RESEARCH DOSSIER
Labor Value and Surplus in Classical Political Economy
Abstract. Classical political economy explains the reproduction of economic society through the relations among production, labor, property, and distribution. This monograph reconstructs labor-value and surplus analysis as a family of related but distinct explanatory projects, extending from the physiocratic net product through Adam Smith, David Ricardo, John Stuart Mill, and Karl Marx to twentieth-century interpretations. Its central finding is that labor requirements, the purchasing power of wages, the production of a net product, and the appropriation of monetary income represent different analytical objects. They coincide only under specified technical, distributive, and institutional conditions. The study uses historical textual interpretation, accounting identities, comparative statics, and explicit numerical counterexamples to determine the conditions under which labor magnitudes explain relative prices and the division of the product. It distinguishes the existence of a physical surplus from its valuation, the generation of new value in Marx's analysis from its redistribution through competitive prices, and monetary labor-income shares from estimates of surplus labor. A two-industry calculation demonstrates how labor values and prices diverge with a positive profit rate even when technology is fixed. The empirical chapter specifies an internationally applicable identification strategy and evaluates the scope of published conflicting claims about price-value correspondence. The resulting synthesis preserves the explanatory contribution of classical political economy while locating each conclusion within its own assumptions. The monograph is a theoretical and historical investigation; its illustrative calculations are not observations of an actual economy.
Keywords: Classical political economy; labor theory of value; labor commanded; embodied labor; surplus product; surplus labor; labor power; surplus value; natural price; prices of production; differential rent; profit rate; reproduction; functional income distribution; input-output analysis
Research Object, Questions, and Method
The problem of value in classical political economy concerns the reproduction of wealth in a society whose members occupy different positions in production and acquire different claims on its results. A commodity has a material or practical character, enters exchanges at a price, and is produced through a sequence of activities that uses living labor and means of production. Each aspect gives rise to a distinct question. The conditions under which resources can reproduce themselves differ from the conditions under which their owners receive income. A theory connecting those conditions must specify both the physical operation and the distributive arrangement.
The subject of this investigation is the relationship among labor requirements, relative value, surplus labor, and the division of the net product into wages, profit, and rent as articulated in classical political economy. Its historical center is the development from Smith and Ricardo to Mill and Marx. Physiocracy establishes the preceding reproduction problem; subsequent interpretation and the marginal-product alternative clarify the competing ways of closing the analysis. The geographical horizon is the world economy, while the reconstructed texts predominantly concern European commercial and industrial development. That distinction matters: the historical range of the sources limits direct generalization, while the formal and empirical questions travel across national boundaries.
A labor-value proposition can perform at least three tasks. It may identify a common measure of reproducible products, explain changes in their relative exchange rates, or provide a measure of the social labor represented by a distributive claim. A successful performance of one task does not establish the other two. A measure of time can be constructed even if prices deviate sharply from it. A successful price prediction does not determine the ethical legitimacy of an income claim. A distributive theory can specify who receives the surplus without representing every price as a fixed multiple of labor time.
The research begins with these distinctions and proceeds by tracing textual claims to their applicable assumptions. The original works are cited at the chapter level, with the dates of first publication distinguished from those of revised editions and translations. For secondary literature available only through abstracts and bibliographic records, attribution is confined to the claims visible in those records. The formal examples are independent textbook-style reconstructions using circular production and competitive pricing. They illustrate consequences of stated premises, without proposing a new theoretical system.
Four types of statement require separate tests. An accounting identity must balance in consistent units. A formal implication must follow from the assumptions and survive a numerical counterexample. A claim about a historical author requires the relevant text. An empirical or causal assertion requires observation, identification, and uncertainty analysis. The method avoids carrying the evidential authority of one type of statement into another. Accordingly, this monograph does not estimate a contemporary global exploitation rate, report a new historical wage series, or treat a hypothetical production system as an actual national economy.
The Classical Problem of Reproduction
Reproduction and advances
The emergence of surplus analysis is inseparable from the recognition that production uses the outputs of prior production. The annual harvest is partly consumed, partly returned to the next agricultural cycle, and partly allocated as revenue. An economy that eats its seed or fails to maintain its instruments may display abundant consumption in a particular year while losing the capacity to sustain production into the next year. The physical prerequisites of reproduction therefore furnish an analytical boundary between total output and disposable output.
In Quesnay's Tableau economique (1759, approximate edition date), the flow of agricultural advances and the revenue of proprietors takes the form of a recurring circuit. The English translation identifies agricultural production as the source of net revenue and distinguishes expenditures that the physiocratic classification treats as sterile. The historical restriction to agriculture is central to the system's content. A modern circular account can retain the distinction between replacement and surplus while recognizing the productive contributions of manufacturing and services.
Consider an agricultural economy that produces 1,000 units of a single grain in one cycle. It needs 200 units for seed and other reproducible material inputs, and households undertaking production require 500 units for the next cycle under a stipulated consumption standard. The physically disposable remainder is 300 units. If the households' consumption requirement rises to 550 units without altering production or seed, the remainder falls to 250. If output increases to 1,100 while reproduction requirements remain 700, the remainder becomes 400. These are conditional consequences of the accounting specification; no price variable has yet entered.
Grain is a particularly convenient illustrative commodity because its gross product, replacement requirements, and consumption can be expressed in the same unit. A modern economy produces heterogeneous products. Steel cannot be added to hours of nursing or to liters of fuel as physical scalars. The net product must generally be represented by a vector. A valuation rule is necessary for a scalar monetary aggregate. That need explains why surplus and value analysis are closely connected despite their distinct objects.
Property and distributive closure
A feasible physical remainder does not determine its recipients. Under communal ownership the residual may finance common provision; under private land ownership it may be paid in rent; under a wage-labor system owners may retain part as profit; and under public ownership an allocation rule may direct it to investment. Each arrangement can exist with the same technical coefficients. The production possibility tells us what can be distributed, while institutions determine the rights under which distribution proceeds.
Reproduction also requires a temporal distinction. Replacement is a physical constraint, whereas scheduled payment is a financial constraint. A producer may possess enough output to replace materials in real terms yet become illiquid before sales occur. Conversely, borrowing may maintain monetary payments for a period even as productive assets deteriorate. A surplus account that aims to explain sustainable accumulation must include depreciation, inventories, timing, and the availability of working capital whenever they are material.
This is the classical problem in its general form: identify the resources reproduced by a continuing process, determine the residual available after specified prior claims, and explain the prices and social arrangements by which the residual acquires an economic distribution.
Adam Smith: Labor Commanded and Natural Price
Labor commanded
Smith's real and nominal price discussion (1776, Book I, Chapter V) relates the exchangeable worth of a commodity to the labor it enables its holder to obtain. In modern notation, if a commodity sells for money price p and one hour of comparable labor costs w, then p/w measures hours of labor commanded. This quantity answers a purchasing-power question. It changes when either p or w changes, even if the labor technically required to reproduce the commodity is unchanged.
Suppose a tool costs 40 currency units, requires two direct and indirect labor hours to produce, and the relevant hourly wage is 10. Its labor commanded is four hours. If the wage increases to 20 with the price fixed, labor commanded falls to two hours. The tool has not become technically easier to reproduce. Its ability to purchase labor has changed. The illustration shows why the two labor magnitudes cannot be used interchangeably.
Smith's discussion should be read in relation to the holder of wealth. The possession of goods confers a capacity to command other activity through exchange, which differs from the amount of activity expended on their production. As wages, skills, customary living standards, and the organization of exchange vary, the conversion between labor embodied and labor commanded may vary as well. A time measure does not by itself solve the problem of heterogeneous labor; a comparison must state which type of labor and wage rate is used.
Production and distributive components
In Book I, Chapter VI, Smith's account of the components of commodity prices (1776) distinguishes wages, profits on stock, and rent in developed society. The price of a processed good incorporates earlier costs and claims along its production chain. This provides a starting point for separating the immediate producer's costs from the economy-wide income distribution.
Consider a baker who purchases flour for 60, pays wages of 25, and sells bread for 110. The simple residual is 25 if other costs are absent. If the miller bought grain for 35 and paid wages of 15, the miller's residual is 10. If the farmer incurred no intermediate cost and paid wages of 20 out of grain receipts of 35, the farmer's residual is 15. The final bread price is matched by total wages of 60 and total residual income of 50. The gross receipts of the three producers total 205. Adding gross receipts and calling that sum the income generated by the chain double counts intermediate sales.
The illustration uses no fixed-capital depreciation or land rent. Including them requires further entries rather than a change to the arithmetic principle. Smith's distributive resolution of prices raises a general analytical question: how do the rates of wages, profit, and rent themselves become established? Defining price as their sum is an accounting operation; supplying a theory of their determination requires additional conditions. The distinction is important whenever price equations are interpreted as causal explanations.
Natural price and market adjustment
Smith's natural and market price analysis (1776, Book I, Chapter VII) organizes the relationship between ordinary distributive conditions and exchange. A sustainable or ordinary price includes the prevailing returns required to bring a commodity to market. Actual exchange prices may depart from that level when supply differs from effectual demand, and market restrictions may prevent rapid adjustment.
The lasting usefulness of the distinction is its treatment of market prices as the outcome of a process. Production takes time. Producers acquire information imperfectly. The inability to expand capacity immediately can produce prolonged high returns even in the absence of a legal monopoly. Barriers to entry and exclusive privileges can sustain a gap after the initial shortage disappears. The natural-price concept is therefore intelligible as a conditional center of adjustment rather than a constant independent of society and time.
A labor-cost relation belongs primarily to the explanation of reproducible supply conditions. It need not predict the price of every transaction. At the same time, demand cannot be removed from an explanation of which outputs are produced, in what quantities, and whether existing producers can sell them. A high labor expenditure on an unsalable article produces a failed investment, not an entitlement to a particular market value. This implication motivates the subsequent separation between technical labor requirements and socially validated production.
David Ricardo: Labor Requirements and Distribution
Labor requirements and the theory of relative value
Ricardo's chapter on value (1817/1821) refines the inquiry by linking relative commodity values to the labor necessary for production. The revised text expressly recognizes that the durability of capital, the mix of fixed and circulating capital, and the length of advances modify a simple proportional relation. Interpreting the mature argument as a universal equation of prices with labor times would discard these qualifications.
Under a restricted production structure, the labor argument has a transparent logic. If two reproducible articles require two and four comparable hours, if the advances and turnover structure are identical in relevant proportion, and if no differential rent or monopoly claim intervenes, the second article can cost twice the first. A technical improvement in the second industry that reduces required hours to three can reduce its relative normal price. The inference operates through altered costs, not through the moral deservingness of either producer.
Now let two products each require ten hours of labor in total. Product A is made with all labor paid at the date of sale, and product B has all labor paid one period before sale. Assume a uniform positive profit rate of 20 percent on advances and a wage rate of one currency unit per hour. Product A has a cost of 10 at sale. Product B has a price of 12 if the advance is financed by capital that earns the normal return. The labor contents are identical; the prices differ. This simplified timing example isolates one of Ricardo's modifications. It abstracts from common operating expenses, differing risk, depreciation, and changes in the price of inputs.
This is not a claim that interest creates material output. A time-dependent price can represent a claim on the existing output arising from the institutional organization of advances. The normal profit rate is an additional price-determining variable in this example. Removing it requires either a zero rate or the same relative time profile in both production processes.
Wages and the cost of reproduction
Ricardo's analysis of wages (1817/1821, Chapter V) connects the ordinary wage to the prices of necessities and customary requirements, while distinguishing the wage paid in the market from this reproducibility condition. A standard of subsistence includes historically formed consumption needs. Treating it as an unchanging biological minimum obscures important variation across places and time.
A constant real wage is compatible with substantial changes in money wages and commodity prices. Suppose a defined wage basket costs 100 currency units in the initial period and 120 in the next. A worker whose nominal wage rises from 100 to 120 receives the same basket, assuming its composition remains fixed. If money wages remain 100, the purchasing power over that basket falls by one sixth. An increase in the money wage is therefore not sufficient evidence of an increase in living standards.
Classical distribution theory uses the wage standard as a condition sustaining the labor force. In an actual economy the standard is shaped by households, public provision, health, education, labor institutions, and access to land or other income. Wages can depart from the customary standard for extended periods. The direction and speed of adjustment depend on institutions and on the evolution of demand and labor supply.
Profits and the wage-goods mechanism
Ricardo's chapter on profits (1817/1821) analyzes the residual available to the capitalist when production and distributive constraints are specified. In a simplified system where net product is fixed and divided only into wage and profit income, the inverse relation follows immediately. For net income N, with wages W and profit P, the identity N = W + P implies that an increase in W reduces P by the same absolute amount if N is unchanged.
This result must be separated from a general causal claim that all wage increases reduce total profits. Productivity improvements can increase N. A change in wage payments can also influence consumption, utilization, relative prices, and investment. For example, N may rise from 100 to 140, W from 60 to 80, and P from 40 to 60. Both incomes increase in absolute terms, while the labor share falls from 60 percent to approximately 57.1 percent. The profit share rises even though workers receive more. The wage-profit relation is a fixed-base comparative statement, not an unconditional prediction about a growing economy.
Within Ricardo's framework a rise in the real cost of customary wage goods places pressure on profits when the worker's reproduction standard is maintained. Agricultural productivity, trading arrangements, and food prices act as mediating variables. The importance of the argument is that conditions in a wage-goods sector can influence the whole distributive system. The labor time spent on a particular commodity alone does not determine the magnitude of profit on all capital advanced to produce it.
John Stuart Mill: Cost, Advances, and Property
Mill consolidates aspects of the classical account while clarifying the place of capital and contractual institutions in the distribution of the product. In Book II, Chapter XV of his analysis of profits (1848/1871), the capitalist advances production expenses and obtains the output under the usual contractual arrangement. The right to dispose of the product is significant. A physical contribution and a legally enforceable right of appropriation are separate facts, even if they coincide under a particular system of property.
Mill's ultimate analysis of costs (1848/1871, Book III, Chapter IV) reviews upstream labor expenditures and profit on successive advances. Differences in the relative significance of advances make profit rates relevant to relative values. The fact that a common profit rate enters every productive process does not imply that a change in that rate raises all relative prices simultaneously: relative price is always expressed against something else.
Consider a workshop that supplies two different services using the same total hours. One has a brief production interval and almost no working capital; the other requires a large stock of instruments and waits a year for payment. Their labor quantities can be the same, while a competitive cost calculation awards different normal returns. The second service also carries risks that may affect the required return separately from the pure timing effect. An analysis must distinguish the value of capital tied up in production from the depreciation charged in that period and from the expected losses priced into a contract.
Mill's position also clarifies a frequent error in the interpretation of classical economics. A residual calculation gives the amount available for a particular distributive claimant after other claims are deducted. It does not establish whether the residual is technically attributable to that claimant independently of the ownership institution. A change in contract can alter the identity of the claimant without changing output. Cooperative production is one such possibility: a group may both provide labor and own the instruments, and the same persons may receive payments classified by different accounting functions.
The division between production and distribution is consequently an important methodological distinction rather than an absolute separation. Property, contracts, incentives, and bargaining arrangements can affect technique selection and investment, while technique sets limits on distributive opportunities. Theoretical isolation of a production system is useful for determining feasibility. Explanation of historical development requires the interaction of production and institutions.
Karl Marx: Labor Power and Surplus Value
Socially necessary labor and the commodity
Marx's analysis of commodities (1867, Chapter I) identifies socially necessary labor time with the labor required under normal productive conditions and average social skill and intensity. An individual producer who spends twice the prevailing hours on an equivalent product does not thereby create twice its social labor value. The definition establishes a relationship between private activity and the social conditions under which products are exchanged.
Labor time in this framework is not simply a physiological clock measure. Skilled and differently intensified labor raise a conversion question. If a skilled hour is counted as several units of simple labor, then the reduction coefficient must be specified. Using the observed wage ratio as its sole estimator risks importing the monetary distribution into the labor-value measure intended to explain it. If the coefficients are wholly inferred from prices, empirical tests of an independent labor-value explanation can become circular.
A useful operational distinction therefore separates observed hours from equivalent standardized hours and treats the conversion as an assumption subject to sensitivity analysis. Data about occupation, training, and work intensity can assist construction, but each such adjustment adds a theoretical dependency. An unadjusted labor-hours estimate remains informative about direct time expenditure without being automatically identical to Marx's concept of socially necessary abstract labor.
Labor power and necessary labor
The decisive move in the purchase of labor power (Marx, 1867, Chapter VI) is the distinction between the capacity to work and labor performed during production. The reproduction cost of that capacity includes historically constituted needs. Under the model of wage employment the owner of capital acquires the use of the capacity for a contracted period. The employment relation makes possible a difference between the value paid for the capacity and the new value attributed to the labor it performs.
Assume an eight-hour working day and a wage basket whose reproduction requires four equivalent labor hours. Under the assumptions of exchange at labor values, the four hours equivalent to the value of labor power are necessary labor. The remaining four hours are surplus labor. The worker receives the agreed wage. The framework locates surplus value in the organization of production and appropriation rather than requiring deception in each commodity transaction.
In the labor and valorization process (Marx, 1867, Chapter VII), the production of useful objects and the production of surplus value are connected aspects of capitalist commodity production. This is a theoretical claim about a specific social organization. It differs from the empirical proposition that any person who works longer than the time needed for physical subsistence is participating in capitalist exploitation. A self-employed farmer may produce a physical surplus without purchasing wage labor, and an unpaid family worker may contribute to reproduction without receiving a wage. The social relation remains essential to the classification.
Constant capital, variable capital, and surplus value
Marx's rate of surplus value discussion (1867, Chapter IX) supplies the distinction among transferred value of means of production (c), value advanced for labor power (v), and surplus value (s). Under a simplified single-period account in which the entire capital advanced is consumed, commodity value is c + v + s; newly added value is v + s.
Let c = 60, v = 20, and s = 20 in common labor-value units. The product value is 100, new value is 40, the surplus-value rate s/v is 100 percent, and a simplified rate of profit s/(c + v) is 25 percent. The rates differ because the denominators represent different analytical questions. The first measures surplus relative to necessary labor under the model; the second measures a return relative to the capital advanced.
If the same labor operation uses c = 100 instead while v and s remain 20, product value becomes 140. The surplus-value rate remains 100 percent; the simplified profit rate falls to 16.7 percent. The mere equality of labor exploitation rates therefore cannot imply equality of capitalist profit rates across producers using different capital compositions. Conversely, equality of realized profit rates under competition does not imply equality of surplus-value production relative to labor power at each firm.
For durable instruments the relevant distinction is sharper. A machine costing 100 and transferring only 10 of value in a period is a stock of advanced capital; 10 is the corresponding cost flow. A profit-rate denominator cannot generally substitute the cost flow for the capital stock. Turnover periods, valuation dates, and depreciation methods matter. The elementary c + v example is valid only under its stated complete-consumption assumption.
Absolute and relative changes in surplus labor
The eight-hour example also separates changes in working time from changes in the reproduction cost of the wage. If the working day rises from eight to ten while necessary labor remains four hours, surplus labor rises from four to six; the surplus-to-necessary ratio increases from 100 to 150 percent. If the day remains eight hours but the wage basket can be reproduced in three equivalent hours, surplus labor reaches five and the ratio rises to approximately 166.7 percent.
The second calculation assumes that technical change reduces the labor content of the wage basket without changing the wage's basket composition. It is not inevitable that workers receive the same basket. They may obtain a portion of the productivity gain as a higher standard of consumption. Collective bargaining, competition for workers, public transfers, and employment conditions influence this outcome. The analytical ratio expresses the stipulated division of new labor value. The historical distribution of productivity gains remains to be investigated.
Labor Values and Prices of Production
A circular representation of labor requirements
The following construction makes the price question explicit using a single-product circulating-capital economy with fixed techniques, homogeneous labor, constant coefficients, and no joint production, taxes, rent, or foreign trade. It is a conventional linear production calculation, not a new theory. Let A be the matrix of physical inputs per unit of industry output, with input commodities arranged in rows and output industries in columns. Let l be a row vector of direct labor hours per output unit. Let lambda be the row vector of total direct and indirect labor requirements.
The labor requirement equation is lambda = l + lambda A. If the nonnegative matrix A has spectral radius below one, I - A is invertible and the nonnegative solution is lambda = l(I - A)^(-1). Its elements count the stipulated direct and indirect labor associated with a unit of final delivery under the chosen technology. This construction excludes a separately modeled production of labor power; adding household reproduction explicitly changes the system boundary.
For a gross production vector x, net output is y = x - Ax. The value identity then follows by multiplication: lambda y = lambda(I - A)x = lx. This states that the labor content of net output equals total direct labor used in the production system. The equation is a consequence of the technology and boundary specification, not an independently observed regularity. If imports, durable capital, or joint production are present, the input structure and labor account require corresponding extensions.
Suppose the economy uses the following coefficients.
| Input or labor | Industry A | Industry B |
|---|---|---|
| Commodity A | 0.20 | 0.40 |
| Commodity B | 0.10 | 0.10 |
| Direct labor hours | 0.50 | 0.20 |
The largest eigenvalue of A is approximately 0.3562, below one. The determinant of I - A is 0.68. Solving the two equations gives lambda_A = 0.47/0.68 = 0.691176 and lambda_B = 0.36/0.68 = 0.529412 hours per unit. The labor-value ratio B/A is approximately 0.765957.
Take gross outputs of 100 units of each commodity. Inputs consume 60 units of A and 20 of B. Net outputs are 40 units of A and 80 of B. Direct labor amounts to 70 hours. Valuing the net output in labor time gives 40(0.691176) + 80(0.529412) = 70 hours, apart from rounding. The example shows exactly how to avoid counting intermediate production twice.
Uniform returns and production prices
Now assume the physical inputs are purchased at the start of the period, wages are paid at the end, and competition establishes a uniform profit rate r on material advances. Let p be a row vector of production prices and w the hourly money wage. Then p = (1+r)pA + wl. If (1+r) times the spectral radius of A is below one, the equation can be solved as p = wl[I - (1+r)A]^(-1), with a normalization fixing the price unit.
Use the same economy and set r = 0.20. The determinant of I - 1.2A is 0.6112. The solution gives p_A/w = 0.464/0.6112 and p_B/w = 0.392/0.6112. Thus the production-price ratio B/A is approximately 0.844828. At r = 0, the ratio was approximately 0.765957. The technology and the total labor requirements did not change; the normal-price structure did.
Fixing p_A = 1 gives p_B = 0.844828 and w = 1.317241. For the stipulated output vector, intermediate purchases are worth 76.896552, the wage bill is 92.206897, and profit is 15.379310. Gross output is worth 184.482759; net output is worth 107.586207. The wage bill and profit exhaust that monetary net output. Profit divided by the value of intermediate advances is 0.20 as stipulated.
| Measure | Result |
|---|---|
| Labor-value B/A ratio | 0.765957 |
| Production-price B/A ratio | 0.844828 |
| Total direct labor hours | 70 |
| Value of material advances | 76.896552 |
| Wage bill | 92.206897 |
| Profit | 15.379310 |
| Net monetary output | 107.586207 |
The example is a counterexample to the unconditional identity of normal prices and embodied labor quantities. It is not a counterexample to labor as a necessary production input, nor to an aggregate labor account. It establishes the logical need to specify the distributive variable and the structure of advances. If wages were advanced at the beginning of the production period, the wage term would also be multiplied by 1+r. With a single uniform period this changes the scale relating wages to prices, while the relative prices in this particular model retain the same ratio at a given r. With different wage-payment dates or turnover times across industries, additional relative-price effects can arise.
A useful sufficient condition for price-value proportionality can be expressed directly. If lambda A is proportional to lambda, then l = lambda - lambda A is also proportional to lambda, and the price equation can have p proportional to lambda for a feasible range of r with an appropriate wage normalization. This condition captures a uniform relationship between direct and previously embodied labor across industries. It is sufficient in the restricted system and should not be presumed in observed industries with heterogeneous techniques.
The transformation question
Marx's posthumous Volume III develops prices of production (1894, Chapter IX) as cost price plus average profit and recognizes the need to account for inputs that themselves have prices incorporating profit. The foregoing circular calculation clarifies why this issue is substantive. If downstream prices are transformed while upstream inputs retain an incompatible valuation basis, the calculated costs, profits, and prices may fail to reproduce the same system.
The price of production is a regulating price within a theory of competition; market prices can diverge from it. Moreover, a uniform return on all capitals is a modeling restriction, not an observation about each industry's realized rate. Sectoral uncertainty, barriers to entry, technological rents, different depreciation schedules, financing risk, and noncompetitive market structures can maintain heterogeneous returns.
The transformation problem consequently contains more than one question. It concerns the determination of reproducible prices, the consistency of input and output valuations, the redistribution of surplus among capitals, and the conditions under which specified aggregate equalities are preserved. A consistent solution to one formulation does not establish that every other formulation has the same assumptions or invariants.
Rent, Profit, Wages, and the Allocation of Net Product
Differential rent
Ricardo's theory of differential rent (1817/1821, Chapter II) analyzes the excess produce obtainable from more productive land or applications of capital relative to the marginal alternative. In a competitive case, the less productive necessary alternative establishes the return available before rent; superior sites can yield an additional claim payable to the owner.
Construct three equal advances, each valued at 60 currency units at the beginning of a cycle, including all required costs and wages. Their grain yields are 120, 100, and 80 units, respectively. Assume a grain price of one currency unit, a common normal profit of 20 on a capital advance of 60, and all three sites required to satisfy demand. Gross returns before rent are 60, 40, and 20 after costs. The associated differential rents are 40, 20, and zero. Each tenant retains profit of 20 when the stipulated rents are paid. Aggregate profit is 60 and aggregate rent is 60. The result balances total sales of 300 with advances of 180 and residual income of 120.
A better site contributes larger physical output with an equivalent advance, but the payment of the differential to the landowner depends on ownership and the lease. A tenant could temporarily retain the differential under a fixed lease. The capability to appropriate a rent and the existence of the physical advantage should be recorded separately.
Real land rents can reflect location, infrastructure, regulatory restrictions, expectations of future development, exclusive access, and improvements made by owners or tenants. The term rent in the contractual sense may also include payment for produced assets. Differential agricultural rent is one mechanism, not a complete theory of every payment called rent. A precise account separates differential productivity, scarcity, ownership, and returns on improvements.
Income shares and stocks of capital
Suppose a set of firms produces net monetary output Y after deducting intermediate consumption and depreciation consistently. In a stripped-down account excluding taxes and mixed income, Y = W + P + R, where W is labor compensation, P profit income excluding rent, and R land rent. If Y equals 240, W equals 120, R equals 30, and P equals 90, the respective shares are 50, 12.5, and 37.5 percent. If the capital stock is 450, P/K is 20 percent. A profit share and a rate of profit answer different questions and need not move together.
Suppose investment doubles capital employed to 900 while net output becomes 360, wages 180, and rent 45. Profits increase to 135, and the profit share remains 37.5 percent, while the profit rate falls to 15 percent. The owner of capital receives a larger absolute income and a lower percentage return. An empirical argument about falling profitability must therefore distinguish stocks from flows, select the accounting period, and establish the treatment of inflation and depreciation.
Many modern monetary claims are secondary distributions of an operating surplus. Interest, dividends, rents, taxes on corporate income, and retained earnings can allocate revenue among claimants without being additive to the original production income. Treating the entire chain of payments as freshly generated surplus violates the accounting boundary. An aggregate distribution study must consolidate intra-sector transactions and specify whether the interest or rent entry is included in an operating surplus measure before deducting it.
Competing Interpretations of the Classical Tradition
The surplus interpretation and its alternatives
Twentieth-century reconstruction renewed the classical question of how relative prices and distributive variables follow from a given system of production. Sraffa's 1960 Production of Commodities is the bibliographic landmark. Roncaglia's (2009) account of Sraffa's contribution describes the simultaneous determination of relative prices and one distributive variable given the other. The simple input-output calculation above independently illustrates that restricted logical structure. It does not reproduce all of Sraffa's analyses of joint production, fixed capital, or choice of technique; the full original book was not available for examination in this research.
Garegnani's (1984) study of value and distribution situates Ricardo and Marx within a surplus tradition distinguished from later demand-and-supply formulations, and its accessible abstract contrasts price-equation and surplus-equation methods. The significance of this interpretive approach is that distribution requires a closure condition once technological reproduction is specified. It is possible to derive competitive prices from technical coefficients and a given wage or profit rate without first assigning a scalar labor value to every commodity. This mathematical property raises the question of what independent explanatory task a labor-value measure is intended to accomplish.
The alternative monetary interpretation developed in Foley's (1982) study of the transformation problem defines a conversion between aggregate money value added and labor time. The publisher abstract states that the magnitude assigned to labor power need not coincide with the labor embodied in workers' consumption once labor values and prices differ. This interpretation preserves a specified aggregate relation through a different choice of definitions. It does not imply that every relative commodity price equals its labor requirement.
One can see the difference by using the two-industry example. Its 70 direct labor hours correspond to net monetary output of approximately 107.586207. Dividing the money wage bill by this aggregate money-per-hour conversion attributes approximately 60 hours to wages and 10 to the residual profit. That result follows from the specified aggregate conversion and accounting boundary. An estimate of the embodied labor in a concrete wage basket would require the quantities purchased with wages. Different basket compositions can yield different embodied labor at the same money expenditure. It is consequently necessary to specify which definition of necessary labor is being used before comparing estimates.
Marginal productivity and distribution
A different route is exemplified by Clark's (1899) analysis of marginal productivity. Under competitive static conditions, the distributive payment to an input is tied to the change in output obtainable from an incremental unit of that input, with other inputs appropriately fixed. This contrasts with a residual distribution after a given wage or other independently specified distributive condition.
In a differentiable constant-returns production function F(K,L), with competitive output price normalized to one and factors paid their marginal physical products, W = L times the partial derivative of F with respect to L and capital income equals K times the partial derivative with respect to K. Euler's theorem gives their sum as F under the homogeneity assumption. For the illustrative function F = K^(0.4)L^(0.6), factor shares are 40 and 60 percent. These shares are consequences of the functional form and factor-pricing conditions. A separately established relationship between observational income shares and the functional parameters is needed before using this example as an empirical explanation.
Where firms have market power, face financing constraints, or employ workers with asymmetric bargaining options, actual factor payments may depart from the competitive first-order conditions. Heterogeneous capital goods also require careful treatment: an aggregate monetary capital measure can change when relative prices and the distributive variable change even if physical equipment is unchanged. The comparison therefore turns on the way the production set, prices, property, and distribution are specified.
Both approaches can describe a surplus in a particular sense. Classical surplus analysis asks what net product remains after reproduction and given distributive claims. Marginal productivity analysis asks what changes in product accompany incremental changes in inputs under a stated technology. These are different operations on a production system. A productive contribution does not by itself establish an ownership right to the resulting increment, and an observed ownership payment does not uniquely identify a physical causal contribution.
What is established by the comparison
The competing approaches should be compared through explicit hypotheses. A classical relative-price proposition imposes constraints on reproducible prices given techniques and distributive conditions. A marginal-pricing proposition imposes constraints on factor choices and payments given technology, prices and a market structure. A monetary transformation interpretation imposes consistency conditions on aggregate labor and monetary accounts. An institutional explanation imposes a relation between rules, bargaining positions and realized distribution. Their predictions can overlap in particular environments. A single observation that is consistent with all of them cannot discriminate among them.
Empirical Observables and Identification
Price-value evidence
Shaikh's (1998) empirical strength argument proposes a relationship in which production prices mediate between market prices and the labor structure of production. Kliman's (2002) statistical critique of price-value correspondence reports an opposing result when controlling for the effect of sector size in United States industry data for 1977-1997. These are published claims about empirical relationships, not independently replicated results of this dossier. The available abstract-level evidence establishes the disagreement and its identification concern. It does not justify choosing a universal numerical error rate or concluding that either paper settles every formulation of labor-value theory.
The size problem can be explained without a regression. Suppose three industries have gross quantities of 1, 10, and 100. Each unit has a labor requirement of one hour. Suppose their prices are respectively 1, 2, and 3. The vectors of sector totals are labor (1, 10, 100) and money (1, 20, 300). They are highly correlated because each contains the common quantity multiplier, even though price per labor hour varies by a factor of three. A correlation of totals is insufficient to establish that relative unit prices are proportional to relative labor values.
A meaningful study must choose an estimand and a unit of observation. It might ask how much the relative price of an industry changes following changes in total labor requirements, conditional on advances and other reproducible inputs. Alternatively, it might ask whether estimated production prices predict normalized market prices better than simpler cost or input-share baselines. It should not switch between these hypotheses once results are observed.
A reproducible research design
For a modern test across countries and industries, the OECD Inter-Country Input-Output tables provide a candidate source of interindustry and international trade flows. A subsequent empirical investigation would need a precisely dated dataset release, matching occupation and hours data, price indexes or physical quantities compatible with the input coefficients, and an explicit treatment of imported inputs. The database is identified as a research infrastructure, not downloaded or analyzed in the present study.
First, establish the production boundary and build a matrix with rows representing inputs and columns representing output activities. Distinguish domestically produced from imported inputs. Specify the valuation basis, such as basic prices, and the treatment of margins and product taxes. In a monetary table the coefficients describe monetary input requirements at observed prices; interpreting them as physical coefficients demands extra information. Using current prices to compute labor values and then claiming independent price prediction can introduce mechanical dependence.
Second, construct direct hours per output unit with transparent estimates of hours worked, multiple jobholding, skill adjustments, and sector coverage. If only monetary output is available, the dimension of the coefficient is hours per unit of observed currency output, not hours per physical article. A study should describe that limitation directly. Where possible, use physical quantity or constant-price measures and document the deflation procedure.
Third, calculate the total labor requirements under explicit productive-system assumptions. Compute a price prediction from the labor-value specification and, separately, from a reproducible-price equation with a specified wage, profit rate and advance timing. Calculate sensitivity to depreciation, industry aggregation, imported labor conventions, labor heterogeneity and alternative normalizations. Retain the original dataset revisions and exact code.
Fourth, compare predictions using relative log prices or deviations after removing common output-scale factors. Report absolute and relative prediction errors, both weighted and unweighted, and check whether estimates are driven by a few large industries. Out-of-sample predictions should be based on information available before the predicted observation. Statistical uncertainty must account for interindustry dependence; industries connected through input-output linkages are not independent replicates.
Fifth, include plausible competing predictors: material cost shares, wage costs, capital intensity, sector-specific markups and price persistence. A model that improves on a naive common-ratio assumption can still perform worse than a conventional cost-based baseline. A transparent study should compare predictive performance and test the assumptions that distinguish explanations, including the possibility that results are sensitive to the aggregation convention.
A causal analysis of productivity and relative prices would require further identifying conditions. Technical improvements may coincide with demand shocks, market entry, or changes in bargaining. An observed correlation between a decline in labor hours and lower prices can reflect a shared cause rather than the specific proposed direction. A research design can exploit credibly exogenous shifts in technology and test pre-existing trends, but the identification assumptions must be justified for the actual data. No such causal estimate is presented here.
Labor-income share and surplus measurement
The ILO's labor-share indicator metadata explains why cross-country comparisons need adjustments for the labor income of self-employed workers. Employee compensation alone can materially understate labor income where own-account work is common. Gross domestic product also includes components that do not correspond directly to the simplified Marxian new-value account.
Consider a country with measured wage payments of 300, mixed household business income of 100, gross operating surplus of 250, and relevant net production and product taxes of 50. GDP in this simplified income presentation is 700. The unadjusted employee-compensation share is 42.9 percent. If 60 of mixed income is estimated as labor income, the adjusted share becomes 51.4 percent. Neither result is an independently observed surplus-value rate. The allocation of mixed income matters, depreciation is included in gross operating surplus, and taxes enter the denominator. The data illustrate an accounting sensitivity, not the composition of any actual country.
Gross and net measures should be distinguished. In the example, subtracting 70 of capital consumption from GDP reduces the net total to 630. If the labor-income numerator is unchanged, the adjusted labor-income share of net output is approximately 57.1 percent. A researcher who changes denominators can therefore obtain a substantially different share with no change in observed compensation.
In the special hypothetical economy in which net value added equals v+s, employee income is exactly v, and the residual equals s, a labor share alpha = v/(v+s) implies s/v = (1-alpha)/alpha. A share of 60 percent gives a surplus-to-necessary ratio of two thirds. The mapping is an algebraic implication of strong definitions. A modern labor share cannot be inserted into it without addressing depreciation, taxes, mixed income, the scope of productive labor, property income and the labor-to-money conversion. A cross-national comparison must keep these distinctions intact.
Historical and Institutional Boundaries
Historical organization of labor
Classical wage-capital-land relations are clearest when the producer works for a wage, the capitalist owns or controls the advance, and the landowner holds a separate property right. The actual organization of production is more varied. Family production, domestic labor, servitude, slave labor, self-employment, cooperative enterprise, and public employment allocate work and income through different combinations of property, legal status and contract. The same technical labor coefficient may coexist with different distribution mechanisms.
A plantation employing coerced labor can produce a saleable surplus, but the absence of a voluntary wage contract changes the historical mechanism of appropriation. A self-employed artisan bears the risks of capital advances and contributes labor; the resulting mixed income cannot be uniquely classified by looking at cash receipts. A producer cooperative can retain earnings for shared investment. The analytic task is to identify the relation that governs appropriation rather than infer it from the use of a familiar income label.
Unpaid household labor is another boundary issue. It contributes to maintaining people and labor capacity, but much of it lies outside measured market output. An economy's ability to reproduce its labor force depends upon activities that a market-based account can leave unpriced. Including a household-production satellite account is one way to study this dependency. It requires an explicit valuation procedure and should not be merged mechanically with existing GDP or with the labor value of market commodities.
Resources, scarce goods, and joint outputs
A labor requirement measure is particularly intelligible for goods produced repeatedly under comparable conditions. Scarce originals, unique locations, exclusive rights and exhaustible natural resources pose different price questions. A unique painting does not have a normal reproduction technique for producing the same original; its transaction price may depend on scarcity, provenance and demand. A resource deposit may command income because its grade or location reduces costs or because access is restricted. In both cases the conditions of ownership and substitutability must enter.
Joint production breaks the simple single-output assumption of the input-output model. If a process yields both product A and product B, its labor hours cannot be uniquely allocated to separate outputs solely from the fact that the outputs are produced together. Different allocation conventions produce different assigned values. Additional processes, technical relations, or an explicit allocation rule are required. A robust investigation should test whether its conclusions survive alternative allocations rather than adopt a single convention without explanation.
Capital goods with multi-period service flows raise similar problems. Their net contribution cannot always be represented by treating each input as fully used within one period. A complete treatment tracks age, depreciation, maintenance, replacement and possible salvage value. Historical cost, replacement cost and market value answer different accounting questions. The determination of the profit rate depends on which capital-stock valuation is used and why it is appropriate for the intended inference.
Global production and institutional heterogeneity
International production networks separate the place where work is performed from the place where income is booked and ultimately spent. A component made in one country may be assembled in another, while the ownership of an intangible asset and financing claims are registered elsewhere. Labor-value accounting must specify whether imported inputs bring foreign labor requirements into the domestic product measure. Distribution accounting must distinguish domestic production income from national income after cross-border property flows.
The relevant real wage also depends on the country's consumption structure, exchange rates, purchasing power and public provision. Comparing nominal wages across countries without an appropriate conversion provides little information about labor reproduction. Comparing adjusted labor-income shares does not settle differences in working conditions, hours, wealth, household production or social welfare. Similar shares can coexist with substantially different living standards and bargaining arrangements.
The classical object remains relevant because it makes production, replacement and distributive claims analytically visible. Its portability across institutional environments depends on rebuilding the actual assumptions rather than exporting a historical class partition unchanged.
Analytical Synthesis
Linking the distinct objects
The evidence and calculations support a synthesis centered on the relationship between reproduction, valuation and appropriation. Reproduction is the technical and social maintenance of the conditions of production. Valuation attaches an economic measure to outputs and inputs under an identified convention. Appropriation distributes the available income and assets under operative institutional rights. These three activities interact, and the distinction makes it possible to identify the location of a disputed proposition.
A physical reproduction question is answered with quantities, technique and the maintenance requirement. An embodied-labor question counts direct and indirect labor through a defined technology. A price question requires a valuation or competitive closure, including the treatment of capital advances and the distribution variable. A surplus-labor question requires a specification of necessary labor and the employment relation. A distribution question requires definitions of the income categories and the rights by which claims are secured. Substituting an answer from one domain for another creates the strongest recurrent errors in the classical debate.
The conditions of explanatory priority
Labor requirements explain an important constraint on reproducible commodities: reducing normally required labor can reduce reproducible cost under specified conditions. Their predictive sufficiency, however, depends on the uniformity of other relevant costs, techniques and advances. The equal-labor timing example and the two-industry circular calculation show that a positive profit rate with heterogeneous relations between material inputs and living labor changes the ratio of production prices while labor requirements remain fixed.
This conclusion leaves room for approximate price-value relationships. Similar cost structures across industries, a modest influence of differential advance timing, or common changes in total labor productivity may generate near proportionality. Approximation must be evaluated against appropriate alternatives. It cannot be inferred solely from the fact that many productive processes use labor or from the scale of sector outputs.
A second conclusion concerns the wage-profit opposition. At a fixed level of distributable net income, a larger share to one claimant reduces the amount available to others. Over time, productivity, employment, investment, relative prices, working time and institutional changes alter the size of the distributable product. Both absolute wages and profits can rise; their shares and rates may move in different directions. The apparent contradiction between these statements disappears once the fixed-base calculation and the dynamic comparison are distinguished.
A third conclusion concerns exploitation and its measurement. Marx's surplus-value account identifies an organizing relationship between the value of labor power and the labor performed under capitalist control. An aggregate ratio based on wage and residual income can represent this relation under a specified monetary interpretation and boundary. The ratio alone does not establish coercion in a legal sense, identify individual unfairness, or determine the distribution of welfare. Such further judgments require additional evidence and normative premises.
The role of institutions in classical explanation
Technology determines feasible production; it does not establish an institutional distribution by itself. The illustrative grain economy can allocate its residual through several forms of ownership. Differential productivity can generate an excess without fixing the contractual recipient. Competitive pricing can reward capital advances without settling the social justification of the corresponding claim. Conversely, institutional arrangements can alter investment, technology, labor utilization, and therefore the physical surplus.
The significance of classical political economy is the ability to examine these interdependencies at the level of a continuing social system. Its arguments are strongest where the production boundary, distributive variables and property arrangements are explicit. The comparisons here identify conditions of applicability rather than a single privileged universal formula. They constitute a synthesis of established analytical resources and independently derived illustrations, with no priority or novel-theory claim.
Further research
A subsequent independent dossier could implement the price-value identification design across country-industry panels and compare normal-price and labor-value predictions on the same harmonized dataset. Another could reconstruct historical wage baskets using archival price and household-budget series, including nonmarket reproduction and regional variation. A third could examine how joint production, fixed capital and changing techniques modify the simple linear representation. These empirical and theoretical extensions would require their own source access and validation.
Conclusion
Classical political economy brings the production and distribution of wealth into a single field of inquiry. From the physiocratic reproduction circuit to Smith's labor-commanded measure and natural prices, Ricardo's labor requirements and distributive conditions, Mill's analysis of advances, and Marx's labor-power and surplus-value distinction, the tradition develops questions that remain analytically meaningful across institutional settings.
The work establishes that the labor embodied in a reproducible commodity and the labor commanded by its price are generally distinct; that physical reproduction and distributable income require an explicit accounting boundary; that uniform positive returns can produce prices different from labor-value ratios; that surplus-value rates, profit rates and income shares have different denominators; and that differential rent requires both a productive differential and a mechanism of appropriation.
The original calculations establish these claims within transparent examples. The review of secondary scholarship locates the continuing disputes over price determination, transformation and statistical identification. The international measurement discussion specifies what evidence would be required to test the corresponding propositions using actual data. It does not present newly measured global aggregates or settle the full empirical debate.
The lasting analytical contribution of the classical tradition is its insistence that production reproduces a social economy and that the product becomes income through economic and institutional relations. Its explanatory force depends on keeping labor requirements, price formation, physical surplus and distributive rights in their appropriate units and connecting them through stated assumptions. This discipline makes possible a rigorous comparative study of value and surplus without converting a historical theory into an unconditional claim about every economy.
References
- Clark, John Bates. (1899). The Distribution of Wealth: A Theory of Wages, Interest and Profits. 1908 reprinting. New York: Macmillan; Liberty Fund/Econlib, Chapter XIII, The Products of Labor and Capital, as measured by the Formula of Rent. Online text or record
- Foley, Duncan K. (1982). The Value of Money the Value of Labor Power and the Marxian Transformation Problem. Review of Radical Political Economics, 14(2), 37-47. DOI: 10.1177/048661348201400204. Online text or record
- Garegnani, Pierangelo. (1984). Value and Distribution in the Classical Economists and Marx. Oxford Economic Papers, 36(2), 291-325. DOI: 10.1093/oxfordjournals.oep.a041640. Online text or record
- International Labour Organization. (n.d.). SDG indicator 10.4.1: Labour share of GDP. SDG Metadata Translation Project, English metadata hosted by the World Bank. Accessed November 17, 2026. Online text or record
- Kliman, Andrew J. (2002). The law of value and laws of statistics: sectoral values and prices in the US economy, 1977-97. Cambridge Journal of Economics, 26(3), 299-311. DOI: 10.1093/cje/26.3.299. Online text or record
- Marx, Karl. (1867). Capital: A Critique of Political Economy, Volume I. Hamburg: Otto Meissner. English translation by Samuel Moore and Edward Aveling (1887), edited by Friedrich Engels; Marxists Internet Archive. Chapter I, Commodities. Online text or record
- Marx, Karl. (1867). Capital: A Critique of Political Economy, Volume I. Hamburg: Otto Meissner. English translation by Samuel Moore and Edward Aveling (1887), edited by Friedrich Engels; Marxists Internet Archive. Chapter IX, The Rate of Surplus-Value. Online text or record
- Marx, Karl. (1867). Capital: A Critique of Political Economy, Volume I. Hamburg: Otto Meissner. English translation by Samuel Moore and Edward Aveling (1887), edited by Friedrich Engels; Marxists Internet Archive. Chapter VI, The Buying and Selling of Labour-Power. Online text or record
- Marx, Karl. (1867). Capital: A Critique of Political Economy, Volume I. Hamburg: Otto Meissner. English translation by Samuel Moore and Edward Aveling (1887), edited by Friedrich Engels; Marxists Internet Archive. Chapter VII, The Labour-Process and the Process of Producing Surplus-Value. Online text or record
- Marx, Karl. (1894). Capital: A Critique of Political Economy, Volume III. Edited by Friedrich Engels. Hamburg: Otto Meissner; English text, Marxists Internet Archive, Chapter IX, Formation of a General Rate of Profit and Transformation of the Values of Commodities into Prices of Production. Online text or record
- Mill, John Stuart. (1848/1871). Principles of Political Economy with Some of Their Applications to Social Philosophy. 7th ed., edited by William J. Ashley (1909). London: Longmans, Green and Co.; Liberty Fund/Econlib digital edition. Book II, Chapter XV, Of Profits. Online text or record
- Mill, John Stuart. (1848/1871). Principles of Political Economy with Some of Their Applications to Social Philosophy. 7th ed., edited by William J. Ashley (1909). London: Longmans, Green and Co.; Liberty Fund/Econlib digital edition. Book III, Chapter IV, Ultimate Analysis of Cost of Production. Online text or record
- Organisation for Economic Co-operation and Development. (n.d.). Inter-Country Input-Output Tables (ICIO). OECD Data, dataset description. Accessed November 17, 2026. Online text or record
- Quesnay, Francois. (1759, date uncertain in the digital edition). Tableau economique. English translation by Arthur Eli Monroe, in Early Economic Thought (1923), pp. 336-348. Cambridge, MA; Marxists Internet Archive. Online text or record
- Ricardo, David. (1817/1821). On the Principles of Political Economy and Taxation. 3rd ed. London: John Murray; Liberty Fund/Econlib digital edition. Chapter I, On Value. Online text or record
- Ricardo, David. (1817/1821). On the Principles of Political Economy and Taxation. 3rd ed. London: John Murray; Liberty Fund/Econlib digital edition. Chapter II, On Rent. Online text or record
- Ricardo, David. (1817/1821). On the Principles of Political Economy and Taxation. 3rd ed. London: John Murray; Liberty Fund/Econlib digital edition. Chapter V, On Wages. Online text or record
- Ricardo, David. (1817/1821). On the Principles of Political Economy and Taxation. 3rd ed. London: John Murray; Liberty Fund/Econlib digital edition. Chapter VI, On Profits. Online text or record
- Roncaglia, Alessandro. (2009). Production of Commodities by Means of Commodities. In Piero Sraffa, pp. 41-59. Palgrave Macmillan. DOI: 10.1057/9780230234697_3. Online text or record
- Shaikh, Anwar M. (1998). The Empirical Strength of the Labour Theory of Value. In Riccardo Bellofiore (ed.), Marxian Economics: A Reappraisal, pp. 225-251. Palgrave Macmillan. DOI: 10.1007/978-1-349-26121-5_15. Online text or record
- Smith, Adam. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Edited by Edwin Cannan (1904). London: Methuen; Liberty Fund/Econlib digital edition. Book I, Chapter V, Of the Real and Nominal Price of Commodities. Online text or record
- Smith, Adam. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Edited by Edwin Cannan (1904). London: Methuen; Liberty Fund/Econlib digital edition. Book I, Chapter VI, Of the Component Parts of the Price of Commodities. Online text or record
- Smith, Adam. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Edited by Edwin Cannan (1904). London: Methuen; Liberty Fund/Econlib digital edition. Book I, Chapter VII, Of the Natural and Market Price of Commodities. Online text or record
- Sraffa, Piero. (1960). Production of Commodities by Means of Commodities: Prelude to a Critique of Economic Theory. Cambridge: Cambridge University Press. African Union Library catalog record. Online text or record