Skip to contents

Compute ratio of last salary to first pension for retired workers

compute_pension_ratio.Rd

For each individual who has retired, computes the ratio of their first pension payment to their last active salary.

Usage

compute_pension_ratio(
  personnel,
  contracts,
  salary_col,
  personnel_id_col = "personnel_id",
  status_col = "employment_status",
  date_col = "ref_date",
  pensioner_value = "pensioner",
  keep_vars = NULL,
  personnel_dt = NULL,
  contract_dt = NULL
)

Arguments

personnel

A data.table (or tibble/data.frame) containing at minimum the columns named in id_col, status_col, and date_col.

contracts

A data.table (or tibble/data.frame) containing at minimum the columns named in id_col, date_col, and salary_col.

salary_col

A single string naming the compensation column to use, e.g. "gross_salary_def" (default), "base_salary_lcu", etc.

personnel_id_col

A single string naming the personnel identifier column. Defaults to "personnel_id".

status_col

A single string naming the employment status column inside personnel_dt. Defaults to "employment_status".

date_col

A single string naming the snapshot/reference date column. Defaults to "ref_date".

pensioner_value

A single string giving the value of status_col that identifies a pensioner record. Defaults to "pensioner".

keep_vars

A character vector of additional contract-level columns to attach via govhr::add_contract_to_event(). Defaults to NULL

personnel_dt

Deprecated. Use personnel instead.

contract_dt

Deprecated. Use contracts instead.

Value

A data.table with one row per retiring individual containing the id_col identifier, ref_date_active (last active date), last_salary, ref_date_pension (first pension date), first_pension, and replacement_rate.

Details

The replacement rate is a standard diagnostic in public sector pension and workforce analysis, and it matters here for a few distinct reasons:

  • Fiscal sustainability: Aggregated across occupation or paygrade, replacement rates feed directly into pension liability projections.

  • Retirement incentive / take-up behavior: Low replacement rates help explain deferred retirement, relevant to calibrating ANNUAL_TAKE_UP rather than assuming 100\ eligibility.

  • Equity diagnostics: Comparing rates across paygrade, occupation, or establishment can surface structural inequities in how the pension formula interacts with career trajectories.

  • Policy reform simulation: Because the function is column-name agnostic, it can be re-run under counterfactual salary or pension formulas or against differently structured client datasets without code changes.