Sunday 18 July 2010

Basic Welfare Economics (notes by Simon Vicary)

1. Preliminaries: the Role of Value Judgements

• The need for evaluative criteria.

• Positive vs Normative statements

• Positive Statements describe or attempt to describe how the world is

• Normative statements are statements of value. They can be aesthetic (beautiful) or ethical (ought)

• Hume's Law: “One cannot derive a ‘ought’ from an ‘is’”

The first section of Book Three of Hume’s Treatise of Human Nature (1740) is entitled “Moral distinctions not derived from reason”. This section concludes with the following:

“In every system of morality which I have hitherto met with, I have always remark'd, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought or ought not, expresses some new relation or affirmation, 'tis necessary that it should be observ'd and explain'd; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. But as authors do not commonly use this precaution, I shall presume to recommend it to the readers; and am persuaded that this small attention would subvert all the vulgar systems of morality, and let us see, that the distinction of vice and virtue is not founded merely on the relations of objects, nor is perceiv'd by reason.”

• Value judgements are what we impose;

• No "objective" justification for welfare/policy statements;

• Paretian approach “weak value judgements”


2. Paretian Welfare Economics

(i) Individualism: Social welfare is based solely on the welfare of the individuals in society
(ii) Non-Paternalism: the individual is the best judge of his or her welfare
(iii) Pareto Improvement: if in moving from state A to B one individual is better off and no one is worse off then state B should be chosen

Note that utilitarianism, in the form that recommends the maximisation of the sum of individual utilities (total community happiness) conforms to these axioms. In this sense it is a special case of Paretianism.

• These value judgements are not trivial.
• Maybe worth seeing where they lead us….The philosopher Rawls describe this process of finding out where value judgements lead as a process of trying to achieve “reflective equilibrium”.


3. Exchange and the Edgeworth-Bowley Box Diagram

• The 2x2x2 economy: not as restrictive as it might appear
• The Edgeworth-Bowley Box Diagram
• Exchange in the Edgeworth-Bowley Box Diagram
• The efficient distribution of goods
• MRSxya = MRSxyb
• The Set of Pareto Optima
• This shown in Figure 1 (lecture summary)


4. Walras’ Law and Competitive Equilibrium

(Skip this section)

Consider an exchange economy. There are j commodities, and each agent i has an endowment of commodities i, and each agent solves the following utility maximisation problem:



subject to:



As you may recall from earlier work, the solution to this problem takes the form of the demand functions for the m commodities:

for j = 1, 2, …….m and i = 1, 2, ……..,n

where

Note that these demand functions are the gross demands for each commodity. The net demand or excess demand for j by individual i is given by:



The aggregate net demand/excess demand for j is given by:



This plays an important role in what follows although we will keep the argument fairly informal. In a single market we have equilibrium whenever demand is equal to supply. For a general equilibrium (the whole economy) we simply extend this idea and require that demand equals supply in each and every market. Put another way, excess demand should be zero for all markets. Put formally:

∀ j, zj(p1, p2, …….., pm) = 0

(This idea can be extended to allow for zero prices and excess supply, but we will ignore this technical issue here). We can talk therefore of a competitive equilibrium. This has a rather precise definition that we will need later. In essence, though, it is simple enough. For an allocation to be a competitive equilibrium we require the following properties:

• Individual firms and consumers face a given set of prices
• Given their endowment of goods and services all consumers have maximised their utility subject to their budget constraint (chosen the best of the available consumption bundles)
• Given the prices faced, all firms have maximised their profits. That is, the technology to which they have access does not enable them to increase their profit beyond what they have achieved
• Demand equals supply for all goods and services. This is the last condition.

In a pure exchange economy everything relevant for finding and analysing an equilibrium is contained within the excess demand functions. With production, matters are a little more complicated. When we get more formal we will keep to an exchange economy, but the key results and ideas we examine carry through to a production economy. One such is a key feature of any general equilibrium system, and one that is widely applied in economic theory:

Walras’ Law
The value of total aggregate excess demand across all commodities is identically equal to zero:

To prove this note that from the budget constraint for individual i we must have:



from which it is easy to see that:



Now sum this equation over all n individuals:



Now change the order of summation so that we first sum over individuals rather than commodities:



Now consider the term:



This is the value of total excess demand for commodity j. Our last equation therefore tells us that the sum of these values across all individuals is equal to 0. This is Walras’ Law. We can write this as:



There are two important consequences of Walras’ Law.

• First suppose that we have achieved equilibrium in the first m – 1 markets. zj = 0, 1, 2,……., m – 1. In this case it must also be true that we have equilibrium in the mth market as well (as long as pm > 0).
• Recall that demand functions are homogeneous of degree zero in prices and income (here only prices really matter). Suppose therefore that we have market equilibrium and divide each price by p1, the equilibrium price for commodity one. In this case it must follow that if:



then



In other words if we are trying to determine equilibrium prices in any general equilibrium system then we cannot uniquely determine the absolute value of prices. We can only hope to determine relative prices uniquely. In a way this helps us in our calculations, because what we can do is to take one good as the numeraire, by setting its price equal to unity (one), and solving for all prices relative to this price. Essentially this is what we see in the final equation.


5. Production and Full Pareto Optimality Conditions

We take a simple economy in which production takes place. This is enough to illustrate the key points we want to make. We assume a 2 by 2 by 2 economy: two individuals; two goods; and two factors of production. We also assume that the supply of factors of production to the economy as a whole is fixed.

To start with we need to make a simple point about production. We do this using an adaption of the standard box diagram we have seen before. Look at Figure 2. Here we have two industries, x and y, the outputs for which are produced using two factors of production L and K. Clearly the economy could not be in a Pareto optimum if it were possible to increase the output of one industry whilst at the same time maintaining the output in the other industry. How can we see such points on our diagram?

A moment’s reflection should convince you that the reasoning we apply to this diagram is exactly the same as that we applied to our original box diagram. Individuals are replaced by industries, indifference curves by isoquants, and commodities by factor of production. The condition for Pareto optimality on this diagram is graphically the same as before. If isoquants are tangential to one another then it is impossible to increase the output of x without at the same lowering the output of y. Hence tangency of isoquants is a condition for Pareto optimality.

The slope of an isoquant has a number of names. Here we call it the marginal rate of technical substitution of factor L for factor K. It measures the amount of factor K we need to replace a given lost of L input if output is to remain the same. In symbols:



MPi stands for the marginal product of factor i in the industry, and justification of the last equality is exactly analogous to the justification of the statement that the MRS is equal to the ratio of marginal utilities. Let us suppose this equality is met. It follows that we can trace out a function relating y to x, such that given any value of x the output of y is maximised. This function if drawn in (x, y) space is called the production possibility curve. Bear in mind one important point when looking at a production possibility curve: it is drawn under the assumption that this condition is met. Put another way, it assumes that that the economy is efficient in production (Here this means no more than that it is impossible to increase the output of any good without at the same time lowering the output of some other good.)

We now turn to a consideration of the full conditions for Pareto efficiency in our economy. Figure 3 shows a production possibility curve, AB. Total community output is at point 0B, thus we assume the economy is productively efficient. An Edgeworth-Bowley Box diagram is inscribed below and to the left of this point, showing how the total amount of x and y produced can be divided up between the two agents A and B. At point P we have equality of the MRS for each of our two individuals. Is this enough to ensure overall Pareto optimality?

Unfortunately the answer is no. As things stand at the moment we have

• Efficiency in exchange
• Efficiency in production

However, consider now the slope of the production possibility curve, the Marginal Rate of Transformation. ( )

Suppose the MRT = 2 and the (common) MRS equals 1. A moment’s thought (albeit a bit tricky moment as we have to think in terms of ratios) should convince you that the economy can gain a Pareto improvement by shifting resources from the x industry to the y industry. Suppose we lower output of x by one unit. In this case the MRT tells us that we will enjoy an increase of 2 in the output of y, assuming we keeping production efficiency. If we lower the output of x by one unit, then one of our individuals will lose a unit of x in their consumption. The MRS condition tells us that in order to compensate our individual for this loss we must give him/her one unit of y. However, we actually have two extra units of y to distribute to the individuals. Thus we can compensate our individual for the loss of x and then give the remaining unit of y to one or other of our individuals to make them better off. Thus somebody can gain, and no one loses. Our original position could not have been Pareto optimal.

This argument can be repeated for any situation in which MRTxy ≠ MRSxy. Hence to the two conditions for Pareto optimality we already have we must add:

MRTxy = MRSxy

If we allow for variable factor supply we would have to add more conditions (Varian discusses this point, and indicates what these conditions involve), but for now these conditions should do:

• Efficiency in exchange
• Efficiency in production
• MRTxy = MRSxy Overall output efficiency condition

So far our discussion has been purely technical. We have said little about how these conditions can be achieved. We now turn to this point.


6. Pareto Optimality and Competitive Equilibrium

The Pareto Optimality conditions are simply conditions for a Pareto optimum, nothing more. They give no institutional detail and as such say nothing about how a Pareto optimum might be achieved. However, if we recall our basic microeconomic theory, there are some hints as to how this could be done.

Consider first the condition;



Suppose x and y are purchased in a market, and suppose that they each pay the same price for the goods. We know from utility maximisation that at the consumer optimum:



Thus if both A and B face the same prices utility maximisation ensures that this particular condition for Pareto optimality is met.

Similar observations apply to the second condition:



Simply assume producers have to buy inputs in a competitive factor market. Suppose the price of factor L is w (the wage rate), and factor K, r (the rental on capital). Cost minimising firms will choose a technique of production such that:



Thus if firms all face the same factor prices, cost minimisation ensures that marginal rates of technical substitution are equalised across sectors of the economy. Hence the second condition for Pareto optimality is met. The economy must be producing somewhere on its production possibility curve.

Overall output efficiency (MRTxy = MRSxy) is less easy to explain, although the underlying principle is the same. The tricky part of the argument is to sort out the value of the MRTxy. If you are interested here it is.

Suppose we decide we want to lower the output of x and raise the output of y, and that we will do this by transferring L of factor L from the x industry to the y industry.

The loss in x and the gain in y will be given by:



It follows from this that:



The minus sign reminds us that x and y must vary inversely.

By applying the same argument to factor K we find:



The question arises as to whether the ratio of the output changes will be the same for the two factors. The answer to this question is yes, if the second condition for Pareto optimality has been met. Recall that this tells us that:



Now by using an argument similar to that for which we get the slope of an indifference curve it turns out that:

and

Given that the MRTS is the same in the x and y sectors, ratio of the change in y to the change in x will be the same regardless of whether it is L or K that is being transferred. Returning now to the equations for , note that this term is nothing other than the (discrete approximation) to the MRTxy. Hence we can write:



Now recall from price theory that:

and

It follows that:



Our conclusion is that the marginal rate of transformation is equal to the ratio of marginal costs. Armed with this conclusion we can now show how prices might guide the economy to a Pareto optimum.

Suppose that the prices consumers pay are the same as those received by firms (suppliers). We know that under perfect competition profit maximising firms equate price to marginal cost. Hence as a result of profit maximisation by firms we can write:



We also know that utility maximising consumers will choose a consumption bundle such that:



Our final conclusion is that if we have perfect competition with profit maximising firms and utility maximising consumers then our third condition for utility maximisation will be met. It appears therefore that there is an intimate relationship between prices in perfect competition and Pareto optimality. In the next section of the lecture we explore more precisely the exact nature of this relationship.


7. The Fundamental Theorems of Welfare Economics

The argument of the last section was suggestive, and points the way to two of the most important results in economic theory: the Fundamental Theorems of Welfare Economics.

Theorem 1
Any competitive equilibrium is also a Pareto Optimum.

Theorem 2
Any given Pareto optimum can be realised as a Competitive Equilibrium.

Varian (p522-523) gives a proof of Theorem 1 for a simple exchange economy. The proof for more complicated economies is similar, and carries through to economies with production. For an exchange economy the argument goes as follows:

• Start with a competitive equilibrium, in which all agents choose an optimal point in their budget set
• Suppose the allocation is not Pareto optimal. This means that there is a Pareto improvement available to the economy with its existing endowment of goods and services
• As some agents are better off with the Pareto improvement, the consumption of goods in that state cannot be affordable given the prices realised in the competitive equilibrium
• Some individuals may be indifferent as between their competitive consumption and their consumption in the Pareto improvement. Under monotonicity of preferences this means that the value (at the equilibrium prices) of their consumption in the Pareto improvement must be no lower than it is in the competitive equilibrium
• Consequently at the prices in the competitive equilibrium the total value of consumption in the Pareto improvement must be greater than in the competitive equilibrium
• But the total value of consumption in both the competitive equilibrium and the Pareto improvement must equal the value of the economy’s endowment of goods and services
• The last two points contradict one another. This contradiction is a consequence of assuming that we had a Pareto improvement over a competitive equilibrium. Hence this assumption is false. Any competitive equilibrium is also a Pareto optimum, and the Theorem is proved Q.E.D.


There are a couple of points to note about this theorem.

o The first can be verified by checking through the argument again. The number of assumptions employed is very limited. In fact there is only one special one (over and above standard assumptions like utility maximisation): monotonicity. This is often weakened to a “better point” assumption, but otherwise we do not even need to make special assumptions like convexity of preferences. Partly this is because our starting point of competitive equilibrium already places some structure on the economy we are dealing with, but nevertheless technically what is needed to prove the theorem remains very limited.
o Secondly, although the Theorem tells us that a competitive equilibrium is Pareto optimal, it does not tell us anything about the distribution of income. Indeed, a Pareto optimum is quite possible if one individual has virtually all the wealth in an economy, and the rest are left with subsistence wages. The actual competitive equilibrium we end up with may not be attractive at all. For this reason the First Fundamental Theorem of Welfare Economics needs to be supplemented with its companion, the Second Theorem.

This theorem is a good deal more difficult in its details, although the principle underlying the proof is simple enough to grasp. The problem lies in the fact that we start from a Pareto optimum. This has as such no institutional detail. In particular, there are no prices to work with. The strategy of the proof is as follows:

• First establish a set of prices. Under a standard set of assumptions, these prices correspond to the prices discussed in section 6 of the lecture
• Show that at these prices firms maximise profits by choosing the input output combinations in the Pareto optimum
• Show that by some suitable choice of consumer endowments consumers will choose the consumption bundles given in the Pareto optimum
• Equality of demand and supply is usually established within the Pareto optimum itself in that total consumption has to equal total endowment (exchange economy), or that total consumption must be feasible given the endowments and technology of the economy. Hence once the last two points have been established, the equality of supply and demand in each market is automatically established

Many more assumptions are needed to prove this second theorem, but the key technical assumption turns out to be convexity both in preferences and in technology (if we are dealing with a production economy). The role of convexity is to establish prices from which the rest of the proof follows. Varian shows how the theorem works in a box diagram, and an analogous argument is illustrated in Figure 4. Point P is the Pareto optimum on the production possibility curve AB. The set of (total) consumption points that can achieve a Pareto improvement on P is shown by the shaded area above P. If individual preferences are convex, then this set is also convex, as illustrated. Convexity in production (either diminishing returns to scale as Varian discuss or constant returns with differing factor intensities as assumed by trade theorists) results in a production possibility curve as illustrated, with the set of feasible outputs (consumption levels) being convex. If in these circumstances point P is a Pareto optimum then the two sets do not overlap, and in this case we can draw a straight line separating them (MN). This idea carries to any number of goods/factors (technically it is a theorem of separating hyper-planes). The slope of the line MN on the diagram gives us a set of relative prices. Using these prices we need to show that point P arises from profit maximisation by firms and, with suitable endowments, utility maximisation by individuals. Doing this completes the proof of the Theorem.

The consequences of this theorem are far reaching as far as the competitive mechanism is concerned. We can in fact specify any Pareto optimum we please, incorporating any notion of fairness or justice in income distribution we like. Regardless of what we want, the competitive mechanism will be able, given suitable initial endowments of goods and factors, to deliver the Pareto optimum we want.


8. Underlying Assumptions

I have mentioned the technical assumptions (monotonicity, convexity) needed to deliver the Fundamental Theorems. Before discussing their significance, it might be useful to list the implicit assumptions they employ:

• Perfect Competition: actually this isn’t really an implicit assumption, as it was assumed in both theorems that all agents were price takers. However, it needs to be stressed that competitive markets are required for the fundamental theorems to be valid.
• Egoism: no individual’s utility is influenced by the utility of other agents (people are neither benevolent nor malevolent).
• No externalities: individual utility is affected neither by the consumption chosen by any other individual, nor by the input-output combination chosen by any firm. Similarly no firm’s output is affected by consumption choices per se of individuals or input-output combinations chosen by any other firm. In short, individual utility is affected only by individual consumption, and firm output only by firm input.
• No problems with uncertainty: It is simplest to begin with to assume no uncertainty. The theorems can actually deal with uncertainty in some contexts, but this raises complicated issues best put aside for the moment (that is, in this module).


9. Comments on the Fundamental Theorems

The two fundamental theorems of welfare economics taken together establish one of the most important and profound results in economic theory. With any such result, though, it is important to appreciate what the theorems tell us and what they do not tell us.

A Bit of History
The idea of the Fundamental Theorems is an old one, going back at least to Adam Smith and his “invisible hand”:

“Every individual...generally, indeed, neither intends to promote the public interest, nor knows how much he is promoting it. By preferring the support of domestic to that of foreign industry he intends only his own security; and by directing that industry in such a manner as its produce may be of the greatest value, he intends only his own gain, and he is in this, as in many other cases, led by an invisible hand to promote an end which was no part of his intention.”
The Wealth of Nations, Book IV Chapter II

The key idea here is that by pursuing one’s one interest (own gain) an individual achieves a desirable social end (the “public interest”) which was “no part of his intention”. This is what happens in the Fundamental Theorems. Each individual maximises his or her own utility, and each firm maximises profits. No one is concerned with the public good, but under the assumptions the outcome is a desirable one. The quotation, however, shows that Smith seems to have had in mind something slightly different, and perhaps a little more restricted, than what we now understand by the term “invisible hand”.

The general idea though is an old one, but if we look at Smith’s quotation, and consider more generally the understanding people took from his arguments in the early 19th century (and bear in mind that The Wealth of Nations was an extremely influential book), there are some loose ends. What exactly do we mean by the “public interest”, and what exactly do we mean by “his own gain”? Kenneth Arrow in 1951 was the first person to formulate and prove the theorems in their modern form, and his great contribution was to make precise (at least in one sense) the terms Smith used. This was an important step in that by making the ideas precise it is easier to see both why the idea works, and the circumstances under which it does work. It also, of course, makes it easier to see when the invisible hand might fail to achieve a fully desirable outcome.

Allocation and Distribution
One aspect of the theorems needs stressing. Suppose we adopt the Paretian approach to welfare economics. As a result of the fundamental theorems we can see that questions involving economic analysis can in principle be split into allocation and distribution issues. Hence if we suppose the distribution of income to basically “fair”, then the competitive mechanism ensures that a Pareto optimum is achieved. The general idea then would be to make the economy more like the “competitive mechanism” of the theorems, and not to worry about the distribution of income. Alternatively, if we suspect that the economy does not work well, and we decide to improve matters by adopting a more competitive mechanism. It is not a sensible argument against such reform that the distribution of income would be “unfair”. By appropriate policies it is in principle possible to correct for unfairness in the distribution of income through some appropriate form of redistribution.

It is not then a good argument against the competitive mechanism that there is undue inequality. Adopting the second theorem we can see that any distribution of income is in principle compatible with this mechanism, should we wish to re-distribute appropriately. We can, then, discuss distributional questions without concerning ourselves about allocation questions.

As you may find if you persevere with public economics the following statement is something of a simplification. However, as a first approximation it is not bad, and is an important consequence of the Paretian approach to welfare economics:

Questions of how resources get allocated as between different industries (or how individual consumption can vary) can be separated from the questions of how resources are distributed amongst the individuals in the economy.

It is worth pointing out that this distinction lay behind the classic division of the functions of a modern state made by Musgrave in his classic textbook The Theory of Public Finance, 1959. These divisions were:

• The allocative branch
• The distributive branch
• The stabilisation branch

These do not of course correspond exactly to the precise divisions of the civil service. Most state activities are a mixture of these three. Nevertheless it is easy to see that some arms of the state are more closely involved with one function.

The Bank of England is clearly concerned predominantly with stabilisation, and indeed the changes in its functions enacted by Gordon Brown when he became Chancellor in 1997 made this clearer: as you know, the Bank was given the task of targeting inflation by means of adjusting interest rates (stabilisation), but its financial regulatory functions (allocation) were transferred to (what is now) the Financial Services Authority.
The Social Security Budget is obviously concerned predominantly with distribution, as its main activities involve the transfer of money to (mostly) the old, but also to the unemployed and disabled.
Defence expenditure is chiefly allocative in nature, however.

The division is not watertight, even in these clear-cut cases. Still less is this so for education and health expenditure. Nevertheless, in helping us to think clearly about public policy the Pareto-Musgrave distinction is a good one to use in the first instance.


What is the Competitive Mechanism?
The competitive mechanism of the fundamental theorems corresponds effectively to the textbook model of perfect competition that you met early on in your study of economics. The theorems themselves confirm what has been hinted at all along: namely that in some sense perfect competition is a “good thing”. However, this observation should remind us to be careful in equating the “competitive mechanism” of the theorem with the market-type mechanisms we observe around us in the so-called real world. There are a number of reasons to suspect that actual market systems may not be quite as efficient as is in principle possible. We examine this question briefly in the next section, but first let us take another dip into the history of economic thought.

One topic much debated in the 1920’s to the 1940’s was the question of socialism. For our purposes it is interesting to note that ideas put forward by two economists Lange and Lerner, who claimed that an appropriate form of socialism (their form) was a better approximation to the textbook ideal of perfect competition than actual market mechanisms. They recommended not Stalinist central planning, but a form of market socialism in which managers ran socialised firms. Managers were supposed to do all the usual things mangers do, such as make output and employment decisions. However, they were not allowed to set prices, this being done by the central “planner”, who would adjust price of all products according to the state of the market. In this way the socialised firms would act as price takers, just like firms in perfect competition, and the economy would, in principle, conform to that postulated by the fundamental theorems.

This is not the place to pursue this idea in depth. Suffice it to say that the practical application of our theorems could take a number of forms. The theorems do not necessarily provide an argument for “market” economies as we observe them. Indeed Friedman’s Capitalism and Freedom (1962), a standard text arguing for markets does not rely on these theorems. Arrow himself is counted as a market socialist and clearly had more in mind than arguing for markets when he introduced the theorems. We will pursue this question in the final section of the lecture.


10. Market Failure and Prices.

The theorems seem to make quite strong assumptions (such as universal perfect competition), and for this reason they could be taken to provide some ammunition for those who are critical of markets. Rather than jumping to any sort of policy conclusions it might be wiser to think of our theorems as providing a framework for thinking about policy. On this argument we see where conditions approximate to those postulated, and there rely on the market to allocate resources; and where conditions are far from those assumed some form of public policy may be desirable. This approach reflects a widely used concept in economic theory: market failure.

To understand this general idea, take the second welfare theorem, and note that part of its proof involves the idea that attached implicitly to every Pareto optimum is a set of prices, or, if we want to be a little les committal, opportunity cost ratios. Bator (1958) took market failure to mean a failure of agents within a market system to use these appropriate opportunity cost ratios as the prices that determined their own allocation decisions. The idea is not too difficult to grasp. We know from an earlier section that attached to each Pareto optimum there exist a set of opportunity cost ratio (MRS = MRT etc.). Each corresponds in some sense to prices that might be observed in an actual market. If the actual prices do not in fact equal these so-called shadow prices, then the economy will not achieve the desired Pareto optimum, and we have market failure. Why this might be so varies, and a quick list here will help.

Briefly, and technically, markets can fail as a result of:

• Failure by existence. It may in fact be impossible to find a set of prices that sustains a desired Pareto optimum. This is an interesting mathematical problem, but not one over which policy makers lose much sleep.
• Failure by signal. There may be prices attached to the desired Pareto optimum, but agents respond to prices in the wrong way. This can happen when there are non-convexities in the economy. The most likely possibility for this to occur is when there are economies of scale
• Failure by incentive. Firms may recognise that at the Pareto optimal prices the profit maximising output is indeed the Pareto optimal output. However, it is possible that they make a loss when they produce this output (so it is really loss minimising). It is better not to produce at all. Firms have no incentive to choose the desired output.
• Failure by structure. In practice, firms choose price. Will they choose the desired price for the Pareto optimum? Not if they enjoy monopoly power. This is also the case with most models of oligopoly as well. The structure of the industry gives firms no incentive to set prices at the desired level. It is (at least potential) failure in this sense that underlies much of Industrial Economics.
• Failure by enforcement. This is the form of market failure we shall mostly be concerned with in this module. The broad category here is externalities. These may stem from the non-existence of proper markets for goods, perhaps through a failure to establish clear property rights. However, in some cases (those we will be mostly concerned with), it may well be impossible to do this. This is particularly so when we are dealing with public goods, to which most of the module is devoted.


The causes of market failure are attributed to some combination of the following:

• Ownership externalities
• Technical externalities
• Public good externalities

No comments:

Post a Comment