This summary was written in the year 2013-2014.
The aim of this chapter is to develop a framework that explains and justifies (or fails to justify) the fact that the state has considerable involvement in the finance and/or delivery of some goods such as health care and education, but leaves others like food to the private market. The main issues concern economic efficiency and social justice. The conventional starting point is the social welfare-maximization problem. The aim of policy is to maximize social welfare subject to three sets of constraints, tastes, technology, and resources, i.e.
Appendix 1
The problem formulated relates to a first-best economy, which implies one of two situations:
Either there are no impediments to efficiency, and also an optimal distribution of endowments
Or government can counter inefficiency or maldistribution with first-best policies (e.g. through lump-sum taxation)
The conditions are stringent, as is the assumption that lump-sum taxation is feasible. A second-best economy faces additional constraints:
Imperfect information is a recurring theme, leading to unrestricted markets to be inefficient or inequitable, and intervention may improve matters.
Externalities are present, and may justify intervention in various forms.
Economic efficiency is about making the best use of limited resources given people’s tastes. It involves the choice of an output bundle:
Appendix 2
Regarding economic efficiency, three conditions must hold simultaneously:
Productive efficiency, which means that activity should be organized to obtain the maximum output from given inputs. Figure 3.2 shows the maximum quantities of the two goods that can be produced with available resources. Productive efficiency means that production is at a point on – rather than inside – the transformation curve (all points on the transformation curve conform with productive efficiency.
However, this is not enough for allocative efficiency, which requires two additional conditions:
Efficiency in product mix means that the optimal combination of goods should be produced, given existing production technology and consumer tastes.
Formally, production is not at any point on the transformation curve in figure 3.2, but at a specific point a, at which the ratio of marginal production costs (i.e. the slope of the transformation curve) is equal to the ratio of marginal rates of substitution in consumption (i.e. the slope of the social indifference curve, I-I).
Efficiency in consumption means that a person’s consumption choices should maximize his/her utility – i.e. the marginal rate of substitution must be equal for all individuals.
This last meaning is summarized in the Edgeworth box in figure 3.3. Economic efficiency, in the sense of a movement to an appropriate subset of the contract curve, is an important aim under all the definitions of social justice as discussed in chapter 2; and in a first-best economy no distribution can be socially just unless it is also efficient.
Pareto efficiency incorporates two value judgments:
Social welfare is increased if one person is made better off and nobody worse off;
Individuals are the best judges of their own welfare.
The fixed-factor case assumes an Edgeworth box of given size, and also that the conditions for efficiency in production and product mix hold. This is equivalent to discussing a first-best solution.
Libertarianism. Welfare is increased by any Pareto improvement.
Utilitarianism. This aims to maximize total utility, and any Pareto improvement increases welfare.
Rawls. The aim is to distribute resources in accordance with social justice. A Rawlsian improvement (RI) benefits the less advantaged individual. In a first-best economy all Rawlsian socially just distributions lie on the contract curve. According to Rawls, goods should be distributed equally, unless any other distribution benefits the least well off.
Socialism. Under one interpretation resources should be shared equally. A movement that reduces relative inequality is a socialist improvement (SI). SI is a subset of RI, and all first-best solutions that are just in a socialist sense lie on the contract curve; and, like Rawls, socialists will favor any movement along the contract curve towards k (figure 3.3).
Relaxing the fixed-factor supply assumption complicates matters because of the need, in the absence of lump-sum taxation, to analyse policies in a second-best economy. It is a standard proposition that lump-sum transfers, being based on characteristics exogenous to the taxpayer, can bring about any desired distribution of income without efficiency loss. Taxes related to income, including any indirect tax whose payment rises with income, are not lump sum, and generally cause inefficiency, inter alia through their effect on individual labor supply. However, attempts to achieve social justice involve redistribution; hence taxation must inevitably relate to income. Consequently, any practicable system of taxation may cause inefficiency in production and/or product mix. Thus there may be a trade-off between efficiency and equity.
In face of this trade-off there will be different views about the desirability of an increase in efficiency, which will not be seen as a welfare gain if its equity cost is too high. To some libertarians it might have zero weight. For utilitarians this might be an open question: a given efficiency gain may (not) increase welfare; and the utilitarian optimum will not necessarily be efficiency (i.e. utilitarians will tolerate some efficiency loss in the interest of greater justice). Rawlsians and socialists give social justice more weight, and will therefore tend to accept a higher efficiency loss to achieve a just distribution. However, note that no theory of society gives social justice complete priority. Even a Marxist would resist the pursuit of distributional objectives if the resulting efficiency costs reduced output to zero.
A first-best economy has the following characteristics:
Perfect competition;
No problems such as externalities, public goods or increasing returns to scale;
Perfect information on the part of buyers and sellers;
Maximizing behavior in the context of well-behaved utility and production functions;
Complete markets;
No distortionary taxation.
The first fundamental theorem shows that in a first-best economy, the operation of perfect competition will lead to a Pareto-efficient allocation of resources, i.e. to a point on the contract curve in the Edgeworth box. The second fundamental theorem shows that in a first-best economy it is possible to reach any desired point on the contract curve by establishing a suitable set of initial endowments, and then letting agents trade.
Deviations from the first-best:
Imperfect competition. It is a standard proposition that a monopolist can be given an incentive to produce the efficiency output either though the imposition of a maximum price (i.e. regulation) or via an appropriate price subsidy.
Public goods. Those exhibit three technical characteristics: non-rivalness in consumption, non-excludability, and non-rejectability, which together imply that the market is likely to produce inefficiently, if at all. The appropriate form of intervention is generally public production.
External effects. These arise when an act of A imposes costs or confers benefits on B for which no compensation from A to B, or payment from B to A, takes place. A technological externality arises when A’s utility function or production function is interrelated with B’s. For an external cost (benefit), the market clearing output will generally exceed (falls short of) the efficient output. The problem might be solved by a merger, negotiation between the two parties, regulation, or an appropriate Pigovian tax or subsidy.
Increasing returns to scale. The presence of increasing returns to scale at all levels of output implies that average cost will exceed marginal cost. The appropriate intervention is subsidy or public production, or both.
However, these deviations only justify residual welfare state, not an institutional welfare state, because it gives no efficiency argument for public production.
Complete information requires at least three types of knowledge:
Quality. The assumption that economic agents have perfect knowledge about the nature of the product implies that individuals have well-defined indifference maps, and firms, similarly, well-defined isoquants. This is not always true. Where information is seriously deficient and/or complex, market outcomes may be less efficient than some sort of administrative solution and there may be a justification for public production and allocation. Markets are generally more efficient (1) the better is consumer information; (2) the more cheaply and effectively it can be improved; (3) the easier it is for consumers to understand available information; (4) the lower are the costs of choosing badly; and (5) the more diverse are consumer tastes.
Asymmetric information. Sometimes one side of the market is less informed than the other.
Price. Efficient choice requires that agents are well informed about prices – that is, that they face a well-defined budget constraint. Where the assumption fails, the market may supply the necessary information, but where the market does not resolve the problem, state intervention via regulation may be necessary.
The future. Inter-temporal utility maximization requires perfect information about the future. In a world of certainty, the welfare state only has a small role: (1) insurance is not necessary, since there is no risk; (2) people save voluntary, borrow in perfect capital markets and, thus, consumption smoothing takes place through voluntary action using private institutions; (3) transient (i.e. temporary) poverty is also dealt with by borrowing or saving; and (4) the only reason for a welfare state in such world is to provide poverty relief for a person who is lifetime poor.
The efficiency advantages of perfect competition are contingent on perfect information.
Deviations from the first best also influence behavior
Bounded rationality. This arises where the complexity of choices leads to an inability to know what to do. Deviations form rationality arise in many guises.
Rules of thumb. In the face of information-processing problems people may adopt rules of thumb. Mental accounting is a particular type of rule of thumb.
Framing. Choices can be influenced by how they are presented. There is also evidence that choices differ depending on whether information is presented as a table of graph.
Irrational choices. Rationality suggests that people should seek to maximize the expected value of assets, which implies that gains and losses are treated together on a net basis. But in practice many people act as though the utility gain from the rise in the value of an asset is smaller than the utility loss if the value of an asset falls (loss aversion)
Bounded will-power. Another problem arises where people know very well what behavior is in their best interests, but do not act on that knowledge. An explanation is hyperbolic utility functions, with a low rate of time preference for the future, but a high rate of time preference in the present. Solutions include commitment devices, where people voluntarily put themselves into a situation where it is difficult to change there mind. Another solution is auto-enrolment – that is, a situation where someone is enrolled automatically, they can opt out, but have to take action to do so.
Markets are often incomplete and fail to supply public goods; certain risks are uninsurable; capital markets may in some circumstances fail to provide loans; there may be no futures market; a commodity may not be supplied because a complementary market is absent; and incomplete contracts. Incomplete contracts are in some respect a form of information failure, arising where neither the individual nor the government can monitor quality effectively. The core argument for private delivery is that ownership affects the incentives that agents face, e.g. to reduce costs, to innovate. Also, the more clearly the government can specify its wishes in a contract with the private provider. Besides, it is easier to monitor that the contractual conditions are being met, and the stronger are the incentives for private providers to maintain quality, e.g. where reputational effects are strong. The case for public delivery arises where one or more of the conditions for private delivery fail sufficiently badly. Competitive pressures to reduce costs may reduce quality. First, imperfectly informed consumers may be unable to monitor quality effectively. Second, consumers may be able to monitor quality, but, because of unequal power, be unable to translate this knowledge into action. Finally, because of the nature of the product, government may be unable to monitor quality effectively. Motivations for government delivery might be poverty relief, ensuring access to core social services, and maintaining high levels of employment.
Taxes cause distortions if they change behavior. A standard proposition is that lump-sum taxation is non-distortionary, since individuals cannot avoid it by changing their behavior (therefore this is compatible with the second fundamental theorem). However, sometimes the state imposes distortionary taxes e.g. for poverty relief issues.
A prima facie case for intervention translates into a case for action only if intervention can improve on an imperfect market outcome. Intervention must be cost effective. This is more likely in the case of a market failure and the more effective and non-corrupt is government. The size of the public sector depends also on demand, which has a political-economy as well as an economic dimension.
Opinions about why redistribution occurs differ. Some writers state that the poor use their voting power to enforce redistribution from the rich, whereas others say that redistribution exists because of altruistic motives. Where redistribution makes some people better off without making anyone worse off, transfers from rich to poor may be justified on quasi-efficiency grounds. In the simplest case a rich (R) and poor (P) individual have the same utility function that is dependent only on his own income.
Appendix 3
Now giving has become a partly public and partly private good. For the latter reason, free-riding becomes less of a problem. Voluntary redistribution alone will be suboptimal unless one believes both that free-riding is not a problem and that the optimal amount of redistribution is what the rich wish to volunteer.
The use of in-kind transfers for distributional purposes is limited unless they are also justified on efficiency grounds.
When individuals have unequal power, leading to horizontal inequity;
In cases of supply-side disruption.
However, sometimes it is politically easier to redistribute in kind. Suppose that the externality is caused not by P’s income but by his consumption, then
Appendix 4
Figure 3.6 shows how merit goods’ their existence can be explained in political economy terms by a consumption externality. The redistribution of income of consumption from rich to poor has to do with vertical equity. Social justice also involves horizontal equity, which concerns equality of opportunity, minimum standards for certain products, and so on. Minimum standards have to do with failure of one of more of the first-best assumptions, and can occur in three ways.
Where agents face information or behavioural problems, and efficient choices are not made.
If agent have unequal power, so that they are not able to enforce their decisions.
There may be externalities.
Where the first-best assumptions hold, the only cause of unequal access is shortage of income. But action to ensure equal access may be justified in particular by imperfect information or unequal power. Equality of opportunity is closely related to equal access.
There are four explanations of the extent of and the growth in government activity: dealing with market failures; redistributing income and wealth; the response to the electorate in the form of coalitions of voters or through pressure groups; and the role of bureaucrats. Figure 3.7 gives an overview of public and private provision.
The term insurance can mean different things:
Sharing risk. This defined insurance in terms of its objective.
Actuarial mechanism, normally organized in the private sector. This defines insurance in terms of a method by which the objective of risk sharing might be pursued.
If a person is risk-averse uncertainty causes disutility per se, hence certainty is a commodity yielding positive marginal utility, for which he/she will pay a positive price. The individuals’ expected income and expected utility are, respectively:
Appendix 5
The bases of insurance are the law of large numbers and the gains from trade. Under the former, persons may face uncertainty, but groups can face approximate certainty. It is the relative certainty about the aggregate probability resulting from the law of large numbers that opens up to individuals the possibility of exploiting gains from trade by agreeing to pool risks. We assume all individuals face the same probability distribution of outcomes, and y, the mean, and the variance of y for each individual are independent of those for every other individual.
Appendix 6
If N identically distributed and independent incomes are pooled, the variance of average income tends to zero as N tends to infinity. By pooling/trading individuals can acquire certainty.
Annuities (i.e. an annual income stream) are another form of pooling. An individual could buy pension of €y per year for a lump sum, A, where A is the present value of the pension stream for the rest of her life, n years, and r is the rate of interest:
Appendix 7
which shows that the annual payment €y depends on the size of the person’s lump sum, A, on the insurance company’s view of the applicant’s remaining life expectancy, n, and on the return, r, the insurance company expect on the assets it holds. If individuals A and B are identical, except A’s lump sum is twice as big, then the insurance company will pay A an annuity that is twice as large as B’s. Also, the insurance company will pay a lower annuity the longer it expects to pay benefit. This is dependent on age, gender, health and marital status. If changes in the price level are not to affect the real value of an annuity, it is necessary to base calculations on the real rate of interest, i.e. the excess of the nominal interest rate over the inflation rate.
The actuarial premium for the ith individual, πi, is defined as
Appendix 8
Pre-existing conditions, in short, are generally uninsurable. It creates inefficiency in the sense of a missing or incomplete market – risk averse individuals who would like to buy insurance but are unable to do so, or unable to buy as much cover as they would like. For the same reason, there is also inequity.
Risk probability of not uncertainty must be known, otherwise the insurer cannot calculate a premium, making actuarial insurance difficult/impossible. Insurance can cope with risk but not with uncertainty. The problem can arise because
Where insured event is rare; large variance.
Of the complexity of the situation; inflation leads to difficulties in estimating future price levels
Of the long-time horizon of certain insured events
Of asymmetric information; where the insurer has less information than the person buying the insurance. Adverse selection and moral hazard.
Suppose there are two groups of people (low and high risk)
Appendix 9
Three problems stand out:
Imperfect consumer information and bounded rationality. It diminishes benefits from competition, may lead to inefficiencies and inequities;
Imperfectly informed insurers. Competitive pressure may create problems in the form of cream skimming, gaps in coverage and third-party incentives to inefficiently high spending;
Administrative costs. With higher administrative loading, α, fewer people will buy insurance. Includes marketing, processing, reimbursement costs and foregone economies of scale that a larger company might enjoy.
If insurance is compulsory, charging all categories of risk the same premium causes little inefficiency in insurance markets, though it might cause secondary inefficiency in related activities. If insurance is voluntary, efficiency requires that insurers should seek to discover who is high and who low risk, and charge premiums accordingly.
Insurance is inequitable since it redistributed from those who do not make claims to those who do. However, the whole point of insurance is that people do not know whether they will need to claim. In Rawlsian terms, people buy insurance behind the veil of ignorance.
The need for insurance arises because in industrialized countries employment is largely a binary phenomenon (i.e. a person is either unemployed or employed) and retirement, similarly, is frequently a discrete event. Thus the risk against which social insurance offers protection are to some extent a social construct. On the supply side, market failures provide both a theoretical justification for and an explanation of, a welfare state that is much more than a safety net. Because of information failures, actuarial insurance cannot cover contingencies such as unemployment, inflation, and important medical risks. Social insurance is one response. Social insurance differs in two respects. First, because membership is generally compulsory, it is possible (though not essential) to break the link between premium and individual risk; a pooling solution is therefore an option. Second, the contract is usually less specific than private insurance, with two advantages: protection can be given against risks that the private market cannot insure; and the risks that change over time. Thus, social insurance, in contrast with actuarial insurance, can cover not only risk, but also uncertainty.
The main aim of cash benefit is the prevention of absolute poverty. But later policy also aimed at alleviating relative poverty. Insurance is concerned with the protection in the face of stochastic contingencies such as unemployment, ill-health, or widowhood. Pure private arrangements include the voluntary purchase of actuarial insurance and voluntary charity. Mixed schemes involve public participation in private arrangements through regulation (e.g. minimum standards for private insurance) and/or through finance (e.g. tax relief). Public schemes include social insurance, awarded on the basis of a contribution record and the occurrence of a specified contingency, such as unemployment on being above a certain age. Social assistance is financed from taxation and awarded on the basis of an income test. Libertarians say that only with non-contributory schemes, others pay benefits, and so the aim of transfers should to prevent absolute poverty – that is, benefits should be at subsistence level. Socialists, in contrast, see income transfers as contribution to their egalitarian aims, and therefore favour publicly organized transfers to prevent relative poverty, and to reduce inequity. Liberals take an intermediate line.
Unemployment benefits provide insurance against income loss where a person is involuntarily unemployed. Sickness and maternity both address temporary incapacity. Old-age, disability, and survivor’s benefits cover long-term risks. A problem in all countries is to assess long-term disability in a way that is fair to the claimant but does not lead to an explosion in the number of recipients.
Where an activity causes an external cost, one form of intervention is a Pigouvian tax. Here, however, the aim is not marginally to influence consumption through marginal price changes, but to prevent non-insurance. Making insurance compulsory (i.e. regulation) is likely to be a more effective way. The major efficiency argument for making membership compulsory is that uninsured losses due to unemployment, illness, or injury may impose costs on others.
Efficiency arguments about the appropriateness of public provision hinge on whether five assumptions hold for the risks covered by social insurance:
Whether individual probabilities of becoming or remaining unemployed are independent is controversial. Simple Keynesian theory suggests that unemployment reduced demand and contributes to further unemployment, thus making unemployment a common risk. Those who believe in a natural rate of unemployment deny this conclusion, except in the short run. Individual probabilities may be partly correlated, though this problem alone is unlikely to make private insurance possible.
The overall probability of being unemployed is less than one, though for some parts of labour force, it may be too high for private insurance to be viable. Individuals returning to the labour force after a long break in employment and unemployed school-leavers cause additional problems.
The average probability of being unemployed is well known – there is simply the aggregate unemployment rate.
Adverse selection: a private insurance company could ask about an applicant’s previous employment record. This is not a complete solution, however: the process is costly; verification is not always possible; and not everyone has a past employment record.
Much the greatest problem is moral hazard. The insured individual may be able to influence, first, the probability of entering unemployment by bringing about his own redundancy. Second, he can influence the probability of leaving unemployment – that is, the duration of unemployment.
Physic cost of unemployment is unobservable and two possible cases can be distinguished. It could be that the physic cost to the individual is high, and that continuing unemployment is caused by a lack of jobs. On the contrary, it could be that the cost is low and that the individual remains unemployed partly by choice. The first case is an insurable risk; the second is not. The efficiency duration of unemployment for the ith individual, xi, is the period tat he would rationally choose if he had to finance his unemployment from accumulated savings or by borrowing in a perfect capital market. Inefficiency arises when an individual chooses to be unemployed longer than xi because insurance has reduced the marginal cost to him of doing so.
If income support is to exist for the unemployed, it will have to be publicly provided. The alternative – no insurance of any sort – is suboptimal: it faces individuals with an inefficient amount of risk; it creates external costs; and it treats technical insurance-market problems as a binding constraint on social policy. The argument for public provision of some sort of insurance can be supported on efficiency grounds because of information problems in insurance markets, of which problems no mention is made in Hayek’s attack on publicly provided benefits. The libertarian predilection for private markets and voluntarism in this instance is untenable. The resulting institutions do not look actuarial.
An alternative approach – known variously as activation policies, welfare to work, and making work pay – subsidizes employment, notably through working tax credits.
Except for the chronically ill, the probability of absence is less than one, and can be estimated. There is no major problem of adverse selection, since a private insurance company can ask about an applicant’s previous health and record of absence. Nor is there a serious problem of moral hazard. The probability of missing work is broadly exogenous, since making oneself genuinely ill is costly, and pretended illness can, at least up to a point, be policed by requiring claimants to provide a doctor’s certificate.
Voluntary maternity insurance may face many problems of adverse selection, making private supply impossible. However, when it is compulsory, it would not necessarily have to be publicly provided.
There are good reasons for arguing that private long-term-care insurance policies, though they exist, are unlikely to be efficient. But also there are strong arguments for suggesting that long-term care is suitable case for social insurance.
The scope for differences in individual choices is limited
These are not risks that fit the actuarial mechanism very well.
The costs of long-term care are much lower than for pensions, because on average people require care for a much shorter period than they require a pension.
Where consumers are well-informed there is no case for intervention on horizontal equity grounds; where they are not, the case for publicly organized insurance can be argued on both efficiency and equity grounds. Regarding public provision based on vertical equity grounds, without efficiency reasons for provision, distributive goals should be pursued through income transfers except, possibly, where there are consumption externalities. In the presence of these, the rich may want the poor to consume insurance, and so impose it as a merit good; and the poor may feel less stigmatized by receiving insurance benefit rather than welfare.
On the face of it, social insurance does not look much like insurance:
In many countries contributions are not related to individual risk.
Entitlement to benefit in many countries does not depend only on the occurrence of the insured event. Benefits (but not contributions) are often higher were the contributor has dependants; contributions and benefits taken together are frequently redistributive from rich to poor; and entitlement to unemployment benefit may be subject to restrictions.
Social insurance is generally not financed on actuarial lines.
The contract is not fully specified, in that the risks covered can change over time. Social insurance, unlike private insurance, offers protection not only against risk, but also against uncertainty.
Though the replacement rate has an effect, labour supply is more influenced by other aspects of the benefit structure, in particular the maximum duration for which benefit can be received. In short, the hypothesis that unemployed benefits exert a substantial upward effect on the level of unemployment receives only limited empirical support.
One of the central roles of old-age pensions is consumption smoothing, but have a range of other objectives. People seek to maximize their well-being not at a single point in time, but over time. Someone who saves does so not because extra consumption today has no value, but because he values extra consumption in the future more highly than extra consumption today. Pensions enable a person to transfer consumption for their productive middle years to their retired years, allowing them to choose their preferred time path of consumption over working and retired life. A pension based on individual saving faces the person with the risk of outliving those savings. However, people could choose to agree to pool their savings together, with each person drawing a pension based on the total amount he/she had contributed to the pool and the group’s life expectancy. This is the essence of annuities, whereby an individual exchanges his pension accumulation at retirement for regular payments for the rest of his life. In a first-best world where nobody is poor on a lifetime basis, these objectives could be achieved by voluntary decisions and private insurance.
Lifetime distribution can be achieved by paying pension to low earners that are a higher percentage of their previous earnings (i.e. a higher replacement rate), thus subsidizing the consumption smoothing of lower earners. Pension systems can also redistribute across generations, or towards families. Pensions also have secondary goals, including economic growth. Excessive public pension spending contributes to high tax rates, putting growth at risk.
There are some problems concerning future consumption by setting aside part of current production for future use. Storing current production is costly in terms of the potential return to savings foregone, and also because storage costs for many commodities are high. A second problem is uncertainty. Third, some services can be transferred over time by storing the physical wealth that generates them; but it is not possible, even in principle, to store devices deriving from human capital, medical services being a particularly important example. The two most common ways of organizing pensions broadly are claims on future productions. So-called funded schemes follow the first; Pay-as-you-go (PAYG) or unfunded schemes the second.
The contributions to fully funded pensions are invested in assets, the return on which is credited to the fund. It is a method of accumulating financial assets, which are exchanges for goods at some later date. Those schemes have two major implications. First, a fully-funded plan always have sufficient reserves to pay all outstanding financial liabilities, which can be done by adjusting contributions to match projected liabilities or by having liabilities defined by available funds. Second, if there is no redistribution across generations, a generation is constrained by its own past savings; thus in present-value terms, a representative individual gets out of a funded system what he has put in – that is, with full funding, a generation is constrained by its own past savings.
PAYG are paid out of current contributions; the working population pays the pensions of the retired generation. A major implication of a PAYG system is that it replaces the constraint that the benefits received by any generation must be matched by its own contributions. Thus a PAYG system can redistribute across generations and can share risk across generations. Pensions also can be partially funded, paying benefits out of a combination of current contributions, returns on assets, and sometimes their sale.
Three approaches are common to answer the question how closely benefits are related to a worker’s previous contributions.
Defined contribution plans (DC). Each member pays into an account a fraction of his/her earnings, which are used to buy assets, which are accumulated in the account, as are the returns earned by those assets.
Defined benefit plans (DB). The plan sponsor premises to pay an annuity at retirement, which depends on the employee’s wage in her final year (of few final years) of work and upon the length of service.
National defined-contribution plans (NDC). NDC plans are conceptually similar to pure defined-contribution plans in that contributions are notionally ‘accumulated’ to determine a balance which is converted into an annuity at retirement, but different, in that they are not fully funded and may be entirely PAYG. NDC plans mimic funded defined-contribution plans by paying an income stream whose present value over the person’s expected remaining lifetime equals his/her accumulation at retirement, but with an interest rate set by government rules, not market returns.
IN a PAYG system, full pensions rights can be built up quickly, since pensions are paid no by one’s own previous contributions, but by those of the current workforce. In a fully contrasted it takes a long time to build up full pension rights, because it takes an individual many years to build up a lump sum sufficiently large to generate an annuity that will support hum fully in retirement. It is argued that PAYG pensions are sensitive to demographic change, but funded schemes are not.
The age-dependency ratio is
P/W
Where P is the number of pensioners and W the number of workers. It is often claimed that funding is insensitive to changes in the dependency ration, because a funded scheme always has sufficient resources to pay the pensions of the members, since the present value of a representative pension stream exactly equals past contributions plus interest.
First tier is generally run by the state, aimed primarily at poverty relief. Countries have
A non-contributory, tax-financed pension; or
A contributory pension aimed at poverty relief, with any of an array of different desings.
Second tier is mandatory, public, or private, intended to strengthen consumption smoothing and includes:
A publicly organized, earnings-related, defined-benefit pension;
An NDC system
An administratively cheap savings plan with access to annuities;
Mandatory, funded, defined-benefit pensions sponsored by industry;
Funded, defined-contribution pensions
Third tier is voluntary, defined-contribution pensions that are organized at the level of industry, the firm or the individual; regulation which is important and difficult.
At first glance, there are no major problems of adverse selection and moral hazard: people do not know when they are going to die, and suicide is costly to the individual, and works in the insurance company's favour. However, there is evidence of adverse selection in that people who buy voluntary annuities tend to live longer. Refining annuity rates to reflect the tendency for better-off people to be longer lived can solve the problem. There is also a potential for moral hazard, since people with large pensions can invest in health so as to live as long as possible. A more significant problem is that the funnel of doubt over the cost of providing a given annuity is increasing, raising the question of whether mortality rates are better thought of as a risk or an uncertainty. If the underlying problem is uncertainty, the likely result is to create unanticipated costs; these costs must fall somewhere. They can fall on the annuitant if annuities are missing, or offer poor value, or if retirement age rises alongside life expectancy, or costs can be shared more broadly, e.g. with the tax payer. Inflation is an uninsurable risk, which affects different pension arrangements differently. PAYG pensions can cope with inflation; they also allow pensioners to share in economic growth that occurs after a worker has retired. Funded schemes can cope with limited inflation, but face major problems beyond a certain level – i.e. it can offer indexation up to some pre-specified level. There is an efficiency argument, at a minimum, for state intervention to assist private pensions to address unanticipated inflation once pensions are in payment. The state is able to offer such a guarantee because it can use current tax revenues on a PAYG basis. Government can address uncertainty about future inflation by issuing inflation-indexed bonds.
The demand side also has to deal with information and behavioural problems. Choice is beneficial, only where the resulting welfare gain is larger than the cost of choice. In addition, the administrative cost of individual funded accounts rend largely to be a fixed cost, an thus bear more heavily on small accounts and in small countries with no economies of scale. Individuals are often poorly informed and have no idea what risks they face; pension pens are frequently complex and incompletely understood.
Behavioural problems arise due to bounded rationality, which leads to poor choices in a variety of ways,
Framing: choices are influenced by the way in which they are presented.
Familiarity: a common but profoundly mistaken choice is for a worker to invest heavily in the stock of his employer. Such behaviour shows a failure to understand the benefits of diversifying risk.
Excessive trading: many people trade too much, worsening their position and also incurring trading costs.
Immobilization: complexity and conflicting information can lead to passive behaviour, where people act like a rabbit in a car headlight. More options can result in lower participation.
and bounded will-power:
Short-term gratification: the purpose of pensions is to ensure retirement income, but some workers pay little attention to the future.
Procrastination: people agree that they should save more retirement but delay saving, do not save, or do not save enough. Similarly, people delay losing weight or giving up smoking.
Inertia: with rational choice, it should not matter whether someone is automatically enrolled in a savings plan or has to take explicit action to join. In practice, automatic enrolment leads to much higher participation.
Three sets of arguments are frequently made for funded pensions: that they provide better insulation against demographic change, that they increase growth rates, and that people regard their property rights as more secure if based on the ownership of financial assets. There are two potential links between funding and growth: funding may increase national savings and/or may improve the effectiveness with which capital markets allocate savings to their most productive investment use. A move to funding increases savings, which increases investment, which increases output by the marginal product of capital. These links often hold, but not always or necessarily. Funding may or may not increase national output. But even if it does, the policy does not necessarily lead to a Pareto improvement, since choices about funding inescapably have effects on the intergenerational distribution of income. Specifically, is it good policy for an economy to increase contributions or reduce benefits now in order to have lower contributions or higher benefits in the future. The PAYG mechanism makes clear both the quarrel over output shares and the dependence of pensioners on younger workers. Funding hides both issues, but does not remove them. It is mistaken that funded pensions are inherently safer. It is always open to government to interfere with investment decisions, or to impose outright nationalization or something close to it. Government can
Increase taxes on corporate profits, thus depressing stock values;
Increase taxes on income, including pensions, or on consumption, thus reducing the value of benefits to current and future pensioners;
Reduce any tax advantages for pension accumulations;
Reduce the value of government bonds through unexpected inflation.
Full funding imposes greater financial discipline; with PAYG, the state makes promises, but may not have to pay anything until later. In a fully-funded scheme, promises of higher future benefits must be matched by increased contributions immediately, thereby, it is argued, guarding against government failure.
Public provision of pensions solely to foster redistribution is justified only by a consumption externality, which the rich confer pensions on the poor as a merit good. Efficiency arguments arise out of the inability of private institutions to guarantee protection against inflation, giving efficiency justification for public involvement at least in indexing pensions, and possibly providing some or all of the pensions. It suggests that using publicly organized pensions for distributional purposes does not necessarily cause substantial efficiency losses.
Redistribution from rich to poor can, and usually does, occur with PAYG pensions. In many schemes there is a formula redistribution, in that individual B with half of the income of individual A generally pays half the contribution, but receives a pension that is more than half of A’s. It is possible to build some redistribution into a compulsory defined-contribution scheme. But, where membership is voluntary, the present value of the annuity received by a representative individual must equal the lump sum accumulated over his working years. This implies, ceteris paribus, that pensions must be proportional to contributions, ruling out systematic redistribution.
Redistribution from men to women is almost universal. Often, in private, defined-contribution schemes, a lower mandatory retirement age for women is inequitable in powerful ways: a woman has fewer years to accumulate pension wealth; and she has to live off that wealth for longer than a man both because in general she has retired five years earlier, and, in addition, because overall women have longer life expectancy.
The efficiency arguments for state involvement rest on externalities, justifying compulsion, and technical failure In the insurance market, justifying public provision at a minimum of some sort of indemnity against inflation. It is agreed that it should be compulsory for people to belong to a pension scheme, at least up to some minimum level, and that efficiency is enhanced where people are able to reallocate consumption over their lifetime. Decisions about consumption smoothing are therefore efficient if inflation can be ignored; but only the state can guarantee full indexation. It can be argued that the increasing role of governments in indexing public and private pensions has contributed to the relative certainty with which individuals can plan for the future, and has therefore increased efficiency, albeit imperfectly because a unified state scheme makes no allowance for differences between individuals in terms of degrees of risk aversion and preferences between present and future consumption.
Modelling the effect of pensions on labour supply is complex, and research came to different conclusions. Recent evidence paints a clear picture that badly designed schemes affect retirement decisions. Two potential distortions have to be considered: decisions about when to retire, and labour market responses earlier in life.
Any solution to the declining population of working age must reduce the demand for goods and services and/or increase their supply. This implies one or more of three outcomes.
Demand can be reduced
By increasing contributions, thereby reducing the average consumption of workers
By reducing benefits, either by paying lower monthly benefits or by paying unchanged benefits at a later age.
Alternatively, on the supply side, workers and pensioners can have the consumption they currently expect, so long as
In theory, raising output involves either or both of two strategies:
Increased output per worker can arise from more and better capital equipment and from improving the allocation of capital; and by raising the quality of the labour force through more education and training and improving the allocation of labour
Increased number of workers can arise from increased labour-force participation by those of working age; from an increase in the retirement age; and/or by importing labour.
In practice, supply-side policies in the face of declining workforce should therefore include some or all of the following:
Introducing policies to increase the quantity and quality of the capital stock, e.g. robots, and to improve the allocation of capital, e.g. through better capital market and tax policies;
Increasing investment in labour through education and training, and improving labour mobility to allocate labour more efficiently;
Increasing labour-force participation by reducing unemployment and by encouraging more women to join the labour force;
Raising the average retirement age;
Importing labour directly, through immigration;
Importing labour indirectly by exporting capital to countries with a young labour force.
Drivers of pension spending include increased longevity and declining fertility. The former makes it harder to finance pensions even in the absence of the latter; the main effect of population ageing is to make the problem worse. As well as being good macroeconomic policy, raising pensionable age is also good social policy. A significant number of people would actively prefer to work longer. And, the policy contains pension spending not by reducing the living standards in old age, but by reducing the duration of retirement. Finally, since adjustment to longer life expectancy could come either from lower consumption or from longer working life, it seems strange to adopt a corner solution by ignoring the latter option.
Four pension design options are pertinent:
A non contributory basic pension. This is also called a citizen’s pension or social pension, that is financed from taxation and paid at a flat rate on the basis of age and residence, rather than of contributions. The argument for such a pension is that they strengthen poverty relief: they can cover everybody and can pay a pension high enough to provide genuine poverty relief. It also has advantages in terms of gender balance, since it is women who tend to have the most fragmented contribution records. Other advantages are the incentive to work effort from a non-contributory pension, which is generally better than those of income-tested poverty relief; and the benefit is fairly well targeted, because age is a useful indicator of poverty; and non-contributory pensions can be made internationally portable on pro-rata basis. There is a range of instruments for making non-contributory benefits affordable, notably the size of the monthly pension; and the age at which the pension is first paid. It is possible also to subject the pension to an affluence test, such that the best-off do not receive the benefit.
Later and more flexible retirement. The argument for later retirement is the separate argument for more flexible retirement. The most efficient and equitable approach is to raise the average retirement age to accommodate aggregate resource pressures, but to accommodate differences across individuals by offering choice over how a person moves from full-time work to full retirement. Such choice and flexibility would be good policy even if there were no problem in paying for pensions. Any well-designed pension has the next four elements
An initial retirement age that makes it fiscally possible to provide a genuinely adequate pension;
A subsequent retirement age that increases in line with rising life expectancy in a way that is rational and transparent, so that people know how long in advance broadly when they will be able to retire;
A flexible labour market that allows people to move from full-time work towards full retirement along a phased path of their choosing;
Public understanding of the simple economics of pensions.
The design of simple, low-cost saving plans. Reduced choice can be welfare enhancing. Literature suggest a number of lessons: voluntarism plus public education are insufficient, automatic enrolment is more effective; keep choice simple; design a good default option for people who make no choice; it is desirable to decouple account management; and in voluntary plans a further option is to allow people to commit now to (reversible) action in the future, thus making use of procrastination to assist policy.
NDC pensions. These plans have range of potential advantages. They are simple. They are centrally administered, hence they have low administrative costs. They do not require institutional capacity to manage funded schemes. They avoid much of the volatility of capital markets. And partially funded NDC schemes can morph over time into simple individual accounts, e.g. Thrift-Savings-Plan. Also, risk can be shared among workers and pensioners currently in the system.
Two distinguish approaches are the starting point in identifying the poor:
Categorical schemes stress the causes of poverty, and institute programmes for specific groups. Non-categorical schemes, in contrast, regard recipients as a spectrum that includes the self-supporting, the very poor, and the large numbers in-between. Such schemes concentrate on outcome rather than cause, and classification is made only in terms of need. The approach is attractive because there are few gaps through which difficult cases can fall, but has the disadvantage of requiring a mean test in one form or another.
Figure 9.1 shows the costs of negative income tax. The simplest negative income tax relates the individual’s tax bill, T (positive or negative), to his income, Y, as
T = t(Y – B)
Where t is the tax rate applied above and below the break-even income, B, as shown by the line G0BA in figure 9.1.
T = tY – G
Where G= tB is a lump-sum transfer to the individual, who then pays tax at a rate, t, on all his other income, Y.
A scheme that gives everyone a lump sum and taxes all other income goes under generic name of guaranteed-income scheme. Guaranteed-income schemes and negative income taxes are analytically equivalent.
Negative income tax has attracted widespread support: from libertarians because cash transfers are compatible with market allocation, and so can loosen what they see as the stranglehold of in-kind transfers under the welfare state, and from socialists because the system guarantees everyone at leas a minimum income as a right of citizenship and without a mean test. Additionally, it is claimed that negative income tax will boost take-up, reduce stigma, and by bringing all individuals under a single umbrella, enable government to be better informed about the economic condition of the population. Possibly the most important claims are that negative income tax can solve the poverty trap and increase the redistributive power of the welfare state. However, implementing a large-scale scheme is costly. The negative income tax may not increase taxation per se so much as replace implicit by explicit taxation. There are two possible cases. In the first, both the level of benefits and their coverage remain unchanged under negative income tax (table 9.1). Identical recipients receive identical benefits; total net expenditure is unchanged; no increase in net revenue is necessary; and nobody’s post-transfer income has changed. Negative income tax has produced an exact mimic of the previous system by a different administrative mechanism: nothing has been done to poverty relief; and the poverty trap and labour-supply disincentives are unchanged. A second case arises where net benefits and/or coverage are increased by negative income tax. This raises the net income of at least some poor people, but net expenditure rises, necessitating an increase in taxed over and above the replacement of implicit by explicit taxation.
Suppose the following hypothetical states in which benefits are paid only to the poor – state A – or to everyone – state B – then
Total tax revenue: gross expenditure, and hence on a revenue-neutral basis also gross tax revenues, must be higher in state B than in A.
Average tax rate (ATR): if tax revenues are greater in B, then ATR must be higher.
Marginal tax rates (MTR): if ATR is higher, then MTR must be higher for at least some groups. Either the increase in MTR is concentrated wholly on the potential labour-supply problem for the poor; or the increase in MTR is concentrated wholly on the rich, in which case the poverty trap is solved, but there is a potential labour-supply problem for the rich.
Thus, negative income tax inevitably increases explicit tax rates.
The cost of ant scheme is the tax/benefit function weighted by the income distribution:
Appendix 10
t can be thought of as the marginal rate applied to all income, or as a weighted average of the dual rates, or, more generally, as the average of any set of marginal rates with the same tax yield. In a scheme that keeps all benefits in the hands of the poor, the more successfully the poverty is alleviated, the more serious is the problem of the poverty trap. The solution to the poverty trap is to withdraw benefits more slowly, in which case the break-even income rises above the tax threshold. Besides, for a given guaranteed income, the poverty trap can be alleviated only by increased taxation of the rich. This risks labour-supply problems. Also, for a given level of benefit it is the shape of the income distribution that creates the poverty trap – changing the system of cash transfers will not solve the problem so long as benefits are conditioned on income. Negative income tax thus increases tax rates at some or all levels of income: it institutionalizes the poverty trap and/or leads to higher tax rates on incomes above the poverty line.
Negative income tax has implications for both horizontal and vertical redistribution. A large-scale scheme paying benefits equal to the poverty line may improve horizontal efficiency in two ways: it can increase take-up; and it can reduce stigma because on system covers everyone. On vertical efficiency, negative income tax scores less well. At first glance it might appear that, by benefiting the poor at the expense of the non-poor, it is strongly redistributive. However, this is not necessarily so. The cost of negative income tax depends on the distribution of income. So too does the redistributive strength of any particular scheme, since a 1% increase in the average tax rate raises more revenue for distribution to the poor the greater the proportion of taxpayers who are non-poor.
Since income distribution is typically heavily skewed towards low incomes, negative income tax will not be a strong re-distributor for two reasons:
As benefits are increased, the average tax rate has to be raised to finance them;
There are many people with low incomes and relatively few with high incomes, so that the tax increase necessary to finance a high guaranteed income therefore falls substantially on those with low and average incomes, clawing back a substantial proportion of the benefit to the poor.
Given the shape of the income distribution and the current benefit levels
There are three potential ways forward:
Small-scale negative income tax in combination with other forms of cash benefits;
The indicator targeting approach;
A combination of both approaches.
The first will not solve poverty, and, therefore, contributes to poverty relief only if it makes other schemes more effective. A key feature of a universal negative income tax is not that it would help the poor directly (because the guaranteed income would be small), but that for the first time most of the population would be under the tax authority’s umbrella, making it possible to search out the poor and hence to increase the take-up of other benefits. The strategy is to use one not to solve poverty, but to enhance the effectiveness of other forms of poverty-relief. The guaranteed income being universal, all other cash benefits become devices only for topping up income, rather than mechanisms for total income support. Also, fixed-period awards would further reduce the poverty trap.
The only feasible role for a pure negative income tax is not as a major redistributor of income, but as a search device to discover low-income families, in order to increase take-up rates. For this purpose, the guaranteed income need be no higher than the cashed-out value of current income-tax allowances; thus, costs are not excessive, reducing problems with the poverty trap.
Empirically, the major correlates of poverty are unemployment, ill-health, large families, single-parent families, old age, and high housing costs; in developing economies, geographical location or being landless may also be highly correlated with poverty. How exogenous are these characteristics:
The endogeneity of the level and duration of unemployment is controversial;
With ill-health, the problem is less acute because genuine ill-health us costly to the individual, and policing is possible through certification procedures;
The incidence of large families is exogenous with respect to benefits unless family size is influenced by the existence of family benefits.
Reaching retirement age is entirely beyond the individual’s control;
Housing costs are far from exogenous under any system that, like housing benefits, bases assistance on actual costs. Matters could be improved if assistance were related to a regional index of average housing costs.
Benefits conditioned entirely on an income test generally run into problems of high tax rates, creating problems of disincentives and administrative costs. Schemes that award benefits only on the basis of indicators of poverty face problems of gaps in coverage, benefits leaking to the non-poor, and problems with defining and administering borderline cases. The most promising way forward is through a mix of the two approaches.
The basic income scheme has several variants. The most promising pays a guaranteed income roughly equal to the cashed-out value of income tax allowances, superimposed on which are social insurance benefits. Some versions include desirable characteristics that benefit is invariant to age, sex, marital status, and employment status, thus minimizing distortionary incentives. The scheme can be regarded as a universal negative income tax whose intercept varies with the characteristics of recipients, and, thus, as a blend of the two approaches.
The participation income scheme is a negative income tax arrangement – but one in which benefit is conditioned on some form of participation. The scheme pays a participation income to everyone aged 18 or over who participates in one of more ways; and it pays a basic income to all children in the place of child benefit. The participation income and social insurance complement and reinforce each other – an explicit mix of the negative income-tax approach and the indicator-targeting approach, with the advantages of both, but with few of the disadvantages.
Working tax credit schemes include the UK working tax credit and child tax credit, and the US earned income tax credit. Such schemes are a mixture of income testing, indicator targeting, and self-targeting. Thus, tax credits can be regarded as a negative income-tax schemes, but one that applies only to lower earners.
A combination of non-contributory pensions and child benefit can be thought of either as a form of negative income tax / guaranteed income apart from working-age adults, or as indicator targeting based on age.
The primary objective of health policy is to improve people’s health, which is to reduce mortality and morbidity. Health derives from many sources, including
Overall living standards
Individual choice and lifestyle
The general external environment
The individual environment
The quality and availability of health care
A person’s inheritance
Allocative efficiency is concerned with producing the quantity, quality, and mix of health interventions that bring about the greatest improvement in health. External efficiency relates both to the overall size of the health sector as a proportion of GDP (macro-efficiency) and to the way resources are divided between alternative uses within the health sector (micro-efficiency). Productive efficiency (or internal efficiency) is concerned e.g. with running medical institutions as efficiently as possible. Equity is more elusive. We can distinguish four definitions of equity in consumption:
Equality of utilization implies that everyone in a given condition should receive the same quantity and equality of outcome implies an unequal allocation such that everyone enjoys an equal state of health or level of educational attainment. Whether or not such an aim is thought desirable, it is not fully feasible. Equity needs to be defined as a form of equality of opportunity, which means that any individual should receive as much health care as anyone else in the same medical condition, regardless of any factors that are thought to be irrelevant – e.g. income. The same argument applies to education.
Measuring the benefits faces three major problems:
Health is hard to measure. Expenditure on health care can be estimated, but health itself (i.e. the outcome) is hard to measure except in terms of broad indicators such as infant mortality, life expectancy, and estimates of the burden of disease.
Causality is complex. The influence of intangible factors is crucial, but impossible to measure. Similarly, improvements in health outcomes may be due to improved diet, reduced smoking, etc. if such factors are ignored, we will tend to overestimate the benefits of health care.
Improved health is hard to value.
In addition, there are intractable problems in measuring the utility benefits arising from any reduction in the physical and emotional suffering of the patient and his/her family. Consequently assessing efficiency is problematical. There are three approaches.
Cost-benefit analysis.
Cost-effectiveness analysis.
Cost-utility analysis.
The last one does not look at the health benefits of treatment but at the utility benefits, based around the idea of a quality adjusted life year (QALY). QALY starts from the premise that the outcome of medical care should be measured in terms not only of the quantity of extra life it produces, but also of its quality. They have the merit of incorporating subjective values explicitly and systematically. It looks at the extra tears resulting from treatment, and divides them by its cost.
An important determinant of health is relative poverty, as opposed to absolute poverty, in part because less well-off people suffer more stress and hence more disease than wealthier people. In addition, poor people will tend to have worse diets and less good access to health care. The benefits of health care perceived by poorer people (MPVP) may be lower than those of the rich (MPVR) (e.g. if the poor have worse information); or the poor might rationally place a lower value on health; or the actual benefits to poorer people might be lower if doctors treat them less effectively than middle-class patients. In addition, the poor face higher costs of health care. See also figure 10.1.
Appendix 11
The system cannot be regarded as a failure unless an alternative system is more equalizing. Also, whether or not the health care system is an equalizing force depends not only on the distribution of benefits but also on whether expenditure is discussed not in isolation, but, more properly, in conjunction with the taxation, which finances it. In practice, measurement raises almost insuperable problems, inter alia because of the difficulty of measuring the incidence of taxes and benefits, but the logic is clear. If (rich) individual R contributes on average twice as much as (poor) P in whatever tax is used to pay for health care, but receives the same quantity of health care, then the system redistributes from rich to poor (i.e. is progressive). But if R contributes twice as much but consumes four times as much as P then the system of health care is said to be regressive.
With medical care:
Much (though not all) the information is technically complex, so that a person would not necessarily understand the information even were it available;
Mistaken choice is costlier and less reversible than with most other commodities;
An individual generally does not have time to shop around if his condition is acute (contrast the situation with a car repair);
Consumers frequently lack the information to weigh one doctor’s advice against another’s;
Emotions can be strong.
Rational choice requires simultaneous knowledge both of prices and of the nature of the product (i.e. of both budget constraint and indifference map); knowledge of prices without adequate information about different types of treatment will not ensure efficiency. Next to imperfect information, there are also the issues of bounded rationality and incomplete contracts. Consumers will choose inefficiently, and it is not clear whether the result will be under- or overconsumption. If the true marginal private valuation (figure 10.2) us shown by the MPV curve, consumer ignorance can result in demand curves D1 (under-consumption) or D2 (overconsumption). In addition, where knowledge and power are systematically related to socio-economic status, there is also inequity.
An important question is whether the private market can supply medical insurance efficiently. The probability of requiring treatment is independent across individuals except during major epidemics. Also, the probability of requiring treatment for a particular type is less than one for ailments like appendicitis or a broken leg. But the condition fails for chronic medical problems arising before a policy is taken out. Besides, the probabilities relevant to medical insurance are generally estimable. However, problems arise with policies whose benefits are long time in the future, clouding knowledge of the relevant risks. Finally, problems arise of both adverse selection and moral hazard. Moral hazard arises through the third-party payment problem. The problem arises because
Patient and doctor both face zero private costs, even though the social costs of health care are positive and frequently large, and both have an incentive to consume all health care that yields any private benefit. The doctor is an agent for two principals – the insurance company and the patient. As a result, the relation between health care and health outcomes is so loose that performance guarantees cannot be given to either principal; this is a kind of information symmetry that faces both ways and that is perhaps even shared by the agent him- or herself. An additional problem is transaction costs. If these are too high, some risk-averse individuals will choose not to insure. This is inefficient if an alternative system could avoid them. Market solutions or different forms of intervention are solutions to the problem. However, there is no complete solution to the moral hazard problem.
Their motives and their economic environment determine doctor’s behaviour jointly. Doctors may not face the costs of their decisions, but are reimbursed by insurers. There is no budget constraint facing doctors individually; and on aggregate they are constrained only by the willingness of consumers or employers to pay insurance premiums. These arguments point towards oversupply of health care in total. In addition, the composition of that total may be distorted in favour of glamorous areas. A solution for non-competitive behaviour is at a minimum regulation required on standards, plus perhaps some regulation of prices and possibly also monitoring doctors’ activities. Another potential solution is the libertarian approach of removing entry barriers to medical practice, thereby enhancing competitiveness on the supply side and largely removing the need for state intervention. Supply-side deviations from competitive behaviour may cause inefficiency; but removing restrictions to competition is unlikely on its own to improve matters.
Lower-income individuals may have less information relevant to choices about health; in addition, they may be less able to make use of any information they acquire. In such cases intervention in the form of regulation or subsidy may improve equity as well as efficiency. Alongside efficiency arguments for public production and allocation, there may be equity motives making it politically easier to make transfers in kind. The rich may favour them for selfish and altruistic reasons; and the poor may prefer them, either because the in-kind transfer is more generous than the cash offer, and/or because they feel less stigmatized by receiving benefits in kind than through means-tested cash transfers. Then there is also the role of giving, such as being a blood donor. The marginal social cost of blood is likely to be low and may be zero. In that case giving might be both morally superior and efficient.
Types of intervention:
Market production and allocation (with or without income transfers);
Public production and allocation;
Intermediate strategies.
Pure market provision. Government intervention destroys the fit between demand and supply, and the destruction is greatest if intervention takes the form of public production. However, health care comes nowhere near to conforming with the first-best assumptions, so that unrestricted market allocation is not a theoretically promising approach. Uncorrected externalities will lead to under- or overconsumption.
Mixed public/private involvement. The desirability of such a package depends on two factors: would it be more efficient and equitable than other methods; and might it be politically more acceptable than national health service-type arrangements? The supply of medical treatment in a private market is crucially influenced by fee for service and third-party payments. The latter one creates a divergence between private and social costs and benefits, and hence cause a particular kind of externality. We can deal with this externality by regulation, or, alternatively, it is possible to internalize the externality by merging the activities of doctor and insurance company, thereby forcing doctors to face the marginal cost of the treatment they prescribe. Also, finance could be organized in on of two generic ways. One possibility is private finance plus residual state finance. But, two difficulties arise: non-insurable risks and the poor. The former include congenital and chronic health problems, the medical needs of the elderly, undertake to family doctors and pregnancy. The state could deal with these problems either by subsidizing private insurance or by paying for treatment, through a residual public insurance scheme or out of tax revenues. The poor could be assisted similarly. However, there is the problem of defining borderlines; policing would be necessary to prevent oversupply; and the poverty trap could arise for the poor. The other option is public finance, where the state pays medical bills from tax revenues or through social insurance. In comparison with actuarial insurance, the advantages of social insurance are twofold: compulsory membership makes it possible with no substantial efficiency loss to gear a premium to ability to pay rather than to risk; and universal coverage avoids problems with borderlines.
Public production, allocation and finance. Consumer sovereignty is appropriate where information is sufficient for rational choice, a process that may be assisted by regulation of quality. The patient’s information is often so imperfect that the individual consumption decision is best made on his behalf by an agent. The third-party problem can also justify public production as a method of controlling the resulting large and inefficient increases in the output of health care. In theory, the strategy is feasible and both efficiency and equity terms.
The macro-efficiency aim is to devote the optimal fraction of GDP to health care. Thus cost containment is not per se a primary objective. However, it has become an important instrumental objective in most countries, given upward pressures on medical spending from a variety of causes.
Demographic factors. The increasing number of older people intensify demand.
Technological advance. Costly new techniques have led to increased possibilities for medical intervention and increased expectations.
Microeconomic incentives. Moral hazard arises where treatment is paid for largely or wholly by the insurance company, and so is free both to patient and doctor. The resulting distortion leads to inefficiently high medical spending.
The relative price has raised the cost of health care by more than the average increase in prices.
Policy-makers wishing to encourage a particular outcome, in this case containing rising medical spending, have two potential instruments:
Incentive-based mechanisms include
Preferred providers mechanism. This mechanism seeks to contain demand. A different approach proceeds via the supply side. E.g. insurers can restrict patients to certain preferred providers, who face competitive pressures to retain approved status.
Prospective payment mechanism. The idea is like any cost-plus contract, open-ended, retrospective reimbursement gives suppliers no incentive to economize, thus imposing the entire task on the payer. If suppliers are paid ex ante they face strong incentives to use resources carefully. Though the logic is appealing, the benefits should not be overstated: prospective payments are a form of price control but expenditure depends on price and quantity: thus controlling price is no automatic guarantee of controlling expenditure.
The other potential instrument is regulation. Publicly organized medical insurance faces the same problems of moral hazard as private insurance. The logic is simple: total spending = price * quantity. Successful cost containment must control total spending directly, or control price and quantity; or use price control to reinforce an overall spending constraint.
The optimal quantity of school education is represented by the quantity X* in figure 3.1 and is subject to certain efficiencies: Allocative efficiency, macro-, and micro-efficiency, and productive efficiency. Child A and B should have an equal opportunity to acquire an education of equal quality and duration, irrespective of e.g. their background.
There are some problems regarding the measurement of the benefits of education:
Output cannot be measured. Output can only be measured in terms of test scores.
It is problematic to connect inputs and output. This is because there are problems measuring inputs; output can only be measured in terms of test scores; and the production function is hard to estimate.
Causality cannot be established. There is no certainty in the causal link between post-primary education and increased individual productivity.
The human capital model attempts to explain the demand for education in terms of its production and utility benefits. The latter benefits arise because education might have consumption benefits in the present as well as investment benefits in the future. The individual return to education also includes non-money rewards such as job satisfaction and the enjoyment of leisure. The initial assumptions of the simplest human capital mode are:
Education raises the individual’s marginal product in the future and therefore his future money income;
This increase in money income is the only benefit from education – i.e. initially we rule out consumption benefits and future non-money returns.
Appendix 12
The screening hypothesis argues that education is associated with increased productivity, but does not cause it. It also argues that education beyond a basic level does not increase individual productivity, and that firms seek high-ability workers but are unable, prior to employing them, to distinguish them from those with low ability (adverse selection). There are various counter-arguments. Where education includes professional training, there is a direct contribution to productivity. In practice, skills and job characteristics are heterogeneous, so that it is necessary to match workers and jobs, giving education and additional social return as a matching device. It is possible to measure the money income benefits – but not the utility benefits – associated wit different levels in education, but the causal relationship is less clear.
Estimates of the private rate of return are suspect because, of necessity, they omit non-money returns. Estimates of the social rate of return are doubly suspect: they omit non-money returns and can take account of screening only very approximately. Because of the tax-dividend point, there is an unarguable external benefit, but it is not possible to show how much.
Is education financed progressively? Education is financed progressively if (rich) individual R consumes (say) twice as much education as (poor) individual P, so long as R pays more than twice as much as P in taxes that pay for education. However, redistributive effects of publicly funded higher education are generally regressive.
Does free schooling equalize access? The idea of tax-financed education is that access should not be constrained by family income. Even in the absence of fees, school attendance typically has costs. In addition, children may drop out of school not for lack of ability, but because their family needs their earnings. School education after the minimum school leaving age is not free, but imposes a cost in foregone earnings that bears disproportionately on lower social-economic groups.
Early child developments matters; it follows, second, that parenting matters as well as education; third, to an important extent, early child development is a one-and-for-all event – it is hard to make up later for earlier deficiencies. The distribution of education does not only affect the individual concerned, but also has generational effects. Differential access to education could in principle be cause entirely by exogenous differences in tastes and abilities, in which no issue of inequality arises. Also, education is not the only driver of economic growth. Also, a different system of education would not necessarily be more egalitarian, in which case egalitarian goals must be pursued by other means.
Children (the immediate consumers) are often not well-informed about the nature of the product, prices, and about the future. Also, parents are likely at best to be only partially informed about such things as the quality of teaching, the characteristics of the other children in the class etc. Consumers of education are likely to differ in the extent of their confidence and articulateness. There is also the problem of missing markets and incomplete contracts. The complexity and multidimensional nature of education creates particular difficulty in defining the product. Thus it is difficult to specify contractually what a school should deliver. The market of education finance is incomplete: from an efficiency point of view we do not normally worry if an individual cannot afford more than x units of a good, but there are losses when an individual cannot afford to buy the socially efficient amount of education. The state could intervene by creating a framework for privately financed loans; it could provide loan finance itself; or it could subsidize education. Education is not a public good; nor does it generally face increasing returns to scale.
Horizontal equity. Parents with little education may have less information about school choice than better-educated parents. In addition, they may be less able to make use of any information they do acquire. In such cases intervention may improve equity as well as efficiency. Regulation would be concerned with the professional qualifications of teachers, minimum physical facilities, school attendance, and, possibly, curricula. Where imperfect information causes underconsumption, a subsidy might be applied to either incomes or prices. Income subsidies usually take the form of cash transfers, but education vouchers can be though of as a form of tied transfer.
Vertical equity. The rich may have an interest in the education of the poor for two reasons. They might support transfers of school education for reasons of efficiency / self-interest: a well-educated workforce might foster economic grow, and/or might reduce social unrest. Alternatively, rich individuals might care about the distribution of education for altruistic / equity reasons. The efficiency arguments for public production and allocation may be reinforced by equity motives that make it politically easier to make transfers in kind. The rich may favour them for selfish or altruistic reasons; and, if the offer is sufficiently generous, the poor might also prefer in-kind transfers.
Types of intervention:
Pure market provision. The failure of many of the assumptions necessary for market efficient is an argument against market production and allocation. It should be clear that a pure market system is likely to be highly inefficient, and also inequitable to the extent that knowledge, power, and access to capital markets are correlated with socio-economic status. Unrestricted market provision of primary and secondary education is theoretically implausible, and, in practice, does not exist in any country.
Mixed public / private involvement. The state regulated education and subsidizes it wholly or in part, but production takes place in the private sector. This is how it goes in most countries.
Public production, allocation, and finance. It is questionable whether public production is, or is not, likely to be more efficient than the private market. The a priori arguments for state intervention in school education are more finely balanced than those about health car. But it is fair to say that the failures of the first-best assumptions are sufficiently great to make public production, allocation, and finance of school education a tenable strategy. To the extent that this is the case on efficiency grounds, it is appropriate to finance education redistributively for reasons of social justice. Since education is not a homogeneous whole, it is necessary to ask whether the same answers apply to all types of education.
There are at least two strategic sets of arguments for spending more on education than in the past.
Technological advance mostly increases the demand for skilled workers and reduces the demand for unskilled workers.
Demographic change creates another argument for spending on education. The rising proportion of older people in many countries presages high spending on pensions and other age-related activities such as medical and long-term care. The solution is to increase output sufficiently. If the problem is that workers are becoming relatively more scarce, the efficiency response is to increase labour productivity.
The idea of education vouchers is to increase choice and competition. There are several variants of the idea.
The latter scheme places more emphasis on distributional goals, in terms of both education finance and schools’ choice of students. By giving parents more choice and strengthening competition, voucher schemes, it is argued, increase efficiency and reduce the risk of political indoctrination. Proponents also claim as an advantage that voucher arrangements avoid the situation where parents who send their children to private school receive no compensation for the costs from which they thereby relieve the state. Opponents state that their efficiency advantages are debatable and their equity effects almost certainly deleterious.
The idea of quasi-markets is to retain public funding but to increase efficiency and responsiveness by introducing market forces within the state system. Underlying idea is to improve efficiency by increasing the accountability of educational system.
It is useful to draw out the different strands of quasi-markets as they apply to education.
Competition. In a competitive regime, schools face a similar incentive to cream skimming. We know that performance is largely determined by socio-economic background, and schools in greater demand can attract more pupils, and hence more funds.
Greater management autonomy for schools. This gives schools more freedom to manage their budgets, allowing them to respond to competitive pressures. Also, if schools are allowed some control over the number of students they accept and the funding formula includes a capitation element such that a pupil signing on triggers a payment to the school, then schools that attract more pupils receive more resources. This implicit voucher arrangement means that school finance is determined in part by parental choice rather than bureaucratic decision.
A national curriculum and national testing. The former can be regarded as regulation education markets to increase equality of opportunity by spreading vest practice and also a benchmark against which national test can operate, thus giving parents information on the performance of individual schools.
So far as school education is concerned, many of the first-best assumption fail, the main problems being imperfect information, incomplete capital markets, and external effects. From an equity point of view, the most important problem is that knowledge about the operation and value of education is likely to be correlated with socio-economic status. Substantial public involvement is therefore essential, although with room for discussing about the theoretical arguments for public products, as opposed to regulation and finance.
Higher education contributes to national economic performances and to the promotion of core values. Therefore, it has a significant cultural dimension that will vary across people within a country, across countries, and over time.
Achieving the primary objective – better educational outcomes – rests on secondary objectives: using resources efficiently and distributing them equitably.
Efficiency:
Allocative efficiency. The totality of resources devoted to education (macro-efficiency) and the division of the total between tertiary and earlier education, the division of resources between further and higher education, the division between subjects and universities, and the spending on universities and on student support (micro-efficiency).
Productive efficiency. The factors as quality of university management.
The specific aims of higher education policy
Freedom.
Intellectual freedom and diversity are core value.
Pursuit of knowledge for its own sake is an objective, partly for intrinsic reasons, but also because it is often unclear in advance whether a particular line of research might have practical benefits.
Quality. Strengthening the quality of teaching and research is important not only or intrinsic reasons, but also because of the economic importance of higher education.
Access. Widening participation by students from disadvantaged backgrounds has several justification, most of it being for equity reasons
Size. It is sometimes argued that higher education is expensive and that a small system might be a more efficient use of resources. The counter agreement is that human capital is even more important a determinant of economic competitiveness than in the past.
Issues regarding information:
Information about higher education is available, and more can and should be made available. There are plenty of university guides and universities publish much information online.
Students can, for the most part, understand and evaluate that information. The process is easier because going to university can be anticipated so that the student has time to acquire the information and advice he needs.
Through the cost of mistaken choice can be large, it is not clear that a central planner would do better. Students are more capable than central planners of making choices that conform with their own needs and those of the economy.
Issues regarding perfect competition and market imperfections:
It is reasonable to regard most higher educations as competitive, an argument reinforced by international competition and distance learning.
Higher education is not a public good, not are there increasing returns to scale.
It is efficient and equitable that the costs of higher education are shared, that is, there should be taxpayer subsidies but also tuition fees. This because externalities arise.
Why student loans? In a first-best world, costs should be shared between taxpayers and the beneficiary; and individuals would finance their share by borrowing against their future earnings. Another reasons for student loans is a head-on collision between the two competing imperatives. On the one hand is the need for more resources to finance large, high-quality systems of higher education. On the other are constraints on public finance arising from competing demands, including ageing populations and spending on health, and from increasingly competitive global economy which limits a country’s capacity to increase taxes.
Loans can be organized in different ways:
Mortgage-type loans. These have fixed repayments of €X per week for a fixed duration.
Income-contingent loans. Repayments take the form of x percent of repaid. Thus, the duration of repayment is variable.
Graduate tax. Repayments continue for life, or until retirement. High-earning graduates may therefore repay more than they have borrowed.
Why not conventional loans? When investing in human capital, there is no physical collateral. This is because of the incomplete market for student loans.
When people borrow for a degree:
They may be imperfectly informed about the nature of the product, if from a family with no graduates;
There is a high risk of failing the degree;
Though the average private return to a degree is positive, there is considerable variance around it;
There is no security for the loan. Thus is the borrower’s income falls unexpectedly, there is no option to sell the degree and repay the loan.
Higher education is free at the point of use, and the loan is designed with in-built insurance against inability to repay, addressing the demand-side capital market imperfection – the uncertainty to the individual about the return to a degree. In such world, government intervention on efficiency grounds has addressed liquidity constraints, and there are no equity grounds for further invention. Intervention to promote access to higher education for equity reasons is justified largely by inequality earlier in the system; and that much of the resulting intervention should take place during earlier education. In addition, if earlier impediments result in some people acquiring an inefficiently small account of human capital, intervention is justified also in efficiency terms.
Competition in higher education is beneficial; and decisions should be taken not through central planning, but as a result of choices by the major stakeholder – students, universities, employers, and government. There are two caveats:
However, the best way of organization is competition with a suitable regulation is the most beneficial.
Core characteristics of student loans:
Loans should have income-contingent repayments, which increase efficiency by protecting borrowers against risks that would otherwise lead to inefficiently low levels of investment, and improve equity, since their inbuilt insurance against inability to repay assists access. And if repayments are collected alongside income tax, such loans also protect the lender from much risk of making an unsecured loan.
Loans should be large enough to cover fees and living expenses, making higher education free at the point of use. The loans are a device to assist consumption smoothing.
Loans should attract an interest rate broadly equal to the government’s cost of borrowing.
Mistaken arguments about tertiary education are:
Higher education is a basic right and should therefore be free. The equity objective is not free higher education, but a system in which no bright person is denied a place because he/she comes from a disadvantaged background.
It is immoral to charge for education. Making something free for everyone can be justified in efficiency terms where market failure make consumer choice problematic and in equity terms where the commodity is consumed by everyone. As a result, taxpayer subsidies are regressive both directly because they crowd out resources that should be used to widen participation.
Elitism has no place in higher education. Intellectual elitism is proper and desirable. The equity objective should be a system in which the ability of the brightest students to study at the most intellectually demanding universities is unrelated to their socio-economic background.
Graduates pay for their higher education through their higher tax payments. Income tax raises only part of government revenue and is paid by many more non-graduates than graduates. Also, there are limits to taxation, not least because of political pressures, which collide with other priorities for public spending. Thus it is no accident that real funding per student declined in many countries as student numbers increased.
Tax finance promotes access. Free higher education not only fails to widen participation, but also uses taxpayer resources to finance what is largely an activity of the better off.
Price signals are useful in education, improving efficiency, and, through competition, making the system more responsive to student and employer preferences.
Demand. Resources are misallocated if students face no price signals.
Supply. Fixed prices also have adverse effects on the supply side. Price ceilings erode incentives to improve quality; price floors erode incentives to increased efficiency. Flat fees, including zero fees, are both a floor and a ceiling, and thus particularly inimical to efficiency gains.
Variable fees have twofold effect:
They bring more resources
Through competitive pressures, they create incentives to use those extra resources efficiently.
Variable fees are fairer:
Redistribution between individuals. Variable fees introduce higher charges for those who can afford them and redistributive polies help poor people to pay those charges.
Redistribution between institutions. Variable fees facilitate redistribution between institutions. With flat fees or tax finance, the volume of resources going to the sector is fixed by government, so that prestigious universities and local institutions compete for the same pot of money in a zero sum game. Variable fees start to address this gridlock.
Directly fairer. In countries with a diverse higher education system, charging everyone the same fee is both unfair and inefficient.
Potential discrimination against home students. Many countries control fees for home students but allow greater freedom for overseas students. Variable fees, by reducing or eliminating the price differential, avoid such discrimination.
Three lessons should be pondered
Fees relax the supply-side constraint. The finance of higher education faces a paradox. Large taxpayer subsidies can create supply-side constraints because of the desire to contain public spending. Where qualified students have no automatic entitlement to a place, the constraint takes the form of a view about student numbers. The result can be a high-quality system, but one that turns away qualified applications. Unless limited taxpayer finance is sufficiently redistributive, however, students from lower-income backgrounds will be deterred from applying. Thus, high subsidies can harm access on the supply side, but their absence can harm it on the demand side.
Big-bang liberalization is politically destabilizing.
No liberalization is also a mistake. In that case quality suffers because the education budget has to compete with other budgetary imperatives. Also, within the education budget, universities compete with nursery education, school education, and vocational training. Consequently, real funding per student declines, and access also suffer. If places are scarce, it will tend to be the middle-class students who get them; and if they are not scarce, the need to finance a mass system typically means that resources to widen participation are limited.
From experience, four lessons are learnt:
Income-contingent loans do not harm interest.
Interest subsidies are expensive.
Positive real interest rates are feasible. With income-contingent loans a higher interest rate does not increase a graduate’s monthly repayments, only the duration of the loan.
Contract design is important. International labour mobility is high and likely to increase, raising question about potential default if a person immigrates.
Evidence suggest three central cause of failure to participate in higher education:
Prior-attainment constraint;
Constraints related to information and aspirations;
Liquidity constrains: for most students well-designed income-contingent loans address this constraint, though additional action might be necessary for some students.
The first two factors can interact.
What policies to widen participation?
Policies to improve attainment. Relevant policies include
Fostering early child development, including action to increase the quality and availability of nursery education;
Action to improve primary and secondary education outcomes;
Policies to prevent high school dropout.
Policies to improve information and raise aspirations. Relevant policies include
Bringing the universities to school.
Additional support for schoolteachers who advise young people on potential pathways.
Better online assistance: well-designed expert system can help to prompt potential applicants – and their advisers – to ask the right questions, and to help them sift through large amounts of information.
Policies to address credit constraints. Relevant policies include
Financial support to encourage people to complete high school;
Better information about available sources of finance;
Loans with income-contingent repayments, large enough to make higher education free at the point of use, and with targeted interest subsidies;
A system of grants for people who underestimate the benefits of higher education, and, hence, are not prepared to take out a loan, even one with income-contingent repayments;
Financial support for part-time study.
Policies to address genuine debt aversion. Correct analysis distinguish at least three reasons why someone might be reluctant to borrow:
Lack of prior attainment. Money intended to widen participation should be spent mainly on raising school grades, preventing drop out, and improving information. The error is to attribute to the credit constraint behaviour that is determined mainly by the attainment constraint.
Lack of information and low aspirations. The primary policy to address the problem is to improve information. there is also a case for full scholarships to finance a student’s first year, on the basis that once a student has successfully completed a year, he/she is likely to be well-informed both about process and outcome, and, hence, prepared to take out a loan for the rest of the degree.
Risk aversion. This is more an information problem than a debt aversion. The primary response, again, is to improve information, and where that fails, to pay grants.
Expanding options for part-time study.
What role for the government?
Finance.
Fees. University set fees, but perhaps subject to a maximum established by government.
Finance for teaching. There is a continued taxpayer support for teaching, justified by external benefits.
Finance for research. This is justified on various grounds, including the fact that research findings are a form of public good.
Organizing student loans. Government action is needed at a minimum to ensure that the administrative structure for a loan scheme is in place, and to provide targeted subsidies for people with low lifetime earnings who, with income-contingent repayments, will not repay their loans in full. Mostly, government will also have to provide the loan capital.
Regulation.
Fees cap. The cap needs to be high enough to bring extra resources into the system, and, by strengthening competition, to improve the incentives to use those resources efficiently, but low enough to maintain long-term political support for the strategy and to allow institution less used to competition the time to develop the necessary management capacity. The longer-term argument has to do with that universities compete in terms of teaching; some universities are also selling access to the student’s network of peers, and, in this latter respect, have an element of monopoly power. This monopoly part may lead to very high level of fees at some universities (in the US).
Ensuring quality assurance.
Setting incentives.
Subsidies for particular subjects. Government is likely to have views about the shape of higher education.
Subsidies for particular students. One approach is for universities to receive additional financial support for students from disadvantaged backgrounds.
Setting boundaries for competition. Government will also have views about the degree of competition it wishes to see.
Redistribution within higher education. Some universities may be at competitive disadvantage. If that disadvantage is a result of low quality or oversupply in a particular location it may be efficient if the institution’s resources decline, or even for the institution to discontinue, with its students looked after by other intuitions.
Promoting access.
Why not tax finance? Beyond that commensurate with external benefits or promoting access, tax finance contributes to non of the core objectives.
Size. Higher education loses in the political competition for public resources in almost all countries. As a result, shortage of finance constrains size in some countries.
Quality. Over-reliance on public finance puts quality at risk in several ways. The shortage of resources harms quality. Separately, tax-financed systems reduce competitive incentives facing universities.
Access. Again, over-reliance on the taxpayer causes multiple problems.
Why not flat fees? A growing number of countries charge fees, but often a fixed fee set by government. The arguments against flat fees are the mirror image of the arguments for variable fees.
Why not a graduate tax? It is problematic for several reasons:
It is public money, ruling out net private finance until cumulative repayments by graduates outweigh cumulative upfront outgoings.
Funding is closed-ended. With a graduate tax, the government controls the funding envelope.
Incentives to quality are muted. Public finance plus close-ended funding resores central planning, muting the competitive incentives facing universities, and hence creating concerns about quality.
A closed-economy model. Unless the graduate tax is small (in which case it is not a powerful instrument in promoting quality) it will create incentives to emigrate.
Politically problematical. If it is compulsory, it encourages emigration; it will violate individual freedom, risking the long-term stability of the system. If it is voluntary, the problem is one of adverse selection.
Why not other forms of private finance? Fiscal constraints and the externality argument both point a system in which public finance is supplemented by private finance. However, there are at least six potential sources of private finance.
Family’s resources are not bad in themselves, but do nothing to promote access.
A student’s earnings while a student are generally small.
A student’s future earnings – that is, loans.
Employers: it is in the interest of employers that people are trained, but in the interest of each employer to try to free-ride on training financed by their competitors. This type of externality did not arise when workers tended to stay with a single employer for many years, but today’s labour markets require mobility. Therefore, contrary to popular belief, employer contributions are likely to be small.
Entrepreneurial activities by universities: the concern here is that net revenues are generally small and there is a risk of diverting scarce institutional capacity to lower-priority activities.
Philanthropy is useful but marginal (except for a small number of US universities).
The welfare state is the outcome of diverse forces over nearly four centuries of developing social policy. Two aspects stand out:
Debates about ideology, including the extent of redistribution
The welfare state’s role in fostering economic efficiency.
To libertarians the primary goal is individual freedom, which is best achieved by unrestricted private markets. Empirical libertarians such as Hayek and Friedman therefore espouse minimal intervention and oppose all but the most austere welfare systems, whose purpose is limited to poverty relief.
Marxists regard the market system by it very nature as incompatible with their primary aim of meeting need. Therefore, they reject it, and give the state a primary role in production and allocation. In part it accords with their aim of meeting need an is therefore to be applauded, yet it serves also to support a capitalist system that they regard as inherently unjust.
Liberals take a more eclectic view. The utilitarian aim is the maximization of total welfare, leaving open how it should be achieved. Rawls argues that good, liberty, and opportunity should be distributed equally unless any other arrangement is to the advantage of the least well off. Where there is a trade-off between efficiency and social justice, their relative weights will vary between libertarians, liberals, and Marxists.
The aims of cash benefits include insurance, consumption smoothing and poverty relief. They also include reducing inequality. Their achievement requires mechanisms to foster self-help and vertical redistribution.
Insurance and consumption smoothing are necessary for an individual who is self-supporting over his lifetime, but needs a mechanism to allow him to redistribute from himself at one time to himself at another. One part of the answer is voluntary private insurance. On its own, however, that is not a tenable strategy. Another part of the answer is consumption smoothing through voluntary private savings plus the eventual purchase of an annuity.
There are strong a priori arguments suggesting that unrestricted private markets for health care and school education will be inefficient, and also inequitable to the extent that information, power, and access to capital markets are correlated with socio-economic status. The precise form of public involvement has two aspects: the allocation issue and the question of production. The theoretical arguments for the public production of health care and school education, though not irrefutable, are strong, largely because of information problems and incomplete markets.
Publicly provided income transfers have an efficiency role where the private market is unable to supply insurance even to individuals able to support themselves on a lifetime basis. To the extent that the welfare state has a substantial efficiency role, opposition by writers such as Hayek and Friedman is misplaced. The single theoretical issue that divides their arguments from those in the book of Barr is their failure to take into account the major implications of information problems, behavioural issues, and incomplete markets, which affect consumers of increasingly complex products, and also arise in important ways in insurance markets. Consequently, it is less public involvement per se that should be controversial than its precise form and the choice of its distributional objectives.
The welfare state is not a monolith, but comprises many smaller components. Many of these components are publicly financed but privately and competitively produced – welfare does not have to be the welfare state. Where the market failures are severe, state action, albeit imperfect, may produce the least-bad outcomes.
It is not controversial to argue that beyond a certain point taxation harms growth. What is controversial is
Also, one can argue that reduced growth is not necessarily a problem.
National boundaries become porous, making competition global and reducing the freedom of any country to conduct an independent economic policy. Many activities are genuinely becoming more global, but the implications for the welfare state are not necessarily apocalyptic.
The world is not wholly global. Not all goods are tradable; nor all factors are mobile.
Western countries can adapt. This can be done in scale and structure.
Newly industrialized countries may also adapt. In industrialized countries social spending has been a superior good: as incomes have risen, electorates have voted to increase the share of such spending in the GDP. In some countries this might have gone too far, but that does not mean that the premise is flawed.
The welfare state is adapting:
There is particular emphasis in many countries on improving the incentive structure of the benefit system;
The generosity of some benefits is being reduced, for both economic, and in some countries, political reasons;
Other reforms, such as raising the age of retirement, are part of the menu options.
The welfare state will continue to adapt. Some of those adaptations can be foreseen:
Demographic and global change will continue to create pressure to contain total spending. Consequently, pensions and health care will continue to face resource constraints, creating further upward pressure on the age of retirement.
Inequality will be a continuing problem. Consequently, there will be pressures to distinguish more carefully the insurance and consumption smoothing functions of cash benefits from their poverty relief function. This may lead to social insurance becoming more actuarial.
There will be mounting pressure for new insurance instruments to cover contingencies such as long-term care.
New lending instruments will emerge.
appendix_part_2.pdf
Add new contribution