Abstract
Background: Socioeconomic status (SES) is an important determinant of health, and SES data is an important confounder to control for in epidemiology and health services research. Individual level SES measures are cumbersome to collect and susceptible to biases, while area level SES measures may have insufficient granularity. The ‘Singapore Housing Index’ (SHI) is a validated, building level SES measure that bridges individual and area level measures. However, determination of the SHI has previously required periodic data purchase and manual parsing. In this study, we describe a means of SHI determination for public housing buildings with open government data, and validate this against the previous SHI determination method.
Methods: Government open data sources (e.g. data.gov.sg, Singapore Land Authority OneMAP API, Urban Redevelopment Authority API) were queried using custom Python scripts. Data on residential public housing block address and composition from the HDB Property Information dataset (data.gov.sg) was matched to postal code and geographical coordinates via OneMAP API calls. The SHI was calculated from open data, and compared to the original SHI dataset that was curated from non-open data sources in 2018.
Results: 10,077 unique residential buildings were identified from open data. OneMAP API calls generated valid geographical coordinates for all (100%) buildings, and valid postal code for 10,012 (99.36%) buildings. There was an overlap of 10,011 buildings between the open dataset and the original SHI dataset. Intraclass correlation coefficient was 0.999 for the two sources of SHI, indicating almost perfect agreement. A Bland-Altman plot analysis identified a small number of outliers, and this revealed 5 properties that had an incorrect SHI assigned by the original dataset. Information on recently transacted property prices was also obtained for 8,599 (85.3%) of buildings.
Conclusion: SHI, a useful tool for health services research, can be accurately reconstructed using open datasets. This method provides more updated data at lower cost. This application highlights the potential for leveraging open data to enable healthcare research.
Introduction
Socioeconomic status (SES) is a fundamental determinant of health outcomes, yet measuring it accurately remains a challenge in health services research. Researchers often face a trade-off between individual-level data—which is sensitive, prone to self-reporting bias, and difficult to scale—and area-level data, which can be too coarse and lead to the "ecological fallacy."
The Singapore Housing Index (SHI), also known as the "room-index," offers a granular, building-level alternative. By assigning an SES score based on the average number of rooms per unit in a specific building, the SHI provides a reliable proxy for asset-based SES in Singapore’s unique housing landscape. However, maintaining the SHI has historically been resource-intensive, requiring the purchase of proprietary datasets and manual data cleaning. This study demonstrates an automated, low-cost workflow to reconstruct the SHI using government open data.
Methods
The researchers developed a programmatic workflow using Python to retrieve and process data from several public Singaporean government APIs. The process involved four primary stages:
- Data Retrieval: The
data.gov.sgportal provided HDB (Housing Development Board) property information, detailing the unit composition (e.g., number of 1-room vs. 5-room flats) for every public housing block. - Geocoding: To bridge the gap between address data and spatial analysis, the SLA OneMAP API was used to match block addresses to precise geographic coordinates and postal codes.
- SHI Calculation: For public housing, the SHI was calculated as a weighted average of rooms per unit, capped at a maximum value of 6.
- Classification and Exclusion: Addresses were categorized into Public Housing, Private Housing (assigned a default SHI of 6), and "Excluded" categories (such as nursing homes or destitute homes) using data from the Ministry of Health (MOH) and Singapore Statutes Online.
The reconstructed dataset was validated against a 2018 proprietary SHI dataset using Intraclass Correlation Coefficients (ICC) and Bland-Altman plots to ensure accuracy.
Results and Validation
The automated workflow successfully identified 10,077 unique residential buildings. The OneMAP API demonstrated high reliability, returning valid coordinates for 100% of the buildings and postal codes for 99.36%.
The validation against the original proprietary dataset showed near-perfect agreement, with an ICC of 0.999. Interestingly, the open-data approach proved more accurate in some instances; the validation process identified five properties in the original "gold standard" dataset that had incorrect SHI values due to manual entry errors.
Furthermore, the SHI was validated against market values. By analyzing 8,599 recent resale transactions from the HDB and URA APIs, the researchers found a strong monotonic relationship: higher SHI levels consistently correlated with higher median log resale prices, confirming the index's utility as an economic proxy.
Discussion and Conclusion
This study highlights the power of open data to democratize health services research. By replacing expensive, static datasets with dynamic API-driven workflows, researchers can maintain highly accurate, up-to-date SES markers at minimal cost.
While the current method is highly effective for public housing, it has limitations regarding private housing granularity. Currently, all private residences are coded as SHI 6, which may mask socioeconomic diversity within the private sector (e.g., comparing luxury condominiums to smaller apartments). Future work could refine these markers by incorporating more detailed private property transaction data or developing composite indices using Principal Component Analysis (PCA).
In conclusion, the SHI is a robust tool for controlling for socioeconomic confounding in Singapore-based research. Its reconstruction via open data ensures that this valuable metric remains accessible and current for the research community.
Summary updated