Ascription, Achievement, and Perceived Equity of Educational Regimes: an Empirical Investigation

This article considers how countries differ in the opinions that citizens hold about the fairness of the social and educational system. From the literature, we derive a typology of four educational " regimes " , based on differences in educational system, labour market and welfare state design. We then use data from the ISSP (2009) to investigate how much weight respondents attribute to ascribed characteristics (e.g., being born in a wealthy family) and individual responsibility (e.g., working hard) to explain success in life or at school. We also examine how these judgments relate to the educational background of the respondents. We consider how these patterns correspond to the existing knowledge on social and economic inequality and what this means for the legitimacy of the social system. In this article, we will analyse perceptions about the fairness of society as a whole, and the school system in particular. Even though often unfulfilled (cf. [1]), many educational and social policies uphold a " meritocratic promise " , maintaining that the allocation of social positions should be based on personal achievement rather than ascription of characteristics relating to social origin. Consequently, every member of society should, regardless of social origins, have the same opportunities to prove himself. This ideology thus aims to discard all barriers between social classes that prevent social mobility between generations. Importantly, this does not mean that social inequality in itself would be unfair. As scholars such as Parsons [2] already defended, some degree of social inequality might be necessary, as society requires different profiles to occupy different positions, and the more important positions have to be the best rewarded. The aim therefore is not equal outcomes, but a particular distribution of possible outcomes that are unrelated to a person's social background. In this paper, we will focus on the criteria according to which social positions are distributed: do they depend on the quality of the ascribed assets (such as social class or wealth of the parents), or are they a function of individual achievement (i.e., a combination of talent and effort). With the concept of " social fairness " , we will then refer to the extent to which the importance of ascribed assets has been overruled by personal achievement. Note that the notion of " achievement " has in turn been interpreted in different ways across political cultures. In a nutshell, the meritocratic school sees personal merit as …

not differentiate between the two variants of achievement: this is due to constraints in the available empirical data.
In a general way, as reflected in Figure 1, social positions ("destinations") are assumed to depend partly on social origin and partly on personal achievement.Both can have a direct effect on social status: for example, access to privileged networks may help children from advantaged families to reach higher social positions (direct effect of origin on destination), while hard work may help to reach a high social status as well (direct effect of achievement on destination).However, in modern societies, the strongest determinant of social success is educational achievement; for example, Bell [5] has argued that information societies put such a high premium on knowledge and skills that educational attainment becomes the ultimate foundation of social success.As strengthening the link between education and destination would imply loosening at least the direct association between social origin and destination, education is seen as the main leverage of social mobility in modern society.
definition that does not differentiate between the two variants of achievement: this is due to constraints in the available empirical data.
In a general way, as reflected in Figure 1, social positions ("destinations") are assumed to depend partly on social origin and partly on personal achievement.Both can have a direct effect on social status: for example, access to privileged networks may help children from advantaged families to reach higher social positions (direct effect of origin on destination), while hard work may help to reach a high social status as well (direct effect of achievement on destination).However, in modern societies, the strongest determinant of social success is educational achievement; for example, Bell [5] has argued that information societies put such a high premium on knowledge and skills that educational attainment becomes the ultimate foundation of social success.As strengthening the link between education and destination would imply loosening at least the direct association between social origin and destination, education is seen as the main leverage of social mobility in modern society.The erosion of ascription as the dominant mechanism for determining life chances, in favour of education, indeed seems to have made current Western society more meritocratic than ever before [6,7].However, there are some concerns that society still is not fully meritocratic.First, cultural and economic disadvantage still appears to affect educational attainment, net of personal merit [8][9][10].Shavit and Blossfeld [11] indeed find that in many countries the link between social origin and educational outcomes has remained remarkably stable throughout the 20th century.Social origin would then still affect destinations, but now through the indirect educational path.Secondly, occupational destinations might still depend on social origin, net of educational achievement, e.g., through access to networks [12,13].

Educational Regimes
As we discussed above, the meritocratic ideal ("only ability and effort should determine social status") can be distorted in two ways: by social origin affecting educational achievement (net of merit), or by social origin directly affecting destinations (net of educational attainment).In the following sections, we will discuss for both pathways how the design of the educational and social system may affect the level of this distortion.

Skill Specificity and Stratification
In the literature, secondary education systems have been mainly analysed in terms of two dimensions [14].First, the dimension of skill specificity has been used to indicate the dominant orientation of the educational system, between a "general" and a "vocational" pole.General education systems are mainly oriented towards supplying generic skills, seeing preparation for further education as their major objective.Vocational systems are mainly oriented towards supplying occupation-specific skills, with the aim of preparing students, in particular those not deemed fit for further education, for direct entry in the labour market.The archetypical example of a general skills The erosion of ascription as the dominant mechanism for determining life chances, in favour of education, indeed seems to have made current Western society more meritocratic than ever before [6,7].However, there are some concerns that society still is not fully meritocratic.First, cultural and economic disadvantage still appears to affect educational attainment, net of personal merit [8][9][10].Shavit and Blossfeld [11] indeed find that in many countries the link between social origin and educational outcomes has remained remarkably stable throughout the 20th century.Social origin would then still affect destinations, but now through the indirect educational path.Secondly, occupational destinations might still depend on social origin, net of educational achievement, e.g., through access to networks [12,13].

Educational Regimes
As we discussed above, the meritocratic ideal ("only ability and effort should determine social status") can be distorted in two ways: by social origin affecting educational achievement (net of merit), or by social origin directly affecting destinations (net of educational attainment).In the following sections, we will discuss for both pathways how the design of the educational and social system may affect the level of this distortion.

Skill Specificity and Stratification
In the literature, secondary education systems have been mainly analysed in terms of two dimensions [14].First, the dimension of skill specificity has been used to indicate the dominant orientation of the educational system, between a "general" and a "vocational" pole.General education systems are mainly oriented towards supplying generic skills, seeing preparation for further education as their major objective.Vocational systems are mainly oriented towards supplying occupation-specific skills, with the aim of preparing students, in particular those not deemed fit for further education, for direct entry in the labour market.The archetypical example of a general skills system is the USA, while the dual model of Germany is the best-known vocational-oriented system [15].
While the specificity dimension mainly describes how differentiation is implemented, the stratification dimension covers the extent to which the system differentiates between pupils.The most salient characteristic in this dimension is the presence of early tracking."Tracking" refers here to the practice of directing pupils via distinct educational trajectories towards different educational and occupational end-points, usually based on some observation of their abilities, but often affected by their social background as well (see below).While all European countries implement separate tracks for pupils above a certain age, this tracking age differs widely between systems: many countries do not track students until age 16, while others, such as Germany, have different tracks starting already at age 10.The earlier the tracking starts, the more it can be supposed to influence the educational career of the students involved.However, Dupriez, Dumay, and Vause [16] have emphasized that the absence of early tracking does not necessarily mean that classes are truly heterogeneous.For example, Southern-European countries often sort out struggling students through extensive use of grade retention, while in most Anglo-Saxon countries students take courses flexibly, at different levels for each discipline (ability grouping).Only in the Nordic countries can classes be considered truly heterogeneous, with differentiated teaching and remediation classes helping all students to master the same common core curriculum until age 16.In practice, the concepts of specificity and stratification are correlated, although not identical.For example, the use of tracking usually implies that large differences are created in the curriculum covered in the different tracks, and the resulting differences in learning outcomes already represent some sort of "skill specificity".The two dimensions thus cannot be easily separated.In this article, we will use a typology of educational systems that relies on combinations observable throughout the Western world, which gives rise to four broad categories of educational systems (see Table 1, columns 1 to 4).

Adding a Labour Market Perspective
It has been repeatedly argued that the educational system design cannot be understood in isolation from characteristics of broader social spheres, in particular the labour market and the welfare state.First, functionalists argue that the observed cross-country variation in educational systems mainly reflects different requirements from the labour market, in particular the demand for either specialized or general-educated employees [20].Thelen [21] has traced the origins of these different skill regimes back to differences in industrial relations at the beginning of the 20th century.In most European-continental countries, she argues, skill supply was at that time still strongly controlled by the traditional artisanal sector.Employers and trade unions from the developing industrial sectors thus had to work together to break down the grip of the artisanal sector in order to create an alternative 1 Countries will be referred to with their two-letter ISO 3166 codes.The participating countries are Austria, AT; Flemish Region (Belgium), BE-F; Germany, DE; Italy, IT; Spain, ES; Portugal, PT; France, FR; Australia, AU; New Zealand, NZ; Great Britain, GB; United States, US; Denmark, DK; Finland, FI; Iceland, IS; Norway, NO; Sweden, SE. channel for skills supply, which gave birth to a strong vocational training sector.In the Anglo-Saxon countries, by contrast, the artisanal sector was less powerful, which made control of occupational skill supply a permanently conflictual issue between employers and trade unions; this gave less room for the development of vocational education, which required the collegiate involvement of both labour market actors.Moreover, Estevez-Abe [18] suggested that this early divergence is reinforced by differences in social protection schemes and economy coordination.She argues that individuals are more reluctant to invest in specific than in general skills, as the advantages associated with the former are tied to a limited number of jobs (only those within a single industry of firm), while the latter are transferable from one job to another.Investments in specific skills thus require some guaranteed "return on investment".In coordinated economies such as those from continental Europe, systems of collective wage-bargaining reduce the individual risk of wage depression: even when changes in labour demand would make a number of jobs redundant, those who were trained specifically for these jobs would still earn a satisfactory wage.Hence, specific skill investments are less risky in coordinated economies than in liberal economies, explaining the deep divergence in skill specificity between continental Europe and the USA.Employers then adapt their product market strategies to the existing skill structure: in specific skill systems, such as Germany, firms specialize in the production of high-quality goods requiring an abundance of medium-skilled technical workers, while the polarized skill structure of general systems (in which an advantaged share of tertiary graduates have high general skills, but those who do not make it lack a strong vocational alternative) leads to a combination of highly innovative branches (e.g., ICT) with low-skilled mass production.This choice for a certain product market strategy of course raises again the demand for workers with the required skills; hence, skill specificity and product market strategies complement each other in what can be called a "skill equilibrium."Hence, in Table 1, a 5th column can be added complementing the educational system component with a labour market regime characteristic.

Education and Power Relations
Educational system design is not simply an objective response to labour market needs.The design choices were often the subject of intense political power struggles as well.Archer [22] argues that the degree of centralisation reflects the social and political conflicts during state formation, and that this is reflected in the degree of stratification as well (as parallel structures survive easier under weak central governments, cf. the grammar schools in England).There may be a strong parallel here with the welfare state typology developed by Esping-Andersen [19], who traced variations in welfare state design back to the structure of the power relations between the different social classes.In particular, he argued that in countries where the political left was fragmented, state intervention remained limited (Liberal world).By contrast, where the left was stronger (mostly due to farmer-worker alliances), it implemented a highly redistributive welfare state (Social-Democratic World).Finally, where Christian-democracy, characterized by a class-cutting constituency, was strong, the emphasis usually shifted from redistribution to insurance.Busemeyer [23] and Sass [24] apply this welfare state perspective to educational system design, arguing that left parties will be supportive of educational policies that benefit the lower tail of the educational attainment distribution, while conservative parties will oppose any drastic expansion of educational opportunities because of budgetary reasons and fears for "expectation inflation" among the working class.Indeed, Braga, Checchi, and Meschi [25] produce strong historical evidence for this correlation between political power and educational positions.By matching educational reforms from the 1930-2000 period in 24 countries with the prevailing political orientation of government, they demonstrate that educational reforms that reduce the dispersion in educational attainment were indeed implemented mostly by left wing governments, while right-wing governments preferred more selective policies.This correlation between welfare state and educational system design is shown in the 6th column of Table 1.
Empirically, indicators from these three spheres in society (education, the labour market, welfare state) have indeed been shown to be strongly interdependent, constituting recognizable "clusters" of countries [26][27][28][29].In this article, we reflect this interdependency in the concept of the "educational regimes" which are mainly identified by their geographical labels, which point both to the educational components of the typology (i.e., the differentiation mechanism and the way vocational education is developed) and to their welfare state and the production regime components (Table 1).

Cross-National Differences in Educational and Social Fairness
In this section, we will explore how the set-up of educational regimes identified above may affect the level of fairness in the educational and social system.We will in particular point at the effect of the two dimensions of educational regimes (stratification and skill specificity) and one socio-economic dimension (economic inequality).

Continental Countries
First, the educational and social systems of the European-continental countries distinguish themselves by their high level of stratification [14].This means that educational attainment determines occupational and social success to a large extent.Hence, the direct effect of social origin on destination, net of educational attainment, is expected to be relatively weak in these countries.
However, the indirect effect of social origin, i.e., mediated through educational attainment, can be expected to have remained rather strong.Continental educational systems rely heavily on early tracking, which has been repeatedly shown to strengthen the link between social background and educational performance [30,31] as it negatively affects the educational opportunities of socially disadvantaged students.This is probably due to two reasons.First, socially disadvantaged students seem to be disproportionally selected into less prestigious tracks, even after accounting for prior performance.This claim has been empirically substantiated in studies from several countries (see [32] for Flanders, [33] for Germany, and [34] for France).An explanation for this effect is that the educational ambitions of young pupils are strongly influenced by the role models they perceive in their environment [35].When tracking decisions have to be made already at a young age, the parental voice is still utterly important, and the impact of socio-economic background on track placement will be stronger [36].Secondly, less prestigious tracks usually offer less stimulating learning environments which may hamper their performance [37].The argument here is that shifting students to a less demanding track, where the curriculum is less challenging and the learning conditions far from optimal, leads to ignoring learning difficulties instead of adequately addressing them [38,39].Moreover, educational resources tend to be unequally distributed across tracks, with the most experienced and most capable teachers often being assigned to the higher tracks, leaving the lower tracks to the less experienced teachers [40].
Within the continental group, a further distinction can be made between Germany and Austria, with rigid tracking starting already at age 10, and the Flemish Region, where a common primary school lasts until age 12 and the onset of tracking follows more gradually.Moreover, while Germany and Austria are well-known for their strong "dual" system training vocational students at the workplace, vocational education in Flanders is predominantly school-based.

Anglo-Saxon Countries
Educational and social policies in the Anglo-Saxon countries usually pursue a strong individualistic meritocratic ideal, in which educational policy aims at removing unjustified barriers that may prevent talented students from achieving success.Hence, while rigid stratification into different tracks at a young age is seen as unfair (as it limits young person's chances based on too limited information on his competences), there is in principle nothing wrong in rewarding individual merit.Hence, flexible grouping of students in differentiated streams is permitted as long as this is done on the basis of their performance, and as long as there are continuous options to promote to a higher level when achievement is satisfactory.This explains the Anglo-Saxon model of ability grouping.However, Green, Green, and Pensiero [41] have argued that precisely this high reliance on individual choices systems fosters social inequity at school: "the greater the variety of different routes through the education system, the greater the likelihood that socially differentiated aspirations and expectations will structure student choices".
This effect of educational system design is probably amplified by the high income inequality in the Anglo-Saxon world, which is characterised by less redistributive tax systems and the more polarized skill structures that are typical for the general skill regimes.High socio-economic inequality indeed has been shown to negatively affect the educational achievement of children from socially disadvantaged parents [42], reflecting an early concern by Boudon [43] that "any lessening of stratification through a reduction of economic inequality is probably more likely to affect educational inequality than any other factor".This is most obvious in the American system, where a high residential inequality combines with a stratified educational market, with a strong opposition between private and public schools.This generates large differences between school systems, school districts, and individual schools in the socio-economic and ethnic composition of the student body [44,45].Relatedly, the expenditure per pupil varies to a much larger extent than in the standardized arrangements common to the continental countries.This may have drastic effects on the fate of disadvantaged students.Recent data indeed point at a strongly reduced social mobility, in particular in the United States [46]; similarly, Jantti, Bratsberg, Roed, Raaum, Naylor, Osterbacka, Bjorklund, and Eriksson [47] report how earnings mobility is nowadays much lower in the USA than in Europe.

Mediterranean Countries
In the Mediterranean countries, labour market outcomes are often less related to educational attainment [17].This is mainly due to the absence of a strong institutional link between education and the labour market.Overall, this is expected to hinder social mobility: when education ceases to play its role as the dominant allocation mechanism, other mechanisms will take over, and these mechanisms are probably more dependent on social origin.As Dronkers [48] has argued, "leaving the (socially inevitable) selection to the labour market instead of the educational system creates the chance that social inequality between students from different strata will become even greater than the inequality that exists within education.After all, selection is even less universalistic (meaning the same criteria apply to everyone) on the labour market than it is in education." Like in the other regimes, there is considerable internal variation in the Mediterranean group as well.For example, France is to some extent a hybrid sharing a number of its characteristics with the Continental group, such as a more extended welfare state [20] and a more pronounced link between education and the labour market (although the strong French preference for abstract knowledge has long hindered the extension of vocational education).

Nordic Countries
Finally, the Nordic countries have always put a strong emphasis on education as this was believed to be "the great equalizer", enabling social mobility and generating equal life chances.The publicly funded comprehensive school system that was established during the 1960s and 1970s in Scandinavia is the educational expression of this policy geared at democracy, equality and progressiveness [49].This educational system had to combine a high average performance level with equal chances for everybody and a low social reproduction of educational achievement.This was pursued by heavy public funding, a well-trained teacher staff and an educational structure that promoted equal opportunities by postponing selection and offering individualized assistance to those in need.Peter, Edgerton and Roberts [50] summarize the social-democratic view on education as follows: "Equalising access to quality education at all levels is vital to equality of condition for citizens.Equal access to comprehensive education is seen as the right of all citizens and as key to social security and a united, prosperous nation."Together with the low level of income inequality, this has led the Nordics to report the highest levels of social mobility and equity [47].
However, over the past two decades, the Nordic model has come under increasing pressure towards reducing its egalitarian objectives in favour of efficiency concerns, most visibly in the proliferation of privately managed schools in Sweden.As a result, differences between schools and pupils have been steadily growing.Hence, while "it is still justified to speak of the five Nordic countries as a rather distinct group, social-inclusive policies have clearly been reformulated and delimited, related to a strengthening of the economic-utilitarian functions of education and a weakening of central education governance" [51].

Relationship with Individual Educational Background
In this article, we will pay particular attention to potential differences in the perceptions of the high-and the low-educated about the fairness of the social and educational system.Overall, it has been argued that the high-educated tend to be less sensitive about remaining inequalities of opportunity.Warikoo and Fuhr [52] and Khan and Jerolmack [53] have provided examples of how students in elite schools believe that their educational success is completely merited, neglecting other factors such as social advantage (and luck).Similarly, Räty, Snellman, Mantesaari Hetekorpi, and Vornanen [54] argue that social representations of educability depend on someone's own school experiences.Shedd and Hagan [55] add to this that segregated learning environments restrict the frame of reference of youngsters; for example, racially segregated schools could lead to an underestimation of ethnic (dis-)advantages.Recently, Mijs [56] suggested that stratification matters as to how pupils attribute failure and success at school.Distinguishing between external factors, such as luck or the quality of the teacher, and internal factors, in particular (a lack of) ability, Mijs (2016) showed that "students in mixed-ability groups tend to attribute their mathematics performance primarily to external factors, whereas segregated vocational-and academic-track students are more likely to blame themselves for not doing well".Moreover, "these differences between mixed-ability group students and tracked students are more pronounced in school systems where tracking is more extensive."Such arguments would suggest that the link between educational background and the perceived fairness would be stronger in countries where the educational system is more segregated, due to curriculum differentiation (early tracking) or school segregation (Anglo-Saxon world).

Data and Method
In examining how citizens evaluate the fairness of the social and educational system in different countries, we will make use of the 2009 wave of the International Social Survey Programme (ISSP).The ISSP is an annual survey, which occasionally covers opinions about social and educational inequality.We make use of two variables.
First, we rely on a number of questions about the perceived importance of different determinants of social upgrading.In particular, respondents had to rate the importance of different assets to get ahead in life (Table 2-A).For each asset, respondents could indicate the perceived importance as "essential", "very important", "fairly important", "not very important", or "not important at all".In this article, we focus on the proportion of respondents who indicated that the asset was "essential" or "very important to get ahead in life".We distinguish between two kinds of assets.First, we consider ascribed assets, i.e., assets for which somebody cannot, in any way, be held responsible: the wealth, educational level or network of one's parents, or one's race.Arguably, fair social systems would attempt to minimize the importance of these assets in determining life outcomes.Secondly, individual responsibility assets refer to characteristics that relate to the meritocratic ideal: working hard, and having ambition.We also classify "having a good education" as an individual responsibility asset, as educational achievement is partly dependent of individual effort and attitude.However, as we noted above, social origin affects educational achievement through a multitude of mechanisms beyond the individual responsibility of the student.For example, lower-educated parents may be less able to help with school assignments, resources needed to perform well at school (a private study room, ICT equipment) may be lacking in disadvantaged homes, or, in the case of school allocation by residence, students from poorer neighbourhoods may be assigned to schools of lower quality.As a consequence, educational achievement could be as much an "ascribed" asset as a matter of individual responsibility.Apart from interpreting results for this variable with additional caution, we also will zoom in on the perceived fairness of the school system by considering a second series of questions on educational fairness.In particular, respondents had to express their agreement to three statements regarding equal opportunities in the access to university (Table 2-B).Respondents could indicate whether they "strongly agree", "agree", "neither agree nor disagree", "disagree", or "strongly disagree" with these statements.We use the share "strongly agreeing" or "agreeing" to the first two statements, and the share not "strongly agreeing" or "agreeing" to the third statement.Hence, for all three variables, the higher the share, the less equal opportunities at school are perceived.
We first consider country-average judgements.Additionally, we also consider how individual judgments depend on the educational background of respondents.To this end, we model, for each country separately, individual judgements (whether the respondent agrees to a particular statement) as a logistic function of his years of schooling.We control for age and sex and restrict ourselves to respondents aged 25 years or above; we also do not take into account respondents reporting more than 25 years of schooling.
Table 3 reports the participating countries, their sample sizes, and the average years of schooling reported.

Country Averages
First, Table 4 presents, for each of the seven assets, the proportion in each country who agreed that the asset was "essential" or "very important" to get ahead in life.Overall, assets referring to individual responsibility are perceived to be more important than ascribed assets.This suggests that, fortunately, most Western citizens are convinced that their social status is distributed in their society in a more or less fair way.However, we will now zoom in on the differences between countries.To this end, the percentages from Table 4 were standardised and plotted into spider diagrams with axes ranging from −2 (the centre point of each diagram) to +2 (indicated by an outside polygon).In the hypothetical example in Figure 2, the inner polygon indicating zero represents the international average.These polygons thus can be interpreted as follows.The four ascribed assets (coming from a wealthy family, having well-educated parents, knowing the right people, and race) are presented on the right side of the figure, the three individual responsibility assets (having a good education, working hard, and showing ambition) at the left.The red polygon represents a hypothetical system in which the share reporting each ascribed asset as important for getting ahead in life is one standard deviation higher than the international average for this asset (Z = +1), while the share agreeing with the individual responsibility asset is one standard deviation lower than the international average (Z = −1).Hence, in an international comparison, this would reflect a relatively unfair social system.The green polygon represents the opposite situation (Z = −1 for the ascribed assets, Z = +1 for the individual responsibility assets), corresponding to a system fairer than average (at least, to the extent that we can consider educational achievement in terms of individual responsibility, see above).Note that the polygons thus represent relative positions, not absolute shares.Figure 3 summarizes all standardized polygons, categorized by type of educational system.Overall, the figure indicates a remarkable correspondence between educational system design and the perceived importance of certain assets to get ahead in life.In the Continental countries, the ascribed assets are perceived to be relatively important: compared to other countries, a rather large share of the respondents report that wealth, parental education, race and a social network are key to get ahead in life.This seems to mirror the relatively strong tie between social origin and educational achievement repeatedly observed in early tracking systems.By contrast, the individual responsibility assets of working hard and showing ambition are perceived to be slightly less important than average.However, educational achievement is deemed to be relatively important for social success in these countries, which reflects the relatively strongly stratified nature of these countries, in which occupational destinations are closely tied to educational background.By contrast, in the Nordic countries ascribed assets are perceived to be far less important to get ahead in life (with a remarkable exception for "race" in Norway).This also matches with the high degree of equal opportunities in their educational systems.In particular individual responsibilities, such as working hard and showing ambition, are perceived to be more important than average.Interestingly, the percentage reporting education to be important to get ahead is lower than average, which could relate to the egalitarian conditions on the labour market, resulting in small income differentials between high-and low-educated occupations.
In the Anglo-Saxon world, meritocratic principles such as working hard and showing ambition are also considered important.The main difference with the Nordics is that in these extremely competitive societies educational accomplishments are also key in being successful (cf.[57]).Moreover, in particular in the United States, with its high level of income inequality, the wealth of one's parents is recognized to be an important determinant of personal success.Figure 3 summarizes all standardized polygons, categorized by type of educational system.Overall, the figure indicates a remarkable correspondence between educational system design and the perceived importance of certain assets to get ahead in life.In the Continental countries, the ascribed assets are perceived to be relatively important: compared to other countries, a rather large share of the respondents report that wealth, parental education, race and a social network are key to get ahead in life.This seems to mirror the relatively strong tie between social origin and educational achievement repeatedly observed in early tracking systems.By contrast, the individual responsibility assets of working hard and showing ambition are perceived to be slightly less important than average.However, educational achievement is deemed to be relatively important for social success in these countries, which reflects the relatively strongly stratified nature of these countries, in which occupational destinations are closely tied to educational background.Figure 3 summarizes all standardized polygons, categorized by type of educational system.Overall, the figure indicates a remarkable correspondence between educational system design and the perceived importance of certain assets to get ahead in life.In the Continental countries, the ascribed assets are perceived to be relatively important: compared to other countries, a rather large share of the respondents report that wealth, parental education, race and a social network are key to get ahead in life.This seems to mirror the relatively strong tie between social origin and educational achievement repeatedly observed in early tracking systems.By contrast, the individual responsibility assets of working hard and showing ambition are perceived to be slightly less important than average.However, educational achievement is deemed to be relatively important for social success in these countries, which reflects the relatively strongly stratified nature of these countries, in which occupational destinations are closely tied to educational background.By contrast, in the Nordic countries ascribed assets are perceived to be far less important to get ahead in life (with a remarkable exception for "race" in Norway).This also matches with the high degree of equal opportunities in their educational systems.In particular individual responsibilities, such as working hard and showing ambition, are perceived to be more important than average.Interestingly, the percentage reporting education to be important to get ahead is lower than average, which could relate to the egalitarian conditions on the labour market, resulting in small income differentials between high-and low-educated occupations.
In the Anglo-Saxon world, meritocratic principles such as working hard and showing ambition are also considered important.The main difference with the Nordics is that in these extremely competitive societies educational accomplishments are also key in being successful (cf.[57]).Moreover, in particular in the United States, with its high level of income inequality, the wealth of one's parents is recognized to be an important determinant of personal success.By contrast, in the Nordic countries ascribed assets are perceived to be far less important to get ahead in life (with a remarkable exception for "race" in Norway).This also matches with the high degree of equal opportunities in their educational systems.In particular individual responsibilities, such as working hard and showing ambition, are perceived to be more important than average.Interestingly, the percentage reporting education to be important to get ahead is lower than average, which could relate to the egalitarian conditions on the labour market, resulting in small income differentials between high-and low-educated occupations.
In the Anglo-Saxon world, meritocratic principles such as working hard and showing ambition are also considered important.The main difference with the Nordics is that in these extremely competitive societies educational accomplishments are also key in being successful (cf.[57]).Moreover, in particular in the United States, with its high level of income inequality, the wealth of one's parents is recognized to be an important determinant of personal success.
Finally, the Mediterranean countries report the less attractive pattern combining a high reliance on ascribed assets with a low grip of individual responsibility on life outcomes.This includes a weak link between education and social success (see above), emphasizing the observation by Dronkers [48] that leaving social selection to the labour market instead of the educational system only serves to reduce, not increase, social mobility.

Relationship with Educational Background
In this section, we consider the educational gradients in the perceived fairness of the social system.Figure 4 illustrates the results graphically, with the standardized years of schooling on the horizontal axis, and the estimated shares finding each respective asset important on the vertical axis (we averaged out the estimates for all countries within a certain regime).For three out of four ascribed assets (family wealth, race, and networks), the gradients are negative, meaning that the higher educated are less willing to attribute success to these assets.In particular, the link with educational background is strongest in the Anglo-Saxon group (with the USA reporting the strongest gradients): the advantaged still seem to believe in the "American dream" of social mobility [58,59], while the disadvantaged appear to be far more critical about it [60].Similarly, in the Continental group, race and networks are perceived as less important to get ahead by the better educated; however, there is only a very small gradient in the perceived importance of coming from a wealthy family.In the Nordic and Mediterranean groups, the reported gradients are much weaker.For the fourth asset, parental education, the gradient is inversed: having educated parents is perceived as more important by respondents who are themselves higher educated.Similarly, the perceived importance of one's own education increases with educational level, except in the Mediterranean group.Finally, hard work and ambition are cited by about equally large shares across the educational spectrum.Finally, the Mediterranean countries report the less attractive pattern combining a high reliance on ascribed assets with a low grip of individual responsibility on life outcomes.This includes a weak link between education and social success (see above), emphasizing the observation by Dronkers [48] that leaving social selection to the labour market instead of the educational system only serves to reduce, not increase, social mobility.

Relationship with Educational Background
In this section, we consider the educational gradients in the perceived fairness of the social system.Figure 4 illustrates the results graphically, with the standardized years of schooling on the horizontal axis, and the estimated shares finding each respective asset important on the vertical axis (we averaged out the estimates for all countries within a certain regime).For three out of four ascribed assets (family wealth, race, and networks), the gradients are negative, meaning that the higher educated are less willing to attribute success to these assets.In particular, the link with educational background is strongest in the Anglo-Saxon group (with the USA reporting the strongest gradients): the advantaged still seem to believe in the "American dream" of social mobility [58,59], while the disadvantaged appear to be far more critical about it [60].Similarly, in the Continental group, race and networks are perceived as less important to get ahead by the better educated; however, there is only a very small gradient in the perceived importance of coming from a wealthy family.In the Nordic and Mediterranean groups, the reported gradients are much weaker.For the fourth asset, parental education, the gradient is inversed: having educated parents is perceived as more important by respondents who are themselves higher educated.Similarly, the perceived importance of one's own education increases with educational level, except in the Mediterranean group.Finally, hard work and ambition are cited by about equally large shares across the educational spectrum.

Country Averages
Table 5 again reports the cross-country shares "strongly agreeing" or "agreeing" that educational opportunities to enter university are not equitable in their country.

What Is
Needed to Enter University?

Country Averages
Table 5 again reports the cross-country shares "strongly agreeing" or "agreeing" that educational opportunities to enter university are not equitable in their country.Figure 5 summarizes all standardized polygons, categorized by educational system type.This figure indicates a remarkable correspondence between educational system design and the perceived inequity in university access.First, in particular in the Mediterranean and, to a somewhat smaller extent in the Continental countries, respondents perceive the access to higher education as relatively unfair (note in particular the high level of inequity observed in France).Tellingly, this corresponds well to the high social inequality in school performance (PISA) observed in educational systems relying extensively on grade retention and early tracking, as described by Dupriez, Dumay, and Vause [16].By contrast, in particular in the Nordic countries, access to university is perceived as being far less unfair.
Soc. Sci.2016, 5, 64 13 of 18 Figure 5 summarizes all standardized polygons, categorized by educational system type.This figure indicates a remarkable correspondence between educational system design and the perceived inequity in university access.First, in particular in the Mediterranean and, to a somewhat smaller extent in the Continental countries, respondents perceive the access to higher education as relatively unfair (note in particular the high level of inequity observed in France).Tellingly, this corresponds well to the high social inequality in school performance (PISA) observed in educational systems relying extensively on grade retention and early tracking, as described by Dupriez, Dumay, and Vause [16].By contrast, in particular in the Nordic countries, access to university is perceived as being far less unfair.

Relationship with Educational Background
Similar to the previous section, we considered how someone's own educational background might affect perceptions about educational inequities.To this end, we again modelled, for each country separately, the probability that a respondent thinks that access to university is unequal (for each of the three statements) as a function of his years of schooling, controlling for age and sex and removing respondents aged below 25 years.Figure 6 illustrates the results, with the standardized years of schooling on the horizontal axis, and the estimated shares perceiving access as unfair on the vertical.

Relationship with Educational Background
Similar to the previous section, we considered how someone's own educational background might affect perceptions about educational inequities.To this end, we again modelled, for each country separately, the probability that a respondent thinks that access to university is unequal (for each of the three statements) as a function of his years of schooling, controlling for age and sex and removing respondents aged below 25 years.Figure 6 illustrates the results, with the standardized years of schooling on the horizontal axis, and the estimated shares perceiving access as unfair on the vertical.
Overall, the observed educational gradients are very strong, in particular for the statements "only students from the best secondary schools have a good chance to obtain a university education" and "only the rich can afford the costs of attending university".Whereas just a small fraction of the high educated consider these statements to be true, there is much more agreement at the lower end of the educational spectrum.By contrast, agreement with the statement "people have the same chances to enter university, regardless of their gender, ethnicity or social background" is much more equally distributed among educational attainment levels.One possible reason is that this question was formulated more generally, with a less explicit reference to educational background (e.g., the reference to "gender" as a criterion for university access could be felt as more important among high-educated women than among low-educated women).Overall, the observed educational gradients are very strong, in particular for the statements "only students from the best secondary schools have a good chance to obtain a university education" and "only the rich can afford the costs of attending university".Whereas just a small fraction of the high educated consider these statements to be true, there is much more agreement at the lower end of the educational spectrum.By contrast, agreement with the statement "people have the same chances to enter university, regardless of their gender, ethnicity or social background" is much more equally distributed among educational attainment levels.One possible reason is that this question was formulated more generally, with a less explicit reference to educational background (e.g., the reference to "gender" as a criterion for university access could be felt as more important among high-educated women than among low-educated women).
Comparing the average gradients from different educational regimes, the Continental (early tracking) countries in particular report a strong effect of educational background on the perceived fairness in university access.While the higher educated in these countries are relatively optimistic about equality of educational opportunity, those with a lower educational attainment do not share this feeling; at the lower end of the spectrum, the perceived inequity in the Continental countries equals or even exceeds that in the Mediterranean countries.

Discussion and Conclusions
In this article, we have considered how citizens from various educational regimes perceive the fairness of society as a whole, and of the educational system in particular.A number of interesting country-differences have emerged.The Mediterranean countries reported a relatively unattractive picture.Relatively large shares of respondents perceived ascribed assets, such as wealth and parental education, to be key to get ahead in life; by contrast, one's own education was considered to be less important.Moreover, the educational systems of these countries, typified by Dupriez, Dumay, and Vause [16] as "uniform integration" systems with a massive reliance on grade retention to deal with heterogeneity, were considered to be less fair, with chances to enter university largely depending on family wealth and school quality.The overall picture of Mediterranean regime therefore is a "doublehurdle" regime, where ascription plays a substantial role within as well as after education.
In the Continental countries, education was considered much more important to be successful in life.However, this did not imply that ascribed assets, such as family wealth, parental education or Comparing the average gradients from different educational regimes, the Continental (early tracking) countries in particular report a strong effect of educational background on the perceived fairness in university access.While the higher educated in these countries are relatively optimistic about equality of educational opportunity, those with a lower educational attainment do not share this feeling; at the lower end of the spectrum, the perceived inequity in the Continental countries equals or even exceeds that in the Mediterranean countries.

Discussion and Conclusions
In this article, we have considered how citizens from various educational regimes perceive the fairness of society as a whole, and of the educational system in particular.A number of interesting country-differences have emerged.The Mediterranean countries reported a relatively unattractive picture.Relatively large shares of respondents perceived ascribed assets, such as wealth and parental education, to be key to get ahead in life; by contrast, one's own education was considered to be less important.Moreover, the educational systems of these countries, typified by Dupriez, Dumay, and Vause [16] as "uniform integration" systems with a massive reliance on grade retention to deal with heterogeneity, were considered to be less fair, with chances to enter university largely depending on family wealth and school quality.The overall picture of Mediterranean regime therefore is a "double-hurdle" regime, where ascription plays a substantial role within as well as after education.
In the Continental countries, education was considered much more important to be successful in life.However, this did not imply that ascribed assets, such as family wealth, parental education or race, were perceived to have ceased to determine life chances.Indeed, educational success was itself perceived to function in a relatively unfair way, particularly among those who did not attain a high educational level themselves.This seems to confirm the idea that education itself is the main "filter" producing social stratification in these countries.
In the Anglo-Saxon world, individual achievement-educational attainment, hard work, and showing ambition-was reported as the most important ingredient for getting ahead in life.However, this belief was not shared by all sections of the population.At the lower end of the educational spectrum, many respondents pointed to the strong dependence of life outcomes on ascribed assets such as family wealth, particularly in the USA.The implicit picture is one of dual societies with winners heralding a meritocratic education system, and losers challenging its fairness.
Fourthly, the Nordic countries emerge from the analysis as the champions of social mobility: working hard and showing ambition were perceived overwhelmingly as the determinants of life success, with family wealth or parental education cited by very few respondents.Moreover, access to university was believed to be largely independent of social background.Maybe most importantly, these beliefs were shared both by the low-and the high-educated, with just a tiny effect of educational background on the level of support.The fact that less successful citizens confirm the fairness of the system reflects a broad consensus among the population about the equity of their social and educational regime.
The observed patterns in subjective, perceived social inequalities largely confirm existing knowledge on the objective, statistically collected inequalities in life chances across different countries.On top of that, however, they are relevant in their own as they show that citizens are aware of these inequalities and the legitimacy of their social system.Recent strands of neo-institutionalist theory [61,62] have argued that institutional design is rather about legitimacy than about efficacy.For the educational system in particular, Tyack and Tobin [63] claimed that educational practices as well "result from a conformity of organizational forms with general public beliefs".Similarly, LeTendre, Hofer, and Shimizu [64] claim that "stratification is legitimated by widely held beliefs about how education should operate.Nation-specific values and attitudes determine which forms of curricular differentiation are legitimated and which contested."This paper also shows that these beliefs are not always shared by all strata of the population.
A number of potentially interesting issues had to be left for future research.First, we mainly were interested in the general relationship between some broad types of educational regimes and the perceived social fairness.Indeed, whereas countries classified in the same cluster often reported reasonably similar patterns of appreciation, the within-cluster variation remaining unexplained should not be ignored.For example, in the reports about what matters to get ahead in life, coming from a wealthy background is reported three times as frequently by US respondents than by respondents from New Zealand, which was classified into the same Anglo-Saxon cluster.Hence, refining the current international comparative analysis by considering more closely how individual countries behave could further add to our understanding of the issue.
Secondly, we treated educational achievement as an "individual responsibility" asset.However, much can be said to qualify this view, as educational achievement is probably influenced by ascribed assets as well.While we also studied the perceived fairness of the educational system in itself, the questions available in the ISSP were not optimal (for example, they solely refer to access to tertiary education).Hence, future research should further elaborate on how educational achievement acts on the cross-roads of social reproduction and individual merit.
Finally, this paper relied on data collected in 2009.As countries may have switched to alternative educational and social policies in the meantime (see, e.g., the increasing emphasis on private schooling in Sweden, which may have effects at odds with its egalitarian tradition), it remains to be seen to what extent the observed patterns accurately reflect the present situation in each country.A new ISSP module on perceptions about social inequality has been announced for 2019.

Figure 3 .
Figure 3. Standardized polygons, representing the shares agreeing that a certain asset is important to get ahead in life.

Figure 3 .
Figure 3. Standardized polygons, representing the shares agreeing that a certain asset is important to get ahead in life.

Figure 3 .
Figure 3. Standardized polygons, representing the shares agreeing that a certain asset is important to get ahead in life.

Figure 4 .
Figure 4.The importance of various assets to get ahead in life, according to educational level (standardized years of schooling).

Figure 4 .
Figure 4.The importance of various assets to get ahead in life, according to educational level (standardized years of schooling).

Figure 5 .
Figure 5. Standardized polygons, representing the shares agreeing that access to university is not fair.

Figure 5 .
Figure 5. Standardized polygons, representing the shares agreeing that access to university is not fair.

Figure 6 .
Figure 6.Perceived inequity in university access, according to educational level (standardized years of schooling).

Figure 6 .
Figure 6.Perceived inequity in university access, according to educational level (standardized years of schooling).

Table 1 .
Typology of educational regimes.

Table 2 .
Importance of various assets to get ahead in life, as surveyed in ISSP 2009. A-

How Much Do You Agree or Disagree with Each of the Following Statements?
Equal opportunities in the access to universityOnly students from the best secondary schools have a good chance to obtain a university education Only the rich can afford the costs of attending university People have the same chances to enter university, regardless of their gender, ethnicity or social background

Table 3 .
Sample sizes of selected countries from ISSP 2009.

Table 4 .
Share in each country agreeing that a certain asset was "essential" or "very important" to get ahead in life.

Table 5 .
Share in each country agreeing that educational opportunities to enter university are not equitable.

Table 5 .
Share in each country agreeing that educational opportunities to enter university are not equitable.