Sequencing the real time of the elderly: Evidence from South Africa

Abstract

1. IntroductionWith an increase in the life expectancy of older adults, a growing body of research has developed in recent decades exploring how the elderly spend their time. In describing the nature, determinants, and consequences of time use among the elderly, studies consistently find evidence of gender differences in time allocation and, particularly, a persistent gender division in housework (cf. Bengtson, Rosenthal, and Burton 1990; Szinovacz 2000; Gauthier and Smeeding 2003; Solomon, Acock, and Walker 2004; Leopold and Skopek 2014).Much of the research that has investigated the time use of the elderly has been conducted in developed-country contexts where there is little variation in household form, and where older adults often live in single-generation households. In this study we investigate the time allocations of the elderly in South Africa, a developing country with a plurality of family types, where many of the elderly live in multi-generational households.We analyse data collected in the 2010 Time Use Survey, a nationally representative household survey which collected a daily time diary from approximately 39,000 individuals in South Africa. This time data is typically analysed using a mean added (or total) time approach. This is a useful and convenient method to describe the daily use of time and how this varies, for example, by gender and age. However, in this study we augment this method by using sequence and cluster data methods to analyse the real time of the elderly. These methods allow us to create a typology of time use behaviour among the elderly, based not only on the total amount of time allocated to different activities but also on the series of activities undertaken during the day, and on their timing and ordering.We focus on gender differences in the daily activities of the elderly in South Africa, but we also consider how time use patterns vary according to other characteristics of the elderly, including the composition and economic status of their household. In the next section we review the South African context, and in section 3 we describe the data that we analyse and the sequence and cluster methods which we use in our analysis. In sections 4, 5, and 6 we present and discuss our empirical analysis, and in the final section we conclude with a discussion of our main findings.2. Context2.1 Postretirement use of timeUntil a few decades ago, retirement was assumed to be the period of rest in an individual's life course, where the elderly would slowly disengage from society (Boudiny 2013). However, with an increase in life expectancy, particularly in developed countries, time spent in retirement has increased considerably, and a large body of literature has developed which investigates how the elderly spend their time. This work is motivated, in part, by the consequences of population ageing for the health care system (for example), and for demand on the public purse more generally.In the late 1990s the World Health Organisation adopted the concept of active ageing (Walker 2002; Boudiny 2013), which can be broadly defined to include not only paid employment and physical activity but also leisure activities that require mental (and not necessarily physical) effort or that involve social interaction, as well as education, participation in community life, for example, through volunteering work, and active engagement in household work and the care of others (Boudiny 2013). Active ageing is seen as a key ingredient to successful ageing and is associated with higher levels of life satisfaction and good health among the elderly (Gauthier and Smeeding 2003).A particular focus of this literature is whether and how patterns of time use and activity vary by gender. Studies have consistently found that among working-age adults, women spend more time on housework and on the provision of care (to children and kin) than men (Baxter 1997; Gerstel 2000; Bianchi, Robinson, and Milkie 2006). …