Penobscot Valley:
Prudent Investments Linking Our Towns III. HOUSING 1 A. Introduction Maine has historically ranked highly in terms of home ownership and housing affordability relative to other states in the United States. In 2000, Maine's homeownership rate of 76.5 percent was second only to that of Michigan (77.2%)2. In recent years, however, unprecedented housing problems have emerged in some regions of Maine (Rooks 2002). For example, the National Association of Homebuilders designated the Kittery region of Southern Maine as one of the ten least affordable housing markets in the United States in 2002 (MSHA 2002, p.5). Closer inspection of many of these housing problems reveals a fundamental and common driver - changes in the spatial distribution of population. At the state-level Maine has experienced modest population growth (3.8% from 1990 to 2000). However, in some cases, shifts in population away from traditional service centers and urban areas to suburban and rural communities have resulted in considerably higher population growth rates in suburban and rural communities (Maine State Planning Office 1997). The communities at the receiving end of these shifts range in their ability to accommodate such growth and hence the extent to which these changes in the spatial distribution of Maine's population nourish housing and other policy issues varies as well. The benefits and costs of these changes in the spatial distribution of population manifest themselves regionally. Interestingly and perhaps not surprisingly, discussions of regional solutions have emerged in response to these recent housing problems. Examples of such activities include the efforts of the Maine Affordable Housing Network, the Southern Maine Affordable Rental Housing Coalition, the Knox County Housing Coalition, the Workforce Housing Coalition, and the MDI Now project (Bouchard 2002; Chapman 2002).3 This section serves as a supplement to discussions of regional housing issues initiated by the Penobscot Valley Council of Governments (PVCOG), Bangor Area Comprehensive Transportation System (BACTS), and the Eastern Maine Development Corporation (EMDC). In the Penobscot Valley Prudent Investments Linking Our Towns (PVPILOT) Draft Regional Plan (BACTS 2002), housing is discussed in the context of three regional business and economic development strategies.4 Three specific housing policy issues were identified in this regional plan: limited incentives for private development of rental housing; housing affordability; and multi-community or regional planning. The remainder of Section III is organized around these three issues. While much of the discussion focuses on recent trends in the provision of housing, emphasis is also given to the rationale for addressing such issues on a regional basis. Private Development of Rental Housing The lack of incentives for private development of rental housing was noted as primary housing policy in Penobscot Valley (BACTS 2002). Incentives for private development of rental housing are closely linked with expected net returns for such investments. When private developers make decisions about the type and location of housing to be constructed within a region, they examine the benefits and costs of various projects. Ultimately, development decisions hinge on which types of housing units and locations offer the highest expected net return. In this report, readily available data are employed to explore the different levels of existing rental housing and rates of new development of multi-family housing units in the communities of the Penobscot Valley. Affordable Housing The second housing policy issue addressed in this report is the supply of affordable housing in Penobscot Valley. There is a perceived shortage of affordable housing in the Penobscot Valley, particularly of starter homes ranging in value from $35,000 to $50,000 (BACTS 2002). Two suggested implications of this shortage are a tighter than usual rental market due to a delay in home ownership and occupancy of sub-standard homes available in rural markets. Affordable housing, by definition, reflects the relationship between income and housing costs in a region. Accordingly, factors that affect the income distribution (e.g., educational attainment, employment opportunities) and housing stock are jointly responsible for shifts in the supply of affordable housing. As noted above in the discussion of rental housing, private provision of new affordable housing units depends on developers' expectations of net returns. In the case of affordable housing, variation in expected net returns depends largely on variation in state and federal programs that act as subsidies to private developers. Furthermore, affordable housing is complicated by additional factors such as not-in-my-backyard (NIMBY) objections. In this report, readily available data are employed to examine housing affordability and rates of new development of affordable housing in the communities of the Penobscot Valley. Multi-Community or Regional Planning The utility of a multi-community planning program is the final housing policy issue addressed in this report. A cursory review of the comprehensive plans of the communities in Penobscot Valley confirms their acknowledgement of the benefits of certain types of cooperative agreements (e.g., education; sewer and water; fire and police; and solid waste). In addition, there appears to be widespread recognition of the regional aspects of myriad land use issues (e.g., natural resource areas; bike trails; transportation services; shoreland zoning). However, few communities address the regional aspects of housing policy issues, aside from acknowledging the variation in burden of providing affordable housing units. In this report, a brief discussion of the applicability of regional cooperation to housing issues is presented based on the future housing objectives of the Penobscot Valley communities. B. Maine Housing Policy Issues Before addressing the three specific policy issues raised in the draft PVPILOT Regional Plan, it is helpful to consider current and emerging ousing policy issues at the state-level. The Maine State Housing Authority (2002) draws attention to the following six housing policy issues: affordable housing in high growth areas; housing for the frail elderly; homelessness; housing in downtowns; preserving existing federal subsidized housing; and increasing the capacity of nonprofit organizations to meet the challenges of providing housing to populations with various needs (p.11). Discussions of these issues center on common themes related to the inter-relatedness of demographic, regional economic, and housing trends. Demographic trends influence housing policy issues in a variety of ways. Notably, different demographic groups have fundamentally different housing needs. As household formation and lifestyle trends change, housing demands are also expected to change. For example, the demand for different types of housing units is expected to vary with age and income. Smaller starter homes, apartments, and mobile homes are often more desirable to younger residents, ceteris paribus. Whereas, larger homes and renovations are often demanded by middle-aged residents, ceteris paribus. Older residents begin to demand housing that is tied with services such as medical treatment, transportation, and meals. Thus, one can expect changes in housing demand as the characteristics of Maine's population changes. Looking into the future, several agencies recognize the consequences of future increases in elderly populations and losses of younger populations (MSHA, 1999; MSHA 2002; EDCM 2002; MECD and MSHA 2002). A second relationship between demographics and housing policy issues involves affordability. Demographic characteristics such as educational attainment are often strongly correlated with income and in turn home ownership. Policy makers recognize that future changes in educational attainment and income can be expected to affect housing ownership trends, housing affordability, and the demand for housing assistance (MECD and MSHA 2002; EDCM 2002). MECD and MSHA (1999) emphasize the fundamental importance of understanding what populations are in need of housing assistance. One way of describing "in need" is to consider families and seniors, making 50 percent or less than the MSA median income and spending more than 30 percent of income on housing. MECD and MSHA (1999) identified 64,000 households in need of rental housing assistance statewide. Of these households, 45 percent (28,590) were seniors. While subsidized housing units were expected to help many of these 64,000 households, approximately 39 percent of these households (24,819) were not expected to be serviced by the existing subsidized rental housing units in their area of residence. The Bangor labor/housing market area had the third highest unmet need of renters (2,182) in absolute terms, after Portland (6,620) and Augusta (2,564). Affordable housing for young people and retirees is anticipated to be a major policy issue in Maine in the years to come (MSHA 2002). Recent data suggest that the Bucksport and Outer Bangor housing/labor market areas along with Somerset County have the most low-income elderly in need of housing statewide (MSHA 2002). The Maine State Housing Authority (MSHA) characterizes housing affordability of rental and owned housing using an affordability index. "The index measures the ability of a family at the median income to purchase a median price home in their labor market (MSHA 1999, p.4)." Values less than 1 indicate lack of affordability; values greater than 1 indicate affordability. Maps 1 and 2 display affordability trends based on this index. Map 1 presents the estimated indices for 2001 at the town level. Kennebunkport (0.39), Brooklin (0.45), Belfast (0.47), Rangeley (0.49), and Southwest Harbor (0.50) were the least affordable towns in 2001. Map 2 presents these same data at the labor market area level for 1997, 2000, and 2001. In 2001, the Belfast (0.64), Kittery/York (0.69), Stonington (0.76), Bar Harbor (0.76), and Biddeford (0.77) labor market areas were identified as the least affordable regions. These data suggest that affordability problems are most intense in the coastal and western regions of Maine. Voter approval of the affordable housing bond initiative in 2001 ($10 million dedicated to the affordable housing crisis in Maine) suggests that citizens support programs that encourage affordable housing. Unfortunately, it is expected that citizens support the concept of affordable housing but are not necessarily supportive when such housing is proposed in their very own neighborhood. Housing policy issues are also closely tied to employment and economic trends. The demand for housing will be a function of employment opportunities. In turn, regional economic development strategies often depend on available housing when encouraging new job creation and business startups. O'Hara (2001) outlines the connections between Maine's jobs, housing, and people, identifying the following strategies for the future health of Maine: (1) create apartments to make Maine attractive to young people; (2) employ neighborhood stabilization and improvement as a property tax strategy; (3) employ retirement housing as an economic development strategy; (4) build up the construction industry; and (5) build neighborhoods and communities instead of units. The unique character of Maine's housing stock distinguishes it from many other areas of the country. In 1999, Maine ranked 7th in terms of having the oldest housing stock (MSHA 1999). Based on the US Census 2000 data, Maine had 651,901 housing units in 2000 and 29 percent of these units were built prior to 1939. From 1990 to 2000, the state experienced a modest increase in housing units of 64,856 (11 percent)). The age distribution of housing varies throughout the state. The older age of Maine's housing complicates housing policy, making rehabilitation programs a logical part of any comprehensive housing policy strategy. The likelihood of outdated heating systems; inadequate plumbing, electrical systems, and insulation; structural decay; and lack of appropriate health and safety systems; and hazards such as lead based-paint are higher in areas where the housing stock is older. Maine ranks highly in terms of home ownership rate. In 2000, 72 percent of the state's housing units were owner occupied. While this is encouraging, it certainly complicates the provision of rental and multi-family units, as low home-ownership costs restrict the payments that can be gleaned in the rental market. Building permit data indicate sparse new development of multi-unit buildings statewide. During the mid-1980s (1984 to 1988), permits for buildings with greater than 5 units averaged 172 per year. Only 27 such permits for buildings with greater than 5 units were issued state wide in 2001, ranking Maine last in the nation in production of multifamily housing units at the state level. C. Penobscot Valley Housing Issues The PV PILOT study area, referred to here as the Penobscot Valley, includes the following municipalities: Bangor, Bradley, Brewer, Eddington, Glenburn, Hampden, Hermon, Holden, Kenduskeag, Milford, Old Town, Orono, Orrington, Penobscot Indian Island, and Veazie. The housing policies in these different communities reflect consistent goals of maintaining and repairing the existing housing stock; maintaining the vitality of neighborhoods, especially those in downtown or village centers; and encouraging the provision of affordable housing. Before addressing rental housing, affordable housing, and multi-community planning, background information on the housing stock of these communities is presented to provide a useful foundation for the consideration of these policy issues. A cursory examination of the housing stock in these communities draws attention to the variation of housing opportunities within the communities of Penobscot Valley. Table 1 presents the total number of housing units in 1990 and 2000 and the associated percentage change in the total number of housing units over this time period by community. In 2000, as the largest community, Bangor (14,587) stands out in absolute terms, with the highest number of housing units, followed thereafter by Brewer (4,064), Old Town (3,686), Orono (2,899), and Hampden (2,288). A review of the percentage changes in housing units distinguishes a different set of communities, as the smaller communities of Indian Island (100%), Glenburn (30%), Hermon (25%), and Bradley (21%) are currently experiencing more rapid growth in housing units. Notably, Holden (-1%) was the only community to experience a reduction in housing units from 1990 to 2000. Tables 2 and 3 display the number of housing units falling into four different housing types: single-family homes, multi-family homes, mobile homes and other housing units such as recreational vehicles or boats. Single-family homes included attached and detached 1 unit homes; whereas, multi-family homes refer to structures with more than 1 unit and include buildings ranging in size from 2 units to 20 or more units. Boats, RVs, and vans are included in the category labeled other. Table 2 summarizes the absolute number of units classified as these different types; Table 3 displays the percentage of total housing units by housing type. In 2000, greater than seventy percent of the housing stock in Hampden (79%), Indian Island (78%), Hermon (73%), Holden (73%), and Eddington (73%) was compromised of single-family homes. In 2000, the highest percentage of multi-family units occurred in the communities of Bangor (50%), Brewer (38%), Orono (44%), and Old Town (33%). In 2000, mobile homes accounted for greater than twenty percent of the housing stock in Kenduskeag (38%), Milford (30%), Glenburn (29%), and Holden (21%). Table 4 recasts the information presented in Table 2 and 3 in terms of changes between 1990 and 2000. It is interesting to note the variation in changes of housing types across communities. While increases in single-family units are observed across all communities in the Penobscot Valley, gains and losses in units of other types are observed in these different areas.5 Note in keeping with the trends of outward migration from urban areas, the higher percentage increases in single-family housing units are observed in Glenburn, Hermon, Bradley, and Orono. Table 5 displays estimates occupancy and vacancy rates by community for 1990 and 2000. Owner-occupancy and renter-occupancy rates are also shown for occupied housing units in Table 5. In 2000, with the exception of Bradley, Glenburn, Holden, and Indian Island, over ninety percent of the housing units are occupied in the communities of Penobscot Valley. Bangor (53%), Orono (52%), Old Town (40%), Brewer (38%), and Indian Island (38%) had the highest percentage of rental-occupied housing units. In contrast, over eighty percent of occupied units are owner-occupied in Glenburn (89%), Holden (88%), Kenduskeag (85%), Orrington (86%), Hermon (86%), Bradley (84%), Eddington (81%), and Milford (80%). Table 6 summarizes the age distribution of the housing stock in these various communities, displaying the percentage of housing units constructed prior to or during 1939, from 1940 to 1969, from 1970 to 1989, and from 1990 to 2000. Thirty-five percent or more of the housing units were built during or prior to 1939 in Bangor (45%), Old Town (38%), and Brewer (35%), implying that rehabilitation may be of particular concern in these communities. In contrast, more than 50 percent of the housing stock in the communities of Holden, Glenburn, and Kenduskeag was constructed from 1970 to 1989. Approximately nineteen percent or more of the housing units in Hermon, Glenburn, Eddington, Milford, and Indian Island were constructed between 1990 and 2000, indicating recent development in these communities. Table 7 displays the percentage of owner-occupied housing units falling into different valuation categories in 2000. Most communities offer a mix of housing in terms of valuation. Bradley (13%) and Indian Island (28%) have the highest percentage of housing units falling in the less than $50,000 category; Holden (12%), Veazie (13%), Orono (8%), and Hampden (7%) have the highest percentage of housing units valued above $200,000. Table 8 presents a similar classification of rental housing units, summarizing the percentage of renter-occupied housing units falling into different rental categories in 2000. Again, most communities offer a mix of rental housing in terms of cost. In 2000, over 50 percent of rental-housing units had a gross rent below $500 in Indian Island (86%), Old Town (62%), Brewer (57%), Bradely (57%), Holden (57%), Bangor (56%) Milford (52%), and Orono (53%). These basic data offer a useful description of the housing stock of the different communities. A few distinctions across the communities are noteworthy. First, Bangor and Brewer, the two largest communities and urban areas, have the greatest number of housing units, highest percentage of multi-family units, and higher rates of renter occupancy. Old Town and Orono exhibit similar trends on a smaller scale as does Hamden. Second, rates of growth in the units of the housing stock are currently highest in the smaller, rural communities, where the housing stock is largely comprised of single-family homes and mobile/other homes. In the next sections, additional data are presented to further the discussions of rental and affordable housing opportunities. D. Incentives for Private Development of Rental Housing As stated in the introduction, incentives for private development of rental housing are closely linked with developers' expectations of the net returns from such projects. When private developers make decisions about the type and location of housing to be constructed within a region, they examine the benefits and costs of various projects. Ultimately, development decisions hinge on which types of housing units and locations offer the highest expected net return. Because of differences in factors such as demographics, accessibility, land use and zoning ordinances, and permitting and design requirements, expected net returns vary within and across communities. Differences in access to and availability of state and federal programs which act as subsidies to private developers and encourage the construction of rental housing also account for differences in incentives. With respect to multi-family units, land use ordinances and permitting requirements have a tremendous influence on the feasibility and expected returns of development projects. Throughout this section, rental housing will often be discussed in the context of multi-unit housing. Readily available data on building permits do not distinguish rental and owner occupied housing. Thus, it is often assumed that multi-unit housing is rental housing. Difficult questions arise when considering how to assess normatively the current distribution of rental housing units. Determining the appropriate levels and rates of new construction of specific types of housing units is a challenging exercise. One approach suggests relying entirely on market forces. If there is excess demand for rental housing, the market should respond (e.g., higher rents) and in turn provide privately increased rental housing. Another approach calls for relying on a mixture of market forces and public actions to determine rental housing opportunities. Under this strategy, market forces are recognized and tinkered with using incentive programs such as subsidies or tax breaks. The second or mixed approach is a logical one for rental housing because of the highly regulated nature of the land market in numerous municipalities. Note that developers often have the option to build single-family or multi-family units. If regulations preclude or restrict greatly the number of units allowed (e.g., density), single-family projects are likely to be chosen over multi-family projects. Currently, there is a perception that developers do not consider private provision of rental housing to be a profitable exercise in the Penobscot Valley. Specifically, it is generally assumed that rental housing is too costly to provide given the expected return. The limited number of new multi-family housing units in the region tends to support this hypothesis. A notable challenge in Maine is that the affordability of non-rental units may cap the rents supported in the private rental market. The subsequent discussion focuses on the incentives for private new and re-development of rental housing, recognizing the perception that developers do not consider rental housing to be a profitable exercise in the Penobscot Valley. Readily available data on housing markets are employed to study the following research hypotheses. First, if the private market is not encouraging the development of rental housing, the number of new rental units constructed is expected to be limited. Second, if there is persistent excess demand for rental housing, contract rents are expected to increase over time. The different levels of existing rental housing as well as different rates of new development of rental housing units across the communities in the Penobscot Valley may reflect to some extent the variation in expected returns and the incentives provided to developers. If there is considerable variation in local incentives provided to private developers, spatial variation in rental housing units is expected across communities, ceteris paribus. Communities with lower costs or higher expected returns may have higher rates of rental housing. Referring back to Table 5, the communities of Bangor, Orono, Indian Island, Old Town, and Brewer have the highest percentage of renter-occupied housing units, with more than 35 percent of their housing stock devoted to this use. From 1990 to 2000, the largest absolute increases in renter-occupied housing units occurred in these same communities. Table 9 displays the number of renter-occupied housing units in 1990 and 2000 and the percentage of occupied housing units that are occupied by renters as opposed to home-owners. These figures do not distinguish new units and owner-occupied units or vacant units that have been converted to renter-occupied units so they cannot be used as a measure of new construction or development. Holden is the only community in which the number of renter-occupied housing units dropped from 1990 to 2000. Referring back to Table 5, the percentage of vacant housing stock varies considerably across communities in the Penobscot Valley. In 2000, higher rates of vacancy occur in Indian Island (39%), Bradley (16%), Glenburn (12%), Eddington (10%), and Holden (13%). Aggregating block-group level census data for 2000, we can examine the number and percentage of vacant units that are rental units. In absolute terms, the highest numbers of vacant rental units are found in Bangor, Brewer, Old Town, and Orono. In relative terms, vacant units for rent accounted for more than 25 percent of total vacant units in Bangor, Milford, Old Town, Orono, and Veazie. Anecdotal evidence suggests that the recent increase in students at University of Maine has resulted in low vacancy rates among rental units in the Penobscot Valley. Building permit data for non-single family homes provide one means to assess the private construction of rental housing. Figures 1a and 1b displays the number of building permits issued annually for residential buildings with more than 1 unit in the Penobscot Valley Communities from 1980 to 2001. Several observations can be made based on these figures. The late 1980s marked a high point in terms of such new construction but little new multi-family housing development has occurred in the region in recent years. Secondly, over this time period, the majority of permits were issued in the larger communities of Bangor (179 permits), Old Town (52 permits), Brewer (47 permits), Hampden (37), and Orono (36). In contrast, over this same time period, zero such permits were issued in Bradley, Eddington, and Kenduskeag and three or less were issued in Holden, Glenburn, Milford, and Veazie. Annual information on contract rents that is consistently organized across communities is difficult to find. Table 10 displays the median value of owner-occupied homes and the median rent of renter-occupied homes by community for 1990 and 2000. The nominal percentage change in median value and median rent is also shown in the far right column. In 2000, Hermon, Veazie, Orono, and Holden had median housing values in excess of $100,000. In contrast, the median housing value was below $80,000 in Bradley, Old Town, and Indian Island. In 2000, the median rent exceeded $500.00 in Glenburn, Orrington, Hermon, Hampden, Veazie, and Kenduskeag. Median housing values increased nominally by more than 20 percent in Kenduskeag, Holden, Hermon, Veazie, and Bradley. Median rents increased nominally by more than 20 percent in Glenburn, Kenduskeag, Orrington, Milford, Hampden, and Hermon. Median values are one measure of central tendency and alone do not describe fully the distribution of rents within a community. What do these data suggest in terms of the research questions that were stated in the outset of this section? From 1980 to 2001, limited building permits were issued to private developers for the construction of residential buildings ranging from 2 units to 5 or more units in the Penobscot Valley. In recent years, this holds especially true. From 1991 to 2001, the number of such permits were as follows: Bangor (8), Bradley (0); Brewer (6); Eddington (0); Glenburn (1); Hampden (8); Hermon (2); Holden (0); Kenduskeag (0); Milford (2); Old Town (3); Orono (1); Orrington (1); and Veazie (0). These data suggest that developers see little expected returns from such projects. The lack of incentives is difficult to gauge without more micro-level data. In some areas, such construction may be precluded by land use and building codes. Table 11 documents the housing policy issues raised in a subset of the Penobscot Valley communities' comprehensive plans. While affordability is a common theme across towns, few towns explicitly focus on rental housing opportunities. Two types of policy initiatives related to rental housing are mentioned. The policies of certain communities emphasize development and provision of new and affordable rental housing (e.g., Bangor, Hampden , Hermon, Milford, Old Town). A second policy emphasis of many towns (e.g., Eddington, Kenduskeag) is the maintenance of the safety and health of existing multi-family units. In cases where the latter is the dominant policy strategy, there appears to be little concern regarding the adequacy of the existing stock to meet the demands of renters. Finally, many communities recognize the connections between their land use and subdivision ordinances as well as design requirements and the incentives to build multi-family units. Throughout the Penobscot Valley, it is encouraging to note the willingness to consider novel planning tools such as cluster development, mixed-use zoning, transferable development rights, and contract zoning to provide the incentives for truly mixed housing stocks. Perhaps this shared appreciation for these tools may serve as the foundation for a regional strategy to provide and maintain rental housing. Hampden, Veazie, and Eddington specifically mention reaching out to a regional entity or propose a regional solution to housing policy issues related to rental housing. Numerous research questions arise in considering the issue of rental housing. For example, how do specific land use and zoning ordinance policies affect the incentives provided to developers? Will increasing rents affect the renewal of subsidized rental housing agreements and/or provide a market-based incentive for future private development of rental housing. Is there a stronger incentive to provide rental housing for specific populations? If so, who is being served (e.g., elderly) and who is not (e.g., poor)? Given limited development of new rental units, how can communities strive to encourage rehabilitation of existing units? If there is an established need for rental units, can their location be coordinated with efforts to revitalize downtown centers? A comprehensive treatment of such questions is beyond the scope of this report. E. Affordable Housing The second housing policy issue addressed in this report is the supply of affordable housing in Penobscot Valley. There is a perceived shortage of affordable housing in the Penobscot Valley, particularly of starter homes ranging in value from $35,000 to $50,000 (BACTS 2002). Two suggested implications of this shortage are a tighter than usual rental market due to a delay in home ownership and occupancy of sub-standard homes available in rural markets. A variety of factors may influence the supply of affordable housing units. As noted above in the discussion of rental housing, private provision of new affordable housing units depends on the expected returns. Variation in expected net returns from affordable housing development is closely related to variation in state and federal programs that act as subsidies to private developers. Note that if a developer has the opportunity to build a higher valued home that will result in a higher expected return, it is logical for the developer to opt not to build affordable housing. For these reasons, different levels of affordable housing and different rates of new development of affordable housing are expected across the communities in the Penobscot Valley. More difficult and interesting questions arise when considering how to assess normatively the current distribution of affordable housing units. In this report, trends in affordable housing are reported on using readily available data for the Penobscot Valley. In addition, information from the comprehensive plans of the various PVPILOT communities is used to assess the extent of concern regarding affordable housing in different communities. Shared and consistent concerns related to affordable housing suggest that affordable housing is a regional policy issue. The data shown in Tables 12 and 13 relate to housing affordability because they describe the percentage of income being devoted to home ownership costs and rental costs. Table 12 summarizes the percentage of owner-occupied housing units falling into different classifications of ownership costs as a percentage of income. Similar data are summarized in Table 13 for rental housing units. Affordability concerns arise when households are spending more than 30 percent of their income on housing. Based on these data, Kenduskeag and Indian Island stand out with higher percentages of income being devoted to home ownership costs on average. Glenburn, Orono, Old Town, and Bangor similarly have a higher percentage of rental households devoting upwards of 30 percent of their income to rent. Housing researchers typically consider a housing unit to be unaffordable to a family when residency requires expenditure of more than 30 percent of their income. Comparing the distributions for 1990 and 2000 yields interesting results. Figures 2 and 3 display the percentage of households devoting upwards of 30 percent of their income to home ownership or rental costs in 1990 and 2000. If this percentage rises over time, affordability of housing is expected to have decreased. Figure 2 shows with few exceptions (Holden and Hermon) that this percentage has risen with respect to owner-occupied homes. The greatest increases in this percentage occur in Glenburn, Kenduskeag, Milford, and Old Town. A more mixed trend is observed in Figure 3 which summarizes rental costs, as numerous communities recorded a drop in the number of renters devoting in excess of 30 percent of their income to rental costs (Bradley, Hampden, Kenduskeag, Milford, Orrington, Indian Island, and Veazie). The highest percentage increases in this category of renters occurred in Glenburn, Holden, Eddington, and Bangor. These data offer some support for a growing concern about housing affordability. It is important to note that affordability is both a function of incomes and housing values. If incomes rise at a faster rate than housing values, affordability increases, ceteris paribus. The data shown in Figures 2 and 3 represent a range of housing values and rents and income classes. Therefore, they do not necessarily target low- and moderate-income households. The Maine State Housing Authority (MSHA) characterizes housing affordability of rental and owned housing using an affordability index. "The index measures the ability of a family at the median income to purchase a median price home in their labor market (MSHA 1999, p.4)." Values less than 1 indicate lack of affordability; values greater than 1 indicate affordability. In 2001, the Maine State Housing affordability indices for the Penobscot Valley towns were: Bangor (0.93); Bradley (1.23); Brewer (1.03); Eddington (1.42); Glenburn (1.26); Hampden (1.05); Hermon (1.31); Holden (0.93); Kenduskeag (1.22); Milford (1.24); Old Town (1.00), Orono (0.90); Orrington (1.28); Indian Island (1.00); and Veazie (0.94). Based on this measure of affordability, Bangor, Orono, and Veazie are considered to be unaffordable. Comparing, 1997, 2000, and 2001, the same affordability indices measured at the regional scale (e.g., the Bangor Labor and Housing Market area) were 1.13 (1997), 1.15 (2000), and 1.09 (2001). While the index has dropped since 1997, the region still is considered affordable especially relative to other regions of Maine experiencing rapid population growth and escalating property values (refer to Map 1). Do we observe a back up of the rental market because starter homes are unavailable? Do we observe occupancy of substandard homes in rural areas? These questions are difficult to address using readily available data but some clues are evident in the census data. Notably, we do see an increase in the number of mobile homes in 1990 and 2000 across the different communities. Increases in these homes are recorded in Bradley, Brewer, Eddington, Glenburn, Hermon, Kenduskeag, Milford, Old Town, and Indian Island. In contrast, decreases in mobile homes are observed in Hampden, Holden, Orono, Bangor, Orrington, and Veazie. These changes may suggest that mobile homes are increasingly serving as starter homes. Assessing the condition of housing units is possible using a lack of adequate plumbing facilities as a proxy for substandard condition. Table 14 presents the percent of occupied housing units lacking adequate plumbing facilities in 1990 and 2000 by community. Based on this index, the occupancy of substandard homes is uncommon and decreasing in the communities of the Penobscot Valley. Returning to Table 11 which summarizes the housing policy issues documented in a subset of the Penobscot Valley communities' comprehensive plans, the expansion of affordable housing is a common objective across the majority of these communities. Bradley is a notable exception, for this town does not recognize affordable housing as an issue. The emphasis given to affordable housing may simply be an artifact of the Growth Management Act requirements for preparing a comprehensive plan. Section G of the Implementation Strategy calls for the following:
However, numerous communities express urgent concerns regarding the provision of affordable housing to elderly, homeless, and low-income populations. As noted in the prior discussion of rental housing, several communities express a willingness to consider the novel land use planning tools listed above. Such openness may prove invaluable in encouraging regional solutions to affordable housing problems. Several communities stand out in terms of proposing regional solutions or perspectives. Eddington specifically mentions developing a housing complex for the elderly with Bradely and Clifton. In addition, Veazie expresses interest in participating in a regional affordable housing program, if the costs and benefits if such programs are shared fairly by the communities in the region. Lastly, Hampden specifically notes expanding its relationship with a regional housing authority or non-profit as a policy objective. Many related research questions arise in considering the issue of affordable housing. For example, how do land use and zoning ordinance policies affect the supply of affordable housing? Will increasing rents affect the renewal of subsidized rental housing agreements? Are specific populations being more adversely affected than others in searching for affordable housing? If so, who is being served (e.g., elderly) and who is not (e.g., poor)? What strategies may encourage the provision of affordable housing in accordance with the GMA goal of 10 percent of new residential units? Are certain sites more appropriate than others for affordable housing? (accessible to employment, service, and cultural centers). While many of these questions are beyond the scope of this report, some are addressed in the subsequent section which focuses on multi-community planning. F. Multi-Community Planning The final housing policy issue discussed here is multi-community or regional planning. The interest in regional planning expressed by the PVPILOT Project (BACTS 2002) parallels that of communities throughout the United States (APA 2002a; APA 2002b). A range of planning activities (e.g., capital, land use, natural resource) could be addressed in a cooperative manner, as noted in this report's discussion of education and capital planning. While housing is affected by myriad planning activities, land use planning has a fundamental and powerful influence on the attainment of housing policy objectives such as the provision of rental and affordable housing. In terms of housing, a regional perspective begins with the coordination of housing priorities and policies. Orfield (2002) views regional housing planning as a means "to provide balance in housing choices within every community of a region (p.121)." Much of the research to date on regionalism and consolidation has focused on metropolitan areas with large urban cores and extensive suburban networks (Orfield 1997; Orfield 2002; Katz 2000 ). In these settings, balance implies having a range of low- to high-income housing in all parts of the metropolitan area, with the dual objectives of maintaining high-income households in urban centers and expanding low- and moderate-income housing opportunities in the suburbs. The Penobscot Valley is fundamentally different from many of the metropolitan areas that have been studied by these researchers. As a result, the details of the regional housing goals are likely to be different here than in well-studied regions such as the twin cities area of Minnesota. Despite these differences, many of the findings of these metropolitan studies are relevant to the discussions of rental and affordable housing emerging throughout Maine. Orfield (2002) notes three complementary policies for encouraging affordable and rental housing opportunities: reduced legislative barriers; fair-share requirements; and financial assistance for low-income housing (p.123). These three policies are germane to the fulfillment of the individual and regional objectives of the Penobscot Valley communities. Reducing barriers to the construction of multi-family housing or conversion of single-family housing to multi-family housing appears to be a common goal of many of the communities. The elimination of many barriers would follow from the revision of land use and zoning policies and building design requirements. Land use policies, design requirements, and permitting processes can significantly affect the costs of a private development and therefore can account for considerable variation in the expected return of a specific project in different communities. Differences in such policies across the region are likely to affect the spatial distribution of new rental and affordable housing units. Fair share requirements ensure that affordable and rental housing are available throughout the region. However, effective fair share requirements recognize differences across communities (population; job availability; past efforts to provide such housing) and account for such differences in setting the specific requirements for each community. The Growth Management Act's goal of 10 percent is one type of fair share requirement. However, there is much potential to design a more innovative system than this specification of a uniform standard. The final suggested policy of a regional housing fund is relevant here as well. While the private market is likely to broaden the choices of housing if barriers can be removed, there is still likely to be little encouragement of very low-income housing. For this reason, regional funds are an important part of regional housing planning. Explicit linkage of fair-share requirements and regional housing funds is a sensible policy consideration. Table 15 summarizes the regional cooperation policy objectives noted in a subset of the Penobscot Valley communities' comprehensive plans. It is encouraging to note that some communities note housing policies in their discussion of regional cooperation objectives. For example, Eddington, Clifton, and Bradely are exploring a regional elderly complex. Glenburn notes working with neighboring communities to set similar goals and priorities in providing affordable and low-income housing. Hampden and Orono express continued interest in working with The Housing Foundation a non-profit low-income housing organization. Lastly, Veazie specifically mentions the regional aspects of affordable housing. A common item concern across communities involves the siting and regulation of new multi-family housing. Emphasis is given to locating such housing in downtown, village, and/or growth areas. Identification of the "best" location for different kinds of housing is a complex process. Increasingly throughout the United States, attention is being given to policies that encourage "smart" growth. From an economics perspective, smart growth is sensible because it is based on the attainment of efficiency. In short, policy-makers acknowledge that different types of housing and developing patterns impose different costs to communities in terms of public services, quality of life amenities, and environmental stress. Smart growth in its most basic realization involves channeling new growth and new housing in growth centers with established infrastructure and creating incentives to prevent new housing development in outlying and rural communities. To work well, smart growth must be implemented on a regional basis. Without regional support, the incentives to locate in service centers and protect outlying and rural areas may be too weak to support the strategy. The Maine State Planning Office has done an outstanding job in conveying the relevancy of smart growth to Maine and initiating discussions with municipalities about the costs of servicing different types of development patterns (Maine State Planning Office 1997). Policies and tools advanced as part of smart growth efforts in other parts of the country may also prove helpful in determining what locations are most appropriate for specific types of housing (APA 2002a; APA 2002b). In the Penobscot Valley, regional agreement on the most appropriate sites for multi-family and special-needs housing may be a first step towards multi-community housing planning. G. Conclusions While regionally cooperative strategies to achieve housing goals have great potential, they are not expected to result in the immediate benefits anticipated from other cooperative activities such as school consolidation. When more severe housing problems emerge in the Penobscot Valley or tighter state regulations are passed designating affordable housing requirements, the incentive for municipalities to pursue regional housing planning strategies will increase and regional housing agreements will be more likely. The discussions and actions of regional housing groups in southern and coastal Maine will offer valuable information to the municipalities of the Penobscot Valley. Watching these regions accommodate changes in the spatial distribution of their population, changes in the demographic characteristics of their population, and changes in their housing stock may enable the PVPILOT communities to be pro-active rather than reactive in addressing future housing issues. In the short-run, improved communication of demographic and housing trends and land use planning tools may prove invaluable to the attainment of regional housing objectives.
References American Planning Association (APA). 2002a. Growing Smart Legislative Guidebook: Model Statutes for Planning and the Management of Change, 2002 Edition. Chicago Illinois, American Planning Association. American Planning Association (APA). 2002b. Planning for Smart Growth: 2002 State of the States. Chicago Illinois, American Planning Association, February 2002. Bangor Area Comprehensive Transportation System (BACTS). 2002. PV PILOT, Draft Regional Plan, March 28, 2002. Eastern Maine Development Corporation, Bangor, Maine. Brookings Institution. 2002. Beyond Merger: A Competitive Vision for the Regional City of Louisville. The Brookings Institution Center on Urban and Metropolitan Policy, Washington, DC. Bouchard, K. 2002. "Report shows housing crunch tightening grip on local renters," Portland Press Herald, September 20, 2002. Chapman, L. 2002. "MDI Towns Work on 10-year Plan: Island's four communities share growth, housing, and zoning issues," Bangor Daily News, June 6, 2002. City of Bangor. 1990. Bangor Comprehensive Plan Revision. June 1990. Prepared by the Planning Division, Department of Community and Economic Development. City of Brewer. 1995. Comprehensive Plan. City of Old Town. 1995. Comprehensive Plan. April. 1995. Economic Development Council of Maine (EDCM). 2002. Partnership for Prosperity: A State of Maine Economic Development Agenda. Economic Development Council of Maine, Augusta, Maine. May 2, 2002. Maine Department of Economic and Community Development and Maine State Housing Authority. 1999. Maine Consolidated Plan, Five Year Plan: 2000-2005, 2000 Action Plan. Maine State Housing Authority, Augusta, Maine. Maine State Housing Authority. 2002. The State of Maine's Housing 2002. Maine State Housing Authority, Augusta, Maine. Maine State Housing Authority. 1999. The State of Maine's Housing 1999. Maine State Housing Authority, Augusta, Maine. Maine State Planning Office. 1997. "The Cost of Sprawl." Augusta, ME: Maine State Planning Office. Katz, B. 2000. "The Need to Connect Smart Growth and Affordable Housing," Speech before the Vermont Affordable Housing Conference. November 29, 2000. Katz, B. and J. Bradley. 1999. "Divided We Sprawl," Atlantic Monthly 284(6): 26-34. Ohara, F. 2001. Houses, Jobs, and Maine People: 2001. Prepared for the 2001 Governor's Affordable Housing Conference. Planning Decisions, Portland Maine. Orfield, M. 2002. American Metropolitics: The New Suburban Reality. Brookings Institution Press, Washington, DC. Orfield, M. 1997. Metropolitics: A Regional Agenda for Community and Stability. Brookings Institution Press, Washington, DC. Ron Harriman Associates. 2002. Eddington, Bradely, and Clifton. Housing Needs Assessment. Prepared by Ron Harriman Associates, Bangor, June 2002. Rooks, D. 2002. "Affordable Housing." Maine Townsman. August/September. Town of Bradley. 1996. Preliminary Assessment. 1996. Town of Bradley. Comprehensive Plan. Prepared by Penobscot Valley Council of Governments. Town of Eddington. 2002. Town of Eddington Comprehensive Plan. Proposed Comprehensive Plan. Draft. Town of Glenburn. 1998. Comprehensive Plan. Town of Hampden. 2001. Comprehensive Plan 2001. Town of Hermon. 1995. Comprehensive Plan. Town of Holden. 1995. Town of Holden 1995 Comprehensive Plan. Town of Kenduskeag. 1991. Comprehensive Plan. Town of Milford. 1995. Comprehensive Plan. Milford Maine. June 1995. Town of Orono. 1998. Town of Orono Comprehensive Plan. January 1998. Town of Orrington. 2000. Comprehensive Plan. Town of Veazie. 2002. 2002 Veazie Comprehensive Plan. January 9, 2002. Go to the next section: Capital Planning For more information concerning this report or for additional copies, please contact Chris Boynton at mcsc@umit.maine.edu.
|
