quality of life in older age: does the context matter? · 2018. 10. 9. · (pittau et al. 2009; di...

Annalisa Donno, Omar Paccagnella & M. Letizia Tanturri Department of Statistical Sciences

University of Padova

Quality of life in older age: does the context matter?

SECOND INTERNATIONAL SEMINAR ON THE ACTIVE AGEING INDEX

27–28 September 2018 Biskaia Aretoia, Bilbao, Spain

Rationale

• As societies age…. – the well-being of the elderly: a priority – active ageing: a new challenge.

• Both active ageing (as a macro-level element) and individual

well-being are complex and multifaceted concepts, influenced and determined by several dimensions

• Complexity has to be taken into account for a comprehensive analysis of the role played by macro-level factors in shaping the perception of individual well-being

Previous literature

• Most of research on active ageing stresses the importance of individual determinants of well-being as: – socioeconomic status, – health status, – participation in socially productive activities

• some scholars suggest that macro factors have also a role: – the welfare provisions and models (Niedzwiedz et al. 2014, Motel-Klingebiel et al. 2008,

Conde-Sala et al. 2017, Esser and Palme 2010), – the level of socio-economic inequality (Mikucka et al. 2017; Roth et al. 2017;

Niedzwieds et al. 2014),

– unemployment rate (Pittau et al. 2009; Di Tella et al. 2003),

– gender equality policies (Van Oyen et al. 2010, Högberg 2018, Palència et al., 2014),

– GDP (Degutis et al. 2010).

Research gaps

• in the quoted studies only single macro dimensions are generally considered,

• a research linking the multiple macro-dimensions of active

ageing to the individual level of well-being is lacking. • The element of novelty of this paper :

The use of composite measures, both at the micro and macro-level

to put individuals in a context

Research question 1

• Do older people living in contexts characterized by high levels of active ageing (measured through the Active Ageing Index) show a higher level of individual well-being, other things been equal?

Research question 2

• In a dynamic perspective a change in the level of active ageing over time could influence the perceived quality of life.

• Do changes in macro-level degree of active

ageing (measured as changes in the AAI levels) play a role in explaining the way older people perceive their well-being?

Research question 3

• ageing and ageing policies are not gender neutral.

• Do AAI levels have a different impact on individual well-being depending on their gender?

• Do men’s AAI levels have an impact on individual well-being of old women, and vice versa?

Macro-level DATA

Micro-level DATA

• The sixth wave of the Survey of Health, Ageing and Retirement in Europe (SHARE 2015) (Börsch-Supan 2017)

• We select a sample composed by 59.267 individuals, – aged 50 and over – living in 16 European countries

• Austria, Belgium, Croatia, Czech Republic, Denmark, Estonia, France, Germany, Greece, Italy, Luxembourg, Poland, Portugal, Slovenia, Spain and Sweden.

C.A.S.P.-12

Control

The dependent variable

Self-realization Autonomy Pleasure

1 2 3 4

3 items 3 items 3 items 3 items

Often Sometimes Rarely Never

i.e. (Control) My age prevents me from doing the things I would like to? i.e. (Autonomy) Shortage of money stops me from doing the things I want to?

C.A.S.P.

CASP-12 to measure individual well-being

All Men Women Mean 36.97 37.45 36.61 St. deviation ± 6.35 Median 38

12 < < 48

The higher the index the higher the well-being

In our sample:

Highest level of well-being

Lowest level of well-being

CASP cross-country variability 0

.05.1

.150

.05.1

.150

.05.1

.150

.05.1

.15

10 20 30 40 50 10 20 30 40 50 10 20 30 40 50 10 20 30 40 50

Belgium Czech Republic Denmark Germany

Estonia Greece Spain France

Italy Luxembourg Austria Poland

Portugal Slovenia Sweden Croatia

Dens

ity

CASP index for quality of life and well-beingGraphs by country

C.A.S.P. 31.84 in Greece < < 41.37 in Denmark

Methods & Models

Multilevel Analysis: we estimate random slope models.

C.A.S.P. -12

• Global Active ageing Index

• Global AAI change

• Men’s AAI • Women’s AAI • Single AAI

domain

II Level variables

•Demographics •SES •Physical and Mental Status •Cognitive abilities

•Social and kin network

Individual Characteristics

Individual well-being

Active ageing at societal level

AAI variable

Model 0

Global AAI

Global AAI variation Interclass

Correlation Coefficient

15.12% 15.18%

*** = 1%; ** = 5%; * = 10%

No, second level variables

Around 15% of the variability in the individual’s well-

being explained by cross-country

differences

Controlling for individuals characteristics

AAI variable

Model 0

Model I

Men Women

Global AAI 0.234 ***

0.268 ***

Global AAI variation

ICC 15.12% 15.18% 10.48% 9.24%

*** = 1%; ** = 5%; * = 10%


cross-country variability is reduced

( > for women )

AAI variable

Model 0

Model I

Men Women


0.268 ***


ICC 15.12% 15.18% 10.48% 9.24%

*** = 1%; ** = 5%; * = 10%


An increase of Global AAI

correlates with an increase of

wellebeing

AAI variable

Model 0

Model I Model II

Men Women Men Women


0.268 *** 0.120 0.153 *


2.032 ** 2.043 **

ICC 15.12% 15.18% 10.48% 9.24% 8.18% 6.93%

*** = 1%; ** = 5%; * = 10%


Further reduction of cross-country

variability

AAI variable

Model 0

Model I Model II

Men Women Men Women


0.268 *** 0.120 0.153 *


2.032 ** 2.043 **

ICC 15.12% 15.18% 10.48% 9.24% 8.18% 6.93%

*** = 1%; ** = 5%; * = 10%


AAI variations seems even more

correlated to individual well-

being.

AAI variable

Model 0

Model I Model II

Men Women Men Women


0.268 *** 0.120 0.153 *


2.032 ** 2.043 **

ICC 15.12% 15.18% 10.48% 9.24% 8.18% 6.93%

*** = 1%; ** = 5%; * = 10%


The level of AAI remains

significant only for women

AAI variable

Model III

Model IV

Men Women Men Women

Men AAI 0.254 *** 0.129 *

Men AAI var.

2.616 ***

Women AAI

0.240 *** 0.184 **

Women AAI

variation 0.992

ICC 10.27% 9.78% 5.57% 9.23%

Men’s well-being more significantly correlated to variation of men’S AAI

Women’s well-being more significantly

correlated to the level of

women’s AAI

AAI variable

Model III

Model IV Model V

Men Women Men Women Men Women

Men AAI 0.254 *** 0.129 * 0.164 ***

Men AAI var.

2.616 *** 2.644 ***

Women AAI

0.240 *** 0.184 ** 0.153

Women AAI

variation 0.992 0.994

ICC 10.27% 9.78% 5.57% 9.23% 10.36% 4.25%

Women’s well-being is

influenced by both men’s AAI

level and variation

The opposite is

not true

The AAI’s domains

“Independent, health and secure living”, is the only AAI domain significantly and strongly correlated with both men and women’s well-being

AAI variable

Model VI

Men Women

Global AAI employment

-0.012 0.015

Global AAI participation

0.082 0.069

Global AAI independence

0.416 *** 0.363 ***

Global AAI capacity

-0.115 -0.071

ICC 5.12% 5.55%

Conclusions

• Results corroborate the hypothesis that the active ageing context matters and cannot be ignored in the analysis of individual well-being.

• living in a context fostering active ageing is positively correlated with high level of individual well-being other things being equal

• The AAI and their variation prove to be good indicators to capture unexplained cross-country variability – level is almost halved when we consider Global AAI and its

variation for both men and women – Level is reduced to a third when we consider different domains

Conclusions

• Policy makers should be encouraged to invest in active ageing, not only to face population ageing at societal level, but also because it is positively correlated with individual level of well-being, ceteris paribus.

• any positive variation of the AAI on the whole is

linked to a higher level of individual’s well-being. • In this sense, any further improvement in active ageing

could be beneficial on the whole for old men and women.

Conclusions

• However, when we try to isolate the correlation between each single dimension of active ageing and individual well-being, the only domain that matters is: – “Independent, health and secure living”, – for both men and women.

Suggestions for policy makers

• Each single dimension of this domain should be specifically supported with targeted policies.

• we suggest the UN to evaluate the

possibility of increasing the explicit weight for the above domain (just 10%) given its robust correlation with old people’s self-rated level of well-being.

Gendered active ageing policies?

• Men’s quality of life seems affected more by variation in men’s employment level, and a public investment in this direction would be beneficial also for women.

• While women’s wellbeing is sensitive to

variations of more domains, therefore policy makers should support many aspects of active ageing to improve women’s quality of life.

Thanks a lot!

COMMENTS and SUGGESTIONS are particularly WELCOME!!!

[email protected]

https://crew-more-years-better-lives.org

https://crew-more-years-better-lives.org/










quality of life in older age: does the context matter? · 2018. 10. 9. · (pittau et al. 2009; di...

Documents