There’s a growing interest in land value taxation. Whilst the idea has been around since Henry George in the 19th century. In the last decade, a land value tax (LVT) been advocated for a long time by the widely respected FT economic analyst Martin Wolf and was supported in the well regarded Mirrlees Review of the future of taxation in the UK. Is it an idea whose time is finally coming? Apparently the Leader of the Opposition favours a form of LVT on developers’ land banks. John McDonnell, shadow chancellor, is a member of the Labour Land Campaign who want to replace business rates with an LVT. It seems that Andy Burnham also wants to replace the Uniform Business Rate with LVT. I suspect that I’d be retired before it ever happens but a debate is emerging.
There was a nice piece by John Kay analysing how the current system of business property taxation essentially results in lower demand for and supply of commercial space since structures as well as land are taxed. He identifies a number of winners and losers if there was a shift to taxing land – in fixed supply – rather than both land and structures.
…the main impact of business rates is on property values. If business rates in London were lower, rents and investment values of central London property would be higher. So if retailers had their way, and achieved substantial reduction in the rating burden, their joy would be shortlived: in the medium and longer term, the beneficiaries would be the owners of central London properties — real estate companies, pension funds, the Duke of Westminster and the Queen.
However, my colleague, Professor Peter Wyatt, highlights a number of practical difficulties associated with implementing such taxes. In a response to some GLA consultation on the topic which focuses on using a land value tax to increase the supply of land for housing, he argues that:-
There are three commonly cited reasons for introducing a land value tax: to raise revenue, to redistribute wealth and to assist land use planning. Obviously these reasons are not mutually exclusive, but what is the main purpose? It would seem from the background paper that it is land use planning – to bring forward more land for residential development. If this is the case, then is a land value tax the ‘best’ way of doing so? Has the GLA considered other ways of bringing land forward? A derelict land tax, grants and other incentives to release such land, changes to business rates relief on empty sites, etc? It might be useful to consider these alternatives alongside a LVT in order to identify the most effective way of meeting the land supply objective.
If a LVT is considered to be the best way of bringing land forward for residential development, then there are some practical hurdles to overcome:
(a) Creating and maintaining a register of land ownership. Does the Land Registry have an up to date register of site-level land ownership for the London area at a level of detail that can be used to levy and land ownership tax? If it does, will it release / is it allowed to release this information for LVT purposes? Presumably a significant proportion of land in London is owned by offshore companies and individuals. How would these owners be identified and how would the tax be levied and collected?
(b) Creating and maintaining a database of site-by-site land use. This requires site-level forward planning. Each site must have its existing use and its ‘optimum’ use assigned. A land use survey would identify existing land use but if the purpose of the LVT is to encourage land in non-residential use to come forward for residential use then an alternative (optimum?) use must be identified too. As you mention in the background paper, the 2013 SHLAA doesn’t identify specific sites because doing so would “compromise wider planning objectives.” Whatever these objective are, they would be seriously compromised by a land use database that allocated each and every site in London to its ‘optimum’ use!
(c) If ‘optimum’ is defined in terms of value then regular revaluations of the tax base will be required. Finding evidence on which to base land valuations is particularly difficult in an urban context because it requires the separation of unimproved land value of the value of improvements to the land. Since most transactions in London will generally relate to improved land then this separation will have to be estimated artificially rather than on the basis of transaction evidence. Valuers use the residual method of valuation to do this and this method is widely criticised in terms of the inherent volatility of the inputs and the sensitively of the output to those inputs. Add to this the problem of allocating hope value between the land and improvement components of real property value and the task becomes difficult and contestable, which leads to (d).
(d) Establishing and running an appeals process. Past experience of land and development taxes have shown this to be a significant cost. This is partly because of the issues discussed in (c). Since land valuations are, to a large extent, hypothetical and require expert input, they are contested; the difficulties faced by policy-makers in trying to pin down development ‘viability’ in the context of affordable housing targets, s106 planning obligations and CIL illustrates this.
A small number of countries have introduced various models land value tax (there’s a particular cluster around the Baltic) despite such challenges. In particular, compared to property taxes like Council Tax or business rates, the difficulties of estimating land values seems particularly problematic. As researchers always say, more research is needed? Although they don’t usually put a question mark there. If you’re interested in this topic, you should Google Ronan Lyons and Andy Wightman, you’ll see that they have done a lot of work on it.