امنیت اجتماعی و مالکیت خانه سالمندان
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|24191||2008||26 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Urban Economics, Volume 63, Issue 1, January 2008, Pages 280–305
Over the last twenty-five years, the homeownership rate of households 65 years and older has risen steadily, while the homeownership rate for 35–64 year old households has remained relatively unchanged. At the same time, the real value of Social Security benefits has risen substantially. Using data from the March 1978 to 2001 Current Population Surveys, this paper documents the evolution and assesses the causal role of the Social Security program in increasing elderly homeownership. To isolate the causal effect, the analysis develops an instrumental-variable approach that relies on the large variation in benefits for birth cohorts from 1900 to 1930 due to double indexation of the system and the so-called Social Security “notch.” For all elderly, the estimated elasticity of homeownership to Social Security benefits ranges from 0.26 to 0.49. Across marital groups, the widowed have the greatest responsiveness to benefits. Increases in benefits also increase household formation among the elderly. Overall, the estimates indicate that between half and as much as all of the time-series rise in elderly homeownership over the last twenty-five years can be attributable to the rise in Social Security benefits and suggest that reductions in benefits would alter homeownership among the elderly significantly.
An important trend in US housing markets has been the sustained increase in the homeown- ership rate of those 65 years and older. Based on data from the US Census Bureau’s Housing Vacancy Survey, the official government source on homeownership rates, the elderly homeown- ership rate was 74.4 percent in 1982 and rose to 80.5 percent by 2003. This 6.1 percentage-point increase in the older households’ homeownership rate represented an 8.2 percent total growth in homeownership between 1982 and 2003. In contrast, homeownership rates for 35–64 year old households essentially have been flat for the last two decades. A potential explanation for the rise in elderly homeownership is the concurrent rise in Social Security benefits. From 1982 to 2003, average Social Security benefits grew 23 percent in real terms. Because there is a strong desire among the elderly to age in place (Venti and Wise [27,28], Commission on Affordable Housing and Health Facility Needs for Seniors in the 21st Century ), these higher benefits may have allowed the elderly to stay in their homes longer, thus rais- ing the elderly home ownership rate. Furthermore, as the level of benefits is a major source of retirement income not related to mortgage interest rates, federal tax policy, changes in the wage distribution, the business cycle or other macroeconomic factors that may explain the time-series patterns in homeownership for the non-elderly (Green ), Social Security may have shielded the elderly from whatever forces affected younger age groups. Importantly, if there is a strong link between Social Security benefits and homeownership in retirement, any benefit reductions through future Social Security reforms could decrease long-run homeownership and housing de- mand by the elderly. The primary aim of this paper is to assess the role of Social Security in driving the increase in elderly homeownership. The existing literature has had relatively little to say about the role of Social Security on elderly homeownership. For example, Hurd [13,14] showed in a life-cycle model with mortality risk and bequest motives that the impact of Social Security on when the elderly exhaust their bequeathable wealth (housing included) is theoretically ambiguous and depends on the form of the utility function. In particular, for some preferences in this model, exogenous increases in Social Security benefits push further into the future the period in which the elderly sell their homes, so that homeownership rates would grow over time as more elderly live in their homes to later ages. On the empirical side, there has been comparatively little recent housing research on the elderly in general, and, to the best of my knowledge, no existing studies of the impact of Social Security on elderly homeownership in particular. 1 The small existing empirical literature on the elderly (e.g., Merrill , Feinstein and McFadden , Reschovsky , Megbolugbe et al. [18,19], among others) consistently has found a positive correlation between homeownership and total income, but with a wide range of implied effects. There are two fundamental problems with the empirical analysis of the impact of Social Se- curity on elderly homeownership. First, there may be measurement error in reported income in micro-data. For example, classical measurement error would impart the standard downward at- tenuation bias to estimates of the impact of Social Security on homeownership. Bound et al.  give a comprehensive survey of estimates of measurement error in survey data for various compo- nents of household income, including earnings, asset income, public assistance, unemployment insurance benefits, and Social Security benefits. For the Current Population Survey (CPS)—the data source for the empirical analysis below—Vaughan and Vuskavage  found that only 7 percent of Social Security beneficiaries correctly reported benefits in the CPS, when compared to Social Security Administration (SSA) records. The mean reporting error was 5 percent of ben- efits, but the mean of the absolute value of the reporting error was larger, 15 percent of benefits. In addition, it is well known that there is substantial misreporting of interest and dividend income in the CPS. Second, homeownership patterns in retirement may be determined in large part by choices made when young. In particular, high-income households when young will transition into home- ownership much before retirement. These households also will have high Social Security benefits when they retire, because benefits are increasing in lifetime earnings. Therefore, it would be natural to observe high rates of homeownership in retirement for households with high Social Security benefits, but this would not be a causal relationship, as the high benefits simply proxy for high lifetime earnings when young. Because lifetime earnings are not observed in the CPS, this would impart a positive omitted-variable bias to estimates of the impact of benefits on elderly homeownership. To circumvent these problems, I draw on the methodology developed in a companion set of papers, Engelhardt et al.  and Engelhardt and Gruber [5,6], on the impact of Social Security on elderly living arrangements, poverty, and broader measures of well-being, respectively, to form an instrumental variable for observed Social Security income. Importantly, when forming the instrument, each individual in the sample is assigned the same lifetime earnings history, so that any differences in homeownership due to economic circumstances when young that persist into retirement will not be correlated with the instrument by construction . The exogenous variation in the instrument instead comes from legislation that induced large Social Security benefit changes for individuals born in 1900–1930. The early cohorts in this range saw enormous exogenous increases in benefits, partly due to double indexation of the system in the early 1970s. This double indexing was ended in the 1977 Amendments to the Social Secu- rity Act that generated the so-called “benefits notch,” which affected the middle cohorts in this range. Specifically, because those born in 1916 would attain the early retirement claiming age of 62 in 1978 when the 1977 law went into effect, the 1977 law grandfathered all individuals born before January 1, 1917, under the old benefit rules, but those born in 1917–1921 received benefit reductions that were as much as 20 percent lower than observationally equivalent individuals in the 1916 birth cohort. The notch birth cohorts received large, unanticipated, and permanent reductions in benefits very late in working life, after most households already had become home- owners. For the later cohorts, those born after 1921, benefits were roughly constant in real terms. As I detail below, drawing on this legislative variation in benefits, the instrument is, by construc- tion, plausibly uncorrelated with actual lifetime earnings and housing choices when young, as well as with measurement error in the CPS. Based on data for the elderly from 298,948 families and 270,773 households born in 1900– 1930 drawn from the 1978 to 2001 March CPS, there are three primary empirical findings. First, the instrumental-variable linear probability model estimates indicate that increases in Social Security benefits increase homeownership: the elasticity of homeownership to Social Security benefits ranges from 0.26 to 0.49 for all elderly, with the less elastic response associated with the richest specification. Across marital groups, the widowed have the greatest responsiveness to benefits. Second, increases in benefits increase household formation among the elderly. Third, these estimates indicate that one half to as much as all of the time-series rise in elderly homeown- ership over the last twenty-five years can be attributable to the rise in Social Security benefits. Overall, the results suggest that prospective reductions in Social Security benefits would sig- nificantly alter elderly homeownership: the largest estimates suggest that over 520,000 elderly homeowners would move to renting if benefits were cut 10 percent.The paper is organized as follows. Section 2 describes the CPS and the construction of the analysis dataset. Section 3 charts the time-series evolution of elderly income, headship, and homeownership from 1978 to 2000. Section 4 provides background on the basic calculation of Social Security benefits to better understand the construction of the instrumental variable, which is discussed in Section 5. Both sections draw extensively on the methods and exposi- tion developed in a companion set of papers, Engelhardt et al. [5,6] and Engelhardt and Gruber [7,8]. Section 6 discusses the estimation results, and Section 7 discusses the implication of the empirical estimates in assessing the impact of prospective benefit reductions on the elderly home- ownership rate. There is a brief conclusion.
نتیجه گیری انگلیسی
Over the last twenty-five years, the homeownership rate of households 65 years and older has risen steadily, while the homeownership rate for 35–64 year old households has remained relatively unchanged. At the same time, the real value of Social Security benefits has risen sub- stantially. Using data from the March 1978–2001 Current Population Surveys, I show these two trends are linked. The estimated elasticity of homeownership to Social Security benefits ranges from 0.26 to 0.49 over all elderly, with the largest effects for the widowed. The estimates in- dicate that one-half to essentially all of the time-series rise in elderly homeownership over the last twenty-five years can be attributable to the rise in Social Security benefits and suggest that reductions in Social Security benefits would significantly alter the elderly homeownership rate. These results are striking, but leave a set of important research questions unanswered: for example, at a finer level of detail, the mechanisms through which increased Social Security ben- efits operate to raise homeownership, such as more home maintenance (Davidoff ), structural modifications to age in place (Commission on Affordable Housing and Health Facility Needs for Seniors in the 21st Century ), longer residence spells, etc.; how changes in Social Security ben- efits, which are real annuities, affect the demand for reverse mortgages and other annuity-based products to liquidate the home equity of the elderly; whether benefits crowd out transfers of time or money to the elderly from, for example, children, targeted toward the upkeep of the home; and how benefits affect the desire of the elderly to spend down their housing equity. Analysis of these types of questions is important for a richer understanding of the economic implications of Social Security for elderly well-being specifically and the role of aging in housing decisions more broadly.