DATASET

Data & Downloads

SynPop-DE provides 40,235,916 synthetic households and 81,629,116 persons across all 400 German districts. The data can be downloaded as CSV or Parquet, or read directly via URL.

82MSynthetic persons

40MHouseholds

400Districts

Downloads

400 Parquet files, one per district, plus the full national dataset.

License CC BY 4.0 Free to use with attribution.

Cite Napiontek, J. & Pichler, P.-P. (2026). SynPop-DE: Synthetic population of 40 million German households using generative neural networks. SocArXiv. https://doi.org/10.31235/osf.io/zha8v_v1

Download by District

All 45 attributes per person and household. File size: ~15–30 MB per district.

The download list requires JavaScript. All files are directly accessible at: data.synpop.de

Full Dataset

All 400 districts in one file. Best for cross-district analyses.

Full dataset (.parquet) (9.1 GB) Zenodo archive

DuckDB and R (arrow) stream via HTTPS byte-range requests — no full download needed. pandas always downloads the entire file for a URL — use a district file instead, or pass fsspec's HTTP filesystem (see the Python example below).

CSV format is available per district only (section above). The national dataset is available as Parquet or via DuckDB/Arrow streaming.

Machine-readable index: catalog.json · DOI: 10.5281/zenodo.20439915

Direct Access

All Parquet files are accessible directly via URL. DuckDB and R (arrow) read only the queried rows and columns via byte-range requests, with no full-file download. pandas needs an explicit HTTP filesystem (fsspec) for this — without it, pandas downloads the entire file.

For filtered or programmatic export. Ready-to-download CSV files are available in the district picker above.

Data Preview

4,000 sample households · 45 attributes

Explore the dataset with 10 random households from every region and all their attributes.

		Household				Persons
		AGS	Building type	Floor Area (m²)	Household income (€/yr)	Age Group	Gender	Education
No rows for this district.

Total households: 0

Page 1 / 1

Data Schema

All 45 columns in the Parquet files — one row per person, household attributes shared within each household.

Column	Level	Type	Description / Values
Identifiers
`household_id`	ID	string	Unique household identifier
`person_id`	ID	string	Unique person identifier
`person_rank`	ID	int	Person's position within household (1 to max. 8)
`ags`	ID	string	AGS district code (5-digit, e.g. "11000" for Berlin)
`state`	ID	int	Federal state — 1=Schleswig-Holstein, 2=Hamburg, 3=Niedersachsen, 4=Bremen, 5=Nordrhein-Westfalen, 6=Hessen, 7=Rheinland-Pfalz, 8=Baden-Württemberg, 9=Bayern, 10=Saarland, 11=Berlin, 12=Brandenburg, 13=Mecklenburg-Vorpommern, 14=Sachsen, 15=Sachsen-Anhalt, 16=Thüringen
`region_type`	ID	int	BKG settlement type — 1=Urban regions, 2=Regions with urbanisation tendencies, 3=Rural regions
`original_household_id`	ID	string	Synthetic household ID from the generation donor pool — no link to real EVS households or their IDs.
`source_household_id`	ID	string	Synthetic source household ID from the generation donor pool — no link to real EVS households or their IDs.
Person attributes
`gender`	Person	int	Gender — 1=Male, 2=Female
`age_group`	Person	int	Age group — 0=0–5, 1=6–14, 2=15–17, 3=18–24, 4=25–29, 5=30–39, 6=40–49, 7=50–64, 8=65–74, 9=75+
`education`	Person	int	Educational attainment — 1=No vocational qualification, 2=Apprenticeship / vocational training, 3=Master craftsperson / advanced vocational, 4=University degree, 5=Doctoral degree (PhD)
`employment`	Person	int	Employment status — 1=Unemployed, 2=Inactive, 3=Employed (Official), 4=Employed (Dependent), 5=Employed (Self-employed)
Household attributes
`household_size`	Household	int	Number of persons in household
`household_type`	Household	int	Household type (child definition: single children under 18) — 1=Single person, 2=Couple without children, 3=Couple with children, 4=Single parent, 5=Other
`household_type_27`	Household	int	Household type as household_type, but with an extended child definition: single children under 27 — 1=Single person, 2=Couple without children, 3=Couple with children, 4=Single parent, 5=Other
`building_type`	Household	int	Building type — 1=Single family, 2=Two family, 3=Multi family (3+), 4=Other
`building_ownership`	Household	int	Ownership — 1=Owner-occupied, 2=Rented
`building_age`	Household	int	Construction period — 1=Before 1949, 2=1949–1990, 3=1991–2000, 4=After 2000
`building_size`	Household	float	Living floor area in m²
`heating_type`	Household	int	Heating system — 1=District heating, 2=Central heating, 3=Single heating
`heating_energy`	Household	int	Primary energy carrier — 0=None / not applicable, 1=Electricity, 2=Gas, 3=Heating oil, 4=Wood / Coal, 5=Renewables
`household_income`	Household	int	Income decile — 1 (lowest) to 10 (highest)
`household_income_num`	Household	float	Annual net household income (€)
`aeq`	Household	float	Adult equivalents per the OECD-modified equivalence scale: first adult 1.0, each additional person aged 15+ counts 0.5, each child (0–14) counts 0.3 (e.g. 2 adults + 2 children = 2.1)
`exp_quota`	Household	float	Consumption ratio — ratio of total annual household expenditure to annual net household income
Expenditures (€/year)
`expenditure`	Household	float	Total annual household expenditure in EUR (sum of all expenditure categories after calibration)
`exp_food`	Household	float	Food and beverages
`exp_clothes`	Household	float	Clothing and footwear
`exp_housing_total`	Household	float	Total housing costs
`exp_housing_rent`	Household	float	Dwelling rent (including operating costs)
`exp_housing_electricity`	Household	float	Electricity
`exp_housing_heating`	Household	float	Heating fuel
`exp_housing_imputed`	Household	float	Imputed rent (including ancillary costs)
`exp_housing_maintenance`	Household	float	Maintenance and repair
`exp_transport_total`	Household	float	Total transport
`exp_transport_fuel`	Household	float	Fuel
`exp_transport_own`	Household	float	Own vehicle purchase and leasing
`exp_transport_public`	Household	float	Public transport
`exp_health`	Household	float	Health
`exp_leisure`	Household	float	Leisure and culture
`exp_telecom`	Household	float	Telecommunications
`exp_furniture`	Household	float	Furniture and home equipment
`exp_hospitality`	Household	float	Restaurants and accommodation
`exp_education`	Household	float	Education
`exp_other`	Household	float	Other goods and services