Saving and assets

The late nineteenth century and the first half of the twentieth century in France, Europe and to a lesser extent the United States was a unique period in terms of the enormous increase in public spending there. A considerable share of that increase could be attributed to the extremely rapid implementation - despite some countries being more gradual than others - of universal pension systems. Pensions are the type of welfare that has probably had the most far-reaching consequences. However, the concrete modalities of the appearance of pension systems are still under-researched, as are the consequences of their rapid development, which are the subject of heated debate between researchers who assert that the emergence of saving behaviour and decisions about when to retire preceded the invention and rapid expansion of pension systems, and those who maintain that industrialisation led to mass pauperisation of the elderly. The studies conducted at INED on France in the nineteenth and early twentieth centuries take the second perspective, by investigating the financial situation of the elderly during that period. They show simultaneously the extreme vulnerability of people aged over 60 - less than half of whom have sufficient savings for their old age - and the advantage enjoyed by those who receive a pension. Those results support the idea that the elderly had no or few opportunities to provide for themselves. Current research focuses on two directions: first on developing the study of pension systems and describing the conditions of their appearance and secondly on analysing the interaction between retirement pensions and another major form of insurance for old age: family. The issue here is at once to identify the motivations behind the adoption and generalisation of pension systems and to clarify how that development influences individual behaviour.

Little research has been done in France on saving behaviour over the life cycle and in particular on possible disaccumulation at the oldest ages. Basic life cycle theory holds that households accumulate wealth until retirement then disaccumulate wealth to zero by the time they die, in order to maintain the same level of consumption despite the drop in income induced by retirement. The data do not support this basic theory. Other models have therefore attempted to understand the reasons for this, by introducing the desire to give money to descendants, uncertainty about lifespan and precautionary saving (for spending on healthcare and dependency). Pursuing that research, studies conducted in France, using a decomposition by age, period and cohort, find there is no disaccumulation during retirement. For that purpose, the surveys on financial investments from 1986 and 1991-1992 and the surveys on assets from 1997-1998, 2003-2004 and 2009-2010 were harmonised to form a pseudo-panel data set. The aim was also to distinguish between real estate and financial assets in total assets, because each responds to a different logic of accumulation. Lastly, special attention is given to differences between cohorts through the identification of cohort effects.

The determinants of pay gaps between men and women have been much researched. Pension gaps are now well documented. Asset gaps have received far less attention in the literature. One of the main reasons is a lack of data, since surveys often measure assets at the level of the household rather than of the individual. Yet studying assets at individual level is important because it fulfils various functions that can affect individual welfare. Assets have utility (providing housing), generate income or can absorb some resource shocks (widowhood, health, etc.). The research work performed at INED on the basis of INSEE’s assets surveys has involved highlighting a wealth gap between men and women, quantifying it and investigating its determinants. Can the gap be attributed to differences in the labour market and therefore in income, to differentiated saving behaviour, or to family structure?