The Public Service Division (PSD) is reviewing an academic working paper alleging that Singapore civil servants disproportionately purchased homes near planned Mass Rapid Transit (MRT) stations before public announcements were made, according to The Online Citizen. The PSD stated that it is examining the data and methodology within the study and will refer any issues to the Corrupt Practices Investigation Bureau if a material basis exists.
Public Service Division Reviews Study on Civil Servants’ Home Purchases
We take any integrity concerns extremely seriously,
a spokesperson for the department said, as reported by Channelnewsasia. The research paper is titled Do Social Norms Substitute for Enforcement? Evidence from Public Officials' Home Purchases in Singapore
and was published by the National Bureau of Economic Research, a United States-based private non-profit research organisation, as detailed by Mothership.
Data and Methodology of the Working Paper
To conduct the study, the researchers assembled private-property transactions from legal filings lodged with the Registry of Land Titles between January 1995 and December 2019. They linked property buyers to the Singapore Government Directory Interactive using snapshots from 2007, 2009, 2011, 2013, 2015, and 2022 to identify civil servants. Station locations for MRT lines announced between 1999 and 2019 were geocoded, and properties within specific distances—such as one kilometre—were classified as target properties.
The researchers also integrated data from the Council for Estate Agencies’ public register of licensed real estate agents and used OpenAI’s GPT 4.0 large language model to classify civil-service job titles into high-rank, mid-rank, and low-rank tiers. According to the paper, civil servants were matched with non-civil servant control buyers sharing similar age, gender, marital status, ethnicity, and residential postal codes.
Findings on Pre-Announcement Purchases and Concentration
The working paper reports that civil servants purchased target properties at higher intensities than matched controls, with the activity concentrated one to two years before official public announcements. The estimated differences were reported as 0.674 percentage points one year before announcement and 0.535 percentage points two years prior. The authors calculated roughly 113 additional purchases across these two pre-announcement years, representing about S$138 million in gross acquisition value at a mean informed-purchase price of S$1,224,150.
The study indicates that this purchasing behaviour was driven primarily by mid-ranking officers and agencies connected to rail planning. Furthermore, survival analysis suggested these properties were sold sooner than comparable transactions, which the authors interpreted as consistent with an investment motive rather than residential use.
Academic Acknowledgements and Reactions
The working paper’s acknowledgements section thanked several current and former academics for helpful comments and discussions, including professors from the National University of Singapore, Harvard University, and the University of Chicago. Following the paper’s release, National University of Singapore professors Ivan Png and Jessica Pan stated they knew nothing of the working paper and asked the authors to remove their names from the acknowledgements. Professor Png noted he was puzzled by the inclusion, while Professor Pan stated she was named by mistake.
