0743-0167#0 $3.00 f 0.00 Pergamon Press plc
Journal ofRural Studies, Vol. 6, No. 2, pp. 175-186, IWO Printed in Great Britain
A Longitudinal Analysis of Neighboring in Rapidly Changing Rural Places E. Helen Berry,* Richard S. Krannich” and Thomas Greiderj*Department of Sociology, Utah State University, Logan, UT 84322-0730, U.S.A.; t Science Applications International Corporation, Las Vegas, Nevada, U.S.A.
Airstmct - A substantial body of research on changes in neighboring patterns has emphasized the impact of ecological and ~p~tion change. This paper examines neighboring in four western U.S. communities experiencing either rapid popuiation increases or declines. Differentials in the sociodemographic composition of rural communities did influence local interaction patterns. Differences in population and ecological growth patterns did not ~~ndently affect neigh~~~g.
people living within a specific area, sharing common ties, and interacting. The idea of community embodies shared locations, experiences, and interactions. Linked to this conceptualization is symbolic identification with the community, or ‘sense of place’, which refers to the meaning of and feelings toward particular places from the perspective of those who live and work there. Sense of place is an element of the experience of community (Eyles, 1985) which has important rami~~tions for community participation, attachment, and action (Firey, 1945; Wilkinson, 1970).
htroduction
has been a central feature in sociological investigations of community and community change for decades. A substantial body of research emphasizes the impact of ecological and ~p~ation change on primary associations, sense of community and community identification, and concomitant activities associated with neighboring. However, most studies examine neighing in urban places, where the number of residents and the tendency for urban social networks to extend beyond the confines of local neighborhoods can diffuse the localized effects of community change (Fischer, 1982; Wellman, 1979). Studies of neighboring in rural places have been limited in both number and scope, with little focus on how neighbo~ng relations may be altered as a result of local economic and demographic changes. Neighboring
Neighboring itself involves elements of both sense of place and sense of community. Neighboring requires two general components: residential proximity and social interaction (Keller, 1968). Since neighborhoods are distinguished from mere residential areas by the level of social interaction (Schwirian, 1983), neighboring can be defined as geographically based social interaction.
This paper examines nei~~ring in four rural towns, each of which has experienced a different pattern of economic and population change. The communities were surveyed at three points in time, proving data on local social conditions during periods of economic boom, bust, or stability and population increase, decrease or stability. These data are used to illuminate the relationship between changing characte~sti~ of place and neigh~~ng.
Neighboring is not, however, merely propinquity. Neighboring is an interactional elaboration of local identi~cation and sense of place. As such, neighboring ties are important reflections of local identification. Local places provide an identity, a basis from which interactions proceed.’ As Lyon phrases it, a minimum level of community rises naturally from propinquity . . . when people live near one another, a level of interaction and common identification is forced upon them (1987, p. 111).
Neighboring and community change The term community has many meanings (Hillery,
1955), but in social science usage generally refers to
‘Thus,
175
neighboring represents one important dimen-
176 sion of the broader array of phenomena community.
E. Helen Berry ef al. identified as
Changes in the extent or function of neighboring or other community relations are often viewed as reflections of shifts in the broader context of social organization (Shevky and Bell, 1955; Greer, 1962). As a result, reductions in localized social interactions have been attributed to increasing societal scale and resultant alterations in status, familism, and ethnicity of residential areas (Nisbet, 1953; Stein, 1960; Wirth, 1938; Vidich and Bensman, 1958). Although Simmel (1950) set forth the initial arguments, Wirth’s (1938) work provides the clearest formulation of the hypothesis that population change, linked to shifts in economic structure, leads to changes in communal relations. Throughout the twentieth century, the predominant social science perspective has asserted that communal associations are ‘lost’ in the context of demographic and economic changes associated with urbanization and modernization (Bender, 1978; Wellman, 1979). However, more contemporary research indicates that neighborly interactions frequently persevere in local urban communities (Gans, 1962; Hunter, 1974; Rubin, 1969; Webber, 1963; Bender, 1978; Suttles, 1972). Research in rural communities has focused less extensively on neighboring per se, but indicates that identification with geographic community and a localized ‘sense of place’ remain strong in rural settings (Munch and Campbell, 1963; Clemente et al., 1974), despite increased vertical ties linking small towns to the broader extralocal society (Warren, 1978). Haga and Folse (1971) examined identification of rural residents with a community and found that though they may be forced to shop, buy gas, or find medical services further away, they tend to identify with the physically closest perceivable community. . . Kinship, church membership, community action, long-term residence, and occupation say more about (the rural residents’) concept of self than the location of his (sic) grocery store or filling station (pp. 49-50).
Thus, local interaction remains important in both urban and rural places, though its form may alter as communities become more or less developed and differentiated. Understanding of the degree to which such alterations occur is limited, however, by the focus and methodology of most research on neighboring and related community phenomena. Most research on neighboring has been conducted in large urban settings and has examined neighboring crosssectionally; relatively little research has documented how neighboring patterns may change, particularly
in smaller communities where ecological change is likely to have a more direct impact (Weber and Howell, 1982). The rural neighborhood is the locale within which social and mutual aid supports among families take place (Smith and Zopf, 1970). Yet rural neighboring is rhoughr (emphasis added) to be disappearing in the face of improved transportation and the consolidation of churches, schools, and stores (Rogers et al., 1988).
Whether, and the extent to which such changes result from economic development and population change remains unknown. This study addresses that uncertainty by examining relationships between community economic and population fluctuations and neighboring patterns. Study approach Research setting The research examines neighboring in four small communities in the U.S. Intermountain West. Three of these communities experienced rapid economic and population dislocations characteristic of energyrelated boomtowns (see Weber and Howell, 1982), while one community has experienced relatively stable population and economic growth for a period of nearly 20 years. The four communities, chosen in 1980 on the basis of their anticipated growth and stability, were surveyed in 1982.1984, and 1986 (see Table 1). In 1982 Evanston, Wyoming, was at the height of an oil exploration and development boom with a population of 11,000. The population increased to 12,427 by 1984, then remained relatively stable from 1984 to 1986.
In 1982 Delta, Utah, was a small farming community in the preliminary stages of a boom related to the construction of a nearby power plant. By 1984 Delta had nearly tripled in size, growing from just under 2,300 to 6,700 residents. By 1986, however, the population had declined by one-third to 4,900. Vernal, Utah, was at the end of a petroleum-related boom in 1982. The population declined from about 7,900 in 1982 to 7,800 in 1984, then dropped further to about 7,000 in 1986. Population growth patterns in Tremonton, Utah, have remained constant, with stable 4-5% growth over the entire period of analysis and no evidence of accelerated growth since 1970.
NeighMug
in Rapidly
ChangingRu~al
177
Places
Table 1. Population estimates and average annual’ rates of growth, 1970-1986
Vernal
Evanston Population 1970 1980 1982 1984 1986
4,462 6,421 ll,ooob 12,427d 12,17P
1,610 1,930 2,285’ 6,67F 4,900’
3,908 6,600 7,883’ 7,792” 7,0@
Average annual growth rates 1980-1982 26.92% 1982-1984 6.10% 1984-1986 -1.02%
Delta
8.88% -0.58% -5.36%
8.44% 53.56% -15.42%
Tremonton
2,794 3,464 3,7648 4,083s 4,xKY
4.15% 4.62% -4.86%
‘The average annual rate of growth is calculated, using assumptions of geometric change, as follows (see Shryock et al., 1976):
where n = P?i P0 r
= = = =
time interval between estimates estimate at later date estimate at earlier date annualized rate of growth.
*Personal communication with Steve Snyder, City Manager of Evanston, Wyoming (April 1982). ‘Estimates: 1980 population of Delta was 21.5% of Millard County, and Vernal population was 32.2% of Uinta County. These percentages were applied to 1 July 1982 (Barber et al., 1983) and 1 July 1984 (Barber et al., 1985) estimates of county $pulations to arrive at these estimates. Buck, 1984. CBoulton, 1984. fU.S. Department of Commerce, Bureau of the Census, special census of Uinta County, Wyoming, 21 August 1986. sU.S. Department of Commerce, Bureau of the Census, population estimates 1 July 1982 and 1 July 1984. ‘Personal communication with Ken Bassett, Vernal City Manager, January 1987. ‘Personal communication with Dorothy Jeffrey, Delta City Recorder, January 1987. ‘Personal communication with Paul Buys, Tremonton City Manager, January 1987.
Survey design Adult
members
of randomly
selected
households
were surveyed during the summers of 1982,1984 and 1986. Sampling frames were based on local water utility listings, supplemented by lists of housing units without separate meters to include dwelling units in mobile home parks and apartment complexes. Questionnaires were distributed to self-designated household heads using drop-off/pick-up procedures. If the household head was unable or unwilling to respond, another adult member of the household was permitted to complete the questionnaire. Sample size for the combined surveys was 2,164, with response rates ranging from 72% to 86% across years and communities (Krannich et ol., 1989).
variables on neighboring, ordinary least-squares multiple regression techniques are used. Each dependent variable was analyzed separately for each community. The unstandardized coefficients (b’s), which are in the metric of the dependent variable, may be compared across the four communities, while the standardized coefficients (betas) indicate which variable ‘drives’ the regression equation.* Pairs of variables related to a single concept (e.g. education and income as indicators of status) were entered singly and jointly. This allowed determination of the individual impact of each indicator and the variation due to each of five conceptual variables: community change, familism, social status, religion, and residential stability. Dependent variables
Statistical procedures To determine
the influence
of the independent
Neighboring activities involve both affiliation/ familiarity with neighbors and social support. Social
178
E. Helen Berry er al.
interactions that provide social support have at least a task component and an two components: emotional component (Thoits, 1982). These several aspects of neighboring were operationalized as three dependent variables.
changes on neighboring, it is important to consider interrelationships involving variables which may also influence patterns of neighboring. These include aspects of familism, social status, ethnicity (as measured here by religion), and residential stability.
A task index of social support consists of responses to a question asking whether respondents would depend on neighbors to (a) watch the house (b) water the lawn (c) run errands and (d) lend money if hospitalized for two weeks. The items were coded 0 or 1, if a respondent indicated that he or she could not, or could, count on neighbors. The index is the sum of the items; 0 indicated no support and 4 indicated full support3
Social area and network community analysis, as well as Suttles’ (1972) typology of neighborhoods, assume that the extent and form of neighboring are related to variables such as familism, social status, and ethnicity. One dimension of familism is measured in this analysis by the number of children residing in the household. A second dimension, housing type, is related to both residential stability and social status and to familism by the tendency for residence in lower-cost and more temporary dwellings to occur during life-cycle periods when children are absent or very young (Michelson, 1970). For this analysis housing type was dichotomized as conventional single-family houses (coded 1) and all other types, including mobile homes and apartments (coded 0).
The emotional support index consists of two items concerning the respondent’s willingness to (a) confide in neighbors about personal problems or crises, and (b) rely on neighbors for emotional support if hospitalized for two weeks. Again responses were coded 0 for no support and 1 for support; the index scores therefore ranged from 0 (no support) to 2 (maximum su~port).~ Familiaritywith neighbors was assessed by responses to the question: Of the 10 houses nearest the respondent, how many has he or she been in? Values range between 0 and 10. Change variables To measure community change, a dummy variable for each survey year (1982, 1984 and 1986) was included in each analysis. The year 1982 served as a reference category, making it possible to evaluate how shifts in population and growth rates corresponding to the different survey periods (Table 1) may combine to influence neighboring in the study communities. Inevitably this approach aggregates many changes which cannot be explicitly addressed due to either limited data or methodological problems (e.g. inclusion of both individual level and aggregate level measures; see Firebaugh, 1979). However, results of alternative multi-level approaches (see Mason et al., 1982) are difficult to interpret. The type of dummy variable analysis used here allows a fairly straightforward comparison of impacts of change over time. The dummy variable for year operationahzes community change, net of shifts over time in the social and demographic characteristics of residents addressed by the several control variables (see Firebaugh, 1979; Krannich et al., 1989; Rao and Miller, 1971). Control variables To isolate the possible effects of rapid community
Years of formal education and family income are included to control for possible differences in neighboring by social status. Ethnic@ is measured by religious affiliation. In the study area, there is little ethnic or racial variation, but there are clear distinctions between Mormons and non-Mormons (Gillespie et al., 1985; Greider and Krannich, 1985; Krannich et al., 1989). Religion was dichotomized as Mormon (coded 1) and non-Mormon (coded 0). Kasarda and Janowitz (1974) demonstrated that intensity of local attachment and involvement are linked to residential stability. Residential stability is measured by length of residence, in months. A second indicator is intention to migrate, assessed by responses to the question, ‘Do you have any plans to move away from this community in the next few years?’ Response categories and the values assigned the responses were: ‘definitely will not move’, coded 1; ‘probably will not move’, coded 2; ‘do not know’, coded 3; ‘probably will move’, coded 4; and ‘definitely will move’, coded 5. Analysis In an earlier cross-sectional analysis, Greider and Krannich (1985) concluded that neighborly interactions continue during rapid population change, although reliance on neighbors for social supports decreased. The mean values of the three neighboring variable%, reported in Table 2, suggest that in most instances community change over time is accompanied by only minor shifts in neighboring interaction. Only in Delta were these changes
Neighboring
in Rapidly ChangingRural
179
Places
Table 2. Means and Fs for neighboring variables, by community: 1982, 1984, 1986 1982
1984
1986
F
Variable Task” Evanston Vernal Delta Tremonton F
1.976 1.933 2.287 2.614 5.1394***
1.766 1.873 1.932 2.443 7.1422***
1.832 1.843 2.172 2.292 4.8796**
1.0238 0.1408 1.8753 1.8879
Emotionalb Evanston Vernal Delta Tremonton F
0.546 0.579 0.750 0.975 8.2229+**
0.507 0.609 0.560 0.904 9.8167***
0.631 0.552 0.736 0.843 4.4498**
1.3481 0.23% 3.2401* 0.9897
Familiarity’ Evanston Vernal Delta Tremonton F
3.964 4.754 6.717 7.262 24.0629***
3.359 4.359 4.2% 6.803 27.3683*“*
3.626 4.467 5.150 6.748 26.7608***
1.4140 0.4801 14.0128*** 0.7343
*p < 0.05. **p < 0.01. ***p < 0.001. ‘Task index of social support, consists of whether the respondent could count on his or her neighbors, should he or she be hospitalized for two weeks, to (a) watch the house; (b) water the lawn; (c) run errands; and (d) lend money. Index ranges from 0 to 4 with the items coded 0 or 1, respectively, if the respondent indicated that he or she could not, or could, count on neighbors, then summed. bThe emotional support index consists of two items indicating whether the respondent was likely to confide in neighbors in times of personal problems or crisis, and whether she or he could count on neighbors for emotional support if hospitalized for two weeks. Again responses were coded 0 for no support and 1 for support, with the summed index yielding a range from 0 for no support to 2 for maximum support. ‘Familiarity with neighbors is operationahxed as: of the 10 houses nearest the respondent, how many has he or she been in? Ranges from 0 to 10.
statistically significant, for both the emotional, and familiarity measures. Otherwise neighboring patterns remained relatively stable within a given community even in the face of considerable change in local socio-economic conditions. Apparently the social and cultural milieu of each community, or alternatively its local ‘identity’, uniquely influence neighboring in each place. Note that the means of all three neighboring variables differ significantly between the communities. Tremonton, the stable cornunity, differed most dramatically from Evanston and Vernal on all three neighboring variables during 1982 and 1984; from Evanston in 1986; and from Vernal on task and familiarity in 1986. Tremonton and Delta had similar neighboring patterns in 1982 (differences in means not statistically different), but in 1984, when Delta’s population increased, there were significant differences between these two places on all three measures.
An alternative explanation is, of course, that the heterogeneity or population mix of a community have more effect on neighboring than does economic change or resultant population growth. To further explore this possibility it is necessary to control for the influence of variables other than change alone. The multivariate analyses reported in Tables 3-5 show that community population change, as measured by the dummy variables for year, had little influence on task, emotional and familiarity measures, net of the influence of the control variables. Neither the individual year variables nor their combined effects contribute to significant variation in the multiple correlation coefficients. Three factors had more effect on task support than year (Table 3). In Evanston, the largest and least homogeneous community, only type of dwelling unit, a measure of familism, was significantly related to task support. In Vernal, Delta and Tremonton, social status as measured by education, income, or
1.221**
0.313 0.098 313
0.024 0.073
0.108
-0.005 0.109
0.097 0.136*
-0.091 -0.072
B
0.005
0.010
0.011
0.022’
0.006
Rz change
0.001 0.031
0.258
0.171 2.68-”
0.081 -0.021
::g
-1.192
b
0.399 0.160 276
0.027
0.008
0.103
0.180’ 0.032
0.095**
’
oOo9
0.001
R’change
0.320** 0.037
0.106 -0.008
0.028 o.ooo
B
Vernal
7.26-” 0.156
0.411
0.035 1.68-05
0.033 0.211
-0.006 0.265
-0.044
b
0.403 0.162 229
0.116 0.124*
0.146
0.063 0.204;:
0.043 0.079
-0.002 0.099
B
Delta
0.038**
0.014
0.045**
oOg7 .
0.010
RZchange
0.329
0.076 1.76-O5
0.042 -0.353
-0.040 -0.371
0.545
b
0.340 0.116 206
0.108 0.064
0.097
0.153’ 0.191**
0.066 -0.099
-0.017 -0.160
B
Tremonton
0.018
0.009
0.070**
0.012
0.020
RZchange
“Task index of social support, consists of whether the respondent could count on his or her neighbors, should he or she be hospitalized for two weeks, to (a) watch the house; (b) water the lawn; (c) run errands; and (d) lend money. The summed index ranges from 0 to 4 with the items coded 0 or 1, respectively, if the respondent indicated that he or she could not, or could, count on neighbors. *&Dummy variables representing year of survey. 1984 is coded 1 for 1984, and 0 for 1982, 1986. 1986 is coded 1 for 1986 and 0 for 1982, 1984. dNumber of children residing in the household. ‘Housing type dichotomized as single-family houses (coded 1) and all other types (coded 0). ‘Years of formal education. ‘Income in dollars. *Religion is dummy coded, 1 = Mormon, 0 = other. ‘Length of residence in the community in months. ‘Plan or intention to migrate is assessed by the answer to the question ‘Do you have any plans to move away from this community in the next few years?’ Response categories and the values assigned the responses were: ‘definitely will not move’, 1; ‘probably will not move’, 2; ‘do not know’, 3; ‘probably will move’, 4; and ‘definitely will move’, 5.
l*p < 0.01. lp < 0.05.
it2 N
Residential stability Length’ 1.65O4 Plans’ 0.064
0.278
-0.003 7.45-06
Social status Education’ Incomes
Religion”
0.078 0.316
-0.174
Familism Children“ Dwelling’
1986’
Communitychange l!w -0.222
Constant
b
Evanston
T4Ae 3. Task index’ of social support regressed on indicators of community change, familism, social status, religion and residential stability for each community
Neighboring
in Rapidly Changing Rural Places
both (respectively) had the most influence. In Vernal and Delta, the communities which experienced the greatest population fluctuations over the study period, residential stability was also significant . The regression on the index for emotional support evidenced association with several independent variables not related to task support (Table 4). However, the year variables again explained little variation. In Evanston, Vernal and Tremonton, emotional support was influenced by the number of children and length of residence. Education is important in Vernal and Tremonton; religion influences support in Delta and Evanston. Familiarity with neighbors involved similar factors in all four communities (Table 5). Length of residence and number of children, both shown by earlier research to influence local participation, social involvement and attachment (see Kasarda and Janowitz, 1974; Michelson, 1970), exhibited significant partial association with familiarity. In Delta and Evanston, type of dwelling is also a significant predictor of familiarity. Education and religion are significant influences in Vernal and Tremonton. Income is a factor in Delta; ‘plans to move’ is related to familiarity in Tremonton. Discussion Change in community scale, as reflected in economic and population growth or decline, does not affect neighboring in these communities. Rather, variables related to the socio-demographic composition of the communities apparently are more important determinants of neighboring activities. Though compositional changes are themselves a result of population and economic change, these findings suggest that it is the local socio-cultural context, not change in community scale, that affects neighboring. Conventional wisdom on community change has held that, cereris p&bus, population increase or decrease will result in an altered sense of community, perhaps a changed sense of place, and altered patterns of local acquaintanceship (Freudenberg, 1986). Although one would expect to see an effect of population change in small communities if such a relationship exists, this study suggests that changes in neighboring levels did not accompany changes in community scale. Instead, neighboring is better predicted by familism, social status, religion or residential stability. These are variables which differ from place to place as does neighboring which varied more between communities than within communities over time. This leads us to suggest that
181
some socio-cultural conditions which differentiate places have more to do with neighboring than changes in the population size or economic base of the community. To elaborate this observation, examination of the control variables influencing neighboring is appropriate. Task support is influenced by status in the three smallest communities (Vernal, Delta, and Tremonton) and by residential stability, or, in the case of Evanston, by the familism/stability variable, dwelling type. Why might task neighboring be more affected by status and stability than community change? Status and household type are frequently cited as factors that influence neighboring (Schwirian, 1983). In the U.S.A., residence in multiple dwelling units or trailers is often viewed as a reflection of both transiency and lower social status. Second, task neighboring may be influenced by place-specific factors, factors which themselves override the effect of other variables. For example, in Evanston only type of dwelling unit has a statistically significant effect. Evanston was also the community with the largest proportions of non-conventional dwelling units (58% in 1982; 45% in 1984 and 1986), with most of these in trailer parks. Trailer park residents have been shown to be among the most isolated in rapidly growing communities (Krannich and Greider, 1984; Moen et al., 1981). In Evanston especially, these units tended to be physically isolated from housing areas where single family units were predominant. In this context the spatial and structural isolation of trailer parks and rental units are likely to override the influence of other status variables. In the communities where the mix of housing tended to favor traditional units, or where housing types were more evenly interspersed, status as measured by either education or income is positively associated with task neighboring. Higher status may lead to greater trust or at least greater reliance on nonrelatives or friends for everyday tasks. Given that higher status persons move more often (Shryock and Siegel, 1976), task reliance on neighbors may be an adaptive mechanism for those who have less access to family or long-term friends. At the same time it is important to note that overall variation in task-type neighboring is poorly explained by either the control or the change variables. Task-oriented neighboring may occur primarily when a need arises due to unusual circumstances or situations, but not as a patterned neighboring activity. Since .this type of neighboring can be perceived as burdensome, neighbors may be less utilized for such tasks than friends or family, if any
0.005 1.55”
0.209
Social status Educatiod Incomes
Religion”
0.378 0.143 313
0.024*
0.017.
0.148*
0.190” -0.014
0.002
0.027**
0.017 0.042
0.157** 0.084
0.001 0.029
0.071
0.085 1.14-06
0.057 -0.086
0.061**
o.tXL?
0.082**
0.018
0.013
1.86-04 0.053
0.288
8:E 229
0.014
0.025*
0. l%*
0.058 0.118
0.019
0.082
0.012
0.080 0.101
0.045 -0.089
-0.090 0.033
B R2 change
and residential
Delta
religion,
0.022 4.22-O6
0.018 -0.012
-0.128 0.045
-0.158
b
social status,
R2 change
8:E 276
0.264** 0.058
0.054
0.300” 0.829
0.142* -0.061
0.067 -0.063
B
Vernal
familism,
9.23-04 0.050
0.138
0.043 4.10-06
0.070 -0.059
-0.044 -0.083
-0.425
b
stability
0.352 0.124 206
0.243** 0.082
0.066
0.142* 0.072
0.178’ -0.027
-0.030 0.058
B
Tremonton
0.068**
0.084
0.029*
0.929’
0.002
R2 change
for each community
< 0.01. ‘p c 0.05. “The emotional support index consists of two items indicating whether the respondent was likely to confide in neighbors in times of personal problems or crisis, and whether she or he could count on neighbors for emotional support if hospitahxed for two weeks. Again responses were coded 0 for no support and 1 for support, with the summed index yielding a range from 0 for no support to 2 for maximum support. *Wmnmy variables representing year: 1984 is coded 1 for 1984,O for 1982 or 1986. 1986 is coded 1 for 1986,O for 1982, 1984. dNumber of children residing in the household. ‘Housing type dichotomized as singIe-family houses (coded 1) and all other types (coded 0). ‘Years of formal education. Qtcome in dollars. hReIigion is dummy coded, 1 = Mormon, 0 = other. $-ength of residence in the community in months. ‘Plan or intention to migrate is assessed by the answer to the question, ‘Do you have any plans to move away from this community in the next few years?’ Response categories and the values assigned the responses were: ‘definitely will not move’, 1; ‘probably will not move’, 2; ‘do not know’, 3; ‘probably will move’, 4; and ‘definitely will move’, 5.
l*p
:a N
Residential stability Length’ 7.04-w Plans’ -0.007
0.069 0.106
Familism Childrend Dwelling’
0.005
0.088 -0.088
-0.11 0.067
Community change -0.014 19846 1986’ 0.089
b
change,
-1.002*
B R2 change
index’ by community
Evanston
support
1.112
b
4. Emotional
Constant
T&e
2.485’*
b
0.005 _. 050
1.307
0.271 -2.61-O6
0.376 0.731
-0.108 0.189
3.67+
b
0.548 0.300 276
0.320** -0.020
0.205’.
0.199** -0.014
0.194’. 0.109
-0.017 0.028
B
0.070”
0.033:’
0.033..
0.045+*
o ’002
0.341
0.008
0.417
0.108 2.52-”
0.314 1.116
-0.954 -0.463
-0.717
b
social status,
R2 change
familism,
Vernal
change,
0.693 0.480 229
0.480** 0.145
0.056
0.073 0.116*
8:;;::
-0.129 -0.066
B
0.201..
0.002
0.021*
0.046**
o .009
R2 change
and residential
Delta
religion
0.005 0.527
1.68
0.504 0.254 206
0.280** 0.193**
0.180+*
0.145’ 0.080
2.04-05
0.183** 0.032
-0.104 -0.138
B
Tremonton
0.135**
0.030**
0.031*
0.032*
0.010
R* change
for each community
0.198
0.322 0.315
-0.675 -0.890
-1.529
b
stability
**p < 0.01. lp < 0.05. “Familiarity with neighbors is in the answers to a single question: Out of the 10 houses nearest the respondent, how many has he or she been in? Ranges from 0 to 10. b*cDummy variables representing year of survey: 1984 is 1 for 1984,O for 1982 or 1986; 1986 is 1 for 1986; 1982 or 1984 is 0. “Number of children residing in the household. ‘Housing type dichotomized as single-family houses (coded 1) and all other types (coded 0). ‘Years of formal education. Qtcome in dollars. hRefigion is dummy coded, 1 = Mormon, 0 = other. ‘Length of residence in the community in months. ‘Pfan or intention to migrate is assessed by the answer to the question, ‘Do you have any plans to move away from this community in the next few years?’ Response categories and the values assigned the responses were: ‘dethtitely will not move’, 1; ‘probably will not move’, 2; ‘do not know’, 3; ‘probably will move’, 4; and ‘definitely will move’, 5.
0.536 0.287 313
:* N
0.116**
0.002
0.045
0.409"
0.003
0.037.’
0.005
R* change
by community
-0.048 0.040
0.168** 0.123*
0.011
0.286
-0.058 6~56-~
0.332 0.700
-0.081 -0.028
B
Evanston
with neighbors
Residential stability Length’ 0.087 Phtll4’ 0.028
Religionh
social status Educatiod Incomes
Llwellin~
Familism Childrend
Community change 19846 -0.484 1986” -0.167
Constant
Table 5. Familiarit
184
E. Helen Berry et al.
are available. Indeed, these are the sort of tasks where reliance on the ‘modified extended family’ (Sussman, 1959) may be most common. That certain variables related to emotional support are not related to task support implies that choosing neighbors for emotional support depends more on shared values, attitudes and beliefs. To explore this, let us examine the contribution of religion to emotional support. Only in Evanston and Delta does religion have an effect. Historically these communities were predominantly Mormon, but both experienced dramatic declines ‘in the proportion of Mormon residents during periods of boom growth. Inmigration brought dramatic changes in the religious composition of these communities, possibly making religious commonality a more important basis for affiliation among those suddenly confronted by a more culturally diverse community setting. No such dramatic change occurred in either Tremonton or Vernal during the study period. In these two places the absence of major shifts in the socio-cultural milieu could result in less emphasis on the importance of religious commonality as a basis for neighboring. Under more stable socio-cultural conditions, other influences come into play. For example, both length of residence and children residing in the household had an effect in all communities except Delta; education was an important predictor of emotional reliance on neighbors in Tremonton and Vernal. Education, religion, and familism are all three life-style related, but length of residence and children are also life-cycle factors. Presumably those at particular stages in the life-cycle exhibit life-styles that are congruent with empathetic neighboring relations (see Michelson, 1970; Timms, 1971). More variance in familiarity is explained by the independent variables than is the case for either emotional or task-type neighboring. This is largely due to the influence of length of residence. In every community, length of residence had the largest partial association with familiarity. Obviously the longer one is in place the greater the probability of meeting one’s neighbors. Having children has a substantial positive effect on familiarity since children increase the number of possible family members who can establish contact with neighbors. Children themselves, because they are usually restricted to the local area by parental proscriptions and limited geographic mobility, have fewer options for contacts outside the local area and are more likely to interact with neighbors. The ties established by children often lead in turn to familiarity among their parents.
In the three smallest communities, status as measured by education and income positively influenced familiarity. In Evanston, dwelling type, also interpretable as a status-related variable, had an effect. As already remarked, Evanston had a large proportion of rental units and trailer courts that were spatially segregated from more ‘established’ residential areas. Since residence in such units is related to status differentials, any influence of other status variables is overridden. Where overt status differences related to residence type are less apparent, education and income come into play. Following classic spatial human ecological arguments (Hoyt, 1939), those with higher status have greater choice in the placement of their residences. Those with more education and income are more able to choose their neighborhoods and perhaps more likely to choose to become familiar with their neighbors. Those with fewer economic choices may be less inclined to interact with neighbors, partly because they have less ability to choose their residential setting. In the communities where religion had no influence on emotional support, religion does influence familiarity. Religion has insignificant effects on familiarity in places such as Evanston and Delta, where community change was most extreme, but enhances the potential for acquaintance among neighbors in more stable places such as Tremonton and Vernal. In rapidly changing communities shared norms and common beliefs, such as those related to religious affiliations, are more important in establishing emotional contacts. In more stable communities, shared beliefs seem to have less influence on such close ties, though they do enhance familiarity. In summary, the array of variables affecting neighboring is complex. The socio-demographic variables considered in this research exhibit shifting relationships with neighboring, depending on the community and the dimension of neighboring under consideration. Moreover, only a modest proportion of variation in the neighboring measures is accounted for by the independent variables, suggesting the operation of other unmeasured variables not available for consideration in this analysis. The most significant finding of this research is that change in community size or scale does not inevitably result in changes in neighboring interactions. Changes in community scale may affect the sense of place, community perceptions, and various interactional contexts (Freudenburg, 1986; Krannich er al., 1989; Krannich and Greider, 1989). Nevertheless, neighboring relations appear to be relatively
Neighboring
in Rapidly Changing Rural Places
‘durable’, and are affected by variables other than shifts in community scale. This suggests that integrative social ties linked to interactions among neighbors can provide an important locus of stability in the context of rapid community change. Efforts to counteract the social disruptions of rapid growth in rural communities might therefore be enhanced by programs that would encourage the development of more extensive neighboring relations among both established and new residents. - The authors wish to acknowledge the contributions of Dr Ronald L. Little and an anonymous reviewer to the study. This research was supported by the Utah Agricultural Experiment Station, Utah State University, Logan, UT 84322-4845, U.S.A. (Project No. UTA 839). Approved as journal paper No. 3681.
Acknowledgements
Notes Communal relations need not be linked with residential proximity (Bender, 1978; Webber, 1%3). Fischer (1975,1976,1981,1982) and Wellman (1979) show that interactions traditionally associated with propinquity continue throughout, even across, major metropolitan areas as well as in rural places though the geographic aspect of neighboring remains a factor in social networks (Fischer, 1982, pp. 97-106). Logistic regression was not used since interest is in the patterns of responses which are best indicated by standardized regression coefficients, not by size of effects. Cronbach’s alpha for this index ranges from 0.61 to 0.81, depending on year and community. For a two item index alpha is an inap~op~ate indicator of item intercorrelations; &i-square indicated that these variables are not statistically independent of one another. References Barber, B.T.. Jensen, KE., Hachman, F.C., Brockert, J.D. and Taylor, J.M. (1983) 1983 population estimates for Utah. Utah Economic and Business Review 43, l-9. Barber, B.T., Jensen, K.E., Hachman, F.C., Brockert, J.D. and Taylor, J.M. (1985) 1984 population estimates for Utah. Utah Economic and Business Review 45-l-9.
Bender, T. (1978) Community and Social Change in America. Rutgers University Press, New Brunswick, New Jersey. Boulton, G. (1984) Delta - small town despite boom of IPP. Salt Lake Tribune November 13, B-l. Buck, G. (1984) Memorandum. Program and fiscal coordinator of the Lincoln-Uinta County Association of Governments, 17 April. Kemmerer, Wyoming. Ciemente, F., Rojek, D. and Beck, E.M. (1974) Trade patterns and community identity: five years later. Rural Sociology 39( 1) (Spring), 92-95. Eyles, J. (1985) Senses of Place. Silberbrook Press,
Cheshire, England. Firebaugh, G. (1979) Assessing group effects: a comparison of two methods. Sociological Methods and Research 7, (May), 384-395.
185
Firey, W. (1945) Sentiment and symbolism as ecological variables. American Sociologic~ Review IO, 140-148.
Fischer, C.S. (1975) Toward a subcultural theory of urbanism. American Journal of Sociology 80, 1319-41. Fischer, C.S. (1976) The Urban Experience. Harcourt Brace Jovanovich, New York. Fischer, C.S. (1981) Public and private worlds of city life. American Sociological Review 46, 306-16. Fischer, C.S. (1982) To Dwell Among Prier&. University of Chicago Press, Chicago. Freudenberg, W.R. (1986) The density of aquaintanceship: an overlooked variable in community research? American Journal of Sociology 92, (July), 27-63. Gans, J.H. (1962) The Urban Villagers: Group and Class in the Life of Italian Americans. Free Press, New York. Gillespie, D., Krannich, R.S. and Leffler, A. (1985) The missing cell: amiability, hostility, and gender differentiation. Social Science Journal 22, 17-30. Greer, S. (1962) The Emerging City: Myth and Reality. Free Press, New York. Greider, T. and Kranmch, R.S. (1985) Neighboring patterns, social support, and rapid growth: a comparison analysis from three western communities. Sociological Perspectives 28, 51-70. Haga, W.J. and Folse, CL. (1971) Trade patterns and community identity. Rural Sociology 36, 42-51.
Hillery, G.A. (1955) Definitions of community: areas of agreement. Rural Sociology 20, 779-791. Hoyt, H. (1939) The Structure and Growth of Residential Neighborhood in American Cities. Federal Housing
Administration, Washington, DC. Hunter, A. (1974) Symbolic Communities: The Persistence and Change of Chicago’s Local Communities. University of Chicago Press, Chicago. Kasarda, J.D. and Janowitz, M. (1974) Community attachment in mass society. American Sociological Review 39, 328-339. Keller, S. (1968) The Urban Neighborhood. Random House, New York. Krannich, R-S., Berry, E.H. and Greider, T. (1989) Fear of crime in rapidly changing rural communities: a longitudinal analysis. Rural Sociology 54, (Summer), 195-212. Krannich, R.S. and Greider, T. (1984) Personal wellbeing in rapid growth and stable communities: multiple indicators and contrasting results. Rural Sociology 49,
(Winter), 541-552. Krannich, R.S. and Greider, T. (1989) Rapid growth effects on rural community relations. In American Rural Communities: Trends and Prospects, Luloff, A.E. and Swanson, L. (eds). Westview, Boulder, Colorado (forthcoming). Lyon, L. (1987) The Co~uni~ in Urban Society. The Dorsey Press, Chicago, Illinois. Mason, W.M., Wong, G.Y. and Entwisle, B. (1982) The multilevel linear model: a better way to do contextual analysis. University of Michigan, Population Studies Center, Research Report No. 32-36, Ann Arbor, Michigan. Michelson, W. (1970) Man and His Urban Environment: A Sociological Approach. Addison-Wesley, Reading, Massachu~tts. Moen, E., Boulding, E., Lillydahl, J. and Palm, R. (1981) Women and the Social Costs of Economic Development: Two Colorado Case Studies. Social Impact Assessment Series, No. 5. We&view Press, Boulder, Coforado. Munch, P.A. and Campbell, R.B. (1%3) Interaction and collective identity in a rural locality. Rural Sociology 29,
(March), 18-34.
186
E. Helen Berry et al.
Nisbet, R. (1953) The Quest for Community. Oxford University Press, New York. Rao, P. and Miller, R.L. (1971) Applied Econometrics. Wadsworth, Belmont, California. Rogers, E.M. Burdge, R.J., Korsching, P.F. and Donnermeyer, J.F. (1988) Social Change in Rural Societies: An Introduction to Rural Sociology (3rd Edition). PrenticeHall, New Jersey. Rubin, I. (1%9) Function and structure of community: conceptual and theoretical analysis. International Review of Community Development 21-22, 111-119. Schwirian, K.P. (1983) Models of neighborhood change. Annual Review of Sociology 9, 83-102. Shevky, E. and Bell, W. (1955) Social Area Annlysis: Theory, Illustrative Application and Computational Procedures. Stanford University Press, Stanford, California. Shryock, H.S., Siegel, J.S. and Associates (1976) The Methods and Mater&& of Demography. Stockwell, E.G. (ed.) (condensed). Academic Press, New York. Simmel, G. (1950) The Sociology of Georg Simmel. Kurt, H. (ed.) (translated). The Free Press, Wolff, New York. Smith, T.L. and Zopf, P.E. (1970) Principles of Inductive Rural Sociology. F.A. Davis Co., Philadelphia. Stein, M.R. (1960) The Eclipse of Community. Harper and Row, New York. Sussman, M.B. (1959) The isolated nuclear family: fact or fiction. Social Problems 6, (Spring), 338. Suttles, G. (1972) The Social Construction of Communities. University of Chicago Press, Chicago. Thoits, P.A. (1982) Conceptual, methodological and theoretical problems in studying social support as a
buffer against life stress. Journal of Health and Social Behavior 23, 145-159. Timms, D.W.G. (1971) The Urban Mosaic. Cambridge University Press, Cambridge. U.S. Department of Commerce (1984) Bureau of the Census, Population Estimates, 1 July 1982 and 1 July 1984. U.S. Government Printing Office, Washington, D.C. U.S. Department of Commerce (1986) Bureau of the Census, Special Census of Uinta County, Wyoming, 21 August 1986. U.S. Government Printing Office, Washington, D.C. Vidich, A.J. and Bensman, J. (1958) Small Town in Mass Society: Class, Power and Religion in a Rural Community. Princeton University Press, Princeton, New Jersey. Warren, R. (1978) The Community in America. Rand McNally and Company, Chicago. Webber, M. (1%3) Order in diversity: community without propinquity. In Cities and Space: The Future Uses of Urban Land, pp. 23-54, Wingo, L. Jr, (ed.). Johns Hopkins University Press, Baltimore. Weber, B.A. and Howell, R.E. (1982) Coping with Rapid Growth in Rural Communities. Westview Press, Boulder. Wellman, B. (1979) The community question: the intimate networks of East Yorkers. American Journal of Sociology 84, 1202-1231. Wilkinson, K.P. (1970) The community as a social field. Social Forces 58311-322. Wirth, L. (1938) Urbanism as a way of life. American Journal of Sociology 44, l-24.