.2 km of road per square kilometer in total road density, and 1.5 km of road per square kilometer in class 2-4 roads. The average commute time to work for the Acadia region was just about a half minute longer than the Belgrade region at 20.5 minutes on average.

The Belgrade region was the third most sprawling region. The Belgrade region had significant growth in population and housing densities, a 31.5 percent increase and a 97.2 percent increase from 1980-2000 respectively. The Belgrade region is very interesting because there has been so much growth, yet the road densities are relatively low at 2.7 kilometers of road per square kilometer in total road density and .15 kilometers of road per square kilometer in class 2-4 road density. The study indicates at this point that road densities have correlated to growth in population density and housing density for the Sebago region and Acadia region. This suggests either that the Belgrade region has utilized current infrastructure and roads, planned their growth, or it may be that the Belgrade region has added much more infrastructure and the data in this study is outdated.

The Baxter region experienced the largest percent decrease in population in the study, (27.8%). Furthermore, this region experienced the smallest percent increase in housing density, an increase of 2.3%. Both road density of total roads and road density of roads class 2-4 were the highest in the study, yet the average commute time in 2000 was only 13.4 minutes. The Baxter region is a contradiction. Despite having the highest road densities, the region has the highest rate of decrease in population density and smallest rate of increase in housing density. The Baxter region was a thriving community during the time when Great Northern Paper Company owned and operated their paper mill in Millinocket.11  The mill had been declining since the late eighties and the company filed for bankruptcy in 2002 then sold and opened as the Katahdin Paper Company. The close of the mill may explain the density decrease along with the small increase in housing density. Similarly, the high road densities suggest that infrastructure was built around the mill and was not needed to access other regions, since the mill was the heart of the community. The average commute time suggests that there was not a great need to travel outside the community since the mill was the center of commerce. As a consequence, the second least sprawling region studied was the Baxter region

Finally, the least sprawling region studied was the Greenville region. This region has the second least percent decrease in population density, 12.5%, also the second smallest percent change increase in housing density, 28.8%. The road density for total roads was the least in the study, while the road density for roads class 2-4 was the second least. The average commute time for the Greenville region was 19.9 minutes, roughly a six minute longer commute than the Baxter region, which had the shortest commute time.

 

Conclusions

The results of my analysis suggest that southern and coastal regions of Maine are currently experiencing the fastest growth rates. The areas of fast growth in my study, Sebago, Acadia, and Belgrade, share similar recreational opportunities and are all situated around regional service centers. Greenville, a gateway community to Maine’s North Woods, also shares similarities to these fast growth communities through the recreational opportunities they provide. Based on my research Greenville also currently has low population density, low housing density, low road density, and a low sprawl index. As a result the area is very likely to present a growth opportunity in the future based on the results collected from other popular lake regions, like Sebago and Belgrade. In order to maintain the current character of Maine’s North Woods growth controls must be implemented.

Some “smart growth” alternatives that cold help sustain the character of Maine’s rural northern communities should focus on encouraging growth towards established communities, increasing development density, and orienting development around transit.12  Building upon towns and cities already in place limits expansion outward and decreases costs for infrastructure renovation. Similarly, other “smart growth” tools that should be considered include residential development caps, litigation, and impact fees.13 

By setting a development cap, a limit is placed on the rate of growth an area can achieve. Development caps can be instituted by local governments and can have the potential to give voters the authorization to approve development permits. In 1999, more than 240 jurisdictions nationwide considered antisprawl initiatives.12  For example, Petaluma, California, limits the number of new residential unit approvals by the city council to no more than 1,500 in any consecutive three-year period, or an average of 500 per year.12

Litigation, similarly, is considered a “smart growth” tool. People that are disapproving of new development and suffer intangible costs still have the liberty to question the law. For example, towns neighboring Shelbourne, Vermont, sued Shelbourne, saying that they would suffer the costs of a new subdivision that was authorized for construction by the city.12 

Another “smart growth” tool that has been effective in Cook County, Illinois is an impact fee. Impact fees are charges that are given to a developer by a municipality or other government unit to compensate the municipality for infrastructure alterations made as a result of new development.13  The use of impact fees shifts the burden of paying for new infrastructure fees to the developer instead of the local government assuming costs.

By utilizing these “smart growth” tools, there is a hope for low-density sprawl to be avoided in the future, while preserving the character of Maine’s North Woods at the same time.

 

Literature Cited

 

1Howard Frumkin, Urban Sprawl and Public Health (Washington, 2004).

2Dolores Hayden, Building Suburbia: Green Fields and Urban Growth 1820-2000 (New York, 2003).

3Evan Richert, "The Cost of Sprawl", in, http://mainegov-images.informe.org/spo/landuse/docs/CostofSprawl.pdf (Augusta, Maine, 1997), 1-22.

4Molly O'Meara Sheehan, City Limits: Putting the Brakes on Sprawl (Washington, D.C., 2001).

5Sperling Forman, Road Ecology: Science and Solutions (Washington, 2003).

6Maine Office of Geographic Information Systems (MEGIS)  US Geological Survey (USGS), Trans (Augusta, Maine, 1989).

7Evan Richert, "Land Use in Maine: From Production to Consumption", in, Changing Maine (Gardiner, Maine, 2004).

8Jym St.Pierre, "Save Moosehead: Plum Creek development plan will cause wildlands sprawl", in, RESTORE: The North Woods (Hallowell, Maine, 2005), 1-2.

9US Census Bureau, "Summary Characteristics for Governmental Units and Standard Metropolitan Statistical Areas", in, 1980 Census Data (Washington, D.C., 1980).

10US Census Bureau, "American Factfinder", in, http://factfinder.census.gov/home/saff/main.html?_lang=en (Washington, D.C., 2005).

11David Woodbury, "The Fall of the Great Northern Paper Company", in, Bangor Metro Magazine (September 2005).

12Elizabeth Gearin, "Smart Growth or Smart Growth Machine?" in J. Wolch, ed., Up Against the Sprawl (Minneapolis, 2004), 279-307.

13Jean M. Templeton, "Land Use Planning Tools in Illinois", in W. Wiewel, ed., Suburban Sprawl: Private Decisions and Public Policy (Armonk, New York, 2002), 80-81.

 

 

 

 

State of Maine's Environment, Colby College, Environmental Studies Program
Content by Students in ES493: Environmental Policy Practicum
Philip Nyhus, Assistant Professor of Environmental Studies
5358 Mayflower Hill, Waterville, ME 04901 USA; Email Us