Profiles in workplace giving: a cluster analysis of ‘types’ of givers within a public university

View author details View Less
  • 1 University of Florida, , USA
Full Access
Get eTOC alerts
Rights and permissions Cite this article

This study examines public employees’ donations to a workplace giving campaign at a large public university in the south-east of the United States. First, we employed logistic regression to predict the likelihood of donating through workplace giving programmes using a sample of employees at a large public university (N = 11,726). Second, we estimated an ordinary least squares regression to identify the significant predictors of donation value with a subsample of employee donors (n=1,832). Third, we developed donor profiles (for example, clusters) of employee benefactors using K-medoids clustering. Factors such as sex, age, education and salary were significant predictors of both being a donor and the donation amount. Additionally, employment duration was significantly related to being a donor and the donation amount, while job classification only predicted being a donor. Employee donors fell into five distinct clusters. These findings contribute to our knowledge of workplace giving campaigns and can be used to develop strategic marketing campaigns.

Abstract

This study examines public employees’ donations to a workplace giving campaign at a large public university in the south-east of the United States. First, we employed logistic regression to predict the likelihood of donating through workplace giving programmes using a sample of employees at a large public university (N = 11,726). Second, we estimated an ordinary least squares regression to identify the significant predictors of donation value with a subsample of employee donors (n=1,832). Third, we developed donor profiles (for example, clusters) of employee benefactors using K-medoids clustering. Factors such as sex, age, education and salary were significant predictors of both being a donor and the donation amount. Additionally, employment duration was significantly related to being a donor and the donation amount, while job classification only predicted being a donor. Employee donors fell into five distinct clusters. These findings contribute to our knowledge of workplace giving campaigns and can be used to develop strategic marketing campaigns.

Introduction

Federated fundraising campaigns raise funds from a specific group – typically a workplace – to distribute to a variety of charitable causes. These campaigns are a common fundraising mechanism to support a broad range of non-profit organisations and their targeted missions (Hager et al, 2002). These programmes are particularly popular in the United States (US) (Shaker and Christensen, 2019), although similar fundraising efforts have been identified in other countries, including Australia and the United Kingdom (Romney‐Alexander, 2002; Haski‐Leventhal, 2013; Wiepking and Handy, 2015). Workplace and community giving programmes are embodiments of federated campaigns, raising more than $3.5 billion in donations annually (Hrywna, 2015). Found in all occupational sectors, federated campaigns set a social expectation and provide an opportunity for employees to act philanthropically. A growing interest in corporate social responsibility and workplace giving from millennial workers showcases an opportunity for more donations (Millennial Impact Project, 2015). However, a shift to contract and part-time employment in the US, independent fundraising campaigns through online platforms, and the rise of convenient tools emulating automated deductions through the payroll, indicate that workplace giving and federated campaigns may be at a turning point (Shaker et al, 2017). Thus, examining giving in the workplace is critical if organisations seek to continue receiving and increasing substantial gifts through federated campaigns.

In the higher education sector, community engagement is increasingly seen as a critical component of education leadership. Workplace giving campaigns serve as strategic public relations tools to elevate public universities’ philanthropic profile while also strengthening community relations (Driscoll, 2009; Hall, 2009; McNall et al, 2009; Scull and Cuthill, 2010). However, traditional community engagement has few structural rewards; and even institutions with strong outreach and community missions – including land-grant universities – grapple with forging partnerships that engage the community (Jones et al, 2018; Harder, 2019). Institutions can leverage employee charitable giving as a benefit to the community, which in turn may improve the community’s perception of the university.

While hundreds of studies have documented philanthropic behaviours in general (Bekkers and Wiepking, 2011a; 2011b; Wiepking and Bekkers, 2012; Christensen et al 2016), there is less research with regards to workplace giving, and more specifically workplace giving at state-run institutions of higher education. We seek to contribute to the literature by addressing two overarching research questions: (1) What are the main characteristics of employee donors? and (2) What are the salient variations in giving among different groups of employee donors? We adopted a quantitative case study research design. Our case was a workplace giving campaign conducted annually at a large public university in the south-east of the US. Donations were given to a variety of non-profit organisations, most of which operated in the community in which the university was located. The campaign was conducted internally but the fiscal agent was the local United Way, an organisation that conducts many federated campaigns throughout the US (Nesbit et al, 2012). We use quantitative analyses to determine significant factors influencing workplace giving and develop homogenous profiles (that is, clusters) of employee givers. In fact, the key contribution of this research is the identification of cluster groups that can inform target marketing strategies.

Literature review

Here we describe the extant literature concerning the characteristics of donors, focusing (where possible) on research related to workplace giving campaigns. To narrow the range of possible characteristics to study, we develop a framework influenced by the work of by Nesbit et al (2012). Specifically, we focused on the micro- and meso-level characteristics that may influence workplace giving in higher education settings. Micro-level characteristics refer to attributes possessed and exhibited by an individual within organisations and include demographic characteristics, preferences and individual motivation to support charities and causes (Nesbit et al, 2012). These characteristics are the characteristics typically studied in workplace giving research (Rimes et al, 2019). Meso-level characteristics encompass the relation between individual donors and their organisations. This can include employees’ organisational identification and commitment to the institution, duration of employment and position (Healy, 2004; Nesbit et al, 2012; Shaker et al, 2014; Christensen et al, 2016). Note, for the purpose of this study, we are not addressing macro-level factors, which, in this case, emphasise the organisational field – or system of organisations – that influences giving.

Micro-level characteristics

Research on the characteristics that may influence workplace giving is sparse; however, extensive research exists on individual characteristics and giving in general philanthropic research (Bekkers and Wiepking, 2011b; Haski-Leventhal, 2013; Borden et al, 2014; Shaker et al, 2016). In general, the literature indicates that charitable behaviours are influenced by many characteristics, including age, gender, salary, religiosity, educational attainment and volunteering (Gittell and Tebaldi, 2006).

Age is typically a characteristic that is significantly correlated with philanthropic giving. Research indicates that increasing age is associated with increased giving (Shaker et al, 2016) and, as donors reach old age, giving declines (Bekkers and Wiepking, 2011b). Studies examining age and giving in the workplace have mixed results, with some studies finding significance (Bekkers and Wiepking, 2011b; Haski-Leventhal, 2013; Borden et al, 2014) and others finding no statistically significant effect (Agypt et al, 2012).

Typically, philanthropic research indicates that women are more likely to donate, while men donate in larger amounts. These findings, however, are highly dependent on the type of organisation receiving the donations (Wiepking and Bekkers, 2012). In workplace giving studies, the relationship between gender and giving is complicated and yields mixed results; some studies have found a statistically significant association (Haski-Leventhal, 2013; Leslie et al, 2013), while others have found no relationship (Agypt et al, 2012; Nesbit et al, 2012).

While not perfectly correlated, salary and income are positively associated with charitable giving behaviours (Romney-Alexander, 2002; Knight, 2004; Gittell and Tebaldi, 2006; Agypt et al, 2012). Income operates primarily as a micro-level characteristic, while, at the same time, salary is often perceived as part of the relational interface between organisation and employee (that is, is implicated as a meso-level characteristic). Scholars posit that the psychological benefits of giving in the workplace, rooted within organisational commitment, may be associated with income and giving (Christensen et al, 2016). In other words, an employee supports the organisation because of their identification with the employer (solicitor), in turn giving the employee (donor) a sense of satisfaction.

Individuals with higher educational attainment tend to donate more to non-profits and charitable causes than those with lower educational attainment (Brown, 2005). Nesbit et al (2012) suggest that an individual giving in the academic workplace is influenced by employee stratification in hierarchically arranged settings, particularly within the staff and faculty strata. Layers of stratum exist as a function of the relationship between socioeconomic status factors, including education and income. Employees in higher strata, reflected by compensation and education, are more likely to give than lower strata employees (Nesbit et al, 2012). Future work, though, is needed to assess potential confounders influencing this relationship (for example, duration of employment and job socialisation).

In the context of workplace giving and micro-level characteristics, research has yielded mixed results, even among commonly studied individual characteristics found in philanthropic giving research (Osili et al, 2011; Nesbit et al, 2012; Haski-Leventhal, 2013; Christensen et al, 2016). Shaker et al (2014) postulate that the variation between workplace and non-workplace giving may be due to the workplace environment’s mediating effects.

Meso-level indicators

A growing theme in the literature seeks to investigate the relationship between organisations and their donors (for example, Schervish and Havens, 1997; Straub, 2003; Healy, 2004). This includes employee–employer dyadic relationships. On a broad scale, workplace giving at the meso level is contextualised by an employee’s organisational commitment and organisational identification (Nesbit et al, 2012; Borden et al, 2014). Organisational identification (or ‘attachment’) argues that ‘the more a member accepts and internalises an organisation’s values, the stronger the identification’ (Nesbit et al, 2012). Closely related, organisational commitment encompasses the bond between the individual and the organisation’s purpose. Organisational commitment and identification are operationalised by characteristics such as employment duration, job satisfaction and position type.

Previous findings regarding meso-level characteristics in the workplace indicate multiple positively associated factors that may indicate giving. For instance, in one study, employment duration, salary status (for example, hourly versus salary or exempt versus non-exempt) and academic rank were significant predictors for donating (Agypt et al, 2011). However, a follow-up study of the same variables indicated that employment duration was the only significant predictor and that tenured and salaried employees donated less than their counterparts (Agypt et al, 2012). Employment duration was found to be significant in a workplace-specific study (Knight, 2004). Alumni status may also represent a strong sense of organisational commitment, identification, or both. In the Knight (2004) study, institutional alumni contributed more than their non-alumni counterparts. However, Borden et al (2014) suggest that multifaceted nuances among alumni and non-alumni employees may be confounding predictors of giving propensity. Shaker et al (2016: 93) surmise that ‘organisational identification may or may not engender giving’. In short, limited empirical findings regarding workplace giving and relational influences reveal mixed results (Knight, 2004; Agypt et al, 2012; Borden et al, 2014; Shaker et al, 2014). In this study, we continue other scholars’ work in connecting donor data and employee records. However, we also (a) focus on a giving campaign that funds external non-profits, and (b) develop clusters of participants to better understand the profiles of workplace giving.

Methods

The case

This case study examines public employees’ donations to a workplace giving campaign conducted annually at a large public university in the south-east of the US. This workplace campaign – a well-established effort dating back to the mid-1990s – is held each autumn over four weeks. The campaign raises approximately $1 million annually on behalf of charities serving the county’s surrounding communities. In 2019, the campaign’s donations totalled $993,564 and represented a 15 per cent employee participation rate. This campaign is considered typical for this size of institution.

The campaign is divided into two focus times: a leadership campaign and a general campaign. During the two-week leadership campaign period, the campaign solicits donations from employees who either have given $500 or more in the previous year’s campaign and/or those with an annual salary of more than $50,000. This campaign strategy allows employee benefactors to make early pledges and donations to charities. Individuals eligible to donate during the leadership campaign season are included in a donation tier-based membership programme (or ‘supporters’ circle’). This donor loyalty programme classifies benefactors as follows: employees whose donations total more than $10,000 become ‘senior leadership’ members; donations between $5,000 and $9,999 are granted the platinum-level membership; contributions between $2,500 and $4,999, the diamond level; donations between $1,500 and $2,499, the emerald level; those who donated between $1,000 and $1,499, the gold level; and those whose contributions fell to between $750 and $999, the silver level. Leadership campaign giving typically accounts for over 75 per cent of total employee donations to the campaign. In 2019, leadership gifts totalled $768,682 through donations from 494 employees.

During the two-week general campaign period, all employees at the higher education institution can participate in the campaign. They receive postcards with detailed information about ways to give and weekly emails with links to the online giving site.

All employees are eligible to participate in the campaign by making donations via payroll deduction, cash, check, credit and stock contribution. Employee donors have the ability to designate some or all the amount donated to one or more participating agencies. If the employee donor wishes to donate to the campaign but does not allocate the donation to a specific agency (that is, undesignated funds), then the funds are divided among all agencies in proportion to the amount received in the campaign.

The campaign has grown from 39 beneficiary organisations in its first year to 100 agencies in 2020. Recipient agencies represent various causes, addressing health, social service, diversity, relief, development or environmental issues of local importance. To participate in the campaign, each agency must fulfil a set of requirements and criteria, including being a registered non-profit organisation and recognised for a substantial programme in health and human services directly benefiting the community within the county region and surroundings.

Research design

We conducted a quantitative study to examine factors associated with workplace giving. The research design and methodological procedures were guided by Creswell and Clark (2017) and Ivankova et al (2006) and approved by the Institutional Review Board at the University of Florida.

Data and sample

Data were collected from public records and restricted-campaign donor records at a large public university in the south-east of the US. Employee record data were obtained from the academic institution’s Office of Public Affairs through a formal public records request. The open-access, copying and inspection of governmental bodies’ public records – including records non-exempt from disclosure generated at public higher education institutions – are afforded under statutory laws (Florida Statutes, Chapter 119). We were granted access to employee data from years 2013 to 2017. The dataset included the following variables: employee identification number, academic unit, annual salary, sex, number of years employed at the institution and highest degree obtained.

Further, anonymised internal donor-level data collected from the workplace giving campaigns in the years 2013 to 2017 were provided by campaign managers in the Office of Community Relations. The internal campaign data collection included the following variables: employee identification number, academic unit, total donation amount (in US dollars), total designated and undesignated amounts, and whether or not the donor was a leadership campaign member. We merged public records and internal donor-level data using employees’ identification numbers. Our analyses used the cross-sectional, full sample of 11,726 employees in 2017, and the subsample was delimited to those employees who donated through a workplace giving campaign in that same year. The subsample of donors in 2017 included 1,832 employees. Table 1 presents the sample profile for all employees and by employee donors and non-donors in 2017.

Table 1:

Sample profile by whether employee donated or not

VariablesTotalDonated in 2017
(N = 11,726)Yes (n = 1,832)No (n = 9,894)
Number%Number%Number%
Sex
Male5,38245.90%78442.80%4,59846.50%
Female6,34454.10%1,04857.20%5,29653.50%
Education level
High school or less2,00017.06%35919.60%1,64116.60%
Some college8277.05%1025.60%7257.30%
Technical training8597.33%1085.90%7517.60%
Bachelor’s degree2,46120.99%40522.10%2,05620.80%
Master’s degree1,91916.37%31117.00%1,60816.30%
Doctoral degree3,66031.21%54729.90%3,11331.50%
Job classification
Faculty3,87733.06%61433.50%3,26333.00%
Exempt3,80432.44%76541.80%3,03930.70%
Non-exempt3,05526.05%45124.60%2,60426.30%
Temporary9908.44%20.10%98810.00%
Leadership programme member
Yes1,28270.00%-
No55030.00%-
Mean (Std. dev.)MedianMean (Std. dev.)MedianMean (Std. dev.)Median
Age47.3 (11.30)47.053.1 (10.2)5445.9 (12.4)46.0
Duration of employment12.1 (9.4)9.417.8 (9.5)17.111.4 (8.6)7.9
Annual income$80,912.64 ($71,171.00)$57,128.23$96,754.70 ($75,137.00)$72,648.81$77,984.85 ($70,025.00)$55,000.00
Amount donated$428.27 ($1,017.00)$130.00-

Data analysis

We undertook a three-step quantitative analysis approach. First, we performed logistic regression to predict the likelihood of donating through workplace giving programmes. Second, we estimated an ordinary least square (OLS) regression to identify donation value predictors. Third, using K-medoids clustering, we segmented donors into homogenous groups (that is, clusters) based on micro- and meso-level characteristics and workplace giving variables. We must note that the first two steps were informed by the analysis strategy used in Shaker et al (2016).

The following characteristics were included in our study: sex, age, educational attainment and salary. The characteristics controlled in our analyses were duration of employment, job classification (or position type) and membership in the leadership programme. Workplace giving indicators, used as dependent variables in the regression analyses, included whether the employee donated through the campaign, the amount donated in US dollars and the logarithm transformation of the amount donated. Table 2 provides the complete list of variables, descriptions, data sources and operationalisations for the quantitative analyses.

Table 2:

Description of variables and measurement

VariablesDescriptionData sourceOperationalisation
Micro-level characteristics
SexEmployee’s sex, as reported by the Human Resource Office.Public recordsBinary variable. Indicator = 1 if the employee is female, 0 otherwise.
AgeEmployee’s age in 2017. Calculated based on birth year. Age = 2017 birth year.Public recordsContinuous measure.
Educational attainmentHighest education level in 2017, per official Human Resource Office records. Raw data included the following categories: (a) none, (b) high school graduate or equivalent, (c) some college, (d) technical school, (e) specialist degree, (f) two-year degree, (g) Bachelor’s-level degree, (h) some graduate school, (i) Master’s-level degree, (j) doctorate (academic and professional) and (k) post-doctorate. For the analyses, we merged categories and created the following levels: high school or less (a and b); some college (c); technical training (d, e and f); Bachelor’s degree (g and h); master’s degree (i); and doctoral degree (j and k). Master’s degrees encompass MS, MBA, MA, MFA and other equivalents. The doctoral degree category includes employees who hold at least one of the following: PhD, EdD, juris doctor degrees or MD/DO degrees.Public recordsDummy variable. Reference group = doctoral degree.
SalaryAnnual individual income in 2017, as reported by the Human Resource Office at the organisation under study.Public recordsContinuous measure
Log annual incomeThe log is taken for all non-zero values of income. All values less than or equal to 0 remain at 0.Public recordsContinuous measure
Meso-level characteristics
Job classificationEmployee’s position type in 2017 as reported by the Human Resource Office.Public recordsDummy variable. Reference group = faculty.
FacultyFaculty members including tenured-track, non-tenured track, as well as clinical and research tracks.
ExemptNon-faculty employees who do not receive overtime pay nor do they qualify for the minimum wage.
Non-exemptNon-faculty employees who are entitled to earn at least the federal minimum wage and qualify for overtime pay.
TemporaryNon-faculty employees hired as contingent workers, contract employees, consultants or seasonal workers.
Leadership programme membershipThe leadership programme is an internal campaign strategy where an employee earns membership if they (i) gave more than $500 in the preceding year (that is, 2016) or (ii) did not give in the preceding year but have an individual salary greater than $50,000 per year.Campaign dataBinary variable, where leadership programme member = 1 and non-member = 0.
Duration of employmentNumber of years employed at the organisation under study.Public recordsContinuous measure
Workplace giving variables
Employee donorWhether the employee donated through the workplace giving programme, regardless of the amount of the donation.Campaign dataBinary variable, where employee donor = 1 and non-donor = 0
Annual donationAnnual donation through the workplace giving programme in US dollars.Campaign dataContinuous measure
Log annual donationThe log is taken for all non-zero values of donation.Campaign dataContinuous measure

Using cross-sectional data from the year 2017, we explored the determinants of being an employee donor via logistic regression using SAS 9.4 software. We conducted the analyses using the full sample of employees in 2017: 11,726 (donors = 1,832; non-donors = 9,894). Logistic regression, a statistical technique widely used in the social sciences, is appropriate for analysing dichotomous categorical variables (in this case, the likelihood of being a donor versus a non-donor) and directly models the conditional probability of binary outcomes (Hilbe, 2009). In our model, the dependent variable was constructed based on whether an employee donated at least $1 through the federated campaign in 2017, where ‘1’ indicates the employee was a donor in 2017, and ‘0’ indicates otherwise. The independent variables included characteristics such as sex, age, education, log-transformed salary, duration of employment, job classification and whether the employee donated in the previous year. Table 2 provides the level of measurement for the variables used in the logistic regression analysis. We must note that we performed the log transformation for the salary variable to adjust for highly skewed data (Wooldridge, 2016).

Second, we examined the factors that significantly influenced the dollar amount donated to charities through the workplace campaign using OLS regression. OLS, one of the most popular statistical techniques used in the social sciences, is used to predict values of continuous response variables (Fox, 2015); in this case, the dollar amount donated through the 2017 workplace campaign. The sample for this analysis included employee benefactors in the year 2017, totalling 1,832 employees. SAS 9.4 software was used to estimate the model. The amount donated through the campaign, the department variable, was log-transformed to achieving approximate homoscedasticity, an assumption in OLS regression (Brooks, 2019). In terms of independent variables, our model included the following characteristics: sex, age, education level and log-transformed salary. Further, duration of employment, job classification and log-transformed amount donated in 2016 were included in the linear model. Finally, we must note that leadership campaign membership was not included in the regression due to multicollinearity (that is, VIF > 5) (Kalnins, 2018).

Third, we performed mixed data cluster analysis to identify salient groups (or clusters) of employees who donated to local charities through payroll deduction in 2017. In the absence of class labels, clustering is a useful machine learning tool that allows for the identification of subgroups from a set of individuals (or subjects) based on shared characteristics or similar attributes, and where members of a particular cluster are more alike to each other than to members of other clusters (Kaufman and Rousseeuw, 1987; Huang, 1998; Park and Jun, 2009).

Partitioning clustering methods1 have been applied to a number of disciplines, including the social sciences and marketing research (Han et al, 2011). Partition clustering’s main objective is to construct k partitions (or clusters) from a set of n objects (for example, a dataset), where each partition must have at least one individual, and each individual must belong to one group only. Homogenous clusters were formed based on employees’ micro- and meso-level characteristics, including sex, age, educational attainment, position type, annual salary in US dollars, employment duration and membership in the campaign leadership programme. The sample used in the clustering analyses was delimited to employee donors who contributed to charities through the workplace campaign in 2017. The analysis was conducted using R statistical software version 3.6.2., package Cluster version 2.1.0. We first estimated the average of partial dissimilarities across employee donors. Then, we identified salient clusters of employee donors using a medoid-based algorithm, namely Clustering LARge Applications (CLARA) (Kaufman and Rousseeuw, 2009). We used T-distributed Stochastic Neighbor Embedding (t-SNE) – a machine learning algorithm for dimensionality reduction and the visualisation of our highly dimensional dataset. t-SNE is a widely used, non-linear dimensionality reduction technique to project the data points onto a lower-dimensional grid (van der Maaten and Hinton, 2008, 2012; van der Maaten, 2014; Krijthe, 2015; Melit Devassy et al, 2020). Finally, the average silhouette width was utilised to determine the clustering quality.

Results

Donor characteristics (logit regression results)

Table 3 presents the logistic regression model results. As shown, all evaluated micro-level characteristics (that is, sex, age and salary) except for education were significant predictors of being a donor. Male employees were less likely to be donors than their female counterparts, where the odds ratio for males versus females was about 0.72. However, the marginal effect of gender on being an employee donor was small (0.54%). On the other hand, older employees and those with higher salaries (log-transformed) were more likely to be donors. Holding all things equal, a one-year increase in age and a dollar increase in logged salary increased the odds of donating through the federated workplace campaign by 0.03 per cent and 0.84 per cent, respectively. Notably, the effects of education on the likelihood of donating to the campaign were not significant.

Table 3:

Logistic regression analysis to predict being an employee donor

VariablesβSEOR95% CI% marginal effect
Micro-level characteristics
Sex (ref = female)
Male–0.331*0.1510.718[0.526, 0.952]–0.54
Age (in years)0.019*0.0081.019[1.004, 1.035]0.03
Education level (ref = doctoral)
High school or less0.3180.3921.375[0.629, 2.936]0.52
Technical training or some college0.2330.3861.263[0.598, 2.722]0.38
Bachelor’s degree0.3900.3341.478[0.772, 2.864]0.64
Master’s degree0.3610.2901.435[0.815, 2.545]0.59
Log annual income0.508**0.1681.661[1.208, 2.333]0.84
Meso-level characteristics
Duration of employment (in years)0.0020.0091.003[0.984, 1.020]0.00
Job classification (ref = faculty)
Exempt0.3660.2911.442[0.815, 2.559]0.62
Non-exempt0.5110.3801.668[0.788, 3.509]0.88
Temporary–2.194*0.9790.111[0.016, 0.766]–5.13
Donated in 2016 (ref = no)7.250***0.1531407.48[1043.14,1899.09]89.11
Intercept–11.250***1.970
Observations11,726
Log likelihood–947.70

Note: Estimated with 2017 year data. * p < 0.05, ** p < 0.01, *** p < 0.001.

A strong predictor in our model was having donated to the workplace campaign in the prior year. Particularly, being an employee donor in 2016 significantly increased the likelihood of being an employee donor in the subsequent year (2017), with a marginal effect of 89.11 per cent. Other workplace variables in this study (that is, job classification and employment durations) were overall not significantly related to donating through the 2017 workplace campaign, although the coefficients were positive. Only temporary workers were less likely to be donors than faculty members (that is, significant at the 5 per cent significance level).

Amount of donation by donor characteristics (OLS regression results)

Table 4 shows the results for the regression model predicting the log-transformed donation amount. As seen, micro-level characteristics, such as sex, age, education and logged salary, were significantly related to the donation amount. As age and logged salary increased, the dollar amount of the donation through the workplace giving programme increased. For every year increase in age, the amount donated increased by 0.7 per cent. Likewise, for every 1 per cent increase in logged salary, the logged donation amount increased by about 0.54 per cent, denoting that the income elasticity was 0.54. Holding other variables constant, the donations for male employees and those with educational levels of high school or less and technical training were, on average, smaller than the donations made by those in their respective comparison groups. In terms of meso-level attributes, a longer duration of employment (as measured by the number of years employed at the institution) predicted lower donation amounts. For every year increase in employment duration, the donation amount decreased by 1.6 per cent. Conversely, job classification had no significant effect on the amount donated through the federated campaign. Additionally, the lagged variable of the amount donated in 2016 (log-transformed) significantly increased the dependent variable. In other words, a 1 per cent increase in the logged amount donated in 2016 increased about 0.50 per cent of the logged amount donated in 2017, ceteris paribus.

Table 4:

OLS regression analysis for annual amount donated (log transformed)

Independent variablesβSE
Micro-level characteristics
Sex (ref = female)
Male–0.113**0.041
Age (in years)0.007***0.002
Education level (ref = doctoral degree)
High school or less–0.393***0.110
Technical training or some college–0.239*0.109
Bachelor’s degree–0.0790.093
Master’s degree–0.1190.080
Log annual income0.542***0.051
Meso-level characteristics
Duration of employment (in years)–0.016***0.002
Job classification (ref = faculty)
Exempt0.0160.081
Non-exempt–0.0790.108
Temporary0.3190.603
Log amount donated in 20160.498***0.012
Constant–3.331***0.591
Observations1832
R20.678
Adjusted R20.676
F statistic318.588*** (df=12)

Note: Estimated with 2017 year data. * p <0 .05, ** p < 0.01, *** p < 0.001.

Grouping employee donors (mixed-data cluster analysis results)

Based on the average silhouette width for the entire dataset, we identified five homogenous clusters. Table 5 presents the aggregated characteristics for all employee donors in 2017 (column 1) and each cluster (columns 2–6). Cluster 1 (n = 273), labelled ‘female faculty’, was mainly composed of female faculty members who held doctoral degrees and belonged to the leadership programme (that is, they donated more than $500.00 through the workplace campaign in the previous year or their annual salaries were greater than $50,000). The average salary and donation in the cluster were $132,175 and $718.50, respectively. Cluster 2 (n = 279), labelled ‘male staff members’, contained male, non-faculty donors with the educational attainment of a Bachelor’s or Master’s degree. The majority in this cluster were members of the leadership programme. The mean salary and donation for this cluster were $87,502 and $337.80, respectively. Cluster 3 (n = 439), labelled ‘non-exempt employees’ represents a group of non-exempt staff members. The average salary and donation were $38,755 and $90.82, respectively. Most of the employees in this cluster were not identified as leadership programme members. More than 50 per cent of the donors in this group were female and had an education level of high school or less. Cluster 4 (n = 357), labelled ‘male faculty’, contained male faculty members with the educational attainment of a doctoral degree. Of the members in this group, 92 per cent were members of the leadership programme. The average salary and donation were $174,894 and $912.10, respectively. This group included a significant number of medical faculty members and/or administrators. Finally, cluster 5 (n = 484), labelled ‘female staff members’, was composed of female, exempt employees with membership of the leadership programme. Of the donors in this group, 46 per cent reported having completed Bachelor’s degrees, and 28 per cent indicated holding Master’s-level degrees. This group’s average salary and donation were $77,081 and $265.90, respectively. Finally, Figure 1 provides a visual representation of the resulting five clusters (that is, donor profiles) in a two-dimensional space, where x and y axes denote the first and second dimensions, respectively. As shown, individuals within a particular cluster (colour coded) were located in similar areas on the plane; this was mainly the case for clusters 1 and 4. Overall, this projection supports the relevancy of the clustering solution.

Table 5:

Composition of sample profile within clusters

VariablesTotal (n = 1,832)Cluster 1 (n = 273), ‘female faculty’Cluster 2 (n = 279), ‘male staff members’Cluster 3 (n = 439), ‘non-exempt employees’Cluster 4 (n = 357), ‘male faculty’Cluster 5 (n = 484), ‘female staff members’
Freq.%Freq.%Freq.%Freq.%Freq.%Freq.%
Micro-level characteristics
Sex
Male78442.8%00.0%279100.0%14833.7%357100.0%00.0%
Female104857.2%273100.0%00.0%29166.3%00.0%484100.0%
Education level
High school or less35919.6%00.0%259.0%29066.1%10.3%438.9%
Some college1025.6%10.4%176.1%5111.6%30.8%306.2%
Technical training1085.9%20.7%207.2%5011.4%10.3%357.2%
Bachelor’s degree40522.1%41.5%12745.5%388.7%41.1%23247.9%
Master’s degree31117.0%5620.5%8129.0%102.3%359.8%12926.7%
Doctoral degree54729.9%21076.9%93.2%00.0%31387.7%153.1%
Average age53.155.1951.3651.8759.549.19
Average annual income$96,754.70$132,175.00$87,502.00$38,755.00$174,894.00$77,081.00
Median annual income$72,648.81$110,948.00$72,685.00$38,462.00$154,018.00$68,651.00
Meso-level characteristics
Average duration of employment17.817.6515.91118.6922.8815.92
Job classification
Faculty61433.5%26296.0%00.0%10.2%34696.9%51.0%
Exempt76541.8%82.9%26494.6%245.5%51.4%46495.9%
Non-exempt45124.6%31.1%134.7%41494.3%61.7%153.1%
Temporary20.1%00.0%20.7%00.0%00.0%00.0%
Workplace giving
Average annual donation$428.27$718.50$337.80$90.82$912.10$265.90
Median annual donation$130.00$270.00$130.00$52.00$270.00$130.00
Leadership campaign member
Yes128270.0%25593.4%23885.3%347.7%32992.2%42688.0%
No55030.0%186.6%4114.7%40592.3%287.8%5812.0%

Note: Analysis performed using 2017 data.

Figure 1:
Figure 1:

Visualisation of the five-cluster solution (that is, clusters of donors) in a two-dimensional space.

Citation: Voluntary Sector Review 2022; 10.1332/204080521X16359450188693

Robustness check

To test for robustness, we used employee-level data from 2016 (n = 11,568) and 2015 (n = 11,151) and estimated two additional logistic regression models to predict the likelihood of giving to charities through workplace giving programmes. The results (direction and significance) were consistent across all three years evaluated (model 1, shown in Table 3, and models 2 and 3 in Table 6): employees with higher salaries, those with longer employment duration, females, older employees, full-time staff (as compared with faculty members) and employees with educational attainment level of high school or less, or a Bachelor’s or Master’s degree (in comparison with those who held doctoral degrees) were more likely to be donors. Similarly, we employed cross-sectional, donor-level data from 2016 (n = 1,791) and 2015 (n = 1,785) and performed OLS regression to examine the determinants of the amount contributed to charities through the workplace giving campaign. Significant predictors and their direction concerning the outcome variable were consistent for all three years evaluated in this study (see Tables 4 and 7). Older employees, those with higher salaries and females were more prone to make larger monetary donations than their counterparts. Employees with longer employment duration and employees with lower educational attainment tended to make smaller monetary donations, on average, than their counterparts. Finally, following the same procedures and analysis described in the methodology section, we clustered employee donors in 2016 and 2015. After performing K-medoid via the CLARA algorithm, we evaluated the clustering quality results. Based on the S ̅(k), we obtained the same five-cluster solution (that is, salient groups of donors) for 2015 and 2016, where the characteristics of groups were proportionally similar and consistent with the results obtained in 2017. Finally, while prior studies have provided evidence of the robustness of K-medoid and its better performance over K-means as it pertains to outlier sensitivity, the K-medoid algorithm suffers from some limitations involving a high computational complexity and its sensitivity to the initial selection of medoids (Ng and Han, 2002; Park and Jun, 2009; Paterlini et al, 2011). To address these limitations, we compared the clustering results with and without an early selection of medoids (see Pérez-Ortega et al, 2017). The clustering solutions obtained from these alternatives were consistent.

Table 6:

Logistic regression analysis to predict charitable giving

VariablesModel 2aModel 3b
βSEOR95% CIβSEOR95% CI
Gender (ref = female)
Male–0.346***0.0580.708[0.632, 0.793]–0.391***0.0590.674[0.603, 0.759]
Age (in years)0.033***0.0031.033[1.027, 1.039]0.031***0.0031.033[1.025, 1.038]
Education level (ref = doctoral)
High school or less0.553***0.1481.739[1.301, 2.324]0.563***0.1491.756[1.310, 2.353]
Technical training or some college0.2140.1471.239[0.929, 1.653]0.2460.1481.279[0.957, 1.709]
Bachelor’s degree0.434***0.1261.543[1.206, 1.974]0.477***0.1271.612[1.258, 2.066]
Master’s degree0.328***0.1081.389[1.123, 1.718]0.377***0.1091.458[1.178, 1.805]
Job tenure (in years)0.05***0.0031.051[1.045, 1.058]0.052***0.0031.054[1.047, 1.061]
Job classification (ref = faculty)
Exempt0.657***0.111.93[1.555, 2.394]0.686***0.111.933[1.557, 2.401]
Non-exempt0.457***0.1461.579[1.187, 2.102]0.488***0.1461.549[1.164, 2.063]
Temporary–3.196***0.7180.041[0.010, 0.167]–1.231***0.7180.043[0.011, 0.176]
Log annual income0.69***0.0651.993[1.756, 2.263]0.758***0.0641.984[1.749, 2.250]
Intercept–12.05***0.761–11.84***0.751
Observations11,56811,151
% predictions76.276.3

Notes: a estimated with 2016 year data, b estimated with 2015 year data. * p < 0.05, ** p < 0.01, *** p < 0.001.

Table 7:

OLS regression analysis for annual amount donated (log transformed)

Independent variablesModel 2aModel 3b
βSEβSE
Gender (ref = female)
Male-0.156**0.055-0.163**0.056
Age (in years)0.016***0.0030.013***0.003
Education level (ref = doctoral degree)
High school or less-0.786***0.147-0.683***0.151
Technical training or some college-0.449**0.146-0.405**0.149
Bachelor’s degree-0.2430.124-0.1820.127
Master’s degree-0.247*0.107-0.237*0.109
Job tenure (in years)-0.011***0.003-0.009**0.003
Job classification (ref = faculty)
Exempt0.0370.1090.0240.110
Non-exempt and temporary-0.1200.1430.020.146
Log annual income0.880***0.0601.078***0.066
Constant-5.208***0.697-7.364***0.760
Observations1,7911,785
R20.3810.389
Adjusted R20.3770.386
F statistic109.41*** (df=10)112.97*** (df=10)

Notes: a estimated with 2016 year data, b estimated with 2015 year data. * p < 0.05, ** p < 0.01, *** p < 0.001.

Discussion and recommendations

This case study found five distinct clusters of workplace givers in a large public university. Each cluster profile illustrates distinct micro-level characteristics of the group members, such as age, gender, salary and education, as well as meso-level indicators such as years of employment. This section discusses the findings from this study in light of previous research, identifies next steps for research and outlines the study’s limitations.

Cluster discussion

The primary contribution of this study is the identification of clusters of givers who behave very differently and, thus, would likely respond differently to campaign marketing strategies. For example, the ‘male faculty’ cluster was primarily composed of male faculty members with doctoral degrees. It had the highest average donation amount as well as the highest mean age and longest mean employment duration. The ‘non-exempt employees’ cluster had the lowest average donation amount and a lower average age and employment duration than the ‘male faculty’ cluster. This group was comprised of both women (about two thirds) and men (about one third), the majority of whom had a high school education or less and worked in non-exempt job positions. Falling between the these two groups in donation amount was the ‘female faculty’ cluster, consisting of mostly female faculty members with a doctoral or Master’s degree. The average age for this group fell between the other two, while employment duration was similar to the ‘non-exempt employees’ cluster. As a percentage of salary, the average gift for ‘female faculty’ was slightly higher than other clusters (0.54% as compared with 0.52 per cent for ‘male faculty’ and 0.23 per cent for ‘non-exempt employees’).

These clusters suggest that employees are behaving differently but in relatively predictable ways. Male and female faculty alike – the highest paid of all university employees, except senior leaders – donate at higher rates, and those who donate typically qualify for the leadership programme. It is possible that these donors are purposefully giving at levels that ensure they are part of the leadership programme, as membership of that leadership programme might signal to their department chair, deans or senior administrators that they are committed to the institution above and beyond the job requirements.

There are other possible explanations as to why faculty are giving. For example, female faculty tend to value environmental condition and department climate. Indeed, scholars have found that, after tenure, female faculty tend to become more focused on contributing to the department than their peers (August and Waltman, 2004). In this study, the ‘female faculty’ cluster’s average annual salary was $132,175. This salary level suggests that most of those who are giving are at the associate or full professor level, and many may also have administrative responsibilities. Thus, it is possible that giving to this campaign is less about a pre-tenure signal of commitment to the university and, instead, part of the general desire August and Waltman (2004) found female faculty had to create an inclusive and supportive university climate.

The pay disparity between the two faculty clusters is also noteworthy. The ‘female faculty’ in this study were paid an average of $42,719 less than ‘male faculty’. This is not surprising, as there is pay disparity (Johnson and Taylor, 2019). However, it is likely that many of the participants in the ‘male faculty’ cluster were located in the medical school, a school that is likely to pay higher average salaries than the rest of the campus. While there is a pay disparity and general lack of gender parity for faculty in medical schools (Kuo et al, 2019; Bernardi et al, 2020), there are also logical reasons why faculty in a medical school would earn more than faculty in non-medical schools, regardless of gender. Thus, these findings cannot be interpreted as a clear indication of pay disparity.

The lowest-paid employees are also giving in interesting ways. Their average gift is small – $90.82 per year, only 10 per cent of the average gift of ‘male faculty’. Their gifts also represent a smaller percentage of their annual income – 0.23 per cent as compared with 0.54 per cent for the ‘female faculty’ cluster. This is intriguing. Many studies have documented that low-income donors give a higher percentage of their income than middle- and high-income donors, typically giving 3–3.5 per cent of their income as compared with the 1 percent given by wealthier individuals (Walker and Pharoah, 2002; Wiepking, 2007; Brooks, 2008; Bennett, 2012; 2018). At first glance, our data do not match these more general findings. However, it is likely that the gifts to the university campaign represent only part of the participants’ overall charitable giving and, thus, if we were to consider all of the participants’ giving, their patterns may be more in line with the general research. Regardless, it is important to note these gifts are important beyond the dollar amount. A 90-dollar average annual gift is more valuable a gift to a non-exempt employee living off of $38,755 than the $912 average annual gift is to a ‘male faculty’ member living off of $174,894. Their reasons for giving are likely also different, and further qualitative research should be performed to parcel out the varying motivations of these groups.

General discussion

Salary and age were positively associated with giving. This is consistent with other literature, which suggests that age – particularly in mid-career and early retirement – is positively associated with giving (Shaker et al, 2016). Education was not positively associated with giving but positively associated with the amount donated. These findings are consistent with research suggesting that salary and educational attainment are correlated positively with giving (Romney-Alexander, 2002; Knight, 2004; Brown, 2005; Gittell and Tebaldi, 2006; Agypt et al, 2012).

The research on gender and giving is less conclusive (Agypt et al, 2012; Nesbit et al, 2012; Wiepking and Bekkers, 2012; Haski-Leventhal, 2013; Leslie et al, 2013); however, this study suggests that, for the most part, women of various demographics are more likely to give and give more. For example, four of the five clusters identified in the sample were comprised of 99 or 100 per cent one gender, while only one cluster (the ‘non-exempt employees’ cluster) had a mix of males (35 per cent) and females (65 per cent); and three of the four homogeneous groups were female.

Employment duration was positively correlated with workplace giving. This supports prior research (Agypt et al, 2011). Interestingly, though, the amount donated through the workplace giving programme decreased as employment duration increased. This might be due to one of two factors. First, it is possible that workplace giving was specified when the employee signed up, and this amount was never adjusted. Second, it is possible that one motivation for giving was recognition by the employer, a motivation that may decrease over time in academic personnel who obtain a permanent tenure status.

Future research

Future research into workplace giving should take into consideration the various profiles of workplace givers in at least four ways. First, scholars should perform a latent class analysis (Magidson and Vermunt, 2002). Scholars could then use the latent class as a variable in direct analysis to see what impact this group membership has on giving donation and amount.

Second, scholars should identify the motivations for giving that drive each of the distinct clusters. This would likely require moving beyond traditional quantitative approaches and including more qualitative methods such as interviews and focus groups. This would be particularly intriguing for investigating the ‘non-exempt employees’ cluster as previously described.

Third, scholars should identify whether such profiles can be used to create targeted marketing campaigns and whether these campaigns might be more successful than generic, organisation-wide campaigns. Our study indicated that these profiles align with both gender and type of position and, therefore, a targeted marketing approach might indeed be successful. The finding of the previously suggested qualitative research on motivations would be useful in creating such marketing campaigns. However, scholars should also take note that these clusters may form and behave differently at different institutions, varying based on factors such as organisational culture and climate (Charoensap-Kelly, 2017).

Fourth, scholars should examine the value and purpose of workplace giving. While it is less attractive in terms of total dollars raised, lower-salaried employees’ and minoritised individuals’ giving patterns and experiences are particularly interesting and warrant future research. For example, we know minoritised groups are more likely to give through workplace giving campaigns (Rimes et al, 2019). Payroll deduction structures that facilitate small donation amounts may be an important community-building tool for the university. This point is not to be taken lightly as it may have far-reaching effects. Exposure to workplace giving campaigns increases the likelihood a household will also give directly to non-profits; and households who give at work also give directly to non-profits at higher rates (Rimes et al, 2019).

Limitations

As with any study, there are a number of limitations to this one. First, this case study analysed giving at one large public university in the south-east of the US. A case study ‘n’ of one can generate valuable data (Donmoyer, 1990); however, generalisability is limited. This study has implications for other universities, particularly universities with sizable medical and law schools if the employee demographics are likely to be reasonably similar. This study is less generalisable in non-academic settings because differing employment demographics would lead to different workplace giving profiles. However, other settings could identify workplace giving profiles, and based on the findings of this study, identification of such groups would generate insights that could be used in marketing the campaign. Second, we exclusively examined micro- and meso-level characteristics. A deeper evaluation of macro-level characteristics (for example, perceptions of and attitudes towards the charities involved in workplace giving and the causes that they represent) and their interactions with micro- and meso-level factors is warranted and should be a compelling avenue for future research in the context of workplace giving.

Conclusions

Workplace giving is at a crossroads (Shaker et al, 2017). On the one hand, social and technological changes suggest that such giving may be declining. On the other hand, workplace giving has the potential to rebound given the increased generational emphasis on corporate social responsibility (Millennial Impact Project, 2015). Workplace giving at public universities is particularly interesting because of the distinct categories of employees, categories which we found led to distinct clusters of types of giver. The findings from this study suggest targeted marketing possibilities for the campaign studied, possibilities that may be of interest to campaigns at other similar universities. This includes strategies for marketing to both current and future donors. Workplace giving at public universities is also interesting because the relationship between the university and the community (often described as the ‘town and gown’ divide). While somewhat transactional in nature, such campaigns have the potential to improve this relationship for the betterment of both parties.

Note

1

Clustering techniques are generally divided into the following categories: hierarchical techniques, density-based techniques, grid-based methods, model-based methods and partitioning methods. In our research, we employed a partitioning-based method (Swarndeep Saket, and Pandya, 2016).

Funding

This work was supported through an internal grant from the Department of Family, Youth and Community Sciences at the University of Florida.

Conflict of interest

The authors declare that there is no conflict of interest.

References

  • Agypt, B., Christensen, R.K. and Nesbit, R. (2011) To Give or Not to Give: Longitudinal Predictors of Donation Decisions in Workplace Giving Campaigns, Toronto, Ontario, Canada, paper presented at the ARNOVA annual meeting, November.

    • Search Google Scholar
    • Export Citation
  • Agypt, B., Christensen, R.K. and Nesbit, R. (2012) A tale of two charitable campaigns: longitudinal analysis of employee giving at a public university, Nonprofit and Voluntary Sector Quarterly, 41(5): 80225. doi: 10.1177/0899764011418836

    • Search Google Scholar
    • Export Citation
  • August, L. and Waltman, J. (2004). Culture, climate, and contribution: career satisfaction among female faculty, Research in Higher Education, 45(2): 17792. doi: 10.1023/B:RIHE.0000015694.14358.ed

    • Search Google Scholar
    • Export Citation
  • Bekkers, R. and Wiepking, P. (2011a) A literature review of empirical studies of philanthropy: eight mechanisms that drive charitable giving, Nonprofit and Voluntary Sector Quarterly, 40(5): 92473. doi: 10.1177/0899764010380927

    • Search Google Scholar
    • Export Citation
  • Bekkers, R. and Wiepking, P. (2011b) Who gives? A literature review of predictors of charitable giving. Part one: religion, education, age and socialization, Voluntary Sector Review, 2(3): 33765. doi: 10.1332/204080511X6087712

    • Search Google Scholar
    • Export Citation
  • Bennett, R. (2012) Why urban poor donate: a study of low income charitable giving in London, Nonprofit and Voluntary Sector Marketing, 41(5): 87091. doi: 10.1177/0899764011419518

    • Search Google Scholar
    • Export Citation
  • Bennett, R. (2018) Financial charity giving behaviour of the working poor: an empirical investigation, Journal of Marketing Management, 34(17–18): 1587607. doi: 10.1080/0267257X.2018.1512516

    • Search Google Scholar
    • Export Citation
  • Bernardi, K., Shah, P., Lyons, N.B., Olavarria, O.A., Alawadi, Z.M., Leal, I.M., Holihan, J.L., Bass, B.L., Jakey, C.E., Kao, L.S. et al. (2020) Perceptions on gender disparity in surgery and surgical leadership: a multicenter mixed methods study, Surgery, 167(4): 74350. doi: 10.1016/j.surg.2019.12.004

    • Search Google Scholar
    • Export Citation
  • Borden, V.M., Shaker, G.G. and Kienker, B.L. (2014) The impact of alumni status on institutional giving by faculty and staff, Research in Higher Education, 55(2): 196217. doi: 10.1007/s11162-013-9318-3

    • Search Google Scholar
    • Export Citation
  • Brooks, C. (2019) Classical linear regression model assumptions and diagnostic tests, in Introductory Econometrics for Finance, Cambridge: Cambridge University Press, pp 182245.

    • Search Google Scholar
    • Export Citation
  • Brown, E. (2005) College, social capital, and charitable giving, in A. Brooks (ed) Gifts of Time and Money in America’s Communities, Lanham, MD: Rowman & Littlefield, pp 185204.

    • Search Google Scholar
    • Export Citation
  • Charoensap-Kelly, P. (2017) Crafting faculty and staff fundraising campaign: predictors of giving, donor motivations and effective strategies, Journal of Education Advancement and Marketing, 2(1): 2037.

    • Search Google Scholar
    • Export Citation
  • Christensen, R.K., Nesbit, R. and Agypt, B. (2016) To give or not to give: employee responses to workplace giving campaigns over time, Nonprofit and Voluntary Sector Quarterly, 45(6): 125875. doi: 10.1177/0899764015619704

    • Search Google Scholar
    • Export Citation
  • Creswell, J.W. and Clark, V.L.P. (2017) Designing and Conducting Mixed Methods Research, Thousand Oaks, CA: Sage Publications.

  • Donmoyer, R. (1990) Generalizability and the single‐case study, in E. Eisner and A. Peskhkin (eds) Qualitative Research on Education: The Continuing Debate, New York, NY: Teachers College Press, pp 175200.

    • Search Google Scholar
    • Export Citation
  • Driscoll, A. (2009) Carnegie’s new community engagement classification: affirming higher education’s role in community, New Directions for Higher Education, 2009(147): 512. doi: 10.1002/he.353

    • Search Google Scholar
    • Export Citation
  • Florida Statutes (Chapter 119) [Online] http://www.leg.state.fl.us/Statutes/index.

  • Fox, J. (2015) Applied Regression Analysis and Generalized Linear Models, Thousand Oaks, CA: Sage Publications.

  • Gittell, R. and Tebaldi, E. (2006) Charitable giving: factors influencing giving in U.S. states, Nonprofit and Voluntary Sector Quarterly, 35(4): 72136. doi: 10.1177/0899764006289768

    • Search Google Scholar
    • Export Citation
  • Hager, M., Rooney, P. and Pollak, T. (2002) How fundraising is carried out in US nonprofit organizations, International Journal of Nonprofit and Voluntary Sector Marketing, 7(4): 31124. doi: 10.1002/nvsm.188

    • Search Google Scholar
    • Export Citation
  • Hall, B.L. (2009) Higher education, community engagement, and the public good: building the future of continuing education in Canada, Canadian Journal of University Continuing Education, 35(2): 1123. doi: 10.21225/D5BC7N

    • Search Google Scholar
    • Export Citation
  • Han, J., Kamber, M. and Pei, J. (2011) Cluster Analysis: Basic Concepts and Methods in Data Mining Concepts and Techniques, 3rd edn, Waltham, MA: Morgan Kaufmann-Elsevier.

    • Search Google Scholar
    • Export Citation
  • Harder, A. (2019) Public value and partnership: critical components of extension’s future, Journal of Extension, 57(3).

  • Haski-Leventhal, D. (2013) Employee engagement in CSR: the case of payroll giving in Australia, Corporate Social Responsibility and Environmental Management, 20(2): 11328. doi: 10.1002/csr.1287

    • Search Google Scholar
    • Export Citation
  • Healy, K. (2004) Altruism as an organizational problem: the case of organ procurement, American Sociological Review, 69(3): 387404. doi: 10.1177/000312240406900304

    • Search Google Scholar
    • Export Citation
  • Hilbe, J.M. (2009) Logistic Regression Models, Boca Raton, FL: CRC Press.

  • Hrywna, M. (2015) United Way revenue ticks down, NonProfit Times, 6 October, http://www.thenonprofittimes.com/news-articles/united-way-fundraising-up-total-revenue-down/.

    • Search Google Scholar
    • Export Citation
  • Huang, Z. (1998) Extensions to the k-means algorithm for clustering large data sets with categorical values, Data Mining and Knowledge Discovery, 2(3): 283304. doi: 10.1023/A:1009769707641

    • Search Google Scholar
    • Export Citation
  • Ivankova, N.V., Creswell, J.W. and Stick, S.L. (2006) Using mixed-methods sequential explanatory design: from theory to practice, Field methods, 18(1): 320. doi: 10.1177/1525822X05282260

    • Search Google Scholar
    • Export Citation
  • Johnson, J.A. and Taylor, B.J. (2019) Academic capitalism and faculty salary gap, Innovative Higher Education, 44(1): 2135. doi: 10.1007/s10755-018-9445-z

    • Search Google Scholar
    • Export Citation
  • Jones, J.A., Pracht, D., Simonne, E., Renfrow, K. and Hunter, C. (2018) Nonprofit partnerships in extension programming: a pilot study, Journal of Extension, 56(2): 19.

    • Search Google Scholar
    • Export Citation
  • Kalnins, A. (2018) Multicollinearity: how common factors cause type 1 errors in multivariate regression, Strategic Management Journal, 39(8): 236285. doi: 10.1002/smj.2783

    • Search Google Scholar
    • Export Citation
  • Kaufman L. and Rousseeuw, P.J. (1987) Clustering by means of medoids, in Y. Dodge (ed) Statistical Data Analysis Based on the L1 Norm and Related Methods, Basel: Springer, pp 40516.

    • Search Google Scholar
    • Export Citation
  • Kaufman, L. and Rousseeuw, P.J. (2009) Finding Groups in Data: An Introduction to Cluster Analysis (Vol. 344), Hoboken, NJ: John Wiley & Sons.

    • Search Google Scholar
    • Export Citation
  • Knight, W.E. (2004) Influences on participation in a university faculty and staff annual giving campaign, The CASE International Journal of Educational Advancement, 4(3): 22132. doi: 10.1057/palgrave.cijea.2140002

    • Search Google Scholar
    • Export Citation
  • Krijthe, J.H. (2015) Rtsne: T-distributed stochastic neighbor embedding using Barnes-Hut implementation, R package version 0.13, https://github.com/jkrijthe/Rtsne.

    • Search Google Scholar
    • Export Citation
  • Kuo, I.C., Levine, R.B., Gauda, E.B., Bodurtha, J., Clements, J., Fivush, B. and Ishii, L. (2019) Identifying gender disparities and barriers to measuring the status of female faculty: the experience of a large school of medicine, Journal of Women’s Health, 28(11): 156975.

    • Search Google Scholar
    • Export Citation
  • Leslie, L.M., Snyder, M. and Glomb, T.M. (2013) Who gives? Multilevel effects of gender and ethnicity on workplace charitable giving, Journal of Applied Psychology, 98(1): 4962. doi: 10.1037/a0029943

    • Search Google Scholar
    • Export Citation
  • Magidson, J. and Vermunt, J.K. (2002) Latent class modeling for clustering: a comparison with K-means, Canadian Journal of Marketing Research, 20(1): 3744.

    • Search Google Scholar
    • Export Citation
  • McNall, M., Reed, C.S., Brown, R. and Allen, A. (2009) Brokering community–university engagement, Innovative Higher Education, 33(5): 31731. doi: 10.1007/s10755-008-9086-8

    • Search Google Scholar
    • Export Citation
  • Melit Devassy, B., George, S. and Nussbaum, P. (2020) Unsupervised clustering of hyperspectral paper data using t-SNE, Journal of Imaging, 6(5): 29. doi: 10.3390/jimaging6050029

    • Search Google Scholar
    • Export Citation
  • Millennial Impact Project (2015) Cause, influence and the next generation: the 2015 millennial report, http://www.themillennialimpact.com/research/.

    • Search Google Scholar
    • Export Citation
  • Nesbit, R., Christensen, R.K. and Gossett, L. (2012) Charitable giving in the public workplace: a framework to understand employees’ philanthropic performance, Public Performance & Management Review, 35(3): 44975. doi: 10.2753/PMR1530-9576350303

    • Search Google Scholar
    • Export Citation
  • Ng, R.T. and Han, J. (2002) CLARANS: a method for clustering objects for spatial data mining, IEEE Transactions on Knowledge and Data Engineering, 14(5): 100316. doi: 10.1109/TKDE.2002.1033770

    • Search Google Scholar
    • Export Citation
  • Osili, U.O., Hirt, D.E. and Raghaven, S. (2011) Charitable giving inside and outside the workplace: the role of individual and firm characteristics, International Journal of Nonprofit and Voluntary Sector Marketing, 16(4): 393408. doi: 10.1002/nvsm.435

    • Search Google Scholar
    • Export Citation
  • Park, H.S. and Jun, C.H. (2009) A simple and fast algorithm for K-medoids clustering, Expert Systems with Applications, 36(2): 333641. doi: 10.1016/j.eswa.2008.01.039

    • Search Google Scholar
    • Export Citation
  • Paterlini, A.A., Nascimento, M.A. and Traina Jr, C. (2011) Using pivots to speed-up k-medoids clustering, Journal of Information and Data Management, 2(2): 22136.

    • Search Google Scholar
    • Export Citation
  • Pérez-Ortega, J., Almanza-Ortega, N.N., Adams-López, J., González-Gárcia, M., Mexicano, A., Saenz-Sánchez, S. and Rodríguez-Lelis, J.M. (2017) Improving the efficiency of the K-medoids clustering algorithm by getting initial medoids, in Á. Rocha, A. Correia, H. Adeli, L. Reis and S. Costanzo (eds) Recent Advances in Information Systems and Technologies, Advances in Intelligent Systems and Computing, vol 569, Cham: Springer.

    • Search Google Scholar
    • Export Citation
  • Rimes, H., Nesbit, R. and Christensen, R.K. (2019) Giving at work: exploring connections between workplace giving campaigns and patterns of household charitable giving in the USA, VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 30(4): 82840. doi: 10.1007/s11266-019-00125-4

    • Search Google Scholar
    • Export Citation
  • Romney-Alexander, D. (2002) Payroll giving in the UK: donor incentives and influences on giving behavior, International Journal of Nonprofit and Voluntary Sector Marketing, 7(1): 8492. doi: 10.1002/nvsm.169

    • Search Google Scholar
    • Export Citation
  • Schervish, P.G. and Havens, J. (1997) Social participation and charitable giving: a multivariate analysis, VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 8(3): 23560. doi: 10.1007/BF02354199

    • Search Google Scholar
    • Export Citation
  • Scull, S. and Cuthill, M. (2010) Engaged outreach: using community engagement to facilitate access to higher education for people from low socioeconomic backgrounds, Higher Education Research & Development, 29(1): 5974.

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G. and Christensen, R.K. (2019) I give at the office: a review of workplace giving research, theory, and practice, International Journal of Nonprofit Voluntary Sector Marketing, 24(1): e1628. doi: 10.1002/nvsm.1628

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G., Borden, V.M., Kienker, B.L. (2016) Workplace giving in universities: a U.S. case study at Indiana University, Nonprofit and Voluntary Sector Quarterly, 45(1): 87111. doi: 10.1177/0899764014565468

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G., Christensen, R.K. and Bergdoll, J.J. (2017) What works at work? Toward an integrative model examining workplace campaign strategies, Nonprofit Management & Leadership, 28(1): 2546. doi: 10.1002/nml.21270

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G., Kienker, B.L. and Borden, V.M. (2014) The ecology of internal workplace giving at Indiana University: a case study of academic and non-academic staff campus campaign fundraising, International Journal of Nonprofit and Voluntary Sector Marketing, 19(4): 26276. doi: 10.1002/nvsm.1501

    • Search Google Scholar
    • Export Citation
  • 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.

    • Search Google Scholar
    • Export Citation
  • Swarndeep Saket, J. and Pandya, D.S. (2016) An overview of partitioning algorithms in clustering techniques, International Journal of Advanced Research in Computer Engineering & Technology (IJARCET), 5(6): 194346.

    • Search Google Scholar
    • Export Citation
  • Van der Maaten, L. (2014) Accelerating t-SNE using tree-based algorithms, The Journal of Machine Learning Research, 15(1): 322145.

  • Van der Maaten, L. and Hinton, G. (2008) Visualizing data using t-SNE, Journal of Machine Learning Research, 9: 2579605. https://www.jmlr.org/papers/volume9/vandermaaten08a/vandermaaten08a.pdf?fbclid=IwA

    • Search Google Scholar
    • Export Citation
  • Van der Maaten, L. and Hinton, G. (2012) Visualizing non-metric similarities in multiple maps, Machine learning, 87(1): 3355. doi: 10.1007/s10994-011-5273-4

    • Search Google Scholar
    • Export Citation
  • Walker, C. and Pharoah, C. (2002) A Lot to Give: Trends in Charitable Giving for the 21st Century, London: Hodder and Stoughton.

  • Wiepking, P. (2007) The philanthropic poor: in search of explanations for the relative generosity of lower income households, Voluntas, 18(4): 33958. doi: 10.1007/s11266-007-9049-1

    • Search Google Scholar
    • Export Citation
  • Wiepking, P. and Bekkers, R. (2012) Who gives? A literature review of predictors of charitable giving. Part two: gender, marital status, income, and wealth, Voluntary Sector Review, 3(2): 21745. doi: 10.1332/204080512X649379

    • Search Google Scholar
    • Export Citation
  • Wiepking, P. and Handy, F. (eds) (2015) The Palgrave Handbook of Global Philanthropy, Houndmills: Palgrave Macmillan.

  • Wooldridge, J.M. (2016) Introductory Econometrics: A Modern Approach, Toronto, Canada: Nelson Education.

  • View in gallery

    Visualisation of the five-cluster solution (that is, clusters of donors) in a two-dimensional space.

  • Agypt, B., Christensen, R.K. and Nesbit, R. (2011) To Give or Not to Give: Longitudinal Predictors of Donation Decisions in Workplace Giving Campaigns, Toronto, Ontario, Canada, paper presented at the ARNOVA annual meeting, November.

    • Search Google Scholar
    • Export Citation
  • Agypt, B., Christensen, R.K. and Nesbit, R. (2012) A tale of two charitable campaigns: longitudinal analysis of employee giving at a public university, Nonprofit and Voluntary Sector Quarterly, 41(5): 80225. doi: 10.1177/0899764011418836

    • Search Google Scholar
    • Export Citation
  • August, L. and Waltman, J. (2004). Culture, climate, and contribution: career satisfaction among female faculty, Research in Higher Education, 45(2): 17792. doi: 10.1023/B:RIHE.0000015694.14358.ed

    • Search Google Scholar
    • Export Citation
  • Bekkers, R. and Wiepking, P. (2011a) A literature review of empirical studies of philanthropy: eight mechanisms that drive charitable giving, Nonprofit and Voluntary Sector Quarterly, 40(5): 92473. doi: 10.1177/0899764010380927

    • Search Google Scholar
    • Export Citation
  • Bekkers, R. and Wiepking, P. (2011b) Who gives? A literature review of predictors of charitable giving. Part one: religion, education, age and socialization, Voluntary Sector Review, 2(3): 33765. doi: 10.1332/204080511X6087712

    • Search Google Scholar
    • Export Citation
  • Bennett, R. (2012) Why urban poor donate: a study of low income charitable giving in London, Nonprofit and Voluntary Sector Marketing, 41(5): 87091. doi: 10.1177/0899764011419518

    • Search Google Scholar
    • Export Citation
  • Bennett, R. (2018) Financial charity giving behaviour of the working poor: an empirical investigation, Journal of Marketing Management, 34(17–18): 1587607. doi: 10.1080/0267257X.2018.1512516

    • Search Google Scholar
    • Export Citation
  • Bernardi, K., Shah, P., Lyons, N.B., Olavarria, O.A., Alawadi, Z.M., Leal, I.M., Holihan, J.L., Bass, B.L., Jakey, C.E., Kao, L.S. et al. (2020) Perceptions on gender disparity in surgery and surgical leadership: a multicenter mixed methods study, Surgery, 167(4): 74350. doi: 10.1016/j.surg.2019.12.004

    • Search Google Scholar
    • Export Citation
  • Borden, V.M., Shaker, G.G. and Kienker, B.L. (2014) The impact of alumni status on institutional giving by faculty and staff, Research in Higher Education, 55(2): 196217. doi: 10.1007/s11162-013-9318-3

    • Search Google Scholar
    • Export Citation
  • Brooks, C. (2019) Classical linear regression model assumptions and diagnostic tests, in Introductory Econometrics for Finance, Cambridge: Cambridge University Press, pp 182245.

    • Search Google Scholar
    • Export Citation
  • Brown, E. (2005) College, social capital, and charitable giving, in A. Brooks (ed) Gifts of Time and Money in America’s Communities, Lanham, MD: Rowman & Littlefield, pp 185204.

    • Search Google Scholar
    • Export Citation
  • Charoensap-Kelly, P. (2017) Crafting faculty and staff fundraising campaign: predictors of giving, donor motivations and effective strategies, Journal of Education Advancement and Marketing, 2(1): 2037.

    • Search Google Scholar
    • Export Citation
  • Christensen, R.K., Nesbit, R. and Agypt, B. (2016) To give or not to give: employee responses to workplace giving campaigns over time, Nonprofit and Voluntary Sector Quarterly, 45(6): 125875. doi: 10.1177/0899764015619704

    • Search Google Scholar
    • Export Citation
  • Creswell, J.W. and Clark, V.L.P. (2017) Designing and Conducting Mixed Methods Research, Thousand Oaks, CA: Sage Publications.

  • Donmoyer, R. (1990) Generalizability and the single‐case study, in E. Eisner and A. Peskhkin (eds) Qualitative Research on Education: The Continuing Debate, New York, NY: Teachers College Press, pp 175200.

    • Search Google Scholar
    • Export Citation
  • Driscoll, A. (2009) Carnegie’s new community engagement classification: affirming higher education’s role in community, New Directions for Higher Education, 2009(147): 512. doi: 10.1002/he.353

    • Search Google Scholar
    • Export Citation
  • Florida Statutes (Chapter 119) [Online] http://www.leg.state.fl.us/Statutes/index.

  • Fox, J. (2015) Applied Regression Analysis and Generalized Linear Models, Thousand Oaks, CA: Sage Publications.

  • Gittell, R. and Tebaldi, E. (2006) Charitable giving: factors influencing giving in U.S. states, Nonprofit and Voluntary Sector Quarterly, 35(4): 72136. doi: 10.1177/0899764006289768

    • Search Google Scholar
    • Export Citation
  • Hager, M., Rooney, P. and Pollak, T. (2002) How fundraising is carried out in US nonprofit organizations, International Journal of Nonprofit and Voluntary Sector Marketing, 7(4): 31124. doi: 10.1002/nvsm.188

    • Search Google Scholar
    • Export Citation
  • Hall, B.L. (2009) Higher education, community engagement, and the public good: building the future of continuing education in Canada, Canadian Journal of University Continuing Education, 35(2): 1123. doi: 10.21225/D5BC7N

    • Search Google Scholar
    • Export Citation
  • Han, J., Kamber, M. and Pei, J. (2011) Cluster Analysis: Basic Concepts and Methods in Data Mining Concepts and Techniques, 3rd edn, Waltham, MA: Morgan Kaufmann-Elsevier.

    • Search Google Scholar
    • Export Citation
  • Harder, A. (2019) Public value and partnership: critical components of extension’s future, Journal of Extension, 57(3).

  • Haski-Leventhal, D. (2013) Employee engagement in CSR: the case of payroll giving in Australia, Corporate Social Responsibility and Environmental Management, 20(2): 11328. doi: 10.1002/csr.1287

    • Search Google Scholar
    • Export Citation
  • Healy, K. (2004) Altruism as an organizational problem: the case of organ procurement, American Sociological Review, 69(3): 387404. doi: 10.1177/000312240406900304

    • Search Google Scholar
    • Export Citation
  • Hilbe, J.M. (2009) Logistic Regression Models, Boca Raton, FL: CRC Press.

  • Hrywna, M. (2015) United Way revenue ticks down, NonProfit Times, 6 October, http://www.thenonprofittimes.com/news-articles/united-way-fundraising-up-total-revenue-down/.

    • Search Google Scholar
    • Export Citation
  • Huang, Z. (1998) Extensions to the k-means algorithm for clustering large data sets with categorical values, Data Mining and Knowledge Discovery, 2(3): 283304. doi: 10.1023/A:1009769707641

    • Search Google Scholar
    • Export Citation
  • Ivankova, N.V., Creswell, J.W. and Stick, S.L. (2006) Using mixed-methods sequential explanatory design: from theory to practice, Field methods, 18(1): 320. doi: 10.1177/1525822X05282260

    • Search Google Scholar
    • Export Citation
  • Johnson, J.A. and Taylor, B.J. (2019) Academic capitalism and faculty salary gap, Innovative Higher Education, 44(1): 2135. doi: 10.1007/s10755-018-9445-z

    • Search Google Scholar
    • Export Citation
  • Jones, J.A., Pracht, D., Simonne, E., Renfrow, K. and Hunter, C. (2018) Nonprofit partnerships in extension programming: a pilot study, Journal of Extension, 56(2): 19.

    • Search Google Scholar
    • Export Citation
  • Kalnins, A. (2018) Multicollinearity: how common factors cause type 1 errors in multivariate regression, Strategic Management Journal, 39(8): 236285. doi: 10.1002/smj.2783

    • Search Google Scholar
    • Export Citation
  • Kaufman L. and Rousseeuw, P.J. (1987) Clustering by means of medoids, in Y. Dodge (ed) Statistical Data Analysis Based on the L1 Norm and Related Methods, Basel: Springer, pp 40516.

    • Search Google Scholar
    • Export Citation
  • Kaufman, L. and Rousseeuw, P.J. (2009) Finding Groups in Data: An Introduction to Cluster Analysis (Vol. 344), Hoboken, NJ: John Wiley & Sons.

    • Search Google Scholar
    • Export Citation
  • Knight, W.E. (2004) Influences on participation in a university faculty and staff annual giving campaign, The CASE International Journal of Educational Advancement, 4(3): 22132. doi: 10.1057/palgrave.cijea.2140002

    • Search Google Scholar
    • Export Citation
  • Krijthe, J.H. (2015) Rtsne: T-distributed stochastic neighbor embedding using Barnes-Hut implementation, R package version 0.13, https://github.com/jkrijthe/Rtsne.

    • Search Google Scholar
    • Export Citation
  • Kuo, I.C., Levine, R.B., Gauda, E.B., Bodurtha, J., Clements, J., Fivush, B. and Ishii, L. (2019) Identifying gender disparities and barriers to measuring the status of female faculty: the experience of a large school of medicine, Journal of Women’s Health, 28(11): 156975.

    • Search Google Scholar
    • Export Citation
  • Leslie, L.M., Snyder, M. and Glomb, T.M. (2013) Who gives? Multilevel effects of gender and ethnicity on workplace charitable giving, Journal of Applied Psychology, 98(1): 4962. doi: 10.1037/a0029943

    • Search Google Scholar
    • Export Citation
  • Magidson, J. and Vermunt, J.K. (2002) Latent class modeling for clustering: a comparison with K-means, Canadian Journal of Marketing Research, 20(1): 3744.

    • Search Google Scholar
    • Export Citation
  • McNall, M., Reed, C.S., Brown, R. and Allen, A. (2009) Brokering community–university engagement, Innovative Higher Education, 33(5): 31731. doi: 10.1007/s10755-008-9086-8

    • Search Google Scholar
    • Export Citation
  • Melit Devassy, B., George, S. and Nussbaum, P. (2020) Unsupervised clustering of hyperspectral paper data using t-SNE, Journal of Imaging, 6(5): 29. doi: 10.3390/jimaging6050029

    • Search Google Scholar
    • Export Citation
  • Millennial Impact Project (2015) Cause, influence and the next generation: the 2015 millennial report, http://www.themillennialimpact.com/research/.

    • Search Google Scholar
    • Export Citation
  • Nesbit, R., Christensen, R.K. and Gossett, L. (2012) Charitable giving in the public workplace: a framework to understand employees’ philanthropic performance, Public Performance & Management Review, 35(3): 44975. doi: 10.2753/PMR1530-9576350303

    • Search Google Scholar
    • Export Citation
  • Ng, R.T. and Han, J. (2002) CLARANS: a method for clustering objects for spatial data mining, IEEE Transactions on Knowledge and Data Engineering, 14(5): 100316. doi: 10.1109/TKDE.2002.1033770

    • Search Google Scholar
    • Export Citation
  • Osili, U.O., Hirt, D.E. and Raghaven, S. (2011) Charitable giving inside and outside the workplace: the role of individual and firm characteristics, International Journal of Nonprofit and Voluntary Sector Marketing, 16(4): 393408. doi: 10.1002/nvsm.435

    • Search Google Scholar
    • Export Citation
  • Park, H.S. and Jun, C.H. (2009) A simple and fast algorithm for K-medoids clustering, Expert Systems with Applications, 36(2): 333641. doi: 10.1016/j.eswa.2008.01.039

    • Search Google Scholar
    • Export Citation
  • Paterlini, A.A., Nascimento, M.A. and Traina Jr, C. (2011) Using pivots to speed-up k-medoids clustering, Journal of Information and Data Management, 2(2): 22136.

    • Search Google Scholar
    • Export Citation
  • Pérez-Ortega, J., Almanza-Ortega, N.N., Adams-López, J., González-Gárcia, M., Mexicano, A., Saenz-Sánchez, S. and Rodríguez-Lelis, J.M. (2017) Improving the efficiency of the K-medoids clustering algorithm by getting initial medoids, in Á. Rocha, A. Correia, H. Adeli, L. Reis and S. Costanzo (eds) Recent Advances in Information Systems and Technologies, Advances in Intelligent Systems and Computing, vol 569, Cham: Springer.

    • Search Google Scholar
    • Export Citation
  • Rimes, H., Nesbit, R. and Christensen, R.K. (2019) Giving at work: exploring connections between workplace giving campaigns and patterns of household charitable giving in the USA, VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 30(4): 82840. doi: 10.1007/s11266-019-00125-4

    • Search Google Scholar
    • Export Citation
  • Romney-Alexander, D. (2002) Payroll giving in the UK: donor incentives and influences on giving behavior, International Journal of Nonprofit and Voluntary Sector Marketing, 7(1): 8492. doi: 10.1002/nvsm.169

    • Search Google Scholar
    • Export Citation
  • Schervish, P.G. and Havens, J. (1997) Social participation and charitable giving: a multivariate analysis, VOLUNTAS: International Journal of Voluntary and Nonprofit Organizations, 8(3): 23560. doi: 10.1007/BF02354199

    • Search Google Scholar
    • Export Citation
  • Scull, S. and Cuthill, M. (2010) Engaged outreach: using community engagement to facilitate access to higher education for people from low socioeconomic backgrounds, Higher Education Research & Development, 29(1): 5974.

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G. and Christensen, R.K. (2019) I give at the office: a review of workplace giving research, theory, and practice, International Journal of Nonprofit Voluntary Sector Marketing, 24(1): e1628. doi: 10.1002/nvsm.1628

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G., Borden, V.M., Kienker, B.L. (2016) Workplace giving in universities: a U.S. case study at Indiana University, Nonprofit and Voluntary Sector Quarterly, 45(1): 87111. doi: 10.1177/0899764014565468

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G., Christensen, R.K. and Bergdoll, J.J. (2017) What works at work? Toward an integrative model examining workplace campaign strategies, Nonprofit Management & Leadership, 28(1): 2546. doi: 10.1002/nml.21270

    • Search Google Scholar
    • Export Citation
  • Shaker, G.G., Kienker, B.L. and Borden, V.M. (2014) The ecology of internal workplace giving at Indiana University: a case study of academic and non-academic staff campus campaign fundraising, International Journal of Nonprofit and Voluntary Sector Marketing, 19(4): 26276. doi: 10.1002/nvsm.1501

    • Search Google Scholar
    • Export Citation
  • 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.

    • Search Google Scholar
    • Export Citation
  • Swarndeep Saket, J. and Pandya, D.S. (2016) An overview of partitioning algorithms in clustering techniques, International Journal of Advanced Research in Computer Engineering & Technology (IJARCET), 5(6): 194346.

    • Search Google Scholar
    • Export Citation
  • Van der Maaten, L. (2014) Accelerating t-SNE using tree-based algorithms, The Journal of Machine Learning Research, 15(1): 322145.

  • Van der Maaten, L. and Hinton, G. (2008) Visualizing data using t-SNE, Journal of Machine Learning Research, 9: 2579605. https://www.jmlr.org/papers/volume9/vandermaaten08a/vandermaaten08a.pdf?fbclid=IwA

    • Search Google Scholar
    • Export Citation
  • Van der Maaten, L. and Hinton, G. (2012) Visualizing non-metric similarities in multiple maps, Machine learning, 87(1): 3355. doi: 10.1007/s10994-011-5273-4

    • Search Google Scholar
    • Export Citation
  • Walker, C. and Pharoah, C. (2002) A Lot to Give: Trends in Charitable Giving for the 21st Century, London: Hodder and Stoughton.

  • Wiepking, P. (2007) The philanthropic poor: in search of explanations for the relative generosity of lower income households, Voluntas, 18(4): 33958. doi: 10.1007/s11266-007-9049-1

    • Search Google Scholar
    • Export Citation
  • Wiepking, P. and Bekkers, R. (2012) Who gives? A literature review of predictors of charitable giving. Part two: gender, marital status, income, and wealth, Voluntary Sector Review, 3(2): 21745. doi: 10.1332/204080512X649379

    • Search Google Scholar
    • Export Citation
  • Wiepking, P. and Handy, F. (eds) (2015) The Palgrave Handbook of Global Philanthropy, Houndmills: Palgrave Macmillan.

  • Wooldridge, J.M. (2016) Introductory Econometrics: A Modern Approach, Toronto, Canada: Nelson Education.

Content Metrics

May 2022 onwards Past Year Past 30 Days
Abstract Views 1 1 0
Full Text Views 9 9 8
PDF Downloads 7 7 6

Altmetrics

Dimensions