Individual and collective outcomes of higher education: a comparison of Anglo-American and Chinese approaches

ABSTRACT The paper compares Anglo-American and Chinese approaches to the outcomes of higher education, primarily but not solely collective outcomes, by examining the Western domain of ‘public good’ and ‘public goods’ and parallel or near parallel activities in China. It reviews scholarly discourses of society, state and higher education in the respective political cultures (‘traditions’), including individualism and collectivism, university autonomy, the critical function, higher education in civil society, and global tianxia and global common good. A key issue in symmetrical cross-cultural comparison is the position from which it is made; and as well as elucidating similarities and differences the paper develops what Sen calls a ‘trans-positional’ view based on integrating the two positional views. The two traditions are not closely aligned. However, aside for the Anglo-American public/private dualism in economics (which occludes collective outcomes), all ideas in both traditions can contribute to transpositional understanding of the individualised and collective outcomes of higher education.


Introduction
What are the outcomes of higher education and how are these understood? These are ongoing questions for governments, scholar-researchers and those who work in the sector or use it. The outcomes of higher education are multiple and complex. Public financing economics has devised tools for measuring some but not all of them (McMahon 2009). Even so, they are specific to institutional mission, vary according to values and priorities, and are interpreted differently between countries. The core topic of this paper is similarities and differences in national-cultural understandings of the outcomes of higher education.
One aspect is common across nations. The outcomes of higher education can be understood as individualised and/or collective (Marginson 2020), even though as shall be discussed, nuances of 'individual' and 'collective' vary. First, individual students acquire benefits distinctive to themselves as single persons, as self-development, knowledge, skills, employability, larger earnings or social status. Some of these individualised benefits are measurable, such as the augmented lifetime earnings associated with graduateness, or access to credential-mediated occupationsthough higher education's larger contribution to individualised agency and capability (Sen 1999a) is harder to assess. Second, higher education contributes to common social resources not confined to specific persons, while impacting the lives of many: for example new scientific knowledge, faculty expertise that supplements the work of government, the social spread of scientific literacy and technological expertise, provision of equitable social opportunities, the preparation of graduates as relational citizens. Some of the individualised and collective outcomes overlap. For example, the preparation of graduates as citizens fosters both individualised capability and collective social relations.
Though both individualised and collective outcomes pose challenges of interpretation, collective outcomes are more difficult (Marginson 2016(Marginson , 2018a. They elude standard treatment. In part, it is a problem of observation. For example, how do we assess and compute the contributions of international education to social tolerance and cross-border understanding? In part, it is a problem of vision, lenses and instruments. For example, economics, sociology and psychology focus on different qualities. In part, it is a problem of national-cultural variation. Ideas, practices and valuations of the collective are determined by the society in which higher education is embedded. Consider the differences between the egalitarian Nordic welfare democracies, the hyper-market and civil society in the United States (US), Singapore's managed global society-economy, and China's partystate order.

Purpose of the paper
This paper compares Anglo-American and Chinese approaches to the outcomes of higher education, by examining the respective political cultures ('traditions'), and higher education within them. It focuses primarily on collective outcomes, which are more illuminating of the global diversity than cross-national comparison of, say, individual rates of return. As a way to collective outcomes, we examine the domain designated as 'public' in Western countries (Dewey 1927) and parallel activities in China. (The implications of using an Anglo-American starting point are discussed below). The Western 'public' is not identical to 'collective'. Some definitions of 'public' include individualised public goods, or privately controlled collective goods. However, 'public' and 'public/private' are useful starting points.

Definitions
By 'political culture' is meant the compound of words, ideas, policies, institutions, regulatory structures, resource configurations and subjectivities that together constitute the social order as relations of power. Social and educational relations are here observed through the lens of political culture because higher education is embedded in government (though the extent of state-determination is one of the variables in play).
The term 'Anglo-America' refers to the US and United Kingdom (UK). 'Higher education' includes all of systems, institutions and disciplines. 'Tradition' refers to current practices that may draw on a long lineage (see below). 'Parallel activities' acknowledges that there are no exact equivalents of Western 'public' and 'private' good(s) in China, though some Chinese concepts overlap with or correlate with heterogeneous Western concepts.

Rationale
Why compare the Anglo-American and Chinese approaches? There are perennial reasons for crosscultural work and reasons for the specific comparison.
First, higher education is partly globalised (Marginson 2011) and its national systems and leading institutions are in ongoing contact. The global science network is expanding rapidly (Wagner, Park, and Leydesdorff 2015). All else equal, convergence in understandings of the outcomes of higher education, including the diversity, facilitates cooperation. Similarities and differences in higher education have practical implications. For example, US-China and UK-China relations are important at global level. In 2018, researchers from the US and China co-authored 55,382 science papers, much the largest collaboration in world science (NSB 2020), yet there are ongoing tensions between the respective governments which could disrupt relations in higher education, science and technology (Lee and Haupt 2020). The development of better knowledge of each other has strategic significance, especially in the Anglo-American countries: Chinese language, ideas and institutions are less well known in the English-speaking countries than these countries are known in China.
Second, such comparisons help the parties on each side to better understand themselves. As Walter Scheidel (2015) states in introducing a comparative review of the Han and Roman empires, comparison provides a way out of parochialism: 'Comparison of alternatives makes the characteristics of one's "own" case seem less self-evident and helps us appreciate the range of possible alternatives' (3). In this regard, the comparison between China and Anglo-America is especially fruitful because multiple differences enable a broad reach, though this creates methodological challenges, as discussed below.
A third reason is that a comparison across diverse traditions helps us to distinguish the common from the variant elements in all higher education. This is most helpful in relation to the difficult problem of collective goods. The possibility that more than one tradition can contribute to higher education studies is intellectually liberating. Although Chinese scholars often draw on both Western and Chinese ideas few Western scholars have done so.
The remainder of this introduction expands on the comparative method and its limits. The next two sections review and distinguish each of the Anglo-American and Chinese traditions, in general and in higher education. The section that follows explores the similarities and differences between them. The conclusion reflects on the comparison.

Theory, method and limits
Our approach is grounded in critical realism (e.g., Sayer 2000). Social relations exist prior to our knowledge of them. The goal is explanation of those relations, including causal explanation. Society is only partly observable using empirical techniques and cannot be exhaustively explained by an assembly of facts alone, necessitating theories and interpretations which must be continually tested and refined. In this paper, our theoretical framework derives from Amartya Sen's recognition of plural cultural identities (e.g., Sen 1999b) and our method of inquiry is taken from Sen's (2002) idea of 'trans-positionality'.

Multi-positionality
The application of Sen's trans-positionality is premised on three steps of reasoning. First, rejection of comparison based on a single cultural standpoint or position. Second, exploration of multiple positions, in this case two: the Anglo-American and the Chinese (Sinic). Third, development of a transpositional assessment. As Sen states: Observations are unavoidably position-based, but scientific reasoning need not, of course, be based on observational information from one specific position only. There is a need for what may be called 'trans-positional' assessmentdrawing on but going beyond different positional observations. The constructed 'view from nowhere' would then be based on synthesizing different views from distinct positions … A trans-positional scrutiny would also demand some kind of coherence between different positional views. (Sen 2002, 467) An older comparative social science sought to explain all societies in terms of Anglo-American or Western norms and trajectories that were seen as universally applicable (Beck 2016). When 'methodological nationalism' (Wimmer and Schiller 2002;Shahjahan and Kezar 2013) frames the comparison, it blocks from view features of the other culture that are different or reworks similarities as isomorphism and differences as pathologies. The multiplication of perspectives is on one hand an act of power, taking in more of the world; on the other an act of comprehension, enlarging the scope for explanation. How then to accumulate and combine the actually existing diversity without negating it? This paper uses multi-positional observation to view parallel phenomena in each setting, privileging one over the other as little as possible. Material can be scrutinised from either starting point. After the multi-positional observations, we develop a transpositional conclusion.

Method
The specific study is of discourses of society, state and higher education, especially of the key animating ideas typical of the two traditions. 'Discourses' refers to scholarly treatment of the social domain, some of which enters policy literatures. We understand discourses as 'knowledge formations' (Bacchi and Bonham 2014) that bridge words and material activity; and also bridge social structure and human agency, being called to the service of each. For Foucault (1972) discourses are more than groups of signs. They are 'practices that systematically form the objects of which they speak' (49). The world is shaped by dyads of words and objects. Ideas matter when they are manifest as discursive practices in systems, institutions and behaviours. However, in the present study, the account is biased to the words side of the dyad in each tradition. We do not match this with a review, for each side of the comparison, of the material practices of higher education. A double comparison of words and objects would be a large study but without it the comparison is incomplete. This is a practical limit of the paper.

Methodological challenges
Further limits are more fundamental to the method. The first is the inherent difficulty of achieving Sen's transpositional 'coherence' across the positional views. A symmetrical comparison based on a common template is impossible. Traditions and discourses do not fall into neat lines of equivalence in internal composition, categories, external linkages and temporalities (Foucault 1972)and in this case, the normal problems of comparison are magnified. There are deep differences between Anglo-America and China, much older than the assumption of power by the Communist Party of China (CPC) in 1949. 'Divergent paths were taken at a number of crucial moments in the development of Chinese and Western cultures. The consequence of this divergence is that the problematics of Anglo-European culture and that of China are really quite distinct' (Hall and Ames 1995, xiii-xiv). This underlines the fact that neither culture can be understood solely and completely through the lens of the other. Though there are certain convergences in the two kinds of modern higher education, the underlying cultural differences lead to contrasting language, concepts, sensibilities and practices of society and of the outcomes of higher education, especially but not only the broader 'public', social and collective aspects.
The two sets of discourses embody different premises and modes of thought and all of these elements are always becoming, evolving (Hayhoe 2017;Hall and Ames 1995). Analytical-rational Western reasoning employs singular and bounded abstractions and fixed categories and elevates theoretical knowledge above practical knowledge. There are diverse claims to universal truth within the canon, yet each is pursued confidently, and most Western thinkers believe that they have the means 'for assessing the value of cultural activity everywhere on the planet' (Hall and Ames 1995, xiv). The older Chinese tradition fosters greater conceptual openness, more extensively employs analogy and correlation rather than linear causal reasoning, naturalises process and change rather than fixing the being and quality of things in the manner of Parmenides and Plato, and places a high value on knowledge for practical uses. It more readily combines heterogeneous ideas, like Confucianism and Daoism, and sustains a resilient continuity in which past ideas are not displaced. Since the final decades of the last Imperial dynasty, the Qing Dynasty (1636( -1911, this robust openness has facilitated partial Westernisation (Qin 1981;Huang 2000;Xia 2014). The Sinic imaginary is layered by ideas from different eras, including Western Zhou Dynasty (1046-771 BCE) statecraft, the successive iterations of Confucianism and monastic Buddhist scholarship, plus Marxist-Leninist Westernisation led by the party-state, American Westernisation, and individualist-consumerist modernity. The endogenous element is especially apparent in the way discursive practices are combined.
While the paper is focused on present approaches to higher education, the differing temporalities, and the varied openness of each tradition to the other, have governed the way the comparison is developed. Despite its long roots, the contemporary Anglo-American tradition took shape in the post-Reformation period and in successive iterations of liberalism. It is more readily discussed as a single piece. The Chinese tradition is a double tradition: on one hand, the long evolution of indigenous language and thought in the Imperial period; and on the other, the modern period of Western influences and hybridities, in which the earlier discourse is still generative in words and actions. The substance of the comparison with Anglo-American 'public' and 'public/private' changes in modern times.
Second, the transpositional method is limited in that prior to the final 'transpositional assessment' it cannot be executed in pure form. There must be a starting point, a basis of comparison. Inevitably this rests on a culturally specific position. In this paper, the comparison starts from the Anglo-American side. It works in English with the Anglo-American idea of 'public' and looks for similarities, parallels and differences in China. If the comparison was Sinic-led the paper's content would be somewhat different. Anglo-American liberty and civil society would partly recede. Confucian humanism (ren), state-managed order, and perhaps relations between humanity and nature, would become more important. There are practical reasons for starting from the Anglo-American side. The modern West influenced China more than vice versa. Discussion of higher education in China is often conducted in primarily Western terms, with Chinese additions and caveats. Nevertheless, it is essential to remember that ideas of 'public' in society and higher education are grounded in the heritage of Hellenic-Judeo-Christian Europe and its settler states, including Anglo-American economics and the civic republicanism of the French and American Revolutions.
There is a third limit in the phrase 'typical of the tradition'. It is ironic that in a paper designed to highlight the diversity between traditions we must limit diversity within them. In China the focus is primarily on Confucian and party-state practices, occluding many other currents, layers of change still present and regional variations. The account of Anglo-American tradition is also narrowly mainstream: at greater length it would expand on social democratic ideas, differences between English and American liberalism, and how some elements of Anglo-American political culture have derived from elsewhere in the West, especially Republican France. Bildung, a generative influence in educational thinking in Britain and the US (Dewey [1916(Dewey [ ] 2011Sijander and Sutinen 2012), is from Germany.

Anglo-American ideas of 'public' and higher education
This section focuses on Anglo-American discourse on the outcomes of higher education. After a necessarily short and schematic summary of the Anglo-American social imaginary and political culture, to underpin the comparison with Sinic discourse, it examines the plural uses of 'public' and the public/private distinction in general and in higher education.
The Anglo-American social-cultural imaginary

The state
Western governance is rooted in divided powers, and despite periodic attempts to establish absolute rule, from feudal monarchs to twentieth-century dictators, the Western state is essentially a limited state and reverts to that default position. This is the legacy of distributed political agency in Republican Rome, the post-Roman division between church and state, the autonomy of mediaeval cities and merchants, and the evolution of the law and later, electoral politics, as both outside executive authority and in continuing relation with it (Scheidel 2019). Perhaps the key moments were the republican revolts against absolutist monarchy in seventeenth-century Holland and England, and eighteenth-century France and the American colonies. The French revolution signalled the death of the inherited aristocratic state, freed commerce and legitimated a broad inclusive space as part of politics. Likewise, in England, Adam  Smith's problem was how to constrain the state and he sought to enlarge the space for both market (Smith [1776(Smith [ ] 1937 and association in civil society (Smith [1759(Smith [ ] 2002, though the separation between them was unclear and each overlapped with the sphere of the household/individual.
Modern Anglo-American society is divided between government-as-state, the seat of political authority with coercive powers; the economic market; public civil society (including churches) in a variable relation with the state; and the individual, who enjoys an ill-defined normative primacy. Within the state, there is a further division of powers between executive, legislature and judiciary. In the Anglo-American world, individual freedom is primarily defined as freedom from state coercion, negative freedom (Berlin 1969;Sen 1992). The boundary between the state and other spheres is endemically contested, tense and unstable. Anti-statism is a core theme of critical political discourse, especially in the US. The primacy of negative freedom overshadows positive freedom, grounded in the capability to act and achieve goals, and the state's role in fostering capability (Sen 1992(Sen , 1999a.

The university
The mediaeval university also emerged as semi-independent within the social division of powers. Like the church the universities saw their mission as universal but evaded absolute clerical domination by embedding themselves also in cities and territorial states. Between church and state, the university established a partial institutional autonomy and academic freedom, protected by legal incorporation; a space for scholarship and later for science. In modern times government reasserted itself but a regulated university autonomy, codified in the Humboldtian ideal and its take-up in the US university, survived. As always in Anglo-American polities, there is ongoing tension at the border with the state. However, the university has proven resilient and flexible. Western nations differ in where it is positioned in society. In the Nordic world higher education is fully engaged in both state and civil society, which are less distinct from each other than in Anglo-America. Though the US university is affected by state funding it is positioned at the junction between the market and civil society. It is unclear whether the British universities are creatures of the state or civil society, though policy has developed a fiction that they are 'private' market corporations. Yet everywhere the university is much the same institution.

Individuals and individualism
Mediaeval Western culture imagined the individual in a unique relation with God in which reflexive self-formation was articulated not via social relations, as in China, but via imagined spiritual authority (Foucault 2012). The Renaissance began to conceive a directly self-referencing person. John Locke ([1690] 1970) saw a rights-bearing property-owning individual who stood alone, like John Proctor in Arthur Miller's The Crucible. The Enlightenment and the French revolution created the Declaration of the Rights of Man and the Citizen (1789), and then liberté, égalité, fraternité, though in Anglo-America, liberty was the most compelling of the three ideas and solidarity the least. In the early twentieth century, the rights-bearing person became all adult men, then all adults. In Anglo-American understandings of the relation between the individual and society, personal liberty is imagined as a distinctive space, still and inviolate, the 'free and autonomous individual separated from roles and communities' (Bell 2017, 565). In ideal conditions each person enjoys absolute self-realisation, providing that no other person is harmed. The first half of this statement has greater normative potency than the second, in contrast with the reverse in Confucianism. Selfregulating liberal individuals sustain self-regulating communities, grounded not in common property but in common rights to separated property and self. Lukes (1973) refers to the individualist vision as: … the actual or imminent realisation of the final stage of human progress in a spontaneously cohesive society of equal individual rights, limited government, laissez-faire, natural justice and equal opportunity, and individual freedom, moral development and dignity. (Lukes 1973, 37) Discussion of individuality has many strands but a feature of Anglo-American thought is the resilience of what Macpherson (1962) calls 'possessive individualism', the central motif of the atomised rights bearing person, pursuing her/his own interests in a competitive world, who is the foundation of social order and prosperity. From the Adam Smith to twentieth-century neo-liberalism (Hayek 1960) this motif repeatedly returns. In mainstream social science, especially economics, it is associated with methodological individualism, 'a doctrine about explanation which asserts that all attempts to explain social (or individual) phenomena are to be rejected … unless they are couched wholly in terms of facts about individuals' (Lukes 1973, 110). One effect of methodological individualism is to suppress from view not just collectivity but social relations as such. 'Society' is defined as the simple aggregation of individuals. In higher education, all outcomes are individualised outcomes.
In the last sixty years, the most influential idea in education policy has been human capital theory (Becker 1964). Here the value of higher education is defined by returns to individual graduates in the labour market. The student is a maximiser of individual utilities as expressed in prices or shadow prices. She/he invests in education up to the point where the cost of education, including foregone earnings, equals the lifetime returns associated with the degree. Ideally, the student pays tuition fees, sustaining market competition in higher education, but where government funding applies, the social value of education is defined as the aggregate of the additional economic productivity of individual graduates. Human capital calculations of the rates of return to degrees are widely used in policy.

Anglo-America meanings of 'public' in society and higher education
In the Anglo-American lexicon use of the terms 'public' and the pairing of 'public/private' are multiple, diverse and confusing, as indicated by the two-column entry for 'public' in the Shorter Oxford Dictionary (OED 1993(OED , 2404(OED -2405. The ambiguity of 'public' signifies its centrality in the political culture. Usage of 'public' fall into three broad categories. The first is the dualistic pairing of public with private as an analytical device. Here public and private are treated as two halves of a whole, exclusive of and opposed to each other. The relation between them is determining. In the second meaning 'public' is a descriptive adjective, not necessarily opposed to 'private', that refers to open and inclusive social relations. The third meaning of 'public' is a general expression of shared beneficence, as in 'the public good'.
The public/private dualism The Anglo-American public/private dualism takes two forms. In one usage 'public' means government or state, as in the term 'public sector', distinct from the 'private' spheres of home, family, economic market and corporate sector. Hence state or public sector education is distinguished from non-state private education. Normatively this dualism is used to underpin a positive public role for social democratic states and also to ground the anti-statist defence of sacrosanct private institutions. In Anglo-American polities, there is no consensus on the extent of government provision and funding. That debate is partly regulated by the second public/private dualism, from economics.
In the economic dualism 'public goods' stands for non-market production, as distinct from 'private goods' produced in economic markets (Samuelson 1954). Goods are 'public' when marketbased production cannot generate a profit because the goods concerned are non-exclusive or non-rivalrous. Goods are non-excludable when the benefits cannot be confined to single buyers, like clean air regulation. Goods are non-rivalrous when they are consumed by any number of people without being depleted, like a mathematical theorem which sustains its value as knowledge indefinitely. Because such goods are subject to market failure they must be financed by states or philanthropy. As Elinor Ostrom (1990) notes, Samuelson's public/private dual assumes that society is composed of two parts: the market setting where private property and commodities are exchanged, and the non-market setting where government-owned property is organised. This embodies the norms of a limited liberal state (Marginson 2018a) in which the purpose is to maximise the space for free individuals and markets by minimising the scope of the state, relations between persons are determined as competition between market actors, and the graduate of higher education is understood as human capital not as an all-round person. By the same token, in the absence of a limited Anglo-American state Samuelson's formula is irrelevant. It has no meaning in a gift economy (Mauss [1954(Mauss [ ] 1990; or a society with state-administered property and limited markets, like pre-1980s China; or any society in which the state has a comprehensive rather than limited mandate, as in the Nordic nations or in China today. It also occludes two elements in many societies, including Anglo-America: the character of goods like education is partly determined by policy (e.g., they can be more or less universal or selective); and market and state can be co-existent, not separated. Governments can install quasi-market mechanisms such as competition and prices into public sector provision.
Nevertheless, together with human capital theory the Samuelson dualism frames Anglo-American policy on higher education. It generates an ideal economic model in which the main product of higher education is individualised human capital. Relations between education and the economy are a continuum of two markets: competition of educational institutions for students, which enhances efficiency, quality and market responsiveness, and the market in graduate labour. Policy seeks to enhance the efficiency of the continuum by maximising the employability of graduates. In this imaginary students transmogrify from consumers in the education market to products for the labour market, without becoming self-forming subjects of education and society. Public goods are formed as 'externalities' or spill-overs (McMahon 2009; Chapman and Lounkaew 2015) generated by investment in private goodsfor example, citizenship which is not rewarded in labour markets and is acquired incidentally during education. Public goods not generated as spill-overs tend to be occluded. In the pure economic model government spends the minimum necessary to sustain the higher education market. The pure economic model is rarely implemented in full, and in Anglo-American polities higher education is variously defined on a spectrum from social democratic free public good, to market-defined private good. In UK and Australia data on the private rates of return regulate a zero-sum private/public split in financing, between student fees and government subsidies that nominally represent the value of the residual public goods (Chapman, Higgins, and Stiglitz 2014). Yet all parties accept the dualism, which entrenches a critical reflexivity based on the limited state. Over time this tends to erode alternate discourses in which 'public' or 'common' outcomes are tied to the state.
The economic dualism generates differential treatment of research and teaching. Basic research is seen as a natural economic public good, non-rivalrous and once published non-excludable, that underpins national technological competitiveness. University research is understood as the 'seed corn' of applications to national defence and the economy (Bush 1945), as in the Triple Helix Model of university/industry/government interaction (Etzkowitz and Leydesdorff 2000). UK policy stretches the Triple Helix idea to include the impact of universities in the public sector and civil society. The US and UK governments expand the funding of basic research at a similar rate to other countries with mature research systems (OECD 2020). This does not mean that any and every university contribution to knowledge is state funded though industry funding chronically falls short of government hopes. Anglo-American research universities are multi-disciplinary organisations in which medicine and engineering draws the largest revenues, natural science enjoys high prestige, and the social sciences and humanities are subordinate and subject to episodic ineffective tests of utility.
In relation to teaching and certification, pure application of the Samuelson formula would suggest that families and students fund higher education to the extent that it generates the positive rates of return defined by human capital theory. In practice, Anglo-American governments spend more, because of blockages to market clearing such as information asymmetry, and to expand and equalise social participation in higher education as an end in itself. Provision of a broad gateway into higher education helps to sustain social order and reduces youth unemployment. Access policy modifies the extent to which higher education is rivalrous and excludable, departing from the pure economic model. Here the public/private economic logic is supplemented by a second and different logic of 'public' as inclusive social relations (see below). Nevertheless, as the enrolment has grown and total costs have expanded the share of costs borne by students and households has increased. The 2009-2011 recession entrenched higher US public sector tuition, and over 1999-2012 the UK moved from free education to a system in which students pay half or more of costs, with access sustained by tuition loans. Given the public/private dualism, each reduction in the public share of costs is read as a change in the abstracted nature of higher education, so that it becomes more of an individualised private good and less collective in outcome.
In short, Anglo-American policy economics consigns non-market higher education to a residual role and further limits its social potentials via methodological individualism. Economics has no agreed means of calculating the value of collectively consumed/produced outcomes such as the effects of graduates on joint productivity at work, scientific and social literacy, reproduction of the professions, and the joint contributions of educated persons to social and international relations. This does not seem to be a concern: the idea of spill-overs allows states to delegate to automatic market mechanisms the responsibility for collective outcomes, while public goods created outside market transactions are neglected, aside from basic research and equitable opportunity. There is pushback against the narrow version of the economic agenda. Drawing on OECD and European policies, 'engagement' policy focuses on higher education's potential in urban and regional development and regeneration (e.g., Goddard et al. 2016). Advocates of the arts and humanities assert social and cultural values. Educators influenced by Bildung (Biesta 2002;Sijander and Sutinen 2012) argue for the broader formation of persons in higher education, beyond the limits of human capital theory, without necessarily excluding graduate productivity and incomes (Marginson 2018b). All such currents find themselves working against the grain of the economic ministries and the media focus on graduate employability and salary levels.

The communicative inclusive public
The second usage of 'public' refers to broad, inclusive or universal assembly (the public, public opinion) and open communications ('going public', public media, public relations). This 'public' is not opposed to the 'private'. Rather, the public setting provides conditions for social interaction between individuals. Individuality within this public can be more or less atomised, or collective and solidaristic, but there is a prima facie bias to inclusion regardless of social or ethnic background. Habermas (1989, 1) notes that 'we call events and occasions "public" when they are open to all, in contrast to closed or exclusive affairs, as when we speak of public places'. He traces the 'public sphere' to seventeenth-century London broadsheets, salons and coffee-houses engaged in critical conversation on state policy and matters of the day. In the eighteenth-century republics, newspapers and urban protest constituted a new collective polity in shared public space. In France it was an expression of the state; in Anglo-America it was constituted more in civil society and the privately owned media. One spinoff was the autonomous eco-system of electoral democracy. Dewey (1927) identifies a communitarian public which defines policy problems that require resolution by the state. While the Anglo-American public oscillates between decision-maker and critic, there is a sense of talking up to power. Castells (2008, 78) defines the 'public sphere' as 'the space of communication of ideas and projects that emerge from society and are addressed to the decision makers in the institutions of society'. However, the inclusive public in this sense is wider than polities. It extends to the 'public company' with shared ownership and tradeable equity. In markets, flat open social inclusion and non-discrimination maximise customers. The dominance of public conversation by privately owned social media further blurs the old Anglo-American lines between public and private, and between polity and economy. These platforms now sustain much the most potent electoral conversation, displacing public meetings and slower inner party debate (Runciman 2018). States call on the platform capitalists to assist governmental projects that require inclusive regulation and information dissemination, for example during the Covid-19 pandemic in New York (Klein 2020).
The inclusive communicative public has several resonances in higher education. Bildung implies the formation of students as public citizens in a shared modernisation project. Sen (1999a) notes education is essential to democratic agency. Higher education institutions were early adopters of the Internet and are among the larger carriers of the communicative public. There is wide support for more open and equal social access and educational participation by different groups. Universities are increasingly seen as accountable to local communities. Calhoun (1992) and Pusser (2011) present universities in Habermas's terms as semi-autonomous adjuncts of government that harbour constructive criticism, policy ideas and transformative social movements. Ignatieff (2018) suggests that within the division of powers, the critically minded university is an analogue to a free media and independent judiciary, a counter to majoritarian populism in the polity. Higher education legislation in New Zealand enshrines the idea of the university as 'critic and conscience' of society.
The universal public good The inclusive communicative public is almost synonymous with 'society' and shades into the third meaning of 'public' as 'the public good', a condition of common and universal welfare, virtue or prospect (Mansbridge 1998). One root of the 'public good' is the feudal European practice of the commons, a shared resource not subject to scarcity and utilised by all, such as a river or a pasture for grazing animals. Normative claims about the public good have rhetorical power. However, in the notion agency is undefined. Western thought embodies many competing claims about universal value. The state carries the only general mandate but in Anglo-America there is skepticism about its claims to embody the public good. State policies may favour strong groups (e.g., in university entrance), while non-state agents can contribute to the public good. Further, the idea of the universal public good is contaminated by the narrower 'public goods' of economic policy.
For these reasons, UNESCO has developed the notion of 'common good' and 'common goods' in education (UNESCO 2015;Locatelli 2018). The ideas are grounded in Western European civic democracy. The collaborative community defines the desired outcomes and engages in joint production and democratic distribution (see also Ostrom 1990). Diversity of objectives and contributions is valued (UNESCO 2015, 78), transferring to negotiation the problem of competing claims to the public good. Dupré (1994, 173) defines common goods as collective non-market goods 'attainable only by the community, yet individually shared by its members'. Common goods augment shared welfare and foster solidarity, inclusion, tolerance, universal freedoms, equality and individual human rights (Deneulin and Townsend 2007, 24). Both government and nongovernment organisations contribute, though 'some kinds of private participation are more defensible than others' (Locatelli 2018, 8) and state funding and regulation may be needed (13). However, common good(s) have greater salience in Western European polities than the English-speaking world.
A key question about public or common good(s) is the boundary of inclusion: city or local region, nation, global region, the world. Shared global goods include ecological security and knowledge. Yet Anglo-American political cultures struggle to imagine goods beyond the national border. Given the public/private dualism, public goods require a state, but there is no global state. The United Nations (UN) Development Programme defines global public goods as 'goods that have a significant element of non-rivalry and/or non-excludability' and are broadly available on a global scale (Kaul, Grunberg, and Stern 1999, 2-3). In a multilateral frame global public goods are transferred national public goods, assembled piece by piece. This marginalises global systems not reducible to nations, such as cross-border scientific networks, and downplays global problems and solutions. The UN agencies and Organisation for Economic Cooperation and Development (OECD) have pan-national perspectives but pursue their agendas via nation-states and with the consent of the powerful countries. The legitimacy of national governments derives from inside not outside their countries (Wang 2017;Zhao 2011). They are under little pressure to address global issues. They prioritise national goods over global goods and have an incentive to free-ride on spill-overs from abroad. While UNESCO's education as a global common good brings cross-border relations and non-government actors into the picture, it has limited conceptual or practical purchase. Universities can range beyond national borders, yet such is the conceptual void of the global that when doing so they are defined as private corporations.
In sum: the Anglo-American approach The Anglo-American social imaginary is a changing patchwork. The individual has a loose illdefined primacy. Beyond the individual, there is no essential primacy of state, market and civil society. The division between them is variable. Lines between government and other spheres, including higher education, are tense. In higher education, the economics of private/public goods forces an individual/collective tradeoff. It limits the scope for collective goods and defines individual benefits as atomised not relational. The core responsibility of the state is reduced to residual collective goods, including research but not necessarily education, rather than the universal 'public good'. Anything more is left to institutions and individual actors. The exception is policy on social inclusion where, consistent with the inclusive-communicative idea of the public, higher education has larger obligations.

Chinese social relations and higher education
This section summarises Chinese discourse on the outcomes of higher education, in the context of the social imaginary and political culture in China. It focuses on the state, individual and collective, the pairing of gong/si (roughly, public/private), and the roles of higher education. As indicated, it proceeds in two sequential sections: the indigenous Sinic imaginary that continues to be determining; and the multiple and hybrid experience in modern China, that is impacted also by Westernisation, where there is a larger intersection with Anglo-American ideas of 'public' and 'public/private' but important differences remain.

The Sinic political-cultural imaginary
The roots of China's political-cultural imaginary are in the Zhou dynasty and Confucian-Daoist ideas of the Spring and Autumn and Warring States times (771-221 BCE). The model is integrated, without a division of powers. The individual is secured inside social relations.
In its classical Confucian form, China's imaginary consists of nested circles ascending from local spheres to wider spheres (Figure 2), on the basis of dyads of smaller self/larger self (xiaowo and dawo). The traditional Sinic family, the primary sphere below the state, is larger than the nuclear family with several generations living in the same locality, often led by a single elder. The system is flexible. Spheres can be larger or smaller. Their scope is not fixed. However, larger spheres have normative primacy over smaller spheres, successively relativising the individual in relation to the collectivities of family, state and society, and tianxia (all under heaven) (Huang 2000;Tu 1985); and relativising the family in relation to the state. This system privileges social order and the family and state as embodiments of order. In the Confucian universe people are loyal to their family and country, love others and maintain the justice of the whole state (Hwang 1999;Li 2008). Classically, the system rests on the ethical formation of persons and self-regulation on the basis of these values.

The state
China's state is not a limited liberal state but a comprehensive state. This was not invented by the CPC. Since the Zhou dynasty state power, politics and statecraft have been customarily supreme over other domains, including the landowning aristocracy, merchants and the economy, cities, professions, the military and religion (Gernet 1996;Zhao 2015). The archetypal Sinic state, both comprehensive and centralising, was the Qin dynasty (221-206 BCE) which suppressed the aristocracy, unified China territorially and standardised language and measures. The Han dynasty that followed codified a state that was both Confucian and legalist (Zhao 2015). Though in pure legalist doctrine the state had no final justification except itself, this was balanced by the Zhou system of accountability. The duties of the state were to secure social order, including defence, and prosperity. If it faltered in these tasks it lost the 'Mandate of Heaven' and popular consent was withdrawn (Zhao 2015, 52-55). However, except during the period of the post-ImperiaI Republic from 1911 to 1949 when Western forms were intermittently used, in China's long history there has been no discursive limit to the authority of the state. Potentially it can intervene anywhere. Correspondingly, civil society in China has always been smaller and more closely managed. Cities and urban elites have always had less autonomy: Whereas public display of political power was central to the Roman world and reflected in the spatial configuration of cities, the exclusionary principle dominated in China. Unlike in Roman cities with their assembly places and theaters, in Han cities people gathered in markets, which served as a conduits of state control. (Scheidel 2015, 8; see also Norena 2015, 181-203) In Imperial China shared values and voluntary consent were mostly sufficient to hold together the vast diverse country. The Imperial state maintained tributary relations with China's neighbours, collected taxes, set rules and regulated property rights but its direct writ stopped above the village. It was not a micro manager. Conformity was secured by cultural compliance with the Confucian moral order (Hwang 1999;Liu 2011). Instead of a division of formal powers China's state evolved forms of devolution which sustained its authority. Centrally formed cadre was deployed as regional officials (Blockmans and de Weerdt 2016). However, states that are both comprehensive and centralising are pulled between the need to lighten the burden and the need to guarantee control. The Imperial state oscillated between periods of opening and civil freedom and periods in which control was tightened. The revolution of 1949 did not lead to new limitations on the state and larger spaces for markets and civil society, as in the French and American revolutions. It created a cohesive and more focused Leninist party-state with closer reach down into the household, and voluntary compliance supplemented by greater state supervision and social engineering. The oscillation between liberalisation and control continues (Muhlhahn 2019).

Higher education
Hence while the mediaeval university in Europe found an autonomy between church and state, higher education took another path in China. From the Zhou dynasty onwards, academies prepared scholar-officials for the Imperial order. The keju, the examination of candidates for merit-based entry into the civil service, emerged in the Sui dynasty (581-618 CE) and became fully consolidated as the mode of selection under the Ming dynasty (1368-1644) (Elman 1991(Elman , 2000. Graduates of the Imperial academies were distributed across China as leaders at provincial and district level. This was the main means of social mobility, though it required advanced cultural capital. Centred on the Confucian classics, the examination took over two decades to master. In the academies, knowledge was valued not as theory above practice, or theology, but for its practical application to governance. As the result of knowledge about China brought to the West by seventeenth-and eighteenth-century Jesuit priests, the Chinese model of higher education nested in state administration, and the use of written not oral examinations, affected reforms in Western Europe, notably the Grandes Ecoles in France (Nakayama 1984;Hayhoe and Liu 2010).
There was a secondary form of higher education with an intermittent and marginalised relation with the state, a lesser role in social mobility and a less instrumental intellectual agenda. The private shuyuan, which originated in the Tang dynasty (618-907 CE) and spread under the Song (960-1279 CE), occupied locally provided land and were mostly independent of the Imperial government.
Influenced by Buddhist monastic scholarship from India, 'they were places of broad and serious learning for the sake of deepened knowledge, not merely career or political advancement' (Hayhoe 2019, 183). They were used by scholar-officials as retreats, and in the later years of the Ming were gathering points for critics of the regime, paralleling the Western public sphere, though shuyuan lacked the protection of a legal charter. There are other precedents for the role of constructive critic, scholars speaking truth to power, stretching back to the Zhou (During 2020). During the Warring States, the intellectually diverse Jixia Academy provided unconstrained advice on statecraft to the kingdom of Qi (Hartnett 2011). From time to time, notably under Tang Emperor Taizong (598-649 CE), officials named jianguan were routinely expected to make comments and criticisms (Cheng 2001;Zhao 2000). The jianguan were granted freedom of expression and protected from punishment (Chen 2001). Emperors and dynasties were not always so generous. The jianguan system was an episodic freedom, at the Emperor's behest, part of the centralised Chinese oscillation between periods of openness and periods of closure.
Individual and collective In Confucian thought an absolute self, separate from society, is impossible (Hsu 1985;Cheng and Yang 2015). In the Imperial period, individuals were not seen as independent social agents. Their rights and liberties were not discussed. There was individuality, but not the normative Anglo-American individualism. The Confucian self (wo) is a relational and role-bearing individual who is successively layered by family, state and tianxia; a member of larger collective groups, especially the family (Ho 1979). Bodde (1957, 66) states: 'Confucian "individualism" means the fullest development by the individual of his creative potentialitiesnot, however, merely for the sake of selfexpression but because he can thus best fulfil that particular role which is his within his social nexus' (see also Bell 2017;Rosemont Jr 2015). Though one strand of Daoism emphasised the separation of self and world and finding one's path rather than social engagement, it was less influential.
A central value of Confucianism is individual development (Lee 2000) through self-cultivation, ethical formation via the working of self on self, first in the family and then in education (Li 2012).
The key moment is the formation in every child, typically at the age of six or seven years, of lizhi, the reflexive commitment or 'will' to learn (163). Confucian learning is not just about knowledge but about how to live in a relational setting, what one should become, and how to improve oneself. Persons do not have fixed qualities or talents that determine their lives. Anyone can succeed. The self is a process and the crucial element is reflexive effort. This includes the cultivation of free will (Cheng 2004). However, Confucianism distinguishes between free will, zhi, the inner self of moral autonomy, and the outer social self. Persons must restrain from enacting their will if there are negative social consequences. Self-determination is absolute but self-realisation is not. In contrast with Anglo-American self-determination (Ryan and Deci 2000), practising free will is not an absolute right but a good thing among good things (Chan 2013).
While the role-bearing Confucian individual is the foundation of Chinese social order, the intrinsic preference for the outer sphere over the inner sphere is the foundation of Sinic collectivism. There is no zero-sum dualism of individual and collective. Nor is the individual suppressed.
The key to this system is the embeddedness at each level, the ascending scale of collectivism. The nesting of persons in the social order is secured by the cultivation of individuality steeped in Confucian values. The gap between family and formal education is narrower than in the West. In the Imperial era schooling was organised in kin groupings (sishu) in the village. Today parents are very active in decisions about higher education.

Tianxia
Tianxia is the unified human and natural sphere, a civilisational zone larger than the state. Tianxia weigong, 'all under heaven is for all', or 'all under heaven belongs to all', carries connotations of universal benefit more central to Chinese thought than 'global public goods' or 'global common goods' in Anglo-America. However, tianxia has more than one meaning. It can refer to the whole world, or to Chinese civilisation and beyond. In the second, China is the kingdom of heaven at the centre with other peoples at the periphery.
Tianxia has various associations in Chinese thought. It embodies the Confucian movement from qin qin (affection for one's kin) to fan ai zhong (affection for all humanity as one community). Wang (2017, 1) contrasts the zero-sum notion in Europe, which embodies an opposition between I and non-I/other, with tianxia which has diverse selves but no 'other'. 'Tianxia refers to a system of governance held together by a regime of culture and values that transcends racial and geographical boundaries'. Tianxia is seen as continually changing (Hall and Ames 1998, 242). It refers to harmony and peace on the basis of respect for diversity, not uniformity. One English version of the Confucian term ren is 'two-man mindedness', awareness of plurality, suggesting a universal humanism that begins in sociability rather than the individual (Liu 2014, 411). 1 Duara (2017) interprets a tianxia system as cosmopolitan. Unity in diversity (heer butong) requires more than tolerance. It needs mutual understanding, respect, dialogue and trust (Fang 2003;Fei 2015).

Correspondences to 'public' and 'private'
In Confucian thought relations between each pairing of smaller self/larger self (xiaowo and dawo), such as the pairing of individual and family, or state/society and tianxia, are understood as a pairing of gong and si. Confucius and Mencius focused on the differentiation of gong and si and the two terms began to include abstract and metaphysical meanings (Huang 1991(Huang , 2005. Gong took on multiple meanings, touching on non-individual, public, common, universal, openness, fairness, all people/humankind, the state. Si connected to private, personal, selfish, and secret. For example, in the Northern Song dynasty, gong referred to righteousness while si stood for private goods and personal desire. Si Maguang (1019-1086) stated that people should prioritise gong (Huang 2005). During the Southern Song, gong referred to the heavenly principle and si represented people's wills. Zhu Xi (1130-1200) argued that gong was legal while si was illegal (Zhu and Lv [1175] 2001). Ideally, there was no conflict between them because the individual internalised social values (Huang 2005). Where there was tension between public and private interests in Imperial China, the task was to find a balance to satisfy both. Where there was continuing conflict, dawo, the larger collective, was supreme (Huff 2009;Watt 1972). However, tensions between gong and si were, and are, insufficiently discussed. Huang and Jiang (2005) find that despite laws and policies to protect private property there is no clear boundary between public and private. Many scholars argue that si is under-recognised in China and the individual is insufficiently protected (Hsu 1985;Huang and Jiang 2005;Lan 2005;Cheng and Yang 2015).
The dual of gong and si correlates to 'public' and 'private' in English in one respect: each movement outwards from a smaller to larger circle is a move to enhanced 'publicness'. When the Anglo-American 'public' references the state, this corresponds to gong embodied in the Chinese state. There are also differences. Gong and si are co-existent and relational, whereas in the dualism, each of public and private signify unique essences that cannot coincide and are related only by being not the other. In the manner of keystone words in Chinese, gong is inclusive, whereas the multiple meanings of 'public' signify not inclusion but ambiguity. Just as the Confucian individual is nested in social relations not ontologically separate, si is nested in gong, not paired with it in a zero-sum relation like the public/private dualism. In China gong, the domain of harmony and social order, has normative primacy over si. In Anglo-American individual liberty might be valued above universal benefit.
Other correspondences to the Anglo-American 'public', 'public goods' and 'private' are more elusive. There is no Chinese equivalent of 'goods'. The literal translation of 'private individual interest', yi ji zhi si, carries a negative connotation of selfishness. Given the weaker civil society in China there is no equivalent of the communicative, inclusive and democratic public domain, as in 'public opinion'. Gong is more readily identified with the Chinese state than is 'public good' with Anglo-American states. In the larger usage of gong it is unclear what is the relation between state and society; the extent to which the state embraces all of society, or it has a border and there is something outside it. Nevertheless, gong implies inclusion, and imagined on the social scale it might take in the Anglo-American 'public good' as universal beneficence. There is a language for discussing gongde, meaning public virtue. The foundation is Confucianism's five constant virtues (wuchang): benevolence and humanity (ren), righteousness and rite (yi), propriety (li), wisdom (zhi), and integrity (xin). The Book of Rites describes a society that is inclusive and equitable in the sense of a meritocracy, foreshadowing a role for education in composing such a society: When Dao prevails, all under heaven is for all, in which selecting criteria were wisdom and ability. Mutual confidence is promoted and good neighbourliness cultivated. Men do not regard as parents only their own parents, nor do they treat as children only their own children … They despise indolence, yet they do not use their energies for their own benefit. In this way selfish scheming are repressed … This is called datong. 2 The terms gongzheng (fairness) and gongping (equity), which pertain to the role of education in fostering social inclusion, include gong. For Confucius education was the route to self-betterment. Any person, from any background, was capable of advanced learning. During the Republic after 1911, Western ideas of equity in education, grounded in equal rights and freedoms, took root (Yang 2011); and after 1949 the egalitarian temper of the CPC reinforced the notion of equal access (Ding 2007;Luo, Guo, and Shi 2018).

Gong and si in modern times
In modern China, pre-Imperial and Imperial tradition are no longer the only determining element and the ascending Confucian circles are no longer sufficient to describe the social order. The autonomous individual has gained greater salience (Yan 2009). At the same time, the comprehensive Chinese state has become stronger, tending to but never completing a monopoly of social relations (Figure 3). Understandings of gong and si are less stable.

The party-state
The CPC is organised on the basis of Leninist democratic centralism (Liebman 1975). Once an issue is resolved the whole Party is committed to its disciplined implementation, enabling a small group  (2000), Tu (1985). of leaders to steer the polity. In the Mao Zedong era, the CPC took Leninist forms into each part of the economy, including rural communes, work units (danwei) in the cities and higher education institutions, thereby more completely conflating state and society (Fewsmith 1999, 70). In Deng Xiaoping's deregulation and opening up after 1978 the party-state retained control. The first entrepreneurs in the new economic zones were mostly party cadre. The party installed itself in all forms of devolution.
The party-state is more effective than was the Imperial state. First, it is more tightly centralised. The separate Imperial authorities of throne, court and officials have been joined. Party and state are formally separated but not independent. This is not a Western-style division of powers: no-one doubts the party is dominant. Second, the party-state is a more effective micro-manager. Democratic centralism is vulnerable to one-way information flows, with ambitious cadre telling the next level what it wants to hear, but the party-state has developed comprehensive surveillance systems, with growing use of automation.
Nevertheless, social and civil communications generate chronic political difficulties. Comprehensive surveillance is burdensome, especially when central control is tightening. In the 1980s the party-state sustained expanded diversity and democratisation inside itselfat times there was open political debate and diverse commentary in the party organs (Vogel 2011)and a 'semi-civil society' (He 1997) outside, without destabilising the social order. Non-government organisations flowered and public arts boomed. Later, however, the local party-state annexed civil organisations to governance, using them as a consultation mechanism while limiting the scope for criticism (Teets 2014). An independent civil order is constrained by the 'overlapping' of the state and the 'fragile autonomy' of intellectuals (He 1997, 147-165). In relation to the inclusive democratic form of 'public' there are three differences with Anglo-America. First, Chinese civil society is not ongoing but episodic and vulnerable to surveillance, suppression and co-option. Second, the means of broad-based discussion such as the Internet, social media, other media and wall posters are restricted. Third, however, there are continuous communications inside partystate networks, which include the universities. In China, the party-state is the wide-reaching collective expression of society, whereas in the West the public sphere of open communication, structured by unequal corporate power but only partly shaped by government, performs that role.

Individual and collective
In the late Qing years and the New Cultural Movement of the 1910s/1920s Western individualism gained ground (Geng 1994, 44). Intellectuals called for a transition from family-based society to an individual-based society (Hu 1918). In the event, Marxist-Leninism offered another path to modernisation (Meisner 1977, 19). The collective character of the CPC, with its unquestioning loyalty of individuals to the larger group, matched Chinese tradition more closely than did Western liberalism (Fu 1974).
Nevertheless, the party-state drove successive political, economic and social upheavals in the Confucian social order. Confucian collectivism was explicitly rejected. Whereas intellectuals in the late Imperial era often criticised the weakness of the state, in the 1950s the ascending circles were replaced by loyalty to the state as the strongest repository of the collective. Mao's rural and urban work groups broke up the traditional kinship networks. Then from 1978 on, Deng Xiaoping's de-collectivisation, partial deregulation, accelerated economic growth and private enrichment triggered another round of transformation. The family partly revived, and some Confucian values were re-endorsed, but market capitalism fostered individuality and a massive migration from the countryside to the cities. Average family size shrank because of migration and the one-child policy (Retherford et al. 2005;Zheng et al. 2009), accelerating the fragmentation of the traditional large kinship family and its replacement by the conjugal family of two or three generations based on the married couple (Yan 2009). This weakened the role of the family in monitoring values, daily life, marriage and career. Single migrant workers in the cities, mostly male, disembedded from their families and with no localised bonds, were freed from social obligations (Lifton 2012;King 2018) or more reliant on horizontal guanxi ties at work.
There are many signs of the rise of the autonomous individual. It is more difficult to identify the smaller self (xiaowo) and larger self (dawo), easing the pursuit of self-interest without Confucian self-commitment to the collective gong. Widespread adoption of the profit motive, plus rising average income, greater economic inequality, the endorsement of policy capitalism and the partial eclipse of communal values, are much discussed (Vogel 2011;Zang 2011;Zhang 1996;Nonini 2008;Zhang and Bray 2017). Yet the individual is nested in a complex of networks based in kin, ethnicity, region, school, work, professional and guild-like structures, religious associations and market transactions. Some such bonds extend abroad. How much has the collective/individual balance actually shifted? Research findings differ. Lu (1998) identifies a tendency to utilitarian individualism but finds that collectivistic values endure. Zhang and Shavitt (2003) state that the rising middle class in the X-Generation is the recipient of individualist values while the masses continue to embody traditional collectivist values. Koch and Koch (2007) confirm this, finding that people from inland China show a more collectivist orientation than those from developed coastal areas. The adolescents studied by Li et al. (2010) are collectivist with their friends while individualist about school achievement. However, the parent-child bond remains very strong, as shown in the level of parental investment in shadow education (Chou 2010;Lee and Xiao 1998;Zhang and Bray 2017). Bodycott and Lai (2012) find that in higher education, the Chinese family is the main decision-maker, not the individual student.
The Chinese individual has a different lineage to the Anglo-American individual. Though there are convergences around modernisation, mobility, economic accumulation and family shrinkage, China's external settings and inner mentalities are distinctive. Yan (2009, 273) refers to 'the Chinese model of individualisation that excludes cultural democracy, welfare state and individualism' in the Western sense. Self-making in China is less a matter of choice of lifestyle and personal politics as often the case in Anglo-America, and more about social status and material life. Personal identity matters when it decides opportunities (288). Like the Imperial dynasties, the party-state allocates rank and station on the basis of membership of social groups. Despite the ease of geographical mobility, rural migrants find it difficult to secure the urban hukou status which provides better health and education (Xu 2020). Meanwhile the party-state continues as a collective dawo, the repository of a shared meta-identity which offers its leading cadre mobility in all forms, including elite universities.

Higher education
After 1978 the party-state built from almost nothing a tertiary education system housing 50% of the school leaver age cohort (UNESCO 2020) and the largest science output in the world (Li 2015;Marginson 2021). In this process, Leninist centralism combined Western modernisation with Chinese tradition. Deng Xiaoping emphasised learning from abroad, especially the US. Foreign universities and scientists were invited to China, students and faculty were sent abroad and universities and disciplines were benchmarked against world leaders (Wang, Wang, and Liu 2011). This 'national/global synergy' (Marginson 2018c) has sustained accelerated development. At first glance, modern Chinese universities closely resemble their Anglo-American counterparts in the degree structure, the curriculum, the doctorate and much internal organisation. Chinese universities have strategic executive leaders and devolved corporate managers. They are partly non-government funded. They also follow the Anglo-American-European map of academic disciplines, though the role of physical sciences and engineering is relatively large (Kirby and van der Wende 2016), and production volume and free scholarship in the humanities and social sciences are politically constrained (Xu 2019;Shambaugh 2013, 244). However, China's universities are Westernised not by external colonisation but via a state-driven project of catch-up in which education and science are firmly nested in national policy. The tensions between Westernisation, and old and new Chinese norms (Yang 2014), are self-imposed and inside the national project. They can be tuned by the party-state.
There is continuity as well as modernisation. The party-state draws on China's inherited political culture, including the Confucian ethic of continuous self-improvement (Ho 1979). Elite universities continue the Imperial mission of preparing graduates for government. The state sets part of curricula, and student numbers. Whether party state-mandated political education is as effective as were the Confucian classics is not known. As before, knowledge generation is focused on practical national needs, now in transport, energy, urbanisation, construction, communications and information. Like its imperial predecessors, the party-state engages more directly in institutional governance than do its Anglo-American counterparts Hayhoe 2016). In Deng's managed autonomy in science and higher education, an essential condition of grass-roots initiative and offshore engagement, democratic centralism is combined with Imperial tradition (Marginson 2018c). Paralleling the Song Dynasty, provincial and university leaders are trained in party schools and posted from the political centre. Dual authority systems in the universities, with party-secretaries alongside specialist leaders at each level, have antecedents in the Ming when the eunuchs of the Eastern Depot monitored and disciplined the scholar-officials (Fukuyama 2011, 309). The alternate higher education, the semi-independent shuyuan, now have little presence. A university wholly outside the state is unimaginable.
Hence in China university autonomy and academic freedom play out within the boundaries of the state rather than on the boundary between the state and society. Though the literal translation of 'university autonomy' is zizhi the term mostly used is zizhu, self-government or mastery. This belies the idea that because universities are closely nested in the state, and the issue of legal separation scarcely arises, they are simply subordinated with limited freedom to act. Rather, they embody zhu (mastery) while interacting with the central state. Faculty exercise social responsibilities that enjoy high status, in the context of the Sinic tradition of knowledge linked to action and contributions to the public good (Hayhoe and Liu 2010;Hayhoe 2011). In China xueshu ziyou (academic freedom) is understood as unconstrained freedom to conduct research as well as sixiang ziyou (intellectual freedom). Nevertheless, faculty in China are more able to accept state regulation as a normal condition, for example, the requirement that they contribute positively to state policy, than are their Anglo-American counterparts.
'Public' and 'private' in higher education In exploring the relation between the Anglo-American 'public' and 'private' and Chinese discourse, Tian and Liu (2019) interviewed state officials, university leaders and faculty. They found a strong sense of an enlarged 'public' as state in China, corresponding to Sinic norms; a weaker use of public/ private in the terms of Anglo-American economics; and interest in ideas akin to global public or common good.
Interviewees understood 'public' primarily in terms of the state, and higher education as part of 'the public service sector'. It serves the state's twin objectives of social order and prosperity and is nested in government. State personnel and university leaders agreed the state has comprehensive responsibility for planning, development and funding, while universities are autonomous in education, student selection and resource management. Faculty especially commented on marketlike elements such as tuition charges, non-state revenues, competition between institutions, and selective entry in elite universities. Some describe higher education as a 'quasi-public good' (zhun gonggong wupin), a term, widely used in China, derived from the Anglo-American economic model. Likewise, economic policy research on higher education uses human capital theory (e.g., Zhang and Zhuang 2011;Li et al. 2012). As in Anglo-America, there are concerns that marketisation erodes the mission of universities and the ethical formation of graduates. However, the discussion is not the same. Given the nesting of si in gong, the self-betterment of individuals in higher education is part of the collective contribution of higher education, rather than zero-sum with it. In the study by Tian and Liu (2019) the interviewees did not see fees and corporate universities as implying a more limited role of the state, or reducing the scope or obligation of higher education to generate collective outcomes. Likewise, in the study by Chen (2020) the main goal of Chinese parents was that their student children embody and fulfil the Confucian self, including its social obligations. They saw no tension between the collectivist ethic and individual self-betterment through education. While relations between gong and si are changing, one does not evacuate the other. Yet measures of the collective outcomes of higher education are no more advanced than in the West. Perhaps the inclusiveness of the state remit, and the absence of an instrumental link between the public/private split of financing and the public/private definition of benefits, makes this less essential. Tian and Liu (2019) find that the Western idea of 'common good' fits higher education in China better than 'public good' because of the ambiguity of 'public good' and the long history of collective forms and grass-roots democracy in the Chinese world (Wang 2012). Some of their interviewees took this to the global level, referencing Xi Jinping's 'a community of shared future for mankind' (renlei mingyun gongtongti) as a Sinic framework in parallel to the Western 'global common good'. The Belt and Road programmes have triggered a discussion about differing visions of tianxia, whether it is the projection of a hegemonic power or fosters diverse agency, and the implications of global projects in higher education and other sectors (Callahan and Barabantseva 2011;Sun and Chen 2016;Wang 2017;Ding 2018;Brook 2019;Feng 2020;van der Wende et al. 2020). With tianxia, China has moved ahead of Anglo-America in imagining a realm of the global and the ecological. In contrast, in Anglo-American universities global higher education is read primarily through the lens of methodological nationalism. Flows of influence are one-directional and Anglo-Americans engaged in global relations do not feel compelled to acquire multiple identities. Very few American and British students learn Chinese and few travel to China, compared to the number of Chinese students abroad. Foreign engagement in China involves plurality. All Chinese students learn English. Large numbers of students move both ways.
With the communicative and inclusive 'public' less developed in China there is limited scope for universities to function as a Habermasian public sphere. As noted, Chinese tradition suggests the jianguan, the sage who speaks truth to power within the circles of the court, and universities and faculty play that role in the modern era. Many personnel from the leading universities are closely engaged in policy making, as autonomous thinkers within the party-state. Professors enjoy higher standing and effectivity in government than their counterparts in Anglo-America. The party-state draws on their capacity for critical thought behind closed doors. Faculty, and more often students, become joined to broader civil and political society only episodically. Here Peking University has played a special role. 'Beida' was the starting point for most twentieth-century political movements, from May the Fourth in 1919 to Tiananmen in 1989 . However, as in 1989 the party-state habitually shuts down the open public potential of the universities when the stability of its rule is in question. Peking University's public political function has never been formally legitimated and in 2018 a university leadership change under Xi Jingping signified the assertion of closer control. In late 2019 references to academic freedom were removed from the laws governing Fudan and Nanjing Universities. Facing the political challenge of managing an educated civil society in which half the young people enter tertiary education, the state has stepped up micro-control of persons, so alien to Western liberalism. Individual agency has been both newly augmented and newly limited. This suggests that a new balance between state, civil society, social obligations and persons will have to be found.

In sum: the Chinese approach
The primacy of the collective gong in China, and the state as a meta-agent of gong, is the default position as in the Imperial period. The state retains a comprehensive perspective and a mandate for arbitrary intervention across its field of observation. It has a hand on all of the outcomes of higher education. Universal Confucian self-cultivation disperses, across society, responsibility for individual outcomes (as in Anglo-America but by a different route). However, in nesting the individual in family and society China avoids tradeoffs between individual and collective, though the scope for individuality is increasing. Higher education is constrained in its capacity to join with a communicative inclusive 'public' except in the global dimension. Limits to the scope of the collective are directly political, not discursive as they are in Anglo-America. China's tradeoff is between different kinds of collectivity. Table 1 summarises the differences of approach in the Anglo-American and Chinese worlds to society, state and the individual and collective outcomes of higher education. In this paper the outcomes of higher education have been investigated via the lens of Anglo-American 'public' and 'private' good(s) and the nearest Chinese equivalents. In discussion of 'public', with its partly contrary meanings, the traditions do not closely align.

Summary and conclusions
In relation to public as government or 'state' there is overlap without equivalence. The Anglo-American concept of the public/private dualism has resonance in discussion of higher education financing in China, but without limiting the state or the potential for collective goods. The informal Anglo-American communicative public, the democratic assembly that is separated from the state executive, fades from view in China where law and civil society are state controlled. This tends to block the potentials of higher education in unmediated social relations. Though there is closer equivalence between tianxia weigong in China and Anglo-American 'global common good' the latter idea is marginal in Anglo-American tradition. The table helps in identifying the presences and the gaps in each tradition, fulfilling Scheidel's (2015) hope that the comparative mirror helps each tradition to see itself more clearly. The two political cultures are each positioned here as one cultural tradition among many.
The comparison suggests that lexically speaking China has the better set of tools for making collective goods in higher education. All else being equal state action has the largest potential to provide universally distributed collective goods. Yet in China, the outcome of state action is more impoverished than this would suggest, while the potential of non-state collective goods is highly restricted. Though in Anglo-America the collective potentials of higher education beyond the state are open, the potentials of state action are discursively truncated. In each case, the limitation is political but fundamental to the tradition.
In lexical terms, both traditions can imagine a socially productive state. At the social democratic extreme, the Anglo-American idea of 'public' looms almost as large as the state in China. However, the Anglo-American state is always contested because of its potential to encroach on individual freedom, because this is defined as separate from the social and protected by the public/private dualism. This critical reflexivity is as strong on the political left as the political right: for the political left, the state has mostly been a creature of property and capital. In higher education the logic of the dualism offers 'public' as a universal and inclusive benefit, while subordinating that common terrain to the superior value of the 'private' individualised outcomes. This installs an ambivalence about the role of the state and the value of non-private collective outcomes that rest on state guarantee. The other source of collective goods, outside the state, is civil society. Yet while civil society is a realm of free association it is highly unequal and fashioned by global communications companies with addictive technologies and no locus of responsibility. This brings the problem of collective goods back to the statewhere there is no answer to the question except mantras about invisible hands and spill-over benefits from market transactions, the use of economic value (e.g., rates of return to degrees) as a poor proxy for all of social value, and the more substantial belief that government carries responsibility for social equity.
The ambivalence and ambiguity are absent in China. The state is loosely equated with the collective gong of society, a universal container that carries all of the individual and collective outcomes of higher education. Western thinkers persistently misunderstand China's political culture. They read it in terms of the Western problematic of individual versus collective. In Anglo-America the individual is curiously separated from and set against the social realm which provides its existence and identity (Dewey 1927). There is no necessary tension between individual and collective in China, where the one is nested within, and expressed within, the other. Likewise, there is no essential tension between the individual benefits of higher education and the common and shared benefits, as testified by Chen's (2020) families. However, the state-shaped gong does not comprehend civil society and thereby misses the larger potential of higher education in social communication and criticism. Whereas in the Anglo-American setting the tension is between the individual and the collective, in China the tension is between the would-be state monopoly of the social, and other forms of collective association beyond the family. In a society of near universal Confucian self-cultivation, with its rich possibilities for the contribution of higher education, blanking out the civil potentials of higher education is an incalculable loss. It also reduces the pressure on the state to deliver. As in Imperial times, the customary contribution of the state is more modest than its authority. It takes responsibility for prosperity and social order, including social equity. From time to time, like its Anglo-American counterpart, it prods the universities into serving their localities and regions as well as the nation. More than its Anglo-American counterpart, the Chinese state is concerned with person formation in higher education, though the family and self-cultivating student really carry this. That is all. China is no more energetic in defining and measuring collective outcomes than is the Anglo-American state. The larger potential for collective goods in higher education is under-realised. This lacuna is disguised. The party-state in China resolves the question of higher education outcomes not discursively but in a directly political manner. Collective good is what is generated by the state; what is generated by the state must be collective good.
That is a reminder that 'words' are not the sole factor that makes and unmakes 'objects' in higher education, to refer again to Foucault's (1972) couple. Yet the comparison in this paper has also shown that discourses do shape and limit imaginings and practices. First, the public/private dualism pulls a whole tradition away from the positive potentials of collective action. Second, the lexical weakness In China of the Habermasian public sphere reproduces the long history of state subordination of the civil order. Third, the absence from Anglo-American political culture of a global realm beyond the nation, and its presence in China, gives Chinese people more resources for tackling the interdependent global setting.
Nothing stays the same. Languages and material practices are constantly changing. Arguably, China is moving to a more balanced combination of individual and collective, in general and in higher education. Given the growing individualism, with the individual still nested in social obligations, it might be possible to move beyond the gong/si framework so that both the individual and the collective have recognised priority, without the need to subordinate one to the otherachieving something parallel to the Nordic coupling of individual rights and social solidarity by another route. This seems appropriate in a society in which half the age cohort is now educated to tertiary education level. It is important to recognise that although the party-state resists the full codification of human rights, state policies in higher education have fostered the enhanced agency of persons on a massive scale. The blockage here is the practise of the party-state as meta-gong, its would-be supervision of the whole of the social; the potency of its control over civil association and political agency. In the perpetual oscillation between liberalisation and tighter control, the universities are currently under closer surveillance, confining their collective potentials. This is not an argument for a multi-party polity in China, an issue outside this paper. It is an argument for something much more fundamental and more directly relevant to higher education institutions and systemsrecognition of the irreducibly multiple quality of social relations, the coexistence of diverse kinds of association and identity (Sen 1999b). Even when the state has primacy in some respects, like the family it is only one circuit of social organisation. There are many kinds of state. Its role in social leadership and coordination may be large as in the Chinese polity, or small as in a minimalist laissez-faire regime. What is impossible is the normative project of a society in which individual, family, economy and every civil network have all been swallowed up inside the state; that is, state and society are seen as identical and ordered on the basis of a single concentrated political identity. 3 The Anglo-American countries are on another path. In higher education, discourse and policy are moving to exclude the public and collective, except in research, and towards a more extreme individualism. Broader person formation gets little attention. Social equity in higher education is seen as access to private goods, with less regard for social distribution. Yet the communicative and relational public suggest a more advanced role for higher education in social dissemination and networks. Even without state funding or provision, the relational civil public takes higher education beyond the economic idea of public goods. The universities continue as places that can initiate social and cultural transformations.
The question is, then, what do we make with this identification of difference? The answer is mutual learning. As Hall and Ames (1987) remark in Thinking Through Confucius: … it is precisely this recognition of significant differences that provides an opportunity for mutual enrichment by suggesting alternative responses to problems that resist satisfactory resolution within a single culture. (Hall and Ames 1987, 5) This two-way process of learning is facilitated by the integration of the two sets of discourses into a transpositional viewpoint, as outlined above, which provides a more inclusive theorisation of the individualised and collective outcomes of higher education.

Transpositional viewpoint
How then can we derive a transpositional viewpoint (Sen 2002) on the basis of the comparison in this paper? By now it will be apparent that neither separate set of constructs can achieve trans-position. Key elements in each tradition, such as the individual and the state, carry differing normative baggage. In the Anglo-American countries, it is impossible to derive a single meaning of public or public/private even within the political culture, let alone apply 'public' in settings where the limited liberal state and the public/private dualism do not apply. Likewise, in China, the pairing of gong and si and the nesting of smaller selves (xiaowo) and larger selves (dawo) is a Sinic-specific framework that fits some other societies better than others. It is a flexible framework in that it enables a variable level of autonomy for each xiaowo. There can be a good deal of space for the individual vis a vis the family and state, depending on how gong/si is practised. Inescapably, however, in this ordinal system the agency of the xiaowo is the gift of the dawo. The gong/si world, which is especially effective in social order, contains an inherent bias to top-down relations. Correspondingly, it underplays bottom-up agency. The independent potential and rights of the individual are underplayed vis a vis the collective. Also, the independent potential of grass-roots communal democracy, as suggested by UNESCO's common good, is underplayed vis a vis the state-led order. To the extent that grassroots democracy is consistent with China's tradition it is only when the gong/si coupling is not applied to full extent (as in the Imperial times but not the CPC era). To fully acknowledge the independent potentials of bottom-up agency, it is necessary to move outside the gong/si coupling.
However, the assertions and omissions in each political culture, while being revealed as culturally specific rather than universal, also suggest potential components of an integrated position. This is where the broad stretch of the comparison between Anglo-America and China is helpful. It brings a larger set of elements to the discussion. Each tradition places strong emphasis on specific domains, from individual to global, and specific social sectors. The Anglo-American tradition highlights the individual, distinguishes state from society, and foregrounds civil forums and organisations as part of its largest 'public'. The Chinese tradition distinguishes the individual and family, highlights the collective, emphasises the positive role of the state rather than defining it as a subtraction from the non-state, and nominates tianxia as a scale. Arguably, all are components of the transpositional viewpoint.
This transpositional framework can be applied to the observation of social relations, and higher education outcomes, in both Anglo-American and Chinese settings. It is an integrated transpositional viewpoint because all identified higher education outcomes, in both traditions, are made visible on an additive basis. No outcomes are excluded by the categories used. The inclusion in this study of Anglo-America ensures that in Table 2, civil and communicative society receives recognition and the state is distinguished from the economy, and from civil and communicative society, despite many co-penetrations between these domains and the state. The inclusion of China ensures the family is distinguished from the individual, that tianxia is a distinct domain, and the domains are not zero-sum but additive, co-relating in complex ways. The table avoids conflict between domains. It is always possible to position social elements in a logic of contradiction, as in the Anglo-American public/private dualism. However, that dualism is incompatible with transpositionality because it hides from view much of the potential for collective outcomes. Hence the public/private dualism is the one discursive element not taken into Table 2. Table 2 also draws attention to the fact that while individualised outcomes of higher education, benefitting a single unit, are usually understood in terms of single persons, there are other domains with a singular aspect, as systems, including the family, the state and even in tianxia, one world as a single interdependent geo-system. Only some domains in the individualised outcomes column have an identifiable single agent: person, family, nation-state. The other and larger domains of economy, civil society and tianxia are singular domains only to the extent that interactive relations are practised as bounded systems. By 'system' is simply meant a set of elements that form an interactive whole within defined boundaries. The contribution of higher education to the 'economy' is one case, though the global science system (Marginson 2021) is a more clear-cut example. However, the singular aspect of systems such as the economy or science is only one of their aspects, hence the use of brackets. Most of the actual activity of higher education and research in relation to the economy, civil society and tianxia falls in the collective-relational outcomes column.

Next steps
Next steps are to extend the study to additional political cultures and higher education systems, beyond those of the Anglo-American and Chinese worldsto enable testing of the elements in Table 2 and the development of further conceptual resources. For example, there is the French tradition of the republican 'public' in which the state is equated with the communicative, inclusive and democratic realm (Carpentier and Courtois 2020). Another is the Latin American autonomous university, the Cordoba model, in which the idea of the university as critically minded public sphere is a counter-hegemonic duty to the nation (Ordorika 2003). There are many cultural variations. Possibly (it has yet to be tried) each variation could be integrated into the transpositional view and thereby further enrich it.
The goal for social science is not just to develop a definitive table of the transpositional elements that comprise the outcomes of higher education but a transpositional theorisation of the worldwide diversity of those outcomes, grounded in unity in diversity (heer butong). Notes 1. China's UN ambassador Zhang Pengchun, vice-chair of the committee that prepared the 1948 Universal Declaration of Human Rights, took this sensibility into the drafting of the Declaration. It was difficult to translate ren, as awareness of the other, into English, though the point was agreed by the committee, but Zhang was able to remove from the document all references to religion, so that it could be universally shared (Liu 2014). 2. 大道之行也，天下为公。选贤与能，讲信修睦；故人不独亲其亲，不独子其子 … 力恶其不出於身 也，不必为己 … 是谓大同. 3. We note in passing that this is not a Leninist vision, it is something else. As Liebman (1975Liebman ( ) explains, in 1917 Lenin told his party to step back and let the masses lead. His State and Revolution, written between the February and October revolutions, barely mentions the Bolshevik Party (198)(199).