By Dr David Comerford, University of Stirling
Recent figures released by Scottish Government and ONS have shown that the distribution of property wealth (GINI = 0.64) is more unequal than the, already skewed, distribution of income (post tax and transfer GINI = 0.31). Therefore, a revenue neutral shift from the really-not-very-proportional-at-all council tax, to a tax levied in proportion to property wealth would, all else equal, have more winners than losers. The broad outlines of the distributional effects are clear. However, as outlined in the Scottish Property Tax Reform principles, the case for taxing property wealth (and land in particular) is also expected to have beneficial effects on average and so the balance of winners relative to losers is even more favourable than a simple zero-sum distributional analysis would suggest. In this post I outline some of the mechanisms by which such a system of taxation could promote economic efficiency.
What is economic efficiency? Inputs are used efficiently if the maximum possible output is produced with these inputs. Economic efficiency is enhanced when labour, capital and land are used in their most appropriate ways; frictions or interventions (such as taxes or subsidies) which prevent their appropriate use (perhaps by shifting the incentives towards an alternative, less productive, use) will damage economic efficiency, though there may be cases where interventions are efficiency enhancing when they mitigate against some market failure.
The case for a shift towards property and land taxation being efficiency enhancing rests upon three main mechanisms which will be considered in turn: (1) the fixed (inelastic) supply of land means that its taxation is not distortionary, and these revenues could be used to reduce other distortionary taxes; (2) the impact of land as an asset class upon the macro-economy; and (3) the impact of home ownership upon labour mobility. The final section then considers the political economy that mitigates against a policy which should be a positive sum game, and which should be supported by the median voter even if it were a zero sum game.
The fixed supply of land
Suppose land is untaxed, you own £100,000 worth of land, and you deem that the best use of your land is to rent it out for £4,000 per year. Sometime later, a land tax is introduced and you are liable for some amount of tax, say £2,000 per year. Whatcha gonna do?
The best use of the land is unchanged so there’s only really two options: continue to rent the land out for £4,000 per year, and receive only £2,000 net of tax, or sell up (the sale price would likely have fallen to around £50,000). The land owner cannot increase the rent to recover the tax as the rental level is set via supply and demand in the rental market – the only way the rent would go up in response to this tax increase is if some (other) land owners decided to take their land off the market and so reduce supply. But if they did this then they’d still be liable for the land tax and they’d now be getting zero rental income – so it is unlikely that the supply of land would be greatly affected by the introduction of the land tax, which means that rental rates would be fairly unaffected, which means that the incidence of the tax would fall almost entirely upon the land owners rather than the land users.
This is clearly a large loss to the land-owners but notice that the level of economic activity is unchanged – the land will likely continue to be used by those who can best use it, and in the case of our example, they are likely to continue to pay around £4,000 per year for the privilege. The loss of spending power by the land owner is a gain in spending power by the government without any direct impact upon the supply of the factors of production being taxed (here land). This is non-distortionary tax: its implementation has not damaged economic efficiency. Other non-distortionary taxes exist e.g. the poll tax is also not associated with any withdrawal of factors of production from productive use – but few would argue that a land tax falls as heavily as a poll tax upon those least able to pay.
In contrast to a properly implemented land tax being non-distortionary, the current system of property taxation is distortionary in a way that is damaging to economic activity. Currently second homes or unlet business premises pay less tax (council tax or business rates) than main homes or let premises. This is effectively a tax subsidy to remove land from the market i.e. the current system says to the owners of a valuable and potentially productive asset, that one of the most privately profitable uses of this asset might be to simply not use it. Given that society as a whole could be made better off if this asset was used in production, or if the empty house was used by a family who needed a house, this framework is clearly crazy from an efficiency viewpoint.
Other taxes can be distortionary, for example the combination of income tax, national insurance, and benefit withdrawal, on top of the extra costs incurred when working like commuting, child care, etc, may cause potential low earners not to seek employment because they are no better off with a job compared with no job. To the extent that a shift in the tax burden towards land is used to reduce distortionary taxes levied in situations like this, we could see an increase in labour supply and so an increase in economic output from a revenue neutral but efficiency enhancing shift towards land taxes.
Land as an asset class
The current system of property taxation subsidises home ownership relative to renting a home. There are various ways in which the system does this including a capital gains tax exemption on your main residence, and no liability to pay tax on the income that you pay as rent from yourself as tenant, to yourself as owner. To see that these constitute a subsidy, consider the same property either owned by the occupants or rented to different occupants. It could be that ownership relative to renting just changed the timing of paying for the housing services that the property provides (e.g. perhaps ownership is more expensive initially but cheaper later on) – in which case there is no subsidy. However, in reality, the government earns more if the property is rented out: receipts of capital gains tax when the non-occupying owner eventually sells; and income tax on some of the rental income that this owner receives whilst in receipt of these rents. A situation in which government income is higher when the property is rented out, can equivalently be viewed as a situation in which government expenditure is higher when the property is owner occupied i.e. the government is paying a subsidy to home owners.
This is a regressive subsidy as it is disproportionately paid to the wealthy who are more likely to own their property. But does it also distort the market in ways that damage efficiency? One way in which it might is by shifting the portfolio choices of individuals: suppose you have savings of £20,000; in the absence of this subsidy you might choose to rent your home and hold these savings in a diversified portfolio of assets; when home ownership is subsidised, you instead use the money as the down payment to obtain a mortgage. This distortion in asset allocation decisions has two obvious costs.
The first is that you are more likely to put yourself into a leveraged position by taking out a large debt that is fixed in nominal terms and secured against a volatile asset (the value of the property). The gross value of the debt is likely many times the net asset position (e.g. 20 times for a 95% mortgage). This has macroeconomic consequences by encouraging individuals to behave in a pro-cyclical manner i.e. it increases the correlation between individual behaviour (we all experience similar house price movements) such that everyone feels wealthier at the same time (when house prices rise but mortgage debt stays constant) or poorer at the same time (when house prices fall but mortgage debt stays constant). To the extent that consumer spending is related to net wealth, this means that economic activity becomes much more cyclical, with all the costs that come with boom and bust cycles. The government should not be subsidising home ownership and encouraging these cycles, and should perhaps be taxing home ownership and discouraging these cycles. To the extent that we are talking about land values overall, and not just land used for residential property, the cyclical movement of real estate values also affects business solvency, and business investment via its use as collateral, all of which adds to the pro-cyclical impact of land values.
The second cost is also related to risk and diversification, but at a less coordinated macroeconomic level. Suppose you live in a town in which your employer is the largest employer. The present value of your future wages constitutes a large component of your implicit “net worth”. An optimal portfolio choice for your savings would be to invest in assets which were not correlated with your own future wages – so that when one asset performs badly, the other performs well to provide some compensation. But if your employer were to go out of business, not only would this damage your future earnings, it would also likely have a strong negative effect on local property values. Therefore, investing in local property is a particularly bad investment decision in this case. Again, it is irresponsible and malign that government policy at present subsidises and encourages individuals to make such sub-optimal choices. (A similar argument pertains against government subsidy of schemes which encourage investment in your own employer’s shares.)
It is not clear that this government subsidy of home ownership raises investment in housing (we do not have an over-supply in the housing stock) but it does likely raise land values overall and therefore shift the proportional composition of wealth holdings towards land, which is an extremely long duration asset, and away from productive investment which has shorter durations. The consequences of this shift in duration is that a greater proportion of wealth is left as inheritances, with consequences for the intergenerational transmission of inequality, if not economic efficiency.
It may also be the case that a policy framework which subsidies home ownership and hence shifts the composition of household savings towards land, crowds out the absolute level of investment in productive capacity for the economy. This needn’t be the case (if you buy land from me, I have to do something with the proceeds – perhaps I’ll start a business or build a factory), but it could be the case if those in receipt of funds from the sale of land are disproportionately likely to consume rather than to invest. This could be the case if it were the old selling to the young and using the proceeds to fund consumption in retirement. This is the mechanism underlying the argument in Weale (2007) in which rising house prices reduce the need to save for retirement, which reduces overall savings, which lowers the rate of investment in productive capital.
High levels of home ownership, and high land values, damage labour mobility. This is a circular process: suppose I have a potential move – a job offer in another location; if we are in a state of low labour mobility then it is likely that the number of potential homes near this new job are relatively low (low mobility means that their current occupants are also less likely to move); but this lack of availability means that I am less likely to take up this opportunity, rather I’ll stay where I am; and this decision to stay itself contributes to a lack of mobility. Low mobility means that people are less likely to do the jobs for which they are most suited because they are more constrained by location. To the extent that they are less productive in the jobs in which they are constrained to stay, this has an economic efficiency cost. A particularly egregious example of such an efficiency loss is where your present state is unemployment, and low mobility means you stay in a location where no jobs are available.
How does home ownership and high land values damage mobility? One way is in differential land values. If the entire land value were taxed away, then a home of equivalent build specification and quality would cost approximately the same, independently of its location. This means that if you have certain expectations and requirements based on your current home, a better job offer in a new location (such that the new wage net of land tax was greater than the old wage net of land tax) needn’t mean much of a change in your housing costs: the sale of your current house should fund the purchase of your new house if they are to approximately the same specification. Without the land component of the property price being taxed away however, property values in productive locations are likely much higher. It is highly likely that the better job offer received will be in a more productive location which means the sale of your current house does not fund the acquisition of a new house. It may be that a higher wage makes up the difference, but this is not necessarily the case. If it is not then you face a trade-off: move to the better job but take an effective pay cut because you cannot replicate the current housing amenity you enjoy; or stay put. Many will stay put – labour mobility has been lowered by differential land values.
Home ownership also damages labour mobility via higher transaction costs. The costs of buying and selling a property (including crazy additional policy measures like transactions taxes e.g. the Land and Buildings Transaction Tax in Scotland), whether measured in monetary, time or stress terms, are greater than the costs of exiting and entering rental lease agreements. The benefits of moving therefore don’t have to simply be greater than zero, they have to be bigger than some positive value that represents all the moving costs. This can be seen in commuting patterns: according to Oswald (1999) home owners commute more and further than renters, which could contribute to transport congestion.
Sounds good – let’s go
In this post I’ve outlined some of the mechanisms whereby a shift in the incidence of tax towards land raises economic efficiency. This means that we would expect average incomes to be higher as the tax burden is shifted towards land relative to the status quo. This is on top of the fact that a revenue neutral shift in the incidence of tax from income to land would benefit more than 50% of the population because of the concentrated nature of land ownership. Therefore there should be a large constituency in favour of this, and its implementation in a democracy should be straightforward, no?
In steady state, the only people who should prefer the status quo over an alternative with greater taxation of land, are children of the very wealthy for whom the inheritances outweigh any efficiency losses. But we cannot jump from one steady state to another, and the transition from the status quo to a policy regime with higher land taxes, mitigates against its implementation. In particular, a majority of the population (65%) are from households who own their own home, and these households will be adversely affected if property values fall. They may benefit on net, but the losses will be immediate and the benefits will accrue over time. There may also be adverse macro-economic consequences if households see their gross asset position hit, without any relief on their mortgages. Funding mortgage relief would add a large cost to the implementation of a shift to land taxes, and may prevent this shift being used initially to reduce other taxes (instead the land tax revenue would be needed to repay the bonds that were issued to fund the mortgage relief).
Although property ownership is concentrated in value terms among the wealthy (with a GINI coefficient of 64%), home ownership itself is widespread, which makes the political economy of the implementation of a shift towards land taxes difficult. However, if wealth inequality continues to increase (as hypothesised by Piketty (2014)) then the political economy of a shift towards land taxes, in terms of whether or not 50% of the population would be better off implementing the policy, transition and all, becomes more favourable over time.