Identify Treatment Groups for Impact Evaluation — identify_treatment

Identifies people who experienced treatment events and collects their complete employment histories for impact evaluation. Uses a person-centered approach that compares entire career trajectories of treated vs. control individuals.

Usage

identify_treatment_events(
  data,
  treatment_conditions,
  event_window = c(-365, 365),
  min_pre_period = 90,
  min_post_period = 90,
  multiple_events = "first",
  require_employment_before = TRUE,
  id_column = "cf",
  date_column = "inizio",
  verbose = FALSE
)

Arguments

data

A data.table containing employment records (output from vecshift)

treatment_conditions

List of conditions defining treatment events. Each condition can be:

A string expression (e.g., "COD_TIPOLOGIA_CONTRATTUALE == 'C.01.00'")
A named list with 'column', 'operator', and 'value'
A function that takes the data and returns a logical vector

event_window

Numeric vector of length 2 defining the event window relative to the treatment date. Default: c(-365, 365) (1 year before/after)

min_pre_period

Minimum number of days required before the event. Default: 90

min_post_period

Minimum number of days required after the event. Default: 90

multiple_events

How to handle multiple treatment events per person:

"first": Use only the first treatment event date
"last": Use only the last treatment event date
"all": Keep all treatment events (creates multiple treatment dates per person) Default: "first"

require_employment_before

Logical. Require employment before the treatment event? Default: TRUE

id_column

Character. Name of the person identifier column. Default: "cf"

date_column

Character. Name of the date column to use for event timing. Default: "inizio"

verbose

Logical. Print debugging information? Default: FALSE

Value

A data.table with ALL employment events for both treated and control people including:

cf: Person identifier
is_treated: Logical indicator: TRUE for people who experienced treatment, FALSE for controls
treatment_event_date: Date of first/last treatment event (NA for control people)
days_to_event: Days from each observation to treatment event (negative = before)
in_event_window: Logical indicator for observations within event window
pre_event_period: Logical indicator for pre-treatment observations
post_event_period: Logical indicator for post-treatment observations
treatment_condition_met: Description of which condition triggered treatment
...: All original employment data columns

Examples

if (FALSE) { # \dontrun{
# Identify people who got permanent contracts and collect their full careers
impact_data <- identify_treatment_events(
  data = employment_data,
  treatment_conditions = list("COD_TIPOLOGIA_CONTRATTUALE == 'C.01.00'"),
  event_window = c(-180, 365),
  multiple_events = "first"
)

# Result contains:
# - ALL employment events for people who got permanent contracts (is_treated=TRUE)
# - ALL employment events for people who never got permanent contracts (is_treated=FALSE)
# - treatment_event_date marks when treated people first got permanent contracts

# Multiple conditions example
impact_data <- identify_treatment_events(
  data = employment_data,
  treatment_conditions = list(
    list(column = "COD_TIPOLOGIA_CONTRATTUALE", operator = "==", value = "C.01.00"),
    list(column = "durata", operator = ">", value = 365)
  ),
  event_window = c(-365, 730)
)
} # }