### Implications of Demographic Change in Enlarged Eu on Patterns of Saving and Consumption and in Related Consumer’s Behaviour

**1. Introduction**

In December 2005 our working group was asked by the European Commission (DG Employment and Social Affairs) to conduct a research project on the effects of ageing on consumption and social expenditure.

The analysis of the effects of ageing on consumption patterns, as well as the effects on the provision of public goods and services, have been widely addressed by both theoretical and empirical literature, we are not aiming, with this work, at providing advancements of the literature, on the contrary, we will build on it to obtain a user friendly and fairly complete model providing consistent answers to policymakers.

The job was planned accordingly, and divided into different steps: first an analysis of Eurostat demographic projection; second the estimation of consumption-age profiles; third the parameterization and radical extension of our already existing projection model, in order to allow for consumption projections; and, finally the running of projections and the analysis of results.

There are many investment and insurance products that can play a role in saving for retirement, generating retirement income, and protecting retirement assets. Several of these products are referenced in the Pillars below. For purposes of this paper, any given product is shown in only one Pillar. In practice, many of these products span multiple Pillars.

**2. The Demographic Analysis**

**2.1 Forecast Assumptions for the Eurostat Population 2005-2051**

EUROSTAT has issued population forecasts for each EU country besides Bulgaria and Romania, by using the ‘cohort-component’ model. The forecast is based on the assumptions made on fertility, mortality and net migration perspectives. Three different assumptions result for each perspective — fertility, mortality and net migration — which are indicated as High or Base or Low, depending on the different levels the corresponding demographic indicators of the three components above mentioned will supposedly register.

In population forecasts, three alternative assumptions are often made, of which the central one (the Base one in this case) is the most probable. Implications emerging from the two other assumptions are still important because they underline (see next section) that in perspective only policies geared to favour and support in particular higher fertility rates, together with lower mortality — such as the ones outlined in the High assumption — will succeed in postponing the EU demographic decline.

As far as mortality is concerned, following standard practice in this field, the High assumption presupposes a wider reduction where the initial levels (2004) are higher and hence in general life expectancy at birth will grow more for males than for females, more in Eastern countries than in other EU countries. The two other assumptions also envisage a reduction in mortality that in both cases will obviously be lower when compared with the High assumption, and also in the Low case when compared with the Base assumption.

With regards to fertility, the High assumption predicts a general growth for each country; this growth is much weaker in the Base assumption and almost non-existent if not decreasing in the Low assumption.

Similarly to what has been said for mortality and fertility, for net migration too the assumptions point to different scenarios in many Eastern countries, where negative net migration would be recorded with a certain variability if compared to all other EU countries, which would be benefiting from positive net migration over the whole forecast timeframe.

As to the Base assumption (the one used in the present work), in particular, it has to be observed that the projections for mortality are totally shareable; this is not the case with fertility and, for different reasons, with net migration.

The fertility forecast per country does not cover the whole timeframe (1/1/2005, 1/1/2051), but shorter periods, after which the forecast proceeds on the constant fertility rates. To set a ‘target year’ — namely the year when the foreseen trend will cease — earlier than the forecast one means making the forecast itself less reliable over time. This timespan is very variable from country to country: in the Netherlands fertility has been constant since 2004; the longest forecast regards Poland where the foreseen trend will continue till 2038.

With reference to migrations we must observe that today’s differential both in economic development and demographic potential between South and North of the planet cannot lead us to think that there will be a decrease in the migratory flow pressure towards northern countries, particularly in the short and medium term; hypotheses that foresee decreasing net migration levels do not seem to be compatible with the present situation.

The assumptions produce analogies for the different countries regarding future trends in mortality, fertility and net migration. These analogies have been outlined through the statistical technique of cluster analysis which has led to a partition of the EU25 plus Bulgaria and Romania into five clusters, which differ considerably from one another, but are internally homogeneous from the point of view of the demographic issues resulting from High, Base and Low assumptions.

Further analysis of the implications of demographic change in an enlarged EU on patterns of saving and consumption and in related consumer behaviour has been conducted taking into account the most representative countries of each cluster, i.e. Bulgaria, France, Germany, Italy, Latvia, Luxembourg, Poland, Spain and United Kingdom; this subset makes up 72% of the population in the EU25 plus Romania and Bulgaria (population data refer to the 1 January 2004).

** 2.2 Baseline Variant of the Eurostat Population Forecast 2005-2051**

Once the three assumptions on future trends of mortality, fertility and net migration are defined, the population is forecast on the basis of plausible combinations of these assumptions; EUROSTAT has analysed six combinations indicated as Trend scenario variants.

Fertility is the main determinant for the future size and composition of the population compared to the effect produced by mortality and net migration. EUROSTAT has highlighted the role of the High assumption for fertility. Should this assumption occur, the future population decline will be avoided and an empirical check of this possibility is given by the three forecast scenarios: High Population, Younger Age Profile Population and High Fertility, all based on the High assumption. At the other end of the spectrum, the scenarios sharing the Low assumptions for fertility are two: Low Population and Old Age Profile Population, show a remarkable decrease of the population in the future. The sixth scenario1, called Baseline, proposes a median forecast compared to the previous ones and originates from a projection of the population in relation to the three Base assumptions for mortality, fertility and net migration. As we have already said, this is the scenario used by the economic analyses developed in the next chapters.

The sum of the results thus obtained points to sharp contrasts deriving both from diverging population trends and demographic structures differing from one another. The comparison provides food for thought because in some cases the issued forecasts can also be construed as a sensitivity analysis tool: as a matter of fact, the High Fertility and Baseline scenarios differ only for the assumptions on fertility, while Low Population and Old Age Profile Population only for the assumptions on mortality. EUROSTAT helps in understanding the extent and in confronting the gaps due to choices different from today’s as a consequence of a population policy aiming more at rebalancing the reproductive equilibrium — until now distorted by the parents’ strong inclination to have only one child — or, on the other hand, the results deriving from the hypothesis that the present imbalance might become even more marked.

In this study EUROSTAT has never taken into account any variants giving migrations some role in stopping or delaying population decline, as underlined by Low Population, Old Age Profile Population and Baseline scenarios; for example: no hypothesis has simulated a population trend combining the Low assumption for fertility with the High assumption for net migration.

Notwithstanding the expected increase of fertility levels as well as the future mortality reduction and the net migration contribution foreseen by the Baseline scenario, it will not be possible to stop the future decline of the EU population, an ongoing process that should continue in Eastern area countries of the EU10 and would take place in the second half of the forecast for almost all EU15 countries.

Fertility, mortality and net migration trends do not only have consequences on population size, but also on its composition by age. A further important result provided by the forecast is that the general reduction in the population will occur alongside a progressive decrease of the infant, youth and adult component and a steady increase of the old and very old component.

The first significant emerging fact is the reduction in the number of babies and very young children (0-2 and 3-5 years). Even if the Base assumption forecasts a certain recovery of fertility (i.e. the average number of children per woman), as shown in the previous section, birth decrease depends on the continuous reduction of women in fertile age, so that, despite the expected moderate increase in the inclination to have children, overall births will keep on declining.

In conjunction with decrease in the number of children, a reduction in the number of young people will also occur and, as years go by, adult age will range up to the most advanced age groups, so that in EU15 the only age group whose size will grow over the whole timeframe of the forecast is the 80 year olds and over: in the EU10 the outlined trends would show a more irregular pattern.

The trend described by the forecast simulates what in demography is known as population ageing process, depending on birth reduction or steadiness and on a simultaneous increase of old (aged 65 and over) and very old people. The forecast points out that ageing in Europe will continue, but with speeds differing from country to country, more so for some of the countries where today’s levels are lower, as they will experience an acceleration in ageing.

Over the forecast, the ratio between the old and the young, which gives a measure of how inter-generation relationships change, will tend to double in almost every country, as the proportion of over 65 year olds probably will. The proportion of people over 80 — the age range where handicap, disability, etc. are more frequent — would tend to triple almost everywhere by 2050. The median population age will grow by nine years on average, with higher rises in the countries where the ageing level is now lower. All these changes will have repercussions on the population dependency ratio whose growth will generally be lower than the corresponding ageing indicators (table 1).

The deep transformations that will change the age structure of the population show the strategic role played, in terms of relative weight, by the more adult age groups of active population, aged 40-64, because they will consist of healthier, more learned and more resourceful people, they will be more motivated to seize new opportunities, compared to their predecessors in previous decades.

All these results provide a picture of what could happen if mortality, fertility and net migration level remain in line with the levels forecast by the corresponding Base assumptions for every country and every year to come. However, there are no data^{2} to decide on the plausibility of these assumptions, nor from the complex of all the formulated assumptions (High, Base, Low) can we infer what is more likely to happen during the next few years. In this case also, the Base assumptions are considered more plausible because this is ‘standard’ practice in population forecasts.

That being said, experience teaches us that in this field the ‘validity’ timeframes of all assumptions (mortality, fertility and net migration assumptions) are generally shorter than the ones considered in the forecast. Assumptions on mortality have a long, slowly decreasing ‘resilience’, extraordinary events apart. Fertility assumptions last about 15 years, then their validity rapidly declines; even shorter is the reliability timeframe for international net migration, but, as we have seen, the impact of this component is more limited than the other two from a quantitative and structural point of view.

Table 1: Indicators of population structure by countries – Baseline variant

Source: Author’s calculations based on EUROSTAT data.

Note a)

*Table 1: Indicators of population structure by countries – Baseline variant
*

Source: Author’s calculations based on EUROSTAT data.

Note b)

2 EUROSTAT (2006): Statistics in focus, Marc.

Tags: consumption patterns, consumption projections, consumption-age profiles, EU demographic change, retirement savings retirement income, savings patterns