Public Finance

§ 2. Regarding land itself as the ‘object’ to be taxed, the most obvious ‘unit’ in a new community would be that of a given area. Assume that none but very fertile land is cultivated, and that only in a simple manner, and the tax by area will be also the just one. Each unit is of about the same value and employs about the same amount of capital and labour. The early taxes on
Jugera in Rome and on ‘hides’ in England were probably at first based on this system, though they soon departed from it, and at present a few of the English dependencies retain it.
*1 But as soon as differences in qualities of soil and in modes of cultivation become noticeable, the method ceases to be fair.

Another form of land tax, that in proportion to the produce, is of greater antiquity. Eastern sovereigns receive their revenue usually in this manner. One-fourth, as in India, one-fifth, as in Egypt, or more frequently one-tenth of the yield was claimed by the monarch. This ‘tithe system,’ as it may be called, arose out of the ruler’s part-proprietorship of the soil. The proportional tax on produce was closely analogous to a
métayer rent. It was partly adjusted to the fertility of the land, and did not press so heavily on the poor soils as the area tax. Under a competitive system its immediate burden would fall on the consumers of agricultural products through the rise of price, though the ultimate effect would be to check cultivation, and therefore to lower rent. As the system has been generally applied to societies in the customary stage the pressure came on the cultivator, who is at once the producer and the chief consumer of those commodities.

These primitive methods are improved, either by arranging land in classes according to its quality and applying a different rate to each class, or by varying the proportion of produce taken according to the method of cultivation. As soon as the elements of fertility and proximity to market begin to tell, it is evident that a uniform rate falls with undue severity on the poor and distant lands, either hindering their cultivation or raising the value of all produce. Consequently we meet efforts at differentiation in various countries. Under the Roman Empire land in some provinces was divided into that of first, second, third, or other fertility, and the rate was adjusted accordingly; in others one-fifth or one-seventh of the yield was taken.
*2 In later times the Duchy of Mecklenburg had its land graded into three classes, with a different rate on each, and some of the Indian assessments have a like idea as their base.
*3 These modifications show some consciousness that the real value of the agent, land, is not to be measured either by its surface area or its gross produce. They are, however, but imperfect attempts at reaching the true aim of a land tax, the value embodied in the object. A tithe or other proportional produce tax does not allow for the expenses of production; as equal amounts of produce are often due to very different quantities of outlay, such a tax discourages the employment of capital and is practically inconvenient in the form of assessment.
*4 A classification of soils gives some, though insufficient, recognition to the influence of natural fertility. Far more is required. The effort to get at the true value of the ‘object’ is attained in respect to the land tax when it is levied on the net yield. The capital of the cultivator and the profits due to it have to be estimated in order to ascertain the income derived from the soil itself. Political equity and financial expediency have both contributed to this result; the fairest and most productive land tax is, on the whole, that which takes the net return as its standard. Fiscal practice tended in this direction. The financial reform of Diocletian seems to have adopted a unit of value, not of area (
Jugum), as the base for taxation of land. The English ‘hide’ came to be regarded as the ‘carucate’ of variable area but constant value.
*5

The mediæval land taxes are so much mixed up with rent and the incidents of tenure that little stress can be placed on their form; they are often parts of the older property tax, and only disentangled from it by degrees. Early English taxation ‘reached the landowner through his cattle, farming stock and corn and other produce of lands,’
*6 and the later subsidies had a rate on land as one of their component parts. In France the
Taille was developed from the feudal dues and became permanent in 1445. But wherever the system of taxing land had to be applied, the idea of taking its value as the real object of taxation came to the front, though the difficulties in carrying it out caused the frequent adoption of the ‘apportioned’ tax, as in the case of the
Taille, both the English ‘tenths and fifteenths, and the subsidies, and, too, in the German and Italian land taxes.
*7 The defects of these systems, with their exemptions and inequalities, made reform essential.

§ 3. It is far easier to point out the conditions of theoretic justice than to overcome the practical obstacles to arriving at the true net yield. The land tax system requires as its basis a valuation, and in the attempt to furnish this requisite various methods have been tried. Perhaps the simplest is that which follows the indications of the market, and uses as its guide the rent at which land is let. There is an obvious advantage in keeping close to the facts, but there is also great difficulty in ascertaining them correctly and following their successive changes. Adam Smith approved of the use of registers of leases, which he would make compulsory, and by their aid assess the land of occupying owners.
*8 The selling value is another possible criterion; it is evidently related to rent as principal is to interest, and for short periods the proportion is steady. A tax directly based on the selling value of land is, however, a tax on property rather than on income.

The difficulties in ascertaining the actual rents, and in some countries the large proportion of occupying owners, have popularised the system of determining the value of land for taxation by official assessment based on survey and valuation. This method is evidently the older one. Thus the Roman provincial land tax had a survey as its foundation followed by valuation.
*9 Domesday Book is a less perfect example of the same kind in England, and in one form or another valuations were common enough in the Middle Ages, but were in general used only for the ruder forms of land taxation, and dealt with the gross produce from the soil or its supposed capital value.

Refinements in fiscal methods require a corresponding elaboration in the valuation, or, to use the serviceable French term,
cadastre,*10 on which they depend. Most of the controversies about the land tax turn on the method of cadastration, and the expediency of its revision at stated periods. For the completion of a
cadastre a series of processes is needed; there must be the measurement of the surface, and its delineation by maps; the boundaries of properties must be marked, and the ownership specified. To this technical work the economic task of valuation succeeds. Estimates of produce and prices and of the cost of cultivation form the
data on which the ‘net annual value’ is calculated. Each of these steps involves much labour, and is liable to error, more particularly in the economic part of the work. Produce must depend on the method and skill employed in cultivation, prices on many different conditions, and both, especially the latter, are fluctuating. Besides, to be really useful, a fiscal survey must deal with minute portions of the soil; each distinct piece (or
parcelle) should be valued and revalued at intervals. Such an inquiry takes a long period to accomplish for any country, and by the time it is completed the results for the districts first treated have become antiquated.
*11 However perfect when first started, a valuation must soon fail to represent the actual position of the land it deals with. The opening of new lines of communication, the adoption of a different style of farming, and the growth of towns will completely alter the old results.
*12 The imperfections of the
cadastre are grave enough from a theoretical point of view, but they also entail much hardship and injustice. Some persons and districts are unduly favoured, leaving to others to make up the amount that they have escaped paying. For example, in France (where the land tax is apportioned), some proprietors are taxed four times as heavily as others. The differences in Italian taxation were still greater, owing to the use of different cadastral bases for different districts.
*13 Between the difficulties that adherence to an old valuation causes, and those due to the expense and confusion of incessant renewals of the
cadastre, it appears that the safer course is to keep the original valuations checked by the actual letting values of land. Apart from the expense of continual revaluation, it is also true that the ‘net annual value’ or the ‘net income’ of official estimations is in a sense hypocritical, as it depends on the accuracy of the assumptions made for the purpose. There is, however, the qualifying fact that a well-executed
cadastre is of use for other purposes. A careful survey is essential for facilitating the transfer of land, so that it is merely the economic part that could in any case be dispensed with. There seems to be no great obstacle to a gradual revision of the general valuation, supplemented by local valuations strictly on the letting value. In this way the former would be a slowly changing norm, while the latter would recognise the actual movements of land value.

§ 4. But whatever may be the hindrances in the way of securing a perfect adjustment of the land tax, there is no doubt that most financial systems use it as a substantial resource. The so-called English land-tax has really been converted into a rent charge;
*14 but Schedule A of the income tax comprises land and houses, the former in 1899-1900 being £52,814,000 in value, yielding at the rate of 8
d. less than £1,800,000. To this sum has to be added the portion of local rates falling on land. Using the proportion of land to houses under Schedule A as a guide to the division of rates between the two classes, we would get about 23 per cent. for the share of land. As the rates for 1898-9 were over £38,600,000 this method would give £9,070,000 as the local charge on land for England and Wales.
*15 As, however, the poor rate valuation differs materially from that employed for income tax, it becomes necessary to consider the estimated distribution of local burdens. According to Sir H. Fowler, land in 1891 bore a little over 15 per cent., while houses contributed nearly 85 per cent.; or, to put the matter in a simpler form, land paid only one-sixth, against five-sixths derived from houses. Thus the true contribution from land in England and Wales would be somewhat over £6,400,000.
*16 Scotland and Ireland show a larger proportion; 50 per cent. would in their case be the share of land, and, we may say that, of the total £7,000,000, £3,500,000 would fall on land. The extent to which the taxation of houses falls on ground rent is—at least for statistical purposes—an insoluble problem; but, omitting it for the present, we get a total taxation of nearly £10,000,000 on land for the United Kingdom. How far this represents a charge on pure land value, as distinct from that on investments of capital, is questionable. We need not in any case hesitate to ascribe the greater part to the value of the land, not to the improvements. When we add to the above the tithes and tithe rent charges, so far as they are devoted to ecclesiastical or other public purposes the total reaches £12,500,000. Allowing for considerable under-valuation in the figures of Schedule A,
*17 it is nevertheless beyond doubt that land contributes largely to the public requirements.

At the same time we must remember that a great deal of this burden is of long standing; the income tax has been for sixty years in continuous operation, and, in the early part of the century, the poor rate was excessive. There is no evidence of new and oppressive charges being imposed. The growth of local taxation, as Lord Goschen has shown,
*18 has chiefly affected the towns, while, until recently, the rent of land was rising.

§ 5. On passing to France we meet with a very different system of land taxation. The old
Taille, whose defects were universally recognised, was supplemented in 1710 by the
Dixième, and from 1748 a
Vingtième was levied. These ‘tenths’ and ‘twentieths’ were rather income, than pure land taxes, but were abolished at the Revolution along with the
Taille, and the modern system was inaugurated.

The decree of December, 1790, established the
Impôt foncier, which was to be apportioned on all landed property in proportion to its ‘net revenue.’ This phrase is evidently due to physiocratic influence, and was explained to mean what remained over after all expenses were deducted from the gross produce. The tax was to be a fixed sum apportioned among the contributories, and to be payable in money. It was not to exceed one-sixth of the net revenue, and, on the loose estimate that 240,000,000 francs would be one-sixth, the contribution was fixed at that amount, with an additional 60,000,000 francs for local taxation. The disturbances of the Revolutionary period hindered the collection of this impost, and the unequal pressure, owing to the absence of proper valuations, was the ground of successive reductions, by which the total amount, from being 240,000,000 francs in 1790-6, fell to 150,000,000 in 1821. In 1835, the increased value of house property, which is included by the law of 1790, was taken into account. The additional
centimes—really an increase of the tax—were given up in 1850, and by 1880 the total amount was almost 174,000,000 francs (£7,000,000). The extra
centimes for the departments and communes were very nearly trebled in amount since 1820; in 1880 they were 94,000,000 francs and 82,000,000 francs respectively.
*19 The loud complaints of agriculturists as to the inequalities and unjust pressure of the
Impôt foncier led to a reform in 1890, by which the house tax was separated from the land tax, and the latter, which had been 118,000,000 francs, was reduced to 103,000,000 francs. By a law of 1897 the smaller properties were relieved. The result has been that, in 1900, even with an extra charge of 8 per cent. on the original general tax, the total taxation on land stands at 253,000,000 francs or £10,120,000. The increase of house property and buildings has supplied a new object for the heavier taxation, as in the case of England. The land tax remains one of apportionment, while the house tax, or more strictly that on land with buildings (
Propriété bâtie), has become rated, and is fixed for the present at 3.20 per cent. The next step in reform will probably be the abandonment of the apportioning of the land tax in favour of the more suitable rated system.
*20

§ 6. The Italian land tax is a development from the taxes of the several Italian States. As the simplest course, 110,000,000 lire was the amount fixed for apportionment among the different divisions. Measures of reform have been since attempted. The tax on buildings was separated in 1865 and made a rated tax, and redistributions of the total charge among the provinces were carried out. The defective scheme of the old
cadastres has led to the enactment of a law prescribing the preparation of a new and uniform one for all Italy. The variations in amount of the land tax have been from 125,000,000 lire to about 96,000,000 lire,
i.e. speaking generally, from about £4,000,000 to £5,000,000, but the local taxation has to be added. Thus for the year 1886-7 the provincial tax was 53,000,000 lire and the communal one 76,000,000 lire, which, with 110,000,000 lire, the general land tax for that year, made a total of 240,000,000 lire (£9,600,000)—a much higher charge than that of France. In qualification it must, however, be noticed that the whole taxation of Italy is far heavier. The most serious grievance is found in the instances of heavily taxed communes, where the greater part of the value of land is absorbed in taxation. So far has this been carried that there have been many cases of evictions by the State.
*21 Inequality in distribution and excessive weight in amount are the gravest possible defects in any tax. The new valuation, though costly, will remedy the former, but the latter is a question of policy as well as finance.

The Spanish land tax, which received its present form in 1845, includes stock, and is therefore more primitive. Owing to the want of a correct valuation, the charges are very imperfectly distributed. The proportion fixed for 1890-1 was 15½ per cent. on those places that have given a satisfactory declaration of value, for others 17½ per cent. The yield for 1900 exceeded 160,000,000 pesetas (£6,680,000), with over 17,000,000 pesetas (£680,000) for local purposes. The law of March, 1900, makes provision for a proper valuation of houses, land, and cattle, which will increase the efficiency of the land tax.

The Portuguese land tax is closely on the lines of the French
Impôt foncier. It was originally rated, but since 1852 has been apportioned; it, is however, proposed to return to the rated method. The yield is nearly 3,000,000 milreis (about £650,000), after paying assessment expenses.

Belgium has a rated tax based on an elaborate valuation. Up to 1867 the method of apportionment was employed. The annual amount for national purposes for 1900 was over £1,000,000, with additional
centimes for local government of nearly £700,000.
*22

Greece, which possessed the tithe system till 1800, has now a rather primitive but yet complicated group of ‘land taxes’ on labouring animals, on area, and certain products, yielding altogether about £500,000.

§ 7. The land taxes are confined to the several States of the German Empire, the imperial revenue being derived mainly from indirect taxation. With numerous differences in detail, there is the general system of basing the tax on official valuation. The Prussian land tax, inherited from the 18th century, was reformed in the period 1810-20; a new valuation was arranged, and inequalities in the distribution between the different provinces modified; but the survivals of the older system of privilege prevented complete success in this object. In 1821 its yield was under £1,500,000. These inequalities were dealt with by the legislation of 1861. The house tax was separated, and for the land tax the amount was fixed at 10,000,000 thalers (£1,500,000) from 1865, and a fresh valuation carried out The new Prussian provinces, acquired in 1866, added 3,200,000 thalers (£480,000) to this fixed sum, giving a total of £2,000,000. The amount of the additional local charges was somewhat uncertain, but for the year 1880-1 the communal and provincial extra land taxes were equal to those of the State in amount (£2,000,000), giving a total burden on land of £4,000,000, independent of the action of the income tax.

Under the legislation of 1893 the state land tax, in common with the house and business taxes, has been surrendered by the Prussian government in order to provide the local subdivisions with adequate objects of taxation. This long-proposed transfer only came into effect for the financial year commencing April 1, 1895.

Each of the smaller German States employs some form of land tax. Bavaria shows a less developed form in its reference to gross produce as the basis of calculation. The cadastral surveys are in most cases elaborate, and serve other than fiscal purposes, such as facilitating the transfer of land. The
communes of the several States also receive contributions through additions to the land tax.
*23

Austria has developed a land tax on a similar type. By the reform of 1817 the valuation of the 18th century was to be replaced by a new one completed in 1856. The house tax was separated in 1820. In 1879 a law for revision was passed, and in 1881 the annual amount was fixed at 35,190,000 florins for fifteen years, a new valuation to be then made. The Hungarian land tax was almost the same sum (35,000,000 florins), and the local charges in Austria levied on land were believed to reach the like amount. Thus the burden on land in Austria proper is under £7,000,000.

Taxation of land in the United States is imposed through the general property tax, which, as we shall see, presses with undue weight on real property, but its discussion belongs to a later chapter.
*24 Nor need the Indian land revenue be again considered.

§ 8. The foregoing notices of the land taxation of some of the principal countries bring out its characteristic features Specially worthy of observation are: first, the considerable amount contributed on the whole, and to both general and local revenues. The absolute amount appears to be highest in England, but everywhere a good percentage of the net annual returns is taken for public use.
*25 Another very common circumstance is the employment of the system of apportionment. A total fixed sum is thus secured, and as each district must pay its part, it has a manifest interest in making all contribute fairly; nevertheless, the method has the great defect of rendering an important part of the tax revenue inelastic, and it is likely to reduce the land tax to a rent charge, as has happened in the case of England. The rated or percentage system is free from these faults, and is therefore the best suited for modern finance. A third question intimately connected with the land tax is that of valuation. If the ‘rated’ system be used, it is necessary in the interests of justice that the basis on which the estimate of value is made should be uniform. Thus
e.g. the English valuation of land is believed to be closer to the true value than the Irish one, from which it follows that the income tax in its A schedule is not the same in the two countries.
*26 The Italian land tax affords a more extreme instance of the same evil. In all countries this inequality must in some degree exist between individuals and smaller districts, but this fact only strengthens the claim for all practicable efforts to secure the removal of proven injustices. Even if it be impossible to alter quickly the particular forms of the tax, there is an advantage in knowing the amount of inequality, which can then be compensated by the adjustment of other taxes.

Finally, the land tax is what has been called a ‘real’ tax; it deals with the object, land, and takes no note of the position of the proprietor. When properly developed it is proportioned to net produce, and therefore allows for the expenses of working the soil. For the same reason it should not take indebtedness into account.
*27 Charges on land are a part of the net return, and have no claim to deduction. A variable land tax may therefore press with great severity on encumbered proprietors who have to pay the tax on the interest of their debts. Any attempt to remedy this evil has the necessary result of creating a partial tax on interest of capital, and, if unaccompanied by taxation of other forms of capital, would either discourage loans to owners of land or raise the interest on mortgages. The conclusion suggested by these facts is that the land tax had best be absorbed in a general income tax, when part of the burden would, as under existing English arrangements, be paid by the creditors. If, on the other hand, the distinct land tax be retained, two courses are open: either to retain it at a fixed amount, when it becomes a rent charge, an undesirable proceeding, or to give it up to local bodies. We have seen that taxes on real property are a good form of local revenue,
*28 and both in France and the United States this treatment, which is in accordance with British practice, has been proposed. The actual condition in Germany with its numerous smaller States partly attains this result, which has been reached in Prussia by the reforms of 1893.
*29

§ 9. The incidence of the land tax is a final question for consideration. In its ruder forms the pressure fell chiefly on the actual cultivators, though the ultimate effect of heavy taxation must have been felt by the proprietors in the check to agricultural improvement and the diminution in their dues. On the hypothesis of competition, a proportional tax on produce,
e.g. a tithe, would tend to raise prices, and thus at first fall on the consumer, unless there was free importation of the article from abroad. Such was Ricardo’s reasoning in respect to tithes, which had to be paid from land at the margin of cultivation, and which consequently yielded no rent. He failed to see the inevitable effect of dearer food in retarding the progress of the community, and thereby preventing the increase of rent. The pressure of a tithe is surely, as time elapses, in greater degree passed on to the landlords.
*30 As soon as net return is taken as the standard for taxation, rent is the element affected. A land tax, therefore, in its developed form, may not inaccurately be regarded as a tax on rent, and the general principles of incidence applied to it. In actual working, however, various complications arise. The action of competition is not always found in full force, and so far as any portion of the pure economic rent is held by the immediate payer—tenant or other—he has to submit to the burden.
*31 A land tax may also affect the interests of labour. If investment of capital in agriculture is checked, and if the rate of wages is easily affected by the action of employers (as has been often the case), taxation on the cultivator may be shifted, not to the landlord by lowering rent, but to the labourer by lowering wages, or in a time of rising prices by preventing their proportional increase in money.
*32 Again, the fact that the land which is the object of taxation often owes its value to the capital sunk in it makes the burden fall on the yield of fixed capital, a point which has been already considered.
*33

A more difficult and disputable point arises in connexion with the incidence of a long-continued land tax. Here it is said that the tax is really a deduction from property. As land is sought for its revenue, whatever lowers its revenue lowers its selling price, and therefore a land tax falls altogether on the possessor at the time of its imposition. Subsequent acquirers take the land subject to the burden, and pay a lower price in consequence. This process of ‘amortisation,’ as it has been called, makes the subsequent removal of the tax undesirable; the persons who have lost by its establishment are not the same as those who gain by its remission. A purchaser buys land at a lower price in consequence of the tax, and gains a like advantage by its removal; in fact, he is allowed for it twice over, once at the time of purchase and again at that of remission.

The elements of truth in this theory, which has received much favour, appear to be the following: (1) as previously pointed out,
*34 when a land tax becomes definitely fixed, so that it can be foreseen, or even capitalised and redeemed, there is no inaccuracy in speaking of it as a charge on land which lowers its selling price; it is just the same as a mortgage, and is so regarded by purchasers; (2) a stable tax of any kind has some of the advantages to which Canard gives such exaggerated importance. Its pressure is more regular, and therefore less felt. An invariable land tax undoubtedly has this in its favour. On the other hand, there is no reason for regarding the modern land taxes as perfectly stable and fixed. In transactions with respect to land there are not merely the existing but the prospective burdens to be taken into account. To assume,
e.g. that the French ‘
centimes additionnels‘ or the English local rates have been ‘amortised’ would be an obvious error. We cannot foresee the future movement of taxation in respect to land, and we cannot expect that the present systems will always continue. Another important consideration is the relation of land taxation to the other forms. If it should happen to be unduly heavy there would be a tendency to depress the value of the land so taxed, just as if it were too light its effect would be the opposite; but this is characteristic of all taxation. Tea, sugar, or any other commodity will have its value for the time being affected by the creation or remission of a special tax on it. But where there is a due proportion of taxation to the several forms of income, the investor in land will only receive the same proportional return as he would obtain in other directions. Any alteration in the land tax ought to have as its motive the effort to secure a more equal distribution of burdens, and to this there can be no valid objection. At the same time, where a tax has been recognised as at once special and definitely fixed, it seems to pass out of the ordinary category of taxes and into that of charges, a transformation only possible in the case of durable productive wealth, and most prominent in respect to land.