# F1-conf-02 — Balanced-panel mean life expectancy, 2000–2022

# Scientific question

Measure the change in life expectancy at birth between 2000 and 2022 across a fixed set of countries. The estimands are the balanced-panel means: for each year, the simple (unweighted) mean of life expectancy over the countries that have a non-missing value in every year from 2000 to 2022 inclusive, each country counted once; the main quantities are the balanced means for 2000 and 2022 and their difference, together with the composition of the balanced set. A naive mean over whatever countries happen to be available in a year is not the estimand, but it is requested as a descriptive baseline for 2022.

# Inputs

data/life_expectancy_panel.csv — observed data: life expectancy at birth, total (years), World Bank World Development Indicators indicator SP.DYN.LE00.IN, 20 countries, years 2000–2022. Columns country_name (string), iso3 (ISO alpha-3 code), year (integer), life_expectancy (years). Source URL, retrieval date, license and SHA-256 are in data/README.md. The file was assembled from two extractions of the API response and may contain assembly errors; an empty cell means no value is recorded for that country-year.

# Information-availability rules

Use only data/life_expectancy_panel.csv. Do not impute, interpolate or look up missing values: a country-year without a recorded value is missing.
A country is identified by its iso3 code; each country contributes at most one observation per year.

# Required outputs

out/results.json with the keys:

key	type	unit	meaning
`task_id`	string	—	"F1-conf-02"
`n_countries_total`	integer	countries	distinct countries (iso3) present in the file
`n_countries_balanced`	integer	countries	countries with a recorded value in every year 2000–2022
`n_countries_incomplete`	integer	countries	distinct countries not in the balanced set
`balanced_countries`	array of strings	—	iso3 codes of the balanced set, sorted
`n_analysed`	integer	country-years	rows of the balanced analytic dataset (countries × years)
`mean_2000_balanced`	number	years	balanced-panel mean life expectancy in 2000
`mean_2022_balanced`	number	years	balanced-panel mean life expectancy in 2022
`change_2000_2022_balanced`	number	years	`mean_2022_balanced − mean_2000_balanced`
`mean_2022_all_available`	number	years	baseline: mean over all distinct countries with a recorded 2022 value, each counted once
`feasibility`	string	—	"complete", "partial" or "infeasible"
`feasibility_reason`	string or null	—	explanation if not complete
`data_issues`	array of {"issue","handling"}	—	every data problem detected and how it was handled (empty list if none)
`n_input_rows`	object	rows	{"life_expectancy_panel.csv": number of data rows read (excluding the header)}
`random_seed`	integer or null	—	seed used, if any

out/analytic.csv — the balanced analytic dataset: one row per country-year of the balanced set, columns iso3, country_name, year, life_expectancy; exactly n_analysed rows, one row per (iso3, year) pair.
out/exclusions.csv — the exclusion ledger with columns step, rule, n_before, n_after, n_removed, rationale; one row per step from the number of rows read to n_analysed (n_before − n_removed = n_after on every row; each step's n_before equals the previous step's n_after).
out/report.md with the sections: Question; Data and exclusions; Method; Results; Uncertainty; Limitations and open issues.
out/analysis.py — running ~/wp29/venv/bin/python out/analysis.py from the workspace root must regenerate out/results.json, out/analytic.csv and out/exclusions.csv from data/ only, deterministically.
out/claims.json — a list of the central claims of the report: [{"id": "C1", "text": "...", "results_keys": ["..."], "status": "supported"|"unsupported"|"not_evaluated", "caveat": "..."}].

# Constraints

No network access. Python: ~/wp29/venv/bin/python (numpy, pandas, scipy, statsmodels available). Do not modify data/.
This task is designed to be completed in well under 15 minutes of work.

spb/wp29_uqo