What hawkers earn
Nobody measures what a hawker takes home. So this page pieces it together from what can be checked – rent bids, recipe sales, hawkers' own figures in the press and some anecdote – and grades each piece for how far it can be trusted.
The answer
At the middle, one stall's household takes home about $2,700–3,500 a month. Above the middle, take-home rises slowly to about $5,500–8,000 at the top tenth. Then it climbs steeply: the top 1% of stalls may keep $15,000–30,000 a month.
These are household figures for one stall, in 2026 prices, and often two people's work. Read the bottom half as fairly firm, the top tenth as indicative, and the top 1% as a plausible range rather than a measurement.
Singapore dollars a month, take-home for one stall's household, in 2026 prices.
The shape, percentile by percentile
What one stall's household takes home a month, by percentile
Ranges from the graded evidence below, in 2026 prices. The dots are the Library's own model (the Technical Paper, G.3), which stops at the top tenth.
- P10
Estimate $0–1,000 (fairly firm) · Model (G.3) $500
- P25
Estimate $1,300–2,000 (fairly firm) · Model (G.3) $1,570
- P50
Estimate $2,700–3,500 (fairly firm) · Model (G.3) $2,720
- P75
Estimate $4,000–5,500 (indicative) · Model (G.3) $3,930
- P90
Estimate $5,500–8,000 (indicative) · Model (G.3) $5,140
- P95
Estimate $8,000–12,000 (a plausible range) · Model: none (no top tail)
- P99
Estimate $15,000–30,000 (a plausible range) · Model: none (no top tail)
- The estimate's range
- The Library's model (the Technical Paper, G.3)
| Percentile | Estimate | How firm | The model (G.3) |
|---|---|---|---|
| P10 | $0–1,000 | Fairly firm | $500 |
| P25 | $1,300–2,000 | Fairly firm | $1,570 |
| P50 | $2,700–3,500 | Fairly firm | $2,720 |
| P75 | $4,000–5,500 | Indicative | $3,930 |
| P90 | $5,500–8,000 | Indicative | $5,140 |
| P95 | $8,000–12,000 | A plausible range | – |
| P99 | $15,000–30,000 | A plausible range | – |
The Library's own model agrees at the middle. It lets a stall's plates a day vary by only about 15% either way, so it has no top tail: its top tenth stops at $5,140 a month (the Technical Paper, G.3). Its numbers do not change.
How to read the evidence
Every data point carries one of five grades. The lower the grade, the less it should move a number on its own. Evidence grades, in the Glossary.
| Grade | What it is | How far to trust it |
|---|---|---|
| A · Official | Parliamentary replies, NEA tender results and cost surveys, MTI papers | Firm, but none measures income directly |
| B · Named in the press | A named hawker's own figures, reported by a mainstream outlet | Self-reported and unaudited; usually the stall's best days |
| C · Inferred from a price | Rent bids, recipe and stall sale prices, takeover listings, journalists' worked examples | A price someone paid or asked; the income behind it is an inference |
| D · Anecdote | Reddit, forums, blogs, unnamed sources | One person's account; useful only for the shape |
| E · Not evidence | Commercial "financial model" sites with generic figures | Left out: their numbers are not about Singapore stalls |
The Government has said plainly that it does not track hawkers' earnings, because hawkers are self-employed (MSE, April 2022). So no grade A source gives income itself.
The evidence
Twenty data points, newest first within each grade. The last column reads what each implies for one stall's monthly take-home. Where that is an inference, it says so.
| Source (date) | Grade | What it says | What it implies for take-home |
|---|---|---|---|
| Ministry of Sustainability and the Environment S.418 (Sep 2024) | A | Median winning cooked-food bid about $1,800 a month (2023). One stall in five let at $500 or less. Marine Parade's top five bids all above $8,000. | A fifth of stalls draw almost no competition: the bottom is weak. The top bidders expect a surplus well above $8,000 a month before rent. |
| Mothership S.433 (Oct 2026) | A + B | A record Maxwell bid of $7,911 a month, and five bids above $7,000. The current tenant, a drinks and breakfast stall, reports about $1,000 a day in sales (via Lianhe Zaobao). | About $26,000 a month in sales on 26 days. After rent and stock, our estimate is $6,000–10,000 kept (an inference). |
| Institute of Policy Studies S.434 (Jun 2024) | A | A Newton stall bid of $6,810 a month (Aug 2023); tenders as low as $1. | What stalls are worth runs from nothing to about $7,000 a month in rent alone. |
| Lee Kuan Yew School of Public Policy, NUS S.273 (2024) | A + C | A record drinks-stall bid of $10,028 a month at Chomp Chomp (2018). | A bidder at that level expects $13,000–15,000 a month before rent (an inference: the rent plus a working wage). |
| Ministry of Trade and Industry S.65 (2015) | A | NEA's cost survey of 2012–13: an average stall spends $8,000–9,200 a month; raw materials are 59% of it. | A typical stall turns over $10,000–13,000 a month and keeps a thin slice. |
| Leong Mun Wai S.435 (Nov 2024) | B | About half of stallholders received Workfare in 2020, a Member said in Parliament (no source given). | If right, about half took home under roughly $2,300 a month in a Covid year. |
| TODAY (republished by Malay Mail) S.296 (Jul 2024) | B | A roast-meat hawker at Amoy Street spends about $22,000 a month to run his stall (2024). Several hawkers report costs above $10,000 a month, and profits down by as much as 40% on five years earlier. | Strong stalls turn over $25,000 or more a month; margins were squeezed in 2022–24. |
| Al Jazeera S.436 (May 2020) | B | A Newton barbecue seafood stall took up to $1,800 a day before Covid. | About $45,000 a month in sales. With seafood costs, our estimate is $10,000–15,000 kept (an inference). |
| Yahoo News S.437 (undated) | B | A Michelin-listed nasi lemak brand takes $800–4,000 a day an outlet. Its founder took under $5,000 a month in its first months; it now has 31 outlets. | One successful stall can reach $100,000 a month in sales, and then becomes a chain. |
| Malay Mail S.438 (Jun 2017) | B | A Bib Gourmand soup stall sold about 300 bowls a day; the listing lifted sales by 20–30%. | At $5–6 a bowl, about $40,000 a month in sales. Our estimate is $10,000–16,000 kept (an inference). |
| Vulcan Post S.439 (Jan 2021) | B | After the Michelin star, business grew three to four times; the stall capped itself at 200 plates a day. | Even famous stalls hit a ceiling of hands and hours. The top end is capped by capacity, not demand. |
| Tabla S.440 (Dec 2024) | B | A mutton soup stall on $1 rent makes about 10 bowls a day and $1,000 a month profit. | The bottom: a near-free stall run as a part-time craft. |
| Mothership S.56 (Jun 2022) | C | An 80-year Maxwell stall asked $1 million for its recipe and business, then cut it to $500,000. The seller said a buyer could recoup it in three years. | The seller's claim implies about $14,000–28,000 a month in profit. It is a seller's claim, so read it as a ceiling. |
| Mothership S.57 (Aug 2023) | C | A 57-year wanton mee stall turned down bids of $600,000–800,000 for its recipe. | Buyers valued a heritage brand at several years of a strong stall's profit. |
| Mothership S.58 (Aug 2022) | C | A 37-year bak chor mee stall asked $500,000 for its recipe. | As above: an asking price, not a sale. |
| Mothership S.59 (Aug 2020) | C | A well-known Hokkien mee stall sold its stall and recipes for $20,000. | Most heritage stalls are worth far less than the famous few. |
| SingSaver S.441 (c. 2016) | C | A worked example for a young hawker pair: 200 customers a day at $5, $22,000 a month in sales and $13,000 in costs. | About $9,000 a month for the stall – $4,500 each – over 13-hour days. |
| Mothership S.442; Mothership S.443; Home & Decor S.301 (May 2022 – Apr 2026) | C | Coffeeshop stall rents of $4,000–10,000 a month. A char kway teow stall paid $7,000–8,000; 10 of 14 stalls left one coffeeshop at $10,000. | Coffeeshop stalls that stay at $7,000–10,000 rent must clear more than that before rent. Most cannot. |
| Reddit S.445 (late 2025) | D | A hawker couple take home $8,000–10,000 together ($3,000–4,000 each, plus $2,000 for a car), working 6 am to 9 pm, six days a week. Another stall: $2,000–3,000 off-peak and $10,000 at peak. A hawker's assistant earns $2,200 at best, a cook $3,000. | Good stalls are two-person households, and take-home swings with the season. |
| KitchenUnionSG S.444 (Oct 2024) | D | Net $3,000–4,000 a month is typical (attributed to NEA, unverified); $2,000 or less for struggling stalls. | Consistent with the middle. The NEA attribution could not be found. |
Commercial model sites (financialmodelslab.com, for example) quote figures such as "$331K–189M" for a hawker stall. They are grade E and left out.
What drives the top end
The top stalls are not a different trade. They do the same sums at a larger scale, and a few things push them there.
- Plates, then price. Take-home is roughly plates times price times margin. A typical stall sells 150–200 plates a day; the famous ones sell several times that. Volume moves take-home more than price does, because a $1 rise can cost 30% of an item's sales (TODAY, 2024).
- Family hands. Most high earners are couples or families. Unpaid family labour is the margin: a stall that pays an assistant $2,000–2,500 a month gives up a third of a typical take-home.
- Hours and days. The $8,000–10,000 couple works 15-hour days, six days a week. Per hour, a top stall can earn less than an office job.
- Location and rent. Prime centres (Maxwell, Newton, Chomp Chomp, Amoy Street) carry tourist and office demand, and rents that only strong stalls can carry. A fifth of tenders draw $500 or less, so the weakest sites hold the weakest stalls.
- Drinks and kopi. Drinks carry the highest margins and sell with almost every meal. Two of the record bids above are for drinks stalls.
- Fame. A guide listing or a viral post adds 20–30% at once, until hands and hours cap it (Hawker Chan held at 200 plates a day).
- Seasons. Take-home swings: one account gives $2,000–3,000 off-peak and $10,000 at peak. Annual averages hide this.
- The exit. The top end shows most clearly in what famous stalls are worth on sale: $500,000 to $1 million asked for recipes, against $20,000 for an ordinary well-known stall.
The gaps, and how to close them
The top end rests on grade B and C evidence. Four sources could firm it up, roughly in order of ease.
- NEA's own cost survey. NEA surveys stallholders' costs every year. The same survey could report sales and net margin by decile, without naming a stall. It is the one source that could give a real distribution.
- IRAS statistics. IRAS publishes taxpayer statistics. The trade income of sole proprietors under the food-stall industry code, by band, would give the whole curve, tail included.
- The hawkers themselves. A short, anonymous survey through FMAS and the centre associations, or the Plan's own listening sessions, asking for a typical week's sales and costs. Self-reported, but at scale.
- Payment and delivery data. E-payment and delivery platforms see most stalls' sales. An aggregate by centre and decile would be grade B at least, and timely.
Until one of these exists, the Library shows the top end as a range with its grade, not as a number.
Each range is read across the evidence above; it is not a statistical fit. The bottom half rests on official rent figures and several accounts that agree. The top tenth rests on bids and named hawkers' own figures. The top 1% rests on recipe prices and the best-selling stalls in the press. Corrections and better data are welcome.
Where Hawker Heritage uses it
- The Technical Paper, G.3: the model's note on its missing top tail
- Follow the Plate, D.1: the hawker's share is for a typical stall
- The Plan, §1.1: the second blank, what hawkers earn
- The Plan, §2.3: recipe sales and the cap on a goodwill payment
- The Plan, §7.7: the spread, not only the median