Erling Häggström Lundevaller

Test of random subject effects in heteroskedastic linear models

A new test of random subject effects in linear regression models is presented. The test is robust against heteroskedasticity and its asymptotic distribution is derived under a sequence of local alternatives. The finite sample properties of the test are studied in a simulation experiment and an empirical example. The results presented show that the new test is to be preferred over earlier test proposed.

Do unequal societies cause death among the elderly? A study of the health effects of inequality in Swedish municipalities in 2006

Background : A lively public and academic debate has highlighted the potential health risks of living in regions characterized by inequality. Research provides an ambiguous picture, however, with positive association between income equality and health mainly being found on higher levels of geographical division, such as nations, but rarely at local level. Methods : We examined the association between income inequality (using the Gini coefficient) and all-cause mortality in Swedish municipalities in the 65–74 age group. A multi-level analysis was applied and controlled for by variables including individual income and average income level in the municipality. The analyses were based on individual register data on all residents born between 1932 and 1941, outcomes were measured for the year 2006. Results: Lower individual income and lower average income in the municipality of residence were associated with significantly increased mortality. We found an association between income inequality and mortality with excessive deaths in unequal municipalities after controlling for mean income level and personal income. The results from the analysis of individual data differed substantially from the analysis of aggregate data. Conclusions: Income inequality was significantly associated with mortality in the age group 65–74 at municipality level. The association is small compared to many other variables, but is not negligible. Even in a comparatively equal society like Sweden, the potential effects of income inequality on mortality at the local level warrant consideration.BACKGROUND A lively public and academic debate has highlighted the potential health risks of living in regions characterized by inequality. Research provides an ambiguous picture, however, with positive association between income equality and health mainly being found on higher levels of geographical division, such as nations, but rarely at local level. METHODS We examined the association between income inequality (using the Gini coefficient) and all-cause mortality in Swedish municipalities in the 65-74 age group. A multi-level analysis was applied and controlled for by variables including individual income and average income level in the municipality. The analyses were based on individual register data on all residents born between 1932 and 1941, outcomes were measured for the year 2006. RESULTS Lower individual income and lower average income in the municipality of residence were associated with significantly increased mortality. We found an association between income inequality and mortality with excessive deaths in unequal municipalities after controlling for mean income level and personal income. The results from the analysis of individual data differed substantially from the analysis of aggregate data. CONCLUSIONS Income inequality was significantly associated with mortality in the age group 65-74 at municipality level. The association is small compared to many other variables, but is not negligible. Even in a comparatively equal society like Sweden, the potential effects of income inequality on mortality at the local level warrant consideration.

The Effect of Rh-Negative Disease on Perinatal Mortality: Some Evidence from the Skellefteå Region, Sweden, 1860–1900

The Rh-negative gene is a well-known cause of perinatal mortality. In this article, we analyze the possible role of Rh disease in perinatal mortality and stillbirths in a particular historical setting: the Skellefteå region in northern Sweden between 1860 and 1900. The data used for the study cover 23,067 children born to 4,943 women. The exact impact is not possible to establish using historical data, but the typical pattern of the disease allows us to make estimations. The expected levels based on knowledge of blood group distribution, the risk of sensitization from Rh incompatability, and the risk of perinatal mortality in births by sensitized mothers are compared with the observed levels. The results show that Rh disease was important for perinatal mortality and clustering of deaths within families.

Gendered death risks among disabled individuals in Sweden: A case study of the 19th-century Sundsvall region

This study follows around 500 disabled individuals over their lifespan to examine their risks of dying in 19th-century society, in comparison to a reference group of non-disabled people. The aim is to detect whether people, due to their disability, had a higher probability of meeting an untimely death. We use Sweden’s 19th-century parish registers to identify people the ministers defined as disabled, and to construct a reference group of individuals who were not affected by these disabilities. By combining the deviance theories from sociology studies with demographic sources and statistical methods, we achieve new insight into how life developed for disabled people in past societies. The results suggest that disability significantly jeopardized the survival of individuals, particularly men, but also that the type of disability had an impact. Altogether, we can demonstrate that the disabled constituted a disadvantaged but heterogeneous group of people whose demography and life courses must be further researched.

The Reationship between Engagement in Leisure Activities and Self-Rated Health in Later Life

ABSTRACT The aim of this study was to examine leisure engagement among people in later life and the potential relationship between leisure engagement and self-rated health. A population-representative sample of 5,435 persons between 65 and 80 years of age, living in northern Sweden and Finland were included. Data were collected by a posted questionnaire survey. Results revealed that levels of leisure engagement decreased progressively between the youngest and the oldest age groups. A significant relationship was found between leisure engagement and self-rated health. The relationship between leisure engagement and health as well as implications for developing health promotion programs are discussed.

Modeling domestic tourism in Sweden.

In this paper it is demonstrated how a microsimulation model based on TDB-data (Swedish Tourist Database provided by Marknadsfakta, Are, Swe¬den) can be used to estimate the number of trips, choice ...