Academic journal article Demographic Research

Sigma and Beta Convergence in Regional Mortality: A Case Study of the Netherlands

Academic journal article Demographic Research

Sigma and Beta Convergence in Regional Mortality: A Case Study of the Netherlands

Article excerpt

1.Introduction

The debate in mortality research about whether differences in mortality levels at the population level decline (converge) or increase (diverge) focuses on between-country differences (e.g., Meslé and Vallin 2002; Vallin and Meslé 2004; Kunst et al. 2004; Moser, Shkolnikov, and Leon 2005; McMichael et al. 2004; Mustard, Derksen, and Black 1999; Singh 2003; Boyle et al. 2004). It is important to widen the scope of this debate to include regional mortality trends, as Vallin and Meslé (2004) also suggested. To the extent that policymakers aim to reduce inequity across regions, it is important to know whether the mortality experience of regions is becoming more or less equal, as this provides essential information for the allocation of central governmental budgets to the different regions in a country. Valkonen (2001) also stressed the need for systematic studies on the trends in differential mortality, including mortality by region.

Convergence of health due to diminishing returns of increases in health expenditures, improvement of education, and economic development is to be expected (Gächter and Theurl 2011). By contrast, divergence may occur due to the 'Matthew effect': regions with high life expectancy may experience even faster improvements. The Matthew effect may occur due to differences in education, as highly educated people may have better access to health care and may benefit more from medical progress (Ben-Shlomo, White, and Marmot 1996; Morris, Sutton, and Gravelle 2005). Another cause of the Matthew effect may be differences in life styles, as unhealthy life styles may have cumulative effects, both within cohorts across the life course and between cohorts, as children from disadvantaged families may have poor health (Ross and Wu 1996; Rigney 2010). Furthermore, selective migration may contribute to divergence, as healthy people tend to move to regions with favourable living conditions (Bentham 1988; Valkonen 2001).

Recent studies show ambiguous results, even when restricted to overall mortality and to low-mortality countries. Both within the United Kingdom (Boyle, Exeter, and Flowerdew 2004; Leyland 2004; Dorling 1997; Leyland 2004; Shaw et al. 1999; 2004), between provinces in Canada (Mustard, Derksen, and Black 1999; Manuel and Hockin 2000), and for the 2,068 counties in the United States (Ezzati et al. 2008), a tendency from convergence in the past to divergence in the more recent past follows from the various studies; in New Zealand, recent divergence has been observed as well (Pearce and Dorling 2006). On the other hand, Gächter and Theurl (2011) observed continuing convergence in Austria. In addition, Valkonen (2001) observed clear differences between several European countries in the trend in regional differences. Finland, Sweden, France, Italy, Romania, and Russia showed a decline after 1970 in the range and the average deviation of life expectancy levels between regions, whereas Spain, Poland, and females in Austria and Denmark experienced an increase (Valkonen 2001). Montero-Granados, de Dios Jiménez, and Martín (2007) observed different results for different geographical scale levels in Spain.

In the above studies, different approaches, dispersion measures, and outcome measures were used to assess convergence. Earlier demographic and epidemiological studies focussed on studying trends in dispersion measures over time. In a few instances scatterplots with the end values versus the initial values were shown (Vallin and Meslé 2001; Caselli, Meslé, and Vallin 2002; Vallin and Meslé 2004). In economic literature, however, a clear distinction is being made between sigma convergence and beta convergence following the work by Barro (e.g., Barro and Sala-i-Martin 1990, 1992) and with recent applications to the study of health convergence (Nixon 2000; MonteroGranados, de Dios Jiménez, and Martín 2007; Gächter and Theurl 2011). Where sigma convergence concerns the formal study of trends over time in cross-sectional dispersion measures, beta convergence formally explores regression towards the mean of the different values over time (Barro and Sala-i-Martin 1992; Nixon 2000; MonteroGranados, de Dios Jiménez, and Martín 2007; Gächter and Theurl 2011). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.