{
  "id": "shadow-price",
  "title": "Shadow Price",
  "lang": "en",
  "source_lang": "en",
  "summary": "A shadow price measures the marginal value of relaxing a constraint or the economic value of a resource when its market price is absent or misleading.",
  "category": "economics",
  "subcategory": "microeconomics",
  "type": "concept",
  "aliases": [
    "Shadow value",
    "Implicit price",
    "Accounting price"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/shadow-price/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/shadow-price/",
      "json": "https://aiwiki.page/zh/shadow-price.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A shadow price is an implicit valuation of a resource, good, or constraint derived from an economic or mathematical model rather than directly observed in a transaction. In mathematical optimization, it usually measures how much the best attainable objective would improve if a constraint were relaxed by a small amount. In cost–benefit analysis, the term also refers to a value intended to reflect a good’s social opportunity cost or marginal social benefit when its market price is absent or distorted. These meanings are closely related, but a shadow price calculated for a private firm is not necessarily a measure of social value."
    },
    {
      "heading": "Economic interpretation",
      "text": "A constrained decision-maker must allocate limited resources among competing uses. The shadow price of a resource expresses the value of having slightly more of it, holding other features of the problem constant and allowing the allocation to be reoptimized.\n\nFor example, if one additional hour of machine capacity increases a factory’s maximum operating contribution by $25, the capacity constraint has a shadow price of $25 per machine-hour. This is not necessarily the machine’s rental price or the cost of operating it. It is the marginal value of additional capacity within the specified model.\n\nShadow prices therefore connect scarcity with choice. A resource can be physically limited without having a positive shadow price: if the available amount already exceeds what is useful at the optimum, a small increase may have no value. Conversely, a resource supplied without charge can have a high shadow price if its limited availability restricts valuable activity.\n\nThe units depend on the objective and the constraint:\n\n- For a profit objective measured in dollars and a labor constraint measured in hours, the shadow price is dollars per additional labor-hour.\n- For a cost-minimization problem with an output requirement, it can be dollars per additional required unit of output.\n- For a utility objective and a monetary budget constraint, it is utility per additional unit of money, not a monetary price.\n\nA shadow price must therefore be interpreted together with its objective, units, and constraint."
    },
    {
      "heading": "Mathematical formulation",
      "text": "Consider the maximization problem\n\n\\[V(b)=\\max_x f(x) \\quad\\text{subject to}\\quad g_i(x)\\leq b_i,\\qquad i=1,\\ldots,m.\\]\n\nHere, \\(x\\) represents the available decisions, \\(f(x)\\) is the objective, and \\(b_i\\) is the permitted level of resource use or activity under constraint \\(i\\). The function \\(V(b)\\) gives the optimal objective value for each vector of constraint limits.\n\nWhen \\(V\\) is differentiable and the relevant regularity conditions hold, the shadow price of constraint \\(i\\) is\n\n\\[s_i=\\frac{\\partial V(b)}{\\partial b_i}.\\]\n\nFor a sufficiently small change,\n\n\\[V(b+\\Delta b)-V(b) \\approx \\sum_{i=1}^{m}s_i\\Delta b_i.\\]\n\nThis is a local sensitivity statement, not generally a valuation formula for arbitrarily large changes.\n\nLagrange multipliers\n\nThe Lagrangian for this maximization problem can be written as\n\n\\[\\mathcal{L}(x,\\lambda) = f(x)+\\sum_{i=1}^{m}\\lambda_i\\bigl(b_i-g_i(x)\\bigr), \\qquad \\lambda_i\\geq 0.\\]\n\nUnder appropriate conditions, the optimal multiplier \\(\\lambda_i^*\\) equals the shadow price of \\(b_i\\). This follows from the envelope relationship between the optimized objective and the constraint parameter.\n\nThe Karush–Kuhn–Tucker conditions include complementary slackness:\n\n\\[\\lambda_i^*\\bigl(b_i-g_i(x^*)\\bigr)=0.\\]\n\nConsequently, a strictly slack inequality has a zero multiplier under these conditions. The reverse is not guaranteed: a binding constraint can have a zero shadow price, for example when it is redundant or when relaxing it does not improve the optimum at the margin.\n\nSigns depend on how the problem is written. Increasing an upper resource limit in a maximization problem cannot reduce the optimal value, provided the original choices remain feasible. Increasing a minimum output requirement in a cost-minimization problem, however, usually raises minimum cost. Multipliers reported by software must be interpreted using the software’s constraint and sign conventions."
    },
    {
      "heading": "Linear programming and duality",
      "text": "Shadow prices have a particularly direct interpretation in linear programming. Consider the primal problem\n\n\\[\\max_{x\\geq 0} c^{\\mathsf T}x \\quad\\text{subject to}\\quad Ax\\leq b.\\]\n\nIts dual is\n\n\\[\\min_{y\\geq 0} b^{\\mathsf T}y \\quad\\text{subject to}\\quad A^{\\mathsf T}y\\geq c.\\]\n\nThe variables \\(y\\) assign implicit prices to the resources represented by the rows of \\(A\\). The dual constraints require that the imputed resource cost of each activity be at least as large as its contribution to the objective.\n\nBy strong duality, when the primal has a finite optimum, the dual also has an optimum, and their optimal objective values are equal:\n\n\\[c^{\\mathsf T}x^*=b^{\\mathsf T}y^*.\\]\n\nThe optimal dual variables are commonly called shadow prices. Where the optimal-value function is differentiable with respect to \\(b\\), they give its marginal changes. At a kink, multiple optimal dual solutions may exist, so a unique two-sided marginal price need not be available.\n\nWorked example\n\nA factory produces goods \\(A\\) and \\(B\\). Their unit contributions, after variable production costs, are $30 and $20. Each uses one labor-hour, while \\(A\\) requires two machine-hours and \\(B\\) requires one. The factory has 100 labor-hours and 140 machine-hours:\n\n\\[\\max_{x_A,x_B\\geq 0}30x_A+20x_B\\]\n\nsubject to\n\n\\[x_A+x_B\\leq100,\\]\n\n\\[2x_A+x_B\\leq140.\\]\n\nThe optimum is\n\n\\[x_A=40,\\qquad x_B=60,\\]\n\nwith a total contribution of $2,400.\n\nThe shadow prices of labor and machine capacity are both $10 per hour. These values satisfy\n\n\\[y_L+2y_M=30, \\qquad y_L+y_M=20.\\]\n\nIf labor availability rises from 100 to 101 hours while machine capacity remains unchanged, the optimal production plan becomes 39 units of \\(A\\) and 62 units of \\(B\\). Contribution rises to $2,410.\n\nThe extra labor-hour is worth $10 because production can be reorganized. Its value is not simply the contribution associated with one product. Nor does the result imply that every additional labor-hour is worth $10: sufficiently large changes alter which constraints determine the optimum."
    },
    {
      "heading": "Shadow prices in economic appraisal",
      "text": "In welfare economics, shadow pricing aims to value project inputs and outputs according to their effects on social welfare rather than financial receipts and payments alone.\n\nMarket prices may diverge from social values because of externalities, market power, taxes, subsidies, price controls, or missing markets. A project’s financial profitability can therefore differ from its net social benefit.\n\nFor a small project in a competitive, undistorted market, the observed price may provide a suitable approximation to social opportunity cost. Where significant distortions exist, an appraisal may use adjusted prices or explicit measures of marginal social costs and benefits.\n\nCommon applications include:\n\n- Labor. A shadow wage estimates the opportunity cost of employing labor. It can differ from the wage paid when a project draws workers from unemployment, informal activity, or other jobs. Unemployment alone does not establish that labor’s opportunity cost is zero.\n- Traded goods and foreign exchange. Border prices and, where appropriate, adjustments for trade restrictions or foreign-exchange distortions can help distinguish resource costs from domestic financial payments.\n- Environmental effects. An appraisal may assign values to pollution, habitat changes, or emissions that have no direct transaction price.\n- Public services and nonmarket benefits. Changes in travel time, recreation, or environmental quality may be valued using evidence about willingness to pay or other appraisal methods.\n\nTaxes and subsidies often contain transfers between groups rather than net resource costs. Their treatment nevertheless depends on the appraisal’s scope, the distortions they create, and any administrative costs; they cannot always be removed mechanically.\n\nCarbon valuation\n\nThe social cost of carbon estimates the monetary value of damages caused by an additional unit of carbon dioxide emissions. It can serve as a shadow value in project appraisal.\n\nA different carbon shadow price arises from a model that minimizes costs subject to a specified emissions limit. The multiplier on that limit measures the marginal cost of tightening the emissions constraint—or the cost saving from relaxing it.\n\nThese quantities answer different questions. A damage-based estimate values the consequences of emissions; a target-based shadow price values scarcity under a chosen emissions constraint. The price of an allowance in an emissions trading system may reflect the latter under suitable conditions, but it need not equal estimated marginal climate damages."
    },
    {
      "heading": "Applications",
      "text": "Business and operations\n\nFirms use shadow prices in production planning, transport, scheduling, and capacity allocation. They indicate which constraints limit the modeled objective and help assess the marginal value of additional resources.\n\nIn a profit-maximization model, a resource’s shadow price can be compared with the incremental cost of obtaining more of it. That comparison is meaningful only if acquisition and operating costs are treated consistently and the proposed change lies within the relevant sensitivity range.\n\nConsumer and dynamic models\n\nIn consumer theory, the multiplier on a budget constraint measures the marginal utility of income. Ratios between relevant marginal valuations can be expressed in money terms, even when the underlying objective is utility.\n\nIn dynamic economic models, shadow values can be assigned to stocks such as capital, inventories, or natural resources. They measure how an additional unit of a stock changes the optimized value of current and future outcomes.\n\nPublic investment\n\nGovernments and development institutions use shadow prices in appraising transport, energy, water, and other projects. Benefits and costs occurring at different times are commonly converted to present values. Estimated social values can depend substantially on discount rates, the geographical scope of the appraisal, and the treatment of distributional effects."
    },
    {
      "heading": "Historical development",
      "text": "The modern mathematical interpretation developed alongside linear programming and optimization during the twentieth century. Leonid Kantorovich’s work on optimal resource allocation connected constrained production planning with implicit resource valuations. Subsequent developments in duality and activity analysis, including work associated with Tjalling Koopmans, clarified the relationship between efficient allocations and supporting prices.\n\nShadow pricing also became important in postwar development economics and project appraisal, where administered prices, trade restrictions, unemployment, and missing markets complicated the use of observed prices as measures of social value. The term now spans both the precise multiplier concept in optimization and the broader practice of estimating economic values for appraisal."
    },
    {
      "heading": "Limitations and interpretation",
      "text": "A shadow price is conditional on the model that produces it.\n\nLocal validity. A marginal value may apply only near the original optimum. Large changes can activate new constraints, remove bottlenecks, or change technology and behavior.\n\nNonuniqueness and discontinuities. Degeneracy, redundant constraints, or changes in the optimal solution can produce multiple multipliers or different marginal values for increases and decreases. Integer decisions and fixed setup costs can make a smooth marginal valuation inappropriate.\n\nModel boundaries. A private firm’s shadow price reflects its own objective and included constraints. It does not automatically incorporate environmental damage, distributional effects, or wider economic responses. A financial budget constraint can have a positive shadow price even when the resulting activity has negative effects outside the model.\n\nDependence on assumptions. Social shadow prices can vary with the assumed counterfactual, market structure, discount rate, and social welfare function. Where distributional weights are used, they also reflect judgments about how benefits and costs to different groups should be aggregated.\n\nDifference from observed prices. A market price records the terms of exchange; a shadow price records an implicit marginal valuation. They can coincide under particular conditions, but neither observed payment nor accounting cost alone establishes a resource’s shadow value."
    }
  ],
  "links": [
    "mathematical-optimization",
    "cost-benefit-analysis",
    "opportunity-cost",
    "scarcity",
    "utility-economics",
    "lagrange-multipliers",
    "karush-kuhn-tucker-conditions",
    "linear-programming",
    "welfare-economics",
    "externality",
    "market-power",
    "taxation",
    "subsidy",
    "price-controls",
    "willingness-to-pay",
    "emissions-trading",
    "profit-maximization",
    "consumer-theory",
    "marginal-utility",
    "present-value"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}