An input-output table is a grid that maps how every industry in an economy buys from and sells to every other industry. Each row shows where an industry’s products go, and each column shows where an industry gets its materials. Together, the table captures the full web of economic relationships, making it possible to trace how a change in one sector ripples through all the others.
How the Table Is Organized
Picture a large spreadsheet. Industries are listed both across the top (as columns) and down the side (as rows). The cell where “steel” meets “automobiles” shows how much steel the auto industry purchases. The cell where “steel” meets “construction” shows how much steel the construction industry uses. Every industry appears twice: once as a seller (in its row) and once as a buyer (in its column).
The table is typically divided into distinct sections, or quadrants. The first quadrant is the inter-industry core. It shows intermediate inputs: the goods and services industries buy from each other as part of their production process. If you read a single column in this quadrant, you see everything that industry consumes from other industries and from itself. Read a single row, and you see how that industry’s output is distributed across all the others.
The second quadrant extends each row outward to show final demand. This is output that doesn’t go to another industry for further processing. It goes to households, government, exports, or investment. The third quadrant sits below the inter-industry core and records primary inputs to production: worker compensation, business profits, imports, and taxes. These are the costs that don’t come from other domestic industries. A fourth quadrant, less commonly discussed, maps primary inputs to final demand categories.
The Math Behind It
The economist Wassily Leontief developed the mathematical framework for input-output analysis, earning a Nobel Prize for it in 1973. The core idea is captured in a simple equation: total output equals internal consumption plus consumer demand. In notation, that’s X = AX + D, where X is total output, A is the matrix of input-output coefficients, and D is final demand from consumers, government, and other end users.
The matrix A contains what are called technical coefficients. Each number represents how much input from one industry is needed to produce one unit of output in another. For example, if producing one unit of agricultural products requires 0.2 units of manufactured goods and 0.1 units of energy, those fractions sit in the corresponding cells of the A matrix. These coefficients make it possible to calculate exactly how much every industry needs to ramp up when demand changes somewhere in the economy.
What Multipliers Tell You
One of the most powerful tools derived from input-output tables is the output multiplier. When an industry increases production, it buys more inputs from its suppliers, who in turn buy more from their suppliers, and so on. This cascading effect is called a backward linkage. The multiplier captures the total economic output generated across all industries from one additional unit of final demand in a single industry.
These multipliers are not fixed. They shift as economies evolve. Bureau of Economic Analysis data shows that U.S. manufacturing multipliers declined between 1972 and 1996. Fast-growing manufacturing sectors had a multiplier of 2.004 in 1972, meaning one dollar of final demand for their products generated roughly two dollars of total economic output. By 1996, that figure had dropped to 1.853. Slow-growing manufacturing went from 2.205 to 2.149 over the same period. The decline reflects a structural shift: manufacturing became less reliant on domestic suppliers, so a spike in demand for manufactured goods generated less activity in the rest of the domestic economy than it once did.
There’s also a forward linkage, which works in the opposite direction. It measures how much additional output an industry must supply when other industries expand. Industries with high forward linkages, like energy or transportation, are deeply embedded as suppliers throughout the economy.
Who Uses These Tables and Why
Governments are the primary producers and consumers of input-output data. National statistical agencies build the tables as part of their system of national accounts. In the U.S., the Bureau of Economic Analysis maintains input-output accounts and regional impact models (known as RIMS II) that state and local planners use to estimate the economic effects of specific events. A city council considering a new baseball stadium, for instance, can use these models to project how much additional economic activity the project would generate across local industries. The same tools help estimate the fallout from a factory closure or a natural disaster.
The OECD maintains an Inter-Country Input-Output database covering 80 economies, which links national tables together so analysts can trace production relationships across borders. This is especially useful for understanding global supply chains, where raw materials from one country become components in another and finished products in a third.
Businesses and policymakers use the tables for tax analysis, regulatory impact assessments, and strategic planning. Because the tables reveal the full chain of suppliers behind any product, they’re a natural fit for understanding how policy changes or economic shocks propagate.
Tracking Environmental Impact
A major modern application extends input-output tables to measure environmental footprints. The U.S. Environmental Protection Agency developed what it calls Environmentally-Extended Input-Output models, which layer environmental data on top of the economic transaction data. The EPA’s version covers 389 industry sectors and tracks land use, water consumption, energy use, mineral extraction, air pollution, nutrient runoff, and toxic releases.
The logic is straightforward. If you know the economic inputs needed to produce a product, and you know the emissions associated with each of those inputs, you can calculate the total environmental cost of that product across its entire supply chain. Amazon has used these models to estimate the carbon footprint of purchased goods and services. General Motors worked with consultants to calculate its supply chain greenhouse gas emissions using the same framework. The EPA also publishes a comprehensive set of supply chain emission factors for every category of goods and services in the U.S. economy, all derived from these extended tables.
How Input-Output Tables Relate to Other Frameworks
Input-output tables are one piece of a larger family of economic accounting tools. Supply and use tables are a closely related format. A supply table shows what each industry produces and what gets imported. A use table shows which products each industry consumes and where final demand goes. National statistical agencies often compile supply and use tables first, then mathematically transform them into symmetric input-output tables where industries (or products) appear on both axes.
A social accounting matrix, or SAM, takes the concept further. While an input-output table focuses on production relationships between industries, a SAM captures the entire sequence of economic accounts: not just who buys from whom in production, but also how income flows to households, how households spend, how government collects and distributes revenue, and how savings become investment. The standard understanding of a SAM often includes extra detail on the household sector, breaking it into income groups to analyze how economic changes affect different populations. Think of the input-output table as the production engine at the center, and the SAM as the full circuit of money flowing through the economy.
The international standards governing all of these tools fall under the System of National Accounts, most recently updated in its 2025 edition. Extended versions of supply and use tables now incorporate an additional dimension for the type of enterprise, allowing analysts to distinguish between, say, domestic firms and foreign-owned subsidiaries operating in the same industry.