Skip to content
Back to search results

Social Accounting Matrix

Introduced by
Richard Stone; popularized by Graham Pyatt & Jeffery Round
Year
1962
Type
Comprehensive, square, double-entry accounting framework
Family
National accounting & multiplier analysis
Font size

What does it do?

A social accounting matrix (SAM) is a square, double-entry table that records all transactions among the production sectors, factors of production, institutions (households, firms, government), and the rest of the world in an economy for a given year.

It extends the input-output table by closing the circular flow of income — connecting how value added becomes factor income, factor income becomes household income, and household income becomes demand — so that every account's receipts (its row) exactly equal its expenditures (its column).

The intuition behind the method

An input-output table shows how industries buy from one another but stops at value added; it never tells you where the wages and profits go. A SAM completes the loop: wages flow to households, households spend on goods, that spending becomes demand for industries, which generates more value added, and round it goes. Because the matrix is square and every row total equals its column total, the SAM is a complete, internally consistent snapshot of who pays whom in the whole economy — and, once you decide which accounts respond endogenously, inverting it yields multipliers that trace an injection all the way around the income circuit.

History

The social accounting matrix grew out of Richard Stone's work on national accounts at Cambridge in the late 1950s and early 1960s, where he formalized the economy as a square, double-entry matrix as part of the Cambridge Growth Project; Stone received the 1984 Nobel Memorial Prize partly for this contribution. Graham Pyatt and Jeffery Round, together with collaborators at the World Bank and the University of Warwick, developed the SAM into a practical tool for development planning and distributional analysis, codified in their 1985 World Bank volume. SAMs subsequently became the foundational database for computable general equilibrium modeling worldwide.

When to use it

Use a SAM when you need an economy-wide, internally consistent picture of income generation and distribution — not just production linkages — and you want to trace how an exogenous injection (export demand, government transfer, foreign aid, a new investment) affects sectoral output, factor incomes, and household groups including induced consumption effects. SAMs are the standard tool for distributional impact analysis and the mandatory database for calibrating computable general equilibrium models. SAM multiplier analysis assumes fixed average expenditure propensities and excess capacity, so like input-output analysis it is a short-run, demand-driven, price-free model; when prices and substitution matter, the SAM should instead be used to calibrate a CGE model.

Limitations and considerations

  • Multiplier analysis assumes fixed average expenditure propensities and idle resources, so prices and substitution are absent.
  • Constructing and balancing a SAM is data-intensive and requires reconciling sources of differing quality, often via estimation techniques.
  • Results are sensitive to which accounts are designated endogenous versus exogenous, a modeling choice that changes the multipliers.
  • Like input-output, it is a base-year snapshot and becomes outdated as economic structure shifts.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Closes the circular flow of income, capturing induced consumption effects that input-output multipliers omit.
  • Disaggregates households and factors, enabling explicit analysis of income distribution and incidence by socioeconomic group.
  • Imposes double-entry balance, forcing reconciliation of disparate data sources into one consistent framework.
  • Serves as the standard calibration database for computable general equilibrium models, bridging accounting and behavioral analysis.

Applications

  • Analyzing the distributional impact of policies across household and socioeconomic groups, including poverty and inequality effects.
  • Calibrating and benchmarking computable general equilibrium models, for which a balanced SAM is the required input dataset.
  • Estimating economy-wide multiplier effects of foreign aid, remittances, public transfers, or export demand including induced consumption.
  • Development planning and structural analysis in low- and middle-income economies where income distribution is a central concern.
  • Constructing environmentally and socially extended accounts that link production, incomes, and resource use.

How it works

How it works

1

The defining property of a SAM is that it is square and balanced: each account appears as both a row (receipts) and a column (expenditures), and the row total must equal the column total. This double-entry discipline guarantees internal consistency and forces the analyst to reconcile data from national accounts, household surveys, and trade statistics into one coherent table.

Formula
\sum_{j} t_{ij} = \sum_{k} t_{ki} \quad \text{(row total} = \text{column total for account } i)
2

To build a multiplier model, accounts are partitioned into endogenous ones (typically production activities, commodities, factors, and households) whose income responds within the model, and exogenous ones (typically government, capital, and the rest of the world) that provide injections x. Endogenous incomes y_n satisfy y_n = N y_n + x, where N captures how endogenous receipts are spent on other endogenous accounts.

Formula
y_n = N\,y_n + x = (A_n y_n) + x
3

Each cell t_ij is divided by the total income y_j of the spending account to give an average expenditure propensity a_ij — the share of account j's income that flows to account i. Collected into the matrix N, these propensities play the same role as technical coefficients in input-output analysis but cover the full income circuit, not just inter-industry use.

Formula
a_{ij} = \frac{t_{ij}}{y_j}, \qquad N = [a_{ij}]
4

Solving the endogenous system gives y_n = (I − N)^{-1} x, where the accounting multiplier matrix M_a = (I − N)^{-1} maps an exogenous injection into total endogenous incomes. Unlike the Leontief inverse, it captures not only inter-industry effects but also the induced consumption effects that arise when factor and household incomes recirculate as new demand.

Formula
y_n = (I - N)^{-1} x = M_a\, x

Common pitfalls

  • Confusing SAM multipliers with input-output multipliers — SAM multipliers are larger because they include induced income effects, which can lead to double counting if mixed.
  • Leaving the matrix unbalanced and forcing closure mechanically (e.g., a single residual cell) instead of using a principled balancing method such as cross-entropy or RAS.
  • Interpreting fixed-propensity multipliers as valid for large shocks, where the linearity and excess-capacity assumptions break down.
  • Aggregating households too coarsely, which hides the distributional conclusions that are a SAM's main advantage.

Frequently asked questions

How does a SAM differ from an input-output table?

An input-output table records only inter-industry flows and stops at value added. A SAM is square and balanced and additionally records how value added becomes factor income, how factor income becomes institutional (household, firm, government) income, and how that income is spent — closing the circular flow. As a result SAM multipliers include induced consumption effects that input-output multipliers, which treat household income as exogenous, leave out.

What does it mean to balance a SAM, and how is it done?

Balancing means making every account's row total (receipts) equal its column total (expenditures), as double-entry accounting requires. Raw data assembled from national accounts, surveys, and trade statistics rarely balance, so analysts use estimation procedures — most commonly the RAS biproportional method or cross-entropy minimization — to adjust the cells minimally while satisfying all row and column constraints.

Why is a SAM needed to build a CGE model?

A computable general equilibrium model requires a consistent benchmark dataset of all transactions among agents in a base year to calibrate its parameters and ensure the model replicates the observed economy at the starting point. A balanced SAM provides exactly this: it is the empirical accounting skeleton on which the CGE's behavioral equations (production functions, demand systems, market clearing) are hung.

Related methods

Sources

  1. Pyatt, G., & Round, J. I. (Eds.). (1985). Social Accounting Matrices: A Basis for Planning. Washington, DC: The World Bank.

    ISBN: 9780821305508
    Copied.
  2. Stone, R. (1962). A Social Accounting Matrix for 1960. A Programme for Growth, Vol. 2. London: Chapman and Hall.

    ISBN: 9780412208300
    Copied.