SPOUSAL INFLUENCE ON EARLY RETIREMENT DECISIONS: ORIGINS AND MECHANISMS

Abstract The interdependence between partners raises considerable interest in the sociology of life course, work and families. Partner influences play a particularly important role in the work domain, because each partner’s work decisions have profound effects on the couple as a whole. In contrast to previous research, this study pays detailed attention to the role partners play in workers’ labor market decisions by using the case of early retirement decisions. We hypothesize that partners’ preferences for older workers’ retirement originate from altruism and self-interest. For example, partners might prefer workers to retire early because the worker’s job is highly stressful or partners might prefer workers to retire early to increase possibilities for joint leisure. Moreover, we expected that partners influence older workers’ early retirement behavior via persuasion and pressure. So, partners might either convince workers to change their preferences, or they might pressure workers to act according to the partners’ preferences, irrespective of workers’ own preferences. To adequately estimate partners’ and workers’ preferences for workers’ early retirement, we used an instrumental variable approach. This was possible due to the multi-actor longitudinal data available from a large representative sample of older workers and their partners in the Netherlands. The results support that partners’ preferences originate in altruism and self-interest and that partners influence workers through persuasion and pressure. Gender differences were marginal, with stronger signs for altruistic origins among female than male partners.

assessments of cognition, depression, personality; surveys of health-related quality of life, medical conditions, number of friends and relatives contacted at least monthly. Of 1037 participants aged 70 to 90 years surveyed at baseline 476 provided network data at all four time-points. Just over half of the participants (n=252, 52.9%) had smaller networks at wave 4 compared to baseline; nearly half of participants had stable (n=65, 13.7%) or larger networks (n=159, 33.4%), some doubling in size (n=65, 13.7%). Chi-square test indicated that more males (59.1%) had decreasing networks [X2(1, 476)=2.26, p=.02, phi=.11]. T-tests indicated: group differences in baseline neuroticism [t(276)=2.31, p=.02, Eta2=.02]; no significant difference in age, education, wave 4 global cognition, depression, quality of life, number of medical conditions. Results complement previous literature, yet challenge assumptions that shrinking social networks are a defining characteristic of older age. Future examination of our mental and physical engagement data may elucidate these differences. There is uncertainty related to whether retirement negatively impacts health--possibly due to the complexity of retirement decisions. The role of lost work opportunity on retirement decisions may help clarify when retirement has a favorable or negative impact on health. Lost work opportunity can be defined as forced retirement or unemployment prompting an earlier than planned retirement. However, 17% of individuals retiring due to the loss of work opportunity (i.e., unemployment, temporary lay-offs, company buy-outs, forced relocations, etc.) do not report either unemployment or involuntary retirement in survey data. We propose a broader conceptualization of late-career unemployment. Using the Health and Retirement Study (HRS), a lost-work opportunity score (LOS) was computed from items indicating unemployment and forced or unplanned retirement. Correlations were computed between this LOS and all continuous variables in the RAND longitudinal compilation of the HRS to determine its convergent and discriminant validity. The LOS demonstrated a Chronbach's alpha of α=0.82 and had convergent validity with constructs of employment (9 variables), finances (36 variables), and health (14 variables), as predicted by the literature on retirement timing. No other continuous variables in the HRS were identified with a moderate or strong correlation to LOS, demonstrating discriminant validity. Further research should explore whether a combination of variables in the HRS can improve the accuracy of measuring retirement voluntariness. Improved precision in measurement, through an expanded conceptualization of lost-work opportunity, may help explicate the retirement-related factors that impact health, to inform policy and support healthy aging decisions at a societal level.

RETIRING AND CAREGIVING IN THE AGE OF STUDENT LOANS: THE IMPACT OF STUDENT DEBT ON RETIREMENT AND LONGEVITY PLANNING
Samantha Brady, 1 Julie Miller, 1 Alexa Balmuth, 1 Lisa D'Ambrosio, 1 and Joseph Coughlin 1 , 1. MIT AgeLab, Cambridge, Massachusetts, United States Saving for retirement and the ability to provide care for a loved one can be dramatically affected by student loan debt. Currently, approximately 44 million people of all ages in the United States carry the weight of over 1.4 trillion dollars of student loan debt. Student loan borrowers of all ages may experience lower financial preparedness for retirement as well as decreased ability to provide care for family members, including aging parents. While older adults hold a relatively small proportion of student loans, they are the fastest growing subset of student loan borrowers and have disproportionately high rates of student loan defaults. As a result of their defaults, the Social Security retirement benefits of Americans ages 65 and older experienced a 500% increase in offsets over the last decade. This presentation will spotlight an MIT AgeLab mixed methods study about how student loan borrowers between the ages of 51 and 75 experience student loans within family systems and perceive and prioritize longevity planning in light of their student loans. Data collected for this study include focus groups and a large national survey. Preliminary findings suggest that for older borrowers, student loans are generally one of several financial constraints that can inform spending and saving decisions. For most, student loan payments are regarded as stunting overall retirement savings while the minority regard the two separately. Older borrowers also tend to have increased financial and familial responsibilities, including caring for aging parents, that compete for borrowers' limited financial and temporal resources. The interdependence between partners raises considerable interest in the sociology of life course, work and families. Partner influences play a particularly important role in the work domain, because each partner's work decisions have profound effects on the couple as a whole. In contrast to previous research, this study pays detailed attention to the role partners play in workers' labor market decisions by using the case of early retirement decisions. We hypothesize that partners' preferences for older workers' retirement originate from altruism and self-interest. For example, partners might prefer workers to retire early because the worker's job is highly stressful or partners might prefer workers to retire early to increase possibilities for joint leisure. Moreover, we expected that partners influence older workers' early retirement behavior via persuasion and pressure. So, partners might either convince workers to change their preferences, or they might pressure workers to act according to the partners' preferences, irrespective of workers' own preferences. To adequately estimate partners' and workers' preferences for workers' early retirement, we used an instrumental variable approach. This was possible due to the multi-actor longitudinal data available from a large representative sample of older workers and their partners in the Netherlands. The results support that partners' preferences originate in altruism and self-interest and that partners influence workers through persuasion and pressure. Gender differences were marginal, with stronger signs for altruistic origins among female than male partners.

WHY DO OLDER WORKERS WITH CHRONIC HEALTH CONDITIONS PREFER TO RETIRE EARLY?
Anushiya Vanajan 1 Ute Bültmann 2 and Kène Henkens 1 ,

Netherlands Interdisciplinary Demographic Institute, South-Holland, Netherlands, 2. University Medical Center Groningen, Groningen, Netherlands
Older workers experiencing chronic health conditions (CHCs) are more likely to retire early. Current literature, however, lacks knowledge on the different pathways through which CHCs stimulate retirement preference. Earlier research is highly fragmented. Some studies have found CHCs to impact vitality, work limitations, or subjective life expectancy. Others have found vitality, work limitations, or subjective life expectancy to predict retirement preferences. We present a comprehensive model in which we hypothesize that the effects of four CHCs -arthritis, cardiovascular disease, sleep disorders, and psychological disorders -on retirement preferences are differentially mediated by vitality, health-related work limitations, and subjective life expectancy. We analyzed data from 6,294 older workers (60 -65 years) in the Netherlands. Effects of CHCs on older workers' retirement preferences were mediated by vitality, health-related work limitations, and subjective life expectancy. The main mediation pathway differed for each CHC. Severe health-related work limitations among older workers with arthritis (65.6% mediated) and cardiovascular disease (44.0%) predominantly guided their retirement preferences. Lower vitality levels mainly mediated retirement preferences of older workers with sleep (59.1%) and psychological disorders (52.9%). Lower subjective life expectancy was a significant mediation pathway (13.7%) for older workers with cardiovascular diseases. Extending working lives is a key public health and policy challenge. We show that health-related work limitations and vitality play a major role in determining retirement preferences of older workers experiencing CHCs. Since both mediators are modifiable, targeted interventions may not only extend the working lives of older workers, but also improve its quality.

SENSORY LOSS AND THE HEALTH CARE SYSTEM
Chair: Nicholas Reed, Johns Hopkins University, Baltimore, Maryland, United States Discussant: Charlotte Yeh, AARP Services, Inc.,

Washington, District of Columbia, United States
Communication is fundamental to patient-centered care. However, sensory impairment may limit communication among older adults. Specifically, hearing impairment strains communication via degraded auditory encoding while vision impairment distresses ability to read and interpret visual cues. The presence of dual sensory impairment, defined as concurrent hearing and vision impairment, may exacerbate these effects. The potential consequences of age-related sensory loss on health care interactions and outcomes are beginning to surface in epidemiologic studies demonstrating poorer patient-provider communication, higher incurred health care costs, increased risk of 30-day readmission, and longer length of stay when compared to individuals without sensory loss. Importantly, these associations may be amenable to intervention via sensory aids; however, uptake to sensory care is low. Notably, less than 20% of persons with hearing impairment have hearing aids and over 55% of Medicare Beneficiaries with reported vision problems have not had an eye examination in the prior year. Affordability and access may contribute to lack of sensory care uptake as Medicare explicitly excludes coverage of vision and hearing services. In this symposium, we will review current and new evidence for whether sensory loss affects health care outcomes, including satisfaction with care and medical costs, and present data on how persons with sensory loss interact with the health care system based on the need and reasons for accompaniment to care visits. Further, we will discuss and provide evidence for how sensory care may mitigate these associations. Lastly, we will place these results within the context of quality care and policy initiatives.

HEARING LOSS AND HEALTH CARE SATISFACTION
Nicholas Reed, 1 Nicholas Reed, 1 Amber Willink, 1 Jennifer Deal, 1 and Frank R. lin 1 , 1. Johns Hopkins University, Baltimore, Maryland, United States Hearing loss (HL) impacts two-thirds of adults over 70 years and affects patient-provider communication which could limit satisfaction. We used two cross-sectional cohorts, The Atherosclerosis Risk in Communities Study (ARIC, n=250) and the Medicare Current Beneficiaries Survey (MCBS,n=12,311) to examine the relationship between HL (subjective and objective measures) and self-report satisfaction with quality of health care using multivariable-adjusted logistic regression. In ARIC, there was an interaction between HL and age such that HL had a greater impact on odds of dissatisfaction as age increased. In an 85-year-old, for every 10 dB increase in HL, the odds of being dissatisfied increased 1.33 (95% Confidence Interval [CI]:0.96-1.83). In MCBS, compared to participants with no trouble hearing, those with a lot of trouble hearing had 1.7 times the odds ( 95% CI = 1.150-2.623) of being dissatisfied. This has implications for patient-centered care planning given that Medicare ties reimbursement to patient-reported satisfaction.

COST-BENEFIT ANALYSIS OF HEARING CARE SERVICES AMONG MEDICARE BENEFICIARIES WITH HEARING AIDS
Amber Willink, 1 Amber Willink, 1 Nicholas S. Reed, 1 and Frank R. Lin 1 , 1. Johns Hopkins University, Baltimore, Maryland, United States Hearing care services for older adults with hearing aids are underutilized and are not covered by the Medicare Innovation in Aging, 2019, Vol. 3, No. S1