Age dependant rewards

Dear Friends,

How can I add productivity loss (job salary lost) that depends on age of the peple in markov model. People should incur productivity loss till 65 years old (retirement age). Cohort runs till the age of 84 years old. 


Thank you,

Best Regards,




  • There are several steps needed to achieve this.

    1. Establish the age of the cohort. Set a starting age for the cohort, then create an age variable that starts with the starting age, then adds to the age based on the cycle length.
      startAge = 50
      age = startAge + _stage (for an annual cycle)
    2. Establish the income lost per cycle. If this is constant, then just create a simple parameter for this. Otherwise, it could be a set of income values that vary by age. You could pull those from a table based on the current age of the cohort.
    3. Place the lost salary cost formula in the appropriate incremental state cost for the state where the cohort cannot work. Even if the lost salary cost is constant, you will still need a formula, so that lost productivity stops at age 65.

    It is unclear how you would want to handle a Dead state where salary is lost for all years after death.


  • Hi Andrew,

    How to handle a Dead state when calculating indirect costs associated with premature death is my question. I have a table of age specific weekly pay. Is there a function or formula I can use to calculate salary lost for all years after death with different weekly pay at every age bracket. The formula I used is not working " tWeeklypay[age]*52*(100-age)*(1-tUnemploymentRate[age])". I want the weekly pay to vary at different.

    Many thanks for your help.

    Best regards,


  • Sorry, a typo - I mean "I want the weekly pay to vary at different age". Thanks. Bing

  • You could create a Dead state that is not actually a terminal node absorbing state. Instead, make it a label node with a single branch that sends the cohort back to the same Dead state.

    You could then accumulate the lost salary for all years from death to age 65 in that Dead state.

  • Many thanks Andrew for your reply.

    Just one more question. How can we set in Treeage that only certain percentage of cohort is employed (e.g. 70% are working in cohort, rest are unemployed). Those who are working only have productivity losses before they reach age of retirement. 

    Thanks again!

    Best Regards,


  • I can't really say where it is best to do that in the model, but you could have different states (alive and dead) for employed vs. unemployed.

    Alternatively, and maybe more easily, you could just add a parameter for employed_pct and multiply the wages by that employment_pct to reduce the overall impact for unemployment.

  • Thanks a lot Andrew!

  • Thank you so much, Andrew. It works very well:-)

  • Happy to help.

Please sign in to leave a comment.

Didn't find what you were looking for?

New post