Heckscher-Ohlin Trade Theory

Table of Contents

    Heckscher-Ohlin trade theory explains how countries can trade because they have different proportions of productive resources, even when they share the same technology. It also helps explain why a change in trade can benefit some owners of labor or capital while harming others.

    The central distinction is between what a country has relatively more of and what a product uses relatively intensively. Connecting those two comparisons gives a conditional prediction about exports. Connecting product prices to factor payments gives a separate prediction about income distribution. Neither is a universal forecast of wages or investment returns.

    What Is Heckscher-Ohlin Trade Theory?

    The Heckscher-Ohlin model, also called the factor-proportions model, explains comparative advantage through differences in relative factor endowments. In its familiar two-country, two-good, two-factor version, a relatively capital-abundant country exports the capital-intensive good, while a relatively labor-abundant country exports the labor-intensive good, under the model’s production and demand assumptions. [1]

    Two Comparisons, Then a Trade Prediction

    Factor abundance compares countries. Factor intensity compares production processes. The model predicts a connection between the two when other possible sources of comparative advantage are held fixed.

    Here, capital means productive equipment and its services. It does not mean a country’s stock-market capitalization or the amount of money in its banking system. Labor means a consistently defined labor input; the basic model treats workers as one homogeneous factor.

    For the opportunity-cost foundation of trade, see comparative advantage and trade. The question here is more specific: how can different resource proportions create comparative advantage when countries have access to the same production methods?

    Factor Abundance vs Factor Intensity

    A country is capital abundant relative to another country when it has more capital per unit of labor. It can have fewer machines in total and still be capital abundant if its labor supply is proportionately smaller. With capital measured as K and labor as L:

    Relative Factor Abundance
    KA / LA > KB / LB
    Country A is capital abundant, and Country B labor abundant, relative to each other.

    A product is capital intensive relative to another product when its production uses a higher capital-to-labor ratio at the same factor prices. This comparison concerns input proportions, not whether one industry uses more machines or employs more workers in total. With flexible production methods, compare industries at a common wage-to-rental ratio and check that their intensity ranking does not reverse. [4]

    For example, an industry using two capital-service units per labor unit is more capital intensive than one using one-half of a capital-service unit per labor unit. The units must be comparable across the countries and industries. Counting differently productive workers as identical units can change the meaning of the comparison.

    The model then links relative endowments to production costs and trade. That link depends on the assumptions below; a large workforce alone does not establish that a country will export labor-intensive goods.

    The Assumptions Behind the Trade Prediction

    The standard teaching model deliberately holds several influences constant to isolate factor proportions. Its basic setup includes common technologies and preferences, with factors able to move between domestic industries but unable to migrate internationally. [1]

    • Two homogeneous factors and two goods. Each country’s capital and labor are comparable in quality; the goods use the factors with different relative intensities.
    • Common technology and constant returns to scale. Each country can use the same methods. Proportionately increasing all inputs proportionately increases output.
    • Competitive markets. Firms take goods prices and factor payments as given; active industries cover their production costs without persistent economic profit.
    • Full employment and domestic factor mobility. Labor and capital can work in either domestic sector, with their total supplies fixed for the comparison.
    • No transport costs or trade barriers. Goods can trade without a wedge between national prices.
    • Common homothetic preferences. In the standard endowment-based trade prediction, countries share preferences whose expenditure proportions do not change merely because all income rises proportionately.

    Continued production of both goods is especially important for the factor-price results below. Identical preferences help close the simple trade model; they are not a separate requirement for solving factor payments once common goods prices and diversified production have already been established.

    A Two-Country, Two-Factor Example

    Consider Alder and Birch, two fictional countries, producing machinery and cloth over one model period. Measure productive capital services and labor services on the same basis in both countries. Neither input can cross the border. All quantities and prices below are hypothetical.

    Both countries use the same fixed recipes: one unit of machinery requires two capital-service units and one labor-service unit; one unit of cloth requires one capital-service unit and two labor-service units. Machinery is therefore capital intensive and cloth labor intensive. Fixed recipes simplify the arithmetic by ruling out substitution between inputs. [2]

    Hypothetical endowments, measured consistently
    Country Capital-service units Labor-service units Capital per labor unit
    Alder 160 140 160 / 140 = 1.1429
    Birch 140 160 140 / 160 = 0.8750

    Alder is relatively capital abundant; Birch is relatively labor abundant. The result comes from their ratios, rather than treating a raw factor total as sufficient evidence.

    1. Find the full-employment production mix

    Let M denote machinery output and C cloth output. Fully using both factors requires:

    Common Technology and Resource Constraints
    2M + C = K
    M + 2C = L
    M = (2K − L) / 3
    C = (2L − K) / 3
    Each good uses both inputs; satisfying only one resource constraint is insufficient.

    For Alder, M = (2 × 160 − 140) / 3 = 60 and C = (2 × 140 − 160) / 3 = 40. Check: machinery and cloth use 2 × 60 + 40 = 160 capital units and 60 + 2 × 40 = 140 labor units. The same calculation gives Birch 40 machinery and 60 cloth.

    2. Specify demand before predicting exports

    Assume residents in each country spend half their income on each good. This is one particular common preference pattern, chosen to make demand checkable. In self-sufficiency, spending equality requires pMM = pCC, so the machinery-to-cloth price ratio is C / M: 40/60, or 2/3, in Alder; 60/40, or 1.5, in Birch.

    Now allow free trade at a common price of 3 money units per unit of either good. Below, we verify that these prices support factor payments of 1 per capital-service unit and 1 per labor-service unit. Each country then earns 300 money units: Alder earns 160 + 140, and Birch 140 + 160. Spending 150 on each good buys 150 / 3 = 50 units of each.

    Production and consumption at the hypothetical free-trade equilibrium
    Country Produces: machinery / cloth Consumes: machinery / cloth Net trade
    Alder 60 / 40 50 / 50 Exports 10 machinery; imports 10 cloth
    Birch 40 / 60 50 / 50 Exports 10 cloth; imports 10 machinery

    World production and consumption both equal 100 machinery and 100 cloth. Each country’s exports and imports are worth 10 × 3 = 30. The capital-abundant country exports the capital-intensive product, with no factor migration.

    In this fixed-recipe example, full employment pins down production, so opening trade changes consumption without changing the output mix. With flexible input proportions, production can also adjust. Do not interpret the table as evidence that trade always leaves output unchanged, or that both countries must consume identical quantities. Those features follow from this example’s specific assumptions.

    Stolper-Samuelson: Who Gains and Who Loses?

    The Stolper-Samuelson theorem concerns a change in a product’s relative price. In the competitive two-good, two-factor model, with both goods still produced and the factor-intensity ranking maintained, a higher relative price raises the real payment to the factor used intensively in that good and lowers the real payment to the other factor. It compares equilibria, without describing the transition. [3]

    Use the same machinery and cloth recipes. Let r be the payment per capital-service unit and w the payment per labor-service unit. Competitive zero-profit conditions are:

    Goods Prices Determine Factor Payments in This Example
    pM = 2r + w
    pC = r + 2w
    r = (2pM − pC) / 3
    w = (2pC − pM) / 3
    These linear formulas apply to the specified fixed recipes while both sectors operate.

    At pM = pC = 3, the equations give r = w = 1, confirming the payments used in the trade example.

    For a separate price experiment, hold the cloth price at 3 and raise the machinery price from 3 to 3.30, a 10% relative price increase, for an economy that continues producing both goods. This isolates a product-price effect; it is not a second world-market equilibrium with the earlier demand assumptions unchanged.

    A Product-Price Change Has a Larger Factor-Price Effect

    Capital: r = (2 × 3.30 − 3) / 3 = 1.20, an increase of (1.20 / 1 − 1) × 100 = 20%.

    Labor: w = (2 × 3 − 3.30) / 3 = 0.90, a decrease of (0.90 / 1 − 1) × 100 = −10%.

    Check both costs: 2 × 1.20 + 0.90 = 3.30 for machinery and 1.20 + 2 × 0.90 = 3.00 for cloth.

    The capital payment rises faster than the price of either good. To check real income, divide each factor payment by each goods price:

    Purchasing power per factor-service unit in the hypothetical price experiment
    Factor payment buys Before After
    Capital: machinery units 1 / 3 = 0.3333 1.20 / 3.30 = 0.3636
    Capital: cloth units 1 / 3 = 0.3333 1.20 / 3 = 0.4000
    Labor: machinery units 1 / 3 = 0.3333 0.90 / 3.30 = 0.2727
    Labor: cloth units 1 / 3 = 0.3333 0.90 / 3 = 0.3000

    Capital gains purchasing power over both goods; labor loses purchasing power over both. Calling this a real-income effect requires more than observing a nominal wage change.

    The result sorts people by factor ownership, rather than simply by their employer’s industry. The homogeneous labor factor receives the same wage in both domestic industries; capital receives the same rental payment. A household owning both factors has a combined income exposure. For the firm-level hiring connection, see factor markets and labor demand.

    Connecting this theorem to trade requires another step: establish how opening trade changes relative goods prices. In the standard H-O setting, trade raises the relative price of the abundant-factor-intensive export good, benefiting the abundant factor and harming the scarce factor. That conditional conclusion does not establish that every worker or capital owner in an actual country experiences the same outcome.

    Factor-Price Equalization and Its Limits

    Factor-price equalization says that trade in goods can equalize payments to identical factors across countries, even when the factors cannot move internationally. Common technology, common goods prices, competition and continued production of both goods provide the connection. It does not require workers to migrate to eliminate the payment difference. [4]

    In Alder and Birch, common goods prices of 3 enter the same two cost equations, giving r = w = 1 in both countries. Their different endowments affect output and trade, while the shared prices and technology identify factor payments.

    There is a useful boundary check. Producing positive amounts of both goods with the stated recipes requires the country’s overall K/L ratio to lie between the sector ratios of 0.5 and 2. Both fictional countries satisfy that condition. An endowment outside that range cannot fully employ both inputs using a positive mix of these two recipes; the diversified calculation would no longer describe that economy.

    More generally, extreme endowment differences can cause complete specialization, removing the pair of active-sector conditions that pins down equal factor prices. Technology differences, barriers and shipping costs also break the common-price/common-technology logic. Measured workers may differ in skills or productivity, so their observed wages do not represent payments to an identical factor. [4]

    The theorem therefore supplies a benchmark for identifying missing conditions. It does not promise equal international paychecks. Its capital rental payment is also different from a stock’s return: shareholders face purchase valuations, financing, depreciation, taxes and business risk.

    The Leontief Paradox: What the Evidence Tested

    Wassily Leontief’s 1953 study examined the capital and labor required for U.S. trade with an average 1947 composition. For each $1 million of exports and competitive import replacements, he calculated direct and indirect U.S. production requirements. Import replacements meant hypothetical domestic production using U.S. methods, rather than observed inputs in foreign factories. [5]

    Leontief’s requirements per $1 million of output, 1947 composition
    Input U.S. exports Competitive import replacements
    Capital, dollars at 1947 prices 2,550,780 3,091,339
    Labor, person-years 182.313 170.004
    Capital per person-year, rounded dollars 13,991 18,184

    The last row divides 2,550,780 / 182.313 = 13,991.21 and 3,091,339 / 170.004 = 18,183.92. Exports used less capital per person-year than import replacements, contrary to the simple expected pattern for the capital-rich United States. This became known as the Leontief paradox. [5]

    The finding did not settle every version of factor-proportions theory. In 1980, Edward Leamer showed that, when a country is a net exporter of both capital and labor services, comparing gross exports’ and imports’ capital/labor ratios is not the appropriate relative-abundance test in his Heckscher-Ohlin-Vanek framework. His comparison of net exports with consumption instead supported U.S. relative capital abundance. [6]

    The practical lesson is to match the test to the proposition. Distinguish physical capital from human capital and land, domestic replacement methods from foreign production, and gross trade from net factor content. A surprising ratio deserves investigation; it does not, by itself, prove that all endowment-based explanations are right or wrong.

    Heckscher-Ohlin vs Ricardian Trade Theory

    Different sources of comparative advantage in the basic models
    Question Basic Ricardian model Basic Heckscher-Ohlin model
    Why do production costs differ? Relative labor-productivity differences across countries Different relative factor endowments with common technology
    How many productive factors? One homogeneous labor factor Two factors, conventionally labor and capital
    Main distribution question The simplest version does not separate labor income from capital income Goods-price changes alter payments to different factors
    What needs separate analysis? Additional factors and adjustment costs Technology differences, factor heterogeneity and other trade drivers

    Both models concern comparative advantage. The difference is the explanation for relative costs, not whether opportunity cost matters. The production possibilities frontier provides the underlying production trade-offs; the existing Ricardian article develops the opportunity-cost calculations.

    How to Analyze a Factor-Endowment Trade Claim

    1. Define the factors and comparison group. Specify physical capital, skill groups, land or resources, with consistent units. A claim of abundance needs a relative benchmark.
    2. Check production intensity. Identify the inputs behind the export activity, including upstream production where relevant. Do not infer intensity from a product’s name.
    3. Separate causes of comparative advantage. Ask whether technology, resources, scale, institutions or supply chains differ, rather than attributing the whole pattern to one K/L ratio.
    4. Establish the product-price change. Distributional analysis needs evidence about relative prices, not just an announcement of higher trade volumes.
    5. Identify factor ownership and adjustment constraints. Skills, retraining, geographic mobility and ownership of several assets can make household outcomes differ from the homogeneous-factor benchmark.
    6. Keep national and financial questions distinct. The composition of exports does not determine the overall trade balance and capital flows, and a factor-payment prediction does not establish an attractive security valuation.

    For an assessment of trade restrictions, use the separate tariffs and quotas framework. For broader income dispersion, income inequality and the Gini coefficient describe outcomes that a two-factor theorem alone cannot measure.

    Common Mistakes

    • Comparing absolute factor totals. More workers in total does not necessarily mean more labor relative to capital.
    • Mixing abundance and intensity. A country is relatively capital abundant; a production process is relatively capital intensive.
    • Changing technology halfway through the example. Different country-specific recipes introduce another source of comparative advantage.
    • Calling a nominal increase a real gain. Compare a factor payment with the prices of goods it can buy.
    • Turning Stolper-Samuelson into a universal wage forecast. Establish the relative-price shock and the conditions supporting the theorem first.
    • Equating capital rents with equity returns. A production input’s rental payment is not a stock investor’s total return.
    • Treating Leontief’s import replacements as foreign-factory observations. The original estimates used U.S. production requirements.

    Limitations of Heckscher-Ohlin Analysis

    A Conditional Benchmark

    The model isolates one source of trade and one channel of income redistribution. Actual trade combines technology, factor quality, geography, policy, scale and firms’ production decisions. An economy’s aggregate gains also do not ensure that people who lose income receive compensation.

    The long-run mobility assumption is particularly consequential. A machine designed for one industry may be difficult to redeploy, and a worker may need new skills or a costly move. A model that assumes both factors can change industries immediately cannot quantify those transition losses. Stolper and Samuelson explicitly set the transition aside in their equilibrium comparison. [3]

    More factors and more goods also weaken simple two-by-two conclusions. Skilled and less-skilled labor need not respond alike, while resources and imported intermediate inputs complicate a single capital/labor classification. WTO’s review discusses how technology differences, nontraded goods and trade costs change empirical factor-content analysis. [1]

    Use the model to make a claim’s assumptions and proposed mechanism explicit. Evaluating an actual country, industry or household then requires evidence about those assumptions and the size of the relevant changes.

    Frequently Asked Questions

    Abundance compares countries’ relative resource supplies, such as capital per labor unit. Intensity compares the input proportions used by different production processes at common factor prices. Heckscher-Ohlin connects the relatively abundant factor to the relatively intensive export product under its assumptions.

    The basic model assumes common technologies so that different relative endowments can explain trade on their own. Country-specific productivity differences introduce an additional mechanism. They can matter greatly in actual trade, but should not be silently inserted into a worked common-technology example.

    No. The theorem connects a relative goods-price change to real factor payments under specified competitive, diversified production conditions. If the labor-intensive good’s relative price rises, the standard result favors labor. Applying it to an actual wage question requires identifying the price change, factor definitions, technology and mobility conditions.

    The benchmark requires comparable factors, common technologies and goods prices, and countries producing both goods. Specialization, trade costs, different skills, productivity and labor-market conditions interrupt that logic. Observed national wages do not automatically measure payments to an identical factor.

    It challenged a simple expected pattern: U.S. exports had lower capital per person-year than U.S. competitive import replacements in the original 1947-based calculation. The appropriate test depends on the model, factor definitions and treatment of trade imbalances. Leamer’s later reanalysis illustrates why the gross export/import comparison alone does not settle every factor-proportions proposition.

    Sources

    1. World Trade Organization: World Trade Report 2008, II.C.1(b), Differences in Resource Endowments, pp. 32–35. Institutional framework, model assumptions and empirical limitations.
    2. Steven M. Suranovic: International Trade: Theory and Policy, Chapter 5, The Heckscher-Ohlin Model. Author’s fixed- versus variable-proportions teaching framework; numerical examples here are original.
    3. Wolfgang F. Stolper and Paul A. Samuelson: Protection and Real Wages, 1941, pp. 58–73, public paper copy. Original assumptions, real-income argument and specialization limits.
    4. Paul A. Samuelson: International Trade and the Equalisation of Factor Prices, 1948, pp. 163–184, public paper copy. Original common-technology analysis and limits of equalization.
    5. Wassily Leontief: Domestic Production and Foreign Trade; The American Capital Position Re-Examined, 1953, especially p. 343, public paper copy. Original 1947-composition capital and labor requirements.
    6. Edward E. Leamer: The Leontief Paradox, Reconsidered, 1980, pp. 495–503, public paper copy. Primary reanalysis of the appropriate factor-content comparison.

    Sources accessed November 15, 2026 UTC. Alder, Birch and the product-price experiment are hypothetical; Leontief’s table is historical evidence with its original period and measurement limits.

    Disclaimer

    This article is for educational and informational purposes only and does not constitute investment, employment or economic-policy advice. Simplified trade models do not forecast an individual’s wages, a country’s exports or a security’s returns. Historical evidence and hypothetical calculations have the qualifications stated above.