zyxwvutsrqpo zyxwvutsrq zyxwvutsrqp zyxwvutsr zyxwvut zyxwvutsrqpon Sconirh Journal of Poliricd Economy, Vol. 36. No. 3. Augusl 1989 0 19R9 Scottish Economic Society zyxwv zy T H E PHYSIOCRATIC T H E O R Y O F TAX INCIDENCE* LEFTERISTSOULFIDIS New School for Social Research The Physiocratic theories of social reproduction and economic policy have been extensively discussed in the economic literature (Ferrara 1958, Samuels 1961, Johnson 1966, Blaug 1978). Nonetheless, the Physiocratic approach to the question of taxation, which is intimately connected to their whole theoretical system, has received only cursory consideration. It is the purpose of the present paper to demonstrate in a rigorous way that the Physiocratic theory of taxation can be gainfully comprehended as a direct application of their theory of social reproduction described in the “Tableau Economique”. In reformulating the latter into an input-output model the effects of different tax scenarios on relative prices, rent rates and growth potentials are addressed. * Quesnay regarded the economy as an ensemble of interrelated sectors which, according to their contribution to the net product, identified as the “wealth” of a nation, were discerned as productive or unproductive. For Quesnay, the net product-whose exclusive form was rent-was defined as the difference between the total output and the total input utilized in agricultural p r o d ~ c t i o n .Consequently, ~ only the agricultural sector was conceived as productive, for “the origin, the principle of all expenditures and all wealth is the fertility of the land”.4 The other sectors, i.e., manufacturing, services, etc. were conceived as unproductive and totally subordinated to agriculture, their activities confined to changing the form of goods already produced in agriculture. Quesnay and the Physiocrats were particularly concerned with the conditions leading to the growth of the net product. This overriding concern was realized with the invention of the “Tableau Economique” which described the distribution of total output among the different social classes and sectors of an economy, and the process of social reproduction. The latter was seen as ’ I wish to thank Willi Semmler. Robert Heilbroner, Persefoni Tsaliki and an anonymous referee of the Journal for their valuable comments on previous versions of this article. Date of receipt of final manuscript: November 14 1988. Luigi Einaudi (1967) in his important paper “The Physiocratic Theory of Taxation” traces out the effects of different taxes in the writings of Physiocrats other than Quesnay. Although his conclusions are similar to our own, he does not attempt a rigorous formulation of the Physiocratic model. In what follows, we treat Quesnay’s writings as if they were representative of the ideas of all Physiocrats, although some of his disciplines held different views on many issues. For the Physiocrats, surplus assumes the exclusive form of rent, while profits are merely different parts into which rent is divided between the agricultural and industrial capitalists. “The ‘Dialogue on the Work of Artisans’ ” in Meek 1963. p. 209. ’ ’ 301 302 L . TSOULFIDIS zyxwvu zyx expanding, contracting or stationary, depending on the way in which total net product was allocated between productive and unproductive consumption, or what amounts to the same thing whether it was invested in agriculture or in industrial production. The more that was spent on productive consumption, the higher would be the scale of the economy’s reproduction; the more that was spent on unproductive consumption, the lower the scale. I THEPHYSIOCRATIC THEORY OF TAX INCIDENCE Based on this theory of social reproduction, Quesnay traced out the effects of different taxes on social classes, as well as on the growth potential of an economy discerning, for the first time in the history of economic thought, the difference between legal and economic incidence. He argued that regardless of where taxes were levied, they were ultimately obtained from the net product, i.e., rent income. “Thus taxes (Quesnay argued) should be laid directly on the net product of landed property; for in whatever manner they be assessed in a kingdom which draws its wealth from its territory, they are always paid by landed property”.’ Hence, the impact of taxation on the growth of an economy was ultimately related to the source that they were levied. Specifically, Quesnay argued that if the tax falls on the resources employed in agriculture, the expenses in this branch of production will be increased, the net product will diminish and the economy as a whole will be reproduced on a lower scale entailing lower rent incomes. The reason is that “the advances of the farmers are sufficient to enable the expenses of reproduction to reproduce at least 100 per cent; for if the advances are not sufficient, the expenses of cultivation are proportionately higher and yield less net product”.6 In similar fashion Quesnay noted that “it is not on the productive wealth of the husbandmen that taxes should be assessed, since this would mean the destruction of the means necessary to produce the annual wealth of the nation”.’ By contrast, if the tax were imposed on rent income in the first place only rent income would diminish, because the outlays in agriculture would not be affected and therefore the economy would continue to reproduce itself on the same scale. A tax on the industrial production or on any other unproductive sector would necessarily increase the expenses of their operation which, in turn, would feed back as an increment in the cost of the agricultural sector and result in the diminution of the net product. Hence, the tax will again be borne by the recipients of the net product, since the industrial sector is zyxwv zy 2. ’“Extract from the Royal Economic Maxims of M. De Sully’’ in Kuczynski and Meek, p. [bid., p. 6. ’“Extracts from ‘Taxation’ ” in Meek 1963, p. 106. zy zyxwvuts T H E P H Y S I O C R A T I C T H E O R Y OF T A X I N C I D E N C E 303 conceived as unproductive, reproducing itself without contributing to the net product .’ A tax on workers, who receive a wage just sufficient to perpetuate their mere existence, would result in higher wages, a burden which would necessarily fall on the employers.’ If the employers are landlords, they will necessarily bear the burden of this tax and the economy as a whole will continue on the same scale. If, however, the employers are industrialists, they will seek compensation either from the landlords, whose incomes will fall leaving the scale of reproduction unaffected, or from the farmers who directly or indirectly will decrease their investible product in order to meet their tax obligations. Consequently, the scale of reproduction in the economy will diminish. Quesnay noted that “an assessment of men rather than on the revenue would be borne by industrial and agricultural costs themselves, and it would be thrown back on the revenue of landed property in a way which involved a double loss, and would rapidly lead to the wiping out of tax receipts”. lo In short, Quesnay argued that an economy reproduces itself on the same scale if the tax falls exclusively on the net product appropriated by the landlords in the form of rent, for they unproductively consume it on unreproductive expenditures, e.g., luxury goods, servants, etc. By contrast, if the tax falls on products or on incomes other than rent, the scale of reproduction of the economy and the rent income will diminish. These results hinge on the Quesnaysian assumption that the composition of government expenditures financed out of tax receipts must be the same as that of the landlords. Specifically Quesnay noted that “the imposition of taxes, taken in this way directly from the net product, causes no change at all in the order of expenditure and distribution; that the cultivators as before receive the sums necessary to pay the revenue and ensure their returns; and that consequently the reproduction is necessarily the same”. Quesnay, however, was aware that changes in the composition of government expenditures affect the scale of society’s reproduction. He argued that “if the nation’s expenditure went more to the sterile expenditure side than to the productive expenditure side, the revenue would fall proportionately, and that this fall would increase in the same progression from year to year successively. It follows that a high level of expenditure on luxury in the way of ornamentation and conspicuous consumption is ruinous. If on the other hand the nation’s expenditure side goes to the productive expenditure side the revenue will rise, and the rise will in the same way ‘The industrialists, who are treated as “wage earners”, sell their products at a price merely sufficient to defray the labour and material costs. “The level of wages, and consequently the enjoyments which the wage-earners can obtain for themselves, are fixed and reduced to a minimum by the extreme competition which exists between them”. ‘The second economic problem’ in Meek 1963, p. 194. ‘““Extract from the Royal Economic Maxims of M. De Sully” in Kuczynski and Meek 1972, p. 1 . ” “The ‘Second Economic Problem’ in Meek 1963, p. 190. ” zyxwv 304 L. TSOULFlDlS zyxwvu increase successively from year to year. Thus it is not that the type of expenditure is a matter of indifference.”’* I1 THEMODELOF TAXINCIDENCE By casting the original “Tableau Economique” into the modern form of a double entry table it can be shown that the Quesnaysian principles of tax incidence hold on their own grounds. Suppose, with Quesnay, that the whole economy is divided into two sectors, e.g., agriculture and manufacturing. Further suppose that, the production relations between these two branches are described with the following input-output table in which, unlike the usual treatments (Phillips 1955), landlords are not treated as a separate branch of production. The reason is that their rent is a residual income accruing to them by virtue of their land ownership and earmarked for their unproductive consumption. The following quotation from Quesnay supports this treatment: “the proprietors are useful to the state only through their consumption; their revenue exempts them from labor; they produce nothing; and if their revenue was not circulated among those in the remunerative occupations, the state would be depopulated through the greed of these unjust and treacherous proprietor^".'^ Thus, it is permissible to treat the landlords as the consumers of the net product whose income is the residual after the payments of all inputs. From Table 1 we observe that only agriculture produces output in excess of all the inputs. Agriculture’s total output is worth five milliards-a “milliard” being the Physiocratic monetary unit of account-whereas its total input is worth three milliards. Hence, agriculture creates a net product equal to two milliards retained by the owners of land in the form of rent. By contrast, manufacturing designated as unproductive does not create any net product, since it “produces” an amount of total output only equal to the cost of its zyxwvutsrqponm zyxwvu zyxwvutsr zyxw TABLE 1 The “Tableau Economique” m on input-outpui table ~~ ~ output Input zy Agriculture Manufacture Rent Total Agriculture Manufacture 2 1 0 2 0 5 2 2 ~ ~~~ Final Demand Total Output 1 1 5 2 9 2 2 l 2 “Extracts from the Royal Economic Maxims of M. De Sully” in Kuczynski and Meek 1972, p. 12. l3 “Extracts from ‘Taxation’” in Meek 1963 p. 104. THE PHYSIOCRATIC THEORY OF TAX INCIDENCE 305 production, e.g., two milliards. Consequently, the rent rate in manufacturing is equal to zero or in Quesnay’s own words: “one must consider industrial wealth in the same light as the wealth which constitutes the costs of agriculture, in order to ascertain the extent and the nature of this wealth. The former costs enable the tradesmen in the towns to get a living; the latter enable the workers in the country to get a living; and both are annually renewed through the products of landed property”. l4 Table 1 can be transformed into the matrix of “technological coefficients” which shows the input requirements of each sector per unit of output, i.e., zyxw zyxwvu zyxwvu zyxwvu zyxwvuts where, matrix A includes the materials, the depreciation and the wages required for the production of one unit of output. Let P = b Pml be the row vector of prices at which the agricultural and manufacturing output exchange with each other and, let “1 ( R ) = [ R0u Rm be the diagonal matrix of the rent rates whose the elements along the main diagonal are the rent rates in agriculture and manufacturing, while all others are zero. Utilizing a linear model, the price equation can be written as follows, P=PA+PA(R) (1) where, P A represents the outlays of the farmers and industrialists in wages, raw materials and depreciation, whereas, P A ( R ) represents the amount of net product that agriculture produces. Equation (1) can be rewritten as zyxwvuts P - P A = P A ( R ) or P [ I - A ] = P A ( R ) or P = PA( R ) [ f - A ] - ’ (2) which is a system of two equations with three unknowns, i.e., the two prices and the rent rate.’’ zyxwv l 4 “Extracts from ‘Taxation’ ” ibid, p. 105. It is worth noting that Quesnay assumes that the landlords spend all the net product they receive. If some of it is hoarded, then Quesnay argued that “failure of the annual net product to return to the productive expenditure class” leads to an economic decline (“Explanation of the Tableau Economique” in Kuczynski and Meek 1972, p. xij). Similarly, if the state spends the tax proceeds on the maintenance of mercenary armies o r as a subsidy of a foreign state, then the tax will have adverse effects on the growth of the economy. (Kuaynski and Meek 1972, p. 3). The transformation of the “Tableau Economique” in a simple input-output table, however, does not address these questions. ‘*The system solves for a unique rent rate and a unique set of relative prices (except a multiplication factor). 306 zyxwvut zyxwvu zyx zyxwv zyxwvuts zyxwvu zyx L. TSOULFIDIS The structure of the Physiocratic system, however, is such that the price of industrial goods is equal to the cost of the inputs utilized into their production, i.e., there is no creation of net product. Hence, the price of industrial output is equal to the price of agricultural goods that it uses as inputs, so that a unit of agricultural product equal to one milliard is transformed into an industrial product of equal value (pa = p m ) . Quesnay argued that “the worker who makes a piece of cloth purchases the raw material and incurs expenses for his needs while he is making it. The payment he receives when he sells it returns him the purchase price and his expenses; the gain which his labour procures for him is confined to the restitution of the expenses he has incurred, and it is through this restitution that he is able to continue getting a living by his labour”.’6 Consequently, we can solve for the rent rate in the following way: A (R) [I-A]-’ Since Rm in the Physiocratic system is equal to zero it follows that agriculture’s rent rate can be estimated from the following equation, p a = pm = 2.5[-4pa -2prnI Ra or p a = p m = 1 and Ra = a 6 6 Considering prices expressed in the unit simplex, we have, + Pa = Pm = -5 which are the normalized equilibrium prices of agriculture and manufacturing respectively.l7 Utilizing the above described system of equilibrium relative prices and rent rate, we can test for the validity of the Physiocratic analysis of tax incidence by addressing different tax scenarios. (a) Direct tar on rent income Suppose with the Physiocrats that the state imposes a tax, i.e., t = 15% on the income from rent. Hence, the new price equation, P ’ =P ’ A ( R ’ ) ( l + t ) [ l - A ] - ’ (3) “Extracts from ‘Taxation’” in Meek p. 105. We define E as the unit vector in the same direction as the row vector of relative prices. nus, 2 E P = g Pi i-1 zyxwvut and the normalized prices are: Pn = P i C Iif, Pi where Pn = [ P a P m ] and P a and Pm are the normalized prices of agricultural and industrial products respectively. zyxw zyxwvuts zyxwvut zyxw THE PHYSIOCRATIC T H E O R Y OF T A X INCIDENCE solves for 307 zyxwv p a = p m = 2*5[-4pa+ .2pm]Ra’(I+ t) p a =pm = 1 which after the proposed normalization condition become: Pa = Pm = .5 Ru’ = -58. The relative equilibrium prices remain the same, whereas the new rent rate (Ra’) has been reduced by an amount equal to tax. Thus, Quesnay correctly argued that “taxes which are laid on the revenue, and which do not fall on cultivation, are not destructive at all, for cultivation will always amply compensate the proprietor, through the revenue which it brings in for him, for the burden of the tax placed on this revenue itself”.’* Consequently, the Physiocrates rightly argued that the after-tax net product will be equal to the pre-tax one whose a portion will be transformed into taxes. The reason is that a tax on rent incomes does not change the input-output structure of the economy and, therefore, prices remain the same. Hence, rent income is reduced by the amount of taxes and landlords fully bear the tax load, whereas, the economy continues to reproduce on the same scale. (b) Indirect faxes Suppose now an indirect tax ( t = 15%) is imposed on the turnovers of both sectors.” The new system of prices and rent rate will be, P’ = P ‘ A ( R ’ ) [ I- A ( I + t ) ] - ’ zyxwv zyx (4) where (I + t ) is a diagonal matrix whose elements of the main diagonal are the indirect tax rates and the elements off the main diagonal are zero, i.e., I l+tu 0 1-15 0 (I+t) = 0 l + t m = [o 1.151. [ Where t = fa = tm = 15%. By solving equation ( 4 ) , [:; [pa’ p m ‘ ]= [ p a ’ p m ’ ] 0 ] [ 3 * 6 3 4-17] lJ[RA’0 Rm’ -83 1.96 A ( R ‘ ) [I -A ( I + r ) ] - ’ we arrive at the following vector of relative prices, [pa’ p m ’ ]= [3.63(.4pa’ + .2pm’)Ra’ 4.17(.4pa’ + -2pm’)Ra’I (4a) The cost conditions in both agriculture and manufacturing have increased, with the industrial products becoming relatively more expensive because the “Extracts from ‘Taxation’ ” in Meek 1963. p. 106. Sernmler (1983) treats indirect taxation in a similar way. 308 zyxwvut zyxwvu zyxwvu zyxwvut zyxwvuts zy L. TSOULFIDIS tax increases their cost relatively more than the agricultural products. Relation (4a) can be rewritten as [pa' pm'] = [3.63 4.171 which after the proposed normalization condition becomes, Pa ' = 465 Pm' = -535. The economic meaning of this result is that the relative price of agriculture reduces by 4 7 5 , whereas, the relative price of manufacture increases by -075, indicating that the burden of taxation passes wholly onto agricultural production. Substituting these prices in relation (4a), we derive, Ru' = *43. Hence, the new rent rate falls by much more had we imposed the tax directly on rents. In conclusion, a uniform indirect tax in both sectors results firstly, in an increase in the price level which is, however, immaterial to industrialists and farmers whose exclusive concern is their relative prices; secondly, in making industrial products more expensive, by exactly the amount of the tax. The reason is that the industrialists, in order to function as such, have no other choice but to charge a price equal to cost; thirdly, in diminishing the net product, since farmers who are deprived of resources available for investment reduce them by an amount equal to the tax; and finally, in lowering the rent income and reducing the economy's scale of reproduction. With an indirect tax only on the turnover of industrial products (rm = 15%), the Quesnaysian price equation changes as follows: P' = P ' A ( R ' ) [ I- A ( I zyxwv zyxwvuts + mZ)]-' (5) where the diagonal matrix of tax rates is, In this case, the price vector will be, [pa' p m ' ] = [2*7(.4pa'+ -2pm')Ra' 3-11(.4pa'+ -2pm')Ra'l (5a) and the row vector of relative prices will be, P' = [ p a ' pm'] = (2.7 3-11] which after our usual normalization becomes, Pa' = 465 Pm' = ~ 5 3 5 zy zyx zyxw zyxwvut zyx THE PHYSIOCRATIC THEORY O F T A X INCIDENCE 309 The rent rate will be: As was expected, industrialists increase their relative prices by an amount equal to their tax obligations, whereas farmers are deprived of a portion of their investible product, equal to the amount of tax, because the higher manufacturing prices feed back onto their cost of production. As a result, the scale of reproduction in the economy as a whole is reduced and the final tax burden rests with the landlords who receive less rent. It is also noteworthy that because the rent rate for the agricultural sector is affected only indirectly it is reduced by less than in the previous case where the rent was affected directly and indirectly. In the case where only the agricultural product is taxed, by solving an analogous system, we amve at the following results: pa’ = p m ’ = 3.22 which after normalization become: and the rent rate reduces to: We observe that relative prices remain the same, for farmers cannot pass this tax onto the industrialists. The rent rate, in this case, has decreased by much more had we imposed the tax on the industrial products, because agricultural production has been taxed directly. As a result of the reduction in the net product the society’s scale of reproduction reduces. I11 CONCLUSIONS The input-output modelling of the Quesnaysian theory of social reproduction and taxation is consistent with the fundamental Physiocratic arguments that a tax levied on resources employed in agriculture disrupts the process of reproduction and diminishes society’s capacity for the creation of net product. In addition, a tax on manufacturing or services falls ultimately on agriculture for these activities do not create any net product. By contrast, a single tax on rent not only minimizes collection costs but moreover, does not disrupt the economy’s growth generating potential. Consequently, this type of tax is beneficial for the society at large, or in Quesnay’s own writings “the form of assessment which is the most simple, the most regular, the most profitable to the state, and the least burdensome to the taxpayers, is that 310 zyxwvuts zyxwvu zyxwvu L . TSOULFIDIS which is made proportionate to and laid directly on the source of continually regenerated wealth”.20 The rationale of such a tax is that the farmers benefit if part of the rent that they pay is converted into taxes, since they enjoy protection from the state without incurring any additional cost. Similarly, landlords are better off if their rent income is being taxed directly, since, as it has been shown in section 11, the rent rate declines at least as much as if the tax were imposed directly on agriculture or manufacturing. Consequently, the society at large is better off if the existing tax system is totally replaced by a single tax on incomes from rent. In proposing their policy for a single tax on the incomes of rent, the Physiocrats did not resort to any normative considerations, but rather pursued the logical conclusions drawn from their economic analysis. The transformation of the “Tableau Economique” into an input-output model enables us to substantiate their principal arguments in a rigorous way. It is a commonplace that the Physiocratic doctrine regarding the “unproductive” nature of manufacturing limited their insights to a great extent. Nevertheless, the theory of social reproduction described in the “Tableau Economique” enabled them to address microeconomic, as well as macroeconomic questions which are usually left out in the modern balanced budget or differential tax incidence analysis. Consequently, by correcting the Physiocratic theory, to include manufacturing and other sectors as productive, the Physiocratic approach to the question of tax incidence may find a new useful application. zyxwvutsrq zyxwvu zyxwvuts REFERENCES BLAUG,M. (1978). Economic Theory in Retrospecf, Cambridge University Press, New York. EINAUDI, L. (1967). The Physiocratic Theory of Taxation. In Economic Essays in Honour of Cwtau Cusself, Augustus M. Kelley. New York. FERRARA, F. (1958). Preface on the Doctrine of the Physiocrats. Internafionaf Economic Papers, Vol. 8. JOHNSON, J . (1966). The Role of Spending in Physiocratic Theory. Quarterly Journal of Economics, Vol. 80. KUCZYNSKI,M. and MEEK,R. (1972). Quesnay’s Tableau Economique, Augustus M. Kelley Publishers, New York. MEEK,R. (1963). The Economics of Physiocracy, Harvard University Press, Cambridge Massachusetts. PHILLIPS, A . (1975). The Tableau Economique as a Simple Leontief Model. Quarterly Journnl of Economics. Vol. 69. SAMUELS,W. (1961). The Physiocratic Theory of Economic Policy. Quarterly Journal of Economics, Vol. 76. SEMMLER. W. (1983). On The Clasical Theory of Taxation. Metroeconomica. Vol. 36. 2”L*Extractfrom the Royal Economic Maxims of M. De Sully’’ in Kuczynski and Meek 1972, p. 2.