might be required in order to increase the ability of the social protection system to cover
retired individuals against the risk of poverty.
The cross sectional analysis of poverty incidence carried out in section 4 does not allow us
to answer to this question. A longitudinal analysis of the individual change of income
associated with the retirement transition is required in order to evaluate whether this
transition makes individuals more likely to become poor.
This is what we try to do in this section. We consider all the individuals that retired during
our sample period and we evaluated the incidence of poverty in the year prior to retirement
an after retirement, and also the proportion of non poor individuals in the year prior to
retirement that become poor after retirement. We do this for the whole population that
retires and for different groups in order to try to evaluate if there are some groups for which
retirement is more likely to imply a situation of poverty. Among the characteristics
considered are gender, early retirement, activity status in the year prior to retirement, and
the situation regarding property of the house.
It is clear from the data that, in general, retirement increases the chances of poverty.
Considering the 33rd percentile low income definition, Table 5-1 shows that the percentage
of persons that live with low income in the year before retirement is inferior to the
percentage of persons with low income in the retirement year.
Women are already poorer than men, on average, but they are not more penalized by
retirement. The difference between the percentages of poor women and of poor men
slightly decreases from the year before retirement to the year of retirement.
As expected, a large percentage of persons who live in social housing are poor, either
before (59%) or after (63,9%) retirement. Nevertheless, the percentage of persons living in
social housing who are not poor before but become poor in the retirement year is
impressing: 38,2%. Being a home-renter is not more typical of those who become poor
with retirement than being a home-owner. Nor is it more typical of those who are poor, in
general.
A Portuguese person is definitely more at risk of becoming poor in the year of retirement if
he/she lives in social housing, was previously an employee working less than 15 hours per
week or was self-employed. Remarkably, the percentage of persons who live with low
14