Academic patronage

Patronage, expansively conceived as covering all forms of bias and discrimination, is pervasive in organisations and professions, including academia. Four key types of academic patronage operate through decisions made, processes used, assistance given to individuals and personal interactions. Some forms of patronage, especially discrimination on the basis of gender and ethnicity, have come under sustained criticism and are officially stigmatised. However, policies for equal opportunity and against conflicts of interest have only begun to address more personal forms of patronage. Some forms of patronage, such as supporting one’s research students, are common and treated as normal; systems without such patronage would seem strange. Looking at patronage in a broad sense can be useful in highlighting abuses that escape the usual lenses of anti-discrimination.


Introduction
At the University of New South Wales, a scandal was reported concerning the Educational Testing Centre (Hicks, 2005; see also Audit Office, 2001). There were allegations of favouritism and bullying. One fact stood out: one quarter of the employees were relatives of each other. This seemed to be a case of nepotism. Could they all have been appointed on merit?
This may be an extreme example, but milder forms are pervasive in nearly every organisation, including universities (McNamee & Miller, 2004). I use the word "patronage" to refer to favouritism based on attributes not connected with merit or performance. The person favoured can be a relative, friend, spouse, protégé or student, or share the same gender, ethnicity, social class, age, religion, political views or other beliefs. This is an expansive definition of patronage, incorporating all sorts of bias and special privilege: see Table 1. Although the word "patronage" has some benevolent connotations -as in a patron of the arts -it also suggests favouritism. Even in something seemingly as innocuous of being a patron of a restaurant -a regular customer -there is always choice involved, with someone potentially gaining advantage over another.
Most attention in academic studies is given to specific types of discrimination, especially those based on gender and ethnicity. My aim here is to use a broader perspective to give some insight into common processes, problems, alternatives and strategies and, in particular, to highlight some common practices that potentially might be considered improper. Table 1.

Some common terms as used in this paper
Bias: treating entities unequally using a non-universal value system. Conflict of interest: a situation in which a decision-maker, due to personal or other relationships, has competing loyalties and is potentially perceived by others as biased or compromised. Cronyism: favouritism for friends and associates. Discrimination: bias or unfair treatment, usually on the basis of membership in a category such as gender, class or religion. Equal opportunity: non-discriminatory treatment of individuals, typically in relation to jobs or careers. Favouritism: synonym for patronage. In-group: a group giving greater advantages to members than non-members; a clique. Insiders: members of an in-group. Merit: worthiness based on capability or performance treated independently of personal characteristics or group membership. Meritocratic: based on merit. Outsiders: non-members of an in-group, typically those working with or competing with in-group members. Patronage: any form of bias or discrimination, either by individuals or groups. Prejudice: an attitude or value system involving bias, often towards particular groups. Unfairness: an unequal outcome without a justification using universal criteria.
I start by describing key modes of patronage and give examples of avenues for patronage. I then present arguments for and against academic patronage, describe some systems with and without patronage and present several possible challenges to patronage. I use the example of honours thesis examining to illustrate struggles over a practice with considerable potential for patronage.
My aim is to use a broad lens to capture types of patronage not fitting into the usual categories of discrimination and equal opportunity. That means I do not cover several topics worthy of analysis, such as the reproduction of patronage -whether and how recipients later become dispensers of patronage -and the social processes by which forms of patronage rise or decline over time.
Patronage practices vary enormously internationally. In some countries, patronage is highly entrenched, with god-professors acting as gatekeepers into an academic career: no one without high-level personal endorsement has any chance of getting ahead. In some places, this becomes what might be called corruption, with university places available for purchase, out-groups denied access to study and excluded from academic jobs, degrees given as favours to powerful figures, and senior academics chosen on political rather than academic grounds. In other places -Australia, for example -where student selection and staff hiring operate with more observance of merit principles and where anti-discrimination legislation is taken seriously, patronage typically takes less blatant forms. Here I look at these issues from a general perspective: the case study is Australian, but the frameworks are meant to apply to practice in a variety of higher education systems.

Types of patronage
Avenues for patronage abound. Consider, for example, allocating offices. Whoever makes decisions about which person gets which office can use this power to dispense favours, giving the largest or most attractive offices to insiders. If you are assessing a grant application or refereeing an article by someone you like, you can give it a more favourable evaluation. If you are deciding the order of authors on a publication, you can manoeuvre to make a favoured person the first author, or even add the name of someone who didn't do any work, as a favour to that person, in what is called gift authorship (LaFollette, 1992, pp. 91-107). If you make decisions about who teaches what, you can assign desirable classes or lighter loads to favoured individuals. Every little thing in academia can be involved, such as priority in having your computer fixed, the words used when your work is cited in a paper and the tone of voice with which you are greeted 1 .
To list avenues is simply to point out that patronage is possible, not that it necessarily occurs or is inappropriate. Whether an action is considered appropriate or inappropriate depends on the culture in an organisation and society. In some academic systems, patronage is the standard way of doing things and peers would consider it highly unorthodox -even dangerous -to support an outsider instead of the usual insider. On the other hand, when expectations of equal treatment prevail -in other words, when meritocracy is the official doctrine -then giving favours may be stigmatised, sometimes through policies and guidelines and sometimes through peer expectations.
If patronage exists within an organisation or network, favourites are advantaged and hence others are disadvantaged. Often it is useful to divide members of a group into three groups: insiders, neutrals and outsiders. Insiders are the favourites. Neutrals are those who are treated on their merits for the most part, except when competing against insiders. Outsiders are actively discriminated against. Outsiders may include critics and challengers, stigmatised groups such as ethnic minorities, and those who do not fit in with the dominant style.
Many academics are members of two or more communities. For example, it is possible to be an insider within a department, a neutral in relation to university-level politics and an outsider in international professional circles.
Methods of active discrimination are just as varied as methods of patronage. They include subtle put-downs, petty harassment, lack of support, ostracism, and direct attacks including public humiliations, formal reprimands, demotions and dismissals. Often the easiest way to get rid of outsiders is to make life unpleasant so they leave in a way that looks voluntary. Potent techniques include preventing them from teaching desirable classes, giving them no recognition for achievements and denying them tenure.
One dynamic underlying patronage is control through obligation: by dispensing favours, others may feel indebted and are more likely to support the patron 2 . For example, members in a research team may co-author papers and cite each other's work, excluding outsiders from both collaboration and citations. Senior figures in the team assist junior researchers, who are likely to reciprocate, especially when they rely on continued opportunities for references, collaboration and access to equipment. Independent researchers have a hard time competing against such an in-group.
Patronage can be categorised into four main types (see Table 2). The first type is making decisions that favour insiders, for example in making appointments and awarding grants. Such actions are fairly obvious and, as such, may be susceptible to challenge in an equal-opportunity environment. Table 2.

Types of academic patronage
The second type of patronage is biasing the processes through which decisions are made, for example appointing insiders to a selection committee or, more subtly, establishing a process that will make appointment of such people to a selection committee more likely. For example, if most full professors are insiders, then requiring a committee to have a majority of full professors will create a bias, although the procedure itself is not biased towards individuals, only full professors. The advantage of this type of patronage action is that outside observers cannot readily see any bias: an intimate knowledge of the group or network is required to understand what is happening.
The third type of patronage is assistance to preferred individuals to develop their knowledge or skills. An example is providing inside knowledge about what is happening in the organisation, how decisions are actually made, the strengths and weaknesses of individuals, and likely developments in the future. This sort of information can greatly advantage those who have it. It is patronage only if it is inside knowledge, namely knowledge that others don't have. Sometimes information is made publicly available, in meetings or emails, but only insiders have the ability to decode it: additional knowledge is needed for grasping the significance of the information. Providing skills to preferred individuals is another facet of this type of patronage. This can include how to deal with conflicts, control emotions, present oneself in interviews, respond to challenges and much else. Again, if skill training is available to anyone on a genuine open-invitation basis, this does not constitute patronage.
The fourth type of patronage is personal interaction, for example the way a person is greeted in a private meeting or public occasion. It includes what is said, ranging from open praise to derogatory comments, and, just as importantly, the way things are said, including tone of voice and body language. Patronage at the interpersonal level is often quite subtle and can be hard to detect, easy to misinterpret, impossible to prove and easily contested. Despite its subtleties, personal interaction can be a powerful motivator for insiders and disincentive for outsiders.
It hardly seems fair to count the twist of a smile or the tone of a comment as manifestations of patronage; yet such minor features of personal interaction should not be discounted because of the difficulties of measurement and documentation. It is natural for most people to respond more positively to those they like and support and to those who respond positively to them. Interpersonal interaction often lays the foundation for other, more overt, types of patronage.
Table 2 also lists disadvantages for outsiders. In many cases, the word "disadvantages" is far too mild: the impact of prejudice, discrimination, bullying and other unfair and adverse treatment can be devastating (Twale & De Luca, 2008;Westhues, 2004;Wilke, 1979).  Table 3. Individual instances of patronage may occur randomly and without any coordination: advantaging a friend here, gender-based bias there, preference for one's student in yet another circumstance. On the other hand, when instances of patronage are systematic or coordinated, it is possible to speak of a patronage system or patronage network. There are two basic types, centralised and decentralised. In centralised systems, most favours are dispensed from a single individual or small core group; in decentralised systems, favours are dispensed at any level and by anyone according to common principles, whether overt or unspoken.

Type of patronage Advantage for insiders Disadvantage for outsiders
Some special types of patronage systems are listed in Table 4. This list is not exhaustive, nor are the types all mutually exclusive. Table 4.

Some types of patronage systems
Anti-performance: high achievers are marginalised or discriminated against because they threaten second-raters running the system. Bullying: out-group members are harassed until they leave or -in some systemsjoin the in-group. Discrimination: in-group membership is dependent on an attribute such as gender, ethnicity, nationality, class, discipline or educational background. Dynastic: patronage is organised around maintaining a dominant group over time.
Godfather: patronage is dispensed by a single individual who expects total loyalty in return.
Information: insiders are told what is happening and why, giving them advantages; outsiders are deprived of information or fed disinformation. Initiation: joining the in-group requires participating in rites of passage. Migratory: a clique moves from place to place, appointing in-group members, either seeking richer pickings or escaping sanctions for abuses. Rule-based: patronage is dispensed according to rules that systematically discriminate against out-groups.
Academic patronage can be considered a facet of in-group preference found at various levels, as shown in Table 5. Psychologically, it can be seen as a manifestation of the way people's sense of identity relates to their membership of groups that are socially significant (Tajfel & Turner, 1979). The contrary process is called here "equity". Table 5.

Forms of patronage and equity at various levels of social organisation
In this picture, patronage is simply another name for manifestations of in-group bias.
The different levels of patronage intersect and overlap, as shown by patronage in academia. The most common general level of organisation in academia is the university: patronage at this level means bias in favour of students, graduates or employees of one's own university. But within a university, individuals pursue their

Arguments against and for patronage
Patronage, by definition, is a violation of the merit principle: insiders are given advantages not deserved by their performance. In an equal-opportunity or meritocratic environment, this is the principal argument against patronage. There are some other arguments too. Some forms of patronage involve a conflict of interest, for example when academics make decisions affecting others with whom they have close personal relationships. Patronage can lead to undesirable consequences, including selection for mediocrity, formation of cliques whose self-seeking behaviour reduces performance, and a lack of diversity that causes a loss of creativity and insight.
Strangely enough, patronage is potentially damaging to patronees, the recipients of favours 3 . In the eyes of others, patronees may be perceived as unworthy precisely because they have sometimes received favours, even when in reality their performance could justify their success. Furthermore, the personal development of patronees may be stultified by an abundance of largess: they may not be presented with sufficient challenge on the path to advancement, missing opportunities for learning that arise from struggling against adversity.
Arguments for patronage are harder to articulate, at least when merit has become the dominant rhetoric. The usual argument for patronage is that insiders are better, but this isn't an argument for patronage but actually a use of the merit principle to deny and disguise patronage. A true argument for patronage justifies support for insiders when they are less accomplished than others. The most common justification is implicit: it's a tradition or an expectation. The eldest son inherits the family business because it's always been done that way: it's assumed to be right and proper; it's a cultural expectation.
When academics receive helpful mentoring, co-author papers with supervisors and obtain jobs with the assistance of good references from supervisors or other supporters, it is natural for them to continue the same sorts of practices, for example by seeking scholarships for favoured students and offering them work as research assistants and tutors, and, when in a more senior position, providing opportunities for junior academics. In this sort of way, patronage is reproduced and seems entirely natural: it seems to be the way academia operates and therefore needs no justification 4 .
Within an academic context, patronage may be justified on the ground that insiders fit in: they help maintain a harmonious work environment and thus increase productivity. A dysfunctional personality may disrupt the group. The trouble is that this rationale is easily used to discriminate on all sorts of grounds.
Another argument is that insiders are trustworthy. They will support the goals of the organisation. They are team players. They won't rock the boat. This sort of rationale may be more common at higher management levels in which social characteristics are more important and performance is harder to judge as an individual accomplishment.
Another argument for patronage is that it works: if income or student numbers or publications are buoyant, this provides an apparent justification for prevailing patterns of internal decision-making. However, seldom is there any benchmarking with other units that have equivalent starting points and similar financial and other support.
There seem to be few empirical studies of the actual performance of patronage systems.

Systems with patronage
In any given academic organisation, it is likely that some types of patronage will be stigmatised and rare while others are accepted and common. Nepotism, with family members gaining advantages over more qualified outsiders, is not so common at the level of academic appointments. It is an obvious form of patronage and hence easier to challenge.
Gender-based patronage is one of the most long-standing forms, directly deriving from patriarchy. In what is called homosocial reproduction, men appoint others who are like themselves -and that usually means men (Kanter, 1977). In the "two-person career", wives or partners assist men to succeed in an academic career (Fowlkes, 1980;Papanek, 1973), with a few instances involving a reversal of the sexes. In peer review, women may be judged more harshly than men (Wenneras & Wold, 1997). Most men receive advice and mentoring from other men whereas women are more likely to be excluded from support systems (Bagilhole & Goode, 2001). At higher levels of management, male characteristics and styles are commonly considered superior. Despite the efforts of feminists, including decades of efforts for equal opportunity and affirmative action, women are still disadvantaged, especially when they are parents. However, within this overall picture, there are pockets of patronage advantaging women, instigated by other women and sometimes by men.
Decades ago, women were considered to be intellectually inferior and barred from many occupations: academic patronage for males was institutionalised. Although beliefs in male superiority persist in a few disciplines such as physics, today in many countries official policies support equal opportunity.
A similar history applies to discrimination on the basis of family, class, religion, ethnicity, schooling, disability and sexuality. Each of these still occurs, but in equalopportunity and meritocratic environments are increasingly stigmatised through official policies and prevailing attitudes.
In some places, a more recent addition to the list of officially stigmatised forms of patronage is that based on so-called close personal relationships, of which sexual and romantic relationships are a common type but by no means the only form. A friendship can be a close personal relationship. If your best friend is applying for a job, you are unlikely to be, or be seen to be, an objective member of the selection committee. However, friendships are hard to quantify, so bias on this score can be hard to identify and police.
Patronage frequently occurs on the basis of disciplines and theoretical orientations. In a school of philosophy and politics, philosophers are likely to favour appointment of more philosophers rather than political scientists. In an economics department, econometricians are likely to support appointment of more econometricians rather than political economists. Likewise with pure mathematicians, organic chemists and environmental lawyers. Battles within and between departments often follow the lines of disciplines, specialities and allegiance to methods and perspectives (Becher & Trowler, 2001).
One of the accepted forms of patronage is support for one's research students, including training, advice, references and job opportunities. It seems natural to give assistance to your own students; this is especially tempting when they are pursuing research aligned to your own. Is this inevitably patronage in the sense of giving these students an unfair advantage? Not always: if all students have equal access to quality supervision, then no one obtains a special advantage. If every supervisor writes equally objective references and has the same level of connections for helping students obtain jobs, then no one obtains a special advantage. But of course this is seldom the case. Supervisors with experience, power, prestige and well-developed networks can do far more for students: their advice and insights give their students a special edge in their careers and their colleagues and supporters are more likely to give these students opportunities. Furthermore, when students follow in the footsteps of their supervisors in terms of research ideas and style, supervisors are more likely to be helpful. (A rebellious student is far less likely to obtain a high level of support.) So in practice supervision is commonly linked to patronage and sometimes to patronage networks based on loyalty to individuals and research traditions.
Struggles against patronage occur at several levels. The narrowest level is challenging an objectionable instance of patronage, for example appointment of an under-qualified man over highly qualified women applicants. Next are challenges to forms of patronage, for example trying to reduce discrimination against women by encouraging women to apply for jobs and ensuring that women sit on all selection committees. Finally there are challenges to standard practices that seek to redefine them as unacceptable. This occurred when exclusion of women from academic posts was challenged. It conceivably could happen with supervisor patronage, for example by saying that supervisors cannot write references for students, the way that today it is seen as inappropriate for a parent to write a reference for a child seeking an academic post.

Systems without patronage
In this picture, the concept of patronage is so broad that it is hard to imagine a world without it, so it is worth giving a few examples. In buying groceries, most customers receive no special deals. In a small corner shop, the owner might give family members free or low-cost food -a classic form of patronage. But in supermarket chains, these sorts of special favours are less common; indeed, family members of the store manager may receive no gifts, certainly not openly, because it might alienate regular paying customers and expose the manager to discipline from higher-level executives. In some shops, employees can obtain special deals on merchandise, but this is separate from favouritism toward customers with connections.
In buying a meal at a restaurant, patronage occurs when special favours -better service, better food, complimentary meals -are offered to people because of who they are, for example family members, police or politicians. In a high-patronage establishment or environment, all sorts of people obtain special favours. If the police receive free meals in a de facto exchange for not prosecuting violations of by-laws, this can be called corruption. Patronage is less likely in highly bureaucratised, highvolume restaurants like McDonald's.
An academic example is students obtaining admission to universities. In a fully bureaucratised process with no special favours, every student is treated identically: admission does not depend on who the student is, only on their grades, test scores and other accomplishments. Many universities have forms of institutionalised patronage, such as easier admission for children of graduates (a common practice in the US, called "legacy preferences"), for local students compared to foreign ones and for students able to afford high fees. Out-groups may be discriminated against, for example on the basis of ethnicity or religion. There can also be individual patronage, for example easier admission for children of the rich and powerful, especially relatives of generous donors.
These examples suggest a negative correlation between patronage and the processes of rationalisation, bureaucratisation and marketisation. The more an organisation resembles a classical bureaucracy (Hummel, 1977;Perrow, 1979), in which employees are interchangeable cogs, the less likely patronage is to flourish, and the more an organisation deals with clients or customers as in a perfect market, without discrimination on the basis of extraneous characteristics, the less frequent patronage will be.
However, bureaucracies and markets seldom conform to theory -they routinely contain forms of bias, and indeed institutionalise them. For example, a university admissions system run purely on market principles to maximise income might give systematic advantage to children from wealthy families; if the student market is built around maximising the scores of incoming students, then a different bias exists. If the goal is giving equal opportunity to students on the basis of how they would perform under equal conditions, then yet a different bias -involving forms of affirmative action -would exist.
The point here is that although there is a link between opposing patronage and bureaucratisation and marketisation, the rise of bureaucracy and markets does not automatically eliminate patronage. For example, it is well known that patriarchy easily coexists with bureaucracy and markets. Patronage needs to be examined on its own terms.
In the past several decades, higher education in many countries has experienced increasing rationalisation, bureaucratisation and marketisation, partly as a result of massification, namely a vast expansion, and partly through competitive pressures linked to globalisation (Marginson & Considine, 2000). Whereas universities once served as upper-class entry to the ranks of lawyers and clergymen, they have now become closer to an extension of high school, a general screening and training venue for a wide range of occupations. This means that previous forms of academic patronage -notably on the grounds of gender, religion, class and ethnicity -are now more commonly opposed. On the other hand, commercialisation of higher education expands the role of a long-standing incentive for and tool of patronage: money. Departmental budgets, salaries, bonuses, salary increments, research grants, consulting fees and investment opportunities can be avenues for giving favours.
Despite the continuing rationalising processes, universities contain many craft elements, especially at the research training level -supervision of research students is individualised. Likewise the appointment process is jealously guarded by local members of disciplines, who claim exclusive rights to judge the quality of others in their fields. These areas offer rich opportunities for patronage. Indeed, many would see a system without patronage as alien.
Imagine an academic operation in which decisions are made without regard to the special characteristics of an individual, only their performance: • Papers submitted for publication would be blinded for referees, as is currently standard practice for many journals, especially in the humanities and social sciences. Furthermore, the author's name would be hidden from the editor by use of an intermediary service, to limit the editor's ability to select referees to give a preferred result. Alternatively, referees would be chosen using an algorithm, for example randomly from a bank of people in the field. Not only would the author's name be removed from manuscripts as well as any selfcitations, but a textual anonymiser would be applied to hide peculiarities in style. For certain types of technical papers, method sections would be judged separately so that results would not bias referees.
• Selection panels for posts would be entirely composed of strangers to the organisation. The selection committee might be asked to select the top candidate, but not told the institution to which the applicant is being appointed. Interviews would be conducted electronically with the identity of applicants disguised by voice-modifying software. Curricula vitae would be anonymised by replacing achievements with generic equivalents, for example a degree from Stanford being listed as a degree from an A-level university -though some might argue that degrees should be ignored in favour of outputs -and a paper in Physical Review being listed as a paper in an A-ranked physics journal. Referees who know the candidate would be ruled out; referees might be chosen randomly from a panel, or refereeing replaced by a computer-based assessment of performance.
• Applications for research grants would be scored by six to ten assessors.
(Alternatively, applications would be given a score based on prior publication and other outputs by the applicant.) The median score would then be used to award a specified number of ballots in a grant lottery. For example, a median score of 80% might give 8 ballots and so forth. Grant recipients would be chosen using the lottery.
What these sorts of hypothetical processes have in common is an attempt to eliminate the role of personal knowledge so that decisions are made using only information about performance. The result may seem absurd today, but perhaps no more absurd than attempts decades ago to eliminate bias on the basis of gender and other personal characteristics (which of course have not been entirely successful). In some places now, applicants for posts or grants do not have to reveal their gender, nationality, date of birth or marital status: these are considered irrelevant considerations in making a decision. When referees do not know the identity of a paper's author, the gender of the author cannot easily play a difference in the referee's response.
Decision-making processes that eliminate knowledge of who applicants are will be opposed by those who want to control the selection process to ensure they are comfortable with the outcome. This is most common in selection for jobs: candidates are sought who will fit in, namely ones incumbents will feel comfortable with. Everyone wants to avoid having to work with a "difficult personality". The trouble is that making compatibility an implicit criterion for appointees opens the door to a host of forms of patronage based on gender, age, discipline and personal style.

Alternative futures
The vision of academic life without patronage may or may not be attractive, depending on one's location and values. In highly rationalised systems like Australia's, getting rid of remaining patronage practices might be logical from a merit perspective, but brings the spectre of a fully bureaucratic, market-driven university in which academics have no power but are merely cogs in a soulless system that churns out graduates and research findings. Having the power to sponsor one's students and choose one's colleagues might seem to one of the few remaining academic privileges in market-driven higher education systems 5 .
From the point of view of those excluded or subordinated by these privileges -for example the talented scholars who can not obtain jobs because competitors have patrons, or junior researchers who have to please team leaders in order to get ahead -any concerns by established academics about loss of their privileges might seem precious. Nevertheless, there is a deeper issue here: is there any way to reconcile the abolition of patronage with a workplace that maintains and strengthens collegiality and mutual support? If the only vision is McDonaldisation of higher education, the situation would be bleak. But there are other alternatives. Here I outline three.
One direction is diversity of institutions and units. Merit can apply within a unit, but units can have quite different goals and expectations. In Australia, all universities are quite similar in their standards, size, mix of disciplines and orientation to teaching and research. In the United States, in contrast, there is a much greater institutional diversity, ranging from elite research universities to community colleges, from enormous state universities to tiny places with only a few hundred students, from research institutions to liberal arts colleges, from mainstream educational approaches to radical alternatives. Among elite institutions, the expectations of academics at Sarah Lawrence College are quite different from those at Caltech. For students, the course structure at the University of Texas is different from that at Evergreen State College. The point of this example is not to hold up the US higher education system as a best practice model, as it has myriad problems, but to illustrate that institutional diversity is one possible future that could combine elimination of patronage with opportunity for collegiality and local control.
However, it should be noted that anti-discrimination initiatives are antithetical to institutional diversity, and many would say rightly so, as when in the 1960s US federal funding was made contingent on desegregation. Institutional diversity, to be compatible with fairness, needs to provide equal opportunity system-wide, for example at equal-quality male-only and female-only units. The more usual pattern is a combination of diversity and unequal opportunity.
Another alternative is an egalitarian educational model. Abbs and Carey (1977), for example, proposed a "new college" in which students and staff jointly make all decisions. In their vision, the curriculum and learning methods would be cooperatively decided, staff have equal salaries and status, and the small institution would be partly self-sustaining through practical work, for example growing food, in which everyone would participate. Abbs and Carey's new college is the antithesis of market-driven uniformity: instead, staff and students are highly bonded in a common enterprise of learning. This model would not be attractive to many of today's academics because it would mean giving up the privileges of salary and power over systems for student learning, but it could be deeply satisfying through personal engagement in meaningful processes for personal growth and mutual support. Egalitarian education, by eliminating academic privilege, eliminates many opportunities for patronage.
A third alternative can be called community education in an intellectual commons. Students, rather than studying a formal syllabus, would develop their own learning plans, which might combine some formal instruction, some self-education (potentially drawing from online curricula) and some practical learning through work and community education, the latter in the spirit of Ivan Illich's concept of deschooling (Illich, 1971). Research would include some traditional targeted projects and some collective enterprises along the lines of free software and Wikipedia (Surowiecki, 2004). Scholars could combine in any way they pleased -for teaching or researchpossibly in collaboration with students and amateurs. There is no standard funding model for this alternative: it could be based on projects or contracts, or based on a guaranteed annual income. The basic idea is that learning would be decoupled from the formal apparatus of courses and degrees and contributing to knowledge creation opened to anyone irrespective of training and credentials.
These alternatives, however unlikely as images of the future, illustrate that intellectual activity can be organised in various ways that avoid the poles of patronage and hierarchical rationalisation. In each alternative, scholars would have the opportunity to engage with like-minded colleagues with some degree of collective control over their activities. But there is a penalty, at least in alternatives two and three: the status of academics would be considerably less. This is a pointer to the core dilemma facing any defence of patronage: in most fields, being a tenured academic is a privileged occupation, with many more capable aspirants than positions.

Challenges to patronage
Patronage can be undermined in various ways. At the policy level, processes are crucial. Examples are journals setting up blind refereeing processes, thesis evaluation procedures ruling out certain examiners and job advertisements being written in a specified fashion. However, policies seldom cover all facets of a process. For example, even when journals use blind refereeing, the editor normally knows the identity of the author, can be influenced by author attributes such as their university affiliation, can select referees predisposed for or against a submission, and can interpret or overrule referee recommendations if desired. Seldom is there any oversight to deal with editor bias. No doubt most editors try to be objective but good intentions do not guarantee a fair outcome.
Many organisations have policies about job selection, for example about the content of job advertisements and the membership of selection committees. But such policies only control some types of bias. Seldom do policies specify enough detail to prevent patronage. A typical scenario is that insiders want to appoint a particular person, often someone already in the position or sometimes a known external person. This can be facilitated by advertising a position in the person's field, writing the advertisement in a way tailored to the person's strengths, picking the selection committee to be sympathetic to the person or acquiescent to patrons of the person, manipulating the committee procedures to rule out competitors, and giving the desired candidate inside information about what to put in the application and say in the interview. Everything is done by the book but the job is stitched up.
Policies and procedures are an important step in reducing patronage but just as important are attitudes and prevailing practices. Attitudes are crucial. Old-fashioned prejudice against out-groups needs to be replaced with personal sensitivity to possible bias. Instead of thinking how to give an advantage to friends and treasured colleagues, thinking might go along the lines of "I'd like them to succeed on their merits". Instead of thinking how to wrangle a scholarship for a favoured student or a job for a mate, it would become more natural to think of all the other students and potential applicants out there -many not even known -who deserve a fair chance.
Prevailing practices make an enormous difference to patronage. Even a single instance of apparent jobs-for-the-boys at the top of an organisation can set the tone for everyone else. Setting an example means doing things in a way that is seen to be fair as well as being fair.
Patronage systems are easily entrenched: policies, attitudes and practices reinforce each other. Changing all of these simultaneously can be extraordinarily difficult.

Examining theses: a case study
In Australia, many undergraduate degrees require three years of full-time study, with an optional fourth year called honours designed for students seeking advanced training in a discipline 6 . The honours year normally includes coursework and a research component involving writing a thesis. Depending on the university and discipline, the thesis may comprise anything from 25% to 100% of the honours course; a typical length for a thesis comprising half the honours mark is 20,000 words.
For students pursuing an academic career, honours is the most common route, and it is a crucial year: a first-class honours result is usually enough to obtain a PhD scholarship but anything less makes subsequent efforts much more difficult. The honours result therefore is a potential point of patronage: even a slight advantage to a student can make a career-changing difference.
Honours theses are usually supervised by a single academic in the field. Supervisors vary considerably in their talents and interest in their role. Some give day-to-day support in everything from thesis topic to grammar. Others take a more stand-offish attitude, letting the student sink or swim based on their own efforts. Some supervisors who take an intense personal interest in a student's progress are grooming the student for success. This might be called patronage, especially when the student's topic follows the supervisor's research agenda or theoretical predilection.
Honours theses are marked by examiners; patronage can occur via the choice of examiners. Consider the following possibilities: • Examiners are chosen randomly from a list of senior academics from other universities. These academics have no interest in supervising additional PhD students.

•
Examiners are chosen by a departmental committee from academics they know or know about at other universities.

•
Examiners are chosen by the supervisor from academics at other universities.

•
Examiners are chosen by a department committee from academics in their own department.

•
Examiners are chosen by the supervisor from colleagues.

•
The supervisor is an examiner.
When the examiner has a closer relationship -actual or potential -to the student, patronage is easier to dispense. This is most obvious when the supervisor is an examiner: if the supervisor personally likes the student and expects that the student will follow the supervisor's research path, then it is very difficult to avoid being supportive. (When the supervisor and student don't see eye to eye or have had a falling out, having the supervisor as an examiner may disadvantage the student.) When a supervisor's colleagues are examiners, they may be influenced by knowing the student personally, perhaps as a teacher in an undergraduate class, or influenced by their relationship with the supervisor.
Examiners also can be influenced by their potential future relationships with students, most importantly by being a prospective PhD supervisor or, more generally, by being a like-minded person in the field, perhaps seeing the student down the track as someone to interact with as a postdoc, a colleague or an important researcher. If the student looks like fitting any of these roles, an examiner may feel extra affinity with the student and be a little bit more favourable in judging the thesis.
In many cases, these considerations are not significant for career purposes, either because the student's performance is not deemed as first-class by even the most sympathetic examiner or because the student's performance is judged as outstanding by even the severest critics. Where bias plays the biggest role is at the margin, when a few extra marks make the difference between an honours result that garners a PhD scholarship and launches a research career and one that leads away from academia.
Administratively, having supervisors as examiners can be said to be a conflict of interest. This is especially true when supervisors have had a strong input into deciding the thesis topic, assisting with data collection or analysis, and polishing the expression. How can supervisors be objective when assessing projects to which they have had considerable input? Having supervisors as examiners is also undesirable pedagogically, because students may be hesitant to be open about their doubts and difficulties, worrying that their supervisors may think less of their abilities.
Being an examiner for one's own student can be awkward. If I try to be scrupulously fair, I will take account of my own bias and, assuming I'm sympathetic to my student, try to counteract this bias by being a bit tougher than I would be with an unknown student. But this creates a new unfairness. If I think my colleagues, examining the theses of their own honours students, are likely to give them an extra advantage or to be strong partisans on their behalf, then by being unbiased myself I'm actually disadvantaging my own student. For example, in writing a reference, if I say my student is in the top 25% while my colleagues rate their own students as in the top 1%, I've penalised my student by being honest -or even by being partisan, just not as blatantly partisan as my colleagues. So to give my student a fair chance, I need to be partisan myself. These dilemmas disappear when examiners have no personal connections with students.
At the University of Wollongong, in quite a few departments it used to be common for honours supervisors to be examiners. This was seen by some as a conflict of interest. Eventually a policy -driven by Deputy Vice-Chancellor Rob Castle, among otherswas introduced specifying that supervisors could not be examiners. However, a couple of departments were so resistant to this policy that they changed the rules for their honours degrees, reducing the weight of the thesis component so it avoided the new regulation. Their argument was that only the supervisor had the unique specialist knowledge to be able to judge the thesis.
Meanwhile, regulations for examination of higher degree theses at Wollongong have gradually been tightened to ensure that examiners are independent of the supervisor and student. Normally two examiners are chosen. For PhDs, both have to be outside the university and cannot be recent graduates. No one who has collaborated with the supervisor or student within the previous five years can be an examiner. A new rule is that a person cannot be used as an examiner more than once a year, intended to prevent research groups using a trusted ally to approve thesis after thesis. The guidelines for selecting examiners give the reason for these and related rules: the intention is to find people who are "at arm's length from the supervisor/s and candidate so as to ensure the maximum degree of objectivity".
Another innovation at Wollongong has been to have research students involved in nominating examiners. The supervisor and student select four or five potential examiners and sign off on the names; the list has to be approved by the academic unit's head of postgraduate studies. The supervisor and head of postgraduate studies then recommend two of those examiners to the director of the Research Student Centre, who will approve them if they meet all selection criteria in the university's guidelines. One goal of this approach is to ensure that unsympathetic people are avoided: essentially, the supervisor and student each have a veto power. In earlier years, when the student was excluded from the process, with the examiners chosen by the supervisor or the department, there was greater scope for pressuring students to conform to the supervisor's research orientation. This process of involving the student in examiner selection has recently been introduced at honours level.
The Wollongong trajectory for thesis examination has been one of gradually excluding obvious conflicts of interest -especially examiners with strong connections with the supervisor -and involving students in the selection process. The current policies discourage patronage, ruling out supervisors and supervisors' close colleagues as examiners. Obviously there is still scope for patronage: a supervisor and student could conspire to choose examiners who are friends or allies, but whose formal relationships are compatible with the guidelines. Nevertheless, the trend over the years has been to minimise opportunities for blatant patronage.

Conclusion
The concept of patronage, taken in a very broad sense, can be used to capture many forms of bias, discrimination and conflict of interest. Academia is commonly seen as an arena in which merit plays a stronger role than in many other occupations.
Although challenges have been made to many forms of academic patronagenotably discrimination on grounds of gender and ethnicity -many others remain, often without much thought about problems or alternatives.
A vision of academic life without patronage can sound utterly alien, for example with elaborate measures to ensure anonymity and independent judgement. To propose such measures may make one sound like a zealot and, more seriously, to be implicitly criticising one's colleagues as not being objective. Furthermore, a fully rationalised university, run like a factory line, would be unattractive to the workers, for whom collegiality and a sense of common purpose make a huge difference. Reconciling elimination of patronage with satisfying work is a challenging task: possible directions include greater institutional diversity, small egalitarian cooperatives, and embedding scholarship in community activity.
An important lesson from the experiences of academic patronage is how difficult it is to bring about change. One reason is straightforward: many academics want to retain their personal capacity to distribute favours, to friends, allies in their field and to students. Research teams are commonly built around powerful leaders who reward those committed to the research directions on which the leaders have built their reputations and sometimes those committed to the leaders personally. Sometimes such research empires are built on discrimination by gender, ethnicity or nationality, but even when these play little role, personal and professional favouritism can be significant. There have been major movements for sexual and racial equality but at the moment there seems to be no equivalent movement to challenge more personal forms of patronage.
in academia, and academic-activist links. He is also the author of 12 books, most recently