65歲以后的支出軌跡及其對退休經濟準備的影響.pdf

《65歲以后的支出軌跡及其對退休經濟準備的影響.pdf》由會員分享，可在線閱讀，更多相關《65歲以后的支出軌跡及其對退休經濟準備的影響.pdf（23頁珍藏版）》請在三個皮匠報告上搜索。

1、Spending Trajectories after Age 65 and Implications for Economic Preparation for RetirementMichael Hurd and Susann RohwedderSusann RohwedderJune 12,20249th International Pension Research Association Conference,ParisStarting Point for Financial Planning for Retirement:“How much will I/we need to fina

2、nce our spending?”Anticipate households spending level and patho Desired living standards at beginning of retirement yearso Anticipated onward trajectorySome things difficult to foresee,sources of uncertainty:o No longer able to work to compensate for shockso Increasing risk of health shocks,long-te

3、rm care(informal,formal?)o Elevated mortality risko Investment risko May lose ability to manage finances Obtain guidance from(i)economic theory,(ii)observed spending patterns.Consistent with each other?Spending Trajectory Implied by Lifecycle Model During retirement:Spending declines due to increasi

4、ng mortality risk Flatter path(smaller rate of decline conditional on age)if higher life expectancy;bequest motive Widowing:one-time drop as single survivor spends less than the married household Interaction with Health:If health good,spend greater share on goods&services complementary to health,les

5、s on health care;If health bad,reverse.On average,health declines with age spending composition changes with age,reduced enjoyment from spending.Data:Health and Retirement Study and CAMS Supplement Representative of U.S.population over age 50 Longitudinal:biennial,1992 onward Spending supplement:bie

6、nnial,2001 onward Household spending elicited in 39 categories Total spending=sum of 39 categories Total compares closely to Consumer Expenditure SurveyLarge Differentials in Spending Levels16,75534,98026,76047,41933,28862,11540,73177,798SINGLEMARRIEDMEDIAN ANNUAL HOUSEHOLD SPENDINGMEDIAN ANNUAL HOU

7、SEHOLD SPENDING AGE 65AGE 65-69$2019 69$2019 Lowest initial wealth quartile2nd3rdHighestShown by initial wealth quartile(measured at or near age 65)Source:Hurd and Rohwedder(2023),Journal of the Economics of Aging.Spending TrajectoriesUse longitudinal data on household spending tooEstimate rates of

8、change in spending by age,marital status and initial wealth(at age 65)oConstruct spending trajectories based on rates of change(Note:cannot use cross-section data because of differential mortality and cohort differences)Assess whether estimated trajectories useful guidance:oAge pattern in households

9、 reporting o being financially constrained,o Dissatisfied with economic situationoChanges in enjoyment of spending on certain types of goods and servicesSpending Declines across ALL Socioeconomic Groups 01020304050607080901006566676869707172737475767778798081828384858687888990Spending Path:Spending

10、Path:Married Households Married Households by Initial Wealth Quartile(assessed at age 65)lowest wealth quartilesecondthirdhighest wealth quartileOverall average annual rate of change:-2.7%2.7%Largest rate of decline among lowest initial wealth group.Flatter spending paths for all other groups with h

11、igher initial wealth;longer lived on average.Similar patterns if stratifying by education.Source:Hurd and Rohwedder(2023),Journal of the Economics of Aging.Updated.Spending Declines ACROSS ALL Sociodemographic Groups01020304050607080901006566676869707172737475767778798081828384858687888990Spending P

12、ath:Spending Path:Single Households Single Households by Initial Wealth Quartile(assessed at age 65)lowest wealth quartilesecondthirdhighest wealth quartileOverall average annualized rate of change:-2.0%.2.0%.Largest rate of decline among lowest two initial wealth quartiles.Flatter spending paths fo

13、r highest two initial wealth quartiles;longer lived on average.Similar patterns if stratifying by education.Source:Hurd and Rohwedder(2023),Journal of the Economics of Aging.Updated.Interpretation of declining spending paths?Are the declining spending paths mostly chosen,that is,consistent with indi

14、viduals preferences?or forced due to unexpectedly tightening budget constraint?Study budget shares(composition of spending)and variation by age Respondents own reports on:Satisfaction with economic position Whether financially constrained(now,how it changed)Changes in enjoyment from some types of sp

15、endingHow does composition of spending change with age?Budget share=fraction spent on good x/total spendingSpending Composition Changes with Age,as Health Declines 00.010.020.030.040.050.0650-5455-5960-6465-6970-7475-7980-8485+Trips and vacationsCouplesSinglesWhile health good,companionship:While he

16、alth good,companionship:Spend more on trips and vacations,leisure activitiesAs health declines,spend less As health declines,spend less on transportation,clothing,durables,trips and vacations,leisure 0.000.050.100.150.200.2550-54 55-59 60-64 65-69 70-74 75-79 80-8485+TransportationCouplesSinglesSpen

17、ding Composition Changes with Age,as Health Declines 00.020.040.060.080.10.120.140.160.1850-54 55-59 60-64 65-69 70-74 75-79 80-8485+Health careCouplesSinglesAs health declines,spend higher fraction on health care.Total spending still declines,except possibly in last year of life.0.000.020.040.060.0

18、80.1050-5455-5960-6465-6970-7475-7980-8485+Gifts and donationsCouplesSinglesGifts and donations:not health dependentIncreasing with age as households time horizon shortens and mortality risk increases.Is Prevalence of Financial Distress Higher at Advanced Ages?No.At advanced ages Larger fraction rep

19、orts(i)Satisfaction w/finances (ii)Little financial constraints0.010.020.030.040.050.055-5960-6465-6970-7475-7980-8485+Satisfaction with Present Financial Situationnot satisfiedsomewhatcompletely/very0.010.020.030.040.050.060.070.055-5960-6465-6970-7475-7980-8485+Financially Constrained Nownot/hardl

20、ysomewhatveryIs Prevalence of Financial Distress Higher at Advanced Ages?Within-person Comparison with Six Years Ago010203040506055-5960-6465-6970-7475-7980-8485+Satisfaction with Present Financial SituationCompared to Six Years Agolesssamemore010203040506055-5960-6465-6970-7475-7980-8485+Current Fi

21、nancial ConstraintsCompared to Six Years AgomoresamelessSatisfaction Percent reporting unchanged satisfaction increased with age Percent reporting less satisfaction decreased with ageFinancial constraints Percent reporting“unchanged”increased with age Percent reporting“more”decreased with ageChanges

22、 in Enjoyment from Spending Special Module in HRS-CAMS 2019Change in enjoyment(marginal utility)of spending-related activitiesCompared to six years ago,how much enjoyment do you(or would you)get today from traveling?5-point response scale:1.Much less 3.Same 5.Much more Same question asked about spen

23、ding on -Going out to eat -Leisure activities -Having new car -Having new appliances0102030405060708055-5960-6465-6970-7475-7980-8485+TravelinglesssamemoreEmpirical Evidence on Change in Enjoyment(CAMS 2019)Compared to six years ago,how much enjoyment do you (or would you)get today from traveling?At

24、 older ages Higher percentage with less enjoyment compared w/6 years ago(within-person)Lower percentage with more enjoymentResults robust to inclusion of controls.Similar results for:-Going out to eat -Leisure activities -Having new car -Having new appliancesWhy does it matter whether spending decli

25、nes?Quantitatively important for finding necessary savings Social Security is a substantial fraction of economic resources Gap between Social Security and pensions and consumption needs to be filled by savings A moderate decline in spending translates into a larger decline in required savingso Examp

26、le:1.Suppose Flat spending path over retirement with expected present value of 100k Social Security finances 50%of spending(50k),savings to cover 50k gap.2.Suppose Declining spending trajectory so that expected present value of spending is 10%lower(90k).Gap to be covered by savings:40k,which is 20%l

27、ess than in scenario with flat spending path.Application:Declining Spending Paths in Assessments of Adequacy of Retirement ResourcesMicrosimulation:Microsimulation:Do households have sufficient economic resources to finance with high probability-consumption to the end of life without running out of

28、wealth?Single and married individuals observed in HRS around age 65-69;reported level of households spending,income and assets Simulate households spending path,anchored to initial spending Account for taxes,asset portfolio allocation(different rates of return and order of decumulation)Stochastic le

29、ngth of life,OOP medical expenditure shocks,heterogeneity by marital status,sex,education.Percent adequately prepared:65.8%Lower among singles,persons of color0.5820.6450.5610.2800.3820.2550.2720.5790.2020.0000.1000.2000.3000.4000.5000.6000.700AllMaleFemaleSingleSingle Individuals,by Race/EthnicityN

30、on-Hispanic WhiteNon-Hispanic BlackHispanic0.8190.8280.8150.6500.6550.6460.6430.6050.6590.0000.2000.4000.6000.8001.000AllMaleFemaleMarried Married Persons,by Race/EthnicityNon-Hispanic WhiteNon-Hispanic BlackHispanicSource:Hurd and Rohwedder(2023),updated.Conclusions Spending declines at older ages(

31、not flat or increasing)Accords with predictions of lifecycle model.Observed for all socioeconomic groups for most,not driven by tightening budget constraint.Self reports:financial constraints not limiting factor for most households Some explanations:Declines in health,reductions in enjoyment Implica

32、tions for resources needed in retirement Overall assessment:For most,economic preparation adequate Simulations:most people adequately preparedConclusions-continued But some more economically constrained over time at older ages Found vulnerable subpopulations with only a minority having a low risk of running out of resources No need to encourage“people”to save more;Instead:Need to encourage(or provide for)“some people.”Focus attention on causes and possible policy solutions.susannrrand.org Thank you!