Land Taxation:
Lessons From International Experience

Colin Clark

[Abridged and reprinted from Land Values, edited by Peter Hall. The Report of the proceedings of a colloquium held in London, 13 and 14 March, 1965, under the auspecies of the Acton Society Trust. At the time, Mr. Clark was Director of the Agricultural Economics Research Institute at Oxford]

EVERY statesman designing a system of taxation has to take into account three quite different considerations, which may often lead to conflicting conclusions, namely social justice, economic efficiency and administrative practicability.

Let us consider land taxation first from the point of view of economic efficiency. The principles of efficiency, applied to taxation, require first that it should give as little discouragement as possible to people's willingness to produce; and further that it should " distort" production as little as possible, i.e., should not encourage producers to produce more of something which consumers want less, at the expense of some other commodity which the consumers want more.

Among the range of practicable taxes, there is none which is so economically "efficient" as land tax. (It is theoretically true that an equal degree of economic efficiency would be attained by a tax on "quasi-rents," i.e., those elements in the incomes of various producers which represent a surplus above the minimum income which would have been required to have evoked their productive efforts; but a tax assessment of this sort cannot even approximately be made in practice.) All taxes on currently earned income, to a greater or lesser degree, discourage production and the earning of income. Taxes on past savings, or on income from them, do not, it is true, have an immediately discouraging effect on current production; but, if carried beyond moderate limits, they will very quickly affect people's willingness to work and save in the future. Anyone, whether wage-worker, salaried worker, manager or proprietor, who is taxed at a rate which he thinks too high, may work less energetically, or take more leisure, or possibly emigrate. A piece of land can do none of these things. On the contrary, its owner is likely to administer it with more skill and care, and put it to the best possible use, if he has to pay a substantial tax on it anyway. And, unlike savings, land is not the fruit of past human effort. It is true that people may use their savings to buy land; but this does not affect the amount of land available, which would remain just the same if nobody or everybody used his savings to buy land (although they would have to pay very different prices for it under these varying circumstances). The amount of land available for agriculture and mining is determined by natural factors, and for urban use (in almost all cases) by the activities of people other than the owner of the land himself.

The above reasoning applies to taxes placed on land itself, and not on buildings and other structures which may have been erected upon it These should be taxed, if at all, as lightly as possible; building is a form of production which we wish to encourage.

This brings us to the question of administrative practicability. At the time when land taxation was first thoroughly discussed in this country, about 1910, there may have been more justification than there is now for the insularity of those who said that a general valuation of all tend was impracticable. In New Zealand, and in three states of Australia, a general valuation of the unimproved value (i.e., excluding all buildings, structures and other improvements) of land has been made for many years, and is used for local taxation in place of the British system of rates. The valuation is perfectly practicable; the economic results of land taxation are satisfactory too. The general instructions to the valuer are that, in valuing any particular property, he must first observe the prices which are being paid in the neighbourhood for similar improved properties (i.e., those carrying buildings, farm improvements, etc., similar to those on the land to be valued), make a valuation of the unproved property, then deduct from this gross total his estimate of the depreciated value of the improvements (bringing the latter up from historic cost to replacement cost if, as is usually the case, there has been significant change in general costs of construction since the time when the improvements were made). The residue represents the valuation of the land.

We have dealt with economic efficiency and with administrative practicability, but we must not neglect social justice. For most people it would seem clear that the gains from land ownership should be specially taxed, most of all because so much of the gain arises out of the efforts of persons other than the owner of the land himself. This is not to imply, as is so often claimed, that speculation in land is economical harmful. Like other forms of speculation, it is on the whole economically beneficial (except perhaps during a period of rapid inflation of the currency, the serious effects of which are aggravated by speculation). Speculation in land is economically beneficial; but hard work in industry is as much or more beneficial to the general economy, and at the same time is highly taxed; it seems clear on all grounds that the proceeds of speculation in land should be highly taxed too.

It is now necessary to say a little about the economics of land, a subject on which so many people hold mistaken ideas, which may do great harm. Land, whether agricultural or urban, has an " economic rent" The scientific meaning of this phrase differs greatly from the meanings often given to it in discussions on, for instance, council house policy. The difference between the economic rents of various pieces of land is determined and measured by the difference between their potential productivity, agricultural and urban, given the present prices, availability of labour, enterprise and other factors of production, access to markets, the prices of their products and other special values such as those arising out of the partial exemption from death duties of agricultural land. Buyers of land generally make fairly good judgments about its potential productivity for this purpose or that, though, like everybody else, they sometimes make mistakes.

The economic rent of rural land in this country (though not in most parts of the United States) includes a non-cash element On most properties there is shooting, and on some fishing - it is true that these can be leased, and a cash value put upon them, but their enjoyment by the owner may be worth more to him than the cash value. Likewise the opportunity for social contacts with "the County," on which some purchasers of rural estates put a very high value indeed. There is also the less definite but very real pleasure of being able to stand on one's doorstep and look out over one's own property. Economic analysis, as well as social observation, shows that there are very real economic rents here, as well as from the net cash income.

Now the most important proposition of land economics, which nearly every body gets wrong, is that economic rents are the cause and land prices the consequence. The price of a piece of machinery or a building bears some relation to its cost of production, at any rate when it is new. Land cannot be produced and has no cost of production. It has a price solely because of revenue (cash or non-cash) expected from it, currently or at some future date.

Countless errors would be avoided if people understood this. The owner of a farm has paid a high price for his land, say the politicians, and, therefore, he is entitled to a reasonable return on it. Some economic historians have blamed high food prices in the past upon the high price of land - a straight confusion of cause and effect, showing how little economics some economic historians know. That high prices of building land are raising the cost of new houses is a favourite proposition with many politicians today. The truth of the matter is that in some parts of the country (though certainly not everywhere) building land is scarce and therefore commands an economic rent. All that we are free to do is to decide whether the rent will go into public or private revenue - but go somewhere it must. The high price of land is just a reflection of the existence of an economic rent.

Now we had better set out the workings of a land tax, in very simple algebra. The proposition with which we conclude is one which many people seem to find it very hard to believe, namely, that a land tax reduces the price of land - which it does in the ratio (tax plus rate of interest)/tax.

  • Let the rate of interest (expressed as a fraction of unity, e.g., 5 per cent as 0-05) be I
  • Let the rate of land tax, levied on the selling value of the land, be r
  • Let the selling value of the land be V
  • Let the full economic rent of the land (including discounted value of expectations of future rises in rent) be E
  • Let the full selling value of the land be F

The full selling value is the appropriate number of years' purchase of the full rent.

F = E/I

The tax levied is rV and the rent remaining to the owner of the land is (E - r V) which when capitalised gives the selling value.

V = (E - rV)/I

which can be simply rearranged as

V = E / (r+i)


E = V (r+i)

i.e., the full economic rent of the land is its selling value times the rate of interest, plus the rates of tax paid, which is a commonsense result. But we can also get the interesting result which cannot be inferred directly, namely, that

V/F = i/(i+r)

Thus, if the average rate of tax is of the same order of magnitude as the rate of interest, men the actual selling value of the land will be about half of its full selling value; and by raising rates of tax sufficiently the selling value of the land can be reduced to almost any figure.

We now consider the available evidence on the revenue which might be obtained by taxing land in [Australia], bearing in mind some relevant information from other countries also.

We may consider agricultural property first The Agricultural Economics Research Institute analyses records of the sale of farm land (which includes farm houses, other buildings, drains, hedges and other improvements to the land) in England and Wales. For our present purpose we require to examine sates with vacant possession. (In the case of land without vacant possession as things are at present the buyer may be taking on a tenant paying a low rent, whom he cannot remove for many years.) This price, which stood at £25 per acre in 1937-39, remained fairly stationary in the neighbourhood of £75 for most of the 1950s- a proportionate rise rather smaller than the rise in the selling price of farm products. With the amendment of the legislation which made it rather easier for owners of land to displace tenants and to raise rents, the price began to rise in 1958. By 1962 it had risen to £134, and a further rapid rise brought it up to £168 in 1963 and to £214 in 1964.

But before we begin exclaiming about this rise in values, we must stop to look at the value of "improvements" which would have to be deducted by a valuer under a system of land tax such as we have described.

Among all the abundant and costly research work in agricultural economics which has been done in the past two decades, there has been practically none throwing any light on the replacement value of farmhouses, other farm buildings, drains, fences and other structures.

Valuers usually give houses an economic life of 100 years or less; but farmhouses and cottages are usually well-built and well-maintained, and when old, often have more than usual attraction to buyers, whether for farming or for residential purposes, so they are here given an average life of 200 years. Of the "fences and gates" it is assumed that three-quarters consist of hedges which, if maintained, do not need to be replaced, and that the remainder would have an average life of ten years. Roads are assumed to need no replacement.

An agricultural economist might object that many fences and gates, … where they are abundant, and indeed some of the buildings, are not really required, and that possibly indeed the selling value of the farm might actually be increased by removing them. The selling value of vacant land … would almost certainly have been higher … by deducting depreciated improvements from the selling value of the farm. This may well be true, particularly if the farm is bought by a neighbour with a view to amalgamation into a larger farm. Nevertheless, any government operating a land tax scheme on the general principle of deducting the replacement value of improvements from the gross selling value of properties would have great difficulty in persuading farmers and public opinion of this proposition, and it is wiser to leave the deduction as it stands.

We now have to consider all land other than agricultural. For the purpose of estimating the amount which might be assessable to tax, we must exclude land owned by governmental authorities or which is used for public purposes - there is no point in asking governments to pay tax to themselves - though it is interesting to have some idea of the amount of such land, and what its market value might be if it were sold. Of the privately owned non-agricultural land, the most important distinction is between that used as sites for residences and their appurtenances, and that used for other purposes.

If we wish to compare land values in different places or countries, or at times when the purchasing power of money is very different, the most convenient way of doing so is to express site values or site rents per head of population, and then to express this latter figure as a percentage of average net national income at factor cost per head.

Considerable interest was aroused by the publication by the Rating and Valuation Association in February 1964 of the results of a trial survey of all land values in the town of Whitstable. The valuer was instructed to express his results in the form of an annual net rent, but in practice he had to work by estimating the selling values of the various lands when vacant, and assuming that rent was 4 per cent, of selling values. …

There is a most ingenious unpublished study which estimates a historical series for total non-agricultural residential site rents in England and Wales by an entirely different method, which was prepared by Dr. Singer, now a high official of the United Nations, when he was a research student at Cambridge in the 1930s working with the present writer. For the period 1845 to 1931 he reviewed the rating assessments, distinguishing reassessment years from others. In the years in which there was not a reassessment, the increase in rateable value showed the new houses constructed, less demolitions; the net value of construction of new houses in the reassessment years was also estimated from the average of the two adjacent years. The remainder of the increase in the reassessment years represented the assessors' estimates of the rise in value of all houses constructed before that year and still standing, due to rising land values plus rising replacement costs less depreciation. Dr. Singer adjusted this figure for physical depreciation, and for rises in replacement costs due to rises in building costs. The residue should have represented increases in urban residential site rents. Dr. Singer estimated an opening value for these rents at £3 million per annum in 1845, and cumulated the subsequent increases. The present writer has raised this opening figure to £6 million, and also made a further small adjustment for the increase in the site value of the land when it is used for residential construction for the first time, which escaped Dr. Singer's calculations. The results are very interesting. They show residential site rents rising from 2 per cent, of the national income in the 1840s to a maximum of 5 per cent, in the 1890s, then falling again to 3 per cent, by 1931. This corresponds to capital values of more than a year's national income in the 1890s, and about 60 per cent, of a year's income' in 1931. This method of analysis can be brought up to date (dealing in more recent years, however, with capital values and not annual values). These are the index numbers of the prices of already-built houses published by the Co-operative Permanent Building Society in their Occasional Bulletin; and data published by the Central Statistical Office (in their annual National Income and Expenditure) of the price of constructing (i.e., exclusive of land values) new dwellings, from which (after making a depreciation allowance) we can estimate changes in the replacement value of the stock of old dwellings.

We require a link between the end of Singer's series in 1931 and the beginning of the new series in 1938 or 1939. It appears permissible to assume that the ratio of residential site rents to net national income fell slightly during this period to 3 per cent; which annual value is capitalised to a selling value of 60 per cent of a year's national income, or £3 billion in round figures.

The Co-operative Building Society, who are greatly to be commended for their initiative in entering this important and unexplored field of research, probably deal On the whole with moderate to low income borrowers; but this should not bias the results, if the character of their clientele has remained about the same throughout the period. In any case, they prepare indexes separately for three types of houses.

Class One - houses built to satisfy modern standards and which will, in the opinion of the surveyor, remain in high demand for owner occupation by virtue of their construction, design and amenities.

Class Two - houses which are satisfactory as to construction, condition and amenities and are in general demand for owner occupation, but which, as a result of maintenance casts, internal design, the changing character of the district or other factors, compare unfavourably with Class One houses and are likely to depredate more rapidly.

Class Three - houses which fall short of present-day requirements and which may not remain in reasonable demand by owner occupiers, in view of maintenance costs, internal design, lack of modem fittings, district or other factors.

These three indexes have been combined with respective weights of 2, 2 and 1. In this way, any really serious bias due to change in the nature of houses sold has been avoided; though not all bias has been avoided.

P. Redfern gives the depreciated stock of housing at 1948 prices (which can be converted to 1938 prices). Redfern depreciated all dwellings by a uniform 1 per cent (straight line depreciation) of their original value for 100 years. A more accurate method of depreciation is that recommended by Bryan Anstey, namely, no real depreciation for the first thirty years of a house's life, followed by a straight line depreciation of 1-1/3 per cent, to extinguish its value in 105 years. Redfern's stock estimates for the end of 1938 and the end of 1952 were adjusted to the new definition of depreciation. The latter figure was the basis for all the post-war figures given below.

The reader may be surprised to see that, in spite of wartime damage, Redfern estimated the real stock of housing at the end of 1947 at 3 per cent, higher than at the end of 1938 (on the new method of depreciation, the post-war figure is higher still). However, according to an official document, the dwellings destroyed or damaged beyond repair during the war up to the end of 1944 were almost precisely offset by new building in the period September 1939 to June 1944 only. Allowing for the remainder of the year 1939, in which building was active, and post-war building beginning mid-1944, the difference can be explained.

The current selling price, inclusive of sites, of the whole stock of houses is based on the 1938 figure of the depreciated stock of houses plus land at the current price, increased first for the rise in the real volume of the depreciated stock of houses, and then increased again according to the index number of the price of existing houses. From this is deducted the current replacement value of the depreciated stock of structures, excluding land, computed first at 1938 prices, then converted to current replacement value by the index number of the price of new construction. The difference between these two results gives an estimate for the value of residential (not commercial) land.

The value of the stock of housing in relation to national income appears much higher than stated by Redfern, who (like Goldsmith in the United States) has applied too high a rate of depreciation to the stock of recently built houses. After reaching an unexpectedly high level in the immediate post-war years, the stock of housing, measured at replacement value (depreciation has been allowed for), is apparently settling down at a level of 1-4 to 1-5 of national income, as against 1-3 in 1938. In contrast to what many have hitherto believed, housing appears to be, in economic parlance, a "superior good," with a long-term income elasticity of demand greater than 1, i.e., as a community becomes wealthier it demands a stock of housing rising more rapidly than national income.

The value of residential land has also moved strangely. It reached a temporary high level in the late 1940s. (It is possible however that new houses then being very difficult to obtain, what appears in the calculation as a rise in site rents may have been a temporary rise in quasi-rents on the houses themselves.) During the 1950s it settled down at about the pre-war ratio to national income. It then began to rise quite suddenly, in about 1959, to an aggregate value of £30,000 million; and the rise may well be still continuing. But I think that a wise man would anticipate that both agricultural and residential land values, which have recently reached anomalously high levels, will soon fall.

We can also measure changes in the average price of sites for new houses, taken as they are, (The average area of site, however, may be rising or falling, and until we know this we cannot estimate the trend of average prices per acre.) Building Society Affairs, published by the Building Societies Association, contains an index, beginning in 1956, compiled by the Ministry of Public Buildings and Works, of the average price of new houses, including sites (compiled from information provided by building societies whose combined assets represent 60 to 70 per cent of all building societies). This can be extrapolated back to 1952 from Co-operative Permanent Building Society data. This appears to be an unweighted index, which will therefore be biased upwards if there has been an improvement in the size or quality of the average house, or an increase in the proportion built in London or other expensive areas.

In 1952-54 the average new house purchased by clients of the Co-operative Permanent Building Society cost approximately £2,000 inclusive of land. As this society may have been biased towards the low-income purchasers, we will raise this figure to £2,500 as a basis for the table, taking an arbitrary £250 as the value of the site then.

Between 1959 and 1963, the average price for a site for a dwelling rose by 72 per cent The aggregate value of sites occupied by the new dwellings constructed over this four-year period can be put at a little under £2 billion. Excluding this, the value of all sites rose from £14-1 billion in 1959 to £24-5 billion in 1963 - a rise of just over 70 per cent. It appears therefore that the orders of magnitude of the increases in the price of new sites and of old sites must have been much the same.

The Co-operative Permanent Building Society data also makes clear bow much this rise has been concentrated in London and the southern region. In 1952, average prices in London were not much above those of the following four regions, though prices were low in the last four regions. Between 1952 and 1964 building costs rose by 28 per cent. In every region house prices rose more than this, i.e., them was a rise in land values, quite marked in Scotland and Northern Ireland and smallest in the North Midlands. But the rise in London was not comparable with that in the other regions.

The proportion of income spent on site rents tended to rise as urban populations became relatively more dense in the nineteenth century. It was the arrival of the electrically driven tram-car in the 1890s, however, which began to chisel apart the compact Victorian city, reducing population densities, and at the same time reducing relative land values. This process occurred also In the United States, but was greatly accelerated by the increased ownership of private cars early in this century.

Recently the greater preferences for certain sites, and the restrictions imposed under the Town and Country Planning Acts, have been making the proportion of income spent on site rents rise again.

While the Whitstable survey estimated more than the national value for residential sites, it almost certainly gave too low a value for industrial and commercial sites, because Whitstable is far from being a major commercial centre.

On the new rating assessments, where a serious effort is now being made to estimate the current potential kiting value of properties (omitting agricultural and publicly owned land), residences represent about 60 per cent of the whole. But the ratio of structure to land value is probably considerably higher in residences than it is in industrial and commercial buildings, because the latter are generally built on more valuable sites, and in any case are generally (though not always) simpler types of structure. We should therefore expect the value of commercial and industrial sites to be of at least the same order of magnitude as the value of residential sites, and probably higher. This has definitely been the case in the United States (see Table 9). We may regard the high figure for 1929 as transitory; but even so, for a long period the value of industrial and commercial sites has substantially exceeded residential. The factors making residential site values lower, in relation to national income, in the United States than they are in Britain should also apply to industrial and commercial site values, though to a lesser extent.

Taxable land values in the United Kingdom therefore should be £4-1/2 billion agricultural, £30 billion residential, and about another £30 billion commercial and industrial. Mr. Anstey makes the total £57 billion in all, including £37 billion residential.

The figures for the United States were compiled by L. Grebler and others and by R. Goldsmith. Residential site values in the United States also reached a maximum in the 1890s, when they stood at 53 per cent of the national income, falling to 29 per cent, of national income in 1929, and remaining about stationary at 16 percent of national income since 1948. The trends are similar, but the whole level is relatively much lower than in this country, because of the higher proportion of the population housed in newer, relatively smaller and less crowded towns where residential site values tend to be lower. In Canada, where these factors prevail even more, the value of residential sites as a percentage of national income was found by O. J. Firestone to fall from 11 percent in 1921 to only 8 per cent, in 1950.

One possible explanation of the recent rise in land prices may be the increasing extent and strictness of planning regulations. The present writer is in favour of Town and Country Planning regulations. But we must not ignore the fact that the application of these regulations has the effect of markedly increasing the price of those lands where building is permitted - the existence of these artificially created values is indeed at prime reason for imposing land taxation. We can draw some conclusions here from the experience of Germany, which has had town and country planning regulations longer, and has enforced them more strictly, than probably any other country except the Netherlands. Any building land in the neighbourhood of Munich (admittedly a rapidly growing city) costs about £35,000 per acre: in England, the few residential sites at a comparable price would fall within the old L.C.C area.

The [data] for the United States …. shows residential site values as a proportion of the national income at their maximum in 1900. During the past decade, their fall has been checked, but not reversed. Industrial and commercial site values were definitely at their maximum, in comparison with national income, in the 1920s, and have shown some recovery in the 1950s. The apparently irreversible movement of residential she values must clearly be attributed to the motor car, already making itself felt even in the 1900s, together with (in the earlier period) improved railway and street car services, which enabled the congested populations of the nineteenth century towns to disperse to the suburbs.

The movement of industrial and commercial site values has been more complex. It has been best analysed in Chicago, a city whose criminal propensity is matched by the excellence of the University of Chicago. In the most central business and shopping area ("the Loop") real land values have been sustained, though not increased, over a long period. For many miles around it however real land values (i.e., allowing for the change in the value of money, but before applying a further deflating factor for the general rise of population and income) have fallen heavily. There are some districts m Chicago now where real land values are actually less than they were at the time of the Great Fire of Chicago in 1871. The recovery during the 1950s is noticeable in Chicago. But it has taken place "inside out." High land values now prevail around actual and potential commercial centres in distant suburbs; and the lowest values are in some of the older districts comparatively near to "the Loop."

It is very probable that such movements will take place in Britain. There are signs that the high land values reached in many parts of London during the recent boom are already beginning to slide. It is probable that a limited number of highly preferred areas (this has been the case with Park Avenue and Wall Street in New York) will retain and indeed increase their real land value; but that almost all other central urban areas will become "grey" or "blighted," with deteriorating quality of land users, and falling land prices.

The idea that "blight" can be met by "urban renewal programmes" is one of the most specious of all remedies - large though the sums may be which have now been devoted to it in the United States. Since the majority of people prefer to live in more distant suburbs, and since shopping and business centres are pushed outwards by traffic congestion and lack of car parking spaces, it stands to reason that very large areas of old urban land must lose their value. Anything which can be done, through fortunate chance or unusual effort, to maintain land values in one place probably only accelerates the decline of land values in neighbouring areas. Most schemes for "urban renewal" in fact represent a transfer of public funds to landowners, to help them hold up a land price which would otherwise have fallen. In Washington. Senator Douglas describes such proposals as running "a welfare state for the wealthy." Many people have supported them without understanding their consequences.

Even more unwise was the proposal, made by one research group in this country, for the state to acquire the sites of existing urban buildings, leaving their present owners to control their use for the reasonable lifetime of the building, with a view to the site subsequently being redeveloped by public enterprise. This scheme was conceived in the expectation that it would make a large profit for the Treasury. All the indications are that it would make a huge loss.

Certain points remain to be discussed regarding the working of a land tax. It has already been indicated that publicly owned land, used for public purposes, should be untaxed. Corporations for charitable, educational and religious purposes should be free from tax on land which they own and use for these specific purposes, though as much as any others they should be liable for tax on any land which they hold or buy as an investment (If we did not make this provision, then there would be a strong incentive for these corporations to use all their available resources for buying land, which would be good neither for them nor for the rest of the country.)

While at is one of the avowed purposes of land taxation to encourage owners of land to make the best possible use of it, and to develop it as quickly as possible, this raises the objection that there may be many buildings and open spaces of great beauty or historic interest whose owners are willing at the moment to preserve them intact, but who would not be so willing if they had to pay a substantial land tax on them. There is obvious force in this objection. The National Trust, whose charter provides that it should hold land and buildings "of historic interest or natural beauty" should clearly be fully exempted from land tax. Such an exemption should under certain circumstances also be accorded to other owners who have not actually handed their properties over to the National Trust (or who have offered them but have had the offer, for one reason or another, refused). It should be possible for any owner to execute a deed or covenant permanently preserving land from building, and upon so doing have the right to be taxed only on the agricultural value of the land. The situation, however, is somewhat different in respect of buildings of great historical or architectural value, or of those which, though not themselves of great architectural merit, nevertheless contribute greatly to the beauty of the landscape or townscape by the siting and spacing. Their owners may have the genuine intention of pre serving them permanently for the delectation of future generations; but no structure lasts for ever, however carefully it is preserved. For such building to be scheduled tax free therefore it would be necessary for the owner to establish a fund to endow the necessary cost of future maintenance, and also to submit a certificate from an independent surveyor indicating that the building was believed to be capable of being maintained in good condition for a long period into the future. In these circumstances, a partial exemption of taxation could be granted, i.e., the building could be taxed in accordance with its current use, and not on the potential value of the site for redevelopment, subject to the condition that the full tax would become payable if the building were eventually demolished because it could not be repaired.

.. Land values may be the subject, as they are in Australia, of both local and state taxation. It is desirable to make the taxes predominantly local, on the grounds that anything which brings an independent revenue to local authorities, and thus makes possible a greater decentralisation of government functions and responsibilities, is good in itself. On the other hand, much of the land revenue in the leading commercial centres, or in some of the wealthiest agricultural areas, will be greater than is needed by the local authority area in which the land is situated.

- Where the source of revenue is to be shared between two taxing authorities, there are some very important practical considerations which are sometimes forgotten, with serious consequences. For instance a scheme in which first the local authority and then the state imposed a tax, the latter allowing the tax paid to the local authority as deduction from the assessment, would have very bad consequences indeed. It would encourage the local authority to take more than they really needed, knowing that, because of full deducibility, their demands would have no effect on the taxpayer, though they would affect state revenue. The system in Australia, which at any rate is workable, is that all land is taxable by the local authority, and that the state taxes only holdings above a certain value in the hands of a single individual or corporation. This of course gives an incentive to distribute ownership of land among members of a family, and leaves the largest land taxes to be paid by corporations such as banks and breweries, which own large numbers of scattered valuable sites. (The above was also true of the Australian Federal Land Tax, introduced in 1910 with the deliberate object of discouraging unduly large individual holdings of land, and recently abolished.) As things are in Britain now, however, a greater aggregation of agricultural land in individual units appears desirable; and aggregation may also be beneficial with urban land. A project for additional tax upon large holders of land should therefore be discouraged. The best device for Britain appears to be a uniform system of valuation, by which land values per head of population should first be ascertained; then the state would impose a land tax which exempted altogether those local authority areas where per-head land values were low, and Which rose in a progressive scale for those with higher land values per head. Each local authority would then also impose its own tax, at a rate decided by itself, the state land tax not being deductible from the assessment This method is open to the theoretical objection that a rise in the local authority tax in a high value area reduces the selling value of the land and thus indirectly reduces the revenue of the state. It will probably be the case however that the land tax demands by local authorities in these high valued areas will not rise to a point where they would have serious effects in this direction.

It remains of course very important that each local authority should be left free to impose its own rate of tax; and that it should have to finance as many as possible of its services out of it. Economists and administrators with tidy minds would often like to impose uniform rates of tax in these circumstances, or, still worse, provide for some system of central collection and apportionment to local authorities. The advocates of such proposals fail to see that they are striking at die very roots of the principles of political responsibility. We want to see each local authority deciding on its expenditure, and then imposing the taxation necessary to meet it in full view of its electors, and responsible to them for its decisions. If any substantial fraction of a local authority's revenue comes as a grant from a higher authority, they win be much less careful in their spending of it, and will always be tempted to devote their energies in agitating for increased grants, which could better have been devoted to economising their expenditure, and increasing their area's prosperity and taxable value.

Speaker's Introduction to Discussion

MR. CLARK said that he was an advocate of land taxation and was more optimistic about it than he had once been. There could be no doubt about its practicability-it had been used m the State of Queensland and worked very well. Because the Ricardian principle of rent was not fully understood we tended to make the very serious mistake of thinking of rents as arising from the price of land. The truth was the exact opposite: the price of land was a consequence of the rent (together, in many cases, with expectations of future rent). Land taxation reduced the price of land: that could be shown both by mathematical demonstrations and practice.

It might be that residential land values, in relation to national in come, were now at their peak, though certain land values still seemed to be rising. There was, he thought, a tendency for urban land Values to fall in the long run. Urban renewal could do nothing to correct this. As Senator Douglas had said in Washmgton, it was a form of "welfare state for the wealthy," and subsidised the owners of formerly high priced land, whose value was failing, and it was, in fact, desirable that land values should fall. Exceptions should be made to the land taxation system in respect of open space and of historic buildings where they were being carefully maintained.


In discussion the following points were raised:

1. It was not clear whether there was a simple relation between the average rate of tax and the selling price of land, and to what extent interest rates entered the computation.

2. It was perhaps not significant to value the whole stock of current property by reference to current selling prices - prices, that is, at which marginal amounts of property changed hands.

3. Mr. Clark had produced evidence of declining site values, but his table excluded agricultural land (which, if included, would in fact strengthen his argument) and also public land (which had become very important over the last sixty years because the amount of public capital applied to land had risen while the amount of private capital had fallen). The inclusion of this public land might affect the argument.

4. The fall in residential values relative to national income might be true of the United States but did not seem to be true as yet in Britain because of planning restrictions. But it seemed probable that eventually Britain would follow the American model.

5. In certain inner areas land values were almost certainly falling, but in wide stretches of inner London they had been rising in recent years, though the situation was precarious. It was suggested that people wanted not so much to live in suburbs as to live in good houses, and that the inner areas might become considerably more attractive after renewal. It was true that inner areas were decaying more quickly than they could be renewed. But virtually no urban renewal scheme could be said to have failed, for every scheme provided better housing than was there before. On the other hand it might be argued that the improved housing would have been better provided on the periphery of the city. We knew very little about people's real preferences in this matter.

6. Agricultural land values were rising around the big cities in Britain, showing a potential rise in residential values there.

7. There was a difference between the Queensland system of land value taxation and the system Mr. Clark had suggested in his paper on site value rating. Under the Queensland system a valuation was made of the existing building on the site and the value of the bare site was deducted from this. The scheme might be practicable if there were sufficient evidence. Mr. Clark had advocated a system of valuing the bare site and estimating what sort of building it could potentially take. The main drawback of the Queensland system described by Mr. Clark was that it did not take future potential value into consideration.

8. Whatever the value of the land, it was suggested that taxation was not the best answer. If the land were lying fallow and were not ripe for development, it could be taxed. If it were ripe, it could be acquired by compulsory purchase. If it were already developed, it could be dealt with by income tax.

Speaker's Reply

Mr. Clark, replying, said that the tax fell on the current price of the land, so that interest did enter the formula. He had treated the interest level as fixed; this was an arbitrary assumption. He thought it was impossible to prove that the marginal value was the same as the average value. His personal judgment regarding land values in central areas of towns was that people wanted space, and would therefore move to low-value land. He accepted urban renewal in theory, but thought that it was quite impracticable (politics being what they were) and that in the United States too much public money had already been invested in it. In Chicago the negroes were convinced that urban renewal was a conspiracy to drive diem out of their old houses, and meanwhile new "grey areas" grew up nearby. An article in the January 1965 issue of Economic Geography showed in the course of fifty years a complete reversal of land values in Chicago, with the lowest values now recorded near the centre and the highest in the outer suburbs. He agreed that local authorities could profitably buy up suburban sites. But there were many administrative and political problems involved in this.

On the basis of valuation, he noted that the Whitstable valuer had postulated no less than three potential shopping centres. In Queensland the land taxation did catch the potential value.

He thought that owners should hold on to any agricultural land they held within seventy miles of central London. Dr. Stone had demonstrated the critical importance of this radius of seventy miles and he himself had confirmed it by studying rural population movements.

It had been asked whether there was any need to tax land if income were taxed effectively: the reply was that income taxes would recover the money only over a very long period of time.