Abstract
Fertilizer application has many known environmental impacts, and household use of fertilizer for lawn cultivation contributes a significant amount toward these impacts. Previous study provides a compelling narrative of the social, political, and cultural factors that drive fertilizer application; however, the psychological factors involved are not well known. This article examines factors that motivate fertilizer application using surveys among a sample of 194 residents within an urban watershed. Measures included aesthetic preferences, financial concerns, social norms, and environmental concerns. A principal components analysis (PCA) revealed four orthogonal factors. Among them, individual interests and social pressures positively predicted fertilizer application, as did the likelihood that children and pets play on a resident’s lawn. Environmental concerns were not predictive. These data contribute to an evolving storyline suggesting that households make tradeoffs between lawn aesthetics and concern for the environment and exposure to chemicals in part due to strong social pressures surrounding lawn maintenance.
Introduction
Numerous studies have attempted to document the ecological impacts of lawn cultivation in the United States and have found results that are considerably mixed. Lawn grass provides the critical functions of sequestering carbon and filtering water in urban and suburban areas that may otherwise be dominated by impervious surfaces (Bandaranayake, Qian, Parton, Ojima, & Follett, 2003; Brabec, Schulte, & Richards, 2002; Milesi et al., 2005; Townsend-Small & Czimczik, 2010; Townsend-Small, Pataki, Czimczik, & Tyler, 2011). On the other hand, maintaining a manicured lawn results in a number of ecological impacts, including the emission of greenhouse gases and air pollutants resulting from gas-powered lawn equipment (Karl, Fall, Jordan, & Lindinger, 2001; Priest, Williams, & Bridgman, 2000), as well a strain on water resources (e.g., Milesi et al., 2005; Yabiku, Casagrande, & Farley-Metzger, 2008). Much of the attention devoted to understanding the environmental impacts of lawn cultivation has focused on the application of fertilizer and other turfgrass chemicals. Nitrogen-containing fertilizer that is applied to lawns to maintain a green, lush, and weed-free aesthetic has a number of known environmental impacts, including the loss of soil nutrients, the acidification of soils and surface water, accelerated loss of biological diversity, contaminated drinking water, and the release of nitrous oxide—a powerful greenhouse gas (for review, see Vitousek et al., 1997).
Although the household share of fertilizer use may appear trivial relative to that used within the agricultural sector, lawn fertilizer accounts for a large proportion of the total nitrogen budget within many urban and suburban watersheds (Groffman, Law, Belt, Band, & Fisher, 2004; Law, Band, & Grove, 2004). Fissore and colleagues estimate that lawn fertilization is the second largest household source of nitrogen input and likely the most effective target for mitigating environmental impacts (Fissore et al., 2011). Without intervention, these impacts are expected to worsen. On a per-hectare basis, more chemical inputs are added to lawns than are used for the purpose of food production (Robbins & Sharp, 2003), a figure that is particularly problematic considering the pace at which lawn cultivation is expanding. Since the Second World War, the proportion of urban land in the United States has roughly quadrupled in size, growing at nearly twice the rate as the population (Lubowski, Vesterby, Bucholtz, Baez, & Roberts, 2006), and the proportion of lot size devoted to lawn space has also increased (Robbins & Birkenholtz, 2003). Both the intensity of chemical use as well as the growth in overall lawn space accounts for growing trends in the use of lawn chemicals. Although there are limited data specifically on household fertilizer use, trends associated with multiple chemicals inputs (including pesticides and herbicides) suggest that household use of chemicals has steadily risen in recent decades even though agricultural use of the same chemicals has declined (for review, see Robbins & Sharp, 2003).
There are a number of policy implications associated with the growing use of fertilizer and other turfgrass chemicals among households. For many reasons, traditional regulatory schemes that have been applied in agricultural or industrial sectors may be less effective or inappropriate for regulating households (Vandenbergh, 2004). A large number of diffuse actors applying fertilizer in small quantities make it difficult to quantify and mitigate the impacts of fertilizer application without a significant investment in resources. Furthermore, traditional regulatory schemes, such as taxes and “command and control” measures, may be unpopular and considered invasive by some private citizens, particularly in the United States. In addition, unlike agriculturalists and professionals within the lawn and landscaping industries, individuals within households rarely receive training on the appropriate methods and rates of applying, storing, handling, and disposing of turfgrass chemicals as well as the environmental impacts of their use. At least one study has shown that only 15% of residents surveyed could correctly report where water goes after running off of their lawns (Nielson & Smith, 2005). This may, in part, explain why lawn cultivation is so highly, chemically intensive on a per-hectare basis. In fact, for a majority of households, nitrogen inputs exceed plant and soil requirements resulting in nitrate leaching, suggesting that a similar landscape aesthetic could be achieved by households with less fertilizer (Fissore et al., 2011; Law et al., 2004). In a study of Oregon residents, Nielson and Smith (2005) found evidence of excess chemical application (both fertilizer and weed control chemicals) relative to what is recommended by water quality specialists.
Although this issue requires more investigation, behavioral interventions such as education, modified labeling schemes, and packaging requirements could be avenues to reducing excessive nitrogen inputs. As such, understanding the factors that motivate fertilizer application among households is an important dimension of managing negative health and environmental impacts. This includes individual aesthetic preferences and economic motivations, as well social influences on lawn-care decision making. The important and often-neglected role of social influences (e.g., social norms and reputational concerns) has recently been highlighted by a number of authors in this field (e.g., Larson, Cook, Strawhacker, & Hall, 2010; Larson & Harlan, 2006). For example, based on their own research, Nielson and Smith (2005) hypothesized that social pressures regarding lawn aesthetics led to excessive water and fertilizer applications and urged for a greater focus on the social dimensions of this behavior. In addition, Werner (2003) has shown that a group-based intervention designed to target attitudes and beliefs about the use of toxic home and landscaping chemicals was a promising means of changing patterns of use and disposal of these chemicals.
In this article, we use surveys to examine motivations for fertilizer application among households in an urban watershed. We consider motivations such aesthetic preferences, economic concerns, environmental concerns as well as social pressures. The underlying relationships among this set of variables, as well as their impact on fertilizer application, are examined.
Factors Influencing Household Fertilizer Use
Like many domains of human decision making, choices about lawn care are often analyzed as highly individual decisions motivated primarily by economic concerns and personal preferences for recreation and aesthetics (Templeton, Yoo, & Zilberman, 1999; Templeton, Zilberman, & Yoo, 1998). When asked, residents often report attractiveness and ease of maintenance as the dominant reason for their landscaping decisions (Larson, Casagrande, Harlan, & Yabiku, 2009; Martin, Peterson, & Stabler, 2002). Although these factors are undoubtedly related to lawn care practices, an analysis that focuses exclusively on individual preferences and expected economic utilities may fail to uncover the highly social environment in which decisions about lawn care are made. For example, opinions about what an ideal yard looks like vary according to socioeconomic status suggesting that personal preferences may be, to some degree, socially constructed. Larsen and Harlan (2006) have demonstrated that residents in the Phoenix area prefer different types of landscapes (e.g., desert, lawn, courtyard) depending on whether they were asked about the front or the backyard. Similarly, when asked about their aesthetic preferences independent of economic considerations, preferences for the ideal front yard were correlated with the individual’s income; however, income was not correlated with preferences for the ideal backyard. Larsen and Harlan’s work suggested that the front and backyards fulfill different utilities, and the choices that construct and maintain these spaces, choices that have significant ecological impacts, may be influenced and constrained by a different set of factors (Larsen & Harlan, 2006). In particular, the highly visible front yard may be constrained not simply by the aesthetic, recreational, or functional utility of the space but also by a desire to fulfill neighborhood norms and to communicate a group membership or social status. In other words, decisions about lawn care extend well beyond the desire to fulfill one’s personal aesthetic preferences to serve as a symbolic expression of the self and to actively maintain or construct identity and status within a community (Larsen & Harlan, 2006).
This interpretation of lawn-care decision making is consistent with social-psychological research on the influence of social norms on behavior (e.g., Keizer, Lindenberg, & Steg, 2008; Schultz, Nolan, Cialdini, Goldstein, & Griskevicius, 2007). In their Focus Theory of Normative Conduct, Cialdini and colleagues (Cialdini, Kallgren, & Reno, 1991; Cialdini, Reno, & Kallgren, 1990) distinguish between two types of social norms: descriptive and injunctive. A descriptive norm describes the typical behavior in a given situation, that is, the extent to which it is typical to apply fertilizer to one’s lawn or to put out recycling in a given community. An injunctive norm describes what is generally agreed to be the ideal behavior in a given situation. In many cases, descriptive and injunctive norms reinforce one another. For example, in many communities, the use of curbside recycling is considered to be the “right” thing to do and may also be very common. However, in other cases, for example exercise habits or smoking, actual patterns of behavior deviate from what is considered socially desirable. Individuals have been shown to conform both to what they perceive to be desirable as well as typical behavior to maintain a positive self-image and to garner social approval or to avoid negative attention from those around them (for review, see Cialdini & Goldstein, 2004). This is particularly true when a norm is salient to the individual (Kallgren, Reno, & Cialdini, 2000), such as when a person feels a strong sense of identity or attachment to the community in which specific social norms operate (e.g., Schofield, Pattison, Hill, & Borland, 2001).
The influence of social norms on behavior has been heavily studied in other areas of environmental decision making and has been discussed by many scholars in the area of lawn-care decision making (e.g., Larson et al., 2010; Robbins, Polderman, & Birkenholtz, 2001). However, to our knowledge, it has not been empirically examined within this context. There is reason to believe that lawn care activities, including fertilizer application, would be particularly susceptible to the influence of social norms. Lawn care is an extremely visible behavior that is often the target of both scrutiny and praise by one’s neighbors. Results from a series of semistructured interviews about lawn care practices in an Oregon community indicated that in addition to concerns about “neatness” and “greenness,” residents frequently cited responsibility to neighbors and fear of neighbor disapproval as reasons for their lawn care practices (Nielson & Smith, 2005). As such, in addition to influence of what is perceived to be normative and desirable, it is likely that the influence of social norms on behavior is also enhanced by the degree to which an individual perceives that one’s reputation is connected to his or her lawn’s appearance or that there may be other social sanctions associated with a failure to appropriately keep up one’s lawn. In some less common cases, lawn care standards may be formally enforced through neighborhood covenants, such as through a homeowners association. Or, more frequently, residents may perceive unwritten rules for how a yard should look, enforced through informal mechanisms such as neighbor-to-neighbor contact. Reputational concerns and the perceptions that there are rules governing lawn care in one’s neighborhood are closely related to neighborhood norms regarding lawn care; however, these factors are conceptually distinct. Neighborhood rules, whether formal or informal, may be a means of enforcing social norms for lawn care; however, social norms have been shown to influence behavior in the absence of the opportunity for any formal or informal sanctions (e.g., when individuals are behaving in private; Goldstein, Cialdini, & Griskevicius, 2008; Kallgren et al., 2000).
The important social obligations fulfilled through lawn care activities may explain why so many households appear to make important tradeoffs between maintaining a desired lawn aesthetic and engaging in behaviors that many households perceive to carry environmental risks (e.g., Larson et al., 2009; Larson et al., 2010; Robbins, 2007; Robbins et al., 2001). Despite the prevalence of which it is used, surveys have shown that individuals perceive at least moderate risks associated with chemical fertilizers (Siegrist & Cvetkovich, 2000). Likewise, work by Robbins and colleagues (2001) suggest that those who use turfgrass chemicals are not unaware of the environmental impacts of their actions; rather, most acknowledge the negative environmental effects of fertilizer (Meyer, Behe, & Heilig, 2001). In fact, there is some indication that the use of turfgrass chemicals may be even higher among those with strong environmental leanings. In a study of lawn care practices among households in the San Francisco area, Templeton and colleagues (1999) found that households in which one or more members belonged to an environmental organization were 1.7 times more likely to apply chemicals to their lawns. Larson and colleagues have also captured decision tradeoffs of a somewhat different variety among residents in the Phoenix area, where they find that individuals often connect green (exotic) lawns with environmental benefits such as heat reduction and improved air quality, despite their water intensity (Larson et al., 2009). In this case, it appears that individuals may be making tradeoffs between multiple environmental outcomes.
Understanding the social environment and the influence it has on lawn care practices is critical to designing appropriate and effective policy interventions. For example, to the extent that individual economic concerns and personal preferences are constrained by the social obligations (whether real or perceived) of maintaining a certain lawn aesthetic, policies that focus on incentivizing or educating the individual without addressing these wider social concerns may fall flat. Educational campaigns that have targeted behavioral changes through the modification of social expectations or misperceptions of what is normative within a community have shown promise (for review, see DeJong et al., 2006; Neighbors, Larimer, & Lewis, 2004; Perkins, 2002). For example, an intervention that targeted the exaggerated perception of the number of people who evade taxes resulted in a reduction in the number of deductions claimed on a subsequent tax filing (Wenzel, 2005). A similar approach that considers the social motivations for applying fertilizer, in addition to economic considerations and personal preferences, may be necessary.
Following calls for a more thorough analysis of individual and household-level emissions and elemental inputs (Chakravarty et al., 2009; Dietz, Gardner, Gilligan, Stern, & Vandenbergh, 2009; Fissore et al., 2011), in this article we examine patterns of household fertilizer use within an urban watershed in Nashville, Tennessee. Using household surveys, we examine set factors that may influence fertilizer application, including variables relating to individual aesthetic preferences, economic considerations, environmental concerns, and social influences on lawn care practices. In particular, we respond to calls for additional research into the social pressures on lawn care maintenance by including a number of variables related to social motivations for fertilizer use, including measures of social norms and perceptions of neighborhood lawn care rules. Although previous literature has revealed a number of key variables, relatively little is known about the relationships among this set of variables as well as their relative influence on fertilizer use. We attempt to add to this literature by examining the underlying dimensions associated with these variables using principal components analysis (PCA). As a first step toward this endeavor, we conduct an exploratory PCA in which the number of factors and factor loadings are not constrained by a priori hypotheses. This analysis is followed by a multivariate regression to assess the predictive value of the derived factors and to gain clarity on the relative influence of these factors on fertilizer use. Understanding the core dimensions of these variables could contribute to a more cohesive and theoretically driven model of lawn care practices.
Method
Research Overview and Study Area
The data presented in the following sections were collected as part of the Nashville Yard Project, an ongoing, interdisciplinary study designed to examine the psychological, economic, social, and legal influences on lawn care practices, as well as the environmental impacts of those practices within an urban watershed. The data reported below were collected using in-person, semistructured interviews administered to 194 residents living within the Richland Creek Watershed. The watershed is 28.5 square miles of urban and suburban neighborhoods in the greater Nashville metropolitan area, of which the majority are residential. These neighborhoods represent a diverse range of residents in terms of race, income, education, and ownership patterns, making it an ideal setting to examine how these factors relate to fertilizer use. The use of a watershed to define the boundaries of this study area also facilitates a hydrological analysis of the environmental impacts of household fertilizer application, which is currently underway.
Sampling Procedure
A two-stage sampling procedure was used to identify prospective participants. First, we used a sampling frame of more than 100 city-defined neighborhoods to stratify the population according to geographic zones within the watershed. Neighborhoods that included fewer than 10 parcels (for example, neighborhoods with one or more large condominium complexes) were excluded resulting in 56 usable neighborhoods. The neighborhoods in our final sample are made up almost entirely of single-family homes with the occasional duplex or other small multifamily unit, with an average of 1.03 dwellings per parcel.
Next, one face block was randomly selected within each neighborhood. To do this, we used a random number generator to select one parcel within a neighborhood. Using this household as the anchor point, a face block was identified by selecting 9 contiguous parcels to the left or right of the anchor parcel as well as 10 contiguous parcels directly across the street. When drawing the boundaries around these face blocks, intersecting streets were avoided where possible (for example, if the randomly selected parcel was at the end of a block of parcels, the face block was identified as the 9 contiguous parcels on the same street and the 10 contiguous parcels directly across the street). However, in cases where this could not be avoided, face blocks did cross intersecting streets. Multiunit structures were included in the sample if the responding household indicated that they were responsible for some portion of a yard space (e.g., upper-level units, apartments without yards, or apartments for which the landlord maintained the yards were removed from the sampling frame).
Participant Recruitment
At the time that these data were analyzed, participant recruitment had taken place in 43 of the 56 face blocks selected for inclusion in this study. All households on a face block were invited to participate in this study. To recruit participants, we used a combination of mailed letters, phone calls, and door-to-door canvassing. First, prospective participants were sent a letter describing the nature of the study with an invitation to participate and instructions for how to schedule an interview (via phone or email). A series of follow-up attempts were made for those who did not schedule an interview after this initial mailing. One follow-up attempt was made by phone, one follow-up attempt was made using a mailed postcard, and one follow-up attempt was made by knocking on the participant’s door during the evening and weekend hours.
A total of 850 households were eligible for inclusion in the study and were invited to participate. If a prospective participant was not reached with any of these follow-up attempts, the household was classified as a “nonresponder.” This characterized 68.8% of households contacted. An additional 8.4% indicated by phone, email, or in conversation with the research assistant that they did not want to participate. The remaining 22.8% of prospective participants agreed to participate in the survey.
The average age of respondents was 53 years. A large majority (89%) owned the home that they lived in at the time the survey was conducted, and the average length of time in residence was 12.4 years. The mean assessed property value was US$443,452. This suggests this was an affluent sample; however, this mean was slightly lower, though not statistically different, than the population mean of US$451,000, indicating that the sample was representative on the basis of income.
Survey Procedure
At the time the interview was arranged, we requested to speak to the individual who makes the majority of decisions about lawn care within the household. In cases where multiple individuals fit this description, the first available individual was interviewed. Surveys were conducted as in-person, semistructured interviews administered at the participant’s home by a trained research assistant. The interview protocol contained both closed- and open-ended items. Interviews lasted roughly 1 hr and the sessions were audio recorded with the participant’s consent. Participants received US$20 as compensation for their time.
Survey Instrument
A series of items were adapted or developed for the purpose of this survey based on the previous literature as well as previous work examining decisions surrounding lawn care practices in the Baltimore area. The items used in these analyses, as they appeared in the survey, as well as scale means (M), standard deviations (SD), and internal reliability coefficients (where applicable) are provided in Table 1.
Summary of Variables and Items as They Appeared in the Survey.
For the dependent variable, we derived a dichotomous outcome variable for whether the household had applied fertilizer on its lawn within the past year. Participants were asked whether they had employed a person or company to apply fertilizer to their lawn during the past 12 months. Participants were also asked whether they or someone within their household had personally applied fertilizer to their lawn over the past 12 months. Participants who responded yes to either of these questions were coded as a “1” to indicate that they had applied fertilizer. Those who responded no to both received a “0” for this variable. Participants were also asked to provide more information about the products used, their provider plan, and so on, and to show receipts or packages where possible. This was to distinguish fertilizer application from the application of other similar chemicals, such as pesticides and herbicides.
Demographic information (i.e., age and education) was collected directly from the participant with the exception of property values, which was retrieved from Davidson County’s publicly available property appraisal and assessment database.
We included questions about the financial and aesthetic motivations an individual may have for using fertilizer. This included the belief that one’s lawn can positively or negatively affect property values (effect of lawn on home value) and the individual’s personal desire to have a lush, green lawn (green lawn importance).
Three questions were developed to examine the effect of environmental concern on lawn fertilization (environmental concern). Intuitively, one would expect those who express a greater concern for the natural environment to be less likely to apply fertilizer to their lawn. Yet, previous study has suggested that environmental concerns may be overlooked in lawn care decisions due to countervailing influences (i.e., concern over property values, etc.), and, in some cases, environmentalism has been found to be correlated with greater lawn chemical use (Templeton et al., 1999). Previously, Robbins (2007) hypothesized from his qualitative work that, in addition to making environmental tradeoffs, households often make tradeoffs involving their own personal safety or the safety of vulnerable others such as pets. To explore this issue, we also include items to measure concerns about personal exposure to chemicals (chemical concern) as well as whether children or pets use the lawn recreationally (children on lawn, pets on lawn).
A series of items were included to measure the social motivations for applying fertilizer. The perceived effect of one’s lawn on his or her personal reputation was measured with one item developed for this survey (lawn reputation). A four-item scale was developed to assess the perception of written or unwritten rules or expectations within one’s neighborhood about how a lawn should be kept (neighborhood lawn rules). Two measures of social norms were also included. Following the definition of Cialdini et al. (1990, 1991), we include measures of both descriptive and injunctive norms. To our knowledge, measures of social norms have not been developed for this context. However, similar items have been used to measure perceived descriptive and injunctive norms in settings involving environmentally significant behavior (e.g., Carrico & Riemer, 2011; Ohtomo & Hirose, 2007).
Finally, we include a measure of neighborhood attachment using the previously validated Brief Sense of Community Scale (Peterson, Speer, & McMillan, 2008). This scale is designed to capture multiple dimensions of neighborhood attachment, including a feeling of belongingness and emotional connection with one’s neighborhood. Due to space and time constraints, only four of the eight items (one representing each dimension) were used in this survey.
Results and Discussion
Overview of Analyses
A series of analyses are presented in the following sections. First, descriptive statistics for fertilizer use are presented followed by a summary of the bivariate relationships among the study variables. We then summarize results from a PCA of the 11 hypothesized predictors of fertilizer application to examine the underlying factor structure associated with these variables. Finally, using factor scores derived from the PCA, we examine the extent to which these factors predict fertilizer application using multivariate binary logistic regression.
Rate of Fertilizer Application
Roughly 48% of the sample (n = 94) reported that they apply fertilizer to their lawn. Among these individuals, the majority (53.2%, n = 50) reported that they self-apply fertilizer, 31.9% (n = 30) reported that a lawn care provider applies fertilizer, and a small minority (14.9%, n = 14) reported both self-applying and using a lawn care provider for fertilization. The average number of applications per year among all fertilizer users was 2.63 (SD = 2.09). These numbers are consistent with earlier studies that have estimated that between 50% and 70% of households apply lawn fertilizer (Fissore et al., 2011; Law et al., 2004; Robbins et al., 2001). The average number of annual applications reported in this sample was somewhat higher than that found in other studies, which have typically reported an application frequency of one to two times per year (Fissore et al., 2011; Law et al., 2004). This could be due to the fact that previous studies included only those who self-apply fertilizer, whereas we include both self-applying and provider-applying households. It could also be because of regional differences in patterns of lawn fertilization (these studies took place in Minnesota and Maryland, respectively), owing to differences in the length of the growing season and other environmental or social factors.
Due to gaps in training and knowledge about lawn care, one might expect differences in the frequency of application between households and trained professionals. We have not measured the intensity of fertilizer applied at each application; however, no differences were found in the frequency of applications within this sample. There was no difference in the number of applications between those who reported self-applying and those who used a lawn care provider, t(74) = 0.93, p = .36. Not surprisingly, those who both self-applied and used a lawn care provider fertilized more frequently (M = 4.21, SD = 2.19) than the remainder of fertilizer users (M = 2.34, SD = 1.95), t(88) = 3.24, p < .01.
Bivariate Correlations
Next, bivariate relations among the 15 study variables were examined using Pearson product–moment correlations (Table 2). Because the vast majority of participants owned their home (89%), the variable for home ownership was excluded from the analysis. Consistent with earlier studies (Robbins et al., 2001), those who applied fertilizer in this sample tended to be older, r(190) = 0.24, p < .01, more educated, r(193) = 0.18, p < .05, and wealthier (as indicated by property values), r(192) = 0.28, p < .01. Fertilizer use was not related to environmental concern, which is consistent with earlier work indicating that environmentalism is unrelated to fertilizer use if not positively predictive (Robbins et al., 2001; Templeton et al., 1999). Fertilizer use was also unrelated to concern about chemical exposure or whether children or pets play on the participant’s lawn.
Bivariate Correlations Among Fertilizer Usage and Study Variables.
Variable is a composite of two or more items.
p < .05. **p < .01.
As expected, fertilizer use was significantly related to the perception that a lawn’s appearance will affect home values, r(193) = 0.23, p < .01, the personal importance of having a lush green lawn, r(193) = 0.25, p < .01, and the belief that having an attractive lawn reflects positively on the resident, r(193) = 0.20, p < .01. Feeling a strong sense of attachment with a neighborhood was also related, r(193) = 0.16, p < .05, as were descriptive and injunctive norms, r(193) = 0.19, p < .05. The perception that there are neighborhood rules about how a yard should be maintained was not related to self-reported fertilizer use, r(193) = 0.14, p = .06.
PCA
As might be expected, small to moderate intercorrelations among the set of predictor variables reported in Table 2 suggest that many of the constructs measured here are not independent of one another. In an effort to explore the underlying dimensions associated with these variables, these data were subjected to a PCA. The initial PCA revealed four components with eigenvalues (eig) > 1.00, accounting for 62% of the total variance (% var), and a visual inspection of the scree plot corroborated this conclusion. Next, an orthogonal varimax rotation with Kaiser normalization was applied. This analysis was also repeated using an alternative orthogonal rotation (quartimax) as well as an oblique rotation (oblimin) with near identical results to the varimax rotation. Only the results of the varimax rotation are discussed here. Table 3 summarizes the factor loadings for the rotated solution. To be conservative, a factor loading cutoff of 0.30 was applied when interpreting the rotated factor solution. However, as can be seen, the pattern of loadings among the 11 variables was relatively unambiguous. In only one case did factor loadings for one variable exceed the cutoff on more than one factor. Neighborhood lawn rules loaded moderately (0.65) onto Factor 3 and slightly onto Factor 4 (−0.41). Given that its factor loading was substantially higher for Factor 3 and that it was conceptually similar to the other three variables that loaded onto this factor, it was categorized as such.
Principal Components Analysis Factor Loadings Using a Varimax Rotation With Kaiser Normalization.
Note: Factor loadings of less than 0.30 are omitted. Boldface is used to indicate the factor on which each item loads most highly.
Variable is a composite of two or more items.
Factor 1 (eig = 1.96, % var = 17.80) comprised of three variables, all of which loaded strongly (>0.72). This factor, which we label as “individualistic interests,” appeared to capture immediate personal interests associated with lawn care decisions. Loading most highly was lawn reputation or the perception that an attractive lawn reflects positively on the resident. The two other factor loadings were similarly high and included the participant’s rated importance of having a lush, green lawn and the belief that one’s lawn affects property values.
Two variables loaded strongly (> 0.92) on Factor 2 (eig = 1.79, % var = 16.25): environmental concern and chemical concern. Although similar, the strong factor loadings among both variables is interesting considering that environmental concern is typically characterized as a somewhat prosocial attitude whereas chemical concern directly references personal chemical exposure. We label this factor as “environmental concerns,” which we hypothesize includes concerns over the integrity of the environment itself as well as the effect of the environment on oneself.
Factor 3, which we label as “social pressures” (eig = 1.74, % var = 15.85), included four variables that loaded moderately to strongly (0.51-0.78). This factor was comprised of variables measuring (in order of magnitude) descriptive norms, injunctive norms, neighborhood lawn rules, and neighborhood attachment. As such, this factor appeared to reflect the neighborhood social pressures associated with fertilizer use, with those scoring higher on Factor 3 living within a neighborhood in which they perceive a high percentage of fertilizer application, strong approval for fertilizer use, written or unwritten social sanctions associated with an unkempt lawn, and a high degree of cohesion within a neighborhood. Considering this cluster of variables, it is somewhat surprising that the lawn reputation variable loaded onto Factor 1 rather than Factor 3, where its factor loading was negligible (<0.10). It is possible that concerns about one’s reputation represents a more individual-level concern, similar to aesthetic concerns and concerns about property values, whereas Factor 3 characterizes the social environment in which lawn care decisions are made independent of one’s reputational concerns. It is also somewhat surprising that neighborhood attachment loaded onto Factor 3, considering it to be distinct from the presence of formal or informal rules or social norms. Although there are distinctions among these variables, it is possible that the presence of strong neighborhood ties facilitates the construction of neighborhood rules and norms and is correlated for this reason. If this is the case, it is not surprising that the loading of neighborhood attachment was somewhat lower than the other variables loading onto this factor. Additional research is needed to understand the relationships among these variables.
Finally, Factor 4 (eig = 1.32, % var = 11.97) comprised of two variables (> 0.68) that capture the degree to which children and pets play on one’s lawn. We refer to Factor 4 as “lawn exposure.” Considering the uniqueness of these items relative to the other variables measured, it is not surprising that they comprised a separate factor. As briefly mentioned earlier, neighborhood lawn rules (categorized as Factor 3) also loaded moderately and negatively (−0.41) onto Factor 4. Those who perceived high levels of neighborhood lawn rules were less likely to report that children and pets play on their lawn. This may be due to the fact that older individuals tend to have fewer pets and children playing on their lawn and also tend to live in neighborhoods that are characterized by a higher degree of lawn rules. However, more research is needed to determine whether this pattern is replicated or if it is simply an artifact of a relatively small sample in this particular location.
Predicting Fertilizer Application
To examine the relationships among these factors and fertilizer application, factor scores were created by summing the variables that loaded on each of the four factors described above. We first examine the bivariate correlations among these factors, as well as with the demographic and outcome variables (Table 4). The correlations among the four factors are expected to be close to zero because they are based on an orthogonal rotation; however, creating factor scores (particularly using a summing method) can reintroduce intercorrelation. In this case, one small but significant correlation was observed between individualistic interests and social pressures. This relation is intuitive as individuals may derive their attitudes and preferences regarding lawn care from their social surroundings to some degree. Similarly, those who have a strong personal preference for maintaining a certain lawn aesthetic may project their own beliefs and preferences on those around them.
Bivariate Correlations Among Study Variables.
Note: ns range from 186 to 193 depending on missing data.
p < .05. **p < .01.
The correlation matrix also indicated that those who scored higher on social pressures tended to be older, more educated, and have higher property values. This may suggest that wealthier neighborhoods are those that exert the greatest social pressures for maintaining a certain lawn aesthetic. There was also a small correlation between lawn exposure and age, not surprisingly, those who were younger tended to report more children and pets played on their lawns.
Two factors showed a bivariate relation with fertilizer use. Those who scored higher on individualistic interests and social pressures reported greater levels of fertilizer use. Environmental concern and lawn exposure were uncorrelated with fertilizer use.
Next, a multivariate binary logistic regression was performed on the dichotomous outcome variable to assess the unique relation between each of the four factors controlling for each other as well as the demographic variables. A two-stage model was run in which age, education, and property values were entered into Model 1. Scores on Factors 1 to 4 were entered as predictors in Model 2. The results, including parameter estimates and model statistics, are summarized in Table 5.
Summary of a Multivariate Binary Logistic Regression Analysis Predicting Fertilizer Use.
Note: OR = odds ratio.
p < .05. **p < .01.
Because participants were recruited at the block level, there is a potential for violating the assumption of independence of errors. To assess this, these analyses were also replicated using a multilevel binary logistic regression. An initial random-intercept model revealed small to moderate intraclass correlation (ρ = .13). However, when the covariates for age, education, and property values were included in the model, the between-group variance dropped to near zero (ρ < .01) indicating that after controlling for demographics, nearly all the variance in fertilizer application could be explained at Level 1. The full model, including both demographic variables and the four factor score variables, replicated the results found with a traditional binary logistic regression. For ease of interpretation, only the traditional logistic regression is reported here.
Model 1 was significant at the p < .01 level, and the Hosmer and Lemeshow statistic indicated that the observed data did not differ significantly from the predicted model, χ2(8) = 5.47, p = .71. Both age and property value contributed unique variance to the decision to apply fertilizer, and in both cases, the effect was positive. Despite a significant bivariate correlation, education had no independent effect on fertilizer application.
The overall model statistics for Model 2 were also significant. The inclusion of the factor scores in Model 2 accounted for a significant amount of variance above and beyond Model 1, χ2(4) = 30.10, p < .01. Again, the Hosmer and Lemeshow test suggested that the observed data did not differ significantly from the predicted model, χ2(8) = 7.99, p = .43.
Three variables significantly predicted fertilizer application in the positive direction. First, there was a significant effect of individual interests. Not surprisingly, those who desired a lush, green lawn and who believed there to be a connection between the quality of their lawn and their reputation and property values were more likely to apply fertilizer to their lawn. This complements previous work in which residents cite aesthetic preferences as a primary reason for their lawn care practices (Larson et al., 2009; Martin et al., 2002).
There was also a significant effect of social pressures, suggesting that decisions to apply fertilizer are also, in part, determined by social pressures felt from one’s neighbors. The fact that social pressures remained significant even after controlling for age, education, and property values indicates that the influence of this factor is independent of the socioeconomic status of the neighborhood and is likely to exist in neighborhoods across the income spectrum, although it may be more prevalent in more affluent communities.
There was no effect of environmental concerns on fertilizer use; those who were concerned about the environment and concerned about their own exposure to chemicals were no less likely to use lawn fertilizer than those who indicated that they do not hold these attitudes. It was also revealed that those who have children or pets that play on their lawns are over two times more likely to fertilize their lawns than their peers. This is true even controlling for the demographic correlates among those who have children or pets. These findings replicate earlier work indicating the tradeoffs that are often made between environmental concerns and lawn care activities (Robbins et al., 2001; Templeton et al., 1999). Individuals seem to use fertilizers despite proenvironmental values and despite the understanding that it may have a negative effect on the environment. With this study, we can add to this conclusion that individuals tend to fertilize despite the fact that children and pets use the lawn recreationally. Presumably, these tradeoffs are made to fulfill other desires such as aesthetic preferences and perceived social obligations, as evidenced here. Although consistent with other findings, this conclusion must be taken with caution, and more work is needed to explore the tradeoffs that are implied in these results. This is especially true with respect to the finding-associated recreational use of the lawn by children and pets. A previous study has shown that individuals perceive a moderate risk associated with chemical fertilizers, comparable to the perceived risks associated with pesticides, DDT, automobile travel, and fire fighting (Siegrist & Cvetkovich, 2000). However, we have not directly measured perceived risk of fertilizer within this sample, and it is possible that this risk is considered negligible and, therefore, no tradeoffs are being made. Similarly, these analyses do not consider precautions that may be taken to avoid exposure of children and animals, or the selection of certain types of fertilizer products that are recommended as safe. As such, additional research is needed to understand the relationship between risk perceptions and fertilizer application and whether it is typical for concerns about safety to be suppressed due to concerns about lawn aesthetics.
Conclusions
These data produced a number of important insights regarding the use of fertilizer among households, as well as motivations associated with fertilizer use. Results from the multivariate analysis suggest that, controlling for age, education, and property values, fertilizer use is associated with individual concerns over one’s property value, reputation, lawn appearance, social pressures as well as recreational use of the lawn by children and pets. However, concerns about the environment and exposure to chemicals were unrelated to fertilizer use. Most importantly, these data speak to the social pressures that shape lawn care decisions, pressures that appear to be independent of personal, aesthetic, and economic considerations but that are as powerful as those concerns. Like other work before it, these results highlight the complex assemblage of social, economic, political, and ecological factors that converge to influence lawn care practices. Likewise, this research fits into a growing body of work that challenges the conception that the quintessential American lawn is simply the product of individuals acting on their own personal preferences and financial considerations. Rather, these data suggest that lawn care practices fulfill a utility that extends beyond the immediate experience of those who live on a property. Instead, maintaining a lawn is an avenue for engaging with one’s neighbors, for fulfilling expectations of what it means to be a positive member of a community, and to communicate a willingness to cooperate in creating and maintaining a shared space. In some cases, these pressures may be overt, such as in the presence of formal neighborhood covenants and lawn care regulations. However, in other cases, they may be more subtle and manifest in the perception, even if biased, of how a “typical” yard is maintained and what would garner approval from most neighbors.
These data also contribute to an evolving storyline in which individuals are negotiating conflicting desires to fulfill their own values and concerns and to achieve a desired lawn aesthetic. Tradeoffs between values and aesthetics have been explored previously in this literature (e.g., Larson et al., 2009; Larson et al., 2010; Robbins, 2007; Robbins et al., 2001). For example, tradeoffs between environmentalism and chemical application have been shown in multiple studies (Robbins et al., 2001; Templeton et al., 1999). This study provides further evidence that motivations for maintaining a green lawn, whether personal, social, or a combination, can overwhelm health and environmental concerns. In this case, environmental concerns were unrelated to fertilizer use; yet, individual interests and social pressures were significantly predictive. Although concerns about exposure to fertilizer in particular were not measured in this study, previous research has revealed that individuals perceive a moderate risk associated with chemical fertilizers (Siegrist & Cvetkovich, 2000). These data suggest that, despite this, concerns about exposure to chemicals were unrelated to decisions to apply fertilizer. Similarly, recreational lawn use by children and pets was associated with an increased prevalence of fertilizer use, rather than a decrease. This latter finding may indicate that tradeoffs are made not only for aesthetic purposes but also for recreational purposes. Larson et al. (2009) have found similar results in the Phoenix area in which households sacrifice concerns about environmental impacts for the heat mitigating benefits of green lawns.
Although this analysis did not address it explicitly, one also cannot ignore the influence of the lawn care and lawn chemical industries in establishing and communicating the qualities of an ideal lawn, qualities that are difficult, if not impossible, to obtain without the intensive use of chemicals and equipment. In addition to the very tangible and reasonable concern over one’s property values, the lawn industry has aggressively campaigned to connect a specific lawn aesthetic with valued ideals such as family, community, and even masculinity (Robbins et al., 2001; Robbins & Sharp, 2003). The source of neighborhood norms and expectations is undoubtedly related to industry-promoted standards, and many of these standards are being incorporated into requirements arising from neighborhood laws and covenants, which residents must abide by if they wish to live in some communities. As such, the ability to choose how to maintain one’s lawn and to influence the ecological impacts of one’s lawn becomes one among a long set of attributes, which the individual must consider jointly prior to purchasing a home. The formalization of lawn care norms and expectations through homeowner’s association covenants and regulations is a trend that could have significant ecological consequences, and is an area that is ripe for further exploration jointly from social scientific, environmental, and legal perspectives.
There are a number of limitations to this study that should also be addressed. First, the relatively low response rate that was achieved (23%) raises concerns that these data may be biased due to nonresponse. It is possible that the individuals who agreed to participate in this survey were more interested in lawn care activities or placed a higher value on their lawn’s appearance. These data may also be biased due to self-report. Participants may have incorrectly reported the history and frequency of their own fertilizer use or confused it with the application of other chemicals. This may especially be true among those who reported hiring a lawn care provider. Future studies should attempt to gather more objective measures of fertilizer application. Finally, these results are based on a relatively small sample size that is culturally and geospatially situated. For example, while people are becoming more aware of the state of the Cumberland River in Nashville, the remedial focus has been on point source pollution such as storm sewer overflows; emphasis has not been placed on how residents’ individual choices lead to nonpoint source pollution. Although Nashville is likely similar to many cities in the country in this respect, it is quite different from places like Baltimore, where a sister project found high levels of environmental awareness about the consequences of household decisions; environmental concern may affect decision making in different ways in different regional contexts. Furthermore, there may be a high degree of variance in the neighborhood norms and expectations for how a lawn should be maintained in contexts beyond Nashville. Therefore, studies such as this should be replicated in other contexts to improve generalizability.
Despite the limitations of this study, these results provide further evidence of the multitude of psychological and social factors that influence the way we manage our environment. Indeed, human decision making surrounding the production and maintenance of the lawn can be conceptualized as a point (moment) that provides an opening to enter into and examine the effects of being intertwined in a constellation of relationships that influence our behavior. In this scenario, like many others, decisions that appear fully reasonable with respect to the goals of fulfilling one’s neighborly duties and maintaining an attractive and functional space can have profound ecological and public health consequences in the aggregate. These data suggest that those neighborly duties, coupled with aesthetic and economic desires, are enough to override concerns about the environment and exposure to chemicals. This fits within a broader literature on human–environment interaction that suggests that proenvironmental values and intentions often fall short of producing proenvironmental actions, and that the social and situational constraints of an actor must be considered. In this regard, our understanding of the factors that influence decisions about lawn care and its environmental impacts can inform our understanding of decisions made in other domains of environmental significance. Furthermore, understanding the motivations and goals that drive decisions may provide some insight into avenues for managing their collective impacts.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project is based upon work supported by the National Science Foundation under Grant No. EAR-0943661. We thank our co-investigators, George Hornberger, Michael Vandenbergh and Kimberly Bess for their helpful comments on the development of this manuscript.