# Avoid the three major pitfalls in Romanian public data

Distinguish domicile from residence, record series breaks and handle geoblocking.

## 1. Choose the right population denominator

For any per-capita calculation (doctors, crimes or income per 1,000 residents), use census resident population (around 19 million), not domicile registration from IDs (over 21.8 million).

[Locality population in the 2021 census](https://mariuscomper.uk/cunoastere-deschisa/en/resurse/rpl-2021-localitati/)

## 2. Verify SIRUTA locality identifiers

Use numeric SIRUTA codes instead of locality names. Communes encompass component villages, and identical names across different counties trigger frequent join errors.

[SIRUTA 2024 — locality identifiers](https://mariuscomper.uk/cunoastere-deschisa/en/resurse/siruta-2024/)

## 3. Overcome network access barriers

When governmental servers restrict foreign IP ranges or return timeouts, rely on locally archived copies or national routing, documenting the exact attempt timestamp.

[data.gov.ro](https://mariuscomper.uk/cunoastere-deschisa/en/resurse/data-gov/)

## Caution

Computing per-capita indicators with domicile counts instead of resident population underestimates studied rates by 10% to 20%, due to counting millions of citizens permanently living abroad.

A methodological compliance note detailing the selected population baseline, SIRUTA codes and source access parameters.