After five years of hard work, I have come to the conclusion that ideal geographical sub-unit for LVT valuation purposes is a postcode sector, which is about 3,000 addresses. To give you a rough idea, at a typical urban/suburban density of twelve homes per acre, that's a circle rather less than a mile across, or ten minutes walking distance from edge to edge. So there wouldn't be a significant difference in values within that area; we take actual recent selling prices, do a bit of averaging, job done.
Sobers begged to differ, swearing blind that there could be huge variations in house prices within one postcode sector, adducing as evidence two physically similar houses in Swindon, postcode district SN3. One is up for auction by the bank which repossessed it, at a starting price of £94,995 and the other is up for a possible sale at £197,500
Nice try, but...
Notwithstanding that one is a repossession (so is being dumped at undervalue) and the other will be sold for rather less than £197,500 or not at all, those two houses are not even in the same postcode sector.
But honour bound, I downloaded the last one thousand recorded sales for the postcode district Swindon SN3 (all sales for the past two years or so) from Houseprices.co.uk and bunged them into a spreadsheet. I selected all semi-detached houses (easiest to understand and most homogeneous), sorted them into six postcode sectors, deleted the bottom and top five per cent in each sector (like all good statisticians do when they are far away at sea - ignore repo's and mortgage fraud) and arrived at the table below (which took me all of half an hour to compile - there are only about 10,000 postcode sectors in the UK, so at this rate, I could work out LVT rates for the whole of the UK in six months).
For clarity:
i. I arranged the postcode sectors by ascending average values.
ii. Standard deviation is a statistical measure. What it means in practice is that if average semi in SN3 2.. is worth £115,000, two-thirds of semi-detached houses in SN3 2... are worth between £103,000 and £128,000, which is not a huge gap,
iii. Just so that you can see I am not cheating, I divided the SD by the average value. The range is fairly tight in each sector and there is no problem with overlap - the most expensive two semi's in the cheapest sector SN3 2.. sold for the same as the cheapest two semi's in the most expensive sector SN3 1..
iv. Assuming average plot size 400 sq yards and a rate of LVT which would be sufficient to replace all other taxes (income tax, National Insurance, VAT, Council Tax, the lot) and leave wiggle room for exempting pensioners and so on, I added the likely LVT rates in the last column. Four of these sectors have rates between £29 and £32 (i.e. hardly worth arguing about). The only noticeable difference is between the cheapest sector, which equates to annual LVT of about £9,200 for a semi, and the most expensive sector, where the annual LVT would be about £15,200. It's unlikely, but possible, that these two sectors are continguous, in which case, no argument.
v. Remember that the flip side of LVT, the Citizen's Income, would be about £11,000 for a two adults-two kids family (education and health vouchers are on top of that), you're not talking about huge sums of money either way.
Click to enlarge:
Once I dipt into the future
1 hour ago
32 comments:
Anyone who knows what a standard deviation is is all right in my book.
My god! I feel guilty now, making you do all that work. Very impressive, I have to say.
I hesitate to ask, for fear of being labelled a troublemaker, but what about commercial property? How will that be included in your calculations? I know for a fact for example that SN3 postcode in Swindon contains a fair bit of industrial/commercial/retail property.
OP, thanks.
S, you keep me on my toes! As to land used for commercial purposes;
1. I see no reason to have special rates for this - is there any reason why the tax bill in the middle of the High Street 'should' be higher or lower if the upper storeys are used as residential or as offices, storage etc? And if so which, higher for residential or for retail/offices?
2. At the other extreme are industrial estates or retail parks, where the buildings themselves are worth very little (being single storey metal sheds), the value is pure location value and two thirds of the area is normally used as car parks.
3. Apparently airports make more money from parking charges than anything else, for example.
4. Working out potential LVT for commercial is dead easy - it will be 'about' twice as much as Business Rates (which would bring it into line with my proposed LVT rate for residential). You can look up each plot size and rateable value at the VOA website (times that by forty per cent to get the Business Rates payable).
5. Reality check: let's say £30/sq yard in SN3. A car park uses about 20 sy yards per parking space on average = £600 per year, or £2.50 per working day. Seems 'about right'.
I suspect as well that your argument (strong as it is) would improve if real square yardage were available (which it presumably would be to the government).
I suspect that the standard deviation would drop significantly if the measure used was price per square yard.
The three primary factors for house buyers have got to be:
- House location
- House type
- Plot size
It wouldn't affect your final answer, since it would come out the same (assuming the 400 sq yd average is correct). But it would add weight to your assertion that house prices within a sector have low-variability.
OP, exactly. The house type is irrelevant for these purposes, it's only location and plot size that matter.
As to plot sizes:
1. HM Land Reg know perfectly well about plot sizes, they sorted out some software a few years ago to do it for them. They already publish the figures online for Business Rates (see previous comment).
2. The £ rate is a 'made up figure', the plot size is a real one. So to prevent silly disputes, instead of charging somebody £30 x 400 sq yards (and some smart arse measures it and appeals because actually it's only 387), they could round down the figure by ten per cent and charge 360 sq yards x £33.33.
3. The 400 sy yards seems a tad on the high side to me, but that's what the government's Generalised Land Use Database says, so who am I to argue.
4. Round where I live there is a weird jumble of tiny houses on tiny plots, big houses on big plots, detached, semi, terrace and flats, the lot.
5. The plot size thing works almost perfectly - the huge house on a 500 sq yard plot next door to me sold for £650,000 and a little terrace across the road on a 200 sy yard plot sold for £280,000 recently. A large-ish semi up the road on a 400 sy yard plot sold for £500,000-odd. A flat up the road that uses pro-rata about 200 sy yards (huge front and back garden) sells for about £250,000 - £300,000.
6. Therefore 'total built value' per sq yard is always around £1,300 x tax rate 8% = £100 (yes, it is a very, very expensive area).
Oooooo my house has a unique postcode. In that area houses are worth very little indeed.
L, your house may well be the only one in that postcode unit but I doubt that it's the only house in that postcode sector?
Philosophically, what I really really love about LVT is that once it's up and running and people start seeing how it works and realise that income tax and vat etc are evil, they'll never ever vote for a government that wants to change back.
The only way government could go back was if it could engineer a war......
For completeness Mark, how is that list updated, when I assume selling prices are altered by the presence of LVT?
L, that's why they started the 1914-18 war, just to distract the little people's attention from The People's Budget :-(
F, I have long been hoping that nobody would ever ask me that question, but here goes:
1. The very long run aim is to get land prices as low and flat as possible, i.e. the same around the whole country. If a residential building plot costs £5,000 or £10,000 to buy, that's not a major issue - at current values of about £100,000 it is an outrage.
2. So once the system is up and running (and assuming we don't get chucked out at the next General Election), every year actual recorded selling prices in each sector are averaged out again.
3. If land and buildings in any sector are being sold for at or lower than rebuild cost (and yes, HMLR even know what the total internal area is for each sale, which we can multiply by an arbitrary figure like £750), the rate is 'too high' and gets reduced by five per cent.
4. In other sectors, land and buildings will still sell for far more than rebuild cost, and the rate in those areas just gets nudged up by five per cent that year.
5. To some extent, people will pre-empt this:
a) Buyers will guesstimate what the likely LVT bill will be in a few years' time (and adjust their offer price down) but
b) If the current owner knows that the bill 'should' be £20,000 but currently it's only £15,000, there is still a six year period in which he gets a discount, worth about £15,000 in cash terms, so he's not going to give the house away (he'd be better off keeping it and renting it out).
So the mechanism is self-correcting and it's a moving target, but hey. After a couple of decades, we'd observe that yer average semi sells for £100,000 (in today's money) whether it's in Dundee (with an LVT bill of £8,000 a year) or in Outer London (with an LVT bill of £28,000 a year).
"(which took me all of half an hour to compile - there are only about 10,000 postcode sectors in the UK, so at this rate, I could work out LVT rates for the whole of the UK in six months)."
My initial thoughts are get to it then, obviously in the really real world you can't just up and spend 6 months preparing this information, I for one would end up bankrupt if I tried. The ease with which it was prepared is definitely another selling point though.
What about postcodes like SW1A 1? It contains lots of buildings which will probably never be sold or belong to the State.
SW, in real life, I can't spend more than half an hour a day on this (wife kids nag), so it'd take me about five years.
EKTWP, if it's a government building, that's a straight in/straight out (no point paying tax to yourself); if it's a no longer used government building (once I've slimmed all these departments down by 90%) then it gets rented out to private sector (so govt. collects all the rent anyway, like Crown Estate buildings).
And if we know the total rental value we can deduct bricks and mortar rent and the balance is the notional LVT, bung that in with the privately owned buildings that are also rented (you can capitalise rents to arrive at theoretical selling price) and Bob's your uncle.
"...then it gets rented out to private sector (so govt. collects all the rent anyway, like Crown Estate buildings)..." Or gets sold off so that it's not the taxpayers problem, they can just collect the LVT.
If only it were true that people would never vote for anything else. Sadly that's exactly what they did in Australia and New Zealand which were big users of LVT in the 1900s. Not to mention Denmark, France, Western Canada and a bunch of other places where it was tried. Admittedly these places did not replace all other taxes with LVT as they should have and there was no Citizen's Income involved, but even so history shows us that achieving the goal is only the first stage. Keeping it is just as tough. People hate land taxes because they are so "in your face". That's why education is so important. You've got to demonstrate to everyone that this ugly duckling is actually a swan.
On the subject of using postcode sectors for valuation, the only problem that I can see is gerrymandering. Once postcode sectors start affecting people's pockets, there will be a clamour to be shifted to the nearest cheaper sector, if I'm any judge. And that's just the sort of issue that politicians pick up and run with.
L, or that.
D, agreed to the first bit, sadly.
As to the second, that is why I chose postcode sectors, which were decided forty or fifty years ago by The Post Office for their own administrative convenience: postcodes hardly ever change and have nothing to do with 'political' boundaries like local council wards or parliamentary constituencies. To be fair, there might be postcode sectors where the discrepancy between 'nice' bit and 'rough' bit is so huge that the inhabitants of the cheaper half apply for the postcode sector to be split for LVT purposes, but the Post Office won't care a toss about this.
What about State owned businesses? Post Office, BNFL, Railtrack, council owned businesses etc etc? Surely they should have to pay LVT, otherwise its a big competitive advantage?
@Derek
Land taxes don't have to be in your face.Start off by altering Council Tax to land value only.Fiddle/ adjust % rate so people pay the same amount.Nobody's going to notice the difference.Further: utilise the Reed/Wadsworth tax reciprocator whereby income tax and land tax are lumped into together,income tax falling as land values increase .House prices going over solicitors' desks in conveyancing to be forwarded to tax office on weekly basis.
S, yes of course the LVT due on such premises will be 'paid', or at least, shown as expense in their accounts and as income in govt accounts (they already use this system for notional rent on government owned buildings used by government departments).
People like Railtrack and power stations currently pay Business Rates, they'll pay LVT instead. Maybe bung a bit on top for the value of the water company's natural monopoly.
DBC, that's true. We agreed once that:
1. LVT could be collected via the payroll (salary or pension) instead of income tax/NI;
2. It could be added to mortgage repayments and collected by the bank;
3. It could be netted with Citizen's Income and only any balance collected via 1. or 2.
and many other such wheezes. That's details.
DBCR, MW, if you guys can come up with sugar for the LVT pill, I'm all in favour. And the reciprocator idea looks sensible.
MW
Oh sh*t - I'm more or less convinced! Mrs U won't be happy now I've got another hobby horse to ride at dinner parties.
U, sorry about that. Four years ago, I suggested the idea of a 1% flat tax on houses as a way of replacing/simplifying Council Tax, Council Tax Benefit, IHT and Stamp Duty Land Tax etc.
I only ever got grief for it (I had my fifteen minutes of fame and hated every one of them), problem was, none of the killer arguments against ever stacked up and the more I was forced to defend the idea, the more rabid I became. Had people said 'Oh, that's nice, good idea' I would probably have left it at that and forgotten all about it.
umbongo/MW Me too. Mrs Lola gets very frosty when I get going at socials. She was used to me boring people to death about motor racing, now she wishes I'd go back to that.
L, Mrs W hates it when I even mention it, it's a guaranteed frosty silence followed by row.
So I'm cheerfully training up the kids - every time we drive past a derelict building or a vacant site, I ask them: "What kind of tax will sort them out?"
MW Hahahha me too. I have four of the little darlings to brainwash, erm, I mean educate.....
MW/L
I'm banished to a separate room by Mrs U when BBC News at 10 comes on. For some reason she objects to my incessant shouting at the TV and the insults hurled at Robinson, Easton, Peston, Shukman . . . . . . She claims she just wants to know what's happening in the world and "cocks a deaf'n" to my arguments that the BBC might present her with a skewed version of events.
Ah - time for "news where you are" - another bout of propaganda for the students seeking to "stop London".
My wife doesn't actually fall out with me but she will resignedly point out that she's heard about it once or twice before. And 25 years of marriage have taught me when to take a hint...
So the mechanism is self-correcting and it's a moving target, but hey.
I thought that was the point.
If land and buildings in any sector are being sold for at or lower than rebuild cost
Okay, that's the info I was after. I remember reading a Michael Hudson piece once where he reckoned using rebuild sometimes gives strange results (like negative land value), but I'm wondering if he was misinterpreting that (by assuming that all land must be valuable, when actually if planning is messed up enough you can wind up in places you really don't want to be but don't really have much choice).
F, negative land values are nothing unusual.
Let's say there's a site that is terribly contaminated and there are buildings on it which are derelict and useless. You might be able to buy the whole site for £1, but it will cost you £10 million to clear it all up, and at the end of the process, you might still only be able to sell it for £1 million. So really, you should only accept the site if the current owner offers a negative price of £9 million (i.e. he gives you money).
However, that negative value breaks down into two things: a 'rebuild cost' of £10 million, and a location value of £1 million. It is still perfectly possible to tax the location value rent. If taxed at 100%, the selling price of the plot falls to negative £10 million - it is then up to the owner whether he wants to pay £100,000s every year in tax, or take a one off hit of £9 million and clear it up (assuming no legal obligation to clear it up).
After a couple of decades, we'd observe that yer average semi sells for £100,000 (in today's money) whether it's in Dundee (with an LVT bill of £8,000 a year) or in Outer London (with an LVT bill of £28,000 a year).
Sorry Mark, you've lost me here. I thought one of the purposes of LVT was to reflect the value of the land. Assuming these semis are identical, and the land is worth the same, why is the tax so different?
Ed, to get the ball rolling, you can base the tax rates on recent selling prices (as these are proportional to the rental value of the location, obviously). That's nice and easy to understand and calculate, and there are enough sales for the results to be fairly reliable.
The rental value of any location does not change because of the tax BUT the cash selling prices do change (the higher the tax rate, the lower the cash selling price).
Ergo, if the annual tax were (conceptually) ninety per cent of the annual rental value (which you can only arrive at by guesswork), the cash selling price of land/location would be a tenth of what it is now, i.e. much the same anywhere in the country - and you then add the bricks and mortar value on top (of course).
Post a Comment