Skip to main content

Table 2 Mean Standardized Bias, Matching Retirees with Non-Retirees, All Well-being Outcomes

From: Employment, late-life work, retirement, and well-being in Europe and the United States

Control group: voluntary part-time workers

 

Before

After

BPL

44.503

4.983

Experienced happiness yesterday

44.467

5.071

Smiled a lot yesterday

44.719

3.950

Experienced stress yesterday

44.650

5.036

Experienced anger yesterday

44.661

4.308

Satisfied with personal health

44.516

4.889

Satisfied with freedom

43.489

5.505

Control group: full-time employees

 

Before

After

BPL

55.708

4.247

Experienced happiness yesterday

55.900

4.580

Smiled a lot yesterday

55.940

4.698

Experienced stress yesterday

56.100

4.319

Experienced anger yesterday

56.011

4.337

Satisfied with personal health

55.875

4.299

Satisfied with freedom

54.583

4.719

Control group: self-employed

 

Before

After

BPL

51.645

7.477

Experienced happiness yesterday

52.056

8.214

Smiled a lot yesterday

51.887

7.539

Experienced stress yesterday

52.054

7.614

Experienced anger yesterday

51.954

7.545

Satisfied with personal health

51.987

7.487

Satisfied with freedom

49.682

6.490

Control group: involuntary part-time workers

 

Before

After

BPL

46.738

5.344

Experienced happiness yesterday

46.935

5.435

Smiled a lot yesterday

46.906

4.131

Experienced stress yesterday

46.950

5.346

Experienced anger yesterday

46.878

5.345

Satisfied with personal health

46.581

6.592

Satisfied with freedom

45.993

5.082

  1. Source: Gallup World Poll, 2009-2010.
  2. Notes: Summary of results from caliper propensity score matching. The treatment group is the group of retirees. The control group is the voluntary part-time workers in the top panel, followed by the full-time employees, the self-employed, and the part-time workers in the bottom panel. Covariates used for matching (and satisfying the balancing property) are: age, gender, whether the respondent is married, an interaction term for marriage and gender, whether the respondent has at least high school education, whether religion is important in the respondent's life, country, and year.