Compute growth decomposition of wagebill

Usage

compute_growth_decomposition(
  data,
  group_cols = NULL,
  measure_col = "gross_salary_lcu"
)

Arguments

data: A data frame containing the data to be analyzed. It should include columns for the grouping variables, a column for the reference date, and a column for the measure of interest (e.g., gross salary).
group_cols: A character vector specifying the names of the columns to group by.
measure_col: A string specifying the name of the column containing the measure of interest (default is "gross_salary_lcu").

Value

A data.table with headcount, compensation, wagebill, wagebill_lag, the continuing-period decomposition (employment/compensation/interaction effects), entry/exit effects, is_observed, and a transition_type label for each row: "start" (panel's first period for this group – left- censored, no baseline available), "continuing" (observed this period and last), "entry" (observed now, not last period – a genuinely new group_cols combination, or a reappearance after any length of absence), or "exit" (not observed now – covers both the period a group first disappears, which carries the real dollar effect, and every subsequent period it remains absent, which correctly carries a zero effect since nothing further changed).