Public Finance

§ 1. At the opposite extreme to capitation or personal taxes are those that are imposed on property. The antithesis between ‘persons’ and ‘things’ or, in economic language, between services and commodities, is apparent in the earliest stages of society. When the period of contributions in kind is passed, the first objects of taxation are the persons and property of the subjects. The property tax is probably older than the separate charges on the yield of land, or capital, or even labour; the sum of existing wealth is an easier, though not so fair an object for imposition. Land, slaves, and oxen—the
res mancipi of Roman law—with household goods generally are the commodities that first fall under taxation: they are on the spot, easily estimated, and in most cases proportioned to the land under cultivation. As society advances and new forms of wealth come into existence, the injustice of the old system becomes evident, and taxation is extended to movable property, either by special taxes, or, more generally, by including it in the category of taxable objects. The difficulties in the way turn out to be too strong: personal property gradually escaped from the duty of contributing.
*67 Such seems to have been the fate of the property tax wherever it has been tried; in ancient Rome, in the various attempts in England, as also with the French
Taille and the later
Dixièmes and
Vingtièmes of the eighteenth century.

This very general tendency to disintegration in the property tax is partly due to its economical defects, partly to technical difficulties in its administration. As regards the former there can be no doubt that, speaking generally, the property tax is merely a form of assessment, the payment being really made out of income. Taxation that falls on capital in the strict sense must diminish the sum of the community’s wealth, a process that cannot continue indefinitely. In using property as the basis for taxation there is always a danger of trenching on the accumulated resources of the society. A second obstacle lies in the fact that property is not really a fair gauge of taxable capacity. Some forms of wealth give a lower return than others, and in special cases may even involve outlay. If, as we saw,
*68 income or revenue is on the whole a satisfactory standard for taxation, a property tax, unless carefully balanced by other charges, is unjustifiable. It is the result of a confused idea as to the true measure of taxation.

The technical difficulties result from the nature of property. In many cases it is only an abstraction obtained by capitalising revenue. This is pre-eminently true of the great mass of property in which the modern stock exchange deals. Shares of companies and public debts are only of value in consequence of their revenue, and their capital value is reached by a process of estimation; it is besides constantly varying in a way that does not allow of precise measurement. Income is a definite receipt during any given period, and is therefore a better object for charge. The difficulty of reaching the multifarious forms of personal property is a further objection. To arrive at the amount of taxable wealth and to assess it fairly is quite impossible. The ‘slipping away’ that always takes place leads to grave inequalities and injustice. The owners of certain forms of wealth are unduly burdened by having to pay the share of those who have evaded their duty. These are sufficient grounds to justify the very general abandonment of the property tax as a leading source of revenue.
*69 Taxes on produce (
Ertragssteuern), such as those discussed in preceding chapters of this book, take its place or survive it, while they in turn tend to develop into the income tax.

§ 2. The property tax has, however, maintained its ground in two countries. Switzerland still possesses, but in most instances with great modifications, this ancient method. Though the central government does not avail itself of the property tax—except in the charge for military exemption—all the cantons employ it. The forms adopted are varied, complex, and often changed.
*70 Their characteristics will be best understood by taking a single canton and examining its system. That used in Zürich divides property-holders into classes. The lowest, those under £800, pay on one-half only; the second, those between £800 and £2,000, pay on one-half of £800, and on three-fifths of the excess. Between £2,000 and £4,000 taxation is imposed on seven-tenths of the excess over £2,000; between £4,000 and £8,000 on four-fifths of the excess. For property under £16,000 only nine-tenths of the excess over £8,000 is charged, while any amount over £16,000 is charged at its full value.
*71 Under such a scale the smaller properties escape very easily. The Zürich method is modified in other cantons. In Graubünden the lowest class is charged at the ‘simple’ rate; in the next class 10 per cent. additional is placed on the entire property; in the third 20 per cent., and so on till in the eleventh class the rate is double. The more primitive canton of Uri has a higher rate of progression; from 1/20 of 1 per cent. on property under £1,200, it rises to 3/20 of 1 per cent. on properties over £16,000. The town canton of Bâle makes but three classes: 1/20 of 1 per cent. is paid by estates under £4,000; 3/20 of 1 per cent. on those between £4,000 and £8,000; and 1/5 of 1 per cent. on those over £8,000. In some cantons there is no progression, all properties being taxed at the same rate. Communal taxation is also in many cases levied on property, but it is rarely progressive (
e.g. in Zürich
communes are forbidden to impose a progressive rate), and generally moderate in amount.

The Swiss system of property taxation suggests several points of financial interest. Though a long-established form, it has been gradually adjusted in accordance with modern ideas, and is used to supply gaps in the other kinds of taxation. The aim of taxing permanent incomes at a higher rate is accomplished by a tax that does not touch pure earnings. Non-revenue-yielding wealth is also reached, and the democratic ideal of reducing the burden on the smaller incomes is in some degree realised. But notwithstanding this tendency, the rates are so moderate that the effect on capital is hardly perceptible. Evasion perhaps accounts for a good deal of this indifference on the part of the wealthy, and shows that the administrative system is far from perfect. Again, the very narrow areas within which the several systems are applied, and the smallness of the populations affected, make the operation of the taxes more difficult to use for generalisation.
*72 They are, in fact, a remarkable form of local taxation, and should be so regarded.

§ 3. One of the many points of likeness between the American States or ‘Commonwealths’ and the Swiss Cantons is their use of the general tax on property. But on closer examination the special differences are more important than the general resemblance. The American tax is not in any case progressive, and is rarely accompanied by anything resembling an income tax. Another feature of difference is the apportionment system adopted in the United States. A given sum has to be divided over the several counties of a State in proportion to their assessment, and the valuation of property is in consequence put by the county officials at the lowest figure admissible. The system adopted in Ohio may serve as an illustration of the general methods. By a constitutional provision all property (with some insignificant exceptions) must be taxed. To carry out this law real property is valued once in ten years by assessors appointed for the purpose, who are to take each plot ‘at its true value in money.’ As the assessors in each county compare results they are probably uniform, but as between different counties there is often great difference, which is corrected, though imperfectly, by a board of equalisation. For personal property an elaborate series of queries is issued to each adult, who is bound to answer them, and to swear to the truth of his return. The number of cattle, watches, pianos, merchandise, money, stocks, bonds, &c., have to be declared, and their selling value stated.
*73 Nothing could apparently be more searching and effective. Other States possess tax laws quite as rigorous. In Georgia both land and personalty are included in the queries issued, which, moreover, contain a question as to evasion. Unfortunately the universal experience is that the greater part of personal property is not returned. Assessors’ reports, Governors’ messages, and reports of tax commissions all dwell on this fact. The New York report of Mr. Wells in 1871 is quite in agreement with the Maryland report of Professor Ely in 1886 while Professor Seligman declares emphatically that ‘the general property tax as actually administered to-day is beyond all doubt one of the worst taxes known in the civilised world.’
*74 The reasons for this general condemnation are not far to seek. They are, first of all, lax administration. Officials elected for short terms cannot be expected to scrutinise very closely the answers of their constituents. Palpably inadequate returns are accepted with little question, and the wealthiest get off best. A second cause is the local nature of the property tax, as compared with the national, or even universal movement of the finer forms of personal property. Bonds and shares are easily moved outside a State during the time of assessment, and more obvious forms of capital have to be leniently treated to avoid their emigration. Mr. Wells has pointed out very forcibly the discouragement to capital that the New York system gave,
*75 in contrast with those of Pennsylvania and other adjoining States; but in practice the pressure is very slight. One fact suffices to establish the defectiveness of the property assessments. It is the decline in the declared value of personal property during a period in which wealth has beyond question increased enormously. The personal property in New York State in 1869 was assessed at $434,000,000, in 1875 it had fallen to $407,000,000, and in 1885 to $332,000,000,
i.e. a decline of over $100,000,000 in the commercial centre of the Union. The similar figures for real property are, for 1869 $1,532,000,000, for 1875 $1,960,000,000 and for 1885 $2,762,000,000, or an increase of nearly $1,230,000,000.
*76

The defects of the American property tax are, it would appear, beyond remedy, and therefore it may be anticipated that it will in the future be transformed into a land tax with additional charges on other selected receipts, and perhaps finally into an income tax.
*77 We may, however, conjecture that a system of state income taxes will also fail owing to the difficulty of localising income. The conclusion already reached
*78 that the income tax is best suited for the national government applies fully to the United States. The most promising sources of state revenue seem to be land and license taxes.
*79 But whatever be the new forms adopted the property tax is decisively condemned.
*80

§ 4. Notwithstanding the weight of past experience, there has been during the last few years a distinct reaction in favour of the taxation of property. Democratic sentiment and the latest financial theories have conjointly supported the reintroduction of a charge on realised wealth as such. The most important instances of the actual adoption of this policy are supplied by Prussia and Holland. In the former country there has been an extensive recasting of the revenue system, which has as one of its salient points the imposition of a tax on property. It should, however, be noticed that this new tax is closely connected with the reform of the income tax,
*81 and is expressly described as a supplementary tax (
Ergänzungssteuer). Its functions, according to its promoters, are (1) to impose heavier taxation on ‘funded’ property, (2) to cover the gaps left by the income tax, and (3) to put the financial position on a sounder basis.
*82 The rate chosen is moderate, amounting to about one mark for each 2,000 marks of property, or to one shilling for £100.
*83 Consequently the anticipated yield for the first year of levy, 1895-6, was 35,000,000 marks (£1,750,000). The actual receipts in 1897-8 were £1,555,000 (31,100,000 marks) in 1899-1900 they reached £1,680,000. Now this, as we shall see, is less than 20 per cent. of the return obtained from the reformed income tax, and hardly seems enough to justify the employment of an intricate and complicated system of taxation.

The Dutch measure, though the outcome of similar tendencies, yet differs in one most important respect. It is intended to be at once an income and a property tax, and is correlated not by a general income tax, but by a professional or vocation tax. It is thus complementary rather than supplementary. A combined income and property tax, in a country like Holland, must necessarily be more productive than a tax on earnings; and accordingly the estimate of the property tax for 1894 was double that of the tax on professional incomes. In this case too, the rate is not excessive. Properties under 13,000 florins escape altogether. Those a little higher pay two or four florins, according as the excess is 1,000 or 2,000 florins. Higher properties pay one-eighth of 1 per cent., the first 10,000 florins being exempt. Possessions beyond 200,000 florins pay one-fifth of 1 per cent. on the excess.
*84 Both the taxes just considered are very slightly progressive, or rather degressive in character, and the Dutch, which, it must be remembered, is intended for both income and property, is the milder. The estimated produce for these taxes for 1901-2 is somewhat under £3,000,000.

§ 5. The failure of property taxes in so many separate cases, and the clearer comprehension of income as the true normal source of taxation, have made the plan of a general tax on revenue or income appear advisable. We have noticed the imperfections and dangers of the single income tax: it is now rather as one of the constituents of a general system of taxation that we have to estimate it. In this aspect we find that the income tax is a distinctly modern product, and one that is likely to grow in importance. A well-balanced financial system will derive a large part of its receipts from direct taxation, as otherwise an approach to just distribution would hardly be possible. Amongst the objects of these direct charges the produce of land, capital, and labour must take their place, and when they have each come under contribution the elements of the income tax are present. The close analogy between the four direct contributions in France and the five schedules of the English income tax is evident, and this resemblance extends to the German ‘produce taxes.’ There is, however one very important difference; the taxes on the several elements of wealth are far less elastic in yield. Thus the French, German, and Italian land taxes have a fixity that is not found in the income tax; and the other produce taxes, though possessing more expansive power, are not yet at all as effective as is desirable: the
Patente expands more slowly than Schedule D. There are besides various gaps in the most developed of the Continental ‘produce taxes.’ State creditors in France escape taxation, while the English and Italian fundholders pay on that part of their revenue. Mortgages and other forms of loanable capital also manage to avoid their proper share, which would be impossible with the income tax. But the actual institution of a tax on income is not due to refined considerations of justice: like most imposts, the income tax is the child of necessity. When other contributions have been carried to their productive limit the financier has perforce to fall back on the direct taxation of income. This method is the more necessary in a country where taxation of the several parts of income is absent or inadequate. Both conditions were combined in the case of the first English income tax (1798),
*85 and were also present in a great degree in Italy in 1864.

The result of this originating cause is seen in the use of the income tax as a complementary receipt, to be employed in cases of pressure and to meet what would otherwise be a temporary deficit. The aim of keeping a correct balance of expenditure and receipts can be best realised by having a varying income tax adjusted to suit the special circumstances of each Budget. Thus in England the rate has varied from 1
s. 4
d. (if we include the earlier income tax from 2
s.) to 2
d. per pound. Italy has been unable to follow the same course, as the highest rate is in her case requisite in order to procure funds, but the desirability of having a movable tax of the kind is indisputable.

Another advantage of the tax on income is the opportunity that it offers for fairly distributing the burden of taxation. Indirect taxation, and particularly that on consumption, falls with greatest weight on the smaller incomes, and lets the rich escape too easily. An income tax with a suitable scale of exemption goes far to correct this inequality, which duties on acts and inheritances also aid in remedying. Both on financial and equitable grounds there is a strong case for the use of the income tax, not as the sole source of compulsory revenue, but in due proportion with other receipts, and with close attention to the special circumstances of the country.

§ 6. The development of the English income tax throws light on many of the problems connected with its general use. Its history is divided into two periods, (1) that of the war income tax (1798-1816), and (2) that of the peace tax since 1842.
*86 The former, preceded by ‘the triple assessment,’ consisted at first of a tax on the sum of income to be ascertained by the taxpayer’s declaration. A lengthy form of return was required, and a number of deductions were allowed, for repairs, support of children, insurance premiums, &c. The yield was about £6,000,000,
*87 at the rate of 10 per cent. on the national income, estimated by Pitt at £102,000,000. Repealed at the Peace of Amiens in 1802, it was reimposed in 1803, with the important change of substituting ‘particular returns of particular sources of income’ for the previous general return. Thus arose the well-known five schedules, and inquiry as to the total amount of income was avoided.
*88 The rate was 1
s. per pound; incomes under £60 were exempt, and those under £150 taxed at a lower rate. The yield for the first year was over £5,000,000. In 1806 the rate was raised to 2
s. in the pound, and several changes in the regulations were introduced. The exemption limit was lowered to £50, and the allowance for children withdrawn, also that for repairs, in Schedule A. The method of stoppage at the Bank was applied to Schedule C. With the high rate of charge the yield was at first £12,000,000; in 1815 it had risen to £15,642,000. On the conclusion of peace the Government desired to continue the tax at half the existing rate, but they were defeated and had to abandon it.

The difficulties of English finance during the succeeding quarter of a century were largely due to this mistaken step. The retention of the income tax would have allowed reforms in other branches to have been carried out with comparative ease. Accordingly competent opinion as expressed by Sir H. Parnell and Sayer advised its reintroduction.
*89 This prudent counsel was adopted by Peel in 1842. His measure—really the old system with unimportant modifications—was enacted for only three years, and the rate was fixed at 7
d. per pound (or under 3 per cent.). The yield in the first year was over £5,600,000, the same as at the rate of 10 per cent. in 1801. At its expiry there was an extension to 1848, and again to 1851. It was voted for one year in 1852; in 1853 it was extended to Ireland, and fixed for seven years by Mr. Gladstone, who held out the prospect of ‘its relinquishment’ at the end of that term. The Crimean war, during which the rate rose to 1
s. 4
d. per pound, prevented this result, and since 1860 it has been continued as an annual tax at rates varying from 10
d. to 2
d. until 1901 when it was advanced to 1
s., and then in the two following years raised to 1
s. 2
d. and 1
s. 3
d. respectively for war purposes. It is now a permanent and, indeed, indispensable part of our financial system.
*90

In studying the English income tax the first noticeable point is its composite character. It is, in Mr. Gladstone’s words, ‘rather a code or system of taxation’
*91 than a single tax. The five schedules may well be regarded as so many distinct taxes, since they deal with separate kinds of revenue. The connexion between them comes out only in cases of exemption or abatement. Inequalities are, however, removed by the comprehensiveness of the tax. Mortgages pay under schedule A by deduction; but there is no inducement to capitalists to put their wealth into the forms included by B, C, or D, as there also they will have to pay on their receipts.

Another important part of the system is the extensive use of stoppage at the source. The result is that a large body of taxpayers never receive the sums due by them to the State. The public funds, dividends, mortgages are all so treated, and evasion and fraud are thereby reduced to a minimum. The separate sources of income are tapped, and supervision is made much easier.
*92

Thirdly, we may bear in mind the very large yield of the tax. At its commencement it contributed £6,000,000 in a time of great pressure, while its latest service has been as a mainstay of the national finances during the recent war. Its contribution at the rate of 14
d. per pound for 1901-2 amounted to £34,800,000. The yield derived from the penny per pound is in fact an indication of national progress; from £700,000 in 1842 it has swelled to £2,500,000 in 1901-2.
*93

This productiveness accounts for its great services both in war and peace. It supplied the means for carrying on the struggle against Napoleon, and it rendered possible the reforms of Peel and Gladstone, besides saving the country from deficits.
*94

The extension of exemptions and abatements is a further interesting point. The original limit of exemption (£60) was soon reduced to £50, in order to include the large class who returned their incomes at £59 10s., and £200 was the point at which the full charge was enforced; this also came down to £150 in 1803. Under Peel’s measure the exemption limit was placed at £150, and, though subsequently lowered to £100, it was again raised to the higher figure. In 1863 a deduction of £60 was allowed from all incomes under £200; in 1873 this was increased to £80 from incomes under £300; and in 1876 to £120 from incomes under £400. In 1894 the exemption was raised to include incomes of £160; a deduction of £160 was allowed from all incomes between £160 and £400, and an entirely new abatement on £100 for incomes between £400 and £500 was introduced. In 1898 this abatement was increased to £150 while a deduction of £120 was allowed on incomes between £500 and £600 and one of £70 on those under £700.

§ 7. The Italian tax on ‘movable wealth’ has strong points of resemblance to the English income tax, and has been much affected by its example. It commenced in 1864, when a sum of 30,000,000 lire (£1,200,000) was apportioned among the several provinces, and raised by a tax on revenue (that from land excepted). In 1865 the amount was more than doubled (66,000,000 lire), and in 1866 the tax was changed from an ‘apportioned’ to a ‘rated’ one, and the rate fixed at 8 per cent. In 1870 it was advanced to 12 per cent., which, with the additional tenth, levied since 1868, made the total 13.20 per cent., in 1894 it was raised to 20 per cent. Many changes have been made in the methods of levy and assessment. The original law of 1864 has been frequently amended; a new and comprehensive measure was passed dealing with the whole subject in 1877, and a further revision of the classes was adopted in 1894. After the English pattern, the contributors are grouped under several schedules, but the arrangement is different, and used for a different purpose. Class A comprises two divisions, (
a) revenue from interest on railways and local government loans assessed at its full value, (
b) other permanent revenue paying only on three-fourths. Class B contains what are called ‘mixed’ revenues, or those in the production of which capital and labour co-operate: these escape with payment on one half of their amount. Class C contains revenue from labour, assessed at nine-twentieths of its total. The incomes of public officials are placed in Class D, and pay only on three-eighths of their amount. To these four classes should be added the Metayers as forming a fifth, paying 5½ per cent. of the land tax. A complicated scale of allowances for small incomes is also part of the system. Incomes under 400 lire in classes B, C, D are exempt, and up to 800 lire the taxable sum is reduced. The declaration of the contributor is the basis of charge, but is tested by inquiry, and, as far as possible, the tax is collected by stoppage.

Notwithstanding the very elaborate provisions of the law, it is found impossible to reach a great deal of the national revenue. Incomes in classes B and C are very generally returned at much below their true amount. Like the property tax in the United States, the Italian income tax is ineffective through evasion. Thus, though the method of stoppage is only applied to a limited set of cases, its receipts are nearly as large as those from direct collection. Again, the proportion paid by companies is about 40 per cent. of the whole, a ratio quite inconsistent with all other available statistics. An analysis of the actual returns of revenue leads to the same conclusion. In 1874 639,302 persons made returns, and out of this number only 986, or 1 out of 640, admitted incomes of more than £1,000
per annum. Even though the wealth of Italy is much less than that of England or France, these figures cannot be accepted as a true representation; they simply prove the existence of fraud on a large scale.

One cause of such widespread evasion is the very high rate of taxation. Twenty per cent. is equivalent to 4
s. 5
d. in the pound,
*95 and so high an income tax would even in England lead to much dissimulation of income. The most obvious remedy is a diminution of the tax rate, combined with greater powers of assessment, more particularly in respect to professional and industrial incomes. The differentiation of the classes of income, which originated in an attempt to apply certain theories very popular at the time in England,
*96 also tends to make the returns inaccurate and to embarrass the officials.

These drawbacks notwithstanding, Italian Finance has found a powerful resource in this form of taxation. The original £1,200,000 of 1865 had increased to £7,000,000 in 1875, to over £8,000,000 in 1885, to over £9,000,000 in 1890, to nearly £9,500,000 in 1894, and to £11,500,000 in 1899. It must besides be remembered that owners of land are excluded from the operation of the tax, since they come under the land tax discussed in a preceding chapter.
*97

The latest addition to the group of income taxes is that introduced into Spain by the law of March 1900. It is framed on the Italian type and comprises three categories (
a) incomes from labour, (
b) incomes from capital, (
c) mixed incomes. In the first, the rates vary from 5 per cent. to 20 per cent., salaries under £60 being exempt. In the second group the variation is from 3 per cent. to 20 per cent., the last applying to certain classes of the public debt. The third class is charged at rates from 2 per cent. to 15 per cent. No additions for local finance are permitted. The estimated yield for 1902 is £4,200,000.
*98

§ 8. The German income taxes are best represented by those of Prussia and Saxony. The Prussian
Einkommensteuer was introduced in 1851 as a development of the older class tax. It was only applicable to incomes over £150, and dealt with them by groups. A sum was fixed for each group amounting to 3 per cent. on the lowest incomes in that group: thus,
e.g., incomes between £600 and £720 paid £18, those between £12,000 and £15,000 paid £360, and all incomes over £36,000 paid £1,080, the highest sum due. Along with the reform of the class tax in 1873 the income tax was settled on a somewhat different scale, but with a general rate of about 3 per cent. The objections to this system as unequal have been so strong as to lead to the reform of 1891, by which the class tax
*99 is absorbed in the income tax, and all incomes under £45 exempted. From that point the rate rises by degrees; between £525 and £1,475 it is 3 per cent.; on incomes over £5,000, 4 per cent. The idea of progression is thus realised, though in a very limited way. The method of assessment has also been changed. It was previously settled by official valuation, based on the materials possessed by the administration, but is henceforth to depend on the declaration of the taxpayer.
*100

It is evident that the Prussian income tax differs in some important respects from those in England and Italy The function of supplementing the other branches of receipts is entirely absent, as the rate is fixed, not movable from year to year. The produce taxes are not brought under the income tax, but are continued quite separately: the taxes on land and industry present, accordingly, apparent cases of double taxation. Finally, the productiveness is much less. In 1864 the amount received was over £500,000, in 1876 it was nearly £1,500,000, in 1884 over £1,750,000; for 1889-90 the estimate was over £2,000,000. Under the new system (with the class tax included) the estimate for 1892-3 was £4,000,000, but the amount obtained was £6,240,000. It fell off slightly in the two following years, but has risen steadily since, and for 1901-2 exceeded £9,300,000.
*101

Saxony set the pattern to Germany of a classified and progressive income tax. Introduced in 1874, it was developed by the laws of 1878 and 1894, the last modification taking place in 1900. Incomes under £20 are free Those between £440 and £2,000 pay 3 per cent. Incomes of £4,000 pay 4 per cent. The yield of the tax in 1880 was over £600,000; in 1890 it rose to £1,030,000, in 1900 it amounted to £1,760,000. Austria adopted a so-called income tax in 1849 which was really a tax, partly on industry, partly on salaries. The ineffectiveness of this system led to the law of 1896, which introduced, besides the industry taxes already mentioned,
*102 taxes on (
a) interest, (
b) personal revenue, (
c) salaries of high officials. In 1900 these taxes brought in over £2,200,000, five-sixths of which was due to the taxes on personal incomes.
*103

To the foregoing may be added the taxes of the Swiss cantons. Their property taxes already mentioned are supplemented by income taxes, in many cases on a progressive scale. No two cantons have adopted exactly the same system in all details, but there are, as might be expected, general points of resemblance. The Zürich income tax follows the pattern of the property tax. The smaller incomes are taxed on a part only of their amount, and at each higher stage the excess over the preceding one is placed under greater pressure until the point of full liability is reached. Graubünden follows its property tax by grouping incomes in classes, and by raising the percentage rate as they get higher. Switzerland is in fact the classical country of progressive income taxes, though the moderation of the rates, and still more of their application, weakens the conclusions that might otherwise be drawn.

§ 9. The income tax was first introduced into the United States during the trying period of the Civil War. At its commencement the rates were moderately progressive—3 per cent. or 5 per cent. according to amount of income, but were soon raised to the higher points of 5 per cent. and 10 per cent., until the close of hostilities allowed of a return to a uniform rate of 5 per cent. in 1867, and reduction to half that amount in 1871, with finally complete abandonment of the tax in 1873.
*104 The highest yield was in 1866, in which year it brought in $73,000,000.

The severe depression in 1893 so affected the United States revenue as to cause a serious deficit for the year 1893-4.
*105 This, coupled with the measure for tariff reform, led to the passage of an income tax fixed at 2 per cent. on incomes over £800 ($4,000), those below that level being exempt. Property acquired by gift or inheritance was to be treated as income. The Supreme Court by a majority of one declared the tax to be ‘unconstitutional,’ inasmuch as, being ‘direct,’ it was not apportioned in accordance with the provision of the Constitution. Though this decision, considering the meaning attached to the term ‘direct’ in the 18th century, is doubtful, it will prevent the employment of the income tax until a counter decision is given by the Supreme Court, or the unlikely expedient of a constitutional amendment is adopted.
*106

Since the establishment of the third Republic several attempts have been made to introduce the income tax into France, but hitherto without success. One reason for this failure is found in the character of most of the proposals, which aim at a progressive system, and are consequently obnoxious to the conservative sentiment of the country. The system of personal declaration which progression necessitates, but which is regarded as inquisitorial, is another reason for the failure. Of late years, however, the feeling in favour of an income tax seems to be increasing, and successive finance ministers have sought to satisfy it. M. Doumer in 1886, M. Peytral in 1898, M. Caillaux in 1900 have each devised a measure, and in 1902 M. Rouvier seems about to follow their example. All such measures must provide for the abolition of the
Personelle mobilière and the door and window taxes, but logic would seem to require in addition the inclusion of the
Impôt foncier and also the
Patente. Were the idea of progression definitely abandoned, it is not unlikely that the income tax would prove a good substitute for (or perhaps more truly a development of) the four direct taxes, to which the tax on movable values might be appended. Still the question is one of great difficulty and complexity.
*107

§ 10. We have deferred a discussion of some fundamental questions relating to the constitution of an income tax until the leading facts of its use were known. Most of them have been already encountered in connexion with the general principles of taxation, but they take a different shape when the income tax is treated as but one part of a larger system, and need to be handled specially with a view to that fact.

One important question is that of progression in the rate of charge. The general conclusion that we reached
*108 as to the inexpediency of any progressive system has to be reconsidered when the income tax is used as a complementary resource. Progression in the case of such a tax may be necessary for true proportional taxation. If the smaller incomes are unduly weighted by taxes on consumption, their exemption, or milder treatment, under the income tax appears so far justified. A variation in the rate of charge is not open to the objection of arbitrariness, as it is determined by reference to the amount of other taxes. The other objections are not so readily refuted. Risk of evasion and unproductiveness may both be urged against the graduation of even a moderate complementary tax. Where the area is a large one, the effect on accumulation and investment will not be serious, as the distribution of taxation will,
ex hypothesi, be equal, but the existing attempts at progression are, it may be said, hardly worth the trouble they involve. The English method of exemption and abatement has great advantages from the technical point of view, since it allows the sources of income to be taxed without reference to their amount.
*109 The treatment of each person’s income as a whole compels recourse to returns of a complicated kind, is disliked as being inquisitorial, and gives opportunities for escape to large masses of income. For these reasons the proportional rate is, on the whole, advisable.

The answer just given helps us in deciding as to the adoption of different rates on different kinds of income. The proposal that life incomes and those derived from industry should be charged at a lower rate has received influential support, and is illustrated by Italian practice. When considering the distribution of taxation we noticed the general arguments as regards the income tax in England. It was for so long looked on as a temporary charge, that the idea of capitalising incomes subject to it gained a good deal of support. The defence of the strict proportional rate rested on two admitted facts: viz. (1) that no ingenuity could avoid some injustice, and (2) that any alterations would mean the destruction of the tax.
*110 Its gradual passage into a permanent charge has greatly strengthened its position in this respect, though the cry to remedy supposed grievances in its distribution may gain greater support.
*111

The working of the Italian tax does not support the system of different rates. The lower scales for profits and salaries are confusing, and account for much of the loss through concealment of incomes. The single general rate would prove advantageous from a fiscal point of view, and with stricter assessment could be effectually carried out. The attempt to group incomes into ‘permanent,’ ‘mixed,’ and ‘temporary’ is, moreover, too rough to give satisfaction or to realise justice.

Allowance for necessary expenditure and repairs is one of the practical difficulties in the administration of the income tax. On principle, as the tax is one on income, not on gross produce, deduction of the expenses of production of the income taxed should be allowed. Where much fixed capital is employed, this is very hard to determine, and we can understand the preference of French administrators for the self-acting rules of the
Patente. In respect to land the English system till recently failed to recognise the cost of repairs; it, therefore, treated this class of revenue with unusual harshness, and gave some support to the view that Schedule A should be regarded as a distinct land tax. The Finance Act, 1894, has partly remedied this grievance, as it provides for an allowance for expenses and repairs.
*112 The exceptional treatment of farmers’ profits is another fact pointing in the same direction; but it may be hoped that this anomaly will be gradually removed.

The exemption of savings has been already discussed, but one method—that of life insurance—appears to be a case of capitalisation; it is in fact turning a life income into a smaller permanent one, just as the purchase of an annuity is the opposite process. Up to a certain limit—one-sixth of the income—the English system allows exemption of insurance premiums, a privilege not extended to savings in general. Though the desirability of encouraging providence may be granted, it would seem that an exemption from duty on transfer after death would be a more fitting mode of bestowing the favour. It cannot be contended that an insurance premium is not a part of income and the principal created by its use will not contribute to the payer’s income in the future. The case is, as Mill puts it,
*113 one of concession to ‘human feeling,’ rather than a sound deduction from general principles.

The problem of assessment is another of the difficulties to be faced: between the Scylla of the contributor’s evasion under self-declaration and the Charybdis of official inquisition it is hard to take an intermediate course; but the dealing with each separate part of income, the combination of declaration and official control, and above all the use wherever possible of taxation at the source of revenue, so characteristic of the English method, are undoubtedly the best safeguards against abuse. The direct contact of the citizen and the tax-collector is the most delicate part of the fiscal machine, needing care and use of the results of experience to prevent friction. Allowing for the inevitable margin of error, the results of the English income tax are eminently satisfactory.

§ 11. Any notice of the question of incidence may seem unnecessary in respect to a tax which falls on all the constituents of revenue. On whom can income receivers in general shift their burdens? Some of the suggested objects are certainly not available. Thus the vulgar idea alluded to by Mill, that the income tax falls on the poor by checking the expenditure of the rich, has no foundation in fact. Nor is there much force in the contention that in so far as the tax is paid out of capital it falls on the labourers,
*114 as this is no peculiar quality of the income tax, but one common to all taxation. The State must obtain revenue, and unless the income tax were specially obstructive to saving, it would produce no peculiar effect. Looking at the subject in a rather different way we obtain a better result. The income tax is composed of taxes on rent, interests, profits, and the higher forms of wages; therefore it may be said that the incidence of these several parts of the tax will, taken together, give the incidence of the whole. This, however, brings us back in a large degree to its non-transferability; for taxes on rent, on the higher kinds of wages, or on employers’ gains, are not easily shifted. Even in the case of interest, unless the growth of capital is checked, a tax tends to remain on the payer. Therefore, speaking broadly, we may say that the shifting of an income tax is not to be expected, and in the rare cases where it does happen is brought about, either by a check in the growth of capital through diminished interest, or by disturbances in the relations of the several industries and trades through its action. A progressive income tax will of course have a stronger tendency to cause the former effect. But though this reasoning is true in respect to an income tax imposed with scientific accuracy on the various components of income it needs to be qualified in considering the actually existent taxes. Thus we should say that the mild treatment of farmers’ profits in England tends to disturb the distribution of capital and affects rent. The incidence of an imperfect income tax can only be traced by analysing the tax into its elements, and examining the course of shifting in respect to each