To assess the evolution of personal relationships in France over the last twenty years, we repeated a study in 2017 already carried out in 2001. Comparison of the results reveals high stability in the composition of “personal networks” (but also allows us to identify certain changes in life cycles and relationships between urban and rural environments).

Et si les réseaux sociaux nous isolaient? (What if social networks isolate us?) was the title given by Le Point in 2017. For Reuters, in 2009, les nouvelles technologies favoriseraient le lien social (new technologies will promote social bonding) while, for Inrockuptibles, in 2015, the smartphone would, quite simply, “ruin it”. Looking at the press, it’s difficult to get an idea of the effects that the Internet, mobile telephony, social media and their democratisation over recent decades have had on our relationships.

The evolution of social life structures is a long-standing debate that was already a concern of the very founders of sociology. This debate was revived in the early 2000s when political scientist Robert Putnam expressed his alarm at the decline in civic life and a disengagement from clubs and societies in the United States (and especially from bowling leagues).

In the same period, the development of electronic communication seemed to make it possible to renew forms of sociability and political investment, this said, some saw it as a device that could destroy family ties and a vector for increased individualisation, which seemed to be corroborated by a growing trend in social isolation among Americans.

However, after several years of debates and methodological reviews, detailed analyses of the data available on the evolution of social relations in the United States point to the idea that, in recent decades, these relationships and the networks that people build around them (what sociologists call their “personal networks”) have actually changed little over time. But what about in France?

---

---

A survey of relationships and networks in France

In 2001, we carried out a survey of the personal networks among a sample of 399 people representative of the Toulouse metropolitan area. We then replicated this survey 16 years later with a new sample of 709 people in order to measure variations in the structure of social life and evaluate changes between the two periods.

These surveys consisted in asking respondents to list the people with whom they maintain certain types of relationships: those they discuss their personal problems with, their hobbies, those that they have had a friendly moment with during the last three months and also those that they could ask for financial assistance in an emergency. We then asked them to describe each of these people using several questions. These new data make it possible to detail how the structure of social life has evolved over the long term in mainland France.

Evaluating whether the democratisation of the Internet, mobile telephony and social media has had an effect on our personal networks is a complicated task as it requires surveys to measure these personal networks today and in the past, data that are very rarely found and, when they do exist, were mainly gathered in North America.

Furthermore, technologies are far from the only factor that can influence personal networks. Large historical, economic or demographic changes can also strongly influence these networks, but to find examples of this one must look beyond regions that have been relatively politically and economically stable for several decades.

This is the case, for example, with countries that have transitioned from centrally planned economies to market economies. In Hungary, throughout the 1990s, a transition towards less repressive regimes led to an increase in the importance of relationships outside the home, particularly with friends.

In China and post-Soviet Russia, in response to the uncertainties surrounding the economic transition, personal networks also seemed to take on great importance throughout the 1990s and 2000s for accessing employment. Extreme events such as the successive lockdowns brought about by the 2020 health crisis and the severe restrictions they induced can also affect personal networks.

During the lockdown of spring 2020, we know that networks refocused on loved ones and homophilic relationships, that neighbourhood relationships were automatically strengthened for certain people and that the youngest members of the community saw their relationships particularly affected.

 

Rather marginal and predictable developments

Our results show that, overall, personal networks have changed little over the past twenty years. Despite the democratisation of communication technologies, the structure of social life has proven surprising stable.

The composition of the networks, that is, the type of people mentioned in the “name generators”, seems almost identical in the two periods for both local services and emergency aid. When we need to ask for assistance, we turn to friends and family in emergencies and to our neighbours for more practical help. Our colleagues become slightly more important when it comes to recreational sociability and sharing personal information, although these roles are still overwhelmingly filled by family members and close friends.

This can be explained by the increase in employment rate among women, but also probably the increase in professional mobility which means that individuals often change jobs several times during their lifetimes, each job providing a new context for building relationships with colleagues.

The size of the networks, measured by the number of names cited in response to the name generators, seems to indicate that this number is significantly greater in 2017 than in 2001. The orders of magnitude remain similar but the increases are significant. The networks are therefore perhaps larger, but due to investigative effects in the collection procedures used, we prefer to limit ourselves to the conclusion that they have not become smaller in 16 years.

The networks are, on the whole, no more geographically dispersed in 2017 than they were in 2001 as a greater number of long-distance relationships are not recorded, with the exception of family members with whom contact is maintained and who live more than an hour away from the respondents.

This could be explained by the slight increase in residential mobility over the last decades, meaning that it’s more common for respondents to live a distance from their parents or children, or by the democratisation of communication technologies which facilitates remote contact with parents, children, cousins, brothers and sisters.

The contexts in which we form our relationships are also very similar in the two periods.

No difference is observed for 30-60 year olds who continue to form their relationships at work, through friends, through their children or spouse or in the neighbourhood. However, it’s less common for the younger age group to meet the people with whom they form relationships through their children or through work, due to the increasing length of studies, a delay in entering professional life and their age when they have their first child.

Those over 60, however, are more likely to meet people with whom they form relationships at work and within associations and less often through their children, due to the shift in retirement age and better overall health. It’s as if we’re seeing a lag in social life of around two years. In other words, we could say that, socially, we’re two years younger in 2017 than in 2001, as we’re more likely to be studying and child-less and, for older members of society, more active (professional or associative).

 

Less dense and more homogeneous networks

However, we can also highlight some more significant changes. There is a decline in network density in rural areas. Density indicates the proportion of our relationships (friends, family, colleagues, etc.) who would be able to contact each other without going through us first. Therefore, the closer this indicator is to 100%, the more the people we know know each other and, in contrast, the closer it is to 0%, none of the people we know are able to contact each other in our network.

In other words, in rural areas, networks are more disconnected or less cohesive in 2017 than in 2001. This decline, which is linked to a disconnection from social circles and especially from work (colleagues are less likely to know our family or friends) could be explained by the transformation of life trajectories in rural areas.

More often than in 2001, people who live in rural areas have had the opportunity to live in a city during their studies, for example, or work in another city, which gives these people a similar lifestyle to the one we see in peri-urban areas.

Another notable change is the increase in the homogeneity of personal networks between 2001 and 2017. The social homogeneity of relationships, the fact that the people interviewed and those they talk about in the surveys share common characteristics, is an indicator of greater or lesser social diversity in those around us.

The networks are particularly more homogeneous for the most qualified: 48.1% of relationships of people with an academic level higher than a bachelor’s degree were with people who had an equivalent level of education in 2001, compared to 61.8% in 2017. This trend can partly be explained by demographics and by the fact that there are more higher education graduates in 2017 than in 2001 in France and in the Toulouse metropolitan area.

But we can also see that the homogeneity of networks seems to increase in terms of age, family situation (single people and those without children are more in contact with people in the same situation as them) and also in terms of academic level, beyond what we would expect given the social composition of French society.

The continuation of already existing trends

Overall, our relationships and networks remain rather stable and we always seem to be able to count on them to obtain services, help or just to have a pleasant time. From this point of view, the social bond does not seem to have deteriorated.

It should be noted that these results were expected, on the one hand because they are similar to those found in the United States and, on the other, because if the democratisation of communication technologies does have an effect, it would above all be an acceleration of trends already at work or it would affect relationships outside the radar of the methods we’ve used here, in particular weak ties (acquaintances or more distant friends). The effect of communication technologies remains relatively marginal compared to demographic and economic developments.

Nevertheless, these trends towards less dense and more homogeneous networks – differences that we usually see when we compare rural environments with urban environments – would therefore support what Georg Simmel, a German sociologist, argued at the beginning of the 20th century, already pointing to a trend involving the segmentation of social circles in modern societies. In other words, our social life is chosen more according to our personal interests but is also more segmented and therefore more segregated.

These developments are neither sudden nor drastic but, observable over a period of 16 years with no major changes in the economic or political system and converging with those observed in the United States, they demonstrate, in our opinion, underlying trends in Western societies.

 

---

For other articles in english

 

---

---