Is the relationship between insurance consumption and its determinants spurious? Is general insurance a luxury service? Do bequest motives matter for life insurance consumption? Is private credit important for the development of life insurance? Do socioeconomic development and informal risk sharing institutions matter for formal insurance consumption? This thesis investigates these and other related issues using international datasets and relatively new panel data method, namely the Common Correlated Effects Pooled (CCEP) estimator. A novelty of the CCEP is that it takes into account the impacts of unobserved common factors. The thesis consists of an introduction, three empirical chapters and conclusions. Chapter 2 studies the relationship between nonlife insurance consumption and income/wealth per capita. Estimation results suggest that income elasticity is below unity and that nonlife insurance is positively related to GDP per capita, the law, risk aversion, infrastructural development, and negatively related to socioeconomic development. Chapter 3 explores life insurance consumption driven by bequest motives. We found that life insurance consumption is positively related to GDP per capita, old age dependency ratio, infrastructural development, and social security and welfare; and negatively related to the extended family institution, savings, inflation, and risk aversion. Estimation results suggest the presence of altruistic, and bequest as exchange old age security motives. Chapter 4 investigates the long run relationship and causality direction between private credit consumption and life insurance development. Life insurance development may be explained by GDP per capita, formal and informal credit consumption, infrastructural development, life expectancy, institutional quality, inflation, and Islam, and Orthodox being the dominant religions. Cointegration test results suggest that life and nonlife insurance consumption and its determinants exhibit a long run relationship; and that there is a long run bi-directional causality relationship between life insurance development and private credit consumption. The thesis concludes that insurance development requires institutional and infrastructural development-in particular- telecommunications infrastructure, to facilitate cost effective insurance supply.

This research examines intrahousehold resource allocation in rural Ethiopia with a focus on education and asset investments in children. Throughout our work we look for potential causes and signs of gender bias in childhood investments. In particular, we first look at whether poverty constrained households substitute the provision of asset and education transfers to their children. For this purpose we use a two-stages-least-squares model of educational demand which controls for the endogeneity of asset transfers. We find a trade-off mechanism between endowments of human capital and bequests to girls but not to boys. Secondly, we investigate the effect of parental transfers on the postmarital residence of children. We look at whether transfers induce old age security from children in the form of a postmarital residence close to the parents. We model postmarital residence using a multinomial logit which controls for the endogeneity of transfers. Our results show that off-springs who receive more assets are more likely to live close to their parents after marriage. On the contrary, more education induces children to leave their home at marriage. Finally, we analyze household choices between education and farm employment of children. We model these choices jointly while looking at how ability affects them. Our findings suggest that in the long-run, more able children accumulate more schooling and less farming experience. However, the latter only applies to boys; farming decisions regarding girls are independent of their ability. In summary, our research addresses four important factors affecting the welfare of children and young people in rural Ethiopia: educational investments, asset transfers, old age security mechanisms and child work decisions.

This study investigates the use of response guidance for errorless learning of a perceptual motor task in normal elderly. It provides normative data for a study with stroke patients using this technique for cognitive rehabilitation. While errorless learning has been shown to be more effective on most tasks than trial and error learning for people with memory impairments, its use with normal individuals has received limited attention. The questions of interest were whether errorless training of the perceptual motor task was more effective for improving and retaining accuracy; and whether both accuracy and response speed were more resistant to the effects of increased cognitive demands. A sample of 43 normal elderly in the United Kingdom, ranging in age from 60 to 77, completed an assessment of intelligence, memory, and attention. They then received training, over two sessions one week apart, to mark the midpoint of Judd Arrows presented on a computer screen using a cross cursor moved by an active force feedback joystick (AFF). During training the errorless group received AFF guidance to the correct midpoint, while the errorful group received none, and both received auditory and visual knowledge of results. There was no AFF during baseline or post test measures. Training was to criterion in each session with a discontinue rule if accuracy did not improve. At the end of session two both groups were given a cognitively challenging task concurrent with the arrow bisection. Results revealed that both groups improved their accuracy through training with the errorless group being significantly more accurate and tending to be faster in the final post tests of both sessions. The errorless group was significantly faster than the errorful group under the cognitive challenge, without sacrificing accuracy. These results suggest not only that AFF is an effective means of implementing errorless perceptual motor learning, but elderly individuals trained in this manner do not sacrifice accuracy for speed. Implications of these results are discussed.

The aim of this thesis was to analyse coexisting disadvantages in the older Swedish population. Coexisting disadvantages are those that occur simultaneously in various life domains. A person who simultaneously experiences several disadvantages may be particularly vulnerable and less well-equipped to manage daily life and may also need support from several different welfare service providers. Concerted actions may be needed for older people who experience not only physical health problems and functional limitations, but also other problems. Research that encompasses a wide range of living conditions provides a basis for setting political priorities and making political decisions. The studies in this thesis used data from two Swedish nationally representative surveys: the Level of Living Survey, which includes people aged 18 through 75, and the Swedish Panel Study of Living Conditions of the Oldest Old, which includes people aged 77 and older. Study I showed that the probability of experiencing coexisting disadvantages was higher in people 77 and older than in those aged 18 through 76. These age differences were partly driven by a high prevalence of physical health problems in older people. In all age groups, coexisting disadvantages were more common in women than men. The longitudinal analyses in Study II indicated that coexisting disadvantages in old age persist in some people but are temporary in others. Moreover, the results suggested a pattern of accumulating disadvantages: reporting one disadvantage in young old age (in particular, psychological health problems) increased the probability of reporting coexisting disadvantages in late old age.   Study III showed that physical health problems were a central component of coexisting disadvantages. The results also showed that being older; female; previously employed as a manual labourer; and divorced/separated, widowed or never married were associated with an increased probability of experiencing coexisting disadvantages. However, the experience of coexisting disadvantages differed: the groups associated with coexisting disadvantages tended to report different combinations of disadvantage. Study IV showed that the prevalence of coexisting disadvantages in those 77 and older increased slightly between 1992 and 2011. Physical health problems became more common over time, whereas limited ability to manage daily activities (ADL limitations), limited financial resources and limited political resources became less common. Associations between different disadvantages were found in all survey years, but certain associations changed over time. The results suggest that in general, the composition of coexisting disadvantages in the older population may have altered over time. In sum, results showed that coexisting disadvantages were associated with specific demographic and socio-economic groups. Physical health problems and psychological health problems were of particular importance to the accumulation and coexistence of disadvantages in old age. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: Manuscript. Paper 4: Manuscript.</p><p> </p>

Amidst widespread concerns about aging on several levels ranging from the personal to the societal, this dissertation examines the construction of meaning in life and older age in contemporary Japan. Based on an ethnographic account of a community salon in Southern Osaka, it explores the experiences of older people and their ideas of the good and meaningful life, while arguing that than an anthropology of the elderly can reveal a far wider scope of issues than aging alone. Drawing on a socio-narratological approach, I show how stories connect people, form a shared body of knowledge, inform our understanding of the everyday, and provide frameworks for our choices. I argue that the capacity of narratives to create coherence and make sense of seemingly random and unconnected events can help to reveal existential issues, and that narrative analysis may therefore be a powerful tool for creating an existential anthropology capable of elucidating and understanding deeply personal dilemmas in their social and cultural context. The ethnography and life stories of elderly salon goers, volunteers and others involved in a local Non-Profit Organisation raise important issues of autonomy and dependence, sociality and isolation, care and concern. People express concern for others through practices ranging from gift-giving, visiting, balanced forms of polite yet friendly discourse, the provision of information, and volunteering in the salon and beyond. I argue that older Japanese are as much providers of care as recipients of it, thereby challenging the constructed image of the elderly as frail and dependent, even though maintaining independence relies paradoxically on cultivating multiple dependencies on others. Navigating the tensions between the benefits of rich social ties and a desired level of separation in which the burden imposed is minimised, or between dependence and freedom, emerges as central to the balancing acts required for living well.

The aim of the research presented in this thesis was to improve the characterisation of the changes in brain structure and function that occur at different stages of Alzheimer’s disease (AD) progression, from pre-symptomatic AD, to mild cognitive impairment (MCI), to clinically evident dementia, using magnetic resonance imaging (MRI) techniques. Baseline structural MRI data from a cohort of healthy older adults who were followed prospectively for ten years, during which time some developed MCI and some AD, were analysed. It was found that structural MRI could detect volume loss in medial-temporal lobes up to 7-10 years before clinical symptoms of AD appear. In addition, volumetric variability of medial-temporal regions detected by structural MRI across cognitively healthy older adults correlated with their performance on a task of visuospatial associative memory, and functional activation of the same regions occurred during successful performance of the same task on functional MRI (fMRI). Three groups of participants - cognitively healthy controls, people with MCI, and patients with probable AD - were then recruited and underwent a multimodal MRI protocol, which included functional sequences acquired at rest and during the execution of two different cognitive tasks (visuospatial associative memory and self-appraisal). Cross-sectional comparisons showed: (i) that successful visuospatial associative memory performance was associated with increased functional activity (measured with task fMRI) in lateral prefrontal regions in AD patients relative to controls and (ii) that increased functional activity overlapped with frontal brain networks showing increased functional connectivity (measured with resting fMRI) in the same AD patients. Further, by demonstrating group- and condition-specific decreased frontal activity in AD patients relative to controls during a self-appraisal fMRI task, it was shown the specific utility of fMRI to unravel cognitive mechanisms underlying specific neuropsychological symptoms such as unawareness of cognitive impairment (anosognosia) in MCI and AD. In conclusion, structural MRI can detect morphological changes in the preclinical stage of AD, possibly earlier than previously described, and these reliably match cognitive functioning in older adults. In the MCI and AD stages, once symptoms of cognitive impairment are clinically evident and measurable, task-related and resting functional MRI can inform on residual brain function detectable over and above the known changes in brain morphology and cognitive performance that have already occurred at these stages, emerging as a sensitive marker of residual ability that could potentially be used to measure the effect of new treatments.

The purpose of this study was to construct and use an instrument to assess the adjustment of aged residents in an institution.

Although institutional care has started to be outmoded in the developed countries with development of different models of care, it still has a considerable place in the developing countries such as Turkey. This is because, changes in the demographic structure, extended family, and urban development of Turkey has brought about several aging problems leading older adults to end up in institutions. Loneliness was one of the significant reasons given in the Social Inquiry Survey of Applicants of Darulaceze Old-Age Institution and the basis for a micro level analysis in this study. Therefore, the main objective of the study was to determine the predictors of loneliness, including age, the state of living alone, functional independence, education, and gender. Analysis of the results indicated that these predictors have significant effects on the loneliness predominantly defined by social factors rather than medical factors. In addition, the meso and macro level analyses were employed to control the micro level analysis and see a general picture of institutional care. Thus, an academic example of diagnosing the main reasons behind the institutional care was presented to understand the context of aging in Turkey.

This dissertation breaks new ground. It examines the perceptions of older adults towards AIDS prevention. Using the National Health Interview Survey, 1988: AIDS Knowledge and Attitudes Supplement, a modified Health Belief Model is developed. Despite the low number of older adults 55+ with AIDS, some extenuating circumstances increase their risk of AIDS contraction. Older adults have lower levels of knowledge about AIDS, weaker immune systems and receive more blood transfusions. Societal influences include educational neglect at the hands of physicians, healthcare workers and social service personnel. The first stage of the dissertation involved establishing older adults as an at-risk population through an extensive literature review. Next, the data was described utilizing frequencies, correlations and factor analysis. Frequencies clearly indicated that older adults in the data set had low levels of AIDS knowledge and did not view themselves at risk for AIDS contraction. Correlations between the variables were minimal. A modified Health Belief Model was developed and tested. Multiple regression determined that minimal variation in the two dependent variables, "Perceived Effectiveness of Effective Methods to Prevent AIDS Contraction" and "Perceived Effectiveness of Ineffective Methods to Prevent AIDS Contraction" was accounted for by the independent variables. Although F ratios allowed rejection of the two null hypotheses, beta weights were low. Adjusted R^2's accounted for only 21% and 16% respectively of the variation in the dependent variables. Finally, discrepancies in the model were determined and recommendations made for further research. Most health belief models concentrate on individual social-psychological variables. Due to AIDS' societal consequences, it is proposed that societal providers of education: physicians, social service workers and healthcare personnel need to be included in the model. Recommendations were made for additional research into sexual behavior of older adults and exploration of available training of physicians, healthcare and social service professionals. Finally, recommendations were made to provide training and education for both professionals as well as the elderly to prevent their growth into an at-risk population.

This study examined the relationship between perceptions of intellectual functioning and measures of cognitive abilities, personality variables and sociodemographic information. One hundred and fifty-two older community residing adults were asked to define their perception of intelligence by completing a questionnaire that asked the extent to which a variety of tasks are: functionally important, contribute to feelings of intellectual vitality and are the object of worry or concern. They also estimated their skill at performing each task. The hypothesis that cognitive abilities would best predict perceptions of cognitive functioning was moderately supported. Personality variables, specifically anxiety, were more predictive of the meaning variables than abilities.