Other definitions of taxation fail through excessive
vagueness. We gain little by being told that taxation
is 'a public charge, a duty imposed on certain things' 1 . Very,
often one or more of the essential elements is omitted.
Thus the fact of taxation falling solely on persons is
neglected in the definition of taxes as ' the enforced pro-
portional contribution of persons or property levied by the
authority of the State for the support of government and
for all public needs ' 2 . Besides the error of including
' property ' as a subject of taxation, this definition brings
in the unessential principle of ' proportionality/ and would
therefore, exclude large groups of what are universally
regarded as taxes. This is a very common defect in the
definition of the term due to the desire to give an exhaus-
tive account of its attributes, or to bring some favourite
theory into its general conception. Professor Ely's elaborate
account, like those of many German writers, illustrates this
danger 3 . The real function of a definition is to give a clear
idea of the nature and limits of the phenomenon denoted
by the term, not to convey in a formal statement all that
is known about it, still less to prejudge the questions that
may arise in the course of further inquiry.

6. The etymologies of the words employed in different
languages to denote this class of public contributions are
full of instruction. The English ' tax/ as also its equiva-
lent in local Finance 'rate,' suggests the estimation or fixing
of the amount of charge. So does the German * Schdtzung?
The assistance or advantage to the State is foremost in the
French ' aide ' and the German 'Steuer.' The idea of com-
pulsion is primary in * impot' and 'Atiflage.' The surrender
by the payer is connoted in * tributuni l dazio' and * Abgabel
while finally the root of taxation in voluntary payment
is evidenced by the words * donum ' and ' benevolence?
Minute investigation may show that there are differences in
the charges described by these several names, but speaking
broadly they all cover what we regard as taxation, and
help to justify the definition given above T .

7. Having determined the meaning of * taxation,' it is
necessary to understand its chief classifications and the
technical terms employed respecting it. First, we may
notice the term ' subject, 1 which is conveniently used to
denote the person who bears its burden, and who must be
distinguished from the immediate payer, e. g. the importer
of wine in England pays the duty on it, but the ' subjects '
of the wine duties are the consumers so far as the charge is
really a pressure on them. The ' subject ' and the payer
may or may not be the same according to the particular
circumstances.

As the 'subject' of taxation is the person affected, so
the 'object' is the thing or fact on which it is imposed,
thus in the example just given of the wine duties, the com-
modity wine would be the object of the duty. Even where
taxation is said to be ' personal * it is assessed on some
object as ' income ' or * produce,' or in the extreme instance
of a capitation or poll-tax on the person as a physical body.
Confusion between the ' subject ' and ' object ' is the cause of 1
the belief that some taxation does not fall on persons 2 .

The ' source* of taxation has somewhat the same relation
to its 'object' as the ultimate bearer or subject to the
immediate payer. The fund created by taxation is. derived
from the resources of the community, i. e. as we shall see
from the income, or in special instances the property, of the
' subjects.' There has been much dispute as to the real
' source ' of the tax-revenue that will need consideration
later on, but there is no doubt as to the proper use of the
term ' source ' in respect to taxation. It is, perhaps, unne-
cessary to mention the terms ' unit ' and ' rate ' which are
employed, the former to describe the quantity of the objects
taken as a standard, the latter the amount of taxation per
' unit.' Where commodities are taxed the unit will be a
measure of weight, e. g. the Ib, as in the British tea duty,
or contents as the gallon in the wine duty, or length as in
the old duties on cottons. A sum of the standard money
is the commonest, as in the system of ad valorem duties *.

8. A much more important set of terms is that con-
nected with the classification of taxation. The division
and grouping of the several kinds of taxes have been varied
to suit particular financial systems, and much of the general
discussions on the subject is concerned with the compara-
tive merits of these arrangements, and the extent to which
they conform to the natural order so far as it can be said to
exist. A preliminary notice of some of the more common
distinctions is desirable at the present stage.

One of the most widely known and frequently used
divisions of taxation is that into ' direct ' and ' indirect ' ;
unfortunately it is used in different senses, thougH with
several points of connexion. That most familiar to English
readers is stated by J. S. Mill in the following terms :

'Taxes are either direct or indirect. A direct tax is onev
which is demanded from the very persons who it is intended
or desired should pay it. Indirect taxes are those which 4
are demanded from one person in the expectation and in-
tention that he shall indemnify himself at the expense of
another V

The difference is here made to turn on the mode of
incidence, a matter often very difficult to determine, and
changing with the special circumstances of each case.
Whatever be its economical importance, it is evidently
useless for administrative purpose's, and probably owes its
origin to the peculiar theory of the Physiocrats respecting
the ' source ' of taxation.

A natural result has been that practical financiers have
adopted a different: basis of distinction, and regard those
taxes as direct which are levied on permanent and recurr-
ing occasions ; while charges on occasional and particular
events are placed under the category of indirect taxation.
On either method the income tax would be ' direct,' and
the excise and customs ' indirect ' : the ' death duties ' would
be ' direct ' from Mill's point of view, and ' indirect ' in the
administrative sense *. (

Another division is that into 'taxes on revenue' and
{ taxes on capital,' or, perhaps better, on ' property.' The
former are paid out of the annual national production ; the
latter encroach on the accumulated wealth of the society.
But in qualification of this statement it must be added that
most of the actual property or capital taxes are so only in
name, being really paid out of the income of the persons
subject to the charge. There is thus a discordance between
the practical and scientific use of these terms as great as in
the case of direct and indirect taxation.

Taxes are often said to be either ' real ' or ' personal,'
and attempts have been made to distribute them into two
classes on this basis. Personal taxes are those in which
the person is taken note of in assessment. They require
lists of the tax-payers (roles nominatives in the language of
French administrators). Real taxes are assessed on objects
other than persons, and without direct reference to the
owners or possessors. Capitation and income taxes are
' personal '; taxes on land, houses, or goods, are ' real.' The
use of these terms has the inconvenience, already noticed,
of obscuring the fact that all taxation is in the last resort
on persons, and further raises a particular form of levy into
undue importance. An income tax is certainly personal,
but Schedule A of the English income tax is very similar
to the French impot fonder that is as certainly ' real.'

In respect to the mode of assessment taxes may be
either '.rated ' or ' apportioned V In the former class the
charge per unit is fixed, but the total yield is always un-
certain, depending as it does on the number of units that
pay. An apportioned tax is one the total amount of which is
fixed, the shares being apportioned among the objects that
are charged. As examples the English income tax and
the French impot fonder will again serve. The former is
' rated,' the latter ' apportioned/ being so divided among
the departments as to make up the previously fixed amount.
This method is decidedly the more primitive : it has dis-
appeared long ago from the English system, and will
probably meet the same fate elsewhere 2 .

9. The foregoing distinctions are too important to be
passed over, but they are also too imperfect to be of much
use in a scientific classification of taxes. Particular aspects
of taxation, the administrative peculiarities of certain coun-
tries, and obsolete or imperfect theories have been the
causes of their employment. It is accordingly advisable to
consider the subject from a more general point of view in
order, as far as possible, to reach a natural arrangement.

In choosing the principle of grouping we have to make a
selection between two contrasted systems which may be
distinguished as (i) the economical and theoretical, and (2)
the empirical or fiscal modes.

The first mentioned depends on the economical theory of
the distribution of wealth, and can be traced back at least
to Adam Smith. He opens his discussion of taxation by
asserting that 'the private revenue of individuals arises
ultimately from three different sources- rent, profit, and
wages, and proceeds, ' every tax must finally be paid from one
or other of those different sources of revenue, or from all of
them indifferently. . . . The particular consideration
of each of these different sorts will divide the second part
of the present chapter into four articles V Nothing can be
plainer and simpler in appearance than this arrangement.
The economic shares in distribution are regarded as so
many sources of revenue on one or more of which every
tax must fall. The later analysis of profit into the com-
ponent parts of * interest ' and ' employer's gain ' would add
one further source, but would not otherwise disturb the
treatment 2 . The great attraction of this method is its
simplicity and the facilities that it affords for employing
the propositions of Economics in deducing the effects of
taxation. To reduce the subject into ' four articles ' even
with * several other subdivisions,' promises a welcome abridg-
ment of labour. English economists in treating of taxation
have, therefore, intended, as far as possible, to follow th : s
course. Ricardo and J. S. Mill are the most prominent
examples. But on closer examination it appears that
neither of them, nor even Adam Smith himself, could
adhere consistently to this over-simple grouping. In
Ricardo's hands the subject requires eleven chapters,
several of which consider the effects of taxes on land,
houses, raw produce, and gold, in addition to those on the
primary sources of rent, profit and wages. Mill goes fur-
ther, and formally limits the division of taxes according to
the economic source on which they are imposed to direct
taxation on income 3 . The taxation of commodities and
such taxes as those on contracts and on communication
are quite outside it. But the Wealth of Nations affords a
stronger proof of the insufficiency of the ground of division
selected by its author. Sections devoted to taxes on pro-
duce of land, on the profit rent, and the ground rent of
houses, to capitation taxes, and taxes on commodities, break
up the compact order that the introduction holds out. It
is evident that the subject-matter refused to fit into the
limited groups that the economic classification required,
and the sound common sense so characteristic of Adam
Smith is shown by his deviations from the theoretic lines
previously traced out by him.

The difficulty arises from the fact that taxation always
has persons for its ' subjects,' and they frequently derive
their income the normal i source' of taxation from more
than one of the different economic shares. The citizen is
not a pure rent, interest, or wages receiver, he often com-
bines all three in his annual receipts. Again, the most
prominent external feature of taxation is the * object ' on
which it is levied. These are, however, very many, and it
is often beyond the power of analysis to decompose the
charge on some commodity or form of receipt into its
economic constituents, e. g. the produce of land may be
due to the co-operation of natural agents, capital, labour,
and directing ability, but to say how much of the taxation
imposed on the result is due to each factor is quite im-
possible.

The obvious conclusion is that the classification is
unsuitable. It is often convenient to use the economic
theorems respecting rent, wages, &c., in our investigations
of the effects of taxation, even though we should never
meet in fact with the pure taxes on those parts of the
product. For the problems of Finance it is also necessary
to remember that these preliminary inquiries are but steps
towards the final result which must deal with realities,
and not with imaginary and hypothetical cases.

10. The defects of the economical mode of classifica-
tion lead us to turn to what we have entiilecLthe 'empirical '
or * fiscal' one, which takes the actual kinds of taxation,
and arranges them in the most convenient way. To this
procedure it may at once be objected that as each country
has its own tax system, varying from time to time, we
cannot attain to a general arrangement valid in all cases.
The classification of taxes suited for ancient Rome would
be inadequate in modern England, and even confining
attention to the present day, the Indian and British tax
systems cannot be easily reduced to the same classification.
This effect of temporary circumstances in limiting general
principles has been already noticed J , and it does at first
sight raise difficulties in the effort to prepare a natural
grouping of taxes. The mode of escapes, however, obvious
on a little consideration. The names and minute details of
taxation vary greatly at different times and places, but
this does not preclude the existence of large categories of
taxation possible in all countries and found in somewhat
different forms in many. The Indian land revenues differ
from the English land tax, and also from the French
impot fonder, but in all three countries there is 'taxation
of land,' which offers a general title, under which they may
be placed in company with the Roman provincial tax and
several others. A like mode can be applied to different
forms of taxes on the produce of industry, and so in other
cases.

The question next arises, How far should this process be
carried and what general categories can we form ? Ran
has boldly grouped all taxes under the two heads of
' estimated taxes' usually charged on goods {Schatzungen)
and ; taxes on expenditure' {Aufwandsteueni) which does
not carry us much beyond the rude divisions mentioned in
8, Hoffmann prefers the division into taxes on posses-
sion (Besitz] and taxes on acts (Handlungen], while Cohn
accepts the tripartite arrangement of Wagner, into taxes on
(a) acquisition (Erwerti), (b) possession (Besitz) and (c] con-
sumption ( Verbraucti) 1 . De Parieu carries out the division
more minutely, and forms five classes of taxes, viz. (i) on
persons, (2) on wealth, (3) on enjoyment, '4) on consump-
tion, (5) on acts. In defence of this arrangement he argues
that, like all natural classifications, it allows of an indefinite
margin between each adjacent group, and that it further
harmonizes with the administrative division between direct
and indirect taxation, classes I, 2 and 3 belonging to the
former and classes 4 and 5 to the latter category 2 .

All the preceding classifications appear to have at least
two defects : for (i) they simply deal with certain external
features of taxes, and do not take note of their essential
characteristics, and (2) like the otherwise very different
arrangement of Adam Smith, they are too simple for the
complexity of the facts to which they are applied. Hock
has attempted to avoid this defect. He starts from the
untenable position that taxation is a compensation for
state services. These services are, he thinks, of three
kinds, to wit; (i) protection of person, (2) of property,
and (3) the performance of special services. To each cor-
responds a 'primitive tax' (Ur 's teuer) : these are (i) per-
sonal taxes, (2) income taxes, (3) taxes for special services
rendered 3 . The practical difficulties in levying these taxes
in their pure form leads to the use of other taxes as sub-
stitutes {Surrogate) in the form of taxes on (a) consumption,
(b] product, (c) customs, (d) special income taxes, (e) fees
and charges on occupations 4 .
Though it is plain that the ground idea of Hock's
division is unsound it yet has the merit of suggesting the
best way of reaching a truly natural arrangement. The
distinction between primitive and derived taxes is a
valuable one, and can be used to combine the economical
and empirical methods of grouping in our arrangement.

11. The position of Adam Smith that taxation must
be derived from the constituents of private income is un-
doubtedly sound. Where it falls on property there is a
diminution of the national wealth which, if continued, must
prove destructive. A true instinct therefore prompted him
in his effort to analyze taxes into those on rent, on wages,
and on profit. On the other hand it is equally true that
the ' objects ' of taxation do not easily allow of this
analysis. Between the taxes of economical theory and
the taxes of actual Finance there is a gulf that appears
hard to bridge over, and one that has retarded the progress
of financial science.

This difficulty is at all events extenuated by the circum-
stance that though the abstract economic taxes are not met
with in fact, they are not wholly imaginary. The economic
tax on rent has some, and often considerable resemblance
to a land tax, or to put it the other way, a land tax often
tends to become a tax on rent. The e tax on profit ' of the
economic text books bears a like relation to the taxes on
business, of which Schedule D of the English income tax,
the Prussian Gewerbesteuer and the French patentes may be
taken as specimens. So with the wages tax, in relation to
actual capitation taxes, or the late Classensteuer of Prussia.
If now we regard taxes on the factors of production, and there-
fore on the shares in distribution as ' primary,' we have a
basis from which to proceed to the investigation of those
secondary taxes that are placed on other ' objects.' By
grouping together the various taxes on land we can con-
sider the play of financial forces in the case of rent. The
industrial taxes will similarly enable us to see the working
of charges on interest and profit, and finally poll and capita-
tion taxes will perform the same service for taxes on wages.
The economic mode of arrangement assigns a place to
taxes on income or revenue which we regard as a combina-
tion of all the primary forms. It might in certain cases be
admissible to break up an income tax into its component
parts, just as it might, on the other hand, be well to com-
bine a series of taxes that together make up an income tax.
Thus the five schedules of the English income tax or the
four of the Italian one might be separately treated, or again
the 'four direct contributions' of the French system might
be taken in combination as nearly equivalent to a general
income tax 1 . Still it is necessary to consider the fiscal
bearings of general income and property taxes, and this
discussion most fitly follows the examination of the taxes
on component parts of income.

When the ' primary,' and if the phrase be admissible,
' quasi-primary ' taxes have been discussed, there remains
no small number of other charges. The whole elaborate /
system of taxation on commodities that has so large a]
place in every country must be dealt with. It may be
regarded as taxation of consumption or of expenditure,
but for practical purposes it includes the two great depart-
ments known to English Finance as * excise ' and ' customs/
So far the taxes enumerated have appeared .to fall on the
production, the distribution, or the consumption of wealth ;
those that directly affect the remaining economic process
of circulation must also be noticed. Taxes on transport ^
and communications come under this head ; so does the yet r ,
more important class of taxes on the transfer of property
and the transactions of commerce, i. e. the ' taxes on acts ' of
De Parieu's arrangement. The taxation of succession after
death may be treated as a particular case of transfer, but it
also has affinities with property and income taxes which
must be carefully considered. In like manner taxes on
necessary commodities often resemble in their effects a tax
on wages as Ricardo with some exaggeration urged. The
other secondary taxes have similar reactions on the consti-
tuents of income, but, nevertheless, their separate treatment
is desirable and indeed unavoidable.

12. We have now obtained what appears, on the whole,
a satisfactory distribution of the several taxes. Briefly
recapitulated it is as follows. The main division is into
' primary ' and ' secondary.' The primary taxes comprise
trifrse on land, on business and capital, on persons and on
labourers' earnings. The combination of these primary
forms gives us the general income and property taxes
which come next in order. Passing to the secondary forms
of taxation we find (i) taxes on commodities, including
both excises and customs, (2) taxes on communication and
transport, (3) taxes on transfer of property, (4) succession
duties, (5) the remaining taxes on commerce and legal
transactions.

But the discussion of the several taxes in the foregoing
order must be postponed until we have studied the opera-
tion of taxation in general and the conditions required for
its satisfactory working. No single tax can be rightly
appreciated without reference to the financial system of
which it forms a part. The chapters immediately suc-
ceeding will be devoted to a study of the characteristics
of taxation in general and the principles that should regu-
late its application. In this part of Finance we meet with
the most difficult theoretical and practical questions which
require the utmost attention for their proper understand-
ing. On some points opinion is sharply divided, and
consequently, while endeavouring to reach a definite judg-
ment on each disputed question, we shall aim at putting
the views of both sides in their strongest form before the
reader.