Sunday, 28 June 2020

By how much would UK tax receipts fall if VAT were scrapped outright?

Physiocrat, here, is very radical on this:

VAT costs and losses

1) Admin - not very much as the costs are transferred to business.
2) Churning - VAT is part of the price index to which pensions, benefits and public sector pay are linked. This makes it a major government cost in its own right.
3) Abstraction from other taxable revenue streams.
4) Welfare costs arising from deadweight losses.
5) Tax revenue losses due to deadweight losses (separate from abstraction).

My piece on the LVTC web site needs to be re-worked with better calculations. I would not be surprised if, because of the above, the Exchequer is out of pocket due to VAT.

I think he's overselling it. We exchanged emails this afternoon, and my summary is this, FWIW. Total VAT receipts are about £130 billion a year in the UK. What happens if we scrap it (which we can now do, post-Brexit) and make no other changes to the tax system?

1) Admin costs

Government HMRC admin costs are supposedly around 1% = £1 bn, double that for 'carousel fraud' etc = £2 bn.

Number of VAT registered businesses (two million) x extra work each quarter adding the extra info to the accounting system and doing the quarterly VAT return (ten hours) x average hourly cost of employing a bookkeeper £50 (wages, NIC, desk space) or lost value of small business owners' time when they could be actually producing something = £4 billion a year. Smallish number, so let's include it with dead weight costs at 5) below.

2) Churning

I'm not sure I agree that "VAT is part of the price index to which pensions, benefits and public sector pay are linked". VAT has little impact on end prices, Phys countered that "End prices would go down if all these people were not handed government money to pay VAT with!" which is true. I'm still not sure what the net impact on government spending would be.

3) Abstraction from other taxable revenue streams

This is the biggest and easiest number.

PAYE (income tax plus two layers of NIC) on average worker = 40%. Higher rate employees PAYE = 49%. Corporation tax is 19%. If profits are paid out as dividends to individuals, there's a bit more income tax on that. VAT cuts into business rents, which are taxed at about 40% via Business Rates. The overall average rate is (say) 33%, so one-third x £130 billion would come back in anyway = £43 billion, assuming no change at all to selling prices and output in real terms (goods and services supplied, hours worked etc).

5) Tax revenue losses due to deadweight losses (separate from abstraction) and 4) Welfare costs arising from deadweight losses

Let's say scrapping VAT boosts the economy and trade by 5% in real terms (low guess - it depends what assumptions you make about price elasticity of supply and demand). Total current tax revenues from economic activity (PAYE income tax and NIC, corporation tax, bank levy, Business Rates etc) about £430 billion a year. 5% of that = £21 billion.

There are about thirty million employees. Let's say that number also goes up by 5% and 1.5 million who are currently un- or underemployed get a part-time job, longer hours, full time jobs. Average unemployment benefit, housing benefit etc (say) £10,000 per person per year = £15 billion saving.


HRMC would lose £130 billion of VAT receipts.

HMRC would save £2 bn admin costs and fraud; get £43 bn more from other taxes anyway (static basis); get another £21 bn from other taxes from losing deadweight costs (dynamic effect); and the DWP would save £15 bn a year in welfare payments, total £81 billion.

That means that the UK government would only lose about £50 billion a year if it scrapped VAT outright.

For convenience and to make this politically sellable, we can play along with the myth that consumers/households pay the VAT.

£130 billion divided by 29 million households = £4,500 per household on average. If we claw back the £50 bn shortfall from total housing values (pre-lockdown) of over £7 trillion, it would require a progressive property tax of about 0.7% on selling prices (or for the purists, a 25% Land Value Tax on residential site premiums) = £1,750 on average.

The average household is £2,750 a year better off. What's not to like?


Woodsy42 said...

Yes, and it would remove some of the grit from the wheels for small traders. If the gov were actually out of pocket I'm sure they could put a few pence on (non-regressive) higher income tax bands, scrap VAT entirely and would still please almost everyone, including higher band tax payers.

Lola said...

Where do I sign up...?

Lola said...

You might also add in a bonus from all the VAT men that could be released from tax consuming posts to find more rewarding wealth creating jobs in private business.

Mark Wadsworth said...

W, higher rate tax of 40% is quite high enough. The additional rate of 45% (once you add on NIC) is past the top of the Laffer Curve, so I am against hiking that. It's a straight swap - pay £4,500 VAT or pay £1,750 LVT. If my figures are over-optimistic, then it might be £2,000 LVT instead of £4,000 VAT. Either way, 90% of households will be better off and the 10% will have to suck it up.

L, you are signed up.

L, that goes under admin costs/deadweight losses.

The ex-VAT people can be diverted to chasing proper tax evaders and doing the land valuations.

The small business owners now have a few more days a year where they can actually provide goods and services for profit instead of fannying about with RTI stuff.

Derek said...

Agreed. Far better way of fixing the economy than messing about with half-arsed "trade deals".

Mark Wadsworth said...

D, ta. I'm very dubious about the merits of 'trade deals', where the emphasis is on the 'deal' and not on the 'trade'.

Shiney said...


Bit late here - but the government could also mitigate the 'loss' of revenue by rolling some or all of the VAT charged on fuel into the duty element if desired. And booze/Fags as well.

Would have the side effect of pleasing the greenies/puritans.

Mark Wadsworth said...

Sh, when I'd finished I remembered that the plan is to merge VAT on fuel with fuel duty so pump price stays the same. But I was too lazy to go back and re-work everything. Taxes on booze and fags are pure politics, not economics.

Physiocrat said...

MW -Your figures are in a similar ball park to mine. I made it 70% on a total of the 2016 figure. Yours (62%) is the same £81 billion on the larger 2019 figure. You seem to have put churning losses down to zero. By churning losses, I am referring to taxpayers' money paid to OAPs, benefit recipients, public sector workers, etc, which they pay back in VAT when they buy things. I made it £16 billion. It nets off to zero for the exchequer.

Mark Wadsworth said...

Phys, I didn't include churn.

The reasons are
- we know that ex-VAT, wages will go up. So govt salaries and state pensions will go up, rather than down
- we know that ex-VAT, prices won't fall much if at all, so no reason to reduce govt salaries and state pensions.