print

Correlational findings

Study Nguyen et al. (2020): study AU 2001

Public
55-75 aged adults, Australia, followed 17 years, 2001-2018
Survey name
AU-HILDA combined waves
Sample
Respondents
N = 7568
Non Response
Assessment
Interview: face-to-face

Correlate

Authors's Label
Retirement status
Our Classification
Remarks
followed 17 years, from 2001-2018 (T1-T18)
Distribution
1: 55.8%, 0:44.2%
Related specification variables
Operationalization
Retired defined as "not in the labour force at the survey time"
I retired
0 non-retired

Observed Relation with Happiness

Happiness Measure Statistics Elaboration / Remarks O-SLW-u-sq-n-11-d DM = +,01 ns retired         M = 8.05
non-retired     M = 8.04
O-SLW-u-sq-n-11-d b-fix = -.01 ns T1-T18 happiness CHANGE by being retired (vs not)

b-fix controlled for
- age, age squared
- migration status
- aboriginal status
- education
- marital status
- number of household members at various age groups
- home ownership status
- state/territory dummies
- survey year
- survey quarters
- local socio-economic background variables
  - regional unemployment rates
  - an index of relative socio-economic advantage/disadvantage
  - a metropolitan dummy
O-SLW-u-sq-n-11-d b-IV-F = +1.0 p < .01 Instrumental Variable used: individual's age is equal or greater than the state pension eligibility age set at the survey time O-SLW-u-sq-n-11-d b-IV-F = +/- s/ns Years since retirement
- 0 year                   b-IV-F = +1.0(01)    
- 1 year                   b-IV-F = +.81(01)   
- 2 year                   b-IV-F = +.50(05)  
- 3 year                   b-IV-F = +.32(ns)  
- 4 year                   b-IV-F = +.04(ns)  
- 5 year                   b-IV-F = +.08(ns)  
- 6 year                   b-IV-F = +.29(ns)  
- 7 year                   b-IV-F = -.02(ns)
O-SLW-u-sq-n-11-d b-IV-F = + p < .01 Male                         b-IV-F = +.99(01)
Female                       b-IV-F = +1.4(01)
O-SLW-u-sq-n-11-d b-IV-F = + p < .10 Married                      b-IV-F = +.97(01)
non-married                  b-IV-F = +1.1(10)
O-SLW-u-sq-n-11-d b-IV-F = + p < .05 Post school/high education   b-IV-F = +1.3(01)
School year 12 or below      b-IV-F = +.82(05)
O-SLW-u-sq-n-11-d b-IV-F = + p < .10 Blue collar                  b-IV-F = +.73(10)
White collar                 b-IV-F = +1.3(05)
O-SLW-u-sq-n-11-d b-IV-F = + p < .05 Top income tercile           b-IV-F = +.73(10)
Bottom income tercile        b-IV-F = +1.3(05)

Similar for analyses
- different retirement definition used
  - retired completely from the labor force
  - excluding "not in the the labor force marginally"from retirement
  - weekly working hours
- including age cubed
- including general physical health
- including general mental health
- including disabled condition
- including non-wage income
- including different age bandwidth
  - 5 years around the pension eligibility age
  - 4 years around the pension eligibility age
  - 3 years around the pension eligibility age

Figure note:
Results for different sub-populations are obtained from separate FE-IV regressions. The (red) triangles indicate the retirement coefficient estimate (and its 95% confidence interval) in the regression for the sub-population mentioned on the yy axis while the(black) circles represent the estimate for the other sub-population. The solid (dash) vertical line shows the retirement coefficient (95% confidence interval) estimates for the whole population.