Dark Background Template

Effects of Attrition and NonResponse in the HRS Pierre-Carl Michaud Arie Kapteyn James P. Smith Arthur van Soest What are the Questions? What is the impact of attrition in the HRS? -on current cross-sections - on longitudinal analysis What is the role of ever out? (Project Funded by SSA) 2 3/23/ DATA Original 1992 age eligible HRS cohort up to 2004 wave 3 3/23/ A Short Detour on Baseline NonResponse Not our focus in this paper but everything is obviously conditional on this HRS attempts to match the CPS by weighting- uses age, gender, race, ethnicity, household composition, and PSUs 4 3/23/ Baseline Non-Response 100 90 80 70 60 50 40 30 20 10 0 81.6 81.1 77.0 % Response Rate Complete Sample African American Supplement Hispanic Supplement 5 3/23/

Baseline Comparisons of HRS and CPS After weighting, baseline HRS and the CPS match very well, which is not much of a surprise- but it also matches well on variables not in the weights Of course, CPS is not necessarily the truth, beauty, and the American way- especially for some types of key HRS content 6 3/23/ Four Types of Subsequent NonResponse Studied Always in the Study Always in Out at least once but back in 2002- Ever out Died Prior to 2004- Died Left and not back in 2004Attritors 7 3/23/ Baseline Characteristics by Type of Participation Sequence 1992-2004 (weighted using Baseline HRS weights) 100 90 80 70 60 50 40 30 20 10 0 60.5 9.4 Always In Ever Out 14.5 Died 15.6 Attritors Between 24.9 and 42.9% of ever out came back, since 96 8 3/23/ Mortality in HRS 16 15.4

14 13.2 12 12.5 10 8 8.6 6 5.9 4 3.8 2 0 1.7 1.7 1994 2.2 2.1 1996 1998 Mortality rate 3.0 2000 3.9 2002 2.9 2004 Weigthed Cumulative mortality rate 9 3/23/ Baseline Characteristics by Type of Participation Sequence 1992-2004 (weighted using Baseline HRS weights) Characteristics Always In Ever Out Died

Attritors Total Demographics in 1992 Age (yrs) Female (%) Born outside U.S. (%) African American (%) Hispanic (%) Married (%) Widow(er) (%) 55.5 55% 9% 9% 5% 78% 6% 54.9 50.2 15.6 15.0 14.1 73.2 5.7 56.4 43% 7% 26% 8% 66% 10% 55.5 52.2 12.9 9.0 7.2 78.2 4.7 55.6 52% 10% 10% 6% 77% 6% Health Status in 1992 Health fair/poor (%) 16% 20.8 47% 16.1 20%

21% 84% 68% 16% 15.2 73.4 69.6 10.9 12% 72% 49% 30% 16.8 83.5 68.8 15.8 19% 81% 66% 17% SES and Employment Status in 1992 College and above % Own house (%) Working (%) Retired or disabled (%) 10 3/23/ Baseline Wealth Distribution by Type of Response 1992-2004 (thousands of dollars) (weighted using Baseline HRS weights) Household Wealth in 1992 Mean 10th pctile 25th pctile Median 75th pctile 90th pctile 276 7.0 55.9 150.6

319.8 606.2 Ever out but in for 2004 281.6 0 17.3 98.5 242 660.4 Died prior to 2004 181.2 0 10.8 82.6 199.2 389.7 Attritor 270.1 8.1 58.7 151.3 329.9 650.2 Total 222.7 2.7 45.3 134 299.8 591.3 Always in 11 3/23/

Baseline Income Distribution by Type of Response 1992-2002 (thousands of dollars) (weighted using Baseline HRS weights) Household Income in 1994 Mean 10th pctile 25th pctile Median 75th pctile 90th pctile Always in 69.1 14.0 29.7 54.9 87.7 131.6 Ever out but in for 2004 68.8 10.6 23.9 48 79.9 117.2 Died prior to 2004 48.5 7.2 15.9 33.3 62.6

97.4 Attritor 68.6 13.8 30.1 53.3 80.2 128.7 Total 66.2 12 26.6 51.1 82.5 126.2 12 3/23/ Multinomial Logits for Panel Inclusion- selective variables, females reference: always in Parameter Estimates - Status 2004 covariates ever out died attritor born outside U.S. 0.281 -0.427 ** 0.300 ** black 0.234 0.071 0.065 hispanic 0.778 ** 0.055 0.322 * single -0.917 ** 0.064 -0.251 some college -0.073 -0.228 -0.370 ** college and above -0.348 -0.204 -0.408 ** own house -0.228 -0.179

-0.129 retired -0.223 0.330 ** -0.268 * 1st wealth quintile 0.320 -0.175 -0.318 * 2nd wealth quintile 0.060 -0.347 ** -0.074 4th wealth quintile -0.032 -0.146 -0.043 5th wealth quintile 0.136 -0.524 ** 0.200 1st hld income quin. 0.257 0.113 -0.047 2nd hld income quin. 0.158 0.162 -0.030 4th hld income quin. 0.250 -0.316 * -0.084 5th hld income quin. 0.076 -0.142 -0.274 * health fair/poor 0.351 ** 0.781 ** 0.059 13 3/23/ Multinomial Logits for Panel Inclusion- selective variables, males Males reference: always in Parameter Estimates - Status 2004 covariates ever out died attritor born outside U.S. 0.213 -0.230 0.580 ** black 0.729 ** 0.338 ** 0.141 hispanic 0.726 ** -0.080 0.041 single 0.262 0.224 0.257 some college -0.038 0.092 0.107

college and above -0.256 -0.172 -0.239 own house -0.117 -0.074 -0.057 retired -0.438 ** 0.149 0.205 1st wealth quintile 0.441 ** 0.417 ** -0.219 2nd wealth quintile 0.210 0.069 -0.019 4th wealth quintile -0.042 -0.018 -0.002 5th wealth quintile 0.296 -0.097 0.084 1st hld income quin. -0.110 -0.107 0.033 2nd hld income quin. -0.037 0.040 0.018 4th hld income quin. 0.070 -0.173 0.013 5th hld income quin. -0.152 -0.207 -0.078 14 3/23/ Use inverse probability weights For example: the weights for the sample including ever out and always in: p ( xi 0 ) p( si ,a | xi 0 ) p( si ,e | xi 0 ) 1 p ( si ,d | xi 0 ) where a is always in e is ever out d is died 15 3/23/ Effects of Weighting on Household Wealth: Samples Excluding and Including Ever out Sequences (thousands of dollars)

Statistic (percentile) Household Wealth in 2004 10th 25th Median 75th 90th Only always in (attrition weights correct for ever out and attritors) Unweighted 3.8 55.5 179.0 448.0 875.0 HRS-92 7.4 69.0 213.2 500.0 969.2 IPW (both ever out and attrit.) 5.6 64.0 203.4 488.0 951.2 HRS-04 7.3 69.8 213.5 500.0 977.0 Always in and Ever out sample (attrition weights correct for attritors only) Unweighted

2.0 48.8 166.6 430.0 864.0 HRS-92 5.0 61.5 200.1 487.0 967.5 IPW (only attritors) 5.0 62.3 200.5 487.0 966.0 HRS-04 5.0 62.0 200.0 487.0 969.2 16 3/23/ Fixed Effects Regression for Log-Wealth Balanced Incl. those who die Excl. those who die Par. Par. z-diff Par. z-diff 3.08 2.85 0.92 2.77 1.28 -0.22 -0.2 -0.85 -0.2 -1.2

age age squared current wave ever had severe ever had mild health good health fair/poor divorced widow(er) previour wave ever had severe ever had mild health good health fair/poor divorced widow(er) Excl. ever out Par. z-diff 3 1.62 -0.22 -1.65 age -0.04 -0.21 -0.17 -0.36 -0.89 -0.63 -0.01 -0.16 -0.18 -0.4 -0.9 -0.68 -0.87 -1.48 0.52 1.26 0.08 1.06 0.03 -0.16 -0.18 -0.38 -0.95 -0.67 -2.36 -1.76 0.76 0.9 0.94 0.99 0.02 -0.17 -0.17 -0.34

-0.99 -0.67 -0.58 -0.71 2.09 3.98 -1.61 0.33 -0.03 0.21 -0.07 -0.12 -0.21 -0.15 -0.11 0.14 -0.09 -0.13 -0.13 -0.05 2.05 2.02 0.83 0.31 -1.18 -2.07 -0.1 0.16 -0.1 -0.14 -0.08 -0.03 2.37 1.84 1.98 0.72 -2.52 -3.24 -0.83 0.19 -0.1 -0.15 -0.12 -0.03 1.25 2.24 0.24 -0.86 -1.42 0.002 z-diff tests difference with first column Last two columns also differ significantly age s curre ever

ever healt healt divor wido previ ever ever healt healt divor wido rho 17 3/23/ Conditional Logits for Home Ownership Balanced Incl. those who die Excl. those who die Par. Par. z-diff Par. z-diff 10.34 9.64 1.12 9.88 0.8 -0.81 -0.74 -1.29 -0.76 -0.93 age age squared current wave ever had severe ever had mild health good health fair/poor divorced widow(er) previour wave ever had severe ever had mild health good health fair/poor divorced widow(er) Excl. ever out Par. z-diff 10.85 2.19 -0.84 -2.3 -0.01 -0.03 -0.18 -0.22 -1.92 -1.22

-0.01 -0.09 -0.14 -0.25 -1.87 -1.17 -0.003 0.78 -1.1 0.633 -0.57 -0.57 0.002 -0.02 -0.13 -0.24 -1.84 -1.18 -0.16 -0.12 -1.38 0.42 -1.01 -0.58 0.03 -0.06 -0.16 -0.21 -1.89 -1.17 0.7 -0.9 -1.16 0.7 -0.97 0.03 -0.3 0.29 0.09 0.01 -0.27 -0.46 -0.37 0.18 0.1 0.04 -0.42 -0.63 0.82 1.38 -0.11 -0.48 1.54 2.09 -0.37

0.21 0.07 0.03 -0.42 -0.55 1.04 1.25 0.73 -0.36 1.79 1.3 -0.33 0.25 0.06 0.02 -0.35 -0.49 0.79 0.92 -0.12 -0.4 1.26 1.22 z-diff tests difference with first column 18 3/23/ Conditional Logits for Labor Force Participation Balanced Incl. those who die Excl. those who die Par. Par. z-diff Par. z-diff age is (ref 50-53) 54/55 56/57 58/59 60/61 62/63 64/65 66/67 68/69 70+ current wave ever had severe ever had mild health good health fair/poor divorced widow(er) previour wave ever had severe ever had mild health good health fair/poor divorced widow(er) Excl. ever out Par.

z-diff -0.25 -0.37 -0.56 -0.86 -1.34 -1.64 -1.93 -2.06 -2.31 -0.26 -0.34 -0.53 -0.83 -1.31 -1.63 -1.89 -2.04 -2.28 0.29 -0.76 -0.94 -0.95 -0.93 -0.33 -0.95 -0.47 -0.86 -0.25 -0.33 -0.52 -0.81 -1.29 -1.6 -1.88 -2.21 -2.28 0.05 -1.14 -1.43 -1.5 -1.48 -1.07 -1.46 -0.95 -1.11 -0.23 -0.33 -0.51 -0.8 -1.29 -1.6 -1.87 -2.02 -2.26 0.05 -1.14 -1.43

-1.5 -1.48 -1.07 -1.46 -0.95 -1.11 -0.19 0.02 -0.04 -0.31 -0.07 0.02 -0.24 0.004 -0.06 -0.35 -0.07 0.03 2.33 0.6 2.27 3.14 0.01 -0.16 -0.23 0.01 -0.05 -0.33 -0.06 0.04 2.16 0.16 1.46 1.75 -0.29 -0.78 -0.2 0.01 -0.05 -0.32 -0.07 0.02 2.16 0.16 1.46 1.75 -0.29 -0.78 0.02 -0.09 -0.002 -0.21 0.04 0.12 0.02 -0.08

-0.01 -0.22 -0.001 0.08 0.11 -0.81 0.76 0.65 1.2 1.65 0.03 -0.08 -0.01 -0.21 0.01 0.08 -0.54 -0.94 0.32 -0.05 1.16 1.72 0.02 -0.09 -0.01 -0.21 0.03 0.09 -0.54 -0.94 0.32 -0.05 1.16 1.72 z-diff tests difference with first column 19 3/23/ Conclusions Cross-sectional HRS analysis in 2004 seems ok with current weights Keeping the ever out in is particularly important for the analysis of wealth and income Balanced samples should be used with caution (e.g. wealth and labor force participation) 20 3/23/

Recently Viewed Presentations

  • Snímek 1

    Snímek 1

    * Historie 2 1853 - důkaz krve - Ludvik Karol Teichmann 1880 - využití teploty těla pro určení doby smrti 1883 - bertillonáž - Louis Alphonse Bertillon 1901 - daktyloskopie (Scotland Yard) 1903 - kapalinová chromatografie M.S.Cvet) 1910 - první...
  • Mutlidimensional Detective Alfred Inselberg Streeable ...

    Mutlidimensional Detective Alfred Inselberg Streeable ...

    Alfred Inselberg Streeable, Progressive, Mutlidimensional Scaling Matt Williams, Tamara Munzner Rylan Cottrell Mutlidimensional Detective Transformation of multivariate relations into 2-D patterns A discovery process for visual data mining Parallel Coordinates Visualize without loss of information.
  • Influences and Foundations of American Democracy

    Influences and Foundations of American Democracy

    The Preamble: When in the Course of human events, it becomes necessary for one people to dissolve the political bands which have connected them with another, and to assume among the powers of the earth, the separate and equal station...
  • Biomes: Definition and Brazilian Examples

    Biomes: Definition and Brazilian Examples

    Biomes: Definition and Brazilian Examples Dr. James A. Danoff-Burg Columbia University A Definition "A biogeographical region or formation - a major regional ecological community characterized by the distinctive life forms and principal plant (or animal in marine biomes) [life forms]."
  • Suki's Kimono - Jefferson County Public Schools

    Suki's Kimono - Jefferson County Public Schools

    Is Suki's kimono made of cotton or silk? Do you hold a handkerchief in your hand or in your mouth? Do dancers move with rhythm or rhyme? Are Suki's dance movements graceful or clumsy? Did Suki and her grandmother go...
  • Value of Information - University of Sheffield

    Value of Information - University of Sheffield

    Costs Some introductory remarks by Tony O'Hagan Welcome! Welcome to the fourth CHEBS dissemination workshop This forms part of our Focus Fortnight on "Costs" Our format allows plenty of time for discussion of the issues raised in each talk, so...
  • The Circulatory System

    The Circulatory System

    Red Blood Cell Antigens and Blood Typing. Rh factor. Antigen D - Rh-positive or Rh-negative. Issues in pregnancy: An Rh− mother exposed to Rh+ fetal blood produces antibodies. This may cause erythroblastosis fetalis in future pregnancies as antibodies cross the...
  • PowerPoint Presentation

    PowerPoint Presentation

    Layoffs Generally. Management generally does not have to bargain with APC about the decision to layoff Unit 4 employees; however, it must bargain with APC about the impact of that decision onUnit 4 employees. "[A] public employer' s decision to...