A Tale of Two Charitable Campaigns

Abstract

In light of the current economic conditions and the subsequent increased pressure on nonprofit organizations to collaborate, many nonprofit organizations are developing and conducting cross-sector workplace giving campaigns to increase philanthropic activity. Although some scholars have focused on the implications of such activities for for-profit organizations, little research has been conducted to better understand employee-level giving behaviors in charitable workplace campaigns. This longitudinal study focuses on workplace givers and the impact of individual-level factors on actual donation amounts in two annual workplace campaigns at a large public university from 2001 to 2008. Results show that salary consistently predicts giving amounts across campaigns; length of service, however, only predicts giving amounts in one campaign. Being promoted and receiving tenure led to employees donating less, whereas being promoted while already tenured led to employees donating more. We close the article with a discussion of the managerial implications of our findings.

Keywords

equity theory hierarchical linear modeling philanthropy workplace giving

Introduction

Many scholars (for example, Nathan, 1996; Paarlberg & Gen, 2009; Van Slyke, 2003) have emphasized the importance of public–nonprofit partnerships in providing services and goods that benefit society. James (2009, p. 661) notes that “[t]he success and funding of nonprofits have become increasingly critical to government organizations’ ability to achieve policy objectives.” However, the public and nonprofit relationship (Salamon & Elliott, 2002; Weisbrod, 2001) is in the midst of challenging economic times as nonprofit organizations struggle to maintain service delivery and government organizations scramble to supplement declining tax revenues. Under these circumstances, strategic relationships may help bridge budget and program service delivery shortfalls. In this article, we focus on the dynamics of one such partnership that is increasingly prevalent: public sector workplace giving campaigns.

Cross-sector partnerships that facilitate and encourage private and government employees’ donations of money to charitable organizations have been in use for some time (Barman, 2006; Romney-Alexander, 2002). Some estimate that the U.S. federal government’s combined workplace giving campaign alone has generated between US$240 and US$270 million in donations each year for the last decade (Bowman, 2003). These annual federal donations represent a small but significant portion of a growing annual total of workplace giving campaign funds ranging between US$3.5 and US$5 billion (Barman, 2006; NCRP, 2003). In addition to federal workplace campaigns, most states have state-wide workplace giving campaigns.

Notwithstanding their prevalence, few researchers have quantitatively studied workplace giving campaigns, and even fewer (e.g., Bowman, 2006) have analyzed actual workplace giving behaviors over time. The purpose of this study is to explore the determinants of workplace giving. Our data give us a rare opportunity to inform the organizational perspective (e.g., what can organization X do to make campaigns more effective as they seek to partner with giving employees in solving social problems), by beginning to more carefully examine individual characteristics of givers over time (e.g., what employee characteristics prompt giving behaviors and how they change over time). In this article, we begin with a literature review of relevant theoretical and empirical scholarship that informs our hypotheses about the effect of these predictors on the amount that one donates to workplace giving campaigns. Next, we discuss the data and the workplace giving campaigns from which they were drawn. We then present the results of hierarchical linear modeling (HLM) analyses that assess the impact of the hypothesized predictors on the amount donated to the workplace giving campaigns. Finally, we discuss the findings as well as the shortcomings and implications of this research.

Literature Review

Because few studies have specifically focused on workplace giving, in developing hypotheses for the present study, we borrow heavily from the general philanthropic studies literature. Our approach is twofold. First, we focus on the common sociodemographic variables that have been identified as important in an individual’s inclination to donate charitably. In the present study, our data allow us to examine sex, age, and income. Rather than treat these variables as controls, we develop formal hypotheses around them because they have been quantitatively untested in the context of workplace giving.

Second, beyond a more individual-level approach that focuses on sociodemographic variables, an important theme in the literature involves the connection between individual donors and their organizations (e.g., Healy, 2004; Schervish & Havens, 1997; Straub, 2003). We develop hypotheses around “dyadic” variables (Barman, 2007) that focus on the relationship between the employee (potential donor) and employer (campaign partner and promoter). In this vein, we are interested in the role of length of service, position, promotion. Each of these variables represents characteristics of the relationship between the individual and the organization (e.g., length of service reflects years given to one’s employer). In our study, as is usually the case, the employer also represents the campaign, asking employees for donations on behalf of the nonprofit partners.

Sociodemographic Determinants: Sex, Age, and Income

Sex and philanthropy

Overall, women tend to be stronger in various measures of philanthropy and giving than men; women give and volunteer more than men (Hodgkinson & Weitzman, 1996). Mesch, Rooney, Steinberg, and Denton (2006) found that single women donated more money overall than single men, and married persons gave more than single persons, likely because marriage socializes men to have more similar giving behaviors to women. In addition, women typically shoulder the decision in the household about whether or not to give and which organizations to give to (Andreoni, Brown, & Rischall, 2003), which explains the similarity in giving behaviors between married couples and single women. The differences in giving between men and women is important, however, because women are gaining control of more of the charitable dollars being spent and the decisions they make can have a large impact on philanthropy (Kaplan & Hayes, 1993). We expect this difference in giving to play out in the workforce, too, especially given that the proportion of women working in the paid labor market is higher than ever before (U.S. Bureau of Labor Statistics, 2008).

Hypothesis 1 (H1): Women will give more money to workplace giving campaigns than similarly situated men.

Age and philanthropy

Many philanthropic behaviors are associated with age (e.g., volunteering, Musick & Wilson, 2008). Retirement-aged households with high assets can be very strong charitable givers (James & Sharpe, 2007). One study found that older Americans were more likely to plan charitable bequests into their final estate settlements on their deaths than younger Americans (James, 2008). It is important to note, however, that some studies find no effect of age on charitable giving when controlling for other factors, such as social networks or household wealth (Brown & Ferris, 2007; Rooney, Steinberg, & Schervish, 2001). In the absence of these control variables in our study, we feel that it is important to investigate the relationships between age and giving in a workplace setting. Although we expect salary to be correlated with age, we also expect that age will have an impact on charitable giving beyond the impact of salary; age is a weak proxy for a more financially secure household with more assets.

Hypothesis 2 (H2): Older employees will give more money to workplace giving campaigns than similarly situated younger employees.

Salary and philanthropy

One of the strongest and most consistent predictors of individual giving and philanthropy is personal or household income. In general, giving increases with income (Gittell & Tibaldi, 2006; Rooney et al., 2001). Many of the earliest studies investigating the relationship between income and giving found a U-shaped relationship between income and donation amounts such that the poorest and richest people give the most (e.g., Clotfelter & Steuerle, 1981). Later studies challenged this finding, claiming that the relationship between income and charitable dollars spent was flatter across income categories with only a slight uptick for the richest people (Schervish & Havens, 1995, 1998). Other researchers have challenged Schervish and Haven’s findings by showing that the samples they used contained upward bias. For instance, James and Sharpe (2007) found stronger evidence for the U-shaped relationship between income and philanthropic dollars given and that many of these poor, low-income givers are actually high-asset households of retirement age. Therefore, we have reason to expect that employee salary will be positively related to giving in the workplace as well.

Hypothesis 3 (H3): Employee salary will be positively related to donation amounts in workplace giving campaigns.

Dyadic Characteristics: Length of Service, Position, and Promotion

Reciprocity, social responsibility, and equity

Extending the previous discussion, individual charitable behaviors are also linked to socioeconomic status. Those in socially high status groups tend to donate more (McPherson, 1981), because philanthropic behaviors may be perceived as status symbols in and of themselves (Ostrower, 1997); higher status individuals tend to be givers. Applied to workplace giving, measures of one’s status within an organization are also “dyadic” indicators that might also be important predictors of giving (e.g., length of employment, receiving a promotion, and position). Our logic is motivated by theories from biology and industrial and social psychology. Research suggests that reciprocity and equity are important determinants of prosocial behaviors, like charitable donations (Valor, 2006).

Triver’s (1971) groundbreaking piece on the evolution of reciprocal altruism suggests that natural selection can actually favor “altruistic behaviors because in the long run they benefit the organism performing them” (p. 35). In short, altruistically reciprocating (behaving benevolently) can be advantageous to the reciprocator. Theorists have examined reciprocity in industrial settings in work on prosocial (Brief & Motowidlo, 1986; Karylowski, 1982; Rushton, 1982) and organizational citizenship behaviors (Smith, Organ, & Near, 1983), and have even suggested that norms of reciprocating beyond the direct benefactor–beneficiary relationship—for example, to any “dependent other” (Brief & Motowidlo, 1986, p. 718; Smith et al., 1983, p. 718)—may, under certain circumstances (Schwartz, 1977), influence altruistic behavior. Scholars refer to this extension of reciprocity as social responsibility (Berkowitz & Daniels, 1963; Krebs, 1970).

In thinking about reciprocity in its direct and broader (i.e., social responsibility) contexts, Triver’s general framework of reciprocal altruism has several advantages when applied to workplace giving. First, it is relation based; it probes how organisms respond to environmental cues. In our study, we ask how employees respond to signals from their employer. Second, it is dynamic, allowing for iterative (evolutionary) behavior as employees develop a relationship with their employing organization over time. Third, Triver distinguishes between those who perform reciprocating behaviors (e.g., donating to a workplace giving campaign) and those who do not, or “cheaters.”

Equity theory informs this final point. Frameworks based on Adams’s (1963, 1965) theory distinguish between inputs—what employees put into their jobs including task and nontask behaviors—and outcomes—what employees receive in return. When the outcome/input ratio is unequal, whether the employee feels overrewarded or underrewarded—the employee feels distress and motivation to equalize the relationship. Huseman, Hatfield, & Miles, (1987) suggested that employees experience different levels of sensitivity to inequities; those with an entitled sensitivity prefer to be overbenefitted in regard to their equity ratio. Those of the benevolent class are less distressed than others when they are underbenefitted by their equity ratio. In terms of Triver’s framework, these are the most likely to altruistically reciprocate.

We neither have direct measures that would allow us to calculate an employee’s equity ratio, nor do we have the ability to calculate individual employee equity sensitivities. Nevertheless, we have several measures that we believe are important proxies as we explore how reciprocity and equity might shape workplace giving, including length of service, position, and promotion.

Length of service

An employee’s relationship or identification with their employer is a function of their time on the job (Bruner, 1957; Dutton, Dukerich, & Harquail, 1994; Foote, 1951). As employees “gain tenure in an organization, they increase the level and breadth of exposure to the collective organizational identity . . . [employees] come to know themselves as members of the organization” (Dutton et al., 1994). As an employee’s collective identity increases we hypothesize that an employee may feel a greater awareness of, and possibly indebtedness to, their employing organization. We draw from reciprocity and equity theories that as one’s length of service increases, one will be more likely to reciprocate (less likely to cheat) and more sensitive to input/outcome inequities in the direction of benevolence (or willingness to be underrewarded). Specifically, employees may see length of service as an outcome for which additional input is warranted.

Hypothesis 4 (H4): Employee’s organizational length of service will be positively related to donation amounts in workplace giving campaigns.

Position

Employee status at work is likely to be an important determinant of workplace giving in other ways (e.g., the position that an employee holds). Employee stratification is particularly prevalent in public organizations where individuals are often hierarchically arranged in the workplace (e.g., entry level, middle management, and upper management). In our study, we have several position-related strata. We focus primarily on the distinction between staff and tenured faculty employees, but recognize others. For example, another stratum differentiates between senior staff (middle/upper administration) and nonsenior staff (entry level). The former are generally salaried, with higher levels of education and compensation. The latter are generally hourly and are subject to greater levels of accountability; these employees, unlike senior staff, have to submit time sheets, leave slips, and deal with more administrative bureaucracy. Within the faculty, nontenure-track, tenure-track, and tenured employees comprise other strata, again generally marked by increasing compensation and experience/prestige.

Extending insights from work organizational identity, we hypothesize that even after including the effects of length of service and salary, position influences workplace giving. Tenured professors and senior staff are more visible members of the workplace and are more likely to have images of themselves as employees that are more tightly connected with images of their employer (Dutton et al., 1994; Turner, 1978). Theories of reciprocity and equity suggest that prestigious employees will be more likely to reciprocate through workplace giving, and more likely to view their position as an outcome for which workplace giving might compensate as an input.

Hypothesis5a (H5a): Full professors will be more likely to give more to workplace giving campaigns than hourly staff.

Hypothesis5b (H5b): Associate professors will be more likely to give more to workplace giving campaigns than hourly staff.

Hypothesis5c (H5c): Salaried staff will be more likely to give more to workplace giving campaigns than hourly staff.

Promotion

Finally, as a direct extension of equity theory, we hypothesize that employees who have received a workplace promotion will generally perceive that promotion as an outcome that changes their equity ratio. Workplace giving may be a way, as an input, to reciprocate or equalize the imbalance.

Hypothesis 6 (H6): Employees who have been promoted will give more to the workplace giving campaign than similarly situated employees who have not been promoted.

Method

Data:

The data for this article come from two separate, annual workplace giving campaigns conducted at a large, public university in the United States from 2001 to 2008. The university is part of a state-wide system. In the 2007/2008 academic year, the university employed more than 2,500 employees, approximately 40% of whom were faculty. Enrollment in that same academic year was more than 22,000 students, 21% being graduate students. Campaign A is a specialized campaign that collects donations for local arts and cultural organizations, and is conducted during the Spring semester each year. All of the money collected in Campaign A stays in the local community. Campaign A is unique to this particular university and is not conducted on other campuses in the state-wide university system. Typically, during the weeks of this campaign, the campus hosts short, free performances from different arts groups that benefit from the campaign, and the campaign is generally considered to be more interesting and “fun” than Campaign B. The messages around Campaign A often focus on improving the quality of life in the local area. Repeat giving to Campaign A is high.

Campaign B is an open campaign that is conducted during the Fall semester every year, and focuses primarily on funding health and human services nonprofits. This campaign is required of all state institutions and is therefore conducted on the campus of each university in the system. This campaign is generally known as a United Way campaign and the United Way is often used as a vehicle for distribution of the funds. In the United States, a local United Way, an umbrella nonprofit organization that provides support and services to other human service organizations, often teams up with employers to ask employees to make donations to local human service organizations. The money is then distributed through the United Way to its member organizations. United Way campaigns have been common in the United States for more than 100 years. In the past, money raised during Campaign B did not always stay in the local area and employees had little say about where the money went. During the past decade and in response to requests and complaints from employees, the campaign has changed to be more open, allow more organizations to participate, and give employees greater control over how they designate their donations. Employees can choose to make an undesignated donation in this campaign, or they can select a specific nonprofit organization to which they can donate their funds, provided that the organization has been through the vetting process. The campaign begins with a kickoff program and agency fair where local nonprofit organizations can set up booths to advertise their organization. Messages in this campaign include helping other people, supporting the local community, and often include video appeals from individuals who have been helped by a local human service organization.

The university strongly advocates both of these programs to all employees. University communications to encourage participation include email announcements (precampaign) and reminders (during the campaign), phone message broadcasts, electronic links to video messages, and a large kick-off event where the university chancellor is present. We obtained giving information from the campaign administrator at the university. This information included the amount of money that the employee donated to either campaign from 2001 to 2008. Because this information was paired with the employee’s unique identification number, we were able to link an individual’s giving to either campaign over this 8-year time period. Using the employees’ unique identification numbers, we also matched this longitudinal giving data to university data on employees, including each employee’s sex, salary, rank and position, and hire date for each academic year from 2001 to 2008.

Participants

Participants were university employees who had donated at some point(s) to the university’s giving campaign(s) between 2001 and 2008. Fifty-eight percent of the sample was female, and the sample’s average age was 48 years (SD = 11.51). Of all of the donations over the 8-year period, 33% came from hourly, lower level staff, 20% came from salaried, higher level staff, 8% came from lecturers, 9% came from assistant professors, 13% came from associate professors, and 17% came from full professors. Our unit of analysis is the donation, by year and campaign. Of the 5,760 total donations, 1.35% came from employees who had been promoted from assistant professor to associate professor, and 1.63% came from employees who had been promoted from associate professor to full professor. In addition, 13% of the donations came in 2001, 10% came in 2002, 13% came in 2003, 13% came in 2004, 11% came in 2005, 14% came in 2006, 15% came in 2007, and 12% came in 2008; donations were evenly spread across the years of the study. The average donation to Campaign A was US$54 (SD = 160.38), ranging from US$0 to US$2,880. The average donation to Campaign B was US$131 (SD = 277.98), ranging from US$0 to US$6,960. The average overall amount that an employee gave per year was US$185 (SD = 393.25), ranging from US$1 to US$6,960. In addition, the average employee salary was US$60,742 (SD = 36,010.57), and ranged from US$122 to US$315,000. Finally, the average employee length of service was 4,059 days (SD = 3,371.90), and ranged from 74 days to 15,944 days.

Measures

For the purposes of this study, our independent variables are all descriptive or demographic variables. At Level 1 of our analyses, we include variables that vary unpredictably between observations within individuals (see analytic technique below), including the year of donation (coded as 2001 = 1, 2002 = 2, 2003 = 3, and so forth on the calendar year during which the campaign occurred), employee salary per year of donation, employee length of service (calculated as the number of days that the employee has worked for the organization until the end of the workplace giving campaign for each year), employee position (dummy coded as salaried staff, lecturer, assistant professor, associate professor, and full professor, with hourly staff as the reference group), and whether or not the employee was promoted (dummy coded for whether or not the employee was promoted from assistant professor to associate professor or from associate professor to full professor). At Level 2 of our analyses, we include variables that do not vary unpredictably between observations within individuals, including age in years and employee sex (coded as male = 0, female = 1).

Our dependent variables were the monetary values that each employee donated to Campaign A, Campaign B, or the overall amount that the employee donated to both campaigns (calculated by adding the amount donated to Campaign A and Campaign B in each year).

Analytic technique

We employed multilevel analysis in HLM 6.0 (Bryk & Raudenbush, 1992) to analyze the effects of the various predictors on the amount that each employee gave to each campaign. Traditional regression analysis operates under the assumption that observations are independent of one another, which is rarely the case, and often leads to small standard errors and a greater likelihood that one will find significant effects (Osborne, 2000). HLM is a multilevel regression technique that accounts for intercorrelated errors and allows for more conservative statistical testing. For example, HLM-related techniques offer the ability to account for varying effects simultaneously that may be correlated (e.g., promotion and salary, position and promotion). HLM is also useful in contexts in which data are nested within larger groups, or, as in this case, observations are nested within one individual over time.

We performed hierarchical linear modeling using a two-level hierarchical multivariate linear modeling (HMLM) approach. HMLM is appropriate for repeated measures data in which observations are nested within individuals, and HMLM is especially useful in this situation because the data are incomplete (i.e., most employees did not donate to a workplace giving campaign during every year within the study’s timeframe). In response to this issue, HMLM allows for estimation of multivariate models despite these incomplete data using an EM algorithm (Dempster, Laird, & Rubin, 1977). For each dependent variable (giving to Campaign A, Campaign B, and overall giving), we conducted a two-level analysis in which we regressed the amount that employees gave onto the Level 1 variables, which included the time interval between measurements (year), salary, length of service, position, and the two promotion variables.

We also regressed the amount that each employee gave onto Level 2 variables, which included employee age and sex. These are Level 2 variables because they do not randomly change between observations, and are therefore not observation-level variables, but individual-level variables. Although age does vary, it varies consistently and nonrandomly, and is therefore treated as a Level 2 variable. Length of service, however, is treated as a Level 1 variable for two reasons. First, a substantial number of employees left the university and returned at a later date. Second, many employees come to the university at a later stage of their career. Both of these affect their length of service in a nonsystematic manner. All hypothesized relationships are modeled in Figure 1, and all independent variables (both Level 1 and Level 2) are included in the model at the same time.

Figure 1.

Modeling the proposed hypotheses

We do not model any interactions between any of the Level 1 and Level 2 predictors because there are no theoretical grounds on which we might hypothesize cross-level interactions. In addition, our analyses are exploratory in nature due in part to the nascent stage of workplace giving research, and we intend for this study to inform future research that examines interactions in workplace giving contexts. Therefore, the Level 2 variables are only modeled as predictors for the Level 1 intercept, and all independent variables are assessed as direct effects. In addition, HLM operates under the assumption of a normal distribution; however, large sample sizes mitigate this issue (Wang, Fan, & Willson, 1996). As a result, HLM as well as structural equation modeling (SEM) have been applied to nonnormal data with success (Bentler & Dudgeon, 1996; Brown, 1990; Micceri, 1989).

Results

We began by calculating descriptive statistics (see Table 1) and intercorrelations for variables per level (see Table 2).

Table 1.

Means and Standard Deviations of Relevant Variables

Variable	Mean	SD
Age (years)	47.66	11.507
Female	0.58	0.493
Salary (dollars)	60,742	36,011
Length of service (days)	4,059	3,372
Salaried employees	0.20	0.402
Hourly employees	0.34	0.473
Lecturer	0.08	0.266
Assistant professor	0.09	0.286
Associate professor	0.13	0.332
Full professor	0.17	0.374
Being promoted from assistant to associate professor	0.03	0.167
Being promoted from associate to full professor	0.03	0.164
Campaign A donation	54.40	160.377
Campaign B donation	130.83	277.978
Total donation	185.22	393.245

Note: Level 1 N = 5,760; Level 2 N = 1,887; Age and sex are treated as Level 2 variables; Salaried employees, hourly employees, assistant professor, associate professor, and full professor are dummy coded for whether or not each employee holds that position; Promotion variables are coded as 0 = was not promoted during that year and 1 = was promoted during that year.

Table 2.

Descriptives and Intercorrelations of Relevant Level 1 and Level 2 Variables

Level 1 variables	1	2	3	4	5	6	7	8	9	10	11	12
Year
Salary	.18**
Length of service	−.02	.23**
Salaried employees	.05**	.08**	−.13**
Lecturer	.01	−.16**	−.04**	−.15**
Assistant professor	.04**	.01	−.19**	−.16**	−.09**
Associate professor	−.03*	.14**	.19**	−.19**	−.11**	−.12**
Full professor	−.01	.54**	.37**	−.23**	−.13**	−.14**	−.17**
Being promoted from assistant professor to associate professor	.07**	.03*	−.03*	−.06**	−.03*	−.04**	−.31**	−.05**
Being promoted from associate professor to full professor	.07**	.16**	.06**	−.07**	−.04**	−.04**	−.05**	.29**	−.02
Campaign A donation	.04**	.38**	.18**	−.03*	−.04**	−.06**	.11**	.20**	−.02	.03
Campaign B donation	−.03*	.41**	.13**	−.02	−.03*	−.05**	.09**	.21**	−.02	.09**	.58**
Total donation	−.01	.45**	.17**	−.02^	−.04**	−.06**	.11**	.23**	−.02	.08**	.82**	.94**
Level 2 variables	1	2
Age
Female	−.09**

Note: * p < .05; ** signifies p < .01. Level 1 N = 5,760; Level 2 N = 1,887; Year is coded as 1-8 for each year in the 8-year study timeframe; hourly employees, salaried employees, assistant professor, associate professor, and full professor are dummy coded for whether or not each employee holds that position; promotion variables are coded as 0 = was not promoted during that year and 1 = was promoted during that year.

Year is positively correlated with salary, salaried employee status, and being an assistant professor, but negatively related (marginally) to length of service as well as being an associate professor. These relationships are quite weak (all but the relationship with salary are below 0.06), which is to be expected because there is no theoretical reason why one type of employee may donate during 1 year more than another. As one may expect, salary and length of service are positively correlated. In addition, salary is positively correlated with salaried employee status as well as being an associate or full professor. This is not surprising because salaried employees typically have higher status in the organization.

Length of service was negatively correlated with salaried employee status, being a lecturer and being an assistant professor; this is also unsurprising in that lecturers and assistant professors typically have short-lived careers at universities prior to either being promoted with tenure or denied tenure and forced to leave the university. Associate and full professors, however, have positive relationships with length of service in that these employees have typically been with the organization for a longer period of time than other types of employees. Cross-level correlations are not included because, to do so, one would have to aggregate Level 1 variables, creating an imprecise analysis. We next turn to the results of our HMLM analysis.

Campaign A’s intercept was significant; however, this should be interpreted with caution because this intercept assumes that one has an age, salary, and length of service of zero (all clearly unrealistic assumptions). The intercept for Campaign B and the total amount donated to workplace giving campaigns was nonsignificant. In addition, year was significant for Campaign A, Campaign B, and total giving (albeit marginally significant for Campaign A), such that the amount that one gives to workplace giving campaigns decreases over time.

H1 states that women will give more than men to workplace giving campaigns. We did not find support for this hypothesis in any of the three models; sex was nonsignificant in its effect on giving. H2 states that age will be positively related to donative behavior in workplace giving campaigns. We also did not find support for this hypothesis; age was nonsignificant in all three models (see Tables 3 through 5).

Table 3.

Hierarchical Linear Model Results for Various Effects on Giving to Campaign A

Variable	Coefficient	SE	t
Intercept	−34.44***	12.411	−2.78
Age	0.08	0.214	0.38
Female	4.16	4.994	0.83
Year	−1.39*	0.735	−1.89
Salary	0.0011***	0.0001	12.47
Length of service	0.0043***	0.0008	5.22
Salaried employees	−7.34	6.447	−1.14
Lecturer	4.61	8.271	0.56
Assistant professor	−12.59	8.165	−1.54
Associate professor	0.77	9.055	0.09
Full professor	−4.48	9.943	−0.45
Being promoted from assistant to associate professor	−19.42	16.214	−1.20
Being promoted from associate to full professor	15.13	15.643	0.97

p < .01. **p < .05. *p < .10. The reference for all positional variables is hourly employees and all coefficients should be interpreted as such.

Table 4.

Hierarchical Linear Model Results for Various Effects on Giving to Campaign B

Variable	Coefficient	SE	t
Intercept	5.77	21.407	0.27
Age	0.06	0.365	0.16
Female	12.18	8.479	1.44
Year	−9.33***	1.288	−7.25
Salary	0.0024***	0.0001	17.58
Length of service	0.0019	0.0014	1.36
Salaried employees	−19.47*	10.814	−1.80
Lecturer	0.76	13.771	0.06
Assistant professor	−35.28***	13.527	−2.61
Associate professor	−3.22	15.744	−0.20
Full professor	−38.07**	17.094	−2.23
Being promoted from assistant to associate professor	−50.57**	25.346	−2.00
Being promoted from associate to full professor	53.87**	24.510	2.20

p < .01. **p < .05. *p < .10. The reference for all positional variables is hourly employees and all coefficients should be interpreted as such.

Table 5.

Hierarchical Linear Model Results for Various Effects on Total Giving Amount

Variable	Coefficient	SE	t
Intercept	−24.11	29.23	−0.83
Age	0.10	0.50	0.20
Female	11.68	11.73	1.00
Year	−11.25***	1.753	−6.42
Salary	0.0036***	0.0002	19.08
Length of service	0.0061***	0.0019	3.16
Salaried employees	−28.34*	14.483	−1.96
Lecturer	5.56	18.430	0.30
Assistant professor	−47.98**	18.526	−2.59
Associate professor	−0.62	21.396	−0.03
Full professor	−48.47**	23.043	−2.10
Being promoted from assistant to associate professor	−77.82**	34.146	−2.28
Being promoted from associate to full professor	72.79**	33.626	2.17

p < .01. **p < .05. *p < .10. The reference for all positional variables is hourly employees and all coefficients should be interpreted as such.

H3 stated that salary would positively predict giving amounts for workplace giving campaigns. We found strong support for this hypothesis across all three models; salary significantly predicted the amount that one gives to Campaign A, Campaign B, and total giving in the predicted direction. In other words, for every additional US$1,000 an employee makes, the employee will donate US$1.10 to Campaign A, US$2.40 more to Campaign B, and US$3.60 more to workplace giving campaigns overall.

We found partial support for H4, which states that employees with a longer length of service will give significantly more to workplace giving campaigns than employees with shorter lengths of service. Length of service significantly predicted the amount that one gives to Campaign A and overall giving; however, although the direction of the effect was as predicted, the effect of length of service on the amount that one donates to Campaign B was nonsignificant (β = .0019, p > .05). For Campaign A and overall giving, however, one can interpret the coefficients as follows: for every 1,000 days that an employee works for the organization, that employee will donate an additional US$4.30 to Campaign A and US$6.10 to workplace giving campaigns in general. The differences across campaigns suggest that workplace giving decisions are not simply a function of being asked by one’s employer to give, but are also influenced by the type of organizations that a campaign aims to benefit. Here the most obvious difference is arts/cultural (Campaign A) organizations versus health and human service organizations (Campaign B).

Hypotheses 5a to 5c stated that tenured faculty (associate professors and full professors) and salaried staff would give more to workplace giving campaigns than entry-level staff. We found countervailing findings when testing Hypotheses 5a to 5c; there was no difference among positions for employee donations to Campaign A, but found the opposite of our hypotheses to be true for Campaign B. Specifically, for Campaign B, untenured faculty (assistant professors) gave significantly less than hourly employees, as did tenured full professors. In addition, associate professors (also tenured faculty members) were not significantly different in the amount that they donated to Campaign B when compared to hourly employees. Finally, the amount that salaried employees (those who are higher in the organization but are not tenured faculty members) gave to Campaign B was less than hourly employees, although the effect was marginally significant. In other words, those lowest in the organization (hourly employees) gave significantly more than faculty without tenure (assistant professors), some faculty with tenure (full professors), and senior staff (salaried employees).

When one examines the effect of position on the amount given in total to workplace giving campaigns (see Table 5), the results are consistent with the effects found in Campaign B (Table 4). Specifically, full professors, assistant professors, and salaried staff gave less when compared to those lower in the organization (i.e., hourly staff). Again, associate professors were no different when compared to hourly staff for the total amount donated to workplace giving campaigns. It seems that position in the organization is actually inversely related to the amount that one donates to workplace giving campaigns; however, to better understand the effects of tenure promotions on giving behaviors, we turn to H6.

H6 states that faculty members who have been promoted will give more to the workplace giving campaigns than faculty who have not been promoted. We again find differences between the two campaigns, in that being promoted does have an impact on giving to Campaign B and overall giving, but not for Campaign A. In addition, we find mixed findings with respect to the impact of being promoted on the amount that one donates to workplace giving campaigns. Specifically, being promoted from assistant professor to associate professor (and thus receiving tenure) actually has a negative impact on the amount that one donates to Campaign B and the amount that one donates in general, revealing the opposite effect of H6. However, being promoted from associate professor to full professor (and experiencing no change in tenure status) leads to an increase in the amount that one gives to Campaign B and the amount that one gives overall to workplace giving campaigns, and thus supporting H6. It therefore appears that receiving tenure has an initial negative impact on one’s giving behaviors that cancels out once one is promoted from associate professor to full professor and has had tenure for an extended period of time.

Discussion

This study explored the impact of various sociodemographic characteristics and dyadic characteristics on employees’ donative behavior in two distinct workplace giving campaigns over an 8-year period. We borrowed from philanthropy, industrial psychology, and biology literature to formulate hypotheses involving variables that include sex, age, salary, length of service, position, and promotion, and tested these hypotheses using hierarchical linear modeling to ensure that effects are isolated to the appropriate level of analysis.

Our findings suggest that age and sex are not significant predictors of giving amounts in workplace giving campaigns. Despite the consistent findings in past scholarship that women donate more than men (e.g., Hodgkinson & Weitzman, 1996), the effect of sex failed to reach significance. We suspect that one factor that could be influencing this result is that we could not directly measure marital status. Previous research has shown that married couples’ giving is correlated and that women often make the giving decisions for the household (Andreoni et al., 2003). Indeed, some dual-career couples might alternate years in which they give to their workplace giving campaigns at their respective workplaces. A second possible explanation is that men are more impulsive givers and are more likely to give than women when other people observe their giving (Schlegelmilch, Love, & Diamantopoulos, 1997). Giving in the workplace is often observable because others in one’s work team or department can be aware of a person’s donation. Men are probably more likely to give in these circumstances, and this could negate the influence of gender on giving in the workplace.

Another possible explanation and more intriguing overall implication is that giving in the workplace might be different than giving in other contexts. The motivations to give might be different, the processes by which people make giving decisions are different, and recipient organizations are different. Little empirical research has been done on workplace giving campaigns, and little is therefore known about how donations may vary between these campaigns and other traditional fundraising methods. It is possible that workplace giving campaigns have the potential to even out some of the sex and age inequalities in giving behaviors, although future research will need to examine how marital status might interact with gender to impact giving in the workplace. In other words, workplace giving decisions may not be influenced by the same predictors as other private giving (e.g., religious giving) decisions that may be heavily influenced by demographic characteristics.

As expected, salary was a strong predictor of giving behaviors across both campaigns and for total giving. Those with greater salaries have a greater potential to donate to the campaigns, and in light of equity theory, they may also experience a sense of obligation to compensate for any notion of receiving greater outcomes than their perceived inputs. Length of service, on the other hand, only operated as predicted for Campaign A and total giving. This highlights the fact that differences in campaigns—how they are run and which organizations benefit—can influence giving to the campaign. As we mentioned earlier, Campaign A is unique in that there is a great sense of loyalty attached with Campaign A; those who donate to Campaign A are likely to donate again. Because it only benefits local arts and cultural organizations, this campaign may attract people with a strong interest in the arts.

Theories of reciprocity and social responsibility raise other possible explanations for the different findings across Campaigns A and B. Indeed, as raised by Schwartz (1977) and others (Batson & Powell, 2003; Bendapudi, Singh, & Bendapudi, 1996; Dovidio, Piliavin, Gaertner, Schroeder, & Clark, 1991) differences in the ways these campaigns are implemented (discussed previously) may influence an employee’s personal norms and therefore their sense of reciprocity and social responsibility in a particular campaign. As indicated earlier, the local nature of Campaign A likely triggers deeper personal altruistic norms. To this point, Dovidio et al. (1991) suggest that proximity of the charitable need influences the employee’s giving norms, whereas Bendapudi et al. (1996), suggest that differences in “message variables” across the two campaigns (in which localized vs. generalized needs potentially play a role) shapes individual charitable giving.

Position varied in its effect between campaigns as well. Although only true for Campaign B and total giving, position operated in contrast to our predictions; those higher in the organization actually gave less than those lower in the organization, even after including the effect of salary and length of service (except for associate professors) in the model. This is especially interesting because people with higher education levels tend to donate more to charitable causes (Andreoni et al., 2003; Brown, 2005), and those higher in the organization would theoretically identify more with the organization, increasing their likelihood to reciprocate through workplace giving. Our findings, however, actually demonstrate the opposite is true; those who are higher in the organization and theoretically identify more so with the organization’s identity actually donate less to workplace giving campaigns.

Following equity theory, employees low in the organization may see this as an opportunity to increase their inputs in hopes that the organization will reciprocate and increase those employees’ outcomes. Employees in lower status positions might also feel more peer pressure to donate to the campaign or might feel more insecure in their jobs if they chose not to give; they may view giving to the workplace giving campaign as a means of protecting their jobs by proving their organizational loyalty. Thus, giving might be motivated more by job insecurity rather than organizational identification and loyalty. Viewing the influence of status on giving in this way fits with our finding regarding the promotion from assistant to associate professor leading to a decrease in giving—lower status individuals in the organization might give to protect their organizational status and job. Similarly, those higher in the organization may identify so highly with the organization that they actually see themselves more as sponsors of the program rather than potential donors to the program.

Another potential explanation for these unexpected findings regarding position is that employees in lower positions tend to have more local roots (potentially closer match between the message variable: beneficiary/solicitor similarity (see Bendapudi et al., 1996, p. 38)) than faculty who tend to move in from other areas. As such, these employees might be more committed to local organizations. In addition, many of these employees likely have friends or family members who have benefited from the work of these local nonprofit human service organizations and they feel a greater desire to give back to these organizations. Again, this could be evidence that workplace giving campaigns have the potential to overcome some of the status bias of philanthropy in other contexts. Because every employee is asked to give, not just those who are wealthier or of higher status, workplace giving campaigns can help to further democratize giving.

Finally, promotion varied in its effect on the amount that employees donated to the two campaigns. Specifically, being promoted from assistant professor to associate professor (presumably attaining tenure in the process) led to a decrease in donative behavior in Campaign B and total giving, whereas being promoted from associate professor to full professor (a promotion in which one’s tenure status goes unchanged) led to an increase in donative behavior in Campaign B and total giving. One reason this may be true is that employees may work toward the goal of tenure by putting forth an extraordinary effort; once they receive tenure, however, they may think that their tenure-track efforts (i.e., inputs) have greatly outweighed their experiences with the organization (i.e., outcomes). Some faculty may also build up the notion of tenure with great grandeur, and once they receive it, may feel somewhat let down by the little relative change that they actually experience in their lives, which subsequently contributes to the imbalance between the employee’s inputs and outcomes described above.

Those who are promoted from associate professor to full professor may contribute a significantly greater amount to Campaign B and total giving because they have experienced an increase in their perceived outcomes relative to their perceived inputs, and act to lessen this perceived dissonance by donating more money to the organization (i.e., the sponsor). Similarly, these employees may have finally experienced the latent effect of receiving tenure, and this effect may only serve to compound the perception that their outcomes are substantially greater relative to their inputs, leading to the increase in donative behavior. In addition, the reason this effect may only occur for Campaign B and total giving may be because, as mentioned previously, Campaign A generates a great deal of loyalty and consistency from its donors because it is such a specialized and targeted campaign. Thus, promotion has less impact on donations to Campaign A, whereas promotion has an extreme impact on donations to Campaign B because Campaign B donors are more selective in their donative behaviors in an “open” call for workplace gifts. This may change as Campaign B moves to a more choice-based model, where donors gain increasing ability to designate their gifts. In addition, full professors are more likely to serve in higher profile leadership positions within the organizations and might feel that they need to lead by example in donating to the workplace giving campaign.

Study Limitations, Strengths, and Future Research

One limitation of this study is that there are several factors that could be related to workplace giving that were not available to us in the data we obtained from the university. This would include such demographic information as marital status, number of children, giving at a spouse’s place of employment, and household assets. In addition, our data limit the variables that we can include in our analyses. For example, we measure employee salary but lack a measure of household income, which is traditionally measured in philanthropy research. Our findings should be interpreted in light of the fact that we were unable to investigate workplace giving as a component of household giving.

Future research will also be needed to investigate the influence of religious attendance and other forms of participation on workplace giving. Although we have tried to incorporate theory to guide us as much as possible, in many ways this is an exploratory study seeking to quantitatively examine workplace giving. Based on our findings and theories discussed herein, we suggest future exploration of several factors that we believe could better establish the field’s understanding of workplace giving behaviors. Such factors include the length of an employee’s residence in the community, whether a friend or family member has benefited from a local human service nonprofit organization, and further information about the campaign’s situational context, such as the mood surrounding the campaign, message content, peer and supervisor pressure from others to give, and timing of the campaign. Measurement of individual traits such as prosocial motivations and equity sensitivity would also facilitate interesting research on workplace giving. In general, although the direct role of sex, age, and other demographic characteristics are central in other giving studies, in workplace giving, future researchers might explore how these moderate workplace giving tendencies.

Another limitation is that our data allowed us to explore causes of giving among employees who had given at least once during the 8-year data collection period. In short, we were unable to examine nongiving behavior—an endeavor we hope to take up in future research, if possible. Nevertheless, because we are not aware of any other longitudinal, behavioral studies of workplace giving, we think it important to start by examining the motivators for workplace giving among those with varying propensities to give. Inasmuch as other data may become available, we intend to contrast the present results with future data that may allow us to contrast givers with nongivers.

This study, however, has several strengths. First, we use actual donation amounts rather than individual’s reported amounts of giving or an individual’s attitudes toward giving. Most studies of philanthropic behaviors are based on self-reported giving and are thus subject to bias because of the inability of the subject to accurately recall their giving behaviors, especially when using a shorter survey questionnaire (Rooney et al., 2001). Studies of actual giving behaviors for an entire organization are rare. In addition, we have actual giving over an 8-year time period. Again, the present study is a unique opportunity to unpack workplace philanthropy, in particular, because most studies of philanthropy are not longitudinal. Finally, we are able to assess the giving across two, distinct workplace campaigns. Campaign A has a specific focus and benefits the university’s immediate community. Campaign B is a more traditional workplace campaign that has followed the United Way model. Although many donations benefit the university’s immediate community, some may go to state or regional causes.

Implications and Conclusion

This study helps researchers to highlight the individual and organizational factors that are important to include in future studies of workplace giving. Our research highlights the relative importance of the dyadic relationship between the individual and the organization over individual demographic characteristics. This relationship will need to be explored in greater detail, especially given that some of our hypotheses, specifically those related to status in the organization, were not supported by the data and, in some situations, the effects were the reverse of what we hypothesized. Nevertheless, our unique findings suggest the importance of treating workplace philanthropy as distinct from general or private philanthropy. For practitioners, our exploratory study allows us to raise not only the importance of the employee–employer (dyadic) relationship, but also the importance of proximity and message in workplace giving.

Organizers of workplace giving campaigns can also use information from this study to help with planning, promoting and executing a workplace giving campaign. Our results indicate that not all campaigns are equal. Thus, the timing and focus of a campaign can influence the amount of money that employees donate. Previous research has shown that the focus of a campaign and the organizations benefiting (Bennett, 2003; Sargeant & Jay, 2004) and the messages that campaign administrators used to motivate donations (Brunel & Nelson, 2000; Hibbert & Horne, 1996) can help to explain differences in campaign success. Because this article focused more on demographic determinants of giving in the workplace, a full exploration of differences in the campaigns and campaign messages is beyond the scope of this article. However, our results indicate that this would be a fruitful area for further research. In addition, employees at different status levels in the university give different amounts to the campaign; perhaps each of these groups of employees might need to hear a unique, customized message to motivate them to donate to the campaign. Thus, more targeted marketing of the workplace giving campaign might be in order.

As philanthropic theories and studies develop, our findings suggest that workplace giving may be a unique setting that warrants special attention. Because workplace philanthropy plays an important role in contemporary governance, we encourage future research and theoretical endeavors to explore this distinct context.

Footnotes

Acknowledgements

This research would not have been possible without the exceptional assistance of university staff and administrators at the university described here in. We are also grateful to anonymous reviewers and to the editor for their insights and suggestions which have greatly benefited this work.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Bios

Brett Agypt, is a consultant for Corporate Executive Board. His research interests include philanthropic behavior, selection and strategy in nonprofit organizations, and time as it is experienced and enacted in organizations. He is also a doctoral candidate in Organizational Science at the University of North Carolina at Charlotte.

Robert K. Christensen, PhD, is an assistant professor of public administration and policy at the University of Georgia. His research focuses on the impact of pro-social (e.g., altruism, public service motivation, voluntary behavior, organization citizenship) and anti-social (e.g., prejudice) attitudes/actions and on public and nonprofit work groups and organizations.

Rebecca Nesbit, PhD, is an assistant professor of nonprofit management and arts administration at the University of North Carolina at Charlotte. Her research interests include civic engagement, volunteering, philanthropic behavior, and nonprofit management.

References

Adams

(1963). Towards an understanding of inequity. Journal of abnormal and social psychology, 67, 422-436.

Adams

(1965). Inequity in social exchange. Advances in experimental social psychology, 2, 267-299.

Andreoni

Brown

Rischall

(2003). Charitable giving by married couples: Who decides and why does it matter? Journal of Human Resources, 38, 111-133.

Barman

(2006). Contesting communities: The transformation of workplace charity. Stanford, CA: Stanford University Press.

Barman

(2007). An institutional approach to donor control: From dyadic ties to a field-level analysis. American Journal of Sociology, 112, 1416-1457.

Batson

C. D.

Powell

A. A.

(2003). Altruism and prosocial behavior. In Millon

Lerner

M. J.

(Eds.), Handbook of psychology, Volume 5: Personality and social psychology (pp. 463-484). Hoboken, NJ: Wiley.

Bendapudi

Singh

S. N.

Bendapudi

(1996). Enhancing helping behavior: An integrative framework for promotion planning. Journal of Marketing, 60(3): 33-49.

Bennett

(2003) Factors underlying the inclination to donate to particular types of charity. International Journal of Nonprofit and Voluntary Sector Marketing, 8(1): 12-29.

Bentler

P. M.

Dudgeon

(1996). Covariance structure analysis: Statistical practice, theory, and directions. Annual Review of Psychology, 47, 563-592.

10.

Berkowitz

Daniels

L. R.

(1963). Responsibility and dependency. Journal of Abnormal and Social Psychology, 66, 429.

11.

Bowman

(2003). Workplace giving: A case study of the combined federal campaign. New Directions for Philanthropic Fundraising, 41, 27-37.

12.

Bowman

(2006). Should donors care about overhead costs? Do they care? Nonprofit & Voluntary Sector Quarterly, 35, 288-310.

13.

Brief

A. P.

Motowidlo

S. J.

(1986). Prosocial organizational behaviors. Academy of Management Review, 11, 710-725.

14.

Brown

(2005). College, social capital, and charitable giving. In Brooks

A. C.

(Ed.), Gifts of time and money: The role of charity in America’s communities. Lanham, MD: Rowman & Littlefield.

15.

Brown

Ferris

J. M.

(2007). Social capital and philanthropy: An analysis of the impact of social capital on individual giving and volunteering. Nonprofit and Voluntary Sector Quarterly, 36(1): 85-99.

16.

Brown

R. L.

(1990). The robustness of 2SLS estimation of a non-normally distributed confirmatory factor analysis model. Multivariate Behavioral Research, 25, 455-466.

17.

Brunel

F.F.

Nelson

M.R.

(2000) Explaining gendered responses to help-self and help-others charity ad appeals: The mediating role of world views. Journal of Advertising, 29(3): 15-27.

18.

Bruner

(1957). Going beyond the information given. In Gruber

Terrell

Wertheimer

(Eds.), Contemporary approaches to cognition (pp. 258-290). Cambridge, MA: Harvard University Press.

19.

Bryk

A. S.

Raudenbush

S. W.

(1992). Hierarchical linear models. Newbury Park, CA: SAGE.

20.

Clotfelter

C. T.

Steuerle

C. E.

(1981). Charitable contributions. In Aaron

Pechman

(Eds.), How taxes affect economic behavior (pp. 403-437). Washington, DC: Brookings.

21.

Dempster

A. P.

Laird

N. M.

Rubin

D. B.

(1977). Maximum likelihood from incomplete data via the EM algorithm. Journal of the Royal Statistical Society. Series B (Methodological), 39(1), 1-38.

22.

Dovidio

Piliavin

Gaertner

Schroeder

Clark

III . (1991). The arousal: Cost-reward model and the process of intervention: A review of the evidence. Prosocial behavior, 12, 86-118.

23.

Dutton

Dukerich

Harquail

(1994). Organizational images and member identification. Administrative Science Quarterly, 39, 239-263.

24.

Foote

(1951). Identification as the basis for a theory of motivation. American Sociological Review, 16(1), 14-21.

25.

Gittell

Tibaldi

(2006). Charitable giving in the U.S. states. Nonprofit and Voluntary Sector Quarterly, 35, 721-736.

26.

Healy

(2004). Altruism as an organizational problem: The case of organ procurement. American Sociological Review, 69, 387-404.

27.

Hibbert

Horne

(1996) Giving to charity: questioning the donor decision process. Journal of Consumer Marketing, 13, 4-13.

28.

Hodgkinson

V. A.

Weitzman

M. S.

(1996). Dimensions of the independent sector: A statistical profile. Washington, DC: Independent Sector.

29.

Huseman

Hatfield

Miles

(1987). A new perspective on equity theory: The equity sensitivity construct. Academy of Management Review, 12, 222-234.

30.

James

R. N.

III . (2008). Health, wealth, and charitable estate planning: A longitudinal examination of testamentary charitable giving plans. Nonprofit and Voluntary Sector Quarterly, 38, 1026-1043.

31.

James

R. N.

III . (2009). The myth of the coming charitable estate windfall. American Review of Public Administration, 39, 661-674.

32.

James

R. N

III Sharpe

D. L.

(2007). The nature and causes of the U-shaped charitable giving profile. Nonprofit and Voluntary Sector Quarterly, 36, 218-238.

33.

Kaplan

A. E.

Hayes

M. J.

(1993). What we know about women as donors: Women as donors, women as philanthropists [special issue] New Directions for Philanthropic Fundraising, 2, 5-20.

34.

Karylowski

(1982). Two types of altruistic behavior: Doing good to feel good or to make the other feel good. In Derlega

V. J.

Grzelak

(Eds.), Cooperation and helping behavior, theories and research (397-413). New York, NY Academic Press.

35.

Krebs

D. L.

(1970). Altruism: An examination of the concept and a review of the literature. Psychological Bulletin, 73, 258.

36.

McPherson

J. M.

(1981). A dynamic model of voluntary affiliation. Social Forces, 59, 705-728.

37.

Mesch

D. J.

Rooney

P. M.

Steinberg

K. S.

Denton

(2006). The effects of race, gender, and marital status on giving and volunteering in Indiana. Nonprofit and Voluntary Sector Quarterly, 35, 565-587.

38.

Micceri

(1989). The unicorn, the normal curve, and other improbable creatures. Psychological Bulletin, 105, 156-166.

39.

Musick

M. A.

Wilson

(2008). Volunteers: A social profile. Bloomington, IN: Indiana University Press.

40.

Nathan

(1996). The “nonprofitization movement” as a form of devolution. In Burlingame

D. F.

Diaz

W.A.

Ilchman

W.F.

& Associtates, Capacity for change? The nonprofit world in the age of devolution (pp. 23-55). Indianapolis, IN: Indiana University Center on Philanthropy.

41.

NCRP. (2003). Giving at work. Washington, DC: National Committee for Responsive Philanthropy.

42.

Osborne

J. W.

(2000). Advantages of hierarchical linear modeling. Practical Assessment, Research & Evaluation, 7(1). Retrieved from http://pareonline.net/getvn.asp?v=7&n=1

43.

Ostrower

(1997). Why the wealthy give: The culture of elite philanthropy: Princeton, NJ: Princeton Univristy Press.

44.

Paarlberg

L. E.

Gen

(2009). Exploring the determinants of nonprofit coproduction of public service delivery. American Review of Public Administration, 39, 391-408.

45.

Romney-Alexander

(2002). Payroll giving in the UK: Donor incentives and influences on giving behaviour. International Journal of Nonprofit and Voluntary Sector Marketing, 7(1), 84-92.

46.

Rooney

P.M.

Steinberg

K.S.

Schervish

P.G.

(2001). A methodological comparison of giving surveys: Indiana as a test case. Nonprofit and Voluntary Sector Quarterly, 30, 551-568.

47.

Rushton

(1982). Social learning theory and the development of prosocial behavior. The development of prosocial behavior, 77-105.

48.

Salamon

L. M.

Elliott

O. V.

(2002). The tools of government : A guide to the new governance. Oxford; UK: Oxford University Press.

49.

Sargeant

Jay

(2004) Fundraising management: Analysis, planning and control. London, UK: Routledge.

50.

Schlegelmilch

B. B.

Love

B. A.

Diamantopoulos

(1997) Responses to different charity appeals: The impact of donor characteristics on the amount of donation. European Journal of Marketing, 31, 548-560.

51.

Schervish

P. G.

Havens

J. J.

(1995). Do the poor pay more: Is the U-shaped curve correct? Nonprofit and Voluntary Sector Quarterly, 24(1), 79-90.

52.

Schervish

Havens

(1997). Social participation and charitable giving: A multivariate analysis. Voluntas: International Journal of Voluntary and Nonprofit Organizations, 8, 235-260.

53.

Schervish

P. G.

Havens

J. J.

(1998). Money and magnanimity: New findings on the distribution of income, wealth, and philanthropy. Nonprofit Management and Leadership, 8, 421-434.

54.

Schwartz

S. H.

(1977). Normative influences on altruism. Advances in experimental social psychology, 10, 221-279.

55.

Smith

C. A.

Organ

D. W.

Near

J. P.

(1983). Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology, 68, 655-663.

56.

Straub

J. D.

(2003). Fundraising and crowd—out of charitable contributions: New evidence from contributions to public radio. College Station, TX: Texas A&M University.

57.

Triver

R. L.

(1971). The evolution of reciprocal altruism. Quarterly Review of Biology, 46, 35-57.

58.

Turner

(1978). The role and the person. American Journal of Sociology, 84(1), 1-23.

59.

U.S. Bureau of Labor Statistics. (2008). Labor force statistics from the current population survey. Retrieved from http://stats.bls.gov/cps/#data

60.

Valor

(2006). Why do managers give? Applying pro-social behaviour theory to understand firm giving. International Review on Public and Nonprofit Marketing, 3(1), 17-28.

61.

Van Slyke

(2003). The mythology of privatization in contracting for social services. Public Administration Review, 63, 296-315.

62.

Wang

Fan

Willson

V. L.

(1996). Effects of nonnormal data on parameter estimates and fit indices for a model with latent and manifest variables: An empirical study. Structural Equation Modeling, 3, 228-247.

63.

Weisbrod

B. A.

(2001). The future of the nonprofit sector: Its entwining with private enterprise and government. In Ott

J. S.

(Ed.), The nature of the nonprofit sector (pp. 399-410). Boulder, CO: Westview.