# An open-source and spatially diverse synthetic population dataset for agent-based modelling and microsimulation in Ireland

## TL;DR

This paper introduces a synthetic population dataset for Ireland, enabling detailed socio-economic modeling while preserving privacy.

## Contribution

The paper presents a spatially diverse synthetic population dataset for Ireland using four generation methods.

## Key findings

- A synthetic population dataset for Ireland was created using the Labour Force Survey and Census data.
- Individuals in the dataset have six characteristics matched to real aggregate statistics.
- The dataset supports agent-based modeling and microsimulation at the Electoral Division level.

## Abstract

Spatial microsimulations, where simulation units represent people or households in a small area, are extremely useful for modelling a wide range of socio-economic scenarios at a fine scale. The characteristics of individuals in these simulations' populations need to accurately represent the real characteristics of the target area to model realistic scenarios. However, individual-level data is not available for the vast majority of populations, Ireland included, due to privacy concerns. Thus, a representative synthetic population for the Republic of Ireland is needed. The data from four methods of generating synthetic populations at the Electoral Division level are given in this paper. Realistic individuals are created by sampling from the Central Statistics Office (CSO) Labour Force Survey. Spatial heterogeneity is achieved by matching the aggregate counts of individuals' characteristics to those from the CSO Census Small Area Population Statistics. Individuals are assigned six characteristics: age group, sex, marital status, house size, primary economic status, and highest level of education achieved.

Image, graphical abstract