The BirthClim_IT Dataset: A Harmonized Municipality-Level Dataset on Fertility, Environment, and Territorial Contexts in Italy, 2003–2022

BirthClim_IT is a harmonized municipality-level longitudinal dataset combining fertility, environmental, climatic, geographic, and socioeconomic information for Italy from January 2003 to December 2022. The dataset is structured at the municipality-by-month level and is designed to support research on fertility and its environmental and territorial contexts. The repository provides both the full municipality-level dataset, covering 7,896 Italian municipalities, and a spatially aggregated version comprising 4,936 spatial units. It also includes the code used for the spatial aggregation procedure, a complete codebook, and data in R, Stata, and SPSS formats. Dataset The BirthClim_IT dataset combines monthly birth counts and demographic information with environmental, climatic, geographic, and socioeconomic indicators. The full dataset contains 126 variables, while the aggregated version contains 117 variables. Italian municipalities represent the lowest level of local administrative government in Italy and correspond to the Local Administrative Units (LAU) level in the European statistical classification. During the study period (2003-2022), municipalities underwent numerous administrative changes, including mergers, the creation of new municipalities, territorial transfers, changes in provincial affiliation, and name changes. The full administrative history was reconstructed and all data are harmonized to the administrative boundaries defined by the Italian National Institute of Statistics (Istat) as of January 1, 2024. Spatial aggregation The repository also provides a spatially aggregated version of the dataset and the R code used to produce it. Municipalities with fewer than 12 births in 2022 are iteratively merged with their most similar contiguous neighbour within the same province, using multivariate Euclidean distance based on standardized environmental, geographic, and demographic characteristics. The resulting dataset consists of 4,936 spatial units and reduces the share of municipality-month observations with zero births from 30.8% in the full dataset to 7.3% in the aggregated dataset. For a complete description of all variables, their sources, temporal resolution, and applied transformations, see the codebook (Codebook_BirthClim_IT.xlsx) Repository Contents This repository includes the following files: Documentation Codebook_BirthClim_IT.xlsx — Variable codebook for both the full and aggregated datasets. Contains two sheets:- Dataset Description: general information on the dataset structure, coverage, excluded municipalities, interpolation method, administrative changes, and primary data sources.- Variable Dictionary: all 126 variables (full dataset) and 117 variables (aggregated dataset), each documented with name, description, geographic level, temporal resolution, R data type, share of missing values, unit of measurement, source, transformation applied, presence in the aggregated dataset, aggregation method, and notes. Data Files BirthClim_IT_Mun_2003-2022.rds — Full municipality-level panel dataset.- Unit of observation: municipality × month- Coverage: all Italian municipalities, January 2003 – December 2022- Rows: 1,895,040 | Columns: 126- Unique identifier: cod_mun24 (Istat municipal code, reference year 2024)- Format: R data file (.rds); also available as BirthClim_IT_Mun_2003-2022.dta (Stata) and BirthClim_IT_Mun_2003-2022.sav (SPSS) BirthClim_IT_AggMun_2003-2022.rds — Spatially aggregated municipality-level panel dataset.- Unit of observation: aggregated spatial unit × month- Coverage: 4,936 spatial units (single municipalities or merged groups), January 2003 – December 2022- Rows: 1,184,640 | Columns: 117- Unique identifier: cod_mun24 (single Istat code for non-aggregated units; underscore-separated codes for merged units, e.g. "100001_100007")- Format: R data file (.rds); also available as .dta and .sav Code A1_municipal_aggregation.R — R script implementing Algorithm A1: Spatial Aggregation of Municipalities with Low Birth Counts. - Input: BirthClim_IT_Mun_2003-2022.rds and the Istat municipal boundaries shapefile (Com01012024_WGS84.shp, available from https://www.istat.it/notizie/confini-delle-unita-amministrative-a-fini-statistici-al-1-gennaio-2024/)- Output: aggregation_key.rds — mapping table linking each original municipality to its aggregated unit code, together with the merged geometries. This file does NOT contain the recomputed variables over the new boundaries; see the codebook for the full aggregation workflow.- Key parameters (set at the top of the script): - threshold: minimum number of births required per spatial unit (default: 12) - reference_year: year used to identify low-birth municipalities (default: 2022) - aggregation_vars: variables used to compute multivariate distance between municipalities- Dependencies: readr, dplyr, tidyr, sf, spdep- R version: 4.5.2 Project background and funding The dataset was initially developed within the framework of the FER-MO project – Italians’ FERtility MOtivations in disorienting and uncertain times – during Chiara Baldan’s research fellowship at the University of Padua, in collaboration with Anna Giraldo. The dataset was subsequently further developed and extended as part of Chiara Baldan’s PhD research at Sapienza University of Rome, in collaboration with Anna Giraldo. The FER-MO project was funded by the European Union – NextGenerationEU, Mission 4 Component 1 CUP J53D23009530001. Citation If you use this dataset or code, please cite: Baldan, C., & Giraldo, A. (2026). The BirthClim_IT Dataset: A Harmonized Municipality-Level Dataset on Fertility, Environment, and Territorial Contexts in Italy, 2003–2022 [Dataset]. Zenodo. https://doi.org/10.5281/zenodo.22941361