Skip to contents

Compute Contract Quality Indicators for Temporal Periods

Source: R/temporal_indicators.R
dot-compute_contract_quality.Rd

Calculates data-driven contract quality metrics within each temporal period using survival analysis methodology. Contract quality is determined by the median survival duration of each contract type - contracts with longer median survival times receive higher quality scores (0.3-1.0 scale). This approach replaces arbitrary fixed weights with empirical employment durability metrics derived from actual survival data.

Usage

.compute_contract_quality(
  data,
  period_type = "quarterly",
  id_var = "cf",
  contract_type_var = "COD_TIPOLOGIA_CONTRATTUALE",
  employment_intensity_var = "prior",
  employment_var = "arco",
  duration_var = "durata",
  start_date_var = "inizio",
  end_date_var = "fine",
  reference_date = NULL
)

Arguments

data

data.table with employment records

period_type

Character. "quarterly" or "monthly"

id_var

Character. Individual identifier column. Default: "cf"

contract_type_var

Character. Contract type column. Default: "COD_TIPOLOGIA_CONTRATTUALE"

employment_intensity_var

Character. Employment intensity column (1=full-time, <1=part-time). Default: "prior"

employment_var

Character. Employment status column (arco >= 1 = employed). Default: "arco"

duration_var

Character. Duration column in days. Default: "durata"

start_date_var

Character. Contract start date column. Default: "inizio"

end_date_var

Character. Contract end date column. Default: "fine"

reference_date

Date. Reference date for periods. Default: NULL (min start date)

Value

data.table with contract quality metrics by person and period including:

  • avg_contract_quality: Duration-weighted average quality score (0.3-1.0)

  • total_employment_days: Total employment days in period

  • n_contracts: Number of contracts in period

  • pct_permanent: Proportion of time in permanent contracts (A.01.00)

  • pct_full_time: Proportion of time in full-time employment

  • median_survival_based_quality: Flag indicating survival-based calculation

Details

The function performs Kaplan-Meier survival analysis on each contract type to estimate median survival times, then normalizes these to create quality scores where:

  • Longest surviving contract types receive quality score = 1.0

  • Shortest surviving contract types receive quality score = 0.3

  • Intermediate contract types receive linearly scaled scores

  • Unknown contract types receive default score = 0.3

  • Part-time contracts receive 0.8x adjustment factor

Survival Analysis Method: Uses estimate_contract_survival_optimized() to compute Kaplan-Meier survival curves for each contract type. Contracts ending at the maximum observation date are treated as censored observations.

Fallback Mechanism: If survival analysis fails, uses simple duration-based quality scores calculated from mean contract durations by type.

Quality Score Range: All scores are bounded between 0.3-1.0 to ensure meaningful differentiation while avoiding extreme values that could dominate analyses.