Risk of error arises in three ways : (i.) Evasion in the tax system itself.
(ii.) Legal omissions from the scheme of tax (i.e., "garden produce" as non-taxable income, "enjoyment" income from mov able property).
(iii.) The basis of capitalisation, viz., the number of years purchase adopted.
This method, generally known as the "Giffen" method, though not invented by him, is the main basis for the valuation for the United Kingdom. It will be found to a limited extent available in the United States, South Africa and other dominions.
(b) Taxation of income on Individual Returns.
Where statistics of this character are available, they may be utilised for capital valuations, but only with some difficulty. If there is a rough division of income into earned income and income from property, it is, of course, of assistance in the capi talisation. The chief defects are : (i.) The considerable extent to which evasion takes place in this particular type of taxation.
(ii.) Omission of all income held or accumulated collectively.
(iii.) Difficulty in determining the ratio of income to capital on the average, which makes capitalisation a far greater difficulty than under (a) (iii.) above.
(2) Based on Material Provided by the Annual Taxation of Capital.
(a) Particular Classes of Property, such as Land or Build ings.
Obviously these details supply a part only of the whole capital valuation, and they more properly belong to the "inventory method" referred to below. Unless the values are regularly re vised on uniform lines, without local differences, they form but a rough basis, and there are always difficulties in determining the extent to which other forms of wealth (i.e., company shares or business profits) duplicate these values. Some of the Continental systems of taxation supply material of this order, and the Aus tralian States have regularly revised valuations which are valuable because they constitute so large a fraction of the total wealth.
(b) General Property Valuation.—The particulars furnished by a system of annual taxation upon all classes of property, should, in theory, form an ideal basis for a valuation. As a matter of fact, however, in practice, even such a tax as the General Prop erty Tax in the United States, is full of defects. The valuations of personal property tend to disappear altogether (as was the case during the eighteenth century with the British "Land Tax") or else to be negligible in amount, leaving real property alone to bear the burden. This real property is assessed on very diverse
lines in different areas, and is admittedly much below the selling values in many States.
(3) Based on Data arising through Taxation of Capital at Irregular Periods.
(a) Statistics of "Estates" chargeable with Duties on passing at Death.
This method has the appearance of being the most satisfactory and scientific of all. A special ad hoc valuation is made periodi cally of all wealth held in individual ownership, and it is only required to ascertain what proportion of the whole comes under review in any given year, or, alternatively, at what intervals of time the same item of wealth will be recharged to duty on the average, in order to compute the total wealth belonging to in dividuals. But this apparently simple task is, in practice, fraught with many difficulties, and the method of ascertainment of the "multiplier," though greatly improved of late years, is still open to doubt or inquiry upon important points. The adequacy of the capitalisation of collective wealth and the impossibility of saying how much is. not reflected in their values are serious drawbacks.
(4) The Inventory Method.
This method aims at a valuation, in the aggregate, of each "form" in which wealth is embodied, without regard to the owner ship by individuals, companies, etc. It is often called the "objec tive" method. It depends for its success almost entirely upon the existence of statistical material compiled for other purposes, e.g., import and export statistics, local government taxation figures., expert valuations of mineral resources, statistical enumerations of objects to which an average value can be applied. Examples of the last mentioned are the valuation of shipping by reference to the total tonnage multiplied by an average value per ton, or of mining capital by the average capital invested per ton of output, or of live stock by the number of each kind multiplied by an average price, or even of business, by a co-efficient. There, are few classes of statistics that have not been pressed or coaxed into service for the "inventory" method, and further illustration can best be seen below.