The margin between costs of production and target sale price will be set by the degree of product differentiation within the industry, compet.i.tive pressures and general market conditions. Each firm will endeavor to sell as much output as it can, but the level of output will be constrained by the size of the firm's market niche and the marketing efforts of rivals.

5.14 A graphical representation of Sraffa's (1926) preferred model of the normal firm So what?

To the non-economist, Sraffa's conclusions might still look like fairly minor points. The supply curve should be horizontal rather than upward sloping; the output of an individual firm isn't set by the intersection of marginal revenue and marginal cost; and marketing and finance issues, rather than cost of production issues, determine the maximum scale of a firm's output. This is a big deal?

Strange as it may seem, yes, this is a very big deal. If marginal returns are constant rather than falling, then the neocla.s.sical explanation of everything collapses. Not only can economic theory no longer explain how much a firm produces, it can explain nothing else either.

Take, for example, the economic theory of employment and wage determination (discussed in more detail in Chapter 6). The theory a.s.serts that the real wage is equivalent to the marginal product of labor. The argument goes that each employer takes the wage level as given, since with compet.i.tive markets no employer can affect the price of his inputs. An employer will employ an additional worker if the amount the worker adds to output the worker's marginal product exceeds the real wage. The employer stops employing workers once the marginal product of the last one employed has fallen to the same level as the real wage.

This explains the economic predilection for blaming everything on wages being too high neocla.s.sical economics can be summed up, as Galbraith once remarked, in the twin propositions that the poor don't work hard enough because they're paid too much, and the rich don't work hard enough because they're not paid enough (Galbraith 1997). The output of the firm is subject to diminishing marginal returns, and thus marginal product falls as output increases. The real wage is unaffected by the output level of the firm. The firm will keep on hiring workers until the marginal product of the last worker equals the real wage.

Since the rational employer stops at that point, the real wage which the employer takes as given determines how many workers this firm employs. Since employment in turn determines output, the real wage determines the level of output. If society desires a higher level of employment and output, then the only way to get it is to reduce the real wage (and the logical limit of this argument is that output will reach its maximum when the real wage equals zero). The real wage in turn is determined by the willingness of workers to work to forgo leisure for income so that ultimately the level of employment is determined by workers alone.

5.15 The economic theory of income distribution argues that the wage equals the marginal product of labor If in fact the output-to-employment relationship is relatively constant, then the neocla.s.sical explanation for employment and output determination collapses. With a flat production function, the marginal product of labor will be constant, and it will never intersect the real wage. The output of the firm then can't be explained by the cost of employing labor, and neocla.s.sical economics simply explains nothing: neither the level of employment, nor output, nor, ultimately, what determines the real wage.

5.16 Economics has no explanation of wage determination or anything else with constant returns Sraffa's critique is thus a fundamental one: if his argument is accepted then the entire edifice of economics collapses.

Clearly, no such thing has happened: economics has continued on as if Sraffa's article was never even published. One might hope that this is because there is some fundamental flaw in Sraffa's argument, or because there is some deeper truth that neocla.s.sical economics discovered to justify preserving the old model with a new explanation. Sadly, neither is the case.

The neocla.s.sical rejoinder.

Sraffa's paper evoked several responses from the economic heavyweights of the time. However, these focused upon another aspect of the paper, his critique of the notion of external economies of scale in the long run. Sraffa's primary argument, that the concept of diminishing marginal returns in the short run is invalid, was ignored so much so that in 1927, 1928 and 1930, Pigou, Robbins and Harrod respectively set out the theory of short-run price determination by rising marginal cost in complete confidence of its validity, and without any reference to Sraffa's paper. Few, if any, conventional economists have since referred to Sraffa's paper.

There are many possible reasons for this complete neglect of a serious challenge to economic orthodoxy. The simplest explanation is that the argument was ignored because its implications, if accepted, were too destructive of conventional economics for neocla.s.sical economists to contemplate. As Chapter 7 argues, this is a not uncommon initial response in all sciences the key difference with economic 'science' being that this can also be the final response. However, it must be acknowledged that even Keynes who was, like Sraffa, critical of the mainstream failed to realize the import of Sraffa's arguments.

The situation has not improved with time. Sraffa's paper is today cited only by critics of economic orthodoxy, while the textbooks teach the theory of rising marginal cost without reference to Sraffa's counter-arguments. It is therefore difficult to put forward a neocla.s.sical response to Sraffa. However, many economists put forward the following arguments when they are informed of Sraffa's paper.

The first is that Sraffa has completely failed to understand the concept of the short run. Neocla.s.sical economics defines three concepts of time: the market period, during which no factor of production can be varied, so that supply is fixed and only price can vary; the short run, during which at least one factor of production cannot be varied, so that output can be varied but only at the cost of diminishing returns; and the long run, during which all inputs can be varied. Since production takes place during the short run, the remainder of the theory follows logically. Diminishing marginal returns will apply, marginal cost will rise, price and quant.i.ty will be jointly determined by supply and demand, and the entire edifice of the theory of production and distribution remains intact.

The second is that Sraffa misunderstands the nature of production in a capitalist economy. Since there is enormous pressure to be compet.i.tive, no firm can survive long with excess capacity. Therefore compet.i.tion will drive all firms towards full capacity, and in this realm diminishing returns will apply.

Time and the short run.

As Chapter 9 points out, time or rather, the absence of time in its a.n.a.lysis is one of the fundamental weaknesses of conventional economics. It is therefore somewhat ironic that economists defend their theory from attack by appealing to the importance of time. However, far from helping to defend economic theory from criticism, the proper a.n.a.lysis of time highlights a critical weakness.

A firm's revenue and costs clearly vary over time, as well as varying as the firm changes its level of output at any one point in time. The economic rule that (in the context of diminishing returns) 'profit is maximized where marginal cost equals marginal revenue' is derived by 'holding time constant' and thus describing revenue and cost as simply a function of the quant.i.ty produced. The gap between revenue and cost is widest where marginal cost equals marginal revenue.

But in fact this rule applies only 'when time stands still' which time never does. Not even an economist can make time stand still (though some victims of economics lectures might dispute that!). Similarly, the rule tells you how to maximize profit with respect to quant.i.ty, but real businessmen are more interested in maximizing profit over both time and output.

It is possible to consider profit as a function of both time and quant.i.ty, as opposed to the economic approach of dividing time into artificial segments, by explicitly acknowledging that profit is a function of both time and quant.i.ty (which the firm can vary at any point in time, and that will also change and hopefully grow over time).8 Profit therefore depends both on the amount a firm produces, and the historical time during which it produces.

Using a rule of mathematics, we can then decompose the change in profit into the contribution due to the progress of time, and the contribution due to changes in quant.i.ty (which will also change over time). This results in the formula: Change in profit equals change in profit due to change in time multiplied by the change in time, plus change in profit due to change in quant.i.ty multiplied by the change in quant.i.ty.

This formula tells us how big a change in profit will be, so if a firm wants to maximize its profit, it wants this number to be as big as possible.

Change in profit due to change in quant.i.ty is the same thing as 'marginal revenue minus marginal cost.' Neocla.s.sical theory argues that profit is maximized when marginal revenue equals marginal cost which we already know is a fallacy but if you followed the neocla.s.sical profit maximization rule here, you would deliberately set this quant.i.ty to zero. Since you get zero when you multiply any number by zero, following this rule sets the second half of the formula (change in profit due to change in quant.i.ty multiplied by the change in quant.i.ty) to zero.

Therefore, economic theory tells us that the change in profit will be maximized when we eliminate the contribution that changes in quant.i.ty make to changes in profit. Change in profit is thus reduced simply to the first half of the formula, where changes due to time alone determine the change in profit. But economic theory has given us no advice about how to make change in profit due to change in time as big as possible.

What's going on? Suddenly, advice that previously seemed sensible (before we considered Sraffa's critique of the notion of a fixed factor of production) looks obviously absurd. Clearly something is wrong: but what?

An a.n.a.logy might help you interpret it. Imagine you have a formula which describes how much fuel your car uses at any given speed, and you want to work out the most economical speed at which to drive. What you need to do is to work out the lowest rate at which to consume petrol per unit of distance traveled per second. If instead you first work out the most economical speed at which to travel, the answer to this first question will be zero miles per hour because at this speed you consume the lowest possible amount of petrol per unit of time, zero. This is an accurate but useless answer, since you're not interested in staying put. If you want to work out the speed that minimizes petrol consumed but still gets you to your destination, you have to handle both problems simultaneously.

The neocla.s.sical theory of the firm ignores time, in the same way that the wrong answer to the 'most economical speed at which to travel' question ignores distance. But time is an essential aspect of economic behavior, in the same sense that distance is an essential aspect of travel. The neocla.s.sical policy for profit maximization is thus false twice: first, it ignores the impact of other firms in an industry on your profit, as the previous chapter showed; then it ignores time. It is thus a kindred spirit to the advice that the cheapest way to get from point A to point B is to travel at zero miles per hour.

There is also an economic way of interpreting this apparent paradox: that advice which appears sound when you ignore (or compartmentalize) time becomes absurd when you take time into account. This is that, by ignoring time in its a.n.a.lysis of the firm, economic theory ignores some of the most important issues facing a firm. Its 'static' emphasis upon maximizing profit 'now' ignores the fact that, to survive, a firm must also grow over time. To grow it must invest and develop new products, and this takes energy and resources. If instead it devotes all its resources to maximizing profit now, then it will not have any energy or resources left to devote to investment and new product development.

If we try to interpret economic theory in the context of historical time, then what the theory is attempting to do is work out the ideal level of output of a product for all time. But in the real world there is no such level of output. The appropriate number of motor cars to produce in 1900 was quite different from the appropriate number to produce in 2000.

This is how something that once looked so right (before Sraffa's critique of the concept of a fixed factor of production) looks so absurd now. The formula discussed above explicitly takes time into account, and is therefore dynamic, while the economic theory of the firm is static: it ignores time, and is therefore only relevant in a world in which time does not matter. But time clearly does matter in our world, and what is right in a static setting is wrong in a dynamic one.

Let's go back to that formula, which is true by definition, and see what it tells us to do.

If the firm's output is growing over time, then the term change in quant.i.ty will be positive. Setting marginal revenue equal to marginal cost means multiplying this positive number by zero which results in a smaller increase in profit than if marginal revenue exceeds marginal cost. With rising sales, you will get a higher profit if 'change in profit due to change in quant.i.ty' is also positive, which requires that marginal revenue be greater than marginal cost. Thus a careful consideration of time argues that a firm should ensure that its marginal revenue is greater than its marginal cost.

This is the position which Sraffa argued actually applies in reality, so that the proper consideration of time strengthens Sraffa's critique, rather than weakening it. It also strengthens the point I made in the previous chapter that the neocla.s.sical 'short-run profit maximization' formula is false; it's false in the long run too.

This is one of the many instances of the phenomenon I mentioned in Chapter 1, that advice derived from static reasoning, which ignores time, is often categorically opposed to advice derived from dynamic a.n.a.lysis, which takes time into account. Since the economy is fundamentally dynamic, static a.n.a.lysis is therefore normally dangerously wrong. I explore these issues in more depth in Chapter 8.

The flaws in economic reasoning pointed out in this chapter and Chapter 4 have a very direct impact on public policy in the area of the pricing of public services. Because economists believe that compet.i.tive industries set price equal to marginal cost, economists normally pressure public utilities to price their services at 'marginal cost.' Since the marginal costs of production are normally constant and well below the average costs, this policy will normally result in public utilities making a loss. This is likely to mean that public utilities are not able to finance the investment they need in order to maintain the quality of services over time. This dilemma in turn interacts with the pressure that economists also apply to privatize public a.s.sets, and to let individuals 'opt out' of the public provision of essential services. The end result, as Galbraith so eloquently put it, is 'private affluence and public squalor.'

TABLE 5.2 Cost drawings for the survey by Eiteman and Guthrie (1952: 8345) Ironically, economic theory also makes economists essentially 'anti-capitalist,' in that they deride real businesses for pricing by a markup on cost, when theory tells them that prices should be set at the much lower level of marginal cost. Industrialists who have to cope with these att.i.tudes in their dealings with government-employed economists are often among the greatest closet anti-economists of all. Maybe it's time for them to come out of the closet.

Compet.i.tion and capacity The argument that compet.i.tion would drive all firms to use their fixed capital at full capacity does look convincing at first glance, but deeper thought reaches precisely the opposite conclusion. A firm with no spare capacity has no flexibility to take advantage of sudden, unexpected changes in the market, and it also has to consider building a new factory as soon as its output grows. Excess capacity is essential for survival in a market economy.

Wrong in fact as well as theory.

Sraffa's critique was entirely based upon an appeal to logic; a defence which might appear to be open to economic theory was, of course, that the facts supported them rather than Sraffa. However, over 150 empirical studies have been conducted into what the costs of actual firms really are, and with rare unanimity, every last one of them has found that the vast majority of firms report that they have very large fixed costs, and either constant or falling marginal costs, so that average costs of production fall as output rises.

One of the most interesting such studies showed factory managers eight drawings of the shape of cost curves, only three of which bore any resemblance to the standard drawings in neocla.s.sical textbooks (see Table 5.2).

When asked to choose which drawings most closely resembled the relationship between cost and output levels in their factories, only one of the 334 companies chose curve 3, the one that looks most like the curve drawn in virtually every neocla.s.sical microeconomics textbook for example, this one in Figure 5.17 from Varian's Microeconomic a.n.a.lysis (Varian 1992: 68) while another seventeen chose curves that looked something like it.

Ninety-five percent of the managers chose drawings that did not conform to the standard textbook model, but instead ill.u.s.trated either constant or falling marginal costs (Eiteman and Guthrie 1952: 837).

Predictably, neocla.s.sical economists ignored this empirical research and in fact the purpose of one of the most famous papers in economics, Milton Friedman's 'as if' paper on methodology (discussed in Chapter 8), was to encourage economists to ignore these empirical results.9 5.17 Varian's drawing of cost curves in his 'advanced' microeconomics textbook This practice of ignoring empirical research continues today, even though the most recent researcher to rediscover these results was not a critic of neocla.s.sical economics, but one-time vice-president of the American Economic a.s.sociation and vice-chairman of the Federal Reserve, Alan Blinder (Blinder 1982, 1998). Blinder surveyed 200 medium-to-large US firms, which collectively accounted for 7.6 percent of America's GDP, and he put his results with beguiling honesty: The overwhelmingly bad news here (for economic theory) is that, apparently, only 11 percent of GDP is produced under conditions of rising marginal cost [...]

Firms report having very high fixed costs roughly 40 percent of total costs on average. And many more companies state that they have falling, rather than rising, marginal cost curves. While there are reasons to wonder whether respondents interpreted these questions about costs correctly, their answers paint an image of the cost structure of the typical firm that is very different from the one immortalized in textbooks. (Blinder 1998: 102, 105; emphases added) The neocla.s.sical model of the u-shaped average cost curve and rising marginal cost is thus wrong in theory and wrong in fact. That it is still taught as gospel to students of economics at all levels of instruction, and believed by the vast majority of neocla.s.sical economists, is one of the best pieces of evidence of how truly unscientific economics is.

TABLE 5.3 Empirical research on the nature of cost curves (summarizing Table 4 in Eiteman and Guthrie 1952: 838) A totem in tatters.

While the neocla.s.sical model of why production costs rise with output is thus fallacious, it is still feasible that, in some instances, price will rise as output rises. Feasible 'real-world' reasons for this include the inflexibility of supply in some markets across some timeframes (something economics attempts to deal with by its concept of the market period, as opposed to the short-run theory debunked in this chapter), firms exploiting high demand to set higher margins, and, in some circ.u.mstances, wage demands rising during periods of high employment.10 But the neocla.s.sical attempt to link higher prices directly to declining productivity is a failure.

This of itself would not be catastrophic, were it not for the extent to which diminishing marginal productivity permeates neocla.s.sical economics. It is a foundation stone which, when it is withdrawn, brings down virtually everything else with it. Sraffa's critique thus provides one more ill.u.s.tration of the remarkable fragility of this outwardly confident social theory we call economics. Economics is not the emperor of the social sciences, but the Humpty Dumpty.

Just as with Humpty Dumpty after his fall, it is impossible to reconstruct the totemic supply and demand diagram after the criticisms outlined in this and the preceding chapters. First, the Sonnenschein-Mantel-Debreu conditions show that 'diminishing marginal utility,' which in theory applies at the individual level and means that individual demand curves slope downwards, doesn't survive aggregation to the market level so that a market demand curve can have any shape at all (apart from doubling back on itself, or intersecting itself). Secondly, the marginal revenue curve derived from this demand curve will be even more unstable. Thirdly, equating marginal revenue and marginal cost isn't profit-maximizing. Finally, diminishing marginal productivity is a theoretical and empirical fallacy, so that for most factories, marginal cost is either constant or falling.

Taken together, these critiques eliminate the 'Totem of the Micro' completely. Virtually every concept in neocla.s.sical microeconomics depends on diminishing marginal productivity for firms on the one hand, and diminishing marginal utility for the community on the other. If both these foundations are unsound, then almost nothing else remains standing. Without diminishing marginal productivity, neocla.s.sical economists cannot explain how a firm decides how much to produce. This alone invalidates their a.n.a.lysis of market structures and income distribution. Without a community utility map, everything from the a.n.a.lysis of optimum output levels to the theory of international trade collapses.

Yet still they teach the standard mantra to their students, and still they apply the same simplistic logic to many other areas of economics.

In the chapters to come, we will temporarily 'forget' the criticisms of these fundamental building blocks, and examine the validity of neocla.s.sical theory as it is applied to specific issues. As you will see, even if we allow, for the sake of argument, that demand falls smoothly as price rises, that production is subject to diminishing marginal returns, and that demand and supply set prices, the neocla.s.sical theories of the distribution of income, the behavior of the macroeconomy, and the role of finance are all intellectually unsound.

6 | TO EACH ACCORDING TO HIS CONTRIBUTION.

Why productivity doesn't determine wages.

One of the most striking aspects of the late twentieth century was the increase in the gap between the poorest worker and the richest. While many bemoaned this increase in inequality, economists counseled that the growing gap merely reflected the rising productivity of the highly paid.

The basis for this advice is the proposition that a person's income is determined by his contribution to production or more precisely, by the marginal productivity of the 'factor of production' to which he contributes. Wages and profits or 'factor incomes,' as economists prefer to call them reflect respectively the marginal product of labor and of capital. The argument that highly paid workers merchant bankers, managers of major corporations, stock and money market traders, financial commentators, etc. deserve the high wages they receive compared to the less highly paid nuclear physicists, rocket scientists, university professors, schoolteachers, social workers, nurses, factory workers, etc. is simply an extension of this argument to cover subgroups of workers. Members of the former group, we are told, are simply more productive than members of the latter, hence their higher salaries.

I'll defer discussion of the proposition that profits reflect the marginal productivity of capital until the next chapter. Here we'll consider the argument that wages equal the marginal product of labor.

Once again, the argument relies heavily on concepts we have already dismissed: that productivity per worker falls as more workers are hired; that demand curves are necessarily downward sloping; that price measures marginal benefit to society; and that individual supply curves slope upwards and can easily be aggregated. Even allowing these invalid a.s.sumptions, the economic a.n.a.lysis of the labor market is still flawed.

The kernel.

Economists prefer to treat everything, including labor, as a simple commodity, subject to the same 'laws of supply and demand' as the simple apple. Yet their own a.n.a.lysis of labor shows that it is fundamentally different. In all other markets, demand decisions are made by consumers and supply decisions by producers. But in the labor market, supply decisions are made by consumers (households supplying labor), whereas labor demand decisions are made by producers (firms hiring labor). Thus the conventional economic a.n.a.lysis of markets, which is suspect enough on its own terms, is highly unlikely to apply in this most crucial of markets. As a result, wages are highly unlikely to reflect workers' contributions to production, as economists argue.

The roadmap.

In this chapter, I outline the economic a.n.a.lysis of labor supply, and the normal economic argument in favor of letting the market decide both wages and the level of employment.

I show that irregularities in the supply of labor when compared to a normal commodity are easily derived from this a.n.a.lysis, yet economists unjustifiably a.s.sume that labor supply will be an upward-sloping function of the wage. However, these labor market irregularities can make the supply of labor 'backward-bending,' so that reducing wages could actually cause the supply of labor to rise rather than fall.

Though economists normally oppose unions, there are economic arguments in favor of a cartel when sellers (such as workers selling their labor) face a buyer with market power. The opposition economists normally present to unions, to interventionist labor market policies, and to attempts to reduce income inequality are thus shown to be unjustified, even on the grounds of standard economic logic.

Labor demand and supply: an inverted commodity.

The economic theory that a person's income reflects his contribution to society relies on being able to treat labor as no different from other commodities, so that a higher wage is needed to elicit a higher supply of labor, and reducing the wage will reduce supply. In fact, economic theory supports no such conclusion. Even economists can't escape the fact that, as commodities go, labor is something out of the ordinary.

The demand for ordinary commodities is determined by consumer incomes and tastes, while supply is determined by the costs of production. However, unlike other commodities, no one actually 'consumes' labor: instead, firms hire workers so that they can produce other commodities for sale. Secondly, unlike all other commodities, labor is not produced for profit there are no 'labor factories' turning out workers according to demand, and labor supply certainly can't be said to be subject to the law of diminishing returns (whatever parents might think!).

These two peculiarities mean that, in an inversion of the usual situation, the demand for labor is determined by producers, while the supply of labor is determined by consumers. Demand reflects firms' decisions to hire workers to produce output for sale; supply reflects workers' decisions about how long to work, on the basis of their preferences for income on the one hand and leisure time on the other.

If economists are to argue that the labor market is to behave like all other markets, then these peculiarities must not complicate the usual totemic duet of a downward-sloping demand curve and an upward-sloping supply curve. Unfortunately for economists, they do.

Marginal workers.

According to economic theory, a firm's labor-hiring decision is determined simply by the impact that each additional worker has on the firm's bottom line. If hiring an additional worker will add to the firm's profit, he is hired; if not, the firm stops hiring.

With a perfectly compet.i.tive labor market, the firm can hire as many workers as it wishes to at the going wage. However, since one input (capital) is fixed in the short run, output is subject to diminishing returns: each additional worker hired adds a lesser amount to output than his predecessor. Diminishing marginal productivity therefore rules the hiring roost.

For each firm, the wage is a constant (set by the labor market in which each firm is an infinitesimally small actor). The amount each worker adds to profits, however, is variable. The firm keeps hiring workers up until the point at which the wage equals the amount for which the last worker's additional output can be sold.

If the industry itself is perfectly compet.i.tive, the additional units can be sold without the firm having to reduce its price (yes, I know that's been debunked already; but let's pretend otherwise). In general, the revenue the firm gains by hiring its last employee is equal to the price for which it sells its output, multiplied by the marginal product of the last worker. The firm's demand for labor is therefore the marginal physical product of labor multiplied by the price of the output.

6.1 The demand for labor curve is the marginal revenue product of labor A disaggregated picture of this is used to explain why some workers get much higher wages than others. They or rather the cla.s.s of workers to which they belong have a higher marginal revenue product than more poorly paid workers. Income disparities are the product of differential contributions to society, and though sociologists may bemoan it, both the rich and the poor deserve what they get.

Aggregate demand.

The demand curves for individual firms are aggregated to form this industry's demand curve for labor, which itself will be a small part of the economy-wide demand curve for labor (since workers can generate many different kinds of output). The real wage is set by the point of intersection of this aggregate demand for labor curve labor's aggregate marginal revenue product curve with the aggregate supply curve.

Aggregate supply, in turn, is simply the sum of the supply decisions of individual workers. According to economists, a worker's decision about how much labor to supply is made the same way he decides how much to consume.

Indifferent workers.

Individual labor supply is determined by the individual's choice between work and leisure. Work is a 'bad' in Bentham's calculus: work is a 'pain' while leisure is a 'pleasure.' Therefore the pain of work must be compensated for by the pleasure of the wage, to make up for the sacrifice of leisure required to earn the wage.

This choice is represented, as always, by indifference curves where potential income is one of the goods, and potential leisure time is the other. The indifference map represents a consumer's preferences between leisure and income, while the budget line represents the hourly wage rate: the higher the wage, the steeper the budget line.

This model has one peculiarity when compared to that applied to normal commodities. With standard commodities, the budget line can be drawn anywhere, so long as it reflects the relative price of the commodities in its slope, and the consumer's income. But with labor, one end of the budget line is fixed at twenty-four hours, since that's the maximum amount of leisure anyone can have in a day. For this reason, all that the budget line can do in this model is pivot about the twenty-four-hour mark, with the slope representing the hourly wage. The distance from zero to the twenty-four-hour mark represents the maximum possible leisure of twenty-four hours a day.

6.2 The individual's incomeleisure trade-off determines how many hours of labor he supplies 6.3 An upward-sloping individual labor supply curve.

As with the consumption of bananas and biscuits, the amount of leisure and income that a consumer will 'consume' is worked out by varying the wage, and seeing what combination of leisure and work the consumer chooses. This generates an individual labor supply curve not a demand curve from this worker.

The individual supply curve is then summed with that of all other workers to produce the market supply curve. We are back in the familiar economic territory of a downward-sloping demand curve and an upward-sloping supply curve intersecting to determine an equilibrium price: the average wage. The 'Totem of the Micro' is once again held aloft.

This argument, which strictly speaking applies to labor in the aggregate, is extended by a.n.a.logy to a disaggregated level in order to explain why some workers get much higher wages than others.

6.4 Supply and demand determine the equilibrium wage in the labor market At a policy level, this model is used to emphasize the futility of minimum wage legislation, demand management policies, and any other attempts to interfere with the free working of the market mechanism in this most political of markets. If a government attempts to improve workers' incomes by legislating a minimum wage, then this will result in unemployment, because it will increase the number of hours workers are willing to work, while reducing the demand from employers because the wage will now exceed the marginal product of labor. The gap between the increased hours offered and the reduced hours demand represents involuntary unemployment at this artificially high wage level.

6.5 Minimum wage laws cause unemployment.

Demand management measures trying to boost aggregate demand to increase employment will also fail, because they can't alter the marginal physical product of labor, which can only be done by raising the productivity of labor on the supply side. Attempts to increase aggregate demand will thus merely cause inflation, without increasing the real returns to firms.

6.6 Demand management policies can't shift the supply of or demand for labor The essential message is that 'you can't beat the market.' Whatever society may think is a fair wage level, or a socially desirable level of unemployment, ultimately the market will decide both income distribution and the rate of unemployment. Moreover, both these market outcomes will be fair: they will reflect individual productivity on the one hand, and the laborleisure preferences of individuals on the other.

Problems.

There are at least six serious problems with this meritocratic view of income distribution and employment determination: the supply curve for labor can 'slope backwards' so that a fall in wages can cause an increase in the supply of labor; when workers face organized or very powerful employers, neocla.s.sical theory shows that workers won't get fair wages unless they also organize; Sraffa's observations about aggregation, noted in Chapter 3, indicate that it is inappropriate to apply standard supply and demand a.n.a.lysis to the labor market; the basic vision of workers freely choosing between work and leisure is flawed; this a.n.a.lysis excludes one important cla.s.s from consideration bankers and unnecessarily shows the income distribution game between workers and capitalists as a zero-sum game. In reality, there are (at least) three players in the social cla.s.s game, and it's possible for capitalists and workers to be on the same side in it as they are now during the Great Recession; and most ironically, to maintain the pretense that market demand curves obey the Law of Demand, neocla.s.sical theory had to a.s.sume that income was redistributed by 'a benevolent central authority' (Mas-Colell et al. 1995: 117) prior to exchange taking place.

Backward-bending supply curves.

Neocla.s.sical economists blithely draw upward-sloping individual and aggregate labor supply curves, but in fact it is quite easy to derive individual labor supply curves that slope downwards meaning that workers supply less labor as the wage rises.

6.7 Indifference curves that result in less work as the wage rises The logic is easy to follow: a higher wage rate means that the same total wage income can be earned by working fewer hours. This can result in an individual labor supply curve that has a 'perverse' shape: less labor is supplied as the wage rises. Economists normally get around anomalies like this by dividing the impact of a higher price into its income and subst.i.tution effects where this time the price of labor is the hourly wage. The subst.i.tution effect necessarily means that you'll provide more labor, since each hour of leisure that you forgo gives you a higher return. It's the income effect which stuffs things up the fact that with a higher wage you can manage to get both a higher income and work fewer hours.

This ruse works when you're considering normal commodities: you simply notionally alter a consumer's income this was the basis of the 'Hicksian compensated demand curve' that played a role in the proof of the Law of Demand for an individual consumer in Chapter 2. However, this is no use when considering labor supply, because while it's quite easy to notionally add or subtract income from a consumer thus varying uniformly the amount of both biscuits and bananas that he can consume it's not possible to add or subtract hours from a day: you can't magically give a worker twenty-eight hours in a day, or take away four.

As a result, it makes no sense to separate the impact of an increase in the wage rate into its subst.i.tution effect and income effect: the fact that the subst.i.tution effect will always result in an increase in hours worked is irrelevant, since everyone will always have twenty-four hours to allocate between work and leisure.