views updated





Marginalism is one of two types of approaches to studying economic and political processes. Each is both a technique of analysis and a view of how both individual economic agents and the economy as a whole function in reacting to change and in causing change, and therefore policy. Marginalism, or incrementalism, gives effect to the methodological individualist perceptions that the meaning of human action is best comprehended at the level of individuals, that change takes place along various margins and not in totals, that decision making/policy is made in a composite manner, as the sum of the results of independent but interacting agents and institutions.

The system comprising all of such agents and institutions moves as a result of marginal adjustments to changes originating within and outside the domain of each. Government, for example, is no single decisional unit; decision making is incrementally developed by a multiplicity of agencies. Systems analysis, on the other hand, gives effect to the methodologic collectivist perception that the meaning of any marginal adjustment, agent or institution is a function of the total system of which it is a part. Systems analysis thus facilitates the study of the totality of interacting agents and institutions, therefore of group processes, and how agents and institutions generate and respond to large-scale social forces, problems, and so on. The systems approach to policymaking centers on the study of the economic and political structures and processes within which individual behavior takes place, just as the incrementalist or marginalist approach focuses on the individual behavior putatively independent of systemic structures, processes, and forces. Uniting the two types of approach enables the combined study of the impacts of system upon individuals and of individuals upon system. Moreover, no single mode exists for modeling either parts or wholes, incremental or systemic change. The variety of schools of thought operating in each approach is enormous; and the range of models runs from the single individual to the world system.

Marginalism gives effect to important aspects of change. One aspect is that a change in a variable is not likely to be a change in its totality; that is, in all of it, but a small increase or a small decrease here and/or there, along one margin but not another. For example, a typical change of tastes does not involve now buying fish and no longer buying pork, but a little more of one and a little less of the other. The change is said to take place at the margin. Another aspect is the impact of a change in one variable at its margin on another variable at its margin. The increase or decrease in purchases of fish may lead to a change in employment in the same direction, and vice versa for pork. The concept of elasticity, for example, relates proportionate changes in quantity to proportionate changes in price or income. A change in aggregate consumption spending will change income in the same direction and may also change investment in the same direction; but not every element comprising the total need change.

Another aspect of marginalism has to do with the tendency of an economy, or parts of it, to adjust so as to reach equilibrium. When consumption increases, leading to an increase in income, consumption further increases, though generally by an amount smaller than the initial increase, and investment also increases; eventually the increases in spending diminish to zero, which will be when spending equals income for one period and income equals spending for the next period; the economy is in equilibrium. As for that increase in investment, the increase in consumption and, with it, income tend to increase the expected rate of profit (marginal efficiency of capital, or MEC) not on all investment but on new investment; that is, investment at the margin; in the example, investment increases as the MEC increases relative to the interest rate, a proxy for alternative uses of money.

Still another aspect of marginalism is its use in optimization. If one defines optimization as when benefit equals cost, then an individual will optimize a line of expenditure at the point of purchase when the additional benefit of an increase in spending equals the cost so that there is no incentive to change spending. A firm will optimize its level of production when the marginal cost of a level of production equals the marginal revenue derived from its sale and when the marginal revenues of different periods and in different markets are equal. An agent is in both equimarginal equilibrium and optimization, when the last dollar spent on each line of spending equals the benefit derived from it, so there is no incentive to change the pattern of spending, given the pattern of tastes. In this simple model, equilibrium and optimization coincide. But, many production processes have increasing returns (marginal cost of another physical unit decreasing). If the good is priced equal to its marginal cost, the total production costs cannot be recovered. Pricing necessarily is an institutional matter of industry cost accounting and other conventions.


Marginal analysis in economics originated in two initial forms. The English economist David Ricardos theory of rent involved decreasing marginal returns in the form of food from additional increases in inputs so that rent was a differential. The returns are still positive, but decreasing, until they reach zero, eliciting no further increases in inputs. In the English jurist and philosopher Jeremy Benthams utilitarianism, an activity is characterized by pain and pleasure, such that an activity will be conducted up so long as the pleasure exceeds the pain (perceived benefits are greater than perceived costs), to the point that the two are equal in magnitude. Whereas Ricardos theory involves physical inputs and outputs, Benthams theory involves subjective feelings of pain and pleasure.

In the late nineteenth century, albeit with many precursors, several economists used marginalism to construct a theory of value different from cost of production. In their view, the value of anything that is bought and sold is at the point where the marginal (last) buyer and seller agree as to the price of purchase and sale. The buyer derives marginal utility from additional acquisitions of the item. The level of marginal utility decreases as more is acquired. Tracking falling marginal utility is falling demand: Because marginal utility is decreasing, the individual buyer will purchase more only at lower prices. Sellers will sell more only if they receive more, due to increasing costs. This is the subjectivist, or Austrian, theory of value in which value is a function of scarcity, as opportunity cost meets subjective demands in the market.

All aspects of marginalist change have proven to be useful either as conjectural explanation, actual explanation, and/or a tool of analysis of how changes in variables occur and their consequences. In many interesting and important respects, the world changes at the margin.


Marginalism, like all concepts and theories in economics, has its limits. One class of limits has to do with the identification of a margin. If one is going to make marginal comparisons, the arrays of margins must be specified. In some respects this is a matter of the physical monetary activity itself. But in other respects it depends on subjective identifications of the lines along which marginal change will be measured. This will depend in part on how firms and households are organized, on the cost accounting conventions adopted by firms and how they are applied, on legal requirements, and so on. The analysis of consumption usually centers on particular commodities or groups of commodities and may include attention to marginal adjustments between complementary goods and between substitutable goods. The analysis of production, undertaken by the typical multi-product firm, is more complicated. Here the assignment of costs between fixed and variable costs can take place along a variety of different identifications of margins and with numbers further influenced by the identification of products and product groups and of block levels of production, instead of single units, in the construction of cost accounting. For nonrival goods, marginal cost of another user is equal to zero. If users are to be charged the cost of the unit they use, some could pay nothing, while others pay some share of the fixed cost of production. Who is designated the marginal user is an institutional matter. Thus, if marginal productivity is the incremental increase (decrease) in total output associated with an increase (decrease) in input by one unit (which may be a block of units), the calculation is dependent on the definitions of product line, input, output, and margin.

Another class of limits involves what some economists perceive to be a superficial rendering of human activity. The question is whether people act as they are portrayed in theory, whether the maximizing or optimizing behavior is more complex than, perhaps even different from, that described in marginalist accounts, whether tastes are given or stable, and so on.

A different class of limits involves the larger system of things in which changemarginal changetakes place. If people make marginal adjustments in their buying and selling and other economic activity, the question involved is, in part, what is the market in which this takes place and how is it formed. Markets are created as firms and governments exercise strategic behavior and attempt to form and structure markets to accomplish their goals. Beyond markets alone, the larger system involves the overall structure of power, the legal-economic nexus, the set of macroeconomic policies pursued by governments, and so on, in which marginal adjustments take place and have effects.

Closely related limitations derive from the specification of the unit of analysis. The unit may be a representative firm, a particular industry, or an element of a particular country or of a particular region, as well as the world system. These and other limits apply to both marginal and systems analysis.

Also closely related is the evolutionary dynamics of the larger economic system, a process of change which manifests in changes at the margin but also, and more importantly, swamps marginal adjustments. Further limitations are imposed by theories centering on asymmetric information, expectations, uncertainty, strategic behavior, corporate goal formation, bounded information, and so on. In practice, the various strengths and limits are trumped by the research interests of particular economists and their respective schools of thought.

SEE ALSO Economics, Neoclassical; General Equilibrium; Substitutability; Trade-offs


Easton, David. 1965. A Systems Analysis of Political Life. New York: Wiley.

Elster, Jon. 1989. Nuts and Bolts for the Social Sciences. Cambridge, U.K.: Cambridge University Press.

Lindblom, Charles E. 1958. Policy Analysis. American Economic Review 48 (June): 298312.

Marshall, Alfred. 1890. Principles of Economics. London: Macmillan.

Robinson, Joan. 1933. The Economics of Imperfect Competition. London: Macmillan.

Samuels, Warren J. 1994. Part-Whole Relationships. In The Elgar Companion to Institutional and Evolutionary Economics, Vol. 2, ed. Geoffrey M. Hodgson, Warren J. Samuels, and Mark R. Tool, 142146. Brookfield, VT: Edward Elgar.

Veblen, Thorstein. 1997. Why Is Economics Not an Evolutionary Science? In Classics in Institutional Economics, The Founders, 18901945 Vol. 1, ed. Malcolm Rutherford and Warren J. Samuels, 227. London: Pickering and Chatto.

Warren J. Samuels