One of the many thought provoking ideas put out in Policy Exchange’s latest report is that VAT is really no better than income tax; both simply reduce the value of your income, either by raising the prices of the things you want to buy (VAT), or taking income directly.
To quote directly:
“Suppose a worker earns £100 and pays £20 tax on it, and there is no VAT or other taxes. Then she has £80 to spend, and her £80 goes on goods with a real vale of £80 … Suppose now that income tax is abolished and instead a flat VAT of 25% is imposed. Then she earns £100 and she uses it to buy goods with a real value of £80 on which a 25% tax is imposed, raising the price from £80 to £100. So the real value of her consumption is again £80”.
I think this thinking is flawed, because it ignores how income tax is not flat, and has thresholds. I have structured the problem around how much of an incentive a person has to do some extra work under two regimes: income tax only, or VAT only. And the tax gained is meant to be the same – no cheating by not making it fiscally neutral.
The thinking is really simple. I have arranged the thresholds so they roughly resemble ours in the UK. There is single product (“fun”) which is going to either cost $1 or some higher number if the tax chosen is VAT. All consumption is spent, thereby avoiding the complexities that PX admirably look at in terms of savings/deferred consumption/etc.
In each scenario, a person has a choice whether to work for an extra $5000 or not. For the Income Tax payer, it is easy to work out the value of working more. Just take the marginal rate at their current rate of income. For the VAT-only guy, he is considering whether to take $5000, but with a higher price for a unit of FUN*.
As you might expect, the VAT system being flat beats the income tax system, particularly when the incomes are close to thresholds. If you make incomes very very high, then the two approximate, because the VAT required to raise the same as an income tax tends towards the 40% rate. If you programme all thresholds to zero, you get the PX thought experiment:
I have graphed how this advantage varies with the income:
but realise that this is a slightly unrealistic way of showing things. In the real world, there are a range of incomes and work-leisure decisions, but only one VAT rate. So if you assume you need an average VAT rate of 20%, and then look at individual work incentives, you get a chart like this:
This is more realistic, because the upper-rate earner faces a choice between his 40% tax rate, and the VAT rate required for the whole of society to pay the income tax. So for him the incentive advantage of changing the overall tax rate towards VAT is undoubtedly good (that flat line is around 38% when the VAT presumed to be needed is 20%). Think it through: if you earned 40k, and the VAT rate for society was 20%, pushing prices up by 20%, you would be able to get $4166 of real goods for your extra $5k of work. In zero-vat world, you would get $3000 after paying 40% on your marginal income.
If I were really clever, or had the time, I would try to model the entire work/leisure payoff, including benefits and tax credits. Then you might be modelling an increase in VAT that could be used to fund a lower tapering of benefits/credits. Since the people at the bottom of the earnings face such high withdrawal rates, I would again argue that the change would greatly improve work incentives.
Now I appreciate that I have been perhaps over-influenced by an OECD finding (June 2009 Economic Outlook) that the worst taxes for growth are corporate taxes, followed by personal income taxes, then consumption taxes, then taxes on immovable property. I often criticize others for extending results gained during non-deflationary periods into today’s crisis; I agree with Robert Reich today that with corporate profits increasing and labour income stagnant, and big companies not knowing what to do with their cash, it is ‘absurd’ to call for a tax cut for big corporates to boost recovery. So I don’t intend to use this OECD result as ‘proof’ that a rise in VAT is better for the economy – no-one should be citing piles of pre-crisis results as proof of anything right now.
But I am unconvinced that this simple thought experiment proves how VAT is really just a disguised income tax**. Sometimes the conventional wisdom is correct.
UPDATE: I have put the scrappy little Excel sheet here, if you want to play with the numbers.
*Note: if total income is 20k, and VAT is 15%, the VAT raised is not 15% of 20k = 3k. If it raises the price by 15%, 100/115 fewer units are bought. To raise £3k from the £20k you need to choose X so that X/(1+X)*20 = 3 ….)
**interestingly, Mark Wadsworth has elsewhere complained that it is a gross margins tax. I have a radical idea. Maybe it is a consumption tax.