California Center for Population Research

The economic crisis at the beginning of the eighties impacted 20th-century Costa Rica in

many different ways. Government programs were successful in reducing the proportion

of poor people from 35% in 1985 to 23% in 2000. This article utilizes official Household

Surveys corresponding to the period 1981-2002 and Population Censuses to perform

an estimation of age, cohort, and period effects, in order to show that poverty among

the elderly can be understood as a cohort effect. Their poverty conditions are

associated with their low schooling, mainly among men. The probability of having the

right to a retirement pension and the fact that individuals with lower educational levels

earned low income during their later working years, intervene in the relationship

between schooling and poverty as a cohort effect.

Soon after the end of the civil war in 1948, Costa Rica underwent political and economic

changes that aided the expansion of the social benefits initialized in the 40’s, such as

universalizing Social Security, promoting health and educational policy, and providing

infrastructure and services to both rural and urban communities. Nevertheless, at the

end of the 70’s, increasing oil prices had a strong negative impact on many countries

around the world. Costa Rica was no exception, and during the 1980 – 1982 period it

went through an economic crisis characterized by hyperinflation, increased

unemployment and underemployment rates, and the declaration of a moratorium on foreign debt payments (Barahona Montero 1999a). The governments after 1982 were

relatively successful in promoting economic recovery by changing the development

model based on import substitution to one promoting non-traditional product exports

and tourism (Barahona Montero 1999a, González Mejía 1999). In spite of the recovery

and public policies designed to combat poverty, since 1991, it has not been possible to

reduce the proportion of poor households below 18%, maintaining an annual average of

around 20%.2 In addition, economists consider that within this one-fifth of the nation’s

population, there is a group of “hard-core poor”, i.e., persons who systematically live on

a minimal income, and who cannot easily climb out of their chronic condition (Proyecto

Estado de la Nación-PEN 2002).

One group in which the proportion of poor is relatively high with regard to the rest of the

population is that of the elderly: 31% of individuals 65 or more years of age live in

households with incomes below the poverty line, according to the 2002 Encuesta de

Hogares (National Household Survey). The objective of this paper is to show that if

poverty among the elderly is associated with structural characteristics within this group

of the population, its incidence can be represented as a cohort effect. In other words, a

large percentage of the Costa Rican elderly would be living in poverty, not because old

age leads to poverty, but because the characteristics that they acquired throughout their

lives – given the historical moments they lived – make them more susceptible to being

poor, in comparison with other groups born more recently. In order to provide separate

estimates of the cohort, age, and period effects, we consolidated the Encuestas de

Hogares from 1980 to 2002, whose basic dependent variable is the proportion of

persons living in poor households. The paper presents variations of these effects

produced by the inclusion of certain independent variables in the model, in particular the

level of education of the cohorts. It also emphasizes the importance of the economic

crisis at the beginning of the 80’s on the incidence of poverty among the elderly, and at

the end of the paper, we relate this phenomenon to educational levels among these

generations and their access to Social Security.