High Participation Systems of Higher Education
The world is rapidly becoming more educated at higher education level. In nearly all countries with per capita GDP of more than about $5,000 per annum there is a longterm tendency to growth of participation. The worldwide Gross Tertiary Enrollment Ratio (GTER) increased from 10% in 1972 to 32% in 2012, and is now rising by 1% a year. By 2012 the GTER had reached 50% in 54 national systems, compared to 5 systems twenty years before, and there were 14 countries with a GTER of 75% or more. The tendency to high participation systems (HPS) is common to countries that vary widely in rates of economic growth, education system structures, and financing arrangements, but share the tendency to urbanization. Possible causes include state policies, economic development, aspirations for social position, credentialism, global factors, and combinations of these. The paper describes the tendency to HPS, explores the possible explanations, and begins to reflect on the implications; on the way reviewing prior discussions of growth in participation including Trow (1974), Schofer and Meyer (2005), and Baker (2011). It closes with suggestions for further investigation.
higher education, national systems, enrollment growth, educational participation, mass education, globalization
[End Page 243]
There is a worldwide tendency to High Participation Systems (HPS) in higher education. Between World War I and the late twentieth century participation in higher education in the United States (U.S.) was very high in comparative terms but the world is now closing the gap. In the last 15 years or so there has been a remarkable surge in enrollment. This is not confined to wealthy countries. It affects nearly every country with a per capita income of over $5,000 USD per head, less than 10% of per capita income in the United States. Figure 1 captures the aggregated tendency. Between 1970 and 2013 students in primary education moved in line with world population.1 Between 1970 and the early 1990s the number of students in ‘tertiary education,’ as the term is used by the United Nations Educational, Scientific and Cultural Organization (UNESCO), equivalent to ‘higher education’ in American usage,2 increased faster than population, at about the same rate as world GDP. Then worldwide participation in higher education began to climb away from GDP, at an accelerating rate of change. Between 1970 and 2013 population multiplied by 1.93, GDP by 3.63 and enrollments in higher education by 6.12.
[End Page 244]
Between 1972 and 1992, the worldwide Gross Tertiary Enrollment Ratio (GTER)3 moved modestly from 10.1 to 14.0%. In the next two decades world GTER more than doubled, reaching 32.0% in 2012. Across all countries one in three young people now enter higher education, more than three in four across Europe and North America (Table 1; UNESCO, 2015; World Bank, 2015). Notions that only some people are capable of higher education are fading. The recurring concerns about over-education and graduate unemployment do not halt growth (Schofer and Meyer, 2005; Teichler, 2009). In some countries aspirations for higher education have become near universal. For example, families and schools in East Asia expect every student to be a high achiever, a strong starting point for tertiary participation (Marginson, 2011). International policy agencies such as the Organization for Economic Cooperation and Development (OECD) advocate open-ended growth. Few governments now set firm limits on the level of inclusion, despite fiscal constraints. The tendency to HPS does not mean that all students receive learning and credentials of high value. The point is that HPS are becoming universal to modern life, like hospitals, bureaucracy and police, transport and communications, piped water and electricity. It is likely that in no more than two generations’ time a majority of all working people will carry tertiary credentials.4 High participation higher education is a major transformation. It deserves more attention than it has received.
This article uses a multidisciplinary and historical method to describe the tendency to HPS and to explore possible explanations, in order to enhance understanding of the implications of this great change in the worldwide role of higher education. A key question is what drives the tendency to HPS. Is it state policies, economic development, the spread of aspirations for social position, credentialism, global factors, or some combination of these? On the way the article reviews contributions by Martin Trow (1974), Evan Schofer and John Meyer (2005), and David Baker (2011) that discuss the tendency to HPS. The conclusion considers possible lines of further investigation.
Many different social science theorizations and analytical and empirical methods can be used to illuminate social phenomena such as the tendency to HPS. Not all theories and methods operate in holistic fashion. In relation to the big picture explanation of HPS, methods developed for localized studies, in which the phenomena under investigation are conceived as a bounded set, are not sufficient to the task. For example, in the social science of higher education, one common approach is to employ a fixed theoretical framework as a lens though which to view [End Page 245] the research problem, and apply a linked methodology to the empirical analysis of primary or secondary data. However, no single theory (or methodology) can suffice for grasping the many-sided problem of HPS that is scarcely smaller than society itself, and encompasses domains that are often understood separately as educational, social, economic, geographic, demographic, psychological, cultural, and political, but at the level of society become combined in complex ways.
In considering the HPS phenomenon, what is needed is an approach that is sufficiently sensitive to local and national specificity and agency, and to emerging causal factors, while at the same time is capable of grasping a complex worldwide picture that is changing over time. This puts a premium on the identification, ordering, and synthesis of broad historical causes and tendencies; the imagining of semi-open intellectual structures (e.g. Appadurai 1996); and the willingness to utilize multiple theoretical lenses at need. It also requires methods that are sufficiently flexible to draw benefit from a range of specific studies that illuminate parts of the reality under investigation. The field of vision of the present article (though hardly its quality) resembles the work of the later Annales school in France, for example Fernard Braudel (1972; 1984). Braudel provided a social history that drew extensively on insights grounded in geography, demography, economics, and cultural and political studies, and distinguished multiple casuality and processes of long-term change from localized and contingent events. For the social historian, the different domains affect each other reciprocally, and no single domain is always determining. Moving between locations, and also over time, the drivers of change may shift from the politics of class or strategies of global competition, to the market economy, to science, and so on. Social structures are continually changing. All generalizations are provisional. Such an approach is helpful in understanding the fast-moving and semi-open world of global higher education, with its imaginative possibilities, and is equally applicable to the study of HPS (see more discussion in Marginson, 2008; Marginson, 2010; Marginson, 2011).
This approach contrasts with analyses that model the social world as a closed system, and social relationships as a set of relationships between discrete variables that are reducible to statistical treatment, for example in multivariate regression analysis. For the social historian, arithmetic, and statistical methods can be very useful, though in the form of auxiliary tools for investigating and illustrating relationships, comparisons and changes, rather than the privileged basis for forming a total picture. By definition, statistical relations between variables cannot constitute an explanatory historical synthesis. Statistical analyses depend [End Page 246] on the separation of elements, and are valid only while the variables are wholly independent of each other and hence unsynthesized (Keynes, 1973, pp. 276–7). In analyzing HPS as the manifestation of higher education, stratified populations, economy and labor markets, state policy and regulation, and global relations, the variables are not independent of each other but constitute a combined system of causes. Bourdieu and Passeron state in another context:
It is the system of factors, acting as a system, which exerts the indivisible action of a structural causality on behaviour and attitudes . . . so that it would be absurd to try to isolate the influence of any one factor, or, a fortiori, to credit it with a uniform, univocal influence at the different moments of the process or in the different structures of factors(Bourdieu and Passeron, 1990/1977, p. 87).
As Andrew Sayer remarks in Realism and Social Science (2000): “Statistical explanations are not explanations in terms of mechanisms at all, merely quantitative descriptions of formal (not substantial) associations” (p. 22). Rather than constituting an historical synthesis as such, multivariate analysis can provide some but not all of the materials used in synthesis. Further empirical observations and theoretical reflexivities also inform that synthesis; and in addition synthetic judgment, which is external to both empirical observation and statistical analysis, must be exercised. Too often multivariate analysis is treated as sufficient in itself, or extended beyond its capacity; for example when probabilistic methods are used to distinguish different degrees of causality among elements that are overlapping and interdependent; or when correlation or coincidence between two elements is held to constitute not a suggestive association between them, triggering further inquiry, but conclusive proof that they are causally related.
Universal Character of the Tendency to HPS
The global GTER is an abstraction. There is no world system of tertiary education. Rather, there are many national systems offering a mix of degree programs and other courses in various institutions, supervised, regulated, and in part funded by national governments. The upward leap in participation shows in the large number of systems affected. According to the UNESCO Institute of Statistics database, in 1992 only 5 national systems had achieved GTERs of 50%. However, by 2012, 54 national systems had passed the 50% mark; in 13 further systems the GTER exceeded 40%; and in many countries the most rapid growth had been in the previous half decade. [End Page 247]
In calculating national GTERs, UNESCO must recognize and define “tertiary education” amid diverse and fragmented provision5 and deal with incomplete country coverage. Nevertheless, the GTER provides the most comprehensive basis for cross-country and longitudinal comparison. Between 1992 and 2012 the GTER increased markedly in each world region except Central Asia (see Table 1). In or prior to 2012 it exceeded 50% across all of Western and Eastern Europe, North America, much of Latin America, and all East Asia except China. Nations with GTERs above 50% in 2012 included Albania, Armenia, Barbados, Bulgaria, Iran, Jamaica, Kazakhstan, Kyrgyzstan, Mongolia, Palestine, Saudi Arabia, Serbia, and Thailand. It seems that when the GTER reaches 50% or more the rate of participation keeps rising. By 2012 the GTER reached 90% in each of South Korea (98.4%), Canada, USA, Finland, Belarus, and Australia. In Cuba, Denmark, New Zealand, [End Page 248] Puerto Rico, Russia, Slovenia, Spain, and Ukraine, among others, it had passed 75%. The U.S. long had the highest GTER in the world but is now third behind South Korea and Canada.6
There are still gaps in participation at world level. In countries with very low per capita incomes, states generally lack the necessary resources and time horizon to mount large-scale tertiary systems. Between 1992 and 2012 the GTER in sub-Saharan Africa doubled but in 2012 it was below 15% in all systems for which data were available except Mauritius.7 However, the GTER is now climbing in the other erstwhile low participation zone, South Asia. In India between 1992 and 2012 it rose from 6 to 25% (UNESCO, 2015). In China the GTER was just below 30% in 2012. The official government target is 40% by 2020. There is regional variation in all large countries, and in China the regional factor remains pronounced: in 2010 the GTER was at 60% in Beijing and Shanghai regions but 25% in Sichuan and 18% in Yunnan (Yang, 2014).
Regional variation within China, and pan-regional variation on the world scale, are the outcome of the process of “combined and uneven development” that typically attends large-scale historical transformations (Naidoo, 2014). Figure 2 lists the 20 largest countries for which we have GTER and Gross Domestic Product (GDP) data. The most spectacular [End Page 249] enrollment growth at scale is in China, Ethiopia, India, Vietnam, Bangladesh, Indonesia, and the Congo, nations that combine high economic growth with incremental absorption of a precapitalist peasant economy and large-scale mobility between the countryside and the expanding cities. Higher education is concentrated in cities. From 1970 and 2010 the urban proportion of the world’s population rose from 36.6 to 51.6%. In South America it climbed from 59.8 to 82.8%, in China from 17.4 to 49.2%, and in Indonesia from 17.1 to 49.9%, though it grew more slowly in India, from 19.8 to 30.9% (UNDESA, 2012). The leading nations in Figure 2 are sometimes tagged as “developing,” as distinct from the “developed” nations. Given the global convergence in urban landscapes, industrialization, energy, transport, government, and education, it is more accurate to refer to these nations as “modernizing.” Higher education is a dynamic part of the modernization process.
Tertiary enrollments are growing more rapidly than GDP in 13 of the 20 cases in Figure 2. At the same time enrollment growth is well below GDP growth in five nations. Of these Russia, South Korea, and Japan have been affected by demographic downturns in the school leaver age cohort. Small annual increases in enrollment have been associated with substantial increases in the participation rate. From 2000 to 2012 the GTER rose by 20.7% in Russia and 15.7% in South Korea (UNESCO, 2015).
Table 2 details the 32 countries experiencing the largest rise in the GTER from 2000 to 2012, led by Turkey where the GTER rose from 25.3 to 69.4%. In all 32 cases the number of tertiary students increased on an average annual basis. In the majority of these nations the size of the school leaver age group declined. It is easier to support participation growth when the size of the age cohort is falling. All else being equal, a lesser level of infrastructure and staffing are required. In Table 2 such HPS include postcommunist Belarus, Ukraine, Estonia, Bulgaria, Romania, Hungary, Poland, Bulgaria, the Czech Republic, Slovakia, Slovenia, Croatia and Russia; and also Spain, Portugal, Ireland, and Mauritius. In some of these HPS the economic growth rate was modest. On the other hand the young tertiary age cohort increased markedly in Cuba, Colombia, Argentina, Chile, Saudi Arabia, Iran, Mongolia, China, Australia, Iceland, and Cyprus. The average economic growth rate rose by 4% or more in all of these except Cyprus (2.2%) and Australia (3.0%), helping to meet the increased costs of expansion. But whatever the conditions, demographic decline or advance, high economic growth or low, the outcome was the inexorable march to high and ever-growing levels of participation.
Figure 3 again focuses on the 32 countries that saw the largest increase in the GTER in 2000–2012. It maps annual growth in student
[End Page 251]
enrollment against annual growth in GDP. The countries are ordered on the basis of GDP growth, and the line of best fit in relation to student growth suggests a moderate association between GDP growth and enrollment growth. As with the large nations in Figure 2, in the majority of countries student numbers grew more quickly than GDP, in some cases by a considerable margin.
Political and Economic Explanations
What drives this ubiquitous expansion of HPS? In both popular and scholarly discussion it is often assumed the number of higher education students simply derives from state policy and regulation. It is true that in most systems, governments control the number of subsidized student places in public higher education institutions, and also in private sectors. In 2012 governments funded 69% of the costs of public and private tertiary institutions in the OECD countries (OECD, 2014, p. 236). Also, government policy and funding are often very visible at the time when national systems move to mass higher education. Take the case of China (Figure 4).
Though the 1980s saw explosive economic growth after the economy was opened up, participation in secondary education fell by almost 12% and tertiary education was locked at 2–3% (UNESCO, 2015). In [End Page 252] the 1990s educational participation started to increase and from 1999 onwards the GTER took off. What changed was that China’s partystate decided to grow higher education. It invested in infrastructure and scholarships, used the 211 and 985 programs to build research universities directly, and encouraged growth in provincial higher education institutions (HEIs). After two decades of economic enrichment there was enough pent-up middle class demand to ensure rapid take-up. Once the dynamic of expanding participation was established, it fed on itself, so that educational growth moved ahead of state targets and outstripped the very rapid and sustained growth in per capita GDP. The policy change is sometimes explained as driven by a shift from low-skill to high-skill manufacturing, but the shift in industry conFiguration followed rather than led educational expansion, with a German-style high skill vocational tertiary sector being created only in 2014 (Postiglione, 2014). A more plausible explanation is that the party-state found in the formation of an HPS a way to cultivate and satisfy popular aspirations at scale and secure a long-term instrumental social connection with the rising urban middle class.
So the state in China was central to the sudden transition to formation of an HPS. But is the state continuing to determine numbers in the same manner? And could it reverse the growth of high education? Probably not. Examination of the UNESCO data shows that significant falls [End Page 253] in participation rates are unusual, and that nowhere in the world, once the HPS dynamic has been released, does any state move to secure a lasting reversal of growth—despite the cost pressures created by expansion. However, there are many cases of states shifting part of the cost of growth to families and students.
The case of Australia also throws light on the state’s role. Figure 5 shows that between 1990 and 2013 in Australia the growth of GDP per capita was steady rather than spectacular, while the trajectory of the GTER was more uneven than in China. Its rise from 35 to 88% was punctuated by two periods of growth, between 1990 and 1996, and after 2009. The state played a different role each time. The first growth period followed more than a decade of steady state enrollment in the universities. The federal government intervened to expand capacity and grow places in both universities and technical and vocational education on the basis of enrollment targets, providing inducements such as income support (Croucher et al., 2013). In the second half of the 1990s there was a change of government, the part-withdrawal of student support payments, and neglect of vocational education. After 2008 another federal government moved again to expand participation. Instead of a planned increase in the number of places, it adopted a “demand-driven system” (Kemp & Norton, 2014) whereby it subsidized all places offered by accredited HEIs, without limit. Participation was determined directly not by government but by social demand, mediated by [End Page 254] institutional supply. Given that the state continued to subsidize higher education through each student place, and low interest tuition loans and student income support, fiscal cost was a continuing issue, but at the time of writing the demand-driven system survived. Between 2009 and 2014 the number of domestic students increased by 24.3% (Australian government, 2015).
Many other examples could be considered. On the basis of China and Australia, it appears the state role in kick-starting the growth of mass education may be larger than the state role after higher participation is achieved. Also, it is necessary to distinguish stated policy rationales from the political imperatives shaping government. Though states have autonomy, they also respond to elites and articulate citizens (Jessop, 2007). If state intervention at times appears to be purposely driven and based on economic or social agendas, at other times states seem to act more as facilitators. It is expedient for states to foster higher education as an opportunity framework. This broadens their legitimacy and political base. At the same time they are able to claim the expansion of higher education to be their own contribution to growth, global competitiveness, social justice, and inclusion.
Government decisions on the expansion of participation are also linked with various short-term triggers. At times the triggers are economic, for example subsidized training provided in a recession; or growth in funded enrollment in the face of particular skill shortages, such as geologists during a mining boom. Other triggers can be social, political and/or cultural. The abolition of apartheid in South Africa in 1990–1994 enabled demand for education among African people previously excluded. However, policy triggers are not long-term causes. There is still the question of whether the long-term drivers of the expansion of participation in higher education are political, economic, social, or something else; that is, if there are long-term drivers common to all HPS.
The public policy world favors explanations in which states routinely facilitate the economy. In the standard policy narrative about higher education, expansion is shaped by government and/ or market forces in response to the need for educated human capital. Higher education expands more or less in step with growing demands for graduate knowledge, skills, and certified professional competences. Economic demand is signaled in the labor markets by the wage returns to marginal productivity (Becker, 1964). Students focus on graduate wages and employability. People (or governments on their behalf) invest in education, in terms of time, income forgone, and tuition, to the point where the lifetime returns to degree holders equal the costs of investment. Since the 1960s the human capital theorization, the subject [End Page 255] of tens of thousands of empirical studies, has held center stage. Economists acknowledge that higher education can lag behind economic need; and it can over-provide graduates (credentialism); but they believe that higher education generally tends to equilibrium with economic demand. If it does not, then there must be a flaw in higher education or government enrollment policy.
However, those who look closely at the motors of participation are not persuaded. In an historical review Schofer and Meyer noted that “the rapid expansion of higher education in the 1960s does not coincide with especially large historical changes in occupational structures, job skill requirements, or labour market demands that would create a need for massive expansion of higher education.” Further, since the 1960s, the apparent association between economic growth and the growth of participation has been weaker than in that first human capital decade (Schofer & Meyer, 2005, p. 900 & p. 916). Table 2 and Figure 5 (above) suggest a moderate association between economic growth and the growth of student numbers but this does not mean that the expansion of higher education was driven by the economy: the causality could be reversed. The surge in participation is associated with a broad range of economic patterns. Economic growth spans from China at 10.1% per annum where the GTER rose 18.9% over 2000–2012, to Portugal at 0.2% per annum where the GTER rose by 21.3%.
This is not to say that economic conditions are irrelevant to participation. As noted, very poor countries cannot build an HPS; and economic growth increases the size of the middle classes, enhancing both the demand for education and the capacity to pay for it. Economic factors structure the supply of higher education, concentrating resources unevenly and stratifying the value of HEIs and degrees (see below). However, the point is that other factors are also at work, as well or instead of economic investment. For Trow (1974), as discussed below, economic demand for skills was less important than the main story which was the lifting of aspirations for social position. Ulrich Teichler (2009) noted that when graduates grow more quickly than professional jobs, graduates move down the occupational scale. “Research has shown that most persons seemingly overqualified did not face major hardships on the labour market but had acquired mostly a position only slightly lower than they had strived for” (Teichler, 2009, p. 28). After a career researching the education/labour market interface, he emphasized that “a ‘match’ between the number of graduates and the corresponding positions, or between the competences acquired during study and job requirements, cannot be expected.” [End Page 256]
It is clear higher education is often associated with social advantage. The question is the extent to which the link takes the form imagined in human capital economics. Many studies of the graduate labor market, including some working in a human capital framework, show that while some students follow earnings-related signals in choosing higher education, others do not. Status effects, status signals, and variations in status by field of study or type of institution, can be stronger than income effects (e.g. Arum & Roksa, 2014, pp. 80–81; Triventi, 2013, pp. 55–57; Zhao, 2012). Prospects of assuming a managerial role seem especially important for graduates from selective institutions, and those with generic degrees working in the public and NGO sectors, which includes many women (Roksa, 2005, p. 207; Hu & Vargas, 2015). Moreover, at point of enrollment students do not always take forgone earnings into account (Thomsen et al., 2013, p. 471). Robst (2007, p. 399) found they mostly know earnings only in their chosen occupation, not in related fields. The fit between higher education programs and occupations is only partly coherent, especially for generic degrees (Roksa & Levey, 2010, p. 391), and for graduates working outside fields of specific training, which often but not always generates income penalties (Melguizo & Wolniak, 2012, p. 383). Geiger (2014) has remarked that labour markets and education markets are on “separate tracks.”
The human capital narrative is accompanied by a counter narrative in economics and sociology, often drawing on similar rates of return data. In the counter narrative the economic role of education is not to impart useful skills but to function as a sorting or screening system for employers (for an early version see Berg, 1971). In this reading of the economy-education relationship, the lifting of participation is an outcome of the social competition for advantage through credentials as signs of employability, plus the vested interest of institutions in multiplying and elevating educational programs and credentials to cater for the demand for advantage. However, it is seen as wasteful to provide training in excess of substantive economic needs.
Social competition and the sorting role of credentials are readily identified in the transition from higher education to work. All else being equal, the zero-sum element in competition for leading social positions tends to lift the bar over time (Frank & Cook, 1995; Hirsch, 1976). However, the argument about credentialism is too one-sided. In a critical review Baker (2011, p. 9) has rightly noted evidence of advancing skill requirements at work, for example in relation to technology. Baker attributed the upgrading of both skills and credentials to a third factor, the “greater institutionalization of education” in society (p. 10). In the “schooled society,” education is the primary source of status attainment [End Page 257] and allocates people to positions in society on the basis of merit. Educational opportunity is seen as the source of social justice, and the extension of education as a sign of progress (pp. 10–13). Against the notion that changes is education are a function of changes in “occupations and the economy . . . external to the system of education,” Baker asserts that education is primary in social transformation (p. 10). But like credentialism this is too mono-causal. Higher education is not the only determinant of social allocations (Marginson, 2016a). Rather than defining institutional higher education as lifting itself without a platform, it would make more sense to develop an explanation in which the selfdetermined institutional advance of education, identified by Baker, is reciprocally joined to the advance of self-determining persons that use education.
Explanations Based on Social Demand
In 1974 Martin Trow published Problems in the Transition from Elite to Mass Higher Education. The argument moves seamlessly between social change and change in higher education. Trow argued that when higher education expanded from an “elite” system to a “mass” system educating at least 15% of the youth cohort, and then to a “universal” system at 50%, quantitative growth led to qualitative change. For example, the purpose of higher education shifted from “shaping the mind and character of the ruling class” (elite), to preparing a larger group in professional and technical skills (mass), to preparing the whole population in “adaptability” to social and technological change (universal). Access shifted from a privilege (elite), to a right (mass), to an “obligation” (universal) for middle class families (Trow, 1974, pp. 5–20). Trow understood his three kinds of higher education as Weberian ideal types, imagining them as historical stages, and also as differing practices in the present. Historically he developed prescient narratives about change, such as the passage of student selection from academic merit (elite), to programs designed to create social equality of opportunity (mass), to open access (universal), “because social inequalities show everywhere a stubbornly persistent effect on educational achievement” (p. 24). By also understanding elite/mass/universal as differing practices in the present, he explained how elite HEIs continued to flourish alongside mass and universal higher education, though he was concerned about potential negative effects of mass higher educational practices in elite universities.
Trow also explained the growth of participation. He noted the expansion of the economy and increased demand for skilled work but did not see them as the principal drivers. Nor was the state the main [End Page 258] source of growth. The motor was family aspirations to maintain and to improve social position. This led Trow to two crucial insights. First, there was no limit to aspirations for social betterment through education. It was not subject to economic scarcity. There would be “continued popular demand for an increase in the number of places in colleges and universities,” he stated. “It seems to me very unlikely that any advanced industrial society can or will be able to stabilize the numbers” (Trow, 1974, p. 40). Despite “loose talk about graduate unemployment or of an oversupply . . . it is still clear that people who had gone on to higher education thereby increased their chances for having more secure, more interesting, and better paid work throughout their lives.” As more people entered higher education, it became “a symbol of rising social status” (p. 41). It became quasi-compulsory. Nonparticipants had less life options. In the universal phase the emphasis shifted from participation for upward social mobility to participation as a defense against declining social position. Like Teichler (2009) later, Trow found graduate unemployment was not a problem because of the “educational inflation of occupations.” Graduate jobs were not fixed but moved down the occupational scale. As the number of graduates grew, they displaced those without college, sometimes using their educated capabilities to enrich the jobs (see also Baker, 2011, pp. 8–10). “What mass higher education does is to break the old rigid connection between education and the occupational structure” that prevented graduates from taking what were nongraduate jobs. Graduates could “seek employment without loss of dignity wherever the jobs may exist” (Trow, 1974, pp. 42–43).
Trow’s second insight was that government policy followed social demand for higher education, not vice versa. Government were under ongoing pressure, especially from middle class families, to facilitate the growth of higher education until saturation was reached, using both expanded supply and financial support for participation. This was plausible in early 1970s California with its politically responsive governor and congress. Trow’s prediction that participation would expand without limit in other jurisdictions (pp. 5–6, p. 40) was bolder, given participation was elsewhere lower than in California, and most governments around the world were enamored of ‘manpower planning,’ as it was called, using rates of return and employment data to plot a rational fit between education and the labor markets (or so they hoped). Few polities were as responsive to popular pressure as was California. Yet in the long run Trow has been proven right. Whether in multiparty or single-party polities, it seems that all nation-states now facilitate social demand for higher education. Further, Trow’s argument that growing [End Page 259] middle class aspirations drive the state-sanctioned expansion of the sector has a more general reach than arguments based on economic demand alone.
Urbanization and the Middle Class
Social demand is not an abstraction. It is rooted in human agents, who are stratified, segmented, and localized. Trow observes that “almost everywhere” the middle classes are the “first to take advantage of increases in educational opportunities of every kind and at every level” (Trow, 1974, pp. 23–24). Upper class people have less need for the status and economic benefits offered by higher education; while lower class people are often underprepared for higher education and lack the economic means (Marginson, 2015). Middle class people want their children to be upwardly mobile or to hold their position, and they have the discretionary income to finance those aspirations through tuition or taxation. The aggregation of the middle class in cities builds a critical mass of upper secondary students, and concentrates political pressure for expanded higher education provision, while enabling economies of scale. This is not to say that every person in cities is part of the middle class, but poor families in cities are much more likely to access higher education than poor families outside cities. [End Page 260]
Figure 6 lists 20 large countries by population in 2014, arranged in order of the extent of urbanization in 2011. The graph compares each nation’s urbanization to its GTER. With exceptions in South Korea, USA, Russia, and Thailand, the rate of urbanization exceeds that of tertiary participation. GTERs are well below the level of urbanization in the poorest countries: Congo, Pakistan, and Bangladesh. However, the main point is that the line of best fit suggests a strong association between urbanization and the GTER. This does not mean urbanization “causes” participation. It is more plausible to suggest that the expansion of the cities and growth of middle classes are both sourced in capitalist development, and growing urban middle classes sustain growth in the social demand for higher education.
In a study for Brookings and the OECD, Homi Kharas and colleagues defined a middle class person as someone living on $10–$100 USD per day in Purchasing Power Parity terms (Kharas & Gertz, 2010). On this basis 1.8 billion persons were middle class in 2009, 28% of world population, and a further 2% were “rich.” Thus 30% of all persons were middle class or above, a level similar to the world GTER of 28% in 2009 (UNESCO, 2015). The authors also found that “over the coming twenty years the world evolves from being mostly poor to mostly middle class.” The middle classes will rise from 1.85 billion persons in 2009 to 4.88 billion in 2030. Most of the growth will be in Asia-Pacific, from 0.53 to 3.23 billion, principally in China and India. In Latin America and the Caribbean growth will be from 181 to 313 million; in sub-Saharan Africa from 32 to 107 million (p. 5). Using a more restricted definition of middle class, the World Bank expects the middle class proportion of global population to more than double by 2030 (p. 6). Even on the basis of the more modest estimate, middle class growth appears to have major implications for the further expansion of participation in higher education.
A World Systems Explanation
One other explanation remains to be considered. In “The worldwide expansion of higher education in the twentieth century,” Evan Schofer and John Meyer (2005) noted the dynamism and ubiquity of the expansion of participation. They emphasized that it was “a worldwide phenomenon” (p. 899, their emphasis). They then moved from this insight to identify expansion as a single process throughout the world. This is more arguable.
Schofer and Meyer (2005) found that “the postwar shift to a liberal, rationalist, and developmental model of society generated a worldwide pattern of increased higher education expansion.” Hence “higher [End Page 261] education expands most rapidly in countries linked in organization and identity to world models” (pp. 903–905). The authors isolated a set of indicators said to reflect a “world model” of society. These were the numbers respectively of international human rights organizations, international scientific associations, countries with a national development plan, states classified as democracies, and international nongovernment organizations (NGOs). The last “reflects the overall structuration of the world polity” (p. 904): the more ties countries have with international organizations, the closer are these countries to the “world polity” (pp. 904–907). Statistically testing these indicators against trends in enrollments, they concluded that the advance of the “world model” has a “strong and significant effect on higher education enrollments” (p. 910). Enrollments expanded faster “in countries with strong links to the international system or the ‘world polity’” (p. 916). They concluded that a new model of society has been institutionalized across the world, “reflected in trends towards increasing democratization, human rights, scientization, and developmental planning. It is this global institutional and cultural change”—not middle class aspirations for betterment—“that has paved the way for hyper-expansion of higher education” (p. 900).
There are a number of problems with this argument. First, the use of statistical methods is open to the critique outlined above. As noted, on pp. 909–913, numerical associations are framed as causal relationships (“strong positive and significant effect”). Quantitative trends are deployed as indicators of qualitative change, and the numerical combination of various quantitative trends is substituted for the more challenging but more rewarding process of historical synthesis. Second, the assumptions underlying Schofer and Meyer’s variables are often weak. For example, the number of scientific associations is not sufficient to prove an increasing breadth and depth of commitment to scientific practices. It may reflect growing ease in cross-border communication and cooperation. Likewise, a quantitative trend in the number of NGOs is not an evidential basis for a “world polity,” any more than growth in the raw number of multinational corporations in itself signifies closer integration of the world economy. Third, the “new model of society” is grounded in an idealized American society based on liberalization, democratization and human rights. Yet the rapid expansion of higher education is occurring in all societies and polities, as the authors note (p. 898, p. 908). In fact recent expansion is as rapid in dynastic singleparty China as in multi-party in India, and more rapid in Russia than the USA. Fourth, educational expansion is modeled as driven by “global” trends from above (p. 899). Much as in Baker’s (2011) idea of institutional education as its own driver, the families that sustain growing [End Page 262] participation are left in shadow. This sits oddly with an emphasis on democratization and human rights as components of the emergent world society. Fifth, Schofer and Meyer blended an abstract globalism with methodological nationalism (Shahjahan & Kezar, 2013). They saw the tendency to HPS as universal Anglo-Americanization. It is another in the long chain of theorizations of globalization that present a culturally-specific global integration as universal, and more overwhelming because it is seen as external to real human agents.
Finally, Schofer and Meyer’s world systems appear to subsume, rather than work with, autonomous national systems of higher education. They rightly noted that in expanding higher education systems there is “a great deal of isomorphism around the world.” HEIs taught similar subjects with “similar perspectives leading to very similar degrees and to credentials that take on worldwide meaning.” The points about convergence in systemic practices, and the spread of science, are suggestive. But the claim that these convergences constitute “a highly expanded, and essentially global, system of higher education” is a step too far. “Every society has a schooled population and institutions that function as a greatly expanded set of receptor sites collecting ideas and practices from world society,” generating “a great deal of global integration” (p. 917). It is a very long stretch from policy mimetics to a single meta-system of higher education, a “world-society,” and a “world polity.” The claim lacks empirical grounding.
Like all bounded or semi-bounded systems, national HPS are positioned in larger settings (King, Marginson & Naidoo, 2011; Rizvi & Lingard, 2010). Worldwide higher education is shaped by (a) crossborder interactions between separate national systems, and (b) global systems of relations. It is in relation to the latter that globalization is most transformative. Arguably, Schofer and Meyer (2005) confused cross-border policy borrowing (a) with global systems (b). In fact states have long functioned as “global competition states” (Cerny, 1997; also see Bayly, 2004), watching each other closely, building capacity, imitating and innovating by turns, in the endless striving for strategic advantage. Cross-border flows encourage parallel evolution and convergence between HPSs, and one form of partial convergence is the expansion of educational participation. Yet the rhythms of growth vary between countries. The common tendency to HPS is articulated through differing state-society-education assemblages. The presence or absence of the features of Schofer and Meyer’s world system can be subjected to empirical test. In the absence of stronger proofs it is more plausible to see each national system as finding its own HPS pathway, within the [End Page 263] larger setting in which imitating behavior is normal, but not driven from a single pivot.
National System Variations
Though the global dimension of higher education is increasingly important, it is not the master dimension. Modern higher education is a creature of the nation-state (Scott, 2011) and national aspects remain primary. Most students and faculty are not mobile across borders. Most HEIs are not research-intensive and not strongly actively in global networks. Nearly all HEIs still draw their main funding and status from national sources. In short, each HPS is both autonomous and part of a larger set of relations, and evolves in a dialectic of sameness and difference with other HPS. The relationship between global patterns and self-forming national systems varies from case to case.
There are many continuing differences between HPS in system structure and political economy. Some systems are unitary, with one mission, the research and teaching university. There are binary systems with two missions, as in Germany and the Netherlands. Some HPS use classifications to sort different missions, as in the United States and China. Private sectors play various roles with no dominant pattern (OECD, 2014, p. 425). In some HPS private HEIs are the main medium of growth: for example Korea, Hong Kong, Brazil, India, and the Philippines, and for-profit HEIs in the United States. In Poland the 1990s growth of the private sector is reversing (Kwiek, 2014). There are also divergent arrangements concerning student fees and financing. In most HPS the household carries part of tuition costs; some systems use income contingent tuition loans, which soften the direct impact of costs by postponing repayment until the means are available; and in some HPS, such as the Nordic systems and Germany, tuition is free and financed from general taxation. Average fees exceed USD $10,000 per year in English-speaking countries but are below $2000 in most of Europe (OECD, 2014, pp. 260–276), where it is widely assumed that heavily subsidized tuition is essential to high social participation. In East Asia, a universal culture of educational achievement sustains majority private costs in Korea and Japan; and there is considerable investment in extra learning outside school hours even in poor families (Bray, 2007). In some but not all HPS, governments use financial incentives to foster the extension of participation to marginal, indifferent and underrepresented social groups.
Differences between HPS in institutional mission, classification, and financing mostly also have implications for variation in the extent and forms of stratification within systems, meaning vertical differences in the status and resources of institutions. Few families using higher [End Page 264] education are unaware of status differences between HEIs, disciplines, and levels of qualification. As systems grow, the question about equity begins to shift from access? to access to what? In “elite” systems in Trow’s (1974) sense, every place carries a reward, and equity turns on inclusion/exclusion. In HPS, with a continuing elite subsector, the binary structure becomes ternary: (1) high value inclusion, (2) low value inclusion, (3) exclusion. There are continuing issues at the boundary of inclusion and also in relation to social mobility, especially for low socioeconomic status and migrant families, rural students, students from underrepresented ethnic and culturally-defined social groups, students with disabilities, and others.
Bourdieu’s (1988) theorization suggested a prima facie tendency for expansion to generate inequality, in that the number of high value student places shrank as a proportion of the total places. This is plausible, but the tendency can be modified by regulated system design. Shavit and colleagues (2007) found that the growth of participation generated greater stratification, and possibly greater inequality of opportunity, though prior structural distinctions between systems affected comparative outcomes. At present the “World-Class University” movement in both established and emerging systems (Salmi, 2009), associated with the concentration of the research function in leading HEIs, tends to promote vertical stratification. However, this movement is nuanced by system. Market competition also promotes vertical differences, but again the use of market forms varies markedly around the world. Empirical evidence in North America has suggested an increasing vertical ‘stretch’ between HEIs in resources and status (e.g. Boliver, 2013; Davies and Zarifa, 2012). Yet enhanced stratification is not universal. The Nordic countries retain relatively flat system structures based on universal free access and high quality. The topic of HPS and stratification in comparative context is too large for this article (however, see Marginson, forthcoming).
Nevertheless, despite varied policy assumptions, financing, structural forms and stratification, participation everywhere advances. This suggests that relationships between system structure and participation, and private costs and participation, are not subject to universal laws alone, but are also system-specific. It is also possible these relationships can vary over time, and between social groups. The fact that the Englishspeaking countries have stepped up private costs without permanently reducing total participation confirms Trow’s point that as participation reaches majority levels it becomes quasi-compulsory. Middle class families cannot absent themselves because of the social and economic costs [End Page 265] of nonparticipation. This suggests that as systems expand, the elasticity of student demand falls. However, this requires empirical test.
History combines long-term evolutions and transformations with localized and episodic events; and combines world phenomena with local and regional articulation and variation (Braudel, 1984). World tendencies have local motors, as well as cross-border and global motors. The formation and growth of High Participation Systems of higher education is common to every nation that achieves minimum affluence and coherence. This is highly significant. The precise forms of HPS, system structures and rhythms of expansion, and the financing of students, vary from country to country. What does not vary is the open-ended character of the growth of participation, as forecast by Trow (1974). There is no limit to the GTER until saturation is reached, as in South Korea. This again is highly significant. A third significant factor is the speed of transformation.
No single causal explanation for the growth and ubiquity of HPS is convincing. Some factors are stronger than others. Governments are often instrumental in growth but also work in conjunction with other drivers. Though the expansion of higher education is simultaneous with the advance of vocational skills and their use, it is likely that causation runs in both directions. The tendency of aggregated economic demand is fragmented, variable, and appears insufficient to explain the ubiquity and dynamism of HPS growth. Explanations based in social demand for opportunity and position look stronger, if only because middle class family aspirations are similar across the modern world. Trow (1974) on California is the earliest but the most potent explanation.
The expansion of urban middle classes appears to be key in the rapid GTER growth in many countries in the last 15 years. However, relations between social demand and the supply of higher education are also mediated by the growth and diversification of economies; by state decisions and often, state investment; and by education institutions and expanding credentials. It is also more strongly expressed in some family cultures than others. Even if aspirations for social betterment are the universal factor at work here, as Adam Smith suggested (1979/1776, p. 441), they are not the mono cause of historical change. Family agency is expressed and realized in tandem with political, economic, and cultural factors. Nested in these contexts, it has varied manifestations across the world. The remarkable fact that distinguishes this period in history is that the ambition for higher education now appears unstoppable, [End Page 266] almost everywhere, while the time span for the realization of ambition is shrinking. (Perhaps too much is expected of higher education). It will not always be so. There is no sign yet of what will happen next.
Matters for Investigation
The foregoing suggests a number of propositions about the evolution of HPS that constitute matters for further historical inquiry and empirical investigation:
1. The tendency to growth of participation in higher education has no natural limit in countries with a minimum level of per capita GDP. The long-term tendency to growth is not subject to economic scarcity; either in terms of labor market options for graduates, or the combined public and private costs of expansion. However, elite places are differentiated on the basis of scarcity and cost.
2. Social elites have no intrinsic interest in the expansion of HPS. However, a key factor driving the growth of participation is effective political and/or economic demand from other families that want higher education for their children.
3. Popular demand for opportunities affects policy on educational participation everywhere, whether the HPS is located in an electoral democracy or not.
4. The expansion of participation is associated with urbanization, and in particular, the growth of urban middle classes.
5. The role of the state varies at different points in system evolution. Government policy and financing plays a stronger role at the early stages of HPS. Once a mass system is in place, popular demand is rising and the costs of nonparticipation are apparent, the state is less crucial to growth and its effective role diminishes.
6. As systems expand, the elasticity of student demand falls.
7. As the systems expand and demand elasticity falls, governments tend to require families to pay a growing share of the costs.
8. As systems expand, between HEIs there is an increased vertical “stretch” in (stratification of) the value of student places.
1. Between 1970 and 2013 the world net enrollment rate in primary education increased from 71.9 to 89.0% but this did not lift the growth of primary enrollment above growth of population because of an upward shift in the age structure (UNESCO, 2015).
2. “Tertiary education” in the standard UNESCO and OECD data sets includes both Type 5A degrees of three years or more and shorter Type 5B programs of two years fulltime equivalent, for example in North American community colleges. [End Page 267]
3. The numerator of the GTER is the total enrollment in two-year programs and above; the denominator is the school leaver age cohort in the national population.
4. This includes participation in uncompleted programs, as well as completed programs.
6. The GTER is boosted by participation additional to the age cohort and appearing in the numerator but not the denominator, including migrants, international students and mature age students. Mature age students can push the ratio over 100% in very high participation systems. Between 2000 and 2012 students studying outside their country of citizenship increased from 2.1 to 4.5 million (OECD, 2014, p. 344). In 2012 the inclusion of onshore international students increased the apparent rate of entry into tertiary education in 2012 by 26% in Australia, 23% in UK, 12% in Austria and 11% in the Netherlands and Switzerland (p. 81 & p. 338). In some countries a net outflow of students reduces the GTER.
7. There were no UNESCO data for Botswana, Lesotho, Nigeria, Sierra Leone, South Africa, Tanzania, and Zambia.