Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Decomposing quintiles by subgroups

    Dear users, I would like to share with you a doubt on which I am thinking about.

    I am looking at how 5 quintiles (containing each the 20% of the workforce, ordered by job quality) change along time (thus, from t0 where they are all equally the 20%, to t1, to see what quintile increases and what decreases).
    Then, I want to see how subgroups of workers (for instance, men and women) contributed to the 'evolution' of each quintile. Here the point is that at t0 men and women are not equally distributed in each quintile (because quintiles are calculated on the whole workforce: when considering separately men and women, they have a different distribution within each quintile in the year we take as reference): in your opinion, how should I calculate the contribution of each subgroup at the quintiles change, taking into account the fact that each subgroup is differently distributed within quintiles?

    As an example, see the following tables:
    Total
    1992 2015
    Q1 22.76 12.09
    Q2 19.83 22.53
    Q3 18.64 21.31
    Q4 18.99 25.23
    Q5 19.78 18.83
    Female
    1992 2015
    Q1 25.04 10.46
    Q2 16.52 19.17
    Q3 11.52 15.48
    Q4 17.46 28.61
    Q5 29.47 26.27
    Male
    1992 2015
    Q1 21.59 13.27
    Q2 21.54 24.97
    Q3 22.33 25.54
    Q4 19.78 22.78
    Q5 14.77 13.44

    Thanks a lot for your feedbacks.

    Best, G,
Working...
X