Community support of a business: The community support of the business is de-
ned mainly in terms of patronage and promotion. Note that other characteristics
of the community, such as the level of trust {TRUST} or the friendliness of the
town {FRIENDLY} are included in the analysis, but not as part of a potentially reciprocal community support relation. The respondents’ subjective assessment of how
much support they receive from the town was reected in ve questions, which
were averaged into the variable labelled {TOWN}. One question directly asked the
respondents to rank the town’s appreciation of their business {appreciate}; fewer
than one-half replied that they felt somewhat or greatly appreciated. Another question asked if residents ‘support local business’ {supportive}; 45% of the respondents
agreed, and 6% strongly agreed. Almost 60% replied that they were satised
( 19% ) or very satised ( 40% ) with their relationships with their own customers
{satised}. Almost two-thirds reported their customer loyalty {loyal} to be predictable ( 36% ) or very predictable ( 29% ) . Respondents felt that people of their town
cared ( 53% ) or really cared ( 6% ) about the fate of the business {care}. The nonmissing responses to these ve ordered categorical variables were averaged to generate {TOWN}, a continuous variable bounded between ( 1, 5) . The term {TOWN}
was used as a measure of unilateral, but potentially reciprocated, community support of the business.
The summary statistics for three of the ve unilateral support variables are presented in table 1. As explained above, a business’s unilateral support of the community
was dened as taking one or both of two forms: cash or in-kind donations by the
business {PROF} ˆ {DONATE} and/or {SUPPORT}; or service or boosterism by the
business person {PERS} ˆ {SERVE} and/or {BOOST}. The denition of the community unilateral support of the business included the community’s patronage of and
attitudes toward the business, as perceived by the business person {TOWN}. The
variables {SUPPORT} and {BOOST} had no signicance as main eects in the
model, which is why their summary statistics are not presented in table 1.