Abstract

The purpose of the study was to explain why people make donations to athletic programs in an effort to increase the effectiveness of athletic fund raising and enrich the fundraising literature with a theoretical framework in donor behavior. Based on a thoughough review of literature, the Giving to Athletics Model (GAM) was developed and tested. The survey research method was utilized to study donors of the Florida State University athletic program. The structural equation modeling (SEM) technique and the Maximum Likelihood estimation procedure indicated several revisions on the Giving to Athletics Model (GAM). The revised model showed that Involvement with Athletics and Emotional Motivation are the two most significant factors that explain giving to athletic programs. Finally, Experience with Sports and Attendance of Athletics were the two factors with an indirect effect on Donations to Athletics.

Introduction

The financial problems resulting from the increasing cost of recruiting, declining enrollment, reduced student financial support through student government, inflation, and Title IX, have lead athletic departments to other sources of revenue such as fund raising. Most athletic departments rely heavily on voluntary donations made by friends and alumni in an effort to balance their budgets. Though athletic fund raising constitutes an essential and significant practice for running college athletics, limited information exists to provide guidance to athletic fund raisers. Little has been written and even less research has been done about fund raising in the athletic domain. There is no elaborated theoretical base that explains giving behavior as well as fund raising and its management. Fund raising practitioners have written about their personal experiences but no effort has been made to provide generalizations that will help in developing fund raising theories. In particular, the most basic questions have not been answered yet by current research: why do people make donations to athletic programs? What factors motivate them to contribute to athletics? The majority of the fund raising and the athletic fund raising literature have been focused on variables such as demographics and attitudes toward giving. There is no study in the literature that includes other factors and no effort has been made to combine these variables in a logical chain of relationships that may explain why people make donations to particular causes/programs.

The purpose of this study was to develop and test a model that may explain donations to athletic programs in an effort to increase the effectiveness of athletic fund raising.

This study was partially exploratory and partially confirmatory. It explored factors and relationships among these factors that have not been studied before in fund raising. Moreover, it investigated factors and relationships that have been researched in the past and attempted to confirm previous findings.

The Giving to Athletics Model (GAM)

Based on the thorough review of literature, a model was developed (figure 1) and proposed to be studied. The model adapted and combined theories from the marketing literature, and results of fund raising studies in an effort to explain why people make contributions to athletic programs. Seven factors were identified that were expected to influence people in making a decision to donate money to athletic programs. The Giving to Athletics Model (GAM) included the following variables: Values, Socioeconomic Status (SES), Involvement with Athletics, Sport Experience, Attendance of Athletic Events, Emotional Motivation, Practical Motivation, and Donations to Athletics.

Values. Though the marketing literature suggests that values may be directly or indirectly related to consumer attitudes and behavior, studies on fund raising have not addressed the issue of whether the values of an individual donor are associated with his/her giving attitude and/or behavior. Two studies only (Yavas & Riecken, 1980; Duke, 1982) have reported that value similarity between a person’s individual values and the values of the charitable organization may affect giving behavior. Bakal (1979) has stated that “our values, beliefs, interests, hopes, fears, and other feelings and habits not only motivate us to give but also determine to which causes we give” (p.43). The marketing literature suggests that values affect directly consumers’ choice criteria, and indirectly their attitudes, intentions and purchase behavior. Thus, it was expected that values will be indirectly related to donors’ behavior and directly related to their choice criteria, represented in GAM as motivation (practical and emotional).

Demographics: income and education. One of the most frequently studied variables in fund raising literature is the role demographics play on giving. Income and education constituted the variable SES in the GAM model that represented the socioeconomic status of the individuals. The majority of the scholars have reported that income is a predictor of giving behavior and/or amount of contribution (Brittingham & Pezzullo, 1990). It is not clear whether or not education is positively related to giving to athletics. It has not been reported if the level of education has a positive or negative effect on giving behavior and specifically giving to athletics. However, Brittingham and Pezzullo’s (1990) summary of research findings on fund raising indicated that the majority of the literature has suggested education as a predictor of alumni giving.

Thus, as Figure 1 indicates an indirect relationship between SES (income and education) and giving was expected with the intervening effects of motivation (practical and emotional). SES was the direct antecedent of practical motivation and emotional motivation and indirect of giving.

With respect to income and education, the sport marketing literature has shown that both are related to sport experience and attendance of sport events. Graham (1994) found that the spectators of two tennis tournaments had an average family income of $86,218 and the majority (82%) had a baccalaureate or graduate degree. In addition, the spectators had a wide variety of recreational and/or competitive sport experiences. Because the above findings are not very strong in terms of the relationships between SES, sport experience and attendance of sport events, the GAM model proposed only one relationship (no causal) between SES and sport experience.

Sport experience. One of the recommendations for future research suggested by Smith (1989) was to study previous participation in athletics and its relationship to athletic giving. Graham studied the spectators attending the U.S. Men’s Clay Court Championship in 1992 and 1994. He found that the majority of the spectators (89.6% and 85.6%) were former or current players. Moreover, they participated in at least four other types of sports (golf, swimming, jogging, and biking), they attended sports (basketball, football, baseball and golf), and they watched sports on TV (tennis, football, basketball, and baseball). These findings indicate that spectators of sport events have prior experience and are highly involved in sports. However, the area of prior experience with sports and its effect on attendance of athletic events as well as its relation to giving to athletics has not been investigated by researchers and is an emerging topic that needs to be studied. The Giving to Athletics Model (GAM) included the variable prior experience with sports in an effort to identify direct and indirect relationships with the other variables included in the model and particularly its role in athletic giving.

Involvement with athletics. Involvement has been defined as “the degree of pertinence and relevance of the cause to the individual” (Engel, 1987, p.48). Engel (1987) has proposed the active and passive reasoning theory of involvement borrowed from the marketing literature to explain donor behavior. According to this theory, when a donor is highly involved with a cause, there is high awareness that leads to careful evaluations of the various alternatives. The donor makes a decision to contribute to certain cause(s). This decision is consistent with other underlying factors (motives) and is perceived as offering personal benefit. Duke (1982) found that donors have better knowledge about the educational institution they fund than non-donors. When involvement is low, there is low level of awareness that leads to action before a decision is weighed carefully. Such decisions however, may not be made in the future.

The important role involvement plays in making donations to educational institutions has been addressed in the fund raising literature from two different perspectives: the undergraduate extracurricular involvement (Nelson, 1984; Hammersmith, 1985) and alumni involvement (Duke, 1982; Hammersmith, 1985). It has also been reported that students involved in extracurricular activities (Greek or professional organizations) are more likely to make donations to their universities and be involved as alumni. It has been also reported that alumni involvement is positively related to the frequency of the contributions made to an educational institution (Blakeley, 1974; Webb, 1989).

Involvement with a product has been found to be related with intentions of future purchases. A study of baseball fans indicated that involvement has a significant positive effect on attendance intentions. Fans who were highly involved in baseball were more willing to attend future baseball games than the low involved (Wakefield & Blodgett, 1994). Thus, involvement with the athletic program was expected to be directly related to attendance at sport events and indirectly to giving to athletics. Athletics may serve as a form that ties individuals emotionally to the university (especially alumni) and influences their giving behavior (Brittingham & Pezzullo, 1990).

Attendance of athletic events. In addition to involvement, attendance at sport events has been found to be related to alumni athletic donations (Hammersmith, 1985). Alumni athletic donors are more likely to attend athletic events than non-alumni. Hammersmith’s study identified attendance of home and away football and basketball games, attendance of bowl games and purchase of season football and basketball tickets as factors related to athletic giving. This may happen because several universities combine their ticket programs with their athletic contribution program (boosters). However, there is no strong evidence to suggest that attendance at athletic events is directly related to donations to athletic programs or vice versa. Thus, it was proposed that attendance at athletic events is indirectly related to donations in athletics. Season ticket purchase and priority seating have been identified as practical reasons that motivate donations and it was expected that attendance at athletic events directly influences such motivations.

Emotional motivation. In emotional motives, factors such as identification and attachment with the cause (institution or program) can be considered. People have strong positive feelings toward an institution and/or a program of that institution that motivates them in supporting it financially (Palmer, 1992). Several researchers have reported that feelings of identification with an institution are positively related with alumni involvement and charitable giving (Blakeley, 1974; Smith, 1989). Moreover, feelings of empathy toward a cause have been found to be related to intentions to give (Griffin, Babin, Attaway & Darden, 1993). Studies have reported that emotional attachment to one’s college is a strong predictor of donors’ giving. People who are more emotionally attached to an institution are more likely to make a contribution. Thus, emotional reactions to a college were expected to be related to the financial generosity of its donors (Spaeth & Greely, 1970; Mosser, 1993).

Practical motivation. Practical motives of giving behavior to athletics may include: tax deductions, priority seating, professional and social contacts. There are contradictory findings regarding the role tax deductions play in motivating giving. It has been argued that tax deductions do not affect giving but only the time and the form a contribution is made (Bakal, 1979). However, Hammersmith found that tax deductions is an important factor in motivating non-alumni donors (1985).

It has been suggested that there is a positive relationship between non-alumni donors and their business interests expressed as social or financial benefits (Hammersmith, 1985). Smith (1989) found that the use of athletic tickets for business purposes was one of the motives for giving to athletics.

Several studies have shown that priority seating for athletic events is a significant factor that motivates giving to athletic programs (Alger, 1969; Kern, 1983; Hammersmith, 1985; Isherwood, 1986; Nelson, 1986; Smith, 1989; Webb, 1989). According to Hammersmith (1985), priority seating is “the most motivating factor determining the amount of the contribution from each donor, especially at the middle contribution level” (p. 181). Smith (1989) reported that 92% of alumni and non-alumni athletic donors rank the opportunity to obtain tickets as one of the most important in making donations. It was also suggested that priority seating may be the only reason for giving by non-alumni.

The purpose of this research was to test the Giving to Athletics Model (GAM). The review of literature assisted in identifying factors and relationships among these factors that may explain how attitudes toward giving to athletics are formed and reasons that motivate donors to make contributions to athletic programs.

Research Hypotheses

Based on the review of literature on fund raising and athletic fund raising, several factors emerged as possible motives for giving to athletic programs. These factors were put together in a set of relationships and constituted the Giving to Athletics Model (GAM). The following statements that reflect the relationships among the factors represented the hypotheses of the study (Figure 1).

The major hypothesis of the study was whether the Giving to Athletics Model explained donations made to athletic programs. The sub-hypotheses of the Giving to Athletics Model were the following:

The study hypothesized that there is a direct positive effect of the following variables to athletic giving:

1. Practical Motivation

2. Emotional Motivation

In addition, the following relationships among the other variables in the GAM were hypothesized:

3. There is a positive effect of Attendance at Athletic events on Practical Motivation.

4. There is a positive effect of Sport Experience on Involvement with Athletics.

5. There is a positive effect of the socioeconomic status (ses) on: (a) Emotional Motivation and (b) Practical Motivation.

6. There is a positive effect of Involvement with Athletics on Attendance at Athletic Events.

7. There is a positive effect of Values on Emotional Motivation.

Subjects

The study utilized the survey research method to study donors of the Florida State University athletic program and other units of the university. A self-administered anonymous questionnaire was sent by mail to 800 athletic donors of the Florida State University.

Measurement Model

The Measurement Model of the Giving to athletics model is shown on Tables 1 and 2. Table 1 presents the measurement model for the exogenous variables whereas Table 2 presents the measurement model for the endogenous variables. All the latent variables consisted of effect indicators except the SES variable which consisted of cause indicators. Cause indicators are observed variables that are assumed to cause a latent variable. Effect indicators are observed variables that are caused by latent variables (Bollen, 1989).

Table 1: Measurement Model of the Giving to Athletics Model: Exogenous Variables

SES (Three indicators)

X1: Education (six levels ordinal variable)

X2: Personal income (six levels ordinal variable)

X3: Household income (six levels ordinal variable)

VALUES (9 indicators grouped into three indicators)

Nine point interval scale: very unimportant - very important

X4: Sense of belonging

X5: Excitement

X6: Self-fulfillment

X7: Being well respected

X8: Fun and enjoyment in life

X9: Security

X10: Self-respect

X11: A sense of accomplishment

X12: Warm relationships with others

SPORT EXPERIENCE (three indicators)

X13: Play sports (5 point interval scale: never - very often)

X14: Number of sports played (continuous variable)

X15: Number of years played (continuous variable)

Table 2: Measurement Model of the Giving to Athletics Model (GAM): Endogenous Variables

INVOLVEMENT WITH ATHLETICS (10 indicators grouped into two components

Cognitive Component:

Y1: Important to me

Y2: Relevant to me

Y3: Means a lot to me

Y4: Valuable

Y5: Needed

Emotional Component:

Y6: Interesting to me

Y7: Exciting to me

Y8: Appealing to me

Y9: Fascinating to me

Y10: Involving

ATTENDANCE OF ATHLETIC EVENTS (2 indicators)

Y14: Attendance

Y15: Sports attended

MOTIVES FOR EMOTIONAL REASONS (11 indicators grouped into three groups)

Y16: Identification

Y17: Affiliation

Y18: Being part

Y19: Support

Y20: Increase prestige

Y21: Meet friends

Y22: Loyalty

Y23: Association

Y24: Believing in the leadership

Y25: Believing in the vision

Y26: Tradition

MOTIVES FOR PRACTICAL REASONS (4 indicators)

Y27: Priority seating

Y28: Tax deductions

Y29: Professional contacts

Y30: Social contacts

DONATIONS TO ATHLETICS (one indicator /continuous variable)

Y31: Actual amount of money being donated

Procedures

Eight hundred donors were randomly selected from the records of the Florida State University Alumni Association office. To identify the amount of money donated by each donor, the subjects were assigned a number different from their existing personal identification code. This number was used for data recording as well as for matching the personal code with the contribution of the donors.

Statistical Analysis

This study used inferential statistics in an effort to test the Giving to Athletics Model and the hypotheses related to the model. In particular, the study utilized the Structural Equation Modeling (SEM) technique and the Maximum Likelihood (ML) estimation procedure. The LISREL 8.12 program was used to analyze the initial model and assess its fit (Joreskog & Sorbom, 1993).

Results

This section will first address the evaluation of model fit and then look at the results in terms of the hypotheses.

Figure 2

The purpose of this study was to test the Giving to Athletics Model (Figure 1) in an effort to explain why people make donations to athletic programs. The Giving to Athletics Model was tested by using the Maximum Likelihood procedure though Weighted Least Squares (WLS) was proposed before collecting the data. The 387 questionnaires received out of the 800 sent did not constitute an adequate sample size for the number of observed indicators needed for running WLS. Thus, the study took a more exploratory approach and a grouping of items had been attempted to reduce the number of observed variables. An exploratory factor analysis was used in grouping the single items of the latent variables. Following the exploratory factor analysis, several revisions took place and the Giving to Athletics Model was modified. Revisions were expected from the beginning of this study due to its exploratory and confirmatory nature. In addition, structural equation modeling is a sensitive technique and LISREL is a program that does not always give solutions as other programs and analyses. Maximum Likelihood (ML) helped in getting results for this study whose data were not always normally distributed. ML is robust to the violation of normality and other less than optimal conditions.

The analysis started by identifying the measurement and structural model. The identification of the models took place after every revision. Both models passed the rules and met the conditions necessary for identification.

The first PRELIS and Confirmatory Factor Analysis (CFA) lead to several revisions for improving the model and its fit. The latent variable SES was excluded from the analysis due to its “unique” relationship (causal) with the observed variables that probably affected the convergence of the solution. LISREL results indicated that the original Giving to Athletics Model (GAM) did not fit the data. The exclusion of the above mentioned variable lead to a new model which converged.

The fit of the model to the data was assessed with LISREL results. The chi-square test along with the global fit indexes, the size of the standardized residuals and the paths being significant in the final model were used for assessing the fit of the model. The chi-square test assessed the magnitude of the discrepancy between the sample and fitted covariance matrices. The fit indexes assessed the relative amount of the observed variances and covariances accounted for by the model.

The initial model results indicated a bad fit to the data since the chi-square was large (547.325) with 98 degrees of freedom. The size of the fit indexes was small whereas the standardized residuals were very large. Four additional paths were added, one from Donations to Athletics to Practical Motivation, one from Attendance of Athletics to Emotional Motivation, one from Involvement with Athletics to Donations to Athletics, and one from Values to Involvement with Athletics.

The new model had a relatively good fit to the data though the results were mixed. The chi-square fit of 269.784 with 94 degrees of freedom and a reject decision indicated that the model did not fit the data well. Of course, such a statement cannot be made with confidence due to the sensitivity of the chi-square test to the sample size. The likelihood of rejecting a valid model increases with large samples. The sample of this study was relatively large (n=387) which made the detection of small discrepancies between the observed and fitted covariance matrices likely.

Due to the sensitivity of the chi-square test to the sample size, other fit indexes (RMSEA, AGFI, CFI) were taken into consideration before making a decision about the model fit. Fit indexes such as Root Mean Square Error of Approximation (RMSEA), Adjusted Goodness Fit Index (AGFI), and Comparative Fit Index (CFI) have the advantage of not being sensitive to sample size.

The AGFI is a measure of the relative amount of variances and covariances that are accounted for by the implied model and has been found to be consistent across different estimation procedures when the sample size is larger than 250 (Hoyle, 1995). The AGFI of the revised GAM (0.886) indicated a good fit (its value was close to the accepted cutoff value of 0.90. Moreover, the CFI value was taken into account when the revided GAM was evaluated. The 0.921 value of CFI indicated a good model fit to the data.

Another consideration taken when evaluating the model fit was the size of the standardized residuals. A value of 3 is usually an acceptable one. The standardized residuals of the revised GAM were as large as 4.588. This value is large comparing to 3 but it was significantly smaller than the size of the residuals in the initial model (11.180).

The last consideration in evaluating the model fit was the significant paths and their effect size. Only five out of the 10 paths of the revised GAM were significant indicating only a fair fit. The sizes of these effects were large ranging from 0.618 to 0.989. All the above results were evaluated and it was decided that the model was partially supported by the data. Probably, the revised GAM was one of the best possible models derived from the given data and explained only some of the variability of Donations to Athletics.

Table 3: Summary of Causal Effects in the Revised GAM

   _________________________________________________________________

* Z-Statistic > 2

Discussion of the Hypotheses Results

The major hypothesis of the study was not supported by the data. The original GAM did not fit the data well. However, the revised version of GAM seemed to have a fair fit. The revised GAM is a parsimonious model that partially explains donations to athletic programs. The LISREL results indicated five significant paths in the revised GAM. Three of these paths were directed or coming from the Donations to Athletics latent variable.

The path from Emotional Motivation to Donations to Athletics was significant at the 0.05 level indicating that Emotional Motivation explained Donations to Athletic programs. Thus, when individuals are emotionally motivated they are expected to make donations to athletic programs. The sign of the path coefficient was positive thus, there was a positive relationship between Emotional Motivation and Donations to Athletics. As a result, when Emotional Motivation increases, Donations to Athletics are expected to increase too. This relationship supports the second hypothesis of the study (+H2) where a direct positive relationship between the two latent variables was expected.

The two way path between Practical Motivation and Donations to Athletics indicated that both variables affect each other. However, only the path from Donations to Athletics to Practical Motivation was significant. That path indicated that Donations to Athletics significantly affected Practical Motivation. The relationship was positive so an increase of Donations to Athletics was expected to lead to an increase in Practical Motivation of the donors. This reverse relationship between the above variables had not been predicted and hypothesized at the original model. The findings did not support the first hypothesis (H1) of the study since the path from Practical Motivation to Donations to Athletics was not statistically significant.

The third significant path was from Involvement with Athletics to Donations to Athletics (0.923). There was a direct positive relationship between the two latent factors that was not included in the original GAM to avoid problems of identification. Such a relationship was expected but it was never stated formally. Thus, the more involved an individual is with Athletics, the more he/she will donate to athletic programs.

Involvement with Athletics and Emotional Motivation were the only latent variables that directly explained Donations to Athletics. Of course they had indirect effects that have been presented in the previous section and on Table 3.

The fourth significant path of the GAM was between Sport Experience and Involvement with Athletics. Sport Experience was directly related to Involvement with Athletics. The relationship between the two variables was positive. This means that the more experience with sports someone has the more involved he/she is with sports. This relationship reflects the fourth hypothesis (+H4) of the study which the results have confirmed.

Emotional Motivation was explained by Attendance of Athletics. This path was significant at the 0.05 level (0.841) and indicated a positive relationship between the two variables. This path was not included in the original GAM though it was expected that people who attend athletic events are emotionally motivated in giving to athletics. The more athletic events they attend the more emotionally motivated they are to give.

The findings of this study have theoretical and practical implications. The revised GAM helps in understanding why some people make donations to athletic programs and it can assist in theory development. Involvement with Athletics significantly affected Donations to Athletics. This information suggests that the fund raisers should identify people involved in sports as well as people having experience in sports as their prospect athletic donors. For example, university fund raisers could identify prospect donors for their athletic programs through the records of the intramural departments. In addition, the study helped in identifying several areas in athletic fund raising that need to be further investigated. Follow-up studies with larger sample sizes in other universities and with different estimation procedures are needed to fully explore the topic and to confirm this study’s findings.

References

Alger, N. R. (1969). A survey of fund raising methods for intercollegiate athletics in the Western Athletic Conference. Unpublished doctoral dissertation, University of Utah, Salt Lake City.

Bakal, C. (1979). Charity U.S.A.: An investigation into the hidden world of the multi-billion dollar charity industry. New York, NY: Times Books.

Blakeley, B. E. (1974). Historical and contemporaneous predictors of alumni involvement. Unpublished doctoral dissertation, Purdue University, Hammond.

Bollen, K. A. (1989). Structural equation with latent variables. New York, NY: John Wiley & Sons, Inc.

Brittingham, B. E. and Pezzullo, T. R. (1990). The campus green: Fund raising in higher education. Washington D.C.: Clearinghouse on Higher Education.

Duke, J. R. (1982). Giving to educational institutions: The relationship between individual and institutional value systems. Unpublished doctoral dissertation, The Florida State University, Tallahassee.

Engel, J. F. (1987). What motivates giving to christian organizations? Fund Raising Management, 18(July), p. 48-52.

Graham, P. (1994). Characteristics of spectators attending professional tennis tournament in two regions of the U. S. Sport Marketing Quarterly, 3(3), 38-44.

Griffin, M., Babin, B. J., Attaway, J. S. and Darden, W. R. (1993). Hey you, can ya spare some change ? The case of empathy and personal distress as reactors to charitable appeals. In L. McAlister, & M. L.Rothschild, (Eds.), Advances in Consumer Research (pp. 508-514). Vancouver: Association for Consumer Research.

Hammersmith, V. (1985). The development of a survey instrument to profile donors to athletics. University of West Virginia. University of Oregon Microfish.

Hoyle, R. H. (1995). Structural equation modeling: Concepts, issues and applications. Thousand Oaks, CA: Sage Publications.

Isherwood, A. C. (1986). A descriptive profile of the fund raising programs in NCAA Division I-A. Unpublished doctoral dissertaition, University of Maryland, Maryland.

Joreskog, K. G. & Sorborm, D. (1993). LISREL 8: Structural equation modeling with the SIMPLIS command language. Lawrence Erlbaum Associates Publishers: Hillsdales, NJ.

Kern, R. W. (1983). A comparative study of ten selected athletic fund raising programs fron institutions of higher education West of the Mississippi river.Unpublished doctoral dissertation, Ohio State University, Colombus.

Mosser, J. W. (1993). Predicting Alumni/ae gift giving behavior: A structural equation model approach. Unpublished doctoral dissertation, the University of Michigan.

Nelson, W. T. (1984). A comparison of selected undergraduate experiences of alumni who financially support their alma mater. Unpublished doctoral dissertation, Indiana University, Bloomington.

Palmer, J. B. (1992). The nature and status of resource development activities in National League for nursing accredited Baccalaureate and Master's degree schools of

nursing. Unpublished doctoral dissertation, The Florida State Univerity, Tallahassee.

Smith, J. Jr. (1989). Athletic fund-raising: Exploring the motives behind private donations. Master thesis. University of Noth Carolina, Chapel Hill.

Spaeth, J. L. & Greely, A. M. (1970). Recent alumni and higher education. New York, NY: McGraw-Hill Book Company.

Wakefield, K. L. and Blodgett, J. G. (1994). The importance of servicescape in leisure service settings. Journal of Services Marketing, 8(3), 66-76.

Webb, J. T. (1989). Analysis of financial donors to Oklahoma State University's athletic program. Unpublished doctoral dissertation, Oklahoma State University, Stillwater.

Yavas, U. & Riecken, G. (1980). Segmentatiing the Market for a nonprofit organization: Profiling the heavy, moderate and nondonors to United Way. Proceedings (p. 281). American Institution for Decision Making.