Research Productivity and Social Capital in Australian Higher Education
Abstract
This study investigates the role of social capital in raising research productivity in academic institutions. Social capital as a strategic resource embedded in social relationships can be utilised towards decreasing pressures from external environmental conditions, such as the global financial crisis. A survey was sent to academic staff in five universities in Victoria, to collect data regarding their frequency of communications and research productivity. The findings indicated that there is a significant and positive correlation between social interactions and research productivity. Regression analysis demonstrated that social interactions as an independent variable predict research productivity of academics.
Introduction
Higher education systems are under extensive pressure to improve their services. There are many trends and demands that put pressure on universities to respond. Among those are decreasing financial assistance from Government, new demands from customers of different ages who seek to upgrade their qualifications, economic and social institutions that expect universities to provide competent graduates, and the base of knowledge which is expanding internationally and exponentially (Clark, 2004). These challenges require academic institutions to expand their activities in the distribution and dissemination of knowledge.
Scholars in the field have creatively elaborated on different methods to decrease these pressures. Clark (2004) argued that universities should adopt an entrepreneurial posture as a response to this turbulent and dynamic environment. Marginson and Considine (2000) debated on the ‘enterprise university’, which carries a tone of marketisation and intentional effort in institution building that requires much special activity and energy. Slaughter and Leslie (1997) argued for the concept of ‘academic capitalism’ to explain the response of academic staff and managers to the new opportunities and pressures to generate revenues. Mignot-Gerard (2003) recognised a change in approach to university management. To respond to these endless demands and pressures, this article suggests that academic institutions should rely on their staff as the main generators and providers of research and teaching services and increase their productivity.
Productivity as a strategy to respond to decreasing Governmental assistance and responding to environmental demands has been debated in the literature (Gates and Stone, 1997). Scholars suggested several strategies at the organisational and individual levels to improve productivity in higher education, including, for example, improving administrative processes, restructuring curricula, more efficient use of classroom time and the effective adaptation of technology (Groccia and Miller, 1998). The article argues that productivity of academic staff as the engine of generating ideas and innovation should be enhanced. It focuses on social capital as an asset embedded in relationships to improve the productivity of academic staff.
Organisational settings as collective entities are conducive to the development of social capital (Nahapiet and Ghoshal, 1998; Burt, 2000). Academic institutions are ripe with social capital: as resources embedded in social interactions, they can create value for participants. The individuals who are connected to networks and have relationships with other members have access to resources that are embedded through these social networks (Putnam, 2004). Academic staff have a lot of social capital in their possession which can be exploited towards individual and organisational objectives. This study attempts to leverage the resources embedded in their relationships to enhance their productivity.
Literature review
Social capital
The concept of social capital has been investigated increasingly in recent years as a useful resource in the form of the co-operative behaviour and trust that is engendered by the fabric of social relationships (Adler and Kwon, 2002). It has been applied to solve many problems with social ties since its appearance in the literature. Such applications include education, public health, economic development, community life, youth behaviour problems and general problems of collective action (Loury, 1977; Coleman, 1988; Kogut and Zander, 1992; Putnam, 1993, 1995; Woolcock, 1998; Fukuyama, 1995a). Originating in sociology, it has gained currency in different disciplines, including organisational studies, entrepreneurship, psychology, economy and politics.
Organisational and management researchers have applied this concept to answer many of the questions they are confronted with in their field of study. The range of organisational issues that have been addressed by social capital has been broad and various, including career success (Burt, 1992; Gabby and Zuckerman, 1998), executive compensation (Belliveau et al., 1996), finding jobs (Granovetter, 1995; Lin and Dumin, 1996), producing a pool of recruits for firms (Fernandez et al., 2000), product innovation (Tsai and Ghoshal, 1998), the creation of intellectual capital (Nahapiet and Ghoshal; Tsai 2003), the formation of start-up companies (Walker et al., 1997), effectiveness of entrepreneurial teams (Weisz et al., 2004) and strengthening supplier relations (Uzzi, 1997; Inkpen and Tsang, 2005). This study, however, examined the role of social capital in enhancing productivity in the higher education system.
Social capital has been defined as ‘the sum of the actual and potential resources embedded within, available through, and derived from the network of relationships possessed by an individual or social unit’ (Nahapiet and Ghoshal, 1998, p. 247). Burt (1992) noted that friends, colleagues and/or general contacts provide opportunities to use both financial and human capital. Thus, members of networks who are connected to other participants have access to resources embedded through such networks (Granovetter, 1992). These valuable resources derive from having connections and relationships that may be mobilised in pursuing individual and organisational objectives. This study attempts to examine the way in which the level of social interactions and the contacts of academic staff can be leveraged to improve academic performance.
Without social ties and interactions no social capital will evolve and, therefore, the resources embedded in social relationships will not be available to individuals or social units. Granovetter (1992) has used the term ‘structural embeddedness’ which is concerned with the properties of the social system and of the network of relations as a whole. In other words, structural embeddedness describes the impersonal configuration of linkages between people or units. Burt (2004, p. 365) informed us that network analysis is concerned with ‘the contacts, ties and connections, the group of attachments and meetings which relate one agent to another. These relations connect pairs of agents to larger relational systems’. People or organisational units can use their personal or collective contacts to obtain information or other resources to facilitate their activities (Fukuyama, 1995b). Thus, in departments and faculties academic staff who have personal contacts within or outside their faculties or departments potentially have access to embedded resources.
Social ties are channels for information and resource flows. Through social interactions, an actor may gain access to other actors' resources. Such access allows innovators to go across formal lines and levels in the organisation to find what they need. The location of an actor's contacts in a social structure of interactions provides certain advantages for the actor. People can use their personal contacts to get jobs, to obtain information or to access specific resources. Nahapiet and Ghoshal (1998) have categorised these behavioural models of social interactions into three facets, including: network ties – the absence or presence of network ties between actors; network configuration – that describes the pattern of connections in terms of measures such as density, connectivity and hierarchy; and appropriable organisation – that is, ties of one kind can be used for different purposes. In the academic context, academicians have several social ties such as colleagues, superiors and industry/business through which they can have access to information and resources which may be exploited for enhancing individual and organisational productivity.
Productivity
Several scholars have maintained that measuring and defining productivity in higher education has been a difficult task (Wood, 1990; Gates and Stone, 1997). The studies on measuring productivity and the factors associated with it have been the focus of research in the literature (Ramsden and Moses, 1994; Groccia and Miller, 1998). Technically, productivity is a measure of output per unit of input and it can be applied in a variety of contexts. The concept of productivity has two dimensions: efficiency and effectiveness. Efficiency refers to the level and quality of service that is obtained from the given amount of resources. If the sector can produce a greater quantity and/or higher quality of output with the same amount of resources, it has improved its efficiency. Effectiveness relates to the extent to which the provider meets the needs and demands of stakeholders or customers. In the higher education sector, these stakeholders include students, faculty, local communities, state governments, industry and the nation at large.
Evaluating the performance of academic staff in universities is somewhat complex and difficult as universities consist of different departments with diverse natures, which affects their performance and its appraisal (Clark, 1998; Marginson and Considine, 2000). There is a wide range of activities that academic staff members are required to engage in that should be evaluated. These range from producing publications such as journal papers to conducting research for business and teaching students. Academic staff in universities are conducting teaching, research and administrative functions, each of which needs special treatment (Wood, 1990).
In the academic community, the most critical indicator of research productivity is publication. As the physical and conventional form of the academic world, publishing books and journal articles is the most fundamental social process of communicating and exchanging research findings (Wood, 1990). In academic environments, publication brings recognition and promotion for both academics and their institutions. In addition, as a unique criterion for obtaining competitive research funds, publishing is evidence of institutional excellence. Leading universities are expending a lot of time and money in publicising the quantity of their books and articles. The growing use of numbers of publications as an indicator of a department's performance indicates its importance in academic institutions.
Measuring productivity
When it comes to measuring productivity, the first task is to define the unit of analysis. For example, an analysis of productivity in higher education could focus on the productivity of an individual (a professor), an organisational sub-unit (an academic department or a school within a university), an organisation (a college or university) or a population of organisations (a state higher education system or the higher education industry as a whole). The specific productivity goals and the key indicators selected will differ depending on which level of analysis is the focus. While there are many levels at which productivity can be analysed, this article focuses on the productivity of academic staff.
An index of research productivity (IP) was defined as the sum of five measurements (3* the number of single or multi-author books) + (the number of papers published in refereed journals) + (the number of edited books) + (the number of chapters in refereed books). This index of productivity, like all others that attempt to provide a single measure of quantity of output that is applicable across different disciplines, is imperfect. It is, however, consistent with the measure reviewed by previous authors (Wood, 1990). In this study, the number of books, refereed papers, conference papers, edited books and chapters that have been produced over the past three years by academic staff form the productivity index.
The index of research activities (IA) was calculated from answers to questions about whether or not the respondent had undertaken each of a series of academic activities such as: received an external, competitive research grant; received an internal, competitive research grant; refereed one or more articles for a journal; and so on. In the context of this study, in addition to some of these elements, research income and consultancy income have been added. As such, publication and research activities were used in this study to measure the productivity of academic staff.
Association between social capital and productivity
Social capital, which resides in relationships between people, can improve the productivity of organisations and individuals. The association between social capital and performance has been the subject of focus at both national and organisational levels. Knack and Keefer (1997) provide empirical evidence to show that ‘social capital’ matters for measurable economic performance, using indicators of trust and civic norms from the World Values Surveys for a sample of 29 market economies. They found that trust and civic norms are stronger in nations with higher and more equal incomes, with institutions that restrain predatory actions of chief executives and with better educated and ethnically homogeneous populations. Putnam (1993) argued that membership of formal groups is associated with efficiency and effectiveness at the national level.
At the individual level, there is some empirical evidence that has established the association between organisational trust and job satisfaction which can improve the individual's performance within organisations. Reagans and Zuckerman (2001), referring to the social capital literature, tested the hypothesis that the diversity of a team's internal network is associated with higher productivity in corporate research and development teams. Hite et al. (2005) investigated the role of multiple and informal networks in improving the performance of educational leaders. Apparently, there are a limited number of studies that investigate this association in the higher education context; this study attempts to fill the gap by developing the following two hypotheses and subjecting them to statistical analysis.
H1. There is a relationship between social interactions and productivity.
H2. Social interactions predict productivity.
Methodology
The sampling frame of this study consisted of full-time academic staff at public universities in metropolitan Melbourne. The population of 5,695 academic staff embraced various levels of academic position, ranging from lecturer to professor. There were 162 males (60.4 per cent) and 106 females (39.6 per cent) in the sample, giving a total of 271 respondents. Data were collected via electronic media, whereby academic staff were encouraged to participate in an online questionnaire via an email hyperlink. The online and electronic media survey approach as amplified by Dillman (2000) was adapted for the study. Ethics approval was obtained from all five participant universities.
Tow indices were developed to capture the productivity construct. An index of research publication (IP) was defined as the sum of five measurements (3* the number of single or multi-author books) + (the number of papers published in refereed journals) + (the number of edited books) + (the number of chapters in refereed books). This index of productivity, like all others that attempt to provide a single measure of quantity of output that is applicable across different disciplines, is imperfect. It is, however, consistent with the more advanced measure reviewed by previous authors (Wood, 1990). Therefore, the publications of academic staff were measured by the sum of the number of books, refereed papers published in academic journals, conference papers, edited books and book chapters that were produced over the past three years. Item analysis procedures were used to help form an internally consistent scale focused on research activity. The current research has operationalised these indices by asking seven questions for research activity (IA) and nine questions for publication (IP). Therefore, the questionnaire for the construct has 16 items. Some of the items for research activities were on a Likert scale with five options from completely disagree (1) to completely agree (5). Several questions have been negatively scored and for other items respondents had to answer the questions by specifying the related number.
The Likert scale of five options was used to operationalise the social interactions of academic staff. The respondents have specified the time they have spent in a week with their networks by choosing one of the options from 0 to 2 hours a week (score 1) to more than 9 hours (score 5).
The characteristics of the statistical sample such as gender, age and experience in the field of study, experience in their institutions, function and position have served as control variables. One way analysis of variance (ANOVA) was used to examine statistically significant differences among groups classified by social interactions and entrepreneurship. A significance level of 0.05 was set for the various analyses. When the ANOVA provided an F ratio that was statistically significant beyond the 0.05 level, post hoc procedure as outlined in Tabachnick and Fidel (2001) was used to compare individual subgroups within a scale in an attempt to locate differences that contributed to the analysis of variance result. To compare females' and males' scores for each research constructs, t-tests were used. Descriptive measures included mean, median, mode and frequency of distributions of the sample. To examine the relationships between research variables, correlation was applied and regression employed to predict the dependent variable (productivity).
Results
Social interactions results
Social interactions of respondents within and outside their departments are indicated in Table 1. As can be seen, 38.4 per cent of academic staff were communicating with their colleagues for between three and five hours a week, about 23 per cent spent more than nine hours a week in communication, 20 per cent between six and eight hours, 17 per cent between one and two hours a week and 1.1 per cent had no communication with their colleagues, with a mean of 3.46 and standard deviation of 1.06. Positive skewness value (0.093) indicates that scores are clustered to the low values and kurtosis value (−0.992) demonstrates a distribution that is relatively flat, with too many cases at the extremes. It means that most academics have relatively good communication with their colleagues.
Social interaction of respondents
| How many hours a week do you spend communicating with your: | None 1% | 1–2 h 2% | 3–5 h 3% | 6–8 h 4% | More than 95% | Mean | SD | Skewness | Kurtosis |
|---|---|---|---|---|---|---|---|---|---|
| 1. Colleagues | 1.1 | 17.3 | 38.4 | 20.3 | 22.9 | 3.46 | 1.06 | 0.093 | −0.992 |
| 2. Administration | 9.2 | 51.3 | 29.9 | 5.5 | 4.1 | 2.44 | 0.887 | 0.906 | 1.130 |
| 3. Superior | 19.6 | 66.4 | 12.2 | 1.1 | 0.4 | 1 | 0.629 | 0.662 | 2.323 |
| 4. Attending social occasions | 37.3 | 54.2 | 7.4 | 0.4 | 0 | 1.70 | 0.616 | 0.377 | −0.127 |
| 5. Colleagues in other departments | 29.9 | 56.5 | 11.8 | 1.5 | 0.4 | 1.859 | 0.700 | 0.721 | 1.351 |
| 6. Colleagues outside institutions | 20.7 | 55.7 | 18.8 | 4.1 | 0.7 | 2.084 | 0.786 | 0.723 | 0.981 |
| 7. International colleagues | 35.8 | 52.8 | 8.1 | 2.2 | 0.7 | 1.788 | 0.744 | 1.126 | 2.436 |
| 8. Business/industry contacts | 31 | 51.3 | 14 | 1.8 | 1.1 | 1.899 | 0.787 | 0.964 | 1.775 |
Question 2 in the social interaction scale asked respondents about the time that they spent in communicating with the chancellery or central administration. As Table 1 shows, 51 per cent of academics in the sample were communicating with central administration for between one and two hours, 30 per cent between three and five hours, 4 per cent more than nine hours, 5.5 per cent between six and eight hours and 9 per cent had no communication with central administration, with a mean of 2.44 and standard deviation of 0.887. Positive skewness value (0.906) indicates that scores are clustered to the left at low values and kurtosis value (1.130) indicates that the distribution is rather peaked. It means that academics have little communication with central administration.
Question 3 in the social interactions scale related to communication with superiors. As can be seen from Table 1, 66 per cent of respondents were communicating for between one and two hours a week with their superiors, 12 per cent between three and five hours, 1 per cent between six and eight hours, 0.04 per cent for more than nine hours and about 20 per cent had no social interactions with their superior, with a mean of 1 and standard deviation of 0.629. Positive skewness value (0.662) indicates that scores are clustered to the left at the low values and kurtosis value (2.323) shows that the distribution is rather peaked and clustered in the centre. It indicates that most academics have normal communications with their superiors – enough to do their normal duties.
Question 4 in the survey related to the participation of academics in social gatherings and occasions. As Table 1 indicates, 54 per cent of respondents spent between one and two hours attending social gatherings and occasions, 7 per cent between three and five hours and 37 per cent of academic staff were not attending any social gathering in their department or faculty. The mean is 1.7 and standard deviation is 0.616. Positive skewness value (0.377) indicates that scores are clustered to the left at the low values and kurtosis value (−0.127) indicates that the distribution is relatively flat with too many cases in the extremes. It indicates that about 40 per cent of academics do not attend social gatherings at all.
Question 5 asked about the time that academics in the sample spent with their colleagues in other departments. As shown in Table 1, 30 per cent of academics had no social interactions with their peers in other departments, while 56 per cent were communicating with their colleagues in other departments for between one and two hours, 12 per cent between three and five hours, 1.5 per cent between six and eight hours and 0.4 per cent more than nine hours, with a mean of 1.859 and standard deviation of 0.7. Positive skewness (0.721) indicates that scores are clustered to the left at the low values and kurtosis value (1.351) indicates that the distribution is rather peaked in the centre. As can be seen, academics have little communication with their colleagues in other departments.
Question 6 in the survey is related to communication of respondents with their colleagues outside their own institutions. As can be seen from Table 1, 55 per cent of respondents spent one to two hours a week communicating with their colleagues outside their institutions, 12 per cent between three and five hours, 1.5 per cent between six and eight hours, 0.7 per cent more than nine hours and 20 per cent had no communication with academics in other institutions, with a mean of 2.084 and a standard deviation of 0.786. Positive skewness (0.723) indicates that scores clustered to the left at the low values and kurtosis value (0.981) reflects the fact that the distribution is rather peaked at the centre.
Question 7 in the social interactions scale asked about communication of respondents with their international colleagues. About 36 per cent of respondents did not communicate with their international colleagues. However, 53 per cent were communicating for between one and two hours a week with their colleagues in other countries, 8 per cent between three and five hours, 2 per cent between six and eight hours and 0.7 per cent more than nine hours, with a mean of 1.788 and a standard deviation of 0.744. Positive skewness (1.126) indicates that scores are clustered to the left at the low values and kurtosis value (2.436) shows that the distribution is rather peaked, clustered in the centre.
Finally, question 8 related to the communications of respondents with industry or business related to their field of study. Table 1 indicates that 31 per cent of academic staff had no communication with business and industry related to their field of study. However, 51 per cent of respondents spent between one and two hours a week in communication with industry/business, 14 per cent between three and five hours, 2 per cent between six and eight hours and 1 per cent more than nine hours, with a mean of 1.899 and standard deviation of 0.787. Positive skewness (0.964) indicates that scores are clustered to the left at the low values and kurtosis value (1.775) reflects that the distribution is rather peaked.
Although normality can be assessed to some extent by obtaining skewness and kurtosis values, other techniques were also available (Pallant, 2001). The results of the normality test showed that the mean score of frequency of communications related to the job for males (17.45) is more than for females (16.83), while the standard deviation is higher for the male group (4.440 for males and 3.447 for females). In addition, the positive value of skewness for males (0.285) indicates that total scores for social communications related to the job are clustered to the left and kurtosis value for males (–0.266) indicates that the distribution is rather flat. For females, the positive value (0.545) reflects the fact that the scores are clustered to the left, with too many cases at the extremes, and kurtosis value (1.328) indicates that the distribution is rather peaked, clustered in the centre.
An independent sample t-test was conducted to compare the frequency of communications scores for males and females. There was no significant difference in scores for males (M = 17.45, SD = 3.44) and females (M = 16.83, SD = 3.44); t (260) = 1.414, P = 0.709.
Correlation results
As can be seen from Table 2, there is a relatively moderate and positive relationship between social interactions and productivity at the 1 per cent level (r-value 0.277, significant at P < 0.0025 and n = 193).
Relationship between social interactions and productivity
| Social interactions | ||
|---|---|---|
| Productivity | Correlation | 0.277 |
| Sig. | 0.000 | |
| N | 193 | |
Regression results
A standard regression was performed between productivity of academic staff as the dependent variable and frequency of interactions as the independent variable. Table 3 indicates that the unstandardised regression coefficients (B) and intercept, the standardised regression coefficients (beta), sr2, R2. R for regression was significantly different from zero, F (2, 185) = 30.481, P < 0.001. For the regression coefficient, 95 per cent confidence intervals were calculated. The confidence limits for frequency of interactions were 0.59–2.483.
Predicting dependent variable
| Model 1 | B | Beta | Sr2 |
|---|---|---|---|
| Social interactions | 1.536** | 0,207 | 0.042 |
- ** P < 0.01.
- Dependent variable: productivity.
- Independent variable: social interactions.
The independent variable contributed significantly to prediction of productivity frequency of interactions 0.042 (sr2. The beta coefficient in Table 3 provides information regarding the level of contribution of the independent variable in predicting the dependent variable. As the standardised coefficients (beta) column shows, the beta coefficient is 0.207. This means that this variable makes a moderately unique contribution to explaining the dependent variable.
Discussion
Academic institutions are under constant pressure to extend their services from a myriad of stakeholders, including Government, the public, and social and economic institutions. In an attempt to reduce these pressures and respond to demands and trends in this challenging environment, this study was undertaken. Academic institutions have much intellectual, physical and social capital in their possession to use in their activities. The focus of the study was on social capital as a strategic resource in fostering the performance of knowledge-based institutions and relieving these pressures.
The richness of social capital in organisational settings persuaded scholars to examine the relationship between the emerging concept and main concerns in organisational studies (Nahapiet and Ghoshal, 1997; Adler and Kwon, 2002; Burt, 2000). In particular, there is much social capital in academic institutions that should be utilised in the interests of individuals and organisational objectives. We focused on social interactions as an important dimension of social capital and stressed the value of social networks. The role of social capital in improving performance was investigated and the results confirm the main proposition of this study.
Analysis of findings
The findings indicated that social interactions, including frequency of interactions, had a relationship with productivity. This finding is in accordance with propositions documented in the literature; for example, Putnam's (1993) study found a strong relationship between social capital and performance in communities. Similar findings were found in the literature; for example, the importance of networks in enhancing productivity was studied by Reagans and Zuckerman (2001) in R & D institutions. They found that people with dense social interactions are more productive in research activities. The benefits of access to networks for utilising opportunities have been reflected in the literature (Burt, 2000; Adler and Kwon, 2002).
The results of regression analysis showed more parsimonious and complex conclusions. The predictability of productivity was evaluated. Regarding the importance of improving the performance of academic staff, the equation for predicting productivity as a dependent variable and as an independent variable was developed in this study. As an independent variable, the frequency of interactions predicts performance. To make sure that the variable still contributes significantly to predict productivity, the effects of some variables such as age and experience in the field of study and institution were controlled for. The results of the statistical analysis indicated no change in the equation.
Again, frequency of interactions plays an important part in predicting productivity of academics and their research activities. The effects of other variables such as age and experience in the field of study were controlled for separately and in combination, and the results indicated that these two independent variables still contribute significantly to predict productivity and research activities. Therefore, given the score on frequency of interaction, productivity and research activities can be predicted.
Conclusion
Higher education systems are under constant environmental pressures to improve their performance. This study attempted to investigate the role of social capital in fostering and enhancing productivity in higher education. It focused on social capital as a powerful resource to improve the performance of academics. The findings indicated that social capital as an asset rooted in relationships has actual and potential benefits for individuals and institutions.
These findings have implications for individuals and knowledge-based institutions. First of all, academics should have more social communication with their colleagues, superiors and central administration, and colleagues in other departments. In addition, they should span more structural holes by having more contacts with business and industry related to their field of study outside their institutions. Academics who are connected to groups beyond their own can expect to find themselves delivering valuable ideas. The finding is in accordance with the literature, which documented the role of networks and weak ties in the prosperity of individuals and communities (Putnam, 1995; Burt, 2004).
In addition to these, the study has an implication for policy makers in knowledge-based institutions. Regarding the role of social communications in fostering productivity, it is recommended that academic institutions facilitate the relationship between their academic staff and business and industry, international colleagues, colleagues and superiors. The more time spent in communications, the more productive an academic is. Also, the environment in academic institutions should be designed in a way that enables academic staff to have more social interactions. Building social capital requires not only establishing more social ties but also nurturing motivation and providing resources. Management should invest in establishing social ties for academics within and outside their institutions. Furthermore, policy makers in Government should pay attention to the social capital available in academic institutions and plan to utilise such capital in the interest of economic and social development.
