11. Social Connections

Abstract

Social Connections address both the quantity and quality of time spent with others, and how much support people feel they have. Despite differences in the amount of time spent socialising, people’s own evaluations of their social connections are mostly positive and fairly similar across OECD countries. On average, people are highly satisfied with their social relationships (8.1 on a 0-10 scale), and 90% feel that they have someone they can count on in times of need. Even though men spend, on average, 40 minutes less than women in social interactions per week, gender differences in satisfaction with social relationships are negligible. Older people spend less time in social interactions and have less social support, but their satisfaction with social relationships is not significantly lower than for younger people. People with lower educational attainment are more likely than their more educated peers to lack social support.

Figure 11.1. Social Connections snapshot: current levels, and direction of change since 2010

Social support

Around 9 out of 10 individuals in OECD countries report having relatives or friends who can help them in times of need, ranging from 78% in Greece, to 98% in Iceland (Figure 11.2). The OECD average level in 2016-18 is almost unchanged from 2010-12. However, the share of the population who feel supported fell in Greece (by nearly 6 percentage points), Poland (-5) and Germany (-4), while over the same time period it rose by more than 4 percentage points in Italy and Estonia, and by 5 points or more in Portugal, Mexico, Latvia, Lithuania and Turkey.

Figure 11.2. 90% of people in OECD countries, on average, have someone they can count on

Time spent in social interactions

Time spent in social interactions considers the number of hours per week spent interacting with family and friends as a primary activity (i.e. it excludes interactions that occur alongside other focal activities such as paid work, caring or studying). Across the OECD, people aged 15 or more spend, on average, 6 hours per week interacting with family and friends (Figure 11.3). This ranges from 2 hours per week in Japan, and around 4 hours in Luxembourg, Hungary and Estonia, to above 7 hours in Italy, New Zealand Turkey and the Netherlands, and more than 9 hours in Austria. Changes in time use since 2005 can be assessed for just seven OECD countries: Belgium, Canada, Italy, Japan, Korea, Turkey and the United States. Over time, average weekly time spent in social interactions has fallen by around half an hour in Canada, Italy and the United States, and by little more than 40 minutes in Belgium.

Figure 11.3. Time spent socialising in OECD countries ranges from 2 to 9+ hours per week

Satisfaction with personal relationships

Satisfaction with personal relationships provides a measure of the perceived quality of social connections. Across the OECD countries with available data, people are generally satisfied with the quality of their personal relations, reporting an average rating (on a 0-10 scale) of 8.1. Cross-country variation spans a fairly limited range, with national averages ranging from just above 7 in Greece to 8.6 in Switzerland, Ireland, Mexico, Austria and Slovenia (Figure 11.4).

Since 2013, average satisfaction with relationships has increased slightly, but this masks diverging patterns across countries – for example, gains of 0.3 scale points or more in Spain, Mexico, Portugal, the Slovak Republic and Estonia, and losses of 0.3 scale points in Latvia, the Netherlands and Denmark.

Despite the relatively high average levels of satisfaction with personal relationships in OECD countries, around 10% of people rate their satisfaction at 5 or below (on a 0-10 scale). This proportion ranges from around 5% in Finland, Switzerland, the Netherlands and Ireland, to above 15% in Hungary, Lithuania and Turkey, and almost 30% in Greece (Figure 11.5).

Figure 11.5. 10% of people in OECD countries, on average, report a low satisfaction with their relationships

Social Connections inequalities: gaps between population groups

Women spend more time in social interactions than men, but there are no gender differences in support or satisfaction

There are no substantial gender differences in social support, or in satisfaction with personal relationships. However, large gender inequalities emerge in time spent in social interactions (Figure 11.6). In the average OECD country, women spend 40 minutes more than men per week in social interactions (6 hours and 20 minutes vs. 5 hours and 40 minutes for men, respectively). The gap in favour of women is especially large in Norway (around 2 hours 20 minutes per week), Australia, Ireland, New Zealand and the United Kingdom (all above 1 hour). Conversely, men spend more time socialising than women in Italy (8 hours 20 minutes per week for men vs. 6 hours 40 minutes for women), and to a smaller extent in Spain and Greece.

Figure 11.6. In the majority of OECD countries, women spend more time in social interactions than men do

Older people feel less supported and often spend less time in social interactions than younger age groups

In most OECD countries, perceived social support declines with age. In Korea, Greece, Chile, Latvia and Portugal, the age gradient in social support is particularly steep (Figure 11.7). For instance, 93% of people aged 15-29 in Korea report having relatives or friends they can count on in times of need, compared to only 63% of those aged 50 or over. By contrast, in France, Iceland, Ireland, New Zealand and the United Kingdom, gaps in social support across age groups are small.

Figure 11.7. Older people have less social support

In the 14 countries with available data, young people (aged 15-29) spend, on average, nearly 2 hours 20 minutes per week more in social interactions than middle-aged people (30-49) (Figure 11.8). The gap is small in Norway and Turkey, but widens in Italy, Ireland and Spain, where young people spend between 3 hours 50 minutes and 5 hours 20 minutes more in social interactions than the middle-aged. On average, in the countries with available data, middle-aged (30-49 years) and older people (aged 50+) tend to spend similar amount of time socialising, although divergent cross-country patterns exist. For example, in Finland, Italy and Norway people aged 30-49 allocate more time to social interactions than those aged 50 and over. By contrast, in Ireland older people spend nearly 1 hour and 40 minutes more per week socialising than those aged 30-49, with this difference being as large as 2 hours 20 minutes in Turkey.

Despite large age gaps in both social support and time spent in social interactions, age differences in satisfaction with social relationships are comparatively small. For the average OECD country, satisfaction with social relationships is 8.3 for people aged 16-29 (ranging from 7.4 in Greece to 8.9 in Slovenia); 8 for the age group 30-49 (ranging from 7.1 in Greece to 8.5 in Austria and Slovenia); and 8 for people aged 50 or above (ranging from 7 in Greece to 8.8 in Sweden).

Figure 11.8. Younger people spend more time in social interactions

People with higher education report better access to social support

For the average OECD country, the proportion of people with only a primary education reporting they have someone to count on in times of need is 9 percentage points lower than for those with a tertiary education (Figure 11.9). In Switzerland, New Zealand and Iceland, the gap is below 2 percentage points, but it exceeds 15 percentage points in Korea, Greece, Turkey and Chile.

Similarly, in the average OECD country, people with a primary education are generally less satisfied with their personal relationships than their more educated peers (Figure 11.10). On average, the difference between people with tertiary and primary education is around 0.5 points (on a 0-10 scale), with the gap being larger for countries with low levels of satisfaction with personal relationships (e.g. Lithuania, Hungary and Italy). By contrast, in Norway, Sweden and Switzerland, where the average satisfaction with personal relationships is high, gaps by education are small.

Figure 11.9. People with lower educational attainment have less social support

Figure 11.10. People with a lower education level are on average less satisfied with their personal relationships

Box 11.1. Measurement and the statistical agenda ahead

An ideal indicator set for Social Connections would provide information about the quantity of social interactions (e.g., frequency and amount of time individuals spend with household members, their family, friends, colleagues, and other known persons), their quality (e.g. satisfaction with social interactions, perceived loneliness), and the support (e.g. emotional and financial) provided by these connections. Measuring both the quantity and quality of social connections is particularly relevant, as the two do not necessarily capture the same phenomena: spending a considerable amount of time interacting with people does not necessarily prevent loneliness or a lack of support. Each of these concepts is captured, to some extent, by the indicators included in this chapter (Table 11.1).

Table 11.1. Social Connections indicators considered in this chapter
	Average	Vertical inequality (gap between top and bottom of the distribution)	Horizontal inequality (difference between groups, by gender, age, education)	Deprivation
Social support	Share of people who report having friends or relatives whom they can count on in times of trouble	n/a	Gaps in social support	Share of people who report that they do not have friends or relatives whom they can count on in times of trouble
Time spent in social interactions	Average number of hours spent in social interactions per week	n/a	Gaps in average hours spent in social interactions per week
Satisfaction with personal relationships	Mean satisfaction with personal relationships, measured on a 0-10 scale	S80/S20 ratio in mean satisfaction with personal relationships	Gaps in mean satisfaction with personal relationships	Proportion of people reporting a score equal to or below 5 on a 0-10 scale (defined by Eurostat as those with “low” satisfaction levels with personal relationships)

Social support refers to the proportion of people responding “yes” to the (yes/no) question: “If you were in trouble, do you have relatives or friends you can count on to help you whenever you need them, or not?” For country averages, data are pooled over all available years for a three-year period (e.g. 2016-18) to improve the accuracy of the estimates; for reporting inequalities, data are pooled over a longer time period (2010-2018). The source for these data is the Gallup World Poll, which samples around 1 000 people per country, per year. The sample is ex ante designed to be nationally representative of the population aged 15 or over (including rural areas).

Time spent in social interactions includes the amount of time allocated to interacting with friends or relatives as a primary activity (e.g. talking with family members or going out with friends) in a typical day (the averages in this chapter were converted into weekly estimates). Therefore, country averages do not exclude people who did not spend any time in social interactions during the surveyed day. Since only the time spent interacting with family and friends as the main or primary activity is considered, time estimates presented in this chapter are likely to underestimate the total amount spent on social activities, as they exclude those interactions that occur alongside a primary activity (e.g. talking around the dinner table, or chatting on the phone while performing unpaid work). These data are sourced from national Time Use Surveys (TUS), which provide detailed information on the amount of time individuals allocate to their daily activities. Respondents typically keep a 24-hour diary during one or more days in which they precisely record each activity.

Some countries (e.g. Colombia, Mexico and, to a smaller extent, Ireland) use a simplified variant of a time-use diary, which results in estimates that are less precise than for other countries. In addition, in the Mexican time-use survey, respondents are asked about their time use during the seven days prior to the interview. Given the large time lapse between the activity and the interview, responses are likely to be rougher estimates of the true time use. For this reason, time-use estimates for Colombia and Mexico are not shown in this chapter.

Ideally, data collection for time-use surveys would be spread over the whole year, and thus contain a representative proportion of weekdays and weekend days, as well as public and school holidays. Some countries, however, cover only particular periods in the week or year: this is the case, to varying degrees, for Australia, Ireland, Japan, Korea and Mexico. Additionally, differences in activity coding may limit cross-country comparability, especially when access to microdata is restricted, as statistical agencies may aggregate very detailed activities into broader categories that may differ, to some extent, across countries. Finally, as the time-use surveys considered in this chapter were administered in different years, with countries at different stages in the economic cycle, this may affect the observed variations between countries. The data shown here have been harmonised ex post by the OECD drawing on the Harmonised European Time Use Surveys, the Eurostat Time Use database, public-use time use survey microdata, and tabulations from National Statistical Offices. These sources are available in the OECD Gender Database. In those countries for which daily time use did not sum up to 1 440 minutes, the missing or extra minutes (around 30-40 minutes usually) were equally distributed across all activities.

Satisfaction with personal relationships: Survey respondents rate their satisfaction with their personal relationships on an 11-point scale, from 0 (not at all satisfied) to 10 (completely satisfied). The variable refers to the respondent’s opinion/feeling about the degree of satisfaction with his/her personal relationships. The respondent is expected to make a broad, reflective appraisal of all areas of his/her personal relationships (e.g. relatives, friends, colleagues from work etc.) in a particular point in time (these days). The sources for this indicator are Statistics Canada (General Social Survey 2016), INEGI (Subjective well-being in Mexico) and Eurostat (EU-SILC, 2018 and 2013). This indicator refers to individuals aged 16 or more, except for Canada (15 or more) and Mexico (18 or more).

Correlations among Social Connections indicators

At country level, there is a positive and significant correlation (0.5) between social support and satisfaction with personal relationships (Table 11.2): in those countries where social support is higher, people tend to rate their satisfaction with personal relationships higher. By contrast, time spent in social interactions is not significantly correlated either with satisfaction with personal relationships or with social support, implying that each metric captures a different aspect of Social Connections.

Table 11.2. Satisfaction with relationships and social support are not correlated
Bivariate correlation coefficients among the Social Connections indicators
	Social support	Time spent in social interactions	Satisfaction with personal relationships
Social support
Time spent in social interactions	0.09 (26)
Satisfaction with personal relationships	0.51*** (29)	-0.00 (20)
Note: Table shows the bivariate Pearson’s correlation coefficient; values in parentheses refer to the number of observations (countries). * Indicates that correlations are significant at the p<0.10 level; that they are significant at the p<0.05 level, and * at the p<0.01 level.

Statistical agenda ahead

The measure of social support included here suffers from a number of limitations: as a simple yes/no question, it provides no information about the frequency, intensity or quality of support received, nor the type of support (e.g. financial or emotional support). It is also not possible to compute vertical inequalities (i.e. the gap between the top and the bottom of the distribution) from a yes/no question, and in several OECD countries, the measure appears to be reaching a ceiling (e.g. 95% of the population or more reporting support), meaning it lacks sensitivity for assessing group differences. Finally, the small sample sizes of the Gallup World Poll raise issues regarding measurement errors, especially when exploring inequalities among population groups and change over time. An extensive psychological literature dating back several decades exists on social support measurement, and National Statistical Offices are taking increasing interest in such measures, but beyond Europe there is little consistency across NSO practices in collecting these types of measures at present (Fleischer, Smith and Viac, 2016[1]).

Time use surveys (TUS) are among the main sources of information on the quantity of time spent in social activities. Despite a growing number of cross-country initiatives (e.g. the Multinational Time Use Study (MTUS) and the Harmonised European Time Use Surveys (HETUS)), guidelines (e.g. UNECE (2013[2]) and UNSD (2005[3])) and international classifications (e.g. the UN International Classification of Activities for Time-Use Statistics (ICATUS)), which all aim to improve the harmonisation of time use surveys, several pending issues still prevent full cross-country comparability. Greater harmonisation is needed across data collection methods, including the length of diary timeslots, and the number of days on which diaries are completed. Moreover, due to the relatively resource-intensive nature of TUS, these are generally conducted at intervals of about five or ten years (with the exception of the United States). In interim years or where their implementation is not feasible, data on the use of time could be collected through survey instruments with a lower collection and response burden, for example, “light” diaries with pre-coded time use categories (UNECE, 2013[2]).

Harmonised surveys on satisfaction with personal relationships are also conducted on an infrequent and ad-hoc basis. Moreover, information on whether social interactions take place face-to-face or via social networks is sparse. However, the frequency of the latter has risen and is likely to continue to do so with increasing digitalisation. Since computer technology may foster a wider network with weak ties, rather than a smaller network with strong ties, its impact on social interactions is likely substantial (OECD, 2019[4]). Most recent time use surveys ask respondents to report the use of technology but, for the time being, this indicator can be computed for only a limited number of countries.

References

[1] Fleischer, L., C. Smith and C. Viac (2016), “A Review of General Social Surveys”, OECD Statistics Working Papers, No. 2016/9, OECD Publishing, Paris, https://dx.doi.org/10.1787/bb54d16f-en.

[4] OECD (2019), How’s Life in the Digital Age?: Opportunities and Risks of the Digital Transformation for People’s Well-being, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264311800-en.

[2] UNECE (2013), Guidelines for Harmonizing Time-Use Surveys, United Nations Economic Commission for Europe, Geneva, http://unece.org/index.php?id=34496.

[3] UNSD (2005), Guide to Producing Statistics on Time Use: Measuring Paid and Unpaid Work, United Nations Statistics Division, New York, http://unstats.un.org/unsd/pubs/gesgrid.asp?id=347 (accessed on 12 December 2019).

╳

Metadata, Legal and Rights

This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.

https://doi.org/10.1787/9870c393-en

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.