Introduction

During the pandemic-related lockdowns introduced by governments worldwide as preventative measures to curb COVID-19 infection among their citizens, schools interrupted their normal operations and families were asked to continue their children’s schooling at home. In France, the first two-month lockdown, which lasted from 16 March to 12 May 2020, shifted the responsibility for school education onto families’ shoulders. This unprecedented situation, which amounted to a real life-size experiment, offered the opportunity to investigate and better understand the effect of parental involvement in children’s schoolwork on inequalities in school achievement.

This inadvertent “experiment” enabled researchers to look at the practices of parents in terms of supervision/support, which had for a long time remained unthought of in the sociology of education – at least in France. In the Bourdieusian tradition, the socialisation of children in the family environment operates by “osmosis” (Bourdieu 1966). This means, in particular, that the transmission of cultural capital takes place through slow and gradual internalisation, “in the absence of any methodical effort and manifest action [taken by parents towards their child]” (Bourdieu 1966, p. 330). This idea suggests that the differences in educational achievement between pupils from different social backgrounds can only be explained by slow, diffuse, and therefore almost impenetrable mechanisms of permeation. Ironically, however, Bourdieu’s analysis in terms of socialisation by osmosis may not fully realise its programme of deconstructing the “ideology of the gift” (Bourdieu and Passeron 1970), because osmotic socialisation retains a certain degree of mystery, of self-evidence, or at least of automaticity in the transmission of cultural capital and educational success.

Yet school education has become a major challenge for parenting (Vincent 2017) in the 21st century. While families let their cultural capital “float”, let it act “by osmosis”, in the 1960s, today they are massively and explicitly involved in the daily schooling of their children. Indeed, research has found daily parental support of their children’s schoolwork to be a determining factor in the making of current educational inequalities.

In France, quantitative surveys show an increase in parental involvement in children’s schooling. Marie Gouyon and Sophie Guérin highlight this clear increase between the early 1990s and the early 2000s (Gouyon and Guérin 2006). This evolution is general and concerns all social backgrounds. More specific work on working-class families in France emphasises the overall rise in educational aspirations (Poullaouec 2010) or the involvement of parents in the practice of homework (Lahire 1995; Kakpo 2012). These trends are not specific to France, Giulia Dotti-Sani and Judith Treas (2016) found a general increase in all developed countries in parental involvement in children’s schooling.

While there is no doubt that pandemic-related lockdowns have contributed to widen-ing the gaps,Footnote 1 the precise nature of the inequalities linked to schooling at home remains unclear. How does parental support, differentiated according to social background, contribute to the creation of gaps in educational achievement? What does the pandemic-related lockdown period teach us about this?

A survey on parenting practices

Seizing the opportunity to investigate and better understand the effect of parental involvement in children’s schoolwork on inequalities in school achievement, Filippo Pirone (University of Bordeaux) and myself launched a survey among parents of primary and secondary schoolFootnote 2 pupils soon after the first lockdown had been implemented by the French government. We obtained ethical clearance from the Ministère de l’éducation nationale (the French Ministry of Education), and written informed consent from all participating parents.

At the beginning of April 2020, we distributed an online questionnaire to parents in order to identify their practices of support for their children’s schoolwork. It was distributed by e-mail, with the help of and through French schools. The schools were contacted beforehand using the directory of contact details for all public (state-run) schools in France, which is available on the website of the French Ministry of Education. The centralisation of the French education system made the operation easy, and the fact that we managed to collect 31,764 responses throughout the month of April 2020 reflects parents’ interest in the subject. This figure refers to individual responses from parents: when parents had more than one child aged 5–13, we asked parents to respond for one particular child only.

The questionnaire contained 120 questions. The first part of the questionnaire (45 questions) asked for information on parents’ socio-demographic profile, their professional conditions during lockdown (e.g. home office or required presence at the workplace), their general approach to their child’s schooling (e.g. general parental involvement). The second part (75 questions) asked them about their practices of supervising the child in doing schoolwork at home during lockdown, specifying details such as material conditions (e.g. digital equipment, places dedicated to schoolwork in the house, etc.) and organisation of work, transmission techniques, resources and pedagogical choices, difficulties encountered, and their relationship with the child.

Some major results, reported on here before completion and refinement of our research findings, already emerged early on. In this research note, we present the results which concern only the parents of pupils from CP (“cours préparatoire”, child aged 5–6, learning to read), up to “5ème” (child aged 12–13, year/grade 2 in lower secondary school). This subsample included a total of n = 19,454. Our reason for starting with CP was that the forms of domestic schoolwork involved for grades preceding CP and learning to read differ too much from what is required at primary and early secondary level; moreover, after 5ème, we noted a quantitative drop in parents’ supervision, as shown in Fig 1.

Fig. 1
figure 1

Share of parents reporting “supervising their child’s work”, by grade.

Source: « L’école à la maison » survey (Université de Bordeaux – CED, May 2020)

Full size image

Our survey was supplemented by interviews (n = 18) with a small sample of parents of pupils from middle- and upper-class backgrounds,Footnote 3 in order to understand the concrete techniques practised by them and to observe what constituted the “educational advantage” of these families’ homes as learning environments.

Several hypotheses are commonly put forward to explain the inequalities between children from different social backgrounds: the digital divide, housing conditions, lack of availability of parents, teleworking, distance from school culture. While one of our aims was to try to document the material dimensions of the inequalities in parental support of their children’s schoolwork, it emerged that the pedagogical dimensions of these inequalities (in particular the ability of parents to transmit the kind of knowledge required for school work, and the ability of children to appropriate such “school knowledge”) seem to be the most decisive.

The rest of this research note is dedicated mainly to measuring the effect of social background on parents’ support practices. In order to do this, I constructed a “social background” variable, based on the “type of occupation” variables filled in by the parents in the first part of the questionnaire. I constructed a normative correspondence table, inspired by the work of Etienne Penissat et al. (2018) on the French nomenclature of professions and socio-professional categories (la nomenclature des professions et catégories socioprofessionnelles, abbreviated PCS) and its uses.Footnote 4

For single-parent families, I chose the profession of the responding parent. For two-parent and step-parent families, we combined the occupations of both parents, again according to a normative correspondence table. We assumed an opposition between working-class backgrounds on the one hand and middle-/upper-class backgrounds on the other, acknowledging that this choice could be criticised for neglecting the internal diversity of each category. Within the lower classes (working class and lower middle class), blue-collar and white-collar workers show different practices of educational support, just as among the upper classes (upper middle and upper class), the forms of parental support are varied (van Zanten 2009). Nevertheless, to keep it simple for this research note and in order to reveal more clear-cut contrasts in practices, we maintain a principle of class homogeneity. Readers interested in a more differentiated approach are invited to read some of the other publications resulting from this survey which question the heterogeneity of parenting practices in middle- and upper class families (e.g. Delès 2021).

A digital divide?

The way in which the questionnaire was administered (online) effectively excluded parents without computer equipment or internet access from our survey. Therefore, this approach was not really a good fit for investigating digital inequalities in relation to school at home. However, it does provide some lessons. First, we found inequalities in terms of equipment: 11.4% of working-class families said they had a connection that posed a problem, compared with 7.9% of families from higher social categories (middle and upper class) who said their internet connection was reliable. Responses to the question “Do you consider your computer equipment and internet access sufficient to carry out the work of the school at home?” revealed the same gap (24.3% “no” versus 17% “yes”). However, this gap is even more visible in parents’ perception of their own computer competence: 45% of middle- and upper-class parents felt “quite capable” of meeting the digital technical requirements of “homeschooling”,Footnote 5 compared with only 31% of working-class parents. These results already seem to suggest that the material dimension of digital inequalities is less significant than the dispositional dimension. The digital divide cannot therefore be reduced to inequalities in terms of equipment that are so often highlighted; what needs to be investigated is the different levels of mastery of digital tools used in the context of educational continuity, depending on the social environment.

Besides technical equipment and competence available in family homes, what was central during lockdown was the use of digital platforms. Therefore, as part of our survey, we explored a corpus of 15 teaching sequences uploaded to a central French platform by teachers of French, and History-GeographyFootnote 6 in secondary schools in France. By proposing “composite” online content, i.e. content that was heterogeneous from a semiotic point of view (texts, videos, images, graphics, podcasts, etc.), pupils were implicitly required to be able to navigate between these different media and to reconstruct their intellectual coherence. Our analysis of these online sequences found that pupils’ competence in in carrying out this task varied greatly, and that parental support was crucial here. This demonstrates that the digital form of school learning often conceals strictly pedagogical issues.

The myth of parental resignation

The sociology of education has long since disproved the conventional wisdom that parents from working-class families are less involved in monitoring their children’s schooling (Lahire 1995; Kakpo 2012; Garcia 2019) than their middle- and upper-class peers. In our survey, all parents said they had the time (around 90% of responding parents) and recognised the importance of following their children’s schoolwork (around 95% of parents) during the period of lockdown, with no strong differences between social backgrounds. Interestingly, we found that the time devoted to homeschooling defied our expectations: lower-class parents spent an average of 3 hours and 16 minutes a day on school support, compared to 3 hours and 13 minutes spent by middle-class parents, and 3 hours and 7 minutes spent by upper-class parents, while teachers spent only 2 hours and 58 minutes on supporting their pupils (Chauvel et al. 2021). We therefore found no “dropping out” of school in working-class families; in other words, they did not abandon schoolwork during lockdown.

There were, however, fairly clear differences, not so much in the amount of time spent on schoolwork during lockdown, but in the precise organisation of this time (see Table 1).

Table 1 Structuration of time spent on homeschooling
Full size table

It is noticeable that working-class families made less use of techniques for structuring school time. Time was less often broken down into stable and formalised sequences. Conversely, we found that middle- and upper class families maintained a dedicated effort to organise time throughout the lockdown period. As for teachers’ structuring of time in their online support of their pupils, most classes followed invariable timetables, small school rituals were set up (roll-calling, drawing up the “weather of the day”, clapping hands to signal the end of “recess”) which aimed to mark the time. The organisation of school time was also designed from a pedagogical point of view: “fundamental learning” (French, mathematics) was placed in the morning, when children have the best capacity for concentration.

In terms of parental pedagogical action, we observed that working-class parents declared more than middle- and upper-class ones that they used direct support techniques, such as monitoring that the child was following instructions (88.0% as against 84.6% in the middle and upper classes); having the child recite the lesson (87.7% as against 84.1%); and doing exercises directly related to the lesson (88,8% as against 86.2%). These activities, which favour the fulfilment of school orders in their most formal aspects, clearly demonstrate working-class families’ support, and maybe even their trust, in the school institution.

Inequalities in the nature of the work undertaken

The above findings suggest that educational inequalities are probably less due to the amount of time spent on homeschooling and more due to the nature of the work involved. On the other hand, in the middle and upper classes, more indirect support strategies can be observed: 49.9% of middle- and upper-class parents declared that they set their children exercises indirectly related to the lesson, compared to 48% parents in working-class environments. Moreover, 27% of middle- and upper-class parents set their children complex exercises that required skills acquired in other subjects, compared with 22.5% of working-class parents. Similarly, we found that middle- and upper-class parents were more likely to use alternative learning materials to those proposed by teachers than working-class parents (25.5% compared with 22.7% respectively). Middle- and upper-class parents – and teachers – therefore did not limit themselves to a kind of mechanical fulfilment of the school tasks. On the contrary, their classroom practice at home shows that they questioned the “technology” of the “school knowledge”, the way in which school knowledge is produced. The more frequent use of indirect or complex exercises in middle- and upper-class families shows that they seemed to penetrate the epistemology of school knowledge and decode its implicit meaning. Figure 2 clearly shows that working-class parents were more active in following instructions for activities already suggested by teachers, whereas middle- and upper-class parents were more active in creating their own activities.

Fig. 2
figure 2

Practised activities: difference in reporting rate between lower (working) and upper (middle and upper) classes. (Source: « L’école à la maison » survey (Université de Bordeaux – CED), May 2020)

Note: Results rely on responses from parents of children attending CP to 5ème (n = 19,454). Each value was calculated as follows: (average reporting rate of working-class parents) – (average reporting rate of middle- and upper-class parents). In locked-down homeschooling, working-class parents sat with their child more often than middle- and upper-class parents (+ 8,9%). All differences are statistically significant (p < 0.001).

Full size image

Here we find the results of the theories of the “relationship to knowledge” (rapport au savoir), developed by French educational sociologists (see, among others, Charlot et al. 1992; Bonnéry 2007; Bautier and Rayou 2013). Their work has sought in particular to analyse the cognitive operations pupils are asked to perform during school learning. To put it schematically, these theories reveal that school learning implicitly assumes that the pupil is capable of conceptualising, i.e. of accessing deep notions hidden underneath contingent school tasks. For instance, in geography, in order to convey the notion of relief to secondary school students, they are asked to colour-code a map (green for low altitude, orange for medium altitude, etc.). In this exercise, it is thus implicitly assumed that students will be capable of generalising this colour code to the notion of relief (the example is taken from Bonnéry 2007). Without always formulating it explicitly, school learning therefore calls upon the abstraction, conceptualisation and “decontextualisation” capacities of the students (and of the parents who help them in their homework). The transmission of school knowledge is therefore based on a method of decoding its implicit meaning that is expected from all pupils, but not in fact equitably distributed between social backgrounds.

Working-class families adopt more direct, more focused and more imitative support practices, which fulfil the formal expectations of school instructions (staying physically next to the child, making him or her recite the lesson, etc. …). Conversely, “expert” parental monitoring (Kakpo and Rayou 2018) involves an understanding of the invisible expectations built into school knowledge and implements support techniques that are more circuitous, but no less effective. This corroborates US-based literature on parental involvement, which suggests that lower-income parents are more focused on traditional (rule-based) forms of learning (Lareau 2003).

The question of effective appropriation of curricular content must therefore be raised, all the more acutely so since homeschooling, by necessity, is deprived of professional face-to-face teaching and direct interactions with teachers. These initial findings led us to extend our quantitative survey with interviews (n = 18, by telephone or in person) with parents of pupils from middle- and upper-class backgrounds and ethnographic observations for a detailed understanding of these mechanisms.

The relationship aspect

In terms of the relational question, there seems to be little difference between social backgrounds in terms of parent–child relationships before the lockdown (our respondents declared about 80% of “good” or “rather good” relationships, irrespective of social background). But the lockdown and monitoring of homeschooling seems to have contributed to the development of relational tensions. We found these tensions to be more frequent in working-class environments: 32.7% of working-class families reported sometimes or often encountering difficulties in their parental relationship with the child during lockdown, compared with 22.2% middle- and upper-class families.

The relational dimension should not be neglected in the effectiveness of learning. A peaceful relationship is obviously more favourable to learning. However, establishing a serene learning climate through peaceful relationships is not a trivial exercise. Parents are expected to be patient: they need to rediscover a form of relational distance, of emotional neutrality towards their children which naturally takes place at school between the teachers and pupils. Indeed, at school, the learning situation is based on a “contract” of institutionalised relationship between the pupil and the teacher, where each respects a social role. For homeschooling to be effective, it is this relational contract that needs to be recreated at home. Parents’ responses to our survey questions concerning the state of their relationships with their children both before and during lockdown demonstrate that not all parents achieve this in the same way.

Conclusion

As we can see, the challenges of homeschooling are multiple. In order to school children effectively at home, it is necessary to benefit from minimum material conditions (in terms of digital equipment, in particular), but it seems that material equipment is not enough; social gaps are also widening in terms of pedagogical skills. In a homeschooling situation where the teacher is, in principle, at a distance, families are left to their own abilities for understanding school instructions. By its very nature, no distance education system can efficiently overcome this type of inequality. The school still needs its teachers ...

Every cloud has a silver lining: “dolorist” philosophies, as Ruwen Ogien (2017) called them, were widely spread during the crisis and particularly in the context of homeschooling. Pupils were said to gain in autonomy, parents in bonding with their children, teachers in professional experience and in mastering distance learning tools, and administration in flexibility. In terms of “lessons learnt” from the inadvertent large-scale “experiment” of pandemic-related homeschooling, one might be tempted to permanently adopt the “best practices” which emerged from the management of the crisis. But even if this experience of pedagogical continuity has demonstrated teachers’ remarkable capacity to innovate and adapt, it is nevertheless necessary to question its effects. We should also remain modest in our efforts to find solutions for the future. What does distant transmission of teaching content actually produce? How do students and their families really appropriate school knowledge? These are just a few of the questions that we will still have to deal with in the months and years to come.

The initial results, presented in this research note, of our survey on parental practices of educational support during the spring 2020 lockdown in France clearly demonstrate the existence of growing inequalities among school children’s chances of educational achievement. The first of the pandemic-related lockdowns already acted as a catalyst for educational inequalities in France, and very likely in many other countries as well. In particular, the differences linked to the appropriation of knowledge became accentuated in a distance learning context. Our initial findings thus strongly suggest that certain inequalities in children’s educational achievement, widely conceived as being the result of a digital divide, are in fact linked to differences in educational support, depending on social environments. In other words, behind the digital divide, there are actually differences in the transmission and appropriation of school content. More precise work on the types of practices expected of parents, combined with more disaggregated categories of social affiliation, should make it possible to further improve our understanding of the mechanisms of educational inequality.