Ageing and consumption in Finland: the effect of age and life course stage on ecological, economical and self-indulgent consumption among late middle-agers and young adults between 1999 and 2014

Previous studies on ageing consumers have mainly focused on chronological age and generational values or studied ageing and consumption with cross-sectional data. Few quantitative studies exist that examine the effect of age together with life course on consumption using longitudinal data. To bridge this gap, the article examines ageing and attitudes towards consumption in Finland, focusing particularly on late middle-agers (46 to 60 year-olds) in comparison to young adults (18 to 30 year-olds) between 1999 and 2014. The article explores three consumption patterns based on attitudinal statements: ecological, economical and self-indulgent consumption. Through analysis of a nationally representative survey study in Finland (N=8543), the article reveals that in all years under examination, late middle-agers reported more ecological attitudes towards consumption than young adults. In 1999 and 2004, the attitudes of late middle-agers appeared more economical, but age-related differences in economical attitudes disappeared between 2009 and 2014. In each year, late middle-agers reported less self-indulgent attitudes than young adults, and these age-related differences did not remarkably change between 1999 and 2014. The results indicate that in 1999 and 2004, ecological and economical attitudes towards consumption were best predicted by age at the year of the study. In later years, ecological attitudes were more closely determined by life course stage, i.e. household type and other socio-demographic determinants. Regarding economical attitudes, generational or cohort effects were pronounced among late middle-agers in 1999. In contrast, the significance of age remained throughout the

statements: ecological, economical and self-indulgent consumption.Through analysis of a nationally representative survey study in Finland (N=8543), the article reveals that in all years under examination, late middle-agers reported more ecological attitudes towards consumption than young adults.In 1999 and 2004, the attitudes of late middleagers appeared more economical, but age-related differences in economical attitudes disappeared between 2009 and 2014.In each year, late middle-agers reported less selfindulgent attitudes than young adults, and these age-related differences did not remarkably change between 1999 and 2014.The results indicate that in 1999 and 2004, ecological and economical attitudes towards consumption were best predicted by age at the year of the study.In later years, ecological attitudes were more closely determined by life course stage, i.e. household type and other socio-demographic determinants.
Regarding economical attitudes, generational or cohort effects were pronounced among late middle-agers in 1999.In contrast, the significance of age remained throughout the

Introduction
Ageing consumers are rapidly becoming the most affluent consumer segment in the marketplace (Biggs et al., 2007;Jones et al., 2008;Drolet et al., 2010).In Finland and worldwide, the income levels of those aged 55 to 65 have risen and this group also consumes more than 30 years ago (Atkinson and Hayes, 2010;Ahonen and Vaittinen, 2015).The factors that affect the consumption patterns of older people are not entirely known.Previous studies of ageing consumers have mainly focused on chronological age (Drolet et al., 2010).The majority of such studies have interpreted the consumption patterns as reflecting generational values (Carr et al., 2012;Eastman and Liu, 2012;Parment, 2013), and studied consumers with cross-sectional data (e.g.Katz-Gerro, 2002;Wilska, 2002).Most quantitative studies on ageing consumers have ignored the effect of life course stage, typically understood as biographical changes and changes in household structures, on consumption (Settersten, 2003;Elder and Shanahan, 2007;Hutchison, 2011).During periods of economic growth, older adults nevertheless have decreased financial responsibility for their children and more opportunities to spend on themselves as daily consumption is no longer affected by the needs of other family members (Jones et al., 2008).Intergenerational relationships, being influenced by macro-economic situations, might, however, be subject to change due to the longstanding economic depression and the increased income inequality among the workingage population (Ahonen and Vaittinen, 2015).
The rest of the article begins with an overview of explanatory approaches to consumer decisions with respect to age, life course and time period.This is followed by an empirical analysis of longitudinal data derived from four surveys carried out in Finland in 1999, 2004, 2009 and 2014.Findings and implications are discussed in terms of age, the life course, and consumption, especially highlighting the findings regarding late middle-agers.

Consumer decisions, age and life course
The ethical value paradigm argues that consumer choices are shaped by personal values like consumerism, described as hedonism, materialism, possessive individualism and narcissism, or sustainable values, described as religious traditions, radical ecology and other values outside the modern capitalist marketplace (e.g.Crompton, 2010;Holt, 2012).Empirical studies show significant differences in attitudes towards consumption among consumers at all ages; younger adults' attitudes are more materialistic and hedonistic than the attitudes of the middle-aged or the elderly, who report more economical, ethical and environmentally conscious attitudes and interest towards cultural consumption (Katz-Gerro, 2002;Wilska, 2002;Atkinson and Hayes, 2010;Purhonen et al., 2011;Berg, 2015).Personal value orientations have shown to change along with ageing, as people in late midlife start to value benevolence, i.e., the promotion of the wellbeing of close relatives, universalism, i.e., taking care of other people and nature, and safety, implying the security of society, personal relationships, and life in general (Charles and Carstensen, 2010;Pulkkinen and Polet, 2010).
Accordingly, consumption practices are explained by the life course approach which addresses the roles of household resources and life-course events on consumption (e.g.Plessz et al., 2016).Biographies and biographical transitions, such as the initiation of cohabitation with a partner or the first child's birth offer opportunities for alterations of new routines and leisure activities (Bisogni et al., 2005;Bove and Sobal, 2006;Plessz et al., 2016;Southerton, 2006).Many consumption practices are linked to parental responsibility, especially in consumption that is related to nutritional and health norms (Régnier and Masullo, 2009;Plessz et al., 2016).Young adulthood, involving transitions such as leaving the parental home, finishing education, entering working life, forming a romantic relationship and becoming a parent (Settersten, 2003;Elder and Shanahan, 2007) differs from late midlife that is characterised by the approach of the zenith of one's career and life's highest social position and income.Important transitions also occur as children move away from home and leisure time increases (Helson et al., 2006;Kokko, 2010).Life transitions in certain cultural, societal, geographical and historical contexts (Pilcher, 1995) might nevertheless be under change due to the risks and constraints related to job markets and decreasing financial resources.In societies with advanced economies, many life transitions related to achieving full independence have generally become postponed, as the life transitions that used to occur earlier, in people's 20s, now often occur in their 30s (e.g.Izuhara, 2013).

Consumer decisions, time period and generational cohorts
As the values and attitudes in comparison to individuals' behaviour show major contradictions, the consumption patterns of individuals are understood in respect to market construction paradigms that specify how market actors repurpose particular cultural, institutional and technological resources to construct and transform markets; hence, the consumption patterns of individuals get constructed as markets evolve (Holt, 2012).The macro-economic situation of society, including economic booms and recessions, necessarily affects material values, attitudes and individuals' behaviour.The economic depression in Finland during 1990-1994 as well as the following economic boom in the late 1990s and early 2000s caused major changes in consumption patterns and attitudes towards consumption (Laaksonen et al., 1998;Wilska, 2002;Räsänen, 2003;Autio and Heinonen, 2004).The global financial crisis in 2008 and 2009 in most developed countries increased poverty and deprivation (Saunders and Wong, 2012).The effect of time period on consumption can thus cause generational or cohort effects on consumption (see also Mannheim, 1952;Turner, 2002).In previous studies, cohort effects were pronounced in the dimensions of early adoption of new products, conscious consumption and personal debt, indicating that members of Generation X, born in the 1960s and 1970s, exhibit the highest rates of overconsumption and competitive consumption while also displaying the lowest rates of conscious consumption (Carr et al., 2012).
In Finland, the Baby Boomers experienced the post-war economic boom and the rise of the welfare state (Karisto, 2007;Purhonen, 2007) and the values of the Boomers are often described as post-materialistic (Wilska, 2011).However, the life course effects may also be in play, and thus consumption patterns can change when moving closer to the retirement age.Many consumption-related values and attitudes, such as attitudes towards ecological and ethical issues, have become established as socially desirable opinions during the past decades (Aro and Wilska, 2014).Thus, recent studies show little variation between age groups or generations in these attitudes (Huttunen and Autio, 2010).

Research questions
The study examines ageing and consumption in Finland by distinguishing the effects of

Data and methods
Ecological, economical and self-indulgent attitudes towards consumption were explained in 1999, 2004, 2009 and 2014 by age at the year of the study and life-course stage including marital status the household type.Control variables were gender, education, employment position, socio-economic position and income per consumption unit.The coding of socio-economic position was conducted on the basis of the ISCO-08 (International Standard Classification of Occupations) classification used by Statistics Finland.The variable 'income per consumption unit' 2 was created using the OECD's scale3 and recoded into five different income quintiles.

Statistical procedures
The analysis started with a factor analysis using attitudinal variables of ecological, economical and self-indulgent consumption that were available in the data from 1999 to 2014.Factor analysis was carried out with the principal axis factoring method and promax rotation to reveal the latent dimension of attitudinal variables.
Three aggregate variables were constituted based on the loadings on each dimension in the factor analysis.To detect the development of the attitudes between 1999 and 2014 among young adults and late middle-agers, the means of the aggregate variables were compared between age groups for each year, by building Two-Way ANOVA models for each year under examination.
The ANOVA model was further executed to examine the effects of age and life -course stage while the effects of other independent variables were controlled for.The ANOVA models were executed separately for the 1999,2004,2009 and 2014 data sets.The factor scores of the consumption attitudes were used as continuous independent variables and the socio-demographic determinants as categorical dependent variables.

Results
2 First converted into 2014 euro.
Factor analysis confirmed the latent dimensions of ecological, economical and selfindulgent consumption.The Kaiser-Meyer-Olkin measure of sampling adequacy was 0.749 and Bartlett's test of sphericity was significant at the level of p=0.000.In each of the three factors, the rotation sums of squared loadings varied from 1.12 to 1.90 based on the factor loadings of the attitude statements.The Cronbach's alphas in the factors 'Ecological' (.645), 'Economical' (.567) and 'Self-indulgent' (.525) were rather low, but this is acceptable in attitudinal questions in social science surveys with large number of questions and respondents.

Ecological attitudes towards consumption
On the basis of the factor loadings, three aggregate variables were constructed: ecological, economical and self-indulgent attitudes.In ecological attitudes towards consumption, the comparison of means between age groups was statistically significant (p<0.000) in the ANOVA model in all years under examination, except in 2009 (p=0.059).Figure 1 shows that the differences in ecological attitudes between late middle-agers (M=3.84) and young adults (M=3.52) were largest in 1999 and 2004.In In the ANOVA model for ecological attitudes between 1999 and 2014, the Adjusted R² values were .071,.098,.083,.050, in chronological order.

Economical attitudes towards consumption
In economical attitudes towards consumption, the comparison of means between age groups was statistically significant (p< 0.000) in the ANOVA model in 1999 and 2004.
In 2009, the difference was significant at the level of p=0.01, and in 2014, at the level of p=0.05. Figure 2 shows that between 1999 and 2009, late middle-agers reported more economical attitudes towards consumption (M=3.16)than young adults (M=2.76).
Differences between age groups disappeared between 2009 and 2014, and in the last year under examination, late middle-agers reported less economical attitudes towards consumption (M=2.89)than young adults (M=3.02).
[insert Figure 2  In the ANOVA model for economical attitudes between 1999 and 2014, the Adjusted R² values were .161,.143,.107,.021, in chronological order.

Self-indulgent attitudes towards consumption
In self-indulgent attitudes towards consumption, the comparison of means between age groups was statistically significant at the level of p=0.000 in the ANOVA model in all years under examination.Particularly in 2014, late middle-agers reported less selfindulgent attitudes towards consumption (M=3.47)than young adults (M=4.06).
Differences between age groups remained almost unchanged between 1999 and 2014.
[insert Figure 3  In the ANOVA model for self-indulgent attitudes between 1999 and 2014, the Adjusted R² values were .067,.056,.088,.082, in chronological order.

Discussion and conclusions
This study threw light on the connection between consumption and ageing by focusing on the development of attitudes towards consumption among late middle-agers and young adults between 1999 and 2014.The results of the study support to some extent the previous findings of young consumers as hedonistic actors with self-indulgent attitudes, and ageing consumers as environmentally concerned with ecological attitudes towards consumption (Katz-Gerro, 2002;Wilska, 2002;Atkinson and Hayes, 2010;Purhonen et al., 2011;Berg, 2015).The study revealed, however, that age in the year under examination explains only certain attitudes towards consumption, and currently, the significance of age is decreasing, particularly in ecological and economical consumption.The significance of presence of children in the household is relatively strong, especially in ecological and economical attitudes towards consumption.The respondent's age remained a significant determinant in ecological attitudes only in 1999 and 2004.In later years, ecological attitudes were better explained by the absence of children in the household and other socio-demographic determinants, such as gender.
These findings suggest that the ecological attitudes of the late middle-agers are increasingly explained by their life stage, such as changes in household structures and values related to life stage, such as benevolence, universalism and the wellbeing of others (Charles and Carstensen, 2010;Helson et al., 2006;Kokko, 2010;Pulkkinen and Polet, 2010).
Furthermore, the study exposed a significant change in economical attitudes towards consumption in both age groups under the years of examination.In 1999 and 2004, late middle-agers reported more economical attitudes towards consumption, but the differences between age groups faded away in later years.In fact, in 2014, young adults reported these attitudes more frequently than late middle-agers.The effect of age on economical attitudes remained significant only in 1999 and 2004, but the household structure was associated with economical attitudes throughout the years.In 1999, there was an economic boom in Finland, but in 2014, a long-standing economic depression persisted.Yet, late middle-agers reported economical attitudes more frequently than young adults in 1999.It is likely that the late middle-agers in 1999, who belonged to the Baby Boomer cohort, possessed economical consumption attitudes regardless of the macro-economic situation.This may indicate a generational or cohort effect on economical consumption.The attitudes of young respondents may be more dependent on economic conditions, and thus the growth in economical attitudes among young adults could be affected by the economic depression that started in 2008.In the whole period, income level was negatively associated with economical attitudes towards consumption.
In contrast, for self-indulgent attitudes age remained significant in all years under examination, and household structure i.e. the absence of children or other sociodemographic determinants were not remarkably associated with these attitudes.
Previous studies have illustrated members of certain generations as particularly hedonistic or prone to status consumption, and interpreted these consumption ideals as typical for certain generational cohorts (e.g.Carr et al., 2012;Eastman and Liu, 2012).
In our study, no generational or cohort effects in self-indulgent attitudes were pronounced, and self-indulgence was associated with age in all years under examination.
There are naturally limitations to the study.Although the data cover a long time period and it is unique in that sense, it is not panel data with the same respondents over time.In survey studies respondents often tend to give socially desirable answers and therefore attitudes towards consumption may not correspond with actual consumer behaviour (see also Holt, 1997;Thompson and Troester, 2002).The results are thus limited to selfreported attitudes regarding certain statements in a Finnish context.Despite these limitations, the attitudes reflect consumption ideals and trends over time and across generations in societies of advanced economies.By examining the development of attitudes in a national context, implications for other similar societies can be postulated.
However, comparative research and even longer longitudinal studies will also be necessary in the future.This article is protected by copyright.All rights reserved.
age and life course stage on consumption among late middle-aged and young adult consumers in 1999, 2004, 2009 and 2014.The research questions are the following: RQ1: How have ecological, economical and self-indulgent attitudes towards consumption developed among late middle-agers (46 to 60 yearolds) and young adults (18 to 30 year-olds) between 1999 and 2014?RQ2: To what extent do age in the year under examination and life course stage (i.e.household type and marital status) explain attitudes towards consumption between 1999 and 2014 when other socio-demographic background variables are controlled for?
2009 and 2014, the differences had slightly decreased, although late middle-agers still had more ecological attitudes than young adults in 2014 (M=3.73 and M=3.48 respectively).[insertFigure1here] The factor scores of the attitudinal dimensions were tested in the ANOVA model in1999, 2004, 2009 and 2014, with  main effect tests for age-related variables and other socio-demographic determinants, separately in the years 1999, 2004, 2009 and 2014.The factor scores were included in the model as dependent variables.The overall statistical significances of the independent variables are indicated by the F values.The unstandardised parameter estimates (B) describe how much the means of the different categories of independent variables deviate from a reference category.The reference categories (0 a ) used in the model were selected on the basis of the lowest value or the reference category with the highest number of cases.The adjusted coefficients of determination (adjusted R²) show the proportions of variance explained by all independent variables together.In ecological attitudes towards consumption, age at the year of the study remained a significant determinant in 1999 and 2004 (F=10.556***,F=21.811***).In 2009 and 2014, age was no longer significant.The absence of children in the household predicted ecological attitudes in 2004 and 2009 (F=4.298,B=.092; F=10.409**,B=.240**).In 2014, age and the presence of children did not remain significant, and other sociodemographic variables predicted these attitudes better than age and life course stage.In the whole period, marital status was not significant.In all years under examination, ecological attitudes were more typical for female respondents (F=24.010***,B=.226***; F=92.580***,B=.338***; F=12.151**,B=.198**; F=5.405*, B=.146*, in order from earliest to latest year).High education (i.e.university degree) was a significant determinant between 2004 and 2014 (F=4.949**,B=.249*; F=7.097***, B=.270; F=2.663*, B=.283*, respectively).In 2004, ecological attitudes were most typical for respondents in a manager position (F=2.979*,B=.126*).Income level was associated with ecological attitudes in all years under examination.Ecological attitudes were least typical for respondents with high income (F=4.164**,B=-.348***; F=6.300***, B=-.267***; F=4.368**, B=-.320**; F=4.164**, B=-.348***, in order from lowest to highest income level).
This article is protected by copyright.All rights reserved.

Table 1 .
Attitudes towards ecological, economical and self-indulgent consumption,