Sven Drefahl is a researcher at the Demography Unit. He studies the relationship between family characteristics, migration and mortality. He is also interested in the mortality of the oldest-old and international trends of health and mortality. As project member of the "Register-based Research in Nordic Demography" his work is based on Swedish register data.
PublikationerI urval från Stockholms universitets publikationsdatabas
2014. Sven Drefahl, Anders Ahlbom, Karin Modig. PLoS ONE 9 (2), e88357
Background: In the beginning of the 1970s, Sweden was the country where both women and men enjoyed the world's longest life expectancy. While life expectancy continues to be high and increasing, Sweden has been losing ground in relation to other leading countries. Methods: We look at life expectancy over the years 1970-2008 for men and women. To assess the relative contributions of age, causes of death, and smoking we decompose differences in life expectancy between Sweden and two leading countries, Japan and France. This study is the first to use this decomposition method to observe how smoking related deaths contribute to life expectancy differences between countries. Results: Sweden has maintained very low mortality at young and working ages for both men and women compared to France and Japan. However, mortality at ages above 65 has become considerably higher in Sweden than in the other leading countries because the decrease has been faster in those countries. Different trends for circulatory diseases were the largest contributor to this development in both sexes but for women also cancer played a role. Mortality from neoplasms has been considerably low for Swedish men. Smoking attributable mortality plays a modest role for women, whereas it is substantially lower in Swedish men than in French and Japanese men. Conclusions: Sweden is losing ground in relation to other leading countries with respect to life expectancy because mortality at high ages improves more slowly than in the leading countries, especially due to trends in cardiovascular disease mortality. Trends in smoking rates may provide a partial explanation for the trends in women; however, it is not possible to isolate one single explanatory factor for why Sweden is losing ground.
2014. Gustavo De Santis, Sven Drefahl, Daniele Vignoli. Population 69 (3), 463-476
Hoem and Muresan (2011a) have recently shown that the most widely used macro-level indicator of fertility, the total fertility rate (TER), can be reconciled with fertility estimates that derive from applications of event history analysis (EHA) to micro-data. The purpose of this paper is to extend their ideas and show that they can be usefully applied to short panels, i.e. when the same people are interviewed in two or more successive rounds over a very limited number of years. This method can also be applied to data collected for general purposes and not strictly for demographic research, including data of an economic nature (employment, income, geographic or professional mobility, etc.). Despite the absence of questions on fertility, group-specific fertility estimates can be obtained that are not otherwise available (e.g. fertility by income level before the birth of the child), which are not biased by memory or selection of respondents and can be made consistent with the TFR observed in that period for the entire population. An application to Italian EU-SILC data in the years 2004-2007 highlights the advantages and the limitations of the method.
2012. Sven Drefahl. Journal of Marriage and Family 74 (3), 462-475
Numerous studies have shown that married women and men experience the lowest mortality. Legal marital status, however, does not necessarily reflect today's social reality because individuals are classified as never married, widowed, or divorced even when they are living with a partner. Denmark is one of the forerunners of developments in coresidential partnerships and one of only a few countries where administrative sources provide individual-level information on cohabitation for the whole population. Using register information from Statistics Denmark on 3,888,072 men and women ages 18–65, the author investigated mortality differences by living arrangement with hazard regression models. Overall, premature mortality was found to be lowest for married persons, followed by cohabiting persons. Adjusting for socioeconomic status reduced excess mortality of nonmarried individuals. Moreover, a mortality-crossover effect emerged in which cohabiters with above-average socioeconomic status had a lower risk of dying than married people. This finding was particularly pronounced for men.
2010. Sven Drefahl. Demography 47 (2), 313-326
I use hazard regression methods to examine how the age difference between spouses affects their survival. In many countries, the age difference between spouses at marriage has remained relatively stable for several decades. In Denmark, men are, on average, about three years older than the women they marry. Previous studies of the age gap between spouses with respect to mortality found that having a younger spouse is beneficial, while having an older spouse is detrimental for one's own survival. Most of the observed effects could not be explained satisfactorily until now, mainly because of methodological drawbacks and insufficiency of the data. The most common explanations refer to selection effects, caregiving in later life, and some positive psychological and sociological effects of having a younger spouse. The present study extends earlier work by using longitudinal Danish register data that include the entire history of key demographic events of the whole population from 1990 onward. Controlling for confounding factors such as education and wealth, results suggest that having a younger spouse is beneficial for men but detrimental for women, while having an older spouse is detrimental for both sexes.