Rural-to-urban migration and the geography of absentee non-industrial private forest ownership: A case from southeast Ohio

Applied Geography 96 (2018) 141–152

Contents lists available at ScienceDirect

Applied Geography journal homepage: www.elsevier.com/locate/apgeog

Rural-to-urban migration and the geography of absentee non-industrial private forest ownership: A case from southeast Ohio

T

Caleb Gallemorea,∗, Darla Munroeb, Derek van Berkelc a

International Affairs Program, 217 Oeschle Center for Global Education, Lafayette College, 730 N. High Street, Easton, PA 18042, USA Department of Geography, The Ohio State University, USA c Center for Geospatial Analytics, North Carolina State University, USA b

A R T I C LE I N FO

A B S T R A C T

Keywords: Private forest ownership Migration Absentee ownership Creative class

There is a growing literature on tropical forests that demonstrates ways in which rural-to-urban migration establishes dynamic connections between forest landscapes and urban areas. In the United States, context, however, studies of the geography of absentee ownership of non-industrial private forest (NIPF) lands focus on urban-to-rural migration for retirement or amenity purposes. Using parcel data sourced from local governments in an 11-county study area in central and southeastern Ohio, along with a range of openly available data, we analyze patterns of absentee ownership of NIPF parcels to determine the characteristics of areas where absentee owners reside. We hypothesize the rural-to-urban migration patterns, particularly of youth, will help explain where absentee NIPF owners of parcels in our study reside. We estimate models for all census tracts in the United States, finding that indicators of migration, creative class employment opportunities, and affluence are strongly associated with finding at least one absentee owner of an NIPF parcel in our study area. Considering these complex connections affecting NIPF parcels in a North American context could support improved forest management education, outreach, and planning efforts.

1. Introduction Absentee landowners are becoming an important part of the future of forests in the United States (US). Individuals and families account for 95% of private forest owners, collectively owning 61% of private forestland (see Fig. 1; Butler, 2008; Butler et al., 2016b; United States Forest Service, 2015). According to the National Woodland Owner Survey (NWOS: 2011 to 2013), approximately 37% of owners of nonindustrial private forests (NIPFs) greater than 10 acres across all states surveyed did not have their primary residence on their forest parcel (Butler, Miles, & Hansen, 2018a), a share that increased since the previous survey round in the early 2000s (Butler, Miles, & Hansen, 2018b). These new geographic arrangements pose important questions for management of NIPF land, where landowners are dealing with challenges related to invasive species (Gandhi & Herms, 2010; Poland & McCullough, 2006; Pyšek & Richardson, 2010), climate change (Chmura et al., 2011; Williams et al., 2010; Zhu, Woodall, & Clark, 2012), fire (van Mantgem et al., 2013), development pressures (Drummond & Loveland, 2010; Radeloff et al., 2010), and growing urban-to-rural migration (Plane & Jurjevich, 2009; Rickenbach & Kittredge, 2009; Xu, 2014). To date there is limited understanding of

∗

Corresponding author. E-mail address: [email protected] (C. Gallemore).

https://doi.org/10.1016/j.apgeog.2018.05.010 Received 26 June 2017; Received in revised form 7 May 2018; Accepted 11 May 2018 0143-6228/ © 2018 Published by Elsevier Ltd.

this segment of forest owners and as Petrzelka, Ma, and Malin (2013, p. 157) put it, absentee landowners remain the “elephant in the room” for US forest policy, despite that their forests are crucial for ecosystem service provisioning (Caputo & Butler, 2017; Petrzelka & Armstrong, 2015). Who are these absentee landowners? will they need to commute large distances to manage their land? what does land inheritance and migration mean for future management? Identifying ways absentee landownership generates novel geographic connections can support conservation efforts (van Herzele & van Gossum, 2008), even at subnational scales. Given that absentee NIPF owners often reside in urban areas (Feldpausch & Higgenbotham, 2006; Hughes et al., 2005), a geographic approach can help characterize which kinds of cities and which areas within cities are more likely to host owners, allowing for more targeted outreach to and characterization of these individuals. Identifying clusters of absentee owners for community building, for example, could allow extension agencies to target master volunteer and peer-learning programs (Kueper, Sagor, Blinn, & Becker, 2014; Sagor, 2012) more effectively. Previous research on absentee ownership emphasizes the role of urban-to-rural migration, particularly for second home or pre-retirement ownership, often driven by amenity considerations (Brown,

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Fig. 1. Family-owned forest lands in the United States (Hewes, Butler, & Liknes, 2017).

decentralization (Drummond & Loveland, 2010; MacDonald & Rudel, 2005; Miller, 2012; Rudel, 2009), and factors contributing to use of various management strategies (Beach, Pattanayak, Yang, Murray, & Abt, 2005; Bourke & Luloff, 1994; Erickson, Ryan, & De Young, 2002; Kelly, Gold, & Di Tommaso, 2017; Silver, Leahy, Weiskittle, Noblet, & Kittredge, 2015; Tian, Poudyal, Hodges, Young, & Hoyt, 2015; West, Fly, Blahna, & Carpenter, 1988). Our concerns here are rather different; rather than asking how NIPF landowners manage their land, a question that is well studied (Butler et al., 2007; Golden et al., 2013; Kelly et al., 2017; Kendra & Hull, 2005; Linghjem & Mitani, 2012; Miller, Snyder, & Kilgore, 2012; Petrzelka, 2012; Petrzelka & Armstrong, 2015; Salmon, Brunson, & Kuhns, 2006; Silver et al., 2015), we are interested in understanding where absentee landowners are more likely to reside. Absentee owners have been found to be less engaged in active management (Golden et al., 2013; Kittredge, 2004; Linghjem & Mitani, 2012; Miller et al., 2012; Petrzelka, Malin, & Gentry, 2012; Rickenbach & Kittredge, 2009), less versed in local ecological knowledge (Eriksen & Prior, 2011), less likely to benefit from the direct social contacts (Kittredge, Rickenbach, Knoot, Snellings, & Erazo, 2013; Mayer & Rouleau, 2013; Petrzelka & Armstrong, 2015; Rickenbach & Kittredge, 2009; Ruseva, Evans, & Fischer, 2014; Sagor, 2012; West et al., 1988), and less likely to be affected by extension efforts (Feldpausch & Higgenbotham, 2006; Hughes et al., 2005; Nielsen-Pincus, Ribe, & Johnson, 2015). Absentee ownership rates are spatially heterogeneous (see Fig. 2). Using NWOS data, Kaetzel, Majumdar, Teeter, and Butler (2012) find that rates of ownership as a secondary residence vary across US geographic regions, with higher rates in North Central, Northeastern, Pacific, and Mountain regions than for the Southeastern and South Central parts of the country. They are also uncertain. Aguilar, Cai, and Butler (2017), for example, find that 25% of the respondents to a forest-use

Johnson, Loveland, & Theobald, 2005; Drummond & Loveland, 2010; Eimermann, 2015; Gosnell & Abrams, 2011; Haugen, Karlsson, & Westin, 2016; MacDonald & Rudel, 2005; Miller, 2012; Rudel, 2009). Our interest here is different. Inspired by work on migration in tropical forest countries (Hecht, Yang, Basentt, Padoch, & Peluso, 2015), we argue that rural-to-urban migration also can set up complex and enduring relationships fundamentally shaping the geography of absentee NIPF ownership that will impact the future management of US forests. The geography of absentee landownership is not just about where absentee parcels are located but also about where their owners reside. Studying these relationships requires comprehensive data on parcel locations, as well as a way to link these parcels to specific absentee owner locations. Here, we use county-level tax administration data drawn from a sample of 11 counties in central and southeast Ohio as of 2014 for this purpose. We use these data to identify both absenteeowned parcels and their owners' places of residence. While this approach cannot provide the high-resolution owner characteristics data of national surveys (Butler et al., 2016b), it complements these data sources by lowering response biases (Butler, Hewes, Tyrrell, & Butler, 2017; Golden, Peterson, DePerno, Bardon, & Moorman, 2013; Rickenbach & Kittredge, 2009) and allowing for more explicit consideration of spatial relationships. We are particularly interested in outmigration of younger cohorts. We hypothesize that migration among younger age groups will tend to concentrate absentee ownership in areas with high levels of creative employment, which attract young outmigrants from rural areas who then inherit properties. 2. Migration, inheritance, and absentee ownership Studies of NIPF management have tended to focus on historical transformation (Kaplan, Krumhardt, & Zimmermann, 2009), urban 142

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Fig. 2. Absentee ownership rates of NIPF parcels greater than 10 acres in the continental United States, from the 2011–2013 round of the NWOS survey (Butler et al., 2018a).

heirs, regardless of age cohort. Markowski-Lindsay, Catanzaro, et al. (2017b), in a survey of NIPF owners in New York, Massachusetts, Vermont, and Maine, similarly found that about two-thirds had some form of estate planning in place for the transfer of their forest land. Inheritance remains an important mode of forestland transfer, particularly as NIPF parcel owners, as a population, age. NIPF parcels acquired via inheritance increased from about 20% of family forest owners in 2006 (Butler, 2008), to 31% in 2011–2013, accounting for 38% of family forest acreage (Butler et al., 2016b). In 2011–2013, 44% of primary NIPF owners were over 65 (Butler et al., 2016b), and, based on NWOS data, Markowski-Lindsay, Butler and Kittredge (2017a) find that about 18% of NIPF owners intend to transfer their land within five years. The probability of transfer increases with age (Butler et al., 2017; Markowski-Lindsay, Butler, et al., 2017a). Like absentee ownership, inheritance is spatially heterogeneous. Kaetzel et al. (2012) find that rates of acquisition by inheritance are much higher for Southeastern and South Central parts of the US than for other regions. In a study of family forest owners in Alabama, Georgia, and South Carolina, Majumdar, Teeter, and Butler (2008) found that 56% of the sample had acquired their forestlands through inheritance, that 19% expected to pass it on within five years, and that the least active managers also tended not to be heirs. Forest landowners' relationship to their forest often extend beyond a mere financial or income calculus with many being highly emotionally attached to their holdings (Dorning, Smith, Shoemaker, & Meentemeyer, 2015). This symbolic identification or sense of place (Creighton et al., 2016) and the inheritance and legacy of these lands play an important role in place meanings, affecting how NIPF owners experience their land and their identity as landholders (Andrejczyk, Butler, Tyrrell, & Langer, 2016; Creighton et al., 2016; Lӓhdesmӓki &

survey in Missouri did not have their primary residence within their forested parcel, while the NWOS estimates the statewide rate to be about 33% (Butler et al., 2018a). The literature on NIPFs and migration in the global North generally focuses on in-migration to rural areas, mostly for lifestyle reasons (Brown et al., 2005; Eimermann, 2015; Haugen et al., 2016), but several studies conducted in the global South hold that bidirectional migration can transform forest management strategies (Carr, 2008; Hecht et al., 2015; Kull, Ibrahim, & Meredith, 2007; Rudel, Bates, & Machinguiashi, 2002; Sunderlin & Resosudarmo, 1999). This work contributes to a growing recognition that migration establishes much more complex and enduring relationships between places (Hecht et al., 2015; Munroe, van Berkel, Verberg, & Olson, 2013). Inspired by these insights, we argue that while the literature on NIPFs and migration in the global North has not sufficiently explored the potential impact of youth migration for the geography of absentee NIPF ownership. While urban-to-rural migration is clearly an important part of the NIPF story in the US, it tends to be a retirement-age phenomenon (Nelson & Sewall, 2003; Plane & Jurjevich, 2009; Xu, 2014). Rural-tourban migration, by contrast, characterizes younger economically motivated cohorts (von Reichert, Cromartie, & Arthun, 2014). While this group might not at first be thought likely to affect NIPF ownership geography, we hypothesize that rural-to-urban youth migration should affect the geography of absentee NIPF parcel ownership as a result of inheritance. In other words, we expect that some migrants leaving an area rich in NIPF parcels seeking employment will subsequently inherit property from family members who remained in the area. Generational transfer of NIPF parcels is emerging as a policy and individual concern (Creighton, Blatner, & Carroll, 2016), and Butler et al. (2017) find that nearly 80% of forest owners worry about their 143

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Expect to Transfer to Family (%)

Inherited (%)

Not Primary Residence (%)

0−34 35−44 45−54 55−64 65−74 75+

0−34 35−44 45−54 55−64 65−74 75+

0−34 35−44 45−54 55−64 65−74 75+

80 First Owner

60 40 20 0 80

Second Owner

60 40 20 0

Age

Fig. 3. Percentages of NWOS respondents owning greater than 10 acres (Butler et al., 2018a) expecting to transfer forestland to family members, who inherited their forestland, and whose primary residence is not on their forest parcel, by NWOS age cohort.

2014; Fiore et al., 2015; Florida, 2003, 2004). Acting together, we would expect a combination of aging ownership, increasing transfer via inheritance, and youth migration to seek employment to generate increasing concentrations of NIPF ownership in creative employment areas, as younger generations migrate to urban areas and then inherit parcels from older generations or perhaps elect to purchase parcels back home due to a sense of legacy. These considerations lead to our primary hypothesis:

Matilainen, 2014). Kelly et al. (2017, p. 887), for example, found that 69% of respondents in a survey of California NIPF landowners cited “pass[ing] land on to my children or other heirs” as an important reason to own forest. This number was surpassed only by “enjoy[ing] beauty or scenery,” at 76%. These numbers are similar to findings from the Forest Service's NWOS, where “legacy” was the third highest ranked objective across owners and second highest when weighted by area, with at least 60% of each forest holding size class above 10 acres rating it important or very important (Butler et al., 2016a). Other evidence, however, suggests legacy falls toward the middle of motivations. Using qualitative coding of open-form responses to the NWOS, Bengston, Asah, and Butler (2011) find that family is mentioned as a motivation for owning woodland much less than income, recreation, or the parcel's status as a home. Belin et al. (2005), similarly, found in a survey of forest owners in Vermont, New Hampshire, and Massachusetts that only 19% list “inheritance for children” as one of their three most important reasons for owning woodland. The importance of family and legacy are not confined to older cohorts. Based on NWOS data, the youngest cohort has one of the highest shares of acquisition by inheritance (Fig. 3), though their absolute forest holdings may be small (Butler et al., 2018a). With the exception of the youngest cohort of first owners, younger cohorts tend to have a slightly higher share of non-primary residence ownership (Fig. 3). They are also quite likely to expect to transfer their forest land to family members (Fig. 3). Indeed, using data from the NWOS, Butler, et al. (2017) find that Generation X forest owners are more likely to list family as an important or very important aspect of their management objectives than older cohorts. There is also evidence inheritance can self-perpetuate. Majumdar, Laband, Teeter, and Butler (2009), for example, use NWOS to demonstrate that landowners who inherited their NIPF are a bit less than half as likely as non-inheritors to rate passing their property to heirs as “not important.” This growth in inheritance rates and the continued importance of inheritance and legacy as part of the identity of NIPF ownership occurs alongside continued rural-to-urban migration among younger rural dwellers (Plane & Jurjevich, 2009; United States Department of Agriculture, 2014). Millennials are particularly attracted to areas with strong creative sectors, with employment in science, engineering, architecture, design, arts, education, law, and management (Bereitschaft,

Hypothesis. Absentee owners of NIPF parcels in our study area will be more likely to reside in areas where a high percentage of the population is employed in the creative sector, controlling for population, income level, and overall employment.

3. Methods and materials To test the hypothesized relationship between rural-to-urban migration and absentee NIPF ownership, we collected parcel data from an 11-county region in central and southeast Ohio, US (see Fig. 4), combining these data with openly available national data from a variety of sources. We utilized logistic, poisson, and negative binomial regression to model our expected relationship (Long, 1997). First, we estimate, for all census tracts in the United States outside our study area, a logistic regression model predicting whether or not an absentee owner of a NIPF parcel in our study area resides in a given tract. We model the residence tracts of absentee owners whose parcels are found in our study area, rather than the presence of absentee owners of NIPF parcels in general, as it is not feasible to collect this sort of parcel-level data for all counties in the United States. Second, we repeat this analysis for all census tracts in our study area except that in this case we predict the total number of absentee owners of NIPF parcels in our study area in a tract. We explain the methods we use to generate these models in more detail in the following sections, but we begin with an overview of our study area. 3.1. Study area Central and southeastern Ohio is a prime example of a region increasingly experiencing forest threats in a context of significant private ownership (Law & McSweeney, 2013). Forest cover in the state has 144

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Fig. 4. Counties in the study area (US Census Bureau, 2015), with percentage changes in population between 2000 and 2010 (US Census Bureau, 2000, 2010). The statewide population change during this period, for comparison, was 1.6% (US Census Bureau, 2000, 2010). Projection: NAD 1983 Ohio South.

3.2. Dependent variables

doubled since the 1940s, now accounting for 30% of all land. Family forest owners hold a large proportion (73%) of this total, 5.8 million acres in the early 2000s (Widmann et al., 2009). Current threats include common invasive species, particularly Hemlock woolly adelgid (Adelges tsugae; Costigan, Soltesz, & Jaeger, 2015) and the loss of important mast-producing tree species (Widmann et al., 2009) as climate change and management choices, especially high grading, have contributed to a shift from oak-to maple-dominated forests. The region is also experiencing a range of dynamic land-use changes. Delaware County, in the northern part of the study area, is undergoing rapid peri-urban development, with employment centered on the state capital, Columbus (Munroe, 2009). The southeastern part of the study area, particularly Hocking County, has a growing amenity-driven service industry, attracting both tourists and migrants (Olson & Munroe, 2012; van Berkel, Munroe, & Gallemore, 2014). Like similar rural areas with high natural amenities (Ulrich-Schad, 2015), the forested part of the study area is experiencing both outmigration, particularly of youth (Cho, Patridge, & Feng, 2015), and inmigration, especially of households attracted to the area for amenity reasons (Goldring, 2014). While its economic and social dynamics are of course regionally unique, parts of the study area also share features in common with other post-industrial “outbacks” (McSweeney & McChesney, 2004), where consumption of nature coexists with economic transition. Because of these common features, we believe that if we can demonstrate that the patterns hypothesized above hold for absentee owners of NIPF parcels in our study area, this provides evidence justifying the effort required to collect data on similar areas to test generalizability. This is particularly the case for those places, such as the Southeastern and South Central US, where interheritance rates are particularly high (Kaetzel et al., 2012).

Existing county-level administrative databases offer efficient ways of analyzing ownership geographies at a high spatial resolution, complementing national surveys (e. g. Butler, 2008; Butler et al., 2016b). We rely on county tax parcel datasets, providing information on parcel location, size, and land cover, and the primary mailing address the owner listed for tax purposes.1 We sought these data from county auditor, recorder, and similar offices in 14 counties in the region as of 2014.2 Of these, data were not available in a usable form from Licking, Perry, or Vinton counties (Fig. 4). Data from all available counties were compiled using the open-source statistical scripting program R 3.4.3 (R Core Team, 2017). When latitude and longitude coordinates for the owner's mailing addresses were not explicitly given in the source data, we geocoded these locations using ArcGIS. We filtered the parcel data to select smallholder NIPF parcels first by excluding owners that include the word “bank” or “mortgage” in their names and by restricting our analysis to properties whose tax mailing addresses were attached to no more than five parcels. Following Butler et al.’s (2016b) definition of NIPF parcels, we limited our sample to parcels with greater than ten acres of forest, based on the National Land Cover Database for 2011 (Homer et al., 2015), resulting in 11,123 1 A few limitations of this approach should be noted. First, not all tax letters are sent to the landowner. Some owners choose to pay taxes from escrow accounts, which means that the tax letters will be sent to the owner's bank or to an escrow firm. Fortunately, these firms show up on a large number of tax documents, allowing us to filter most of them out based on their high number of ownerships. 2 The region was defined purposively to include Franklin county, where the state capitol, Columbus, is located, all surrounding counties, and the amenity migration counties in southeast Ohio.

145

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Fig. 5. Location of absentee-owned NIPF parcels and connections with absentee owners in the study area. Created with ggplot2 (Wickham, 2009) in R 3.4.3 (R Core Team, 2017).

3.3. Independent variables

complete observations (mapped in Fig. 5). To identify absentee landholders, we computed the great circle distance3 between the NIPF parcels and their respective owners' mailing addresses (Nychka, Furrer, Paige, & Sain, 2016). A one mile distance between parcels and owners is commonly used in US Forest Service studies (Butler, 2008) to define absentee ownership; however, given the size of parcels in many of our counties, we selected parcels located at least five miles (8.05 km) from their owners as absentee, amounting to 28.1% of the forested parcels in our dataset. Fig. 5 presents a map of the connections between of absentee NIPF owners and their parcels across the study area. Lines indicate ownership connections between absentee owners, depicted in white, and NIPF parcels, depicted in black. There is a clear pattern across the study area of ownership connections between rural areas and the Columbus metropolitan area. We find a particularly dense set of ownership connections between Columbus and the forested areas to the southeast, which are sites of significant of amenity-led exurban migration (Olson & Munroe, 2012). Fig. 6 presents the residence locations of owners of NIPF parcels in the study area across the contiguous US. The bulk of ownerships are found within the state of Ohio, despite that some of the study area is near Pennsylvania and West Virginia. There are also clear clusters of owners around major metropolitan areas, particularly New York, Los Angeles, and Washington, DC., and the large number of ownerships in Florida is suggestive of retirement migration. To test our hypothesis regarding creative employment and absentee ownership, we predicted whether or not a given 2010 census tract outside our study area (N = 69,994) is listed as a mailing address for at least one absentee-owned NIPF parcel in our study area using logistic regression. To assess the stability of our findings across different spatial extents, we also estimated models predicting the number of absentee owners of NIPF parcels in our study area for 2010 census tracts within the study area itself (N = 439), using poisson and negative binomial regression.

Studying how patterns of absentee ownership of NIPF parcels in our study region reflect rural-to-urban migration patterns required measures of creative sector employment, retirement age population, and a range of controls. Because sales-date data were not available for many of the counties in our study area, and because where these data were available they generally only recorded the most recent sale, it was not possible to use panel regression methods, so we attempted to account for the dynamic nature of the migration process discussed above by using data for several years prior to our parcel data collection year (2014) as independent variables. In other words, we use measures of past migration characteristics to predict absentee NIPF ownership patterns some years later. Our primary independent variable of interest is creative sector employment. We tested the association between absentee ownership and creative industries using the percentage of employment in creative industries, at the county level, between 2007 and 2011. This measure was derived from data used to create the United States Department of Agriculture's Economic Research Service's (2014) Creative Class County Codes and was calculated as the proportion of total county employment in occupations identified by creative class theorists: management, business and finance, computing and mathematics, architecture and engineering, natural and social sciences, law, education, arts, and marketing (McGranahan & Wojan, 2007, 2011). We associated with the presence of absentee owners of NIPF parcels in our dataset. As we expected creative employment migration to have an effect on the geography of absentee NIPF ownership over and above that of migration in general, it was necessary to include controls for overall migration patterns. To accomplish this, we used the US Internal Revenue Service's SOI Migration Data, computing the total outflows of household returns between 2000 and 2012 from our study counties to the counties in which each census tract was located (Pierce, 2015). To control for other tract characteristics, we also include the natural logarithm of the total tract population, the natural logarithm of income per capita, and the percent non-white population for all

3 That is, Euclidean straight-line distance taking into consideration the curvature of the Earth's surface, calculated using longitude and latitude coordinates.

146

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Fig. 6. Spatial pattern of absentee ownership of NIPF parcels in the study area for the contiguous United States.

census tracts,4 also from American Community Survey 2006–2010 estimates (Minnesota Population Center, 2011). We also control for employment opportunities using the natural logarithm of mean total employment at the commuting zone level from 2007 to 2011 (United States Department of Agriculture Economic Research Service, 2014). As the rural population ages, we should expect some retirement migration leading to absentee ownership (West, Cole, Goodkind, & He, 2014; Xu, 2014), so we also control for the percentage of the census tract population over 65, taken from the American Community Survey 2006–2010 mean estimates (Minnesota Population Center, 2011). To control for specialization in major employment categories other than the creative sector (Munroe & York, 2003) we compute the portion of total county-level employment in the primary sector (farming, forestry, fishing, and mining), manufacturing, and, as a control for any additional bank ownerships not removed from the dataset by the filtering techniques above, the financial sector (US Census Bureau, 2012; 2014). As further controls, we included the natural logarithm of the total employment for the commuting zone in which the tract was found, the reciprocal of the great circle distance between the centroid of our study area and the centroid of each tract, in hundreds of kilometers, and a binary variable coding whether or not the tract was located in Ohio. See Table 1 for an overview of variables.

3.4. Statistical models We modeled the locations of absentee owners of NIPF parcels in our study area, using logistic, Poisson, and negative binomial regression (Long, 1997), which can be thought of as an extension of Poisson regression (Flowerdew & Lovett, 1988; Lovett & Flowerdew, 1989) that permits the mean and variance of the dependent variable to be different (Hoekman, Frenken, & van Oort, 2009). The negative binomial distribution is given by: 1/ α

Γ (y + 1/ α ) ⎛ 1 ⎞ ⎜ ⎟ Γ (y + 1) Γ (1/ α ) ⎝ 1 + αμ ⎠

⎛ αμ ⎟⎞ ⎝ 1 + αμ ⎠

y

⎜

where:

lnμ = β0 + β1 x1 + ...+βk xk is the regression component that links the independent to the dependent variables (Long, 1997). An additional term, 1/α in the above equation, models the difference in variance of the outcome from what would be assumed in a Poisson regression. 4. Results Estimates of the presence of absentee owners of NIPF parcels in our study areas are reasonably fit across all three models, as measured by a range of model fit metrics (Table 2). Models indicate a positive relationship between the home address of absentee NIPF and locations with creative employment and high proportion of retirees. In our logistic model, for example, an increase in creative sector jobs as a share of county employment from 0% to 5% is associated with an increase of 13% in the odds of observing at least one absentee landowner of a NIPF

4 The concern here was that the effect of median income might be overestimated due to rural caucasian migrants avoiding non-white areas, a sort of inversion of the “white flight” phenomenon that contributed to the development of affluent caucasian suburbs around several major US cities (Crowder & South, 2008; Frey & Liaw, 1998).

147

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Table 1 Summaries of variables used in model estimations. Variable

Tracts Outside Study Area (N = 69,994)

Absentee NIPF Owner Presence

μ = 0.0151 σ = 0.122 Min = 0 Max = 1

Absentee NIPF Owners

Percent of County Employment in Creative Sector

ln Tax Returns Filed in Study Area in Previous Year, 2000–2012 + 1

ln Total Commuting Zone Employment

Percent of County Employment in Primary Sector

Percent of County Employment in Manufacturing Sector

Percent of County Employment in Financial Sector

Percent of Tract Population Over 65

ln Tract Income Per Capita

Non-White Percentage of Tract Population

ln Total Tract Population

1/Distance from Tract Centroid to Study Area Centroid

Tract is in Ohio

ln Absentee NIPF Parcels in Tract + 1

μ = 24.8 σ = 6.72 Min = 4.17 Max = 50.0 μ = 4.27 σ = 3.07 Min = 0.00 Max = 10.2 μ = 13.6 σ = 1.53 Min = 6.82 Max = 16.1 μ = 2.84 σ = 4.52 Min = 0 Max = 64.7 μ = 7.61 σ = 4.87 Min = 0 Max = 55.1 μ = 4.91 σ = 2.13 Min = 0 Max = 23.1 μ = 13.4 σ = 7.64 Min = 0 Max = 100 μ = 8.11 σ = 0.554 Min = 4.75 Max = 12.1 μ = 26.3 σ = 25.6 Min = 0 Max = 100 μ = 8.24 σ = 0.493 Min = 2.89 Max = 10.6 μ = 0.146 σ = 0.156 Min = 0.0288 Max = 6.51 μ = 0.0358 σ = 0.186 Min = 0 Max = 1

Tracts within Study Area (N = 439)

μ = 3.80 σ = 4.91 Min = 0 Max = 37 μ = 27.8 σ = 5.34 Min = 15.0 Max = 40.5 μ = 14.6 σ = 1.35 Min = 11.8 Max = 15.5 μ = 13.7 σ = 0.861 Min = 11.0 Max = 14.0 μ = 1.19 σ = 1.66 Min = 0.223 Max = 7.14 μ = 6.56 σ = 4.63 Min = 1.26 Max = 28.4 μ = 5.86 σ = 1.93 Min = 2.22 Max = 8.21 μ = 11.0 σ = 5.51 Min = 0.00 Max = 35.8 μ = 8.01 σ = 0.506 Min = 5.68 Max = 9.84 μ = 23.6 σ = 24.9 Min = 0.00 Max = 98.6 μ = 8.25 σ = 0.470 Min = 6.34 Max = 9.59 μ = 9.38 σ = 11.5 Min = 0.808 Max = 93.4

Source

Presence of at least one absentee owners of NIPF parcels in the study area residing in the tract, computed using parcel data collected from study area counties combined with 2011 forest cover data from the National Land Cover Database (Homer et al., 2015) Number of absentee owners of NIPF parcels in the study area residing in the tract, computed using parcel data collected from study area counties combined with 2011 forest cover data from the National Land Cover Database (Homer et al., 2015) Percent of total employment in which the tract is found accounted for by creative sector occupations, pooled values for 2007–2011 (United States Department of Agriculture Economic Research Service, 2014) Natural logarithm of the sum, plus one, of the total outflows of household returns from our study counties to the counties in which each census tract was found between 2000 and 2012 (Pierce, 2015) Average total employment, 2001–2011, in the 2000 commuting zone in which the tract is found (United States Department of Agriculture Economic Research Service, 2012, 2014) Average percentage of total employment, 2001–2011, in the county in which the tract is found that is in the primary sector (US Census Bureau, 2012; 2014) Average percentage of total employment, 2001–2011, in the county in which the tract is found that is in the manufacturing sector (US Census Bureau, 2012; 2014) Average percentage of total employment, 2001–2011, in the county in which the tract is found that is in the financial sector (US Census Bureau, 2012; 2014) Average percentage of total employment, 2001–2011, in the county in which the tract is found that is in the primary sector (United States Department of Agriculture Economic Research Service, 2012, 2014) Estimate of mean natural log of income per capita for the tract, 2006–2010, using American Community Survey (Minnesota Population Center, 2011)

Estimate of mean percentage of the tract's population that is non-white, 2006–2010, using American Community Survey (Minnesota Population Center, 2011) Estimate of mean natural logarithm of the tract's population, 2006–2010, using American Community Survey (Minnesota Population Center, 2011)

The reciprocal of the great circle distance to the study area centroid, in 100s of kilometers

Binary variable indicating the tract is located in Ohio

μ = 0.761 σ = 1.32 Min = 0 Max = 5.71

Natural logarithm of the count, plus one, of the number of absentee-owned NIPF parcels in the tract

previously filed in our study area, we expect the odds of observing an absentee owner of a study area parcel to increase by 13%. We also find positive and significant associations across all models with the percentage of the tract population over 65, tract income per capita, and total tract population across all models. For the national model, we additionally find a positive association with proximity to the study area and the location of the tract in Ohio. Finally, we find a statistically significant negative association with the percentage of the tract population that is non-white across all models. The study area models exhibit some different patterns, as compared

parcel in our study area. This term is not, however, statistically significant in the models for the study area. This is not particularly surprising, as the only major creative employment center in the study area is Columbus, which is also the main urban center. As a result, the population and employment variables are highly correlated with the creative employment variable, making it difficult to discern effects at the study area scale. Turning to our migration control variable, we find a statistically significant relationship at the national scale. For every 1% increase in the number of tax returns from 2000 to 2012 filed by people who had 148

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

Table 2 Estimated models.* = sig. at 0.05,** = sig. at 0.01,*** = sig. at 0.001. Standard deviations in parentheses. Variable

Absentee NIPF Owner Presence, Tracts outside Study Area

Number of Absentee Owners, Tracts within Study Area

Number of Absentee Owners, Tracts within Study Area

Type of Regression Model Intercept

Logistic −19.3*** (1.39) 0.0252** (0.00968) 0.103** (0.0337) −0.180*** (0.0564) 0.0142 (0.0140) −0.0101 (0.107) −0.0284 (0.0261) 0.0239*** (0.00417) 0.738*** (0.0641) −0.0214*** (0.00276) 1.18*** (0.103) 2.03*** (0.479) 2.10*** (0.300)

Poisson −10.4*** (2.99) −0.0679 (0.0953) −0.225 (0.239) 0.126 (0.195) −0.261 (0.162) −0.0142 (0.0230) 0.208 (0.108) 0.0489*** (0.00712) 0.625*** (0.0953) −0.0175*** (0.00146) 1.07*** (0.0645) −0.00130 (0.00247)

Negative Binomial −11.4*** (3.01) −0.0787 (0.0410) −0.255 (0.273) 0.115 (0.230) −0.276 (0.176) −0.108 (0.0265) 0.229 (0.121) 0.0558*** (0.00632) 0.759*** (0.0823) −0.0135*** (0.00130) 1.13*** (0.0736) 0.000457 (0.00225)

0.134*** (0.0176)

2500 0.259

0.162*** (0.0307) 1.531*** (0.169) 2036 0.403

439

439

Percent of County Employment in Creative Sector ln Tax Returns Filed in Study Area in Previous Year, 2000–2012 + 1 ln Total Commuting Zone Employment Percent of County Employment in Primary Sector Percent of County Employment in Manufacturing Sector Percent of County Employment in Financial Sector Percent of Tract Population Over 65 ln Tract Income Per Capita Non-White Percentage of Tract Population ln Total Tract Population 1/Distance from Tract Centroid to Study Area Centroid Tract is in Ohio ln Absentee NIPF Parcels in Tract + 1 ln(1/α) Bayesian Information Criterion McFadden's Pseudo-R2 (Negative Binomial Model compared against Poisson) Area under the Receiver Operating Characteristic Curve Tjur's Coefficient of Determination N

8225 0.262 0.854 0.154 69,994

with the model for tracts outside the study area. First, we find no statistically significant relationship with any of the county-level variables. As noted above, this could simply be a result of the relatively small sample size at the scale of the study area, which comprises only 11 counties, only one of which is highly urbanized. Further evidence supporting this interpretation comes from the fact that, despite these differences, several of the tract level variables retain the same sign and significance they have in the model for tracts outside the study area. This is true for the percentage of the population over 65, income per capita, the percentage of the population that is non-white, and the total tract population.

to be home to absentee owners than would be expected given rates of migration from our study area to these places. Furthermore, the statistically significant, positive association between our migration variable and the presence of absentee owners indicates the importance of rural-to-urban migration for the geography of NIPF ownership more broadly. This association is substantively quite significant: we estimate that a 1% increase in the number of tax returns filed in a census tract's county that were previously filed in a county in our study area increases the odds of observing an absentee owner residence in a tract by 10.8%. If these patterns are found in other areas with high level of forestland inheritance, they could lead to even greater numbers of absentee landowners in the future, as inheritance or retirement drive further ownership transitions. Increasing numbers of young absentee landowners could produce a mosaic of management philosophies in American forests, as young urban residents adopt an Arcadian landscape aesthetic stemming from an idealized sense of place associated with their family forest. Creighton et al. (2016), for example, find that generational land transfer is associated with changing value orientations. Once highly traditional utilitarian management perspectives may be confronted with these new attitudes as urban educated individuals inherit their family forests. These findings are quite consistent with the most recent NWOS's evidence that family forest owners are increasingly educated, affluent (Butler et al., 2016b) and have ecological objectives for owning their land (Markowski-Lindsay, Butler, et al., 2017a; Markowski-Lindsay, Catanzaro, et al., 2017b). There are several potential policy implications of these findings. First, as Kittredge (2004, p. 16) notes, “the audience [of extension services] is a moving target” and increasing numbers of family forest

5. Discussion While much literature discusses the role of amenity migration and second-home ownership as drivers of absentee NIPF ownership patterns, we find evidence that rural-to-urban migration, in addition to urban-to-rural migration, also explains rising levels of absentee ownership. Three variables in our model of absentee owner residence locations outside the study area are of particular interest. First, residence tracts of those owning an NIPF parcel in our study area have high percentages of employment in the creative sector (positive, significant relationship). Hosting locations also have higher percentages of population above retirement age. This is also suggestive of migration, though in this case toward the end of, rather than early in, the life cycle. We find evidence that these patterns hold even when controlling for a common indicator of migration, indicating that areas with high creative employment and high populations over 65 are even more likely 149

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

References

owners can strain extension agents (Feldpausch & Higgenbotham, 2006; Redmon et al., 2004). Given that outreach often fails to connect with many landowners, approaches like ours could support extension work - and land governance more broadly. While any given county's absentee NIPF owners may be spread out across a wide area, there may be opportunities for economies of scale in absentee-owner education by identifying urban areas where such owners are likely to live that can be targets for community building and peer-to-peer education programs. If, as studies indicate, embeddedness in local social networks can make important contributions to land management, it may be that one way to reach urban absentee landowners would be to try to build social networks where those individuals are found (Kueper et al., 2014). In addition, our findings suggest that one important avenue for further research on NIPF ownership may be to try to forecast who future owners will be - and where they will be. As forestland owners in the US age, decisions about what to do with forested land may increasingly be made by heirs who have spent considerable parts of their lives in urban areas and who may have very heterogeneous management objectives. Finding ways to support these new landowners may be a critical part of forest management policy for regions with high levels of transmission via inheritance in the coming years.

Aguilar, F. X., Cai, Z., & Butler, B. (2017). Proximal association of land management preferences: Evidence from family forest owners. Public Library of Science One. http:// dx.doi.org/10.1371/journal.pone.0169667. Andrejczyk, K., Butler, B. J., Tyrrell, M. L., & Langer, J. (2016). Hansel and gretel walk in the forest, landowners walk in the woods: A qualitative examination of the language used by family forest owners. Journal of Forestry, 114(1), 52–57. Beach, R. H., Pattanayak, S. K., Yang, J.-C., Murray, B. C., & Abt, R. C. (2005). Econometric studies of non-industrial private forest management: A review and synthesis. Forest Policy and Economics, 7(3), 261–281. Belin, D. L., Kittredge, D. B., Stevens, T. H., Dennis, D. C., Schweik, C. M., & Morzuch, B. J. (2005). Assessing private forest owner attitudes toward ecosystem-based management. Journal of Forestry, 103(1), 28–35. Bengston, D. N., Asah, S. T., & Butler, B. J. (2011). The diverse values and motivations of family forest owners in the United States: An analysis of an open-ended question in the National Woodland Owner Survey. Small-scale Forestry, 10(3), 339–355. Bereitschaft, B. (2014). Neighbourhood change among creative-cultural districts in midUS metropolitan areas, 2000-10. Regional Studies, Regional Science, 1(1), 158–183. van Berkel, D. B., Munroe, D. K., & Gallemore, C. (2014). Spatial analysis of land suitability, hot-tub cabins and forest tourism in Appalachian Ohio. Applied Geography, 54, 139–148. Bourke, L., & Luloff, A. E. (1994). Attitudes toward the management of nonindustrial private forest land. Society & Natural Resources, 7, 445–457. Brown, D. G., Johnson, K. M., Loveland, T. R., & Theobald, D. M. (2005). Rural land-use trends in the conterminous United States, 1950-2000. Ecological Applications, 15(6), 1851–1863. Butler, B. J. (2008). Family forest owners in the United States, 2006: A technical document supporting the forest service 2010 RPA assessment. PA: United States Forest Service Northern Research Station, Newtown Square. Butler, B. J., Hewes, J. H., Dickinson, B. J., Andrejczyk, K., Butler, S. M., & MarkowskiLindsay, M. (2016a). Family forest ownerships of the United States, 2013: Findings from the USDA forest Service's national woodland owner survey. Journal of Forestry, 114(6), 638–647. Butler, B. J., Hewes, J. H., Dickinson, B. J., Andrejczyk, K., Butler, S. M., & MarkowskiLindsay, M. (2016b). USDA forest service national woodland owner survey: A technical document supporting the forest service update of the 2010 RPA assessment. PA: United States Forest Service Northern Research Station, Newtown Square. Butler, B. J., Hewes, J. H., Tyrrell, M. L., & Butler, S. M. (2017). Methods for increasing cooperation rates for surveys of family forest owners. Small-scale Forestry, 16(2), 169–177. Butler, B. J., Miles, P. D., & Hansen, M. H. (2018a). National woodland owner survey table maker. Amherst, MA: United States Department of Agriculture, Forest Service, Northern Research Station. https://apps.fs.usda.gov/nwos/tablemaker.jsp. Butler, B. J., Miles, P. D., & Hansen, M. H. (2018b). National woodland owner survey tabler. Amherst, MA: United States Department of Agriculture, Forest Service, Northern Research Station. https://apps.fs.usda.gov/nwos/tablemakerVersion1.jsp. Butler, B. J., Tyrrell, M., Feinberg, G., VanManen, S., Wiseman, L., & Wallinger, S. (2007). Understanding and reaching family forest owners: Lessons from social marketing research. Journal of Forestry, 105(7), 348–357. Caputo, J., & Butler, B. (2017). Ecosystem service supply and capacity on U.S. family forestlands. Forests, 8(10), 395. http://dx.doi.org/10.3390/f8100395. Carr, D. (2008). Population and deforestation: Why rural migration matters. Progress in Human Geography, 33(3), 355–378. Chmura, D. J., Anderson, P. D., Howe, G. T., Harrington, C. A., Halofsky, J. E., Peterson, D. L., et al. (2011). Forest responses to climate change in the northwestern United States: Ecophysiological foundations for adaptive management. Forest Ecology and Management, 261, 1121–1142. Cho, C., Patridge, M. D., & Feng, B. (2015). A reportcard on rural and urban Ohio. Columbus, OH: Swank Program on Rural-Urban Policy. Costigan, K. H., Soltesz, P. J., & Jaeger, K. L. (2015). Large wood in central Appalachian headwater streams: Controls on and potential changes to wood loads from infestation of hemlock woolly adelgid. Earth Surface Processes and Landforms, 40(13), 1746–1763. Creighton, J., Blatner, K. A., & Carroll, M. S. (2016). For the love of the land: Generational land transfer and the future of family forests in western Washington state, USA. Smallscale Forestry, 15, 1–15. Crowder, K., & South, S. J. (2008). Spatial dynamics of white flight: The effects of local and extralocal racial conditions on neighborhood outmigration. American Sociological Review, 73(5), 792–812. Dorning, M. A., Smith, J. W., Shoemaker, D. A., & Meentemeyer, R. K. (2015). Changing decisions in a changing landscape: How might forest owners in an urbanizing region respond to emerging bioenergy markets. Land Use Policy, 49, 1–10. Drummond, M. A., & Loveland, T. R. (2010). Land-use pressure and a transition to forestcover loss in the Eastern United States. BioScience, 60(4), 286–298. Eimermann, M. (2015). Lifestyle migration to the north: Dutch families and the decision to move to rural Sweden. Population, Space and Place, 21(1), 68–85. Erickson, D. L., Ryan, R. L., & De Young, R. (2002). Woodlots in the rural landscape: Landowner motivations and management attitudes in a Michigan (USA) case study. Landscape and Urban Planning, 58(2–4), 101–112. Eriksen, C., & Prior, T. (2011). The art of learning: Wildfire, amenity migration and local environmental knowledge. International Journal of Wildland Fire, 20, 612–624. Feldpausch, A., & Higgenbotham, B. (2006). Texas absentee landowners managing for wildlife: Their goals, interests, and educational needs. Proceedings, 11th triennial national wildlife & fisheries extension specialists conference, October 14-18, Big Sky, MT.

6. Conclusion Existing literatures on migration and land use in the US context have tended to focus on how migrants affect land-use patterns in the places to which they move, but it is also necessary to consider how these movements create novel bidirectional connections between places (Seto et al., 2012). We have used our sample of counties in central and southeast Ohio as a case that allows us to better understand the geography of absentee ownership of NIPF land in the United States. We find evidence that rural-to-urban, in addition to urban-to-rural migration, may be an important factor explaining the geography of absentee NIPF ownership. We further hypothesized that some of this migration may be driven by a search for employment in creative industries, a pattern we have associated primarily with younger demographics. Though labor migration in the United States declined in the 2000s (Partridge, Rickman, Olfert, & Ali, 2012), as rural populations continue to age we would continue to expect rural-to-urban migration and inheritance to continue to affect absentee NIPF ownership patterns for some time. Interesting questions remain about this process, as owners always have the option of selling land upon moving or inheriting. Key questions for future research in this area, therefore, relate to how migrants or inheritors make the choice to retain or sell rural parcels. In our study area, larger parcels tend to have been in the same family longer, and are therefore more likely to be inherited, than smaller parcels, which tend to have been subdivided for development (Olson, 2012). Ultimately, improved understanding of how migration impacts the geography of absentee NIPF ownership in the United States is required to build comprehensive forest policy and target forest management outreach effectively. As this new form of landownership continues to grow, understanding the socioeconomic factors driving the decisions to become absentee NIPF owners may help to develop more effective ways to build connections among these individuals and families to support effective forest management. Studies taking a geographic perspective and using publically available data like those analyzed here may be quite helpful in addressing the “elephant in the room” (Petrzelka et al., 2013, p. 157) in US forest policy. Acknowledgements This research received support from the US National Science Foundation, Award 1010314, “CNH: Collaborative Research: Explaining Socioecological Resilience. Following Collapse: Forest Recovery in Appalachian Ohio.” 150

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

van Mantgem, P. J., Nesmith, J. C. B., Keifer, M., Knapp, E. E., Flint, A., & Flint, L. (2013). Climate stress increases forest fire severity across the western United States. Ecology Letters, 16, 1151–1156. Markowski-Lindsay, M., Butler, B. J., & Kittredge, D. B. (2017a). The future of family forests in the USA: Near-term intentions to sell or transfer. Land Use Policy, 69, 577–585. Markowski-Lindsay, M., Catanzaro, P., Bell, K., Kittredge, D., Leahy, J., Butler, B., et al. (2017b). Estate planning as a forest stewardship tool: A study of family land ownerships in the northeastern U.S. Forest Policy and Economics, 83, 36–44. Mayer, A. L., & Rouleau, M. D. (2013). ForestSim model of impacts of smallholder dynamics: Forested landscapes of the Upper Peninsula of Michigan. International Journal of Forestry Research520207. http://dx.doi.org/10.1155/2013/520207. McGranahan, D., & Wojan, T. (2007). Recasting the creative class to examine growth processes in rural and urban counties. Regional Studies, 41(2), 197–216. McSweeney, K., & McChesney, R. (2004). Outbacks: The popular construction of an emergent landscape. Landscape Research, 29(1), 31–56. Miller, M. D. (2012). The impacts of Atlanta's urban sprawl on forest cover and fragmentation. Applied Geography, 34, 171–179. Miller, K. A., Snyder, S. A., & Kilgore, M. A. (2012). An assessment of forest landowner interest in selling forest carbon credits in the Lake States, USA. Forest Policy and Economics, 25, 113–122. Minnesota Population Center. (2011). National historical geographic information system. Minneapolis, MN: University of Minnesota. http://www.nhgis.org, 2.0. Munroe, D. K. (2009). Pattern-based evaluation of peri-urban development in Delaware County, Ohio, USA: Roads, zoning and spatial externalities. In A. Páez, J. Gallo, R. N. Buliung, & S. Dall’erba (Eds.). Progress in spatial Analysis: Methods and applications (pp. 149–169). New York, NY: Springer. Munroe, D. K., van Berkel, D. B., Verburg, P. H., & Olson, J. L. (2013). Alternative trajectories of land abandonment: Causes, consequences and research challenges. Current Opinion in Environmental Sustainability, 5, 471–476. Munroe, D. K., & York, A. M. (2003). Jobs, houses, and trees: Changing regional structure, local land-use patterns, and forest cover in southern Indiana. Growth and Change, 34(3), 299–320. Nelson, P. B., & Sewall, A. (2003). Regional comparisons of metropolitan and nonmetropolitan migration in the 1970s and 1980s: Age and place implications. Professional Geographer, 55(1), 83–99. Nielsen-Pincus, M., Ribe, R. G., & Johnson, B. R. (2015). Spatially and socially segmenting private landowner motivations, properties, and management: A typology for the wildland urban interface. Landscape and Urban Planning, 137, 1–12. Nychka, D., Furrer, R., Paige, J., & Sain, S. (2016). Fields: Tools for spatial data. R Package. Version 8.3-6https://CRAN.R-project.org/package=fields. Olson, J. L. (2012). The evolution of urban-rural space(Doctoral Dissertation). Columbus, OH: The Ohio State University. Olson, J. L., & Munroe, D. K. (2012). Natural amenities and rural development in new urban-rural spaces. Regional Science Policy & Practice, 4(4), 355–371. Partridge, M. D., Rickman, D. S., Olfert, M. R., & Ali, K. (2012). Dwindling U.S. internal migration: Evidence of spatial equilibrium or structural shifts in local labor markets? Regional Science and Urban Economics, 42(1–2), 375–388. Petrzelka, P. (2012). Absentee landowners in the Great Lakes Basin: Who they are and implications for conservation outreach. Society and Natural Resources, 25(8), 821–832. Petrzelka, P., & Armstrong, A. (2015). Absentee landowners of agricultural land: Influences upon land management decision making and information usage. Journal of Soil and Water Conservation, 70(5), 303–312. Petrzelka, P., Malin, S., & Gentry, B. (2012). Absentee landowners and conservation programs: Mind the gap. Land Use Policy, 29, 220–223. Petrzelka, P., Ma, Z., & Malin, S. (2013). The elephant in the room: Absentee landowner issues in conservation and land management. Land Use Policy, 30(1), 157–166. Pierce, K. (2015). SOI migration data: A new approach. IRS Statistics of Income Bulletin (Summer)https://www.irs.gov/uac/SOI-Tax-Stats-Migration-Data. Plane, D. A., & Jurjevich, J. R. (2009). Ties that no longer bind? The patterns and repercussions of age-articulated migration. Professional Geographer, 61(1), 4–20. Poland, T. M., & McCullough, D. G. (2006). Emerald ash borer: Invasion of the urban forest and the threat to North America's ash resource. Journal of Forestry, 104(3), 118–124. Pyšek, P., & Richardson, D. M. (2010). Invasive species, environmental change and management, and health. Annual Review of Environment and Resources, 35, 25–55. R Core Team. (2017). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. https://www.R-project.org/. Radeloff, V. C., Steward, S. I., Hawbaker, T. J., Gimmi, U., Pidgeon, A. M., Flather, C. H., et al. (2010). Housing growth in and near United States protected areas limits their conservation value. Proceedings of the National Academy of Sciences of the United States of America, 107(2), 940–945. Redmon, L. A., Clary, G. M., Cleere, J. J., Evers, G. W., Haby, V. A., Long, C. R., et al. (2004). Pasture and livestock management workshop for novices: A new curriculum for a new clientele. Journal of Natural Resource and Life Science Education, 33, 7–10. von Reichert, C., Cromartie, J. B., & Arthun, R. O. (2014). Reasons for returning and not returning to rural U.S. communities. Professional Geographer, 66(1), 58–72. Rickenbach, M., & Kittredge, D. B. (2009). Time and distance: Comparing motivations among forest landowners in New England, USA. Small-scale Forestry, 8(1), 95–108. Rudel, T. K. (2009). How do people transform landscapes? A sociological perspective on suburban sprawl and tropical deforestation. American Journal of Sociology, 115(1), 129–154. Rudel, T. K., Bates, D., & Machinguiashi, R. (2002). A tropical forest transition? Agricultural change, out-migration, secondary forests in the Ecuadorian Amazon. Annals of the Association of American Geographers, 92(1), 87–102.

Fiore, A. M., Niehm, L. S., Hurst, J. L., Son, J., Sadachar, A., Russell, D. W., et al. (2015). Will they stay or will they go? Community features important in migration decisions of recent university graduates. Economic Development Quarterly, 29(1), 23–37. Florida, R. (2003). Cities and the creative class. City & Community, 2(1), 3–19. Florida, R. (2004). America's looming creativity crisis (Harvard Business Review. October). Flowerdew, R., & Lovett, A. (1988). Fitting constrained Poisson regression models to interurban migration flows. Geographical Analysis, 20(4), 297–307. Frey, W. H., & Liaw, K.-L. (1998). Immigrant concentration and domestic migrant dispersal: Is movement to nonmetropolitan areas “white flight”? Professional Geographer, 50(2), 215–232. Gandhi, K. J. K., & Herms, D. A. (2010). Direct and indirect effects of alien insect herbivores on ecological processes and interactions in forests of eastern North America. Biological Invasions, 12, 389–405. Golden, K. E., Peterson, M. N., DePerno, C. S., Bardon, R. E., & Moorman, C. E. (2013). Factors shaping private landowner engagement in wildlife management. Wildlife Society Bulletin, 37(1), 94–100. Goldring, S. A. (2014). Migration and inequality in the rural United States: Connecting urban to rural and local to global. Sociology Compass, 8(3), 324–355. Gosnell, H., & Abrams, J. (2011). Amenity migration: Diverse conceptualizations of drivers, socioeconomic dimensions, and emerging challenges. GeoJournal, 76(4), 303–322. Haugen, K., Karlsson, S., & Westin, K. (2016). New forest owners: Change and continuity in the characteristics of Swedish non-industrial private forest owners (NIPF owners) 1990-2010. Small-scale Forestry, 15(4), 533–550. Hecht, S., Yang, A. L., Basentt, B. S., Padoch, C., & Peluso, N. L. (2015). People in motion, forests in transition: Trends in migration, urbanization, and remittances and their effects on tropical forests. CIFOR Occassional Paper, No. 142Bogor, Indonesia: Center for International Forestry Research. van Herzele, A., & van Gossum, P. (2008). Typology building for owner-specific policies and communications to advance forest conversion in small pine plantations. Landscape and Urban Planning, 87(3), 201–209. Hewes, J. H., Butler, B. J., & Liknes, G. C. (2017). Forest ownership in the coterminous United States circa 2014: Distribution of seven ownership types - geospatial dataset. Fort Collins, CO: Forest Service Research Data Archive. https://doi.org/10.2737/RDS2017-0007. Hoekman, J., Frenken, K., & van Oort, F. (2009). The geography of collaborative knowledge production in Europe. Annals of Regional Science, 43, 721–738. Homer, C., Dewitz, J., Yang, L., Jin, S., Danielson, P., Xian, G., et al. (2015). Completion of the 2011 National Land Cover Database for the conterminous United States - representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, 81(5), 345–354. Hughes, G., Measells, M. K., Grado, S. C., Dunn, M. A., Idassi, J. O., & Zielinske, R. J. (2005). Underserved forest landowner workshops: Opportunities for landowners and extension. Journal of Extension, 43(4), 5. https://www.joe.org/joe/2005august/a5. php. Kaetzel, B. R., Majumdar, I., Teeter, L. D., & Butler, B. L. (2012). Regional differences among family forest landowners using National Woodland Owner Survey results. Southern Journal of Applied Forestry, 36(3), 141–145. Kaplan, J. O., Krumhardt, K. M., & Zimmermann, N. (2009). The prehistoric and preindustrial deforestation of Europe. Quaternary Science Reviews, 28(27–28), 3016–3034. Kelly, E. C., Gold, G. J., & Di Tommaso, J. (2017). The willingness of non-industrial private forest owners to enter California's carbon offset market. Environmental Management, 60(5), 882–895. Kendra, A., & Hull, R. B. (2005). Motivations and behaviors of new forest owners in Virginia. Forest Science, 51(2), 142–154. Kittredge, D. B. (2004). Extension/outreach implications for America's family forest owners. Journal of Forestry, 102(7), 15–18. Kittredge, D. B., Rickenbach, M. G., Knoot, T. G., Snellings, E., & Erazo, A. (2013). It's the network: How personal connections shape decisions about private forest use. Northern Journal of Applied Forestry, 30(2), 67–74. Kueper, A. M., Sagor, E. S., Blinn, C. R., & Becker, D. R. (2014). Extension forestry in the United States: Master volunteer and other peer-learning programs. Journal of Forestry, 112(1), 23–31. Kull, C. A., Ibrahim, C. K., & Meredith, T. C. (2007). Tropical forest transitions and globalization: Neo-liberalism, migration, tourism, and international conservation agendas. Society & Natural Resources, 20(8), 723–737. Law, J., & McSweeney, K. (2013). Looking under the canopy: Rural smallholders and forest recovery in Appalachian Ohio. Geoforum, 44, 182–192. Linghjem, H., & Mitani, Y. (2012). Forest owners' willingness to accept compensation for voluntary conservation: A contingent valuation approach. Journal of Forest Economics, 18(4), 290–302. Long, J. S. (1997). Regression models for categorical and limited dependent variables. Thousand Oaks, CA: Sage. Lovett, A., & Flowerdew, R. (1989). Analysis of count data using Poisson regression. Professional Geographer, 41(2), 190–198. Lӓhdesmӓki, M., & Matilainen, A. (2014). Born to be a forest owner? An empirical study of the aspects of psychological ownership in the context of inherited forests in Finland. Scandinavian Journal of Forest Research, 29(2), 101–110. MacDonald, K., & Rudel, T. K. (2005). Sprawl and forest cover: What is the relationship? Applied Geography, 25(1), 67–79. Majumdar, I., Laband, D., Teeter, L., & Butler, B. (2009). Motivations and land-use intentions of nonindustrial private forest landowners: Comparing inheritors to noninheritors. Forest Science, 55(5), 423–432. Majumdar, I., Teeter, L., & Butler, B. (2008). Characterizing family forest owners: A cluster analysis approach. Forest Science, 54(2), 176–184.

151

Applied Geography 96 (2018) 141–152

C. Gallemore et al.

United States Forest Service (2015). Who owns America's trees, woods, and forests? Results from the U. S. Forest service 2011-2013 national woodland owner survey. Newtown Square, PA: U. S. Forest Service Northern Research Station. US Census Bureau. (2000). 2000 census summary file 1. Washington, DC: US Census Bureau. US Census Bureau. (2010). 2010 census summary file 1. Washington, DC: US Census Bureau. US Census Bureau. (2012). North American industry classification system. Washington, DC: US Census Bureau. US Census Bureau. (2014). North American industry classification system. Washington, DC: US Census Bureau. US Census Bureau. (2015). TIGER/Line shapefiles: Technical documentation. Washington, DC: US Census Bureau. West, L. A., Cole, S., Goodkind, D., & He, W. (2014). 65+ in the United States: 2010: Special studies. Washington, DC: United States Census Bureau. West, P. C., Fly, J. M., Blahna, D. J., & Carpenter, E. M. (1988). The communication and diffusion of NIPF management strategies. Northern Journal of Applied Forestry, 5(4), 265–270. Wickham, H. (2009). ggplot2: Elegant graphics for data analysis. New York, NY: Springer. Widmann, R. H., Balser, D., Barnett, C., Butler, B. J., Griffith, D. M., Lister, T. W., Moser, W. K., Perry, C. H., Riemann, R., & Woodall, C. W. (2009). Ohio forests, 2006. Newtown Square, PA: US Forest Service. Williams, A. P., Allen, C. D., Millar, C. I., Swetnam, T. W., Michaelsen, J., Still, C. J., et al. (2010). Forest responses to increasing aridity and warmth in the southwestern United States. Proceedings of the National Academy of the Sciences, 107(50), 21289–21294. Xu, Z. (2014). Spatial and longitudinal patterns in county age-specific net migration in the United States, 1950-2010. Professional Geographer, 66(4), 641–652. Zhu, K., Woodall, C. W., & Clark, J. S. (2012). Failure to migrate: Lack of tree range expansion in response to climate change. Global Change Biology, 18, 1042–1052.

Ruseva, T. B., Evans, T. P., & Fischer, B. C. (2014). Variations in the social networks of forest owners: The effect of management activity, resource professionals, and ownership size. Small-scale Forestry, 13(3), 377–395. Sagor, E. S. (2012). Personal networks and private forestry: Exploring extension's role in landowner education(Doctoral dissertation). Minneapolis, MN: University of Minnesota. Salmon, O., Brunson, M., & Kuhns, M. (2006). Benefit-based audience segmentation: A tool for identifying nonindustrial private forest (NIPF) owner education needs. Journal of Forestry, 104(8), 419–425. Seto, K. C., Reenberg, A., Boone, C. G., Fragkias, M., Haase, D., Laganke, T., et al. (2012). Urban land teleconnections and sustainability. Proceedings of the National Academy of Sciences, 109(20), 7687–7692. Silver, E. J., Leahy, J. E., Weiskittle, A. R., Noblet, C. L., & Kittredge, D. B. (2015). An evidence-based review of timber harvesting behavior among private woodland owners. Journal of Forestry, 113(5), 490–499. Sunderlin, W., & Resosudarmo, I. A. (1999). The effect of population and migration on forest cover in Indonesia. Journal of Environment and Development, 8(2), 152–169. Tian, N., Poudyal, N. C., Hodges, D. G., Young, T. M., & Hoyt, K. P. (2015). Understanding the factors influencing nonindustrial private forest landowner interest in supplying ecosystem services in Cumberland Plateau, Tennessee. Forests, 6(11), 3985–4000. Ulrich-Schad, J. D. (2015). Recreational amenities, rural migration patterns, and the Great Recession. Population and Environment, 37(2), 157–180. United States Department of Agriculture (2014). Rural America at a glance: 2014 edition. Washington, DC: United States Department of Agriculture. United States Department of Agriculture Economic Research Service (2012). Commuting zones and labor market areas. Washington, DC: United States Department of Agriculture. United States Department of Agriculture Economic Research Service (2014). Creative class county codes. Washington, DC: United States Department of Agriculture.

152