Side project / 2026
The Pay Packet
An open-source explainer that makes Singapore's ministerial salary framework legible without presenting benchmarks as individual pay.
Context
Singapore’s ministerial salaries are widely discussed and rarely easy to inspect. The public framework combines a published reference point with fixed pay, an annual variable component, and performance and national bonuses. A headline figure can be technically accurate while still suggesting something stronger than the underlying policy supports.
I built The Pay Packet as an independent, open-source explainer. It lets a reader choose an MR4 minister or the Prime Minister, move the assumptions, and see how the components add up. The project is not an official government service or a disclosure of any individual’s remuneration.
Problem
The job was to make the framework inspectable without replacing it with a simpler but misleading story.
The most consequential distinction is between a revised benchmark and immediate pay. The 2026 review sets a new MR4 benchmark of S$1.8 million, while existing officeholders receive a one-off adjustment of up to 9% from 15 October 2026, depending on individual circumstances. Those are related, but they are not interchangeable. The calculator had to put that distinction where a reader would see it, not leave it as a footnote.
What I shaped
The main interface starts with a compact calculation rather than a policy document. Readers can compare the previous and 2026 benchmark frameworks, select a role, and adjust bonus assumptions. The breakdown then shows fixed and variable components separately.
The same total can be viewed as cash, chicken rice, or annual median household income. These are deliberately illustrative scales rather than attempts to make unlike things equivalent. The household-income view is based on SingStat’s 2025 median monthly household market income of S$12,446, annualised to S$149,352. For very small totals, the cash view switches to S$2,000 copper coins and the household view reveals a partial HDB block. The numbers remain visible when the images stop being useful.
I also included a personal-salary comparison. It stays in the browser and uses the same illustrative assumptions. It is a way to explore the model, not an employment contract.
Key decisions and trade-offs
Make the caveat part of the interaction. The revised benchmark and the October implementation adjustment appear together. A calculator that only showed the S$1.8 million figure would be easier to explain, but would hide the point most likely to be misunderstood.
Use source-backed reference points, not personal claims. The implementation uses published PSD, PMO and SingStat material, linked in the project’s expandable methodology. It models published norms and assumptions, not actual salaries, grades, or bonuses for named officeholders.
Keep detailed controls optional. The first control changes linked bonuses. Individual performance and national-bonus controls sit behind an expandable section for readers who want to test a scenario without making the initial view feel like a spreadsheet.
Publish the method with the tool. The code, tests, source links and CC BY-NC 4.0 licence are public. Others can inspect or adapt the work non-commercially with attribution. Policy sources remain subject to their own terms.
What launched
The project is available on GitHub, with a TypeScript calculation layer and tests for the salary, comparison, and input-handling logic. Its methodology records the figures’ verification date, 15 September 2026, so readers can judge whether a later policy change needs a refresh.
The page makes one thing easier: seeing what the numbers mean before arguing about them.