Longevity of Southern European retiree by income and socio-demographic characteristics

Diego Ramiro-Fariñas, Consejo Superior de Investigaciones Científicas (CSIC)
Francisco Viciana, Institute of Statistics and Cartography of Andalusia
Víctor Montañés Cobo, Institute of Statistics and Cartography of Andalusia
Rosa Canovas, Institute of Statistics and Cartography of Andalusia
Margarita Montesó, Institute of Statistics and Cartography of Andalusia

Across Europe, governments are considering a major overhaul of public pension systems, because of doubts about their sustainability in a context of population aging and increasing government revenue reduction. Many of the arguments used to justify these reforms deal with the growth of life expectancy. However, there are few works which establish the possible relationship between the increase in the amount of the pension and increased life expectancy in southern Europe. This paper tries to quantify the combined effect on the longevity of income levels controlling by educational attainment and other socio-demographic and economic characteristics. We investigate the mortality experience of a cohort of 700,000 people over 65 years old living on the 1-1-2011, residing in Andalucía and receiving a lion share of their income from the Spanish social security public pensions system. We follow them to the end of 2012 using the Longitudinal Population Register of Andalusia which has the information of these individuals, pension levels reported by the Social Security and other socio- demographic characteristics derived from the population census. The results show the existence of a significant gradient in mortality associated with the amount of pensions for men. The gross effect of income is a reduction of about 1% in mortality for every increase of 100 € for the amount of the pension. The magnitude of the effect is halved, but remains significant (below 0.5 %), when controlling for other socio - economic and educational attainment. For women, the ratio of income to mortality, although statistically significant, is not as obvious or linear as in the case of men. The paradoxical effect of low female mortality associated with low levels of pensions is explained by the confounding effect of widowhood that influences both mortality risk and on the amount of pensions in the Andalusian female population.