D.7Scoring the Proposals

Every proposal is scored for the three parties the plan names (diners, hawker households and the public purse) and for the two it must not enrich by accident, private landlords and suppliers. The table shows a tendered NEA stall; the simulator runs any stall.

Diners

change in what this stall's customers pay in a year, plus the value of any vouchers they receive.

Hawker household

change in income and savings: income after tax, plus Workfare (the Government's top-up for low-wage workers) and retirement matches paid into CPF.

Cash

the part the household can spend now.

Programme cost

what the Government or the programme spends on the proposal.

Net public cost

programme cost, plus any extra Workfare paid out, minus the extra income tax collected and any money the state gets back as landlord.

Leverage

the return per dollar: (what the hawker household gains + what diners gain) ÷ programme cost, in income and savings and in cash only.

D.7.1What the scores say

• Price freedom moves money from diners to hawkers, and loses some on the way. A 5% price rise adds +$7,205 a year to the hawker household and costs diners $10,345; the $3,139 gap is customers who stop coming, and tax. It costs the state nothing, which is why it needs the vouchers beside it.
• Relief cooks return the most per programme dollar (leverage 2.17), because a covered day keeps a stall's takings that a closed day loses. That assumes the relief cook keeps the stall's custom, which is exactly what the Plan, §2.7's documented-recipe rule is for.
• The Protection Floor returns about a dollar per dollar, almost all of it as savings (leverage 0.94, cash 0.19). That is the design: its value is retirement and insurance the hawker cannot buy alone, not spending money. The match is counted for 50% of stallholders, the most who could be under 55. From 55, a stallholder who passes the existing scheme's income and home tests can use the Matched Retirement Savings Scheme (MRSS) instead (D.7.5).
• The Buying Circles cost the Budget nothing, and their saving comes from serving many small buyers more cheaply. Ingredients 3% cheaper add +$2,780 a year to the household, power and gas 10% cheaper +$1,179, and delivery at a 25% group rate instead of the 30% list rate +$833 (D.16). The suppliers' column shows what circle members pay them less (−$2,830): a lower cost of serving one pooled order instead of many, shared between supplier and cook, not money clawed back from margins that, on the evidence, have not widened (D.5.3). The circles are the hawkers' own, so they have no programme cost and no leverage (benefit to hawkers and diners for each dollar of programme cost) to show; the Seed Bank's small part in them (a part-time partnerships role, the open price board and the circle leads' honoraria) is in the operating budget (the Budget Book, B.1).
• The Government pays the table-cleaning fee subsidy, and the model counts all of it as a cost. At a pilot-centre stall the Government pays $2,400 a year towards the stall's cleaning fee, whoever holds the cleaning contract (at a Town Council centre, the hawkers' association), and the household gains +$2,357, leverage 0.98. No landlord loses income. At an operator-run centre the operator only passes the fee on and may not profit from it, so the Government pays there too. The true cost may be lower, because re-tendered cleaning contracts may cost less (D.14). The plan asks for it at the pilot centres only, so a pilot-centre stall gets it and the national tables count only those stalls.
• The named deputy gives time, not money. It gives back 302 hours a year at no cost to anyone, so its row is zero in every dollar column; what it moves is the hawker's hourly earnings, from $9.69 to $10.49 (the Plan, §2.7, D.4.1).
• Rent bands are an almost pure transfer from the state as landlord to the new entrant (leverage 0.83): the right tool for entry, and only where the state is the landlord. Its cost accumulates as each year's entrants join, by about $1.0M a year.

The package for a sitting tenant of a certified heritage stall, the stalls D.15 frees to price: price freedom (+5%), Makan Vouchers, Relief-Hawker Corps and the Protection Floor, with the Buying Circles' ingredients, power and delivery (D.16), the table-cleaning fee subsidy at the pilot centres (D.14) and the named deputy (the Plan, §2.7). The HDB tender lever is not in it: it is a way to pay for the Protection Floor's match at HDB-rented coffeeshops, not a second match (D.7.5). The diners column is this stall's customers only; the vouchers are spread across all stalls, so at one stall in five the average diner gains (D.7.3). Mall food courts do not accept CDC vouchers, so diners there pay the higher prices with no voucher, and the mall landlord takes a quarter of the extra, because the rent is 25% of sales.

D.7.2The same $200 a month, attached to the stall or to the person

Across all 14,200 stalls, stall-attached support (paid to the stall, so a landlord can take it back at renewal) of $200 a month would cost $34.1M a year. At renewal, landlords would raise rents and take back $12.6M of it: $2.3M returns to the state where it is the landlord, and $10.3M (30%) goes to operators and private owners. Hawker households would keep $19.8M. Paid to the person as a retirement match, the same money reaches hawkers in full, as savings. The capture rates are assumptions; the direction is not. This puts numbers on the rule in Hawkers and Social Protection, C.7: help should go to the person it is meant for, not to whoever owns the property.

D.7.3Do Makan Vouchers cover price freedom?

D.7.4At national scale

How to read the national table.

• Take-up. The rates are assumptions: price freedom at certified and new-entrant stalls (one in five); Corps capacity after scale-up; the Protection Floor at 70% enrolment; the Buying Circles at one stall in four; the fee subsidy at the pilot centres only, three government centres with 408 cooked-food stalls in all, each managed by NEA or a Town Council; the named deputy, which moves hours rather than money, at three stallholders in ten and not shown; rent bands for the 2–4% of tendered stalls that turn over each year, shown for one year's intake. Single-policy effects are summed; an exact combination differs by under 1% nationally.
• The Protection Floor. Its row, and the total, use its stretch scope: every stallholder at 70% enrolment, $7.92M a year, of which $5.96M is the retirement match for stallholders under 55, a ceiling, and $1.95M the hospital-cash cover. The base, Hawkers and Social Protection's $4.14M, is the match alone for every public-centre stallholder under 55 (D.7.5). Its net public cost equals its programme cost, $7.9M: the model credits it with no extra Workfare (the Government's top-up for low-wage workers).
• Why no Workfare is added. The Government pays Workfare, its top-up for low-wage workers, only to people who are paying their MediSave. A hawker cannot renew the Singapore Food Agency (SFA) stall licence while behind on MediSave (in arrears), unless he is on GIRO (a bank's automatic monthly payment), an automatic monthly bank deduction (SFA; CPF, 4 Nov 2025). So the self-employed's one in four behind (MOM, Feb 2025) does not describe hawkers. The model assumes one in twenty behind, with or without the floor; MediSave as you earn changes the timing of the bill, not the Workfare.
• The HDB tender lever. It is shown below the total, below the total, for its cost to HDB alone: the tender rent HDB may give up, up to about $1.7M a year at half take-up (the Budget Book, B.7). Operators pay the $1,200 for every stallholder, but only the part for those under 55, about $0.8M, can stand in for the match; the part for those 55 and over goes into their own retirement savings beside MRSS. So it pays for part of the Protection Floor's coffeeshop match, gives no hawker a second $1,200 and is not added to the total.
• The fee subsidy. Its row is what the Government would pay of the table-cleaning fee at the pilot centres, $200 a stall a month, counted as programme cost in full, an upper bound (D.14, D.2.2). Its national row, below the total, is the same subsidy at every stall in a public centre: a later decision, not part of the plan, shown so that its size is known before it is asked for. The plan's own budget for Years 0–3, about $19.68 million (the Plan's Part VII), funds the pilots, not this national roll-out.

At national scale the plan's programmes would cost about $79M a year and return about 1.06 dollars to hawkers and diners per programme dollar in income and savings, or 1.00 in cash. Leaving out price freedom, which diners rather than taxpayers fund, the public programmes return 1.24. Most of the cost is the Makan Voucher line, a cost-of-living transfer to diners, and the next largest is the Protection Floor. Private landlords barely move. Circle members pay suppliers $11.8M a year less, $7.2M net of the extra custom the other proposals bring them: a lower cost of serving many small buyers, shared between supplier and cook (D.5.3). The plan pays the person, not the property.

D.7.5The Hawker Protection Floor: one list

Hawkers and Social Protection costs the retirement match at about $4.1M a year; the table above shows $7.9M. They are one proposal at two scopes. The match is for stallholders under 55. From 55, the existing Matched Retirement Savings Scheme (MRSS) already matches cash top-ups to a member's CPF Retirement Account, dollar for dollar, up to $2,000 a year and $20,000 over a lifetime. To qualify, a member must have less than the Basic Retirement Sum in that account (as a MediSave-only hawker does), average income up to $4,000 a month, a home with an annual value (the rent it would fetch) up to $21,000, and no more than one property. So the plan helps hawkers of 55 and over who qualify claim it, through the Seed Bank and the Hawker Council, rather than paying twice; that help is outreach, not a cost line. The base is a $1,200 yearly retirement match for every stallholder under 55 in a public centre (NEA subsidised, NEA tendered and operator-run centres): at most 3,450 of 6,900 stallholders, $4.14M a year. The stretch reaches every stallholder under 55, coffeeshops and food courts included, at 70% enrolment, and adds shared hospital-cash cover for every enrolled stallholder of any age: $7.92M a year, of which $5.96M is the match and $1.95M the cover. Both are ceilings. How many stallholders are under 55 is not published; the median hawker is 60 (MSE, Nov 2024), so fewer than half are, and the model counts 50%. If it were paid at every age, the match would cost about $8.3M (base) or $13.9M (stretch) a year, and would pay twice for those aged 55 and over whom the MRSS already matches. The Budget Book, B.9.5 asks how much of this could be paid from gifts and licence fees.

Stallholders: those matched (under 55, at most 50%), except in the cover row and the stretch total, which count every enrolled stallholder. The hospital-cash premium is $197 a stall a year: a 6% yearly chance of a 21-day episode at $120 a day, plus 30% for administration (assumptions). Because the stretch assumes 70% enrolment everywhere, it counts fewer public-centre stallholders than the base; with every public-centre stallholder enrolled as well, the stretch would cost $9.57M. The tender lever is one way to pay for the coffeeshop part of the stretch, not an extra benefit. The stretch already pays every enrolled coffeeshop stallholder under 55 the $1,200 match: $1.18M a year for at most 980 stallholders at HDB-rented coffeeshops. If HDB agrees, it forgoes up to $1.7M a year of tender rent for 1,400 stalls. Operators pay $1,200 for every one of those stallholders, but only the part for those under 55, about $0.8M, can stand in for the match; the part for those 55 and over goes into their own retirement savings beside MRSS and is not counted. That pays for about 71% of the coffeeshop match, so the lever is not added to the package or to the total in D.7.4 (the Budget Book, B.9.5).