Is GST regressive? The economics of Singapore's 9 percent tax.

Every time GST goes up, the same argument plays out. Someone points out that the tax falls hardest on the people who can least afford it, and they are not wrong about that. At 9 percent, GST is the tax you feel most directly, on the groceries, the haircut, the meal out, and from an individual's point of view nobody likes it. No consumer ever does. But the popularity of a tax and the soundness of a tax are two different questions, and the second one is the interesting one. So let me take you through what GST actually is, why it is regressive, and why I think the system around it, on balance, is a good one.

There is one fact to fix in your head before anything else. As at June 2026, Singapore's GST is 9 percent, and it comes with GST Vouchers and U-Save rebates that are used to offset the impact on lower-income households. Hold those two halves together, the tax and the transfers, because the whole argument turns on the fact that they are designed as a pair. Judge the tax alone and you will reach the wrong conclusion. Judge the system, and it looks rather different.

What GST is: an indirect tax on spending

Start with the category. GST is an indirect tax, a tax on a transaction rather than on income. You do not pay it because you earned something; you pay it because you bought something. The seller collects it and passes it to the government, which is why it is called indirect: the person who hands over the money and the person who bears the burden are reached through the price of the good, not through a tax return. Income tax and corporate tax, by contrast, are direct taxes, levied straight on earnings and profits.

The defining feature of GST, and the one that matters for everything below, is that it is broad-based. It applies to almost all goods and services, so almost everybody pays it on almost everything they buy. That breadth is exactly what makes it efficient as a way to raise revenue: it is hard to avoid, it does not distort one market relative to another the way a narrow tax would, and it does not punish work or enterprise the way a high income tax can. A broad-based consumption tax is, in revenue terms, one of the cleanest taxes a government has.

Why GST is regressive

Here is the charge against it, stated fairly. GST is a flat rate, 9 percent on the price, the same rate for everyone. But a flat rate on spending is not flat as a share of income, because lower-income households spend a much higher proportion of what they earn, while richer households save a large slice of theirs. If you spend almost all of your income, you pay GST on almost all of it; if you save a third of yours, a third of your income never meets the tax. So measured against income, GST takes a bigger bite from the poorer household than the richer one. That is the precise meaning of regressive, and it is true.

This is the part of the argument that is correct and that you should never try to wave away in an essay. An indirect tax on consumption is regressive by its nature. The mistake is to stop the analysis there, as if regressiveness alone settles whether the tax is a bad idea. It does not, because regressiveness is a problem with a known and used solution, and because the alternative taxes you would have to raise instead carry costs of their own.

Mr Toh's take: judge the system, not the tax

Let me give you my honest view, because students ask me whether GST is unfair and they expect me to either defend it or condemn it. I will do neither, because that framing is the trap. From an individual's view, of course nobody likes GST. No consumer does, and I do not either when I am the one paying it. But GST is a broad-based tax, and the whole point of a broad-based tax is that everybody pays, which is exactly what lets the government keep personal income taxes and corporate income taxes low. That is the trade you are actually being offered. It is not GST or nothing. It is GST, low income tax and low corporate tax, versus a higher income tax, a higher corporate tax and no GST.

So the real question, the one worth arguing, is whether it is a good trade-off to raise GST to 9 percent in exchange for those low direct taxes, and then use transfer payments to deal with the regressiveness. And my view, on balance, is that this system is a sound one. A broad-based consumption tax raises revenue efficiently and keeps the economy attractive to workers and firms, which matters enormously for a small open economy that competes for both. The regressiveness is real, but it is dealt with on the other side of the ledger, through the GST Vouchers and U-Save rebates that channel money back to exactly the households the tax weighs on most. You do not fix the regressiveness by making the tax worse at its job. You fix it with a separate, better-aimed tool.

It is not GST or nothing. It is GST with low income and corporate taxes, and transfers to offset the rest.

The zero-rating trap

There is a popular fix that sounds obviously right and is mostly wrong, and you should know why, because it is exactly the sort of evaluation that separates a good script from a thin one. The idea is to reduce GST's regressiveness by zero-rating necessities, charging no GST on essentials like food, electricity and water, so that the poor are spared the tax on the things they cannot do without. It feels compassionate and it feels targeted. It is neither.

The problem is hidden in the word necessities. Yes, lower-income households spend a higher share of their income on essentials. But in absolute dollar terms, higher-income households consume more of almost everything, including electricity and water, because they live in bigger homes, run more appliances and simply use more. So if you zero-rate electricity and water, you hand a tax saving to every household, and the richer household, consuming more units, pockets a larger saving in dollars than the poorer one you were trying to help. You have spent government revenue spraying relief at everyone, and more of it landed on the people who needed it least.

Read the figure as the mechanics behind the whole debate. The tax wedges supply upward by 9 percent, the price consumers pay rises, the price producers keep falls, and the gap is the revenue the government collects. The regressiveness is about who, by income, ends up bearing that burden across all their spending. The transfers are the second move the diagram does not show: that collected revenue flows back out to lower-income households, which is how a tax that is regressive at the till can, once you count the GST Vouchers, leave the overall system progressive. Zero-rating, by contrast, never collects the revenue in the first place, so it has nothing to aim, and the saving leaks to the rich.

This is why I keep telling students that targeted transfers are the better tool than zero-rating. They are not just kinder; they are more precise. A GST Voucher can be sized by income and household, given to the people you mean to help and withheld from those you do not. A zero-rated good cannot tell a millionaire's kettle from a cleaner's. If your goal is to protect lower-income households at the least cost to the public purse, you want the scalpel, not the broadcast.

Want this on paper? Grab the free 112 page Summary and Diagrams pack.