HELPING VALUES AND CIVIC ENGAGEMENT Open Access

Helping other people has many connotations but it can be defined as a contribution to the achievement of something. In the context of civic engagement helping might include: organising a community event, supporting a particular campaign or voluntary group or assisting a neighbour. It is this very general notion of helping, viewed as a form of civic engagement that we are interested in here.

Helping other people, when attached to positive and socially desirable outcomes is vital to the functioning of civil society. It is part of the infrastructure of community, both in terms of how relationships develop and how they are maintained. Helping is the embodiment of civic society and an underlying notion of helping could be considered to underpin civic participation. It is notable, for example, that Plagnol and Huppert (2010) have identified that informal volunteering (providing help to those outside a person's family, work or a voluntary group) is positively associated with formal volunteering (helping in a charity group). Helping is arguably part of the set of human values. Qualitative research by Roberts and Devine (2004) found that people often understand their voluntary work in terms of helping out and as just part of their everyday activity. Whilst Lie et al. (2009) in their qualitative research on volunteering amongst older people have highlighted how volunteering is seen as an expression of citizenship.

Helping in the form of voluntary activity is increasingly being seen as an aspect of citizenship, renewing democracy and also delivering services. Governments are increasingly focussing on the voluntary sector and the role of voluntary activity of citizens as a key aspect of the restructuring of the welfare state (Macmillan and Townsend 2007; Milligan and Conradson 2007). In many countries, including the UK, the responsibility of citizens has become part of the national school curriculum though the teaching has taken different forms (Callan 1997). A range of European Union policies have increasingly focused on engaging citizens in civil society. Ilcan (2009) has pointed to the mobilization of responsible citizenship and it is important to consider the wider context of citizenship in the liberal democratic state and the balance between rights, responsibilities and obligations. See Janoski (1998) and Marinetto (2003) for discussion.

However, a number of measures point to a decline in citizen involvement in many advanced democracies (Putnam 2000) For example, in the UK Curtice et al. (2010) argue that the British electorate can no longer be relied upon to vote simply out of duty or habit. In relation to volunteering in the UK evidence from the Citizenship Survey has shown that levels of volunteering at least once a year have declined by 10 percent since 2001 – from 74 to 64 percent. The notion of a long-term decline is debated (Stolle and Hooghe 2005). For example, Inglehart (1997) and Dalton (2008) amongst others have pointed to the new types of civic engagement activities people are involving themselves in. Hall (1999) concluded that in the UK, contrary to Putnam's findings in the USA, aggregate levels of civic engagement have been maintained in the UK.

Across many European countries there are considerable variations in voluntary activity rates (Morales 2009). Associational involvement and voluntary work varies from almost all the population in Norway having some kind of associational involvement to 28 percent of the population in Russia. Badescu and Neller (2007) have shown that across European countries middle aged men with higher qualification levels and income are the most likely to be involved in voluntary associations. Dekker et al.'s (2007) research in The Netherlands using survey and time use diaries has shown that overall rates of participation in voluntary work are likely to remain static in the future and young people's rates of volunteering are in decline.

In this paper we consider general measures of helping and specifically the importance people attach to helping and caring for the well being of others and the extent to which they report helping in practice (help with or attending local activities). The key research questions are: how does the importance people attach to helping other people vary by: age, gender, education, country, type of area and the local context of help (the extent to which a person thinks other people living locally help each other)? Do those people who see helping others as important help in practice? Are those people who see helping people as important more likely to be involved in civic engagement activities?

Our analysis focuses on the European Social Survey (ESS). The ESS collects a range of data on attitudes and reported behaviour patterns. Multivariate multilevel modeling techniques are used to unravel variations and relationships in the indicators of help and civic engagement at the individual, region and country levels.

The relationship between individual values and behaviour is complex. From a psychology perspective Fishbein and Ajzen's (1975) model highlights the importance of understanding the influences on intention to act in a certain way i.e., the individual's views of the activity and its outcomes. Research by Berkowitz (1972) concludes that the helpfulness norm is only a weak determinant of help giving of most people in many situations.

Evidence suggests that there appears to be a natural predisposition to empathize with and help others but that this is affected by learned experiences and the specific context (Dovidio et al. 2006). Research has also highlighted the development of concern for others amongst young children (Schwartz and Bilsky 1990; Zahn-Waxler et al. 1996; Inglehart et al. 2004). Trivers (1971) has described what he sees as reciprocal altruism – where a person is altruistic because the individual is likely to later be the recipient of similar altruistic acts. Batson (1998) also highlights how the motivations to help can also be down to self-interest, as people feel better about themselves after helping. Experimental research in the 1960s found that the extent to which young, working class men gave help was associated with the help they had received in the past (Berkovitz 1968). Young people have been shown to be more likely to volunteer if their parents volunteer (Rosenthal et al. 1998).

In terms of volunteering, research has shown that a range of factors are associated with such activity including: being male, older people, those with higher qualifications, those who are married and have children, those people who live with someone who volunteers and those who attend religious services regularly being more likely to volunteer (Wilson 2000). Though these associations can vary by type of volunteer activity. It is notable that in research by Batson (1998) intelligence and religion were not found to be strong predictors of whether a person would help a stranger. The opportunities for volunteering are also important to consider. For example, those with children are more likely to be asked to volunteer, as they are often part of social networks where other people are already volunteers (Brady et al. 1999). Levine et al. (2008) in their comparative experimental study of different measures of helping in three American cities, described in terms of prosocial behaviour, highlight the importance of community norms in helping.

In this article we consider help and helping in three inter-related contexts: (i) in the context of personal characteristics such as age, gender and educational attainment; (ii) in the context of geography such as the country or region in which the individual lives, including the extent to which individuals feel other people in their local area help one another; and (iii) in the context of other civic engagement activities such as contacting a politician, signing a petition and voting. We use a general, low resource cost measure of helping in practice so that we can limit the impact of other factors that may be associated with people providing help in specific circumstances.

The ESS is a survey of people's attitudes, beliefs and behaviour. The 2006 wave includes data from 23 European countries. The requirement is for random probability samples of the eligible residential populations. The required sample size is 1500 or 800 in countries with populations less than 2 million. The target response rate is 70 percent. The ESS is a high quality survey. Questions are translated and tested in country specific languages and this has been subject to independent verification and testing (ESS 2007).

We have considered seven indicators of help and civic engagement measures and a measure of the extent to which the respondent perceives people living locally help one another (see Table 1).

We began by using single level logistic regression models taking each of the seven outcomes as separate recoded dichotomous variables and relating this to country with the UK as the reference category and included: age gender, educational attainment and the perceived local context of help. All single-level analyses were weighted to take account of the population size of the different countries and also in relation to the country specific sampling strategies.

In order to examine the overall extent of between country variation in each of the outcome variables, we fitted bivariate logistic multilevel models (Goldstein 2003). This set of models was used to investigate the inter-relationships between the outcomes at the three different levels: individuals, regions and countries. For each person (indexed by i) in each region (indexed by j) in each country (indexed by k), there are seven outcome variables considered. y_ijk is a n×2 matrix containing any two of the seven outcome measures for each of the n individuals in the sample. p_ijk is a n×2 matrix of predicted probabilities for the two outcomes given the explanatory variables, that is, p_ijk=pr(y_ijk=1 | X_ijk). e_ijk is an error term.

v_1k is the country level error term for the first help/civic engagement outcome and u_1k is the region level error term for the first help/civic engagement term. These can be similarly defined for the second outcome. All error terms are assumed to be uncorrelated between levels. p_ijk is a matrix of 2 probabilities for any pair of civic engagement outcomes e.g., p_1ijk for help in practice and p_2ijk contact a politician. z_1ijk and z_2ijk are dummy variables to indicate outcome 1 (e.g., help in practice) or outcome 2 (e.g., contact a politician). $σν12$ are respectively: the country level variance of the first outcome, the country level variance of the second outcome and the country level covariance between the first and second outcomes, from which we can calculate the country-level correlation.

Firstly, in terms of measuring helping in practice it is clear that ‘Helping out with or attending activities organized in your local area’ is only one aspect of helping. Such a measure may not capture those who are too busy to help with or attend activities in their local area because they are committed to other helping activities. However, it is a low resource cost form of helping in terms of time and commitment. As such, whilst it does carry some ambiguity and is general, we feel it is a good indicator. Local activities might include a community gathering or fundraising event but we believe the measure provide some valuable insights into helping behaviour in a general sense.

Secondly, the social desirability of giving a certain response may also be a factor as respondents may state that they see helping others as important or state that they help in practice because they feel that is the acceptable answer.

Thirdly, whilst in the single level models we control for age, gender and educational attainment levels, we restricted part of the multi-level analysis to null models, as we are interested in the inter-relationships between the response variables at the different levels. We are aware that other factors may be at work here in relation to each individual outcome including for example: health, governance infrastructure, electoral system and marginality.

Finally, it could be that there are variations in how the key concepts are understood in different languages. However, the ESS is a high quality survey and questions are translated and tested in country specific languages (ESS 2007). In relation to measuring values cross nationally and some of the limitations see Schwartz (1994) and Davidov et al. (2008).

We consider these important limitations in the interpretation of our findings.¹

Nearly two-thirds (61 percent) of people across the countries in the ESS state that they feel helping other people is important (the value of help) (Figure 1). There are however considerable differences between countries.

There appears to be some clustering by liberal democratic tradition and social welfare regime type which other research has also highlighted. However, the countries having the highest rates of the value of help are quite mixed including Spain, Switzerland and Slovenia. It is notable that Slovenia is considered to be one of the most democratic of the former Communist states (Freedom House Index 2010).

In relation to helping in practice (helping with or attending a local activity) the overall rates are much lower as shown below. Just over a quarter (27 percent) of people state that they help others in practice (Figure 2). The gap between reported attitudes and reported behaviour are substantial.

The country level differences are again striking. At both ends of the scale there appears to be clustering by democratic tradition and welfare regime type with certain Scandinavian countries contrasted with post Communist Eastern European countries and Southern-European countries.

It is also notable that the countries with the highest proportions of people stating that helping others is important are not necessarily the countries with the highest percentages of people who help in practice (helping with or attending local activities). The difference in the value of help and helping in practice in Bulgaria is notable. Supporting our argument that the extent to which people state ‘helping others is important’ has the qualities of a human value is that this shows the least between-country variation. Moreover, the proportion of people stating that helping others is important has been stable over recent times across Europe. In 2002, 63 percent of people in the ESS stated that helping others is important and in 2004 this was 65 percent.

In relation to civic engagement across Europe data from the ESS suggests that: 12 percent of people had contacted a politician in the last year; 20 percent had signed a petition; 7 percent had taken part in a demonstration; 22 percent had been involved in voluntary work and 75 percent of people had voted in the last election. Table 2 provides a summary of the between country variances and the three highest/lowest rates of each outcome.

This analysis highlights the higher rates of helping and civic engagement outcomes across certain Scandinavian countries and lower rates across certain Eastern European countries.

In relation to the extent to which people think helping others is important and whether they help in relation to organising or attending local activities there are some differences between men and women and by age (Figures 3 and 4). Women across all age groups are more likely than men to state that helping other people is important (64 percent compared to 57 percent).

Conversely in relation to helping in practice we see that overall women are less likely to help than men (26 percent compared to 28 percent). Though statistically significant, the difference is small. In general older men and women aged 25+ are more likely to state that helping others is important. Older men and women are not the most likely to state that they help in practice as measured here. This may be a consequence of the barriers that older people may face in implementing their values.

In terms of the local context of help across countries in the ESS, just under half of all respondents (48 percent) stated that people in their local area help each other to some extent. Only 9 percent of people stated that people in their local area helped each other a ‘great deal’.

We began by producing single level logistic regression models for each of the outcomes including countries as fixed effects with the UK as the reference category and control for key demographics. We use the UK as an example country in order to produce a picture of the relative standing of the other countries represented in the ESS. The UK is a valuable case study country it has relatively high levels of people stating that helping other people is important and also relatively high levels of helping people in practice. We restrict the results to 5 percent (or less) significant differences in the coefficients.

Table 3 compares the differences in help as a value, helping in practice and the civic engagement outcomes across European countries compared to the UK and also controlling for age, gender and educational attainment.

The findings support the descriptive analysis. Older people, women compared to men and those with higher educational qualifications are more likely to state that it is important to help other people. In relation to helping in practice (help with or attending local activities) there is a slight negative relationship with age (which may signify a person's ability to help) and a negative relationship with gender perhaps indicating that women may face additional barriers to helping in practice. Women are also less likely than men to volunteer as measured here though this result is not statistically significant. There is a positive relationship between education level and helping in practice. In relation to the other measures of civic engagement those with higher levels of education are more likely to contact a politician. Men are more likely than women to contact a politician, take part in a demonstration and vote. Older people are more likely to vote and contact a politician compared to younger people. Older people are less likely to take part in a demonstration or sign a petition.

Comparing across different countries in the ESS we found considerable differences in the aggregate levels of helping and civic engagement activities. In relation to the value of help we found that almost all countries were significantly less likely to see helping other people as important compared to the UK. Only Spain was significantly higher than the UK. Considerable differences were also apparent in relation to help in practice. Only in Austria was the likelihood of helping in practice significantly higher than the UK. In relation to the other civic engagement activities there was also considerable country level variation compared to the UK. Of particular note is the significantly higher likelihood of voting in many countries compared to the UK, accept for a number of post Communist Eastern European countries, and the lower likelihood of signing a petition across all countries compared to the UK.

These findings suggest that there are important country level contextual issues to consider when looking at the value of help and helping in practice and other measures of civic engagement across Europe.

We now look specifically at one aspect of context – the extent to which people feel other people around them help one another. Table 4 below summarises the results of modelling each of the seven outcome measures after controlling for age, gender and educational attainment. The reference category is – people think other people in their local area do not help one another at all.

We find a generally positive relationship with the value of help and the local context of help. This suggests that as people perceive more help around them they are also more likely to state that helping others is important and to help in practice, although from this analysis we cannot establish the causality. In relation to the other civic engagement activities, as a person's perception of the extent to which people in their local area help one another increases so does, on the whole, their likelihood of voting, contacting a politician and signing a petition.

We now consider the region and individual level correlations between the help and civic engagement outcomes, calculated as explained in the methodology section. In Tables 5 and 6 the correlations marked with a * are statistically significant at the 5 percent level (based on Wald tests of the underlying covariance).

At the country level there is evidence of relatively high levels of correlation between help as a value and help in practice (helping with or attending local activities) and the other measures of civic engagement. The associations of the other civic engagement outcomes with the value of help are weaker than they are with helping in practice. It seems that the value of help at the country level can be detached from what people may do in their everyday lives.

Though the correlations are smaller at the individual level than at the country level, we can again see relatively strong associations between help in practice, volunteering and other civic engagement outcomes. The correlations are weaker with the value of help than with helping in practice. Perhaps not surprisingly help in practice is strongly related to voluntary activities at the individual level. More generally we note that again all correlations are positive. These correlations are in general small mainly because these variables are binary at the individual level.

People who see helping others as important and who help in practice are also likely to be involved in voluntary activity. This is perhaps not that surprising at first glance but other civic engagement activities and in particular voting and taking part in a demonstration are less associated with helping in practice.

At the regional level the results are more varied. The negative associations between the value of helping and signing a petition, taking part in a demonstration and voting may suggest that at the region level the context in relation to particular cultural or political traditions and governance infrastructures can exert a general influence. The strong positive associations of helping in practice with volunteering and contacting a politician suggests that there may be some area effects perhaps relating to the role of individual representatives and the opportunities for participation more widely.

Finally, it is important to consider how other types of civic engagement are associated at different levels. These inter-relationships are often ignored, or the information is lost when the variables are combined into single indices. Table 7 highlights the associations between the other civic engagement outcomes.

In terms of civic engagement at the country level there are strong associations between signing a petition and contacting a politician and volunteering. This perhaps reflects a certain type of political culture and types of civic engagement activities. At the individual level there is a positive association between signing a petition and contacting a politician and between signing a petition and taking part in a demonstration, though this is not so evident in relation to voting. As other research has shown, at an individual level voting is an action linked to a range of additional contextual factors such as incumbency, campaigning and marginality.

The analysis of generalised helping has shown that the value of help underpins civic participation activities both conceptually and also in practice. However, the associations are not as direct as might be expected. Overall, people are much more likely to see helping others as important than actually help in practice even in terms of the general and low resource cost measure used here.

Our analysis suggests that the way people see the importance of helping others has the quality of a human value that can transcend contextual factors, though there are still considerable between country variations. However, it seems that the importance people give to helping others is detached from what many people do in their everyday lives. Help as a human value is a belief that can motivate action but helping in practice can be dependent on a person's individual characteristics and circumstances. Moreover, arguably we have identified a limit to the amount of helping people can undertake. Even people who help out a lot have a limit to what they can do. It is notable that few people live in areas where they feel other people help each other a great deal.

Women across all age groups are more likely than men to state that helping other people and caring for their well being is important. Women are however less likely to state that they help in practice as measured here (though the difference is small). In general older men and women aged 25 years and older are more likely to state that helping others is important, though the association is weak. Those with higher educational attainment levels are more likely to state that helping other people is important and to actually help in practice compared to those with lower levels of education. It is notable that Anheier and Salamon (2000) also found evidence of higher rates of volunteering across Europe amongst those people in higher socio-economic occupations.

The gap between the value of helping and helping in practice is likely to be, in part, a consequence of the barriers certain people face and the lack of opportunities for implementing their values and becoming more civically engaged. For example, older people may not have the opportunities, the resources, or be in good enough health to help in practice. As such, the ageing population in many Western countries is a crucial issue in relation to the future patterns of helping. Women are perhaps involved in other types of helping activities but not helping with or attending activities in their local area. It may also be that women face additional barriers to translating their values into practice, for example due to other caring responsibilities and a lack of time, resources and opportunities. It is notable that research by Amato (1993) reveals that people are most likely to help friends and family members and that the individual characteristics of individuals are more closely associated with planned helping.

Survey measurement issues may be a factor here in relation to how the notion of helping is perceived. Women and men, as a consequence of perceptions of gender roles, may be responding to the questions on helping along social desirability lines. Hopefully, the general helping in practice measure we have used has minimised this. But the gender differences are still striking. In relation to social class and education level it is important to note that research in the UK by Williams (2003) suggests the neighbourliness and mutual support in lower income areas are not always captured in surveys.

Across the countries in the ESS our findings suggest that the importance people attach to helping others and the extent to which they help in practice varies considerably. This is the case even given the generalised measures of helping used here and the testing that has been done on translating the survey questions and concepts in each country. The low levels of the value of help and helping in practice in many post Communist Eastern European countries are particularly striking. Country and context specific differences such as in the infrastructures for civic engagement and the public and cultural perceptions of the role of social welfare are clearly important. Cross national research by Parboteeah et al. (2004) has highlighted how formal volunteering levels tend to be higher in countries where there is greater liberal democracy and where wealth and education levels are higher. Morales (2009) concludes that people with similar social and cognitive attributes are not equally likely to be politically active in all western countries because of contextual differences. Huppert et al. (2009) and Plagnol and Huppert's (2010) research on volunteering and wellbeing have highlighted the low levels of volunteering in many Eastern European countries. The recent history of the state structures are thought to be an important factor in explaining this including compulsory volunteering and its impact on social norms and values, a lack of voluntary sector infrastructure and the impact of economic transition. Kuti (2004) suggests that the shared history of military rule and authoritarianism (including the enforcement of volunteering) did not engender the development of the infrastructure of civic society. Bartowski and Jasinska-Kania (2004) have highlighted the positive association between levels of voluntary association membership and activity with indicators of human development and also (though to a lesser degree) indicators of economic freedom and civil liberties across European countries. They also suggest that that the development of democracy and a capitalist economy in the short term can serve to reduce community activities. Meier and Stutzer (2008) point to a substantial drop in volunteering following reunification with West Germany due to the collapse in the state infrastructure of volunteering in the East. Such contextual issues are likely to explain the variations we identified at the regional level. As we have explained, the negative associations between the value of helping and other types of civic engagement suggests that particular cultural or political traditions and governance infrastructures are a factor. Moreover, the strong positive associations of helping in practice with volunteering and contacting a politician suggests that there may be some area effects perhaps relating to the role of individual representatives and the opportunities for civic participation within an area.

Research has also examined the impact of psychological factors, levels of social integration and cultural resources on the likelihood of volunteering (Plagnol and Huppert 2010). Whilst associations were identified with people with a positive outlook on life, those with a more active social life and those with hedonistic values and achievement values were more likely to volunteer, such characteristics did not fully explain the differences in volunteering rates across Europe. Anheier and Salamon's (2000) critique of single factor theories of the non-profit sector (which by definition involves voluntary participation) including: market failure and government failure, supply side issues, levels of trust and size of welfare state, points to the social origins of the voluntary sector. Such an approach is clearly important here in trying to fully explain the cross-national differences in helping. The social origins approach highlights the combined importance of the social, political, and cultural as well as economic structures both across and within countries.

In relation to helping and other types of civic engagement activities, those people who help in practice in terms of helping with or attending a local activity are also likely to be involved in other voluntary activities. Whereas the other types of civic engagement such as voting and taking part in a demonstration are less associated with the value of help and helping in practice. This raises important issues in relation to democracy and the ways in which citizens may take a critical stance towards certain forms of engagement. Research across European countries by Badescu and Neller (2007) found that higher levels of dissatisfaction with the way democracy works tends to be associated with higher levels of involvement in voluntary associations. However, Dekker and van den Broek (1996) found a positive association between volunteering and political participation across European countries. Putnam (2000) also suggests that volunteers are more interested in politics and less cynical about politicians when compared to people who do not volunteer. Research by Dalton (2008) has highlighted that alongside the decline in citizen duty in the USA there has been an increase in informal independent participation i.e., being active in a community project and buying products for political reasons.

Highlighting the importance of context, helping in practice, as measured here, is associated with the extent to which people feel that other people in their local neighbourhood help each other. Moreover, as people perceive help around them they are also more likely to say that helping others is important. In relation to other civic engagement activities, as a person's perception of the extent to which people in their local area help one another increases so does, on the whole, their likelihood of voting, contacting a politician, signing a petition and being involved in voluntary activity.

We are conscious of the limitations of the generalised helping measure used in the analysis. In fact our findings on the importance people give to helping others and helping in practice may in fact be an over estimate due to the social desirability of certain responses and the low resource cost measure of helping we have used. In addition, whilst we cannot prove the direction of causality our contention is that the value of help is an important focus for understanding the development of civic society.

Across Europe the relatively high levels of people seeing helping others as important is a potential building block for renewing civil society. Engendering the value of help remains important for the creation of a civic society but there is also scope for policy innovation to assist people overcome the barriers they may face in translating their values into action rather than just generally encouraging people to become more involved in civic activities.

K. Purdam is a Research Fellow at the University of Manchester, UK

M. Tranmer is a Senior Lecturer in Social Statistics at the University of Manchester, UK