A Multidimensional Approach to Intergenerational Balance Measurement: A Cross-Sectional Study for European Countries

Abstract

The demographic shift experienced by developed countries inevitably results in a change in intergenerational relations. However, despite some attempts to evaluate intergenerational balance in quantitative terms, there is still a significant literature gap in this respect. This paper aims to propose a conceptual and empirical framework to measure intergenerational balance in a cross-country perspective which can serve for the comparative assessment of the outcomes of various policies. It addresses the research question in which countries selected generations are privileged over others in socio-economic terms. Our empirical study includes the investigation of the (relative) situation of different generations as well as comparisons of gender differences in terms of studied welfare state performance across generations to examine whether gender equality in some generations is more promoted than in others. As a criterion, we employ multivariate statistical analysis methods to group 25 European countries into clusters using the intergenerational balance in terms of poverty, income, housing, labour market, education and health. We distinguish four patterns in this respect: ‘Supporting young’, ‘Supporting adult’, ‘Discriminating against elderly’ and ‘Supporting elderly’. Our findings reveal that shifting the perspective from inputs to outcomes and including gender perspective gives a somewhat different picture of Europe in terms of intergenerational balance.

Appendix

See Tables

Table 5 Eurostat indicators employed in the empirical analysis.

5,

Table 6 Variables included for synthetic indicators composition

6,

Table 7 Values of intergenenerational balance partial indicators—relative postion of given generation in comparison to other generations in the same country

7,

Table 8 Values of intergenenerational balance partial indicators—intragenerational gender differences

8 and

Table 9 Aggregate synthetic indicators of intergenerational balance

9.

Example 1

To illustrate this problem of the sensitivity to the demographic structure with reference to the measurement of relative position of each generation (ratios in Table 6: RPR, RSMD, RMI, RS80/S20, RHCO, RMHC, ROR, RUnemp, REdu, RUMN) and justify our assumption that the ratios should not account for the demographic differences between countries, we present below a simplified example, where we apply the formula expressing the relative position of a given population in terms of the poverty rate.

Let’s assume that there are two countries A and B. In both countries, the corresponding generations have the same poverty rates (PR) given in %, i.e.:

PR_{A <18} = PRB _<18 = 20.

PR_{A 18–64} = PR_{B 18–64} = 20.

PR_{A ≥65} = PR_{B ≥65} = 40.

We can observe that elderly generations in both countries have poverty rates two times greater than young and working-age generations. We can also see that in both countries the situation is precisely the same. Intuitively, we assess both countries as equal in terms of the poverty situation of each generation as compared to the remaining generations.

Now, when constructing the indicator of the relative position of a given population in terms of poverty rate, i.e. RPR, let us compare the two approaches:

First—sensitive to the age structure:

$$\begin{array}{*{20}c} {{\text{RPR}}_{{{\text{young}}}} = \frac{{{\text{PR}}_{ < 18} }}{{{\text{PR}}_{{{\text{total}}}} }}} &\quad {{\text{RPR}}_{{{\text{adult}}}} = \frac{{{\text{PR}}_{18 - 64} }}{{{\text{PR}}_{{{\text{total}}}} }}} &\quad {{\text{RPR}}_{{{\text{elderly}}}} = \frac{{{\text{PR}}_{ \ge 65} }}{{{\text{PR}}_{{{\text{total}}}} }}} \\ \end{array}$$

where PR_total is the poverty rate reported for the total population in the country.

Second—insensitive to the age structure:

$$\begin{array}{*{20}c} {{\text{RPR}}_{{{\text{young}}}} = \frac{{{\text{PR}}_{ < 18} }}{{{\text{PR}}_{{{\text{mean}}}} }}} &\quad {{\text{RPR}}_{{{\text{adult}}}} = \frac{{{\text{PR}}_{18 - 64} }}{{{\text{PR}}_{{{\text{mean}}}} }}} &\quad {{\text{RPR}}_{{{\text{elderly}}}} = \frac{{{\text{PR}}_{ \ge 65} }}{{{\text{PR}}_{{{\text{mean}}}} }}} \\ \end{array}$$

where PR_mean is obtained according to the formula (PR _< 18 + PR₁₈₋₆₄ + PR_≥ 65)/3.

Additionally, let’s assume that the age structure in country A is:

Proportion of young generation A = 10% (of the total population).

Proportion of working-age generation A = 80%

Proportion of elderly A = 10%

and in country B:

Proportion of young generation B = 25% (of the total population).

Proportion of working-age generation B = 25%

Proportion of elderly B = 50%

As a result of such different demographic structures, poverty rates for the total population PRtotal, in both countries will significantly vary as they are affected by the strong prevalence of different generations (in country A—working-age, in country B—elderly):

(to simplify, we calculate PR_total as PR average weighted by age structure) PR_{A total} = 22, whereas PR_{B total} = 30.

This way, in the first approach, we obtain:

For country A –

$$\begin{array}{*{20}c} {{\text{RPR}}_{{{\text{young}}}} = \frac{20}{{22}} = 0.90} &\quad {{\text{RPR}}_{{{\text{adult}}}} = \frac{20}{{22}} = 0.90} &\quad {{\text{RPR}}_{{{\text{elderly}}}} = \frac{40}{{22}} = 1.80} \\ \end{array}$$

For country B –

$$\begin{array}{*{20}c} {{\text{RPR}}_{{{\text{young}}}} = \frac{20}{{30}}0.66} &\quad {{\text{RPR}}_{{{\text{adult}}}} = \frac{20}{{30}} = 0.66} &\quad {{\text{RPR}}_{{{\text{elderly}}}} = \frac{40}{{30}}1.33} \\ \end{array}$$

Thus, if we apply the age-structure sensitive measurement, the indicators of the relative position reveal entirely different pictures of both countries. At the beginning of this example, we observed that equally in both countries elderly generation has the poverty rate two times greater than the remaining generations, but after including age structures, in country B, the situation of the elderly seems to be significantly better than in country A, which is a misleading conclusion. Consequently, we perceive the impact of the demographic structure as a bias of the analysis, and we employ the measurement free from this impact.

In the second approach, when we replace the denominator PR_total with PR_mean we will obtain precisely the same results for both countries, which adequately reflects the similarity of both countries:

For countries A and B –

$$\begin{array}{*{20}c} {{\text{RPR}}_{{{\text{young}}}} = \frac{20}{{27}} = 0.74} &\quad {{\text{RPR}}_{{{\text{adult}}}} = \frac{20}{{27}} = 0.74} &\quad {{\text{RPR}}_{{{\text{elderly}}}} = \frac{40}{{27}} = 1.48} \\ \end{array}$$