FROM WELFARE STATE TO SOCIAL STATE

About the Author
Wilfried Prewo is chief executive of the Hannover Chamber of Industry and Commerce in Hannover, Germany, and a board member of CNE. He is grateful to Michael Seitz of the Hannover Chamber for his help.

INTRODUCTION

A little over a century ago, Bismarck pioneered the modern welfare state. Pension, unemployment and health insurance provided social and political stability in a new nation state, kept socialism at bay and seemed to be the model for the maturing industrial society. As workers were required to move from job to job and region to region, and as the close bonds between old and young generations within families became looser, the welfare state replaced the old intra-family generational contract with a societal contract. It was even credited with initiating a virtuous cycle: For poorer parents, the prospect of a pension made it unnecessary to rear many children as a form of old age provision; instead, more could be saved which, in turn, fuelled growth.

The cost of the welfare state was, initially, bearable. Benefits were puny compared to today's. Incomes were growing, health was improving thanks to leaps in innovation, and the pension payments for a small number of retired people with still-low life expectancies could easily be financed by a growing number of young people soon to enter the labour force. The "age tree" had its ideal, pyramidal shape.

The success of this model affordable, near-universal protection, against Lord Beveridge's five evils (note 1) made us blind to its basic design flaw. Whether in the forms shaped by Bismarck in the 1880s, Roosevelt or Beveridge in the 1930s and 40s, the welfare state treated the citizen as the recipient of an entitlement bestowed on him and provided by government, whether the latter was acting on a democratic mandate or, as in Bismarck's case, as a benign autocratic patron.

With more and more generous benefits, politics whetted - but did not sate - the appetite of a demanding electorate under the quiet understanding that a major part of the cost could be shifted to yet-to-be-born generations. The demographic trend toward more and more old and fewer young people appeared on the horizon, especially after World War II, but this time bomb had a long fuse. Costs did rise substantially but did not quickly bankrupt the system, since the predominantly national economies were not exposed to severe foreign labour cost competition, especially not from non-welfare state countries.

Increasing global competition has now exposed the Achilles heel of the welfare state. In world markets, we no longer can command prices that are generous enough to finance it. Jobs are lost to more frugal countries. With higher unemployment, the cost has become onerous. Finally, with an increasing number of pensioners, the welfare state sucks up the savings that we would urgently need as we leave the machine age and have to invest and grow into the information age. The welfare state is the albatross around our neck. We are now caught in a vicious cycle.

There is no easy way out. While we can devise an efficient social security system, a major difficulty in implementing it is that, in the political arena, equity considerations take precedence. Until recently, cutting benefits and costs was practically not possible; where it happened, success often was short-lived, with cut benefits soon reinstated via political pressure. Now, the situation has changed; the notion that the welfare state cannot be sustained finds increasing recognition. However, this general awareness does not yet translate into widespread support to cut specific benefits. The distributive conflict, -who gains, who looses?- is not calming down, but is heating up and is standing in the way of real reform. The resulting reform paralysis is frustrating and can only be resolved if efficiency and equity considerations are addressed simultaneously.

Economically sound reform proposals will not always be of the simple win-win-variety. The win-lose-situation is more likely, in which this or that group's distributive status quo is threatened, but the overall gain outweighs such partial losses. If the reform proposals are promoted only by alluding to the overall net gain and do not entail a compensation scheme for the losers, they are brushed aside and, in the political debate, dealt the fatal blow that they lack compassion for the socially weak. To have any chance for a political hearing, reform proposals must, at the outset, guarantee that everybody can gain, that potential losses can be compensated. A revolutionary system change must not be bloody.

In reconciling both efficiency and equity, the cornerstones of this proposal are financial empowerment and individual responsibility: To hand the individual the money required to purchase the current level of benefits, and to leave it to him, within bounds, whether he will do so. This will induce a behavioural change because people will find it in their own interest to economize on social security spending. While guaranteeing that everybody can buy the current benefits, the savings from restraint will be the consumer's to keep. Preserving the no-loss rule is a basic tenet of the empowerment model. This, together with the incentive to save, defines the velvety characteristics of the system change. It must be politically palatable, since it cannot be attacked for worsening anybody's legitimate status quo (note 2)

This system change replaces the welfare state's top-down, paternalistic and heady governance with a market solution, which stands the social system on its feet, those of the constituents. Individual behaviour is changed but the goal of social safety is not called into doubt.

A market solution unlocks efficiency gains and makes benefits more affordable. Elimination of waste helps to re-establish societal consensus about the necessity of a social safety net. But it will also make clear that, as a society, we can only afford a social system which we stand ready to finance ourselves. Gaming the system or shifting costs into future generations is anything but socially responsible.

II. THE WELFARE STATE TRAP: LABOUR COST AND GLOBALISATION

Even with moderate growth, Western Europe is suffering from record-high unemployment. Growth over the last three decades has produced no or only few new jobs. This is in stark contrast to the U.S., where the number of jobs rose by about 70 per cent from 1970 to 2002, while in Western Europe the increase was less than 30 per cent over this period (Figure 2). Unemployment will only recede, if labour costs can be lowered significantly.

Since the 1960's, unemployment in Germany, the EU-country we mostly refer to in this study, has been ratcheting upwards: The additional unemployment generated during a recession was, in the following upswing, hardly reduced, since the cost structure was not corrected. Instead, labour cost kept increasing. With the next recession, unemployment jumped again (Figure 3).

While this relationship has been observed for some time, politicians have closed their eyes and hoped that unemployment would fade away in the next cycle. In fact, the long growth period covering much of the 1980's saw some job creation also in Europe, not exciting by U.S. standards, but very comfortable relative to past European experience. The oil price shocks had been accommodated and there was optimism that new service industries would create more new jobs than would be lost in the inevitable restructuring of old industries. On top of that, the European Single Market brought new hope even for stagnant industries.

On a global sphere and until the end of the 1980's, the main markets and competitors of Western European companies were in the triad of Western Europe - North America - Japan. China and South East Asia were much talked about in the 1980's, but were not yet a major investment location except for a few large companies. Africa was totally out of the picture and South America not yet back in.

In this situation, high European labour costs were, still in the 1980's, (barely) bearable, as labour cost premiums of up to 35 per cent within the triad could be partially compensated for by productivity and other locational advantages; they were usually not the deciding factor for investment decisions. Consequently, successful European companies could remain solidly embedded in their home markets, producing mostly there and exporting from there. "Made in...." was a selling point, be it "Germany" for cars and machine tools or "Italy" for clothing.

With the melt-down of the Iron Curtain and the onset of the information age, this situation has changed completely. The labour cost benchmark is now set by emerging markets in South East Asia and Central and Eastern Europe. These not only offer low wages (many countries do) in the magnitude of one tenth of European labour cost, but also productive labour. They are attractive as production sites for labour-intensive high-quality and, soon, high-tech products.

The exodus to these countries has begun. For Western European firms, countries such as Slovakia or the Baltic countries are the "Hong Kong around the corner"; they are attractive for mid-size companies that, in the past, had not considered going to far-away countries. As a result, employment in Western Europe will stagnate or shrink further. In post-Cold-War Europe, the increased competitive heat melts the financial pillars of the welfare state. In the Bismarckian variety of social security systems, financing is linked to employment and, via payroll taxes that are partly borne by employers, directly becomes a part of labour cost, regardless of whether the covered risks are employment related or not; examples are Germany, France, Belgium, and the Netherlands. In the Beveridgean systems (examples are the U.K., Denmark or Sweden), where social security such as a national health system is financed out of government revenues, the costs do not directly add to (indirect) labour cost, but the effect is eventually the same: In the Bismarckian countries, payroll taxes make jobs too expensive. In the Beveridgean countries, jobs are eliminated as other taxes drive up gross wage demands and make the products too expensive on world markets. The main causes of sluggish job growth are the high cost of labour, especially indirect labour cost (Figure 4) (note 3), and a large government sector (see again Figure 1).

In Germany, the archetypical Bismarckian welfare state that can be used as a point of reference, indirect labour cost in manufacturing industries amounts to 81,2 per cent of direct, gross wages in 2001 (note4), which is about double the U.S. share. Indirect labour costs in Germany have increased almost twice as fast as gross wages since 1966, and this increase is closely correlated with the rise in the unemployment rate (Figure 5). Consequently, lowering indirect labour cost appears as a most promising avenue to employment creation.

In 2004, the payroll taxes in Germany amount to: 19.5 per cent of gross wages for pension insurance, an average of 14.3 per cent for health and 1.7 per cent for nursing (old age) care, 6.5 per cent for unemployment insurance: a total of over 41 per cent of gross wages! Figure 6.

Half of this is paid by employers, adding to indirect labour cost and making Germany the country with the second highest labour cost, which for a German industrial worker was 22 per cent higher than his American counterpart in 2002 (note 5) For the German worker, his half of the payroll tax, together with his taxes, reduces his take-home pay to 16 per cent below the American level. Higher labour cost does not necessarily yield a higher standard of living. Figure 7.

Consider this comparison: A German plumber has to work five hours (at a net wage of Euro 8-9 per hour) to hire a painter for one hour (at a cost of Euro 38-40 per hour) and vice versa. Neither of them can afford that. Turning to do-it-yourself is one reaction, though not a problem for society. But it is a problem when the reaction takes the form that more and more unemployed do actually not seek re-employment, instead supplementing their unemployment benefits with black market income and finding the combined to be higher than their previous net wage. As a result, the welfare state is haemorrhaging. Fewer and fewer actively employed people have to pay more and for more social security recipients, raising labour cost even further and adding to the job loss spiral.

On the investment side, the welfare state's consumption spending drains resources needed for capital formation. In 1980, investment outlays still exceeded public welfare expenditures in the major European countries; since then, public social spending in these countries has exceeded investment, and the gap has grown. (Figure 8). In the U.S., the increase in the ratio of social spending has been far more moderate, and is still dwarfed by investment.

The problem is exacerbated by demographic trends: In 1950, Germany still had, with the exception of war-related indentations, a population pyramid with a broad base of young and a narrow cap of old people; by 2002, the "age tree" has become a pumpkin with few young people, many mid-lifers (still in the active labour force) and a growing, though not yet dominant number of retirees. By 2040, it will resemble a mushroom: very few young, few mid-lifers, many old people: The pyramid will stand on its head! (Figure 9).

Under such conditions, pensions cannot be financed by a pay as- you-go-scheme where the actively employed pay for the pensions, health and nursing care of the elderly. By about 2030, there will be about one pensioner for one active worker with the consequence that the worker's pension payroll tax would, on average, equal the average pension (note 6). Ask today's 25 yearolds whether they believe that their children will be ready for that. The "generational contract" underlying this system is an unsustainable chain-letter. The only distinction from other pyramid schemes - schemes that also portend to have a broad and supportive pyramid base, but in fact are built on sand - is that the pay-as-you-go pension system is legal, and the foulest of all.

The welfare state impoverishes the current generation, deprives future generations of growth opportunities, and even shifts further liabilities onto a weakened future generation. As well-intentioned as the welfare state's design may have been, the results are perverse: Rather than guaranteeing social security, it is now endangering it. The welfare state's fate will mirror that of the equally morally corrupt and economically exhausted socialist states at the end of the 1980's. Without reform, it will implode just as Eastern Europe did in 1989. Resisting such reform is the truly anti-social posture.

III. GOVERNMENT IS THE PROBLEM

The question is not whether but how - and, derived from that, how much - social security is provided. This is where the systemic flaw of the welfare state's social security systems becomes apparent: Each is a uniform plan that disenfranchises the individual. Benefits packages are standardized, one size fits all. You cannot pick a lesser plan for a lower premium. Consumer responsibility and choice have no material place; the individual is bestowed with an entitlement determined by the government, he is not a consumer/buyer of social insurance products.

Social security plans are, by and large, not insurance systems. The contribution or premium is not actuarially calculated. Through the institutional tools of either the payroll tax or outright socialization, individual consumption of social security benefits is unrelated to individual expenditure. This gives rise to moral hazard: The insured individual exploits the system by extracting as many benefits as possible, since the costs that he inflicts on the community of the insured are only insignificantly borne by himself because premiums increase for everybody. As more and more behave this way, the system becomes grossly wasteful; worse yet, as prudent behaviour is ridiculed, solidarity loses its moral base.

While any insurance is exposed to the moral hazard problem, insurers try to eliminate or, at least, limit it with deductibles and co-payments, differentiated premiums, incentives for prudent use, or penalties for abuse.The welfare state, in contrast, invites moral hazard and makes itself hostage of entitlement mentality. The claims against the entitlement system are never lowered, only raised, all the more since claims can be conveniently filed at the voting booth. Bowing to such pressure is not only costless to politicians; those that are best at cheering on the electorate in its call for more, the animators of the Club Med welfare state, can hope to carry the grand prize of "taking care of people".

When business plans fail, they are scrapped or the firm goes bankrupt. When government plans fail, they are expanded. In attempting to control costs, the social security systems have become increasingly politicised. Even those not run outright by the government, but conceived as friendly workers' societies or employer-employee partnerships under self-autonomy, have, under increasing government intrusion, been converted to government plans.

The welfare state's construct - how to provide social security - answers, "how much? always more, but, in any case, the same for everybody. While it is possible to stop the entitlement spiral temporarily, there is no lasting success to such efforts: Although German health care underwent many reforms, all of them failed in their goal of redressing excessive health cost inflation. They are the problem, not the cure.

IV. THE EMPOWERMENT REFORM

Breaking out of the welfare state's vicious cycle requires discarding the top down, social engineering approach of government-run plans and organizing social security in a bottom-up, consumer (incentive) driven way. Instead of disenfranchising the individual, we must empower him to evolve from a spoon-fed recipient of a uniform, one-size-fits-all entitlement to a sovereign and picky buyer of social insurance products, (note 8) while:
- enabling him to buy the current benefits,
- requiring him to buy only a minimum preventing the worst case (becoming a welfare recipient), and
- allowing him to pocket any savings resulting from prudent behaviour.

The empowerment reform, or "consumer model", contains five elements:

1. Empowerment requires, first of all, purchasing power in the hands of the individual that enables him to finance his choice. But fresh money is not needed. All we have to do is hand him the money currently spent on his behalf. This preserves the no-loss rule.

   To this end, all current contributions to social security systems should be deposited in an individual Social Savings Account (SSA), which is established in the insured beneficiary's name. In the Bismarckian systems, this means that the current employer and employee payroll taxes are deposited in the SSA. This transfer must be tax-free, to the extent that current employer contributions are also tax-deductible. In the Beveridgean countries where social systems are - partially or fully - financed out of general tax revenues, the government must transfer the corresponding monies into the SSA of the individual beneficiary. In either case, the change is revenue neutral for the government; it should not lose, either, but should also not get additional revenues.

   The individual SSA's must be corrected for current cross subsidies or implicit solidarity transfers. For example, a health insurance payroll tax, being dependent on salaries only, is not an actuarially calculated premium. Lower income earners, large families, pensioners, or women, each as a group, are typically subsidized by high income earners, singles or unmarried couples with two incomes and no children, the young, or men.

   Those belonging to the former group have a payroll tax falling short of an actuarially calculated premium, while those of the second group typically pay more (note 9). At this point, we do not open a debate on the merits of current transfers, but, strictly adhering to the no-loss rule, require that these transfers are preserved: Consequently, the first group will have amounts added to its own SSA deposits, the second group will have corresponding amounts deducted; both will then have an SSA balance that is sufficient to buy current benefits at full actuarial value. Nobody - government, employers, employees - has been made worse off

   .

2. Allowing the beneficiary to evolve from recipient to buyer requires giving him the option of buying less (or more) than current benefits and letting him pocket any savings resulting from his choice. Instead of only making hollow appeals for prudent behaviour, the call for self-responsibility must be magnetised, if we want to hear an echo. Whoever demands less should pay less.

   There should be limits to choice. We may regret that, but individual choice should have its limit where carelessness would produce a welfare case. The community would not gain, if an individual would not insure himself against catastrophic illnesses in order to save on his premium and, in the worst case, become a burden to society. But this also defines an upper limit to mandatory coverage: It can be as low as to prevent the welfare case. In the case of health care, this implies, for example, mandatory coverage for catastrophic illnesses and for ambulatory care beyond a sizeable annual deductible. In the case of unemployment insurance, mandatory coverage need not start on day one, but, say, after six weeks of unemployment (note 10).

   The individual, not a third party, controls his social savings account deposit. He can use it to buy coverage for the full current benefits or for as little as the mandatory level. The premium savings in the latter case are a major deterrent to "gaming the system". It is not necessary to police socially unacceptable behaviour by bureaucratic control, as is often suggested. The incentive of a lower premium in the case of prudent behaviour is the best weapon against moral hazard.

3. There is free choice among providers of health care, unemployment, sick leave or pension insurance. Free choice among providers applies not only to optional coverage, but also to the mandatory part. Thus, mandatory coverage does not mean government as the monopoly provider. Analogous to mandatory automobile liability insurance, where the insured can select among competing insurers, free choice among providers will unlock a vastefficiency potential on the supply side. Moral hazard on the demand side is, by far, not the only source of savings.

4. The premium savings resulting from less-than-full coverage or from having chosen a more efficient provider will be the individual's to keep. However, unspent balances cannot be withdrawn for consumption purposes.11 The individual can only draw on them either for expenses not covered, e.g., a deductible, or for other social security purposes, such as nursing care or additional old age insurance. There can be wide latitude in defining social insurance, allowing any investments deemed to secure social safety. This could include equity in a home or in a company. The reasons for proposing forced savings in the sense of limiting the spending of SSA balances to social insurance purposes are two-fold:
   First, as the SSA balances build up over time, the limits for mandatory coverage can be lowered and become a variable depending on unspent balances, thus increasing individual choice and incentives to economize without the threat of becoming a welfare case.
   Second, the SSA addresses more than one area of social insurance; in particular, we can and will need unspent SSA balances, whether they come from less spending on health care or other areas of social security, to build up a fully funded pension system. While forced savings may have a negative connotation to some, they do, in this case, pave the way towards a market-oriented pension system. Needless to say, being vested in the individual, unspent SSA balances must, at the time of death, go to the holder's estate, exempt from inheritance tax (note 13).

5. Finally, the Social Savings Account may be managed by current social security administrations. There are pragmatic reasons for this: We should allow the current pension, health or unemployment administrations to seek a role in the new system. They and the sizable body of their employees would fight the new system, tooth and nail, if we gave them no role and would just abolish them.

   The German public health system (Gesetzliche Krankenkassen) alone has about 150,000 administrative employees (note 14), a formidable force; it is not necessary to make enemies with them. We should give these institutions the opportunity to also evolve into a competitive provider. If they change successfully, all the better. However, there must be no monopoly care-taker of the SSA. If an individual sickness fund, for example, offers to manage the account, it does so in competition with others. The individual decides where he does his SSA banking. The current system of monopoly care-takers of health, unemployment, of pension insurance entails considerable administrative overhead that can be saved by competition among SSA portfolio managers.

In a nutshell:
1. Privatisation and efficiency on the demand and supply sides.
2. Equity with the no-loss rule.
3. Safeguard.

First, the empowerment reform privatises the social systems, unlocking the efficiency potentials that materialize in privatisation schemes: Consumers gain from more choice and economize by picking those services that appear optimal to them on cost-benefit grounds; and on the supply side, privatisation and competition bring better products faster to market at the lowest price. Second, the no-loss rule assures equity in providing everybody with the same incentives to act prudently. Third, mandatory minimum insurance is the safeguard.

It should be added that the savings account proposed here is not a novelty: It has been inspired by the Medisave accounts in Singapore and related concepts that have been discussed and experimented with in the U.S. since the early 1990�fs . Archer medical savings accounts (MSA), health reimbursement arrangements (HRA), flexible spending accounts (FSA), and, finally, the health savings accounts (HSA) that have been established by the 2003 Medicare reform bill. Of these, HSA�fs appear to become a potent instrument toward consumer-driven health care (note 15).

V. APPLICATION

The general criteria of the empowerment model need to be applied to health, unemployment and pension insurance. It has to be shown how features that are genuine to each of the three areas can be embedded. In addition, the reform of current sick leave payments is addressed: First, although the financing of sick leave differs from the other three areas, it is related to them, in particular to health insurance. Differences in the duration and cost of sick leaves among countries, whose societies enjoy a similar health status, highlight the importance of institutional arrangements. Second, since employer paid sick leave is typically limited to a month or so, even the most drastic reform - abolishing it - would not slash the social safety net. The argument that reform would threaten "social peace" is ridiculous. Finally, the proposed sick leave reform is ideal in showing the basic workings and strengths of the empowerment model. Therefore, we address it before health reform with its additional complexity of cross-subsidies or "solidarity transfers".

As health reform is - and will remain - at the top of the social policy agenda, it is dealt with in more detail. The contrast with the increasingly politicised government health plans shows the empowerment reform's ideological departure from the welfare state's social engineering approach (note 16) and its promise in containing costs through its reliance on and respect for individual choice.

Critics of self-responsibility often argue that it would lead to less than adequate preventive care. Such arguments do not rely on logic, but usually dramatically allude to catastrophic illnesses. The application of the empowerment model to unemployment insurance shows that the opposite is likely to happen: Given the right incentives, people will seek more jobtraining, which is the best preventive care against unemployment. Pension insurance is treated last, as here we have to show how SSA balances can ease the - admittedly burdensome - transition to a fully funded pension plan. Welfare - aid to the poor - is not specifically addressed here, not because reform is not urgent, but since its financing is not related to employment and labour cost. The empowerment reform would, however, substantially reduce the welfare "problem": Easing the burden that the welfare state places on labour cost would increase the take-home pay of low-wage earners relative to welfare receipts. Second, welfare recipients would not enjoy the monetary benefits entailed in the Social Savings Account. Both aspects would make it more attractive to leave welfare rolls and seek employment, while concentrating society's resources on the poor that cannot be integrated into the labour force.

51. Sick Leave

Typically, European employers are required to continue paying employees that are on sick leave as much or nearly as much as if they were working. The duration of employer-paid sick leave is not open-ended, but limited to a period of, typically, a few weeks to three months per incidence. After that, government or health plans take over, often at a lower compensation level. Employer-paid sick can amount up to 100 per cent (e.g., Germany); and it commences either on the first day (corresponding to a zero deductible), as in Germany and Denmark, or after a waiting period of one to three days (e.g., 1 day in Belgium and Sweden, or 3 days in Italy or France), during which the employee is "at risk" (note 17).

When, as in Germany's case, the employer pays 100 per cent of income beginning with the first day of illness, the moral hazard problem is most severe. As a result, about 4.2 per cent of working days are lost to sick leaves in Germany, in contrast to two per cent in the United States. (note 18) The cost to German industry amounts to 3.9 per cent of direct wages (note 19). International differences in sick leave, varying, as shown, by a factor of 2 or 3 - have little to do with the employees' health status (Figure 11).

Gaming the system, or moral hazard evoked by full or near-full compensation, is the plausible cause (note 20). Consequently, the cost savings, or efficiency gains, from replacing the current system with the empowerment model would be exceptionally large. In keeping with the reform criteria, the reform steps would be:

1. Employers transfer an amount equalling as much as their current sick leave expenses - in German industry, on average, up to 3.9 per cent of direct wages - into the SSA of the individual employee (note 21). In turn, the individual employee is then responsible to cover himself against income lost during sick leaves. A pay-out corresponding to previous employer expenses is sufficient to buy insurance at the current level of sick leave benefits.

2. The individual is not required to cover all income losses of sick leaves. The empowerment model only calls for mandatory coverage at the level where uncovered income losses would send the employee to seek public assistance. Since only a few weeks (of currently employer-paid sick leave) have to be covered, it is quite conceivable that there would be no need for any mandatory sick-leave coverage at all. Many people have some savings, and for those that do not, their employer's deposit into their SSA would build them up fast. However, granting that some minimum coverage would be desirable even for such a limited risk, mandatory insurance at the level of unemployment coverage (around 60 per cent of net income) would certainly suffice (note 22). Another option would be to start mandatory coverage after a grace period of two weeks.

3. Sick leave insurance can be obtained from any insurer of the beneficiary's choice. Private insurance companies can offer sick leave insurance - they already do so, currently to a clientele of self-employed. The public health system (sickness funds) also could be a willing insurer; it has the necessary actuarial information, as it processes the health bills and already compensates income losses after employer- paid sick leave runs out.

4. A policy with such co-payments (e.g., at a 60 per cent compensation level instead of 100 per cent) or deductibles (two unpaid weeks instead of coverage from day one) would cost far less than currently. The savings would remain in the SSA and could either compensate income shortfalls in the case of illness or be used for other social security purposes, such as building up a fully funded pension plan. It does not take a clairvoyant to predict that, in countries with above-normal sick leaves, their duration and cost would come down to international benchmarks. Sick leaves vary with leave benefits, as shown by an international cross section in figure 13.

The Swedish experience is an interesting example, as it highlights how, within the same country, behaviour (length of sick leaves) changes when benefits are changed. In late 1987, employer-paid sick leave was raised to 100 per cent (from 90%) and started on day one (previously, the first day was not compensated). As a result, sick leave in Sweden rose from 7.7 in 1986 to 9.7 per cent in 1988, its peak (Figure 14). In the wake of Sweden's deep crisis, successive cutbacks were made in 1991, 1992 and 1993: the first day went unpaid again, compensation was dropped to 75 per cent for the second and third days, then 90 per cent for days 4 to 14. Concomitantly, sick leave fell from its 1988 peak of 9.7 to 3.8 per cent in 1997. With only one unpaid day of sick leave and with mandatory benefits at 75 to 90 per cent, the Swedish reform policy of the early 1990�fs was certainly not asking for too much sacrifice. Yet, these relatively mild corrections resulted in a 60 per cent drop in sick leaves.

The reform movement was met with a political backlash, and in the second half of the nineties, the pendulum in Sweden swung back! Sick leave benefits were increased again and sick leaves rose steadily to 6.1 per cent by 2002. This is a persuasive example of the impact of government policies on the behaviour of a rational citizen.

Employers would gain, even if they transfer the full current cost of sick leaves into the Social Savings Accounts of the employees: With reduced absenteeism, machine utilization increases, the cost of capital per unit of output decreases, and additional costs of hiring replacement workers are saved (note 26). More important, the past trend of the cost of sick leave would be broken: During the period 1998 to 2002 alone, companies' total costs for sickness have increased by 17.4 per cent (note 27).

During the same time, average income in Germany increased by 10.0 per cent (note 28); under the reform proposal, it would, in the future, only increase at the rate of wages. Furthermore, both employers and employees would reap health cost savings, since the documentation of sick leaves requires a visit to the doctor's office.

52. Health

All European countries are exposed to excessive health cost inflation (Figure 15). This is, to a large part, unavoidable: Aging drives up health costs, and medical innovation comes at a price. But these are costs that we are willing to bear. On the other hand, all national health plans suffer from inefficiency; unnecessary costs are a waste to society. In Germany, for example, the administrative cost of the public health system has risen even faster than medical expenses, although information technology and standardized billing should have led to a decrease in the cost of overhead.29 Instead, health care "reforms" have bloated administrative expenses and posed disincentives to medical and pharmaceutical innovation.

a. Health: The Problem

The core problem of a national health plan is that it is a plan. The problem is not that we are planning - in this case, providing for the prevention and treatment of illness. Every day, we are formulating plans, what and how to work and how we spend our time and money. Our plans are highly differentiated: We may get up early or work late; we may engage in sports or watch it from the couch. Everybody plans individually according to his desires and circumstances. The problem of national health plans is that each is a uniform plan, one size fits all. Our shoes are different, fitting the size of our feet, matching our clothing, or being suited for work, dance, or sports. But our health plans are standardized. Not we as individuals, but the government as our presumed agent determines what is and what is not covered. Consumer responsibility and choice have no material place. You cannot pick a lesser plan for a lower premium, a higher deductible or co-payment in exchange for a lower premium, or choose less dental coverage than currently provided in exchange for other options.

A system that does not reward restraint but disenfranchises the individual, puts itself under permanent pressure: More and more is demanded. Everybody tries to milk the system as much as he can. Individual consumption is not related to individual expenditure, neither in the Beveridgean system with its general tax finance nor in the Bismarckian system with its payroll tax. Beneficiaries behave like diners at an eat-as-much-as-you-can restaurant: Each time we go there, we tend to eat too much. Afterwards, we regret it.

On the supply side, doctors, dentists, pharmacists, and hospitals do not compete for our, the customers', business, but render their services. Their remuneration is dictated by cartels or the government. More and more, government plans fix detailed prices, prescribe services and formularies and subordinate everybody under a budget.

Here the beneficiaries - there the providers, and each group behaves like bears sitting around the honey pot. Everyone digs in, grabbing with one hand as much as he can, while trying to restrain the neighbour with the other hand. In astonishment, politicians recognize that each time a fresh pot has been served, it is gone in record time. Reform then first tries to appeal to table manners in the form of round table talks or concerted actions - with little avail; under the next reform the honey is stretched and, finally, the contents are reduced by a false bottom. Every time, however, the honey becomes more expensive. And the ground rules remain the same: Everybody gets the unitary brand and as much as he wants - as long as supplies last.

We are still hesitant to discard this absurd plan with its foredoomed cost spiral and replace it with a variety of individual plans. We are hesitant because we view health as a basic, elementary need. Everybody - rich or poor - should have access to health care. Nothing less than universal access to health care is suggested here. The fallacy is that we believe that health must be government provided if it is to be available to all.

Clothing, housing and food also fill basic needs. We do not want anyone, for reason of poverty, to be without clothes, shelter or food. Yet clothing, housing and food are organized differently from health care. We do not have the government outfitter from which we obtain the one-size-fits-all coat. We do not have the central quartermaster that provides standardized housing. Nor do we eat the same menu in the people's canteen. In the case of health care, however, (too) many still believe that the goal - universal access to health care - requires a unitary plan. Consequently, we do not speak of a health care market in Europe, but of its provision. The insured are not purchasers, customers and clients, but beneficiaries and "members" of plans.

We did not have to lose any sleep over this as long as we spent a small fraction of GDP on health care. But spending around a tenth of national income on health should give cause to consider whether, in a market economy, we should organize health care as a market or ration it socialism-style. Too frequently, we talk about the market economy as if it were inconsistent with social policy and that the latter always requires at least a good dose of socialism. This ideology has drawn us deeper into the hole. Seemingly harmless interventions were followed by more heavy-handed dictates and, finally, formularies for drugs, fixed budgets, mandated prices and fee schedules.

As many have said before: Social policy and a free market economy are not contradictions. To the contrary, the efficient and qualitatively best social policy is the one with a market orientation.

b. Health: Reform Proposal

The proposal follows our guidelines and mirrors the reform steps for sick leave benefits. In addition, the health care proposal has to address the current cross subsidies between different risk groups:

1. Each current beneficiary of a health plan receives - in the form of a deposit in his Social Savings Account - the money necessary to finance his current benefits (Figure 16). In turn, the individual himself is then responsible for his health coverage, either by buying health insurance or by contracting directly with a provider, be it a fee-forservice physician or a managed care provider, such as a health maintenance organization (note 30).

In the Beveridgean countries, the deposit into the individual's SSA would come from the government which, in turn, saves the corresponding amounts by discontinuing to finance its current national health plan. In the payroll tax-financed Bismarckian countries, the money comes from the employees and their employers according to their current shares in the payroll tax. In some countries, such as in Germany, the payroll tax is shared evenly by employers and employees; in others, such as in France, Belgium, Italy, Spain, or Portugal, employers pay a larger share.31 Of course, in mixed Beveridgean/ Bismarckian systems, money would come from all three current financiers: government, employers and employees. Furthermore, to the extent that current payroll taxes are tax deductible - in particular, this holds for the employers' shares -, contributions into the SSA must be tax-free in order to preserve revenue neutrality.

In the Bismarckian countries, the payroll tax paid by an employee and his employer is, except by coincidence, not equal to full actuarial value of current benefits. It either exceeds or falls short of that. Younger employees, singles, dual income households with no children, higher income earners32, and male employees tend to have a payroll tax exceeding the full actuarial value of benefits.33 Correspondingly, the contribution (payroll tax) paid by or on behalf of pensioners, households with only one income and with dependents, low income earners, the unemployed, or women tends to fall short of the full actuarial value of their current benefits. This is especially so for the elderly, large families and low income earners (Figure 17).

For example, German retirees have health costs that are much higher than those of the actively employed and their co-insured. On the other hand, while they also contribute to the health plans, the payroll tax rate is assessed, like for everybody else, only on their current income, their pensions, not on their previous salaries (note 34).

In 2002, the contributions on behalf of pensioners amounted to Euro 27.9 billion in Germany, while their health expenditures came to Euro 63.4 billion, i.e., pensioners were subsidized to the tune of Euro 35.5 billion or close to 60 per cent of their costs. Underlying this subsidy is an implicit intergenerational contract, analogous to that of the pension system itself (note 35). The subsidization of the elderly will become more severe as the proportion of senior citizens increases.

Families with only one income and several dependents constitute the other major group whose payroll tax falls short of this group's health expenditures. In Germany, all family dependents receive the same benefits as the payroll tax-paying head of the household, whereas the payroll tax is independent of family size. For every ten working employees paying the payroll tax, there are 5.3 non-paying dependents. Non-paying dependents had health expenditures of Euro 18.9 billion in 2002.

With health expenditures unrelated to income, low income earners also are subsidized; this also holds for the unemployed. And women tend to have higher health expenditures than men. In a pure Bismarckian system, where the health expenses of all beneficiaries are financed by the payroll tax, these subsidies are covered entirely within the system, by the respective counterparts of young people, singles, high income earners etc.

The empowerment model's no-loss rule requires that the deposit into each SSA is at full actuarial value of current benefits, i.e., it must suffice to buy the current full benefits of each beneficiary, regardless of his and his employer's contribution. Otherwise, the choices offered would not unfold the same incentives for everybody. Therefore, the payroll tax deposits into the SSA's have to be adjusted for the current subsidies. Those with payroll taxes exceeding the full actuarial value of current benefits will have amounts deducted from their SSA's, while the others - pensioners, families with dependents etc. - will have additional amounts deposited into their individual SSA's. No fresh money, however is required in this process, since these withdrawals and deposits even out: Currently, this cross-subsidization goes on "behind the scene", within the sickness funds, but unobserved by the individual beneficiary. All that would change is that we lift the curtain and throw light on this process; the transfers become explicitly and individually observable, as they show up as credits or debits on the individual SSA's. After the transfer, everybody will have an amount in his SSA that is at full actuarial value of the benefits that he currently enjoys. The no-loss rule is preserved. This is critical for allowing the empowerment model to unfold its economizing incentives.

In turn, the individual employee, having sufficient funds in his SSA, is then required to cover himself against health risks.

2. The individual is, however, not required to buy a health plan covering all current benefits. The mandatory or minimum coverage would encompass those risks which, if uninsured, could force him to seek welfare assistance and become a burden to society. For the empowerment model to unfold its strength, it is also not necessary to limit mandatory coverage to absolute bare-bones essentials. A comfortable level of mandatory coverage, with the right structure of deductibles and co-payments, would already provide significant incentives for prudent use of health care. It is, therefore, not necessary to engage in a debate of what is exactly the lowest possible mandatory coverage, below which the welfare risk is imminent. We can be generous and, rather, err on the high side.

   In the case of health care, coverage of catastrophic illnesses would, of course, be mandatory. Most treatments requiring hospital stays would be included, possibly with a co-payment or deductible. Chronic illnesses requiring continuous or frequent treatment would be likewise covered. In the case of outpatient medical treatment, mandatory coverage could begin after a fixed annual deductible or a thirty per cent co-payment. Also, the aggregate of all annual co-payments and deductibles would be capped, for example, at 4000.00 Euro. Putting an upper limit on the total of annual risk sharing is a backstop guarantee that the beneficiary is absolutely safe from falling through the net. On the other hand, aside from new co-payments and deductibles, the mandatory package would not include medical expenses that, even if currently covered, can be borne by the individual. Examples are eye glasses or dental care.

3. There is free choice of health care providers. Only the extent of mandatory coverage is prescribed. Provider and form of that coverage are up to the individual to choose. The provider need not be an insurance company, but could also be a health maintenance organization or a preferred provider network. The individual may choose a fee-forservice plan that allows him to see any doctor of his choice or he may subject himself to restrictions that might come with managed care. Of course, these plans will have different costs. The respective premium savings are a strong incentive to "shop" for the best plan. What is "best" is determined by the individual, not by politicians or bureaucrats (note 36).

   Beyond mandatory coverage, both scope of coverage and providers are optional. The individual may, with his SSA deposit, buy the current benefits or as little as the mandatory level. And he can buy benefits that are currently not covered if he realizes savings in other areas.

4. Any savings will belong to the individual and stay in his SSA. Premium savings come from two sources, from choosing either less coverage than currently provided or a less costly provider. The individual can use his SSA balance to pay (self-insure) health care expenses that he has not covered (for example, eye glasses or a copayment), or he can use the SSA balance and invest in a retirement fund or other instruments raising his social security.37 The savings resulting from prudent behaviour in the areas of sick leave, health care or unemployment insurance can ideally be used for building up a fully funded pension plan.

5. As there is free choice of providers and plans, the individual is also free to choose his SSA fund manager. No qualified institution should be excluded from offering its services as an SSA manager. The current government health plans or the legally mandated sickness funds may also offer to manage the individual SSA's. It is quite conceivable that they will not be shunned by the individual. For, in a market environment, these previous bureaucracies will quickly change, just as mail and telecom providers did change with privatisation. Sickness funds even have a distinct advantage over new private SSA managers: Currently, they account for all of the cross-subsidies between old and young, rich and poor. This information gives them head-start opportunities compared with other providers. The question is whether they take a defensive attitude and try to resist change or whether they view it as an opportunity. We should leave that decision - and its consequences - up to them.

c. Health: Quality and Price

The empowerment model improves health care and reduces its cost. Since everybody may choose his provider - insurance or capitation-based HMO, fee-for-service plan or managed care - everybody gets a plan that, in his view, not some politician's, serves him best. And since beyond mandatory coverage everyone can determine the extent of coverage, he will spend the money in those areas that he desires. Value is in the eye of the beholder. Nobody has to pay for a plan or for extra coverage that he does not want.

Using our analogy again: We replace the fixed-price, unitary food-plan, monopoly eatery with highly varied, fiercely competing restaurants. There are those with table cloths, flower arrangements and extensive wine lists; and there are the family restaurants, the chains, and the serve-yourself cafeterias. Instead of a unitary diet, each restaurant allows the customer to choose from an exhaustive menu and compose his own order. Only a minimum diet (mandatory coverage) is prescribed. The previous unitary plan is also included in the offering, at no more than its previous cost. Which restaurant arrangement would you prefer?

Whatever level of quality we choose, it will become available at least cost, since everyone will try to preserve his - his own and valued - SSA balances and spend only as much as necessary. Since different plans come at different costs, with more expensive plans leaving less to be added to his SSA, he will contract only with such providers whose services he values at least as much as their cost. Waste which was caused by careless (not cost-conscious) use will be eliminated, since, with a deductible or co-pay, services are no longer perceived to be free.

Imprudent individual behaviour is no longer externalised, shifted onto the community of all beneficiaries. Unnecessary visits to doctors would vanish. A patient would not embark on an odyssey visiting specialist after specialist without regard for cost, but will consider the cost of duplicate exams and lab tests. He will ask his doctor and, first of all, himself, whether he has to return for a follow-up visit after his nose stops running. He will no longer ask the doctor to give him a sick leave slip for a week, when his tennis elbow aches on Monday morning.38 Selfpayment of small bills that fall under a deductible would relieve the insurer of substantial administrative cost; hence, the premium cutback would be more than the actuarial value of the deductible. Likewise, an insurer could offer the beneficiary to submit bills for reimbursement only quarterly or semi-annually, or when the total has exceeded a set amount, with respective premium savings passed on to the individual (note 39).

On the supply side, consumer choice would, above all, bring about an efficient and competitive structure: Benefits and prices will be market-determined, not bureaucratically dictated or negotiated between supplier and provider cartels. Health care will no longer be structured along guild-like professional or corporatist rules. Why should, for example, the fee-for service physician practice be the only or dominant form of providing ambulatory care? Cost decreases can be achieved in a service-industry: Health maintenance organizations or preferred provider organizations will be offered to and chosen - or rejected - by the individual. The fee-for-service practice will not disappear, but each form of service will carry its price tag. We will have a variety of plans fitting a variety of individual needs. Again, we do not prescribe, but let the individual choose his preferred option (and pay or save accordingly).

d. Health: Competition and Supply Structure

For outpatient care so far, Germans see only a fee-for-service, often solo-practicing, physician. Hospitals are, typically, owned and operated by communities or charities. Doctors in private practice cannot operate in hospitals; ambulatory care by doctors in private practice and stationary care by hospitals are separated, with each side guarding its turf. As of 2004 hospitals may also treat patients on an outpatient basis. Ownership of a pharmacy was until 2004 limited to one per pharmacist. Now, a pharmacist may own up to four pharmacies. Legal and professional regulations, some of them rooted in guild-like structures had prohibited multi-specialty group practices, chain or mail-order pharmacies. State control over opening or expanding a hospital allowed little room for private hospitals (note 40).

As a consequence, HMO's are still unknown, preferred provider networks do not exist, there are no pharmaceutical benefit managers, no physician practice-management companies41, while in other countries, above all in the U.S., there has been rapid structural change and experimentation with new organizational forms over the last two decades. It would not be necessary to have the government promote new forms of service. All we would have to do is allow them: Give people free choice of providers and lift restrictive regulations. Structural change would follow immediately, especially when cost savings from alternative forms of service and a variety of health plans become apparent. Consumer choice means supplier competition. Today, doctors and hospitals spend an inordinate amount of energy on how to extract remuneration out of the public health systems. In the empowerment model, they will compete for the patient's business and strive to offer alternative, better or less expensive services.

The empowerment model emphasizes choice instead of the imposition of a unitary form of service. Therefore, we do not take sides - for or against managed care, for or against traditional pharmacies, for or against public hospitals, for or against fee-for-service-practices, for or against physician practice-management companies. We leave this choice up to the consumer. Different plans entail advantages or disadvantages for the individual and will come at different prices.

Overall, health care costs may or may not absolutely decrease, as aging and technological advances tend to push costs up. But whatever total costs we will have, they will be at the minimum for the level of care desired by society and markedly below the cost trend of the current system.

e. Health: Criticism and Objections to reform proposals

The empowerment model finally puts the patient, as a consumer, on a level playing field with the suppliers (doctors, hospitals, insurers, managed care providers). Critics are quick in raising objections:

• (i) Patients lack medical information to make sound judgments. Patients are, in most cases, not medical experts. Granted. Do you have to be an automotive engineer or a computer expert when buying a family car or a home computer? When the car breaks down, we get it repaired; we do not need to self diagnose the problem and prescribe the repair (therapy). We order the outcome ("get it fixed"). We know from our own or others' experience where we get the best service, whom we can trust, where we get our money's worth. Sometimes we make the wrong choice and are disappointed. We will remember and not repeat that mistake. The parallel to health care is obvious. Not possessing the same information as a medical expert, seemingly puts us at a disadvantage. But in the empowerment model, our distinct advantage is that we can take our business elsewhere, and advise our friends likewise. Competition and freedom of choice of supplier forces the latter to always try to use his best judgment and, together with the looming threat of liability, do anything to prevent malpractice. There is no guarantee that we always receive the best service. But, overall, such a system minimizes bad calls.
• (ii) Patients are not mature.

  There are many critics that hold the individual not mature enough to be a wise shopper of health services, when offered the option of a sizeable deductible, even when limited by an annual cap of about 4.000 Euro, as suggested here. This is grotesque. Each owner of a car bears the cost of a tune-up, an inspection, tires, and depreciation. These costs easily add up to 4.000 Euro annually. We certainly hold every car owner, rich or poor, responsible for these costs. Imagine how we would drive if our auto insurance would pay for burnt-up engines or wornout tires. We even hold the individual mature enough to buy a car and decide all by himself whether he wants to buy, on top of mandatory liability coverage, collision insurance for self inflicted damage to his car. For the smallest new car, this is a risk of 10.000 Euro.

  We also entrust our citizens to decide on their own whether they want to marry and have children - decisions that do not only have grave financial consequences. It is, indeed, grotesque to claim that an individual is not competent to decide about a 4.000 Euro deductible. Strangely enough, those that contend so, will say that they themselves, of course, could make such a decision. The guardians of the welfare state lose their elite status when the troops become smart.

• (iii) People will save themselves sick.

  In striving to save money and increase his SSA balances, the individual would, critics claim, cut down on necessary medical services. This is a serious argument. But there are ways to prevent a prudent consumer from skimping on necessary purchases. First, the forced savings concept of the SSA prevents the individual from using his SSA balances for a vacation or a car purchase. Second, deductibles and co-payments can be structured in such a way that they offer incentives for preventive care. For example, there are insurers which waive the deductible if the patient has had his routine check-up. Finally, preventive care that is related to catastrophic illnesses can be included in the mandatory package.

• (iv) Young people will underinsure themselves.

  Young and healthy people may restrict themselves to the mandatory package and buy a frugal plan. When they get older, they may realize, with regret, that switching to a more extensive and comfortable plan is very expensive, since actuarial rates will then be much higher. This argument is valid, but where is the problem? Those that decided on a minimum plan in their earlier years have saved a lot. Their SSA balances allow them, as they become older, to step up to a more expensive plan or to cover extra benefits out of their balances. Or they have, instead of additional health coverage, bought more life insurance or used their SSA balances to invest in a house or other assets on which they can fall back, if necessary. And to repeat: The minimum mandatory coverage is the back-stop that prevents a drop to welfare levels.

• (v) Optional coverage has no effect on hospital expenses.

  Tertiary illnesses - such as cancer or stroke - would be included in the mandatory package. Hospital care takes the largest slice of the health cost pie - in Germany 32.1 per cent (2002).42 If hospital care is in the mandatory package, how would the empowerment reform save money in this important area? In the case of hospital care, there is relatively little moral hazard on the demand side; a deductible makes little sense here: Nobody - with or without a deductible - wants to undergo unnecessary surgery or stay in the hospital for excessive periods. There is, however, considerable waste on the supply side. In a competitive system, such as among U.S. hospitals, hospital stays have been shortened tremendously. In the noncompetitive public health systems of Europe, hospital stays are being shortened now, but the trend has set in much more slowly and we are still far behind U.S. benchmarks. In Europe, there are more hospital beds and hospitalisation occurs more frequently and the average stay is longer than in the U.S.: The average stay for acute care in the U.S. has been 5.7 days as compared to 11.6 in Germany or 9.2 in Switzerland (Table 2).

The German public health system has, until 2004, reimbursed hospitals for the length of a patient's stay, irrespective of treatment. Depending on the type of hospital - general care in a rural region or teaching hospital in a city - a day was billed on average at somewhere about 320 Euro. Per-diem reimbursement have tended to lengthen hospital stays: In order to break even, a patient was kept longer than necessary. Consequently, hospitals had little incentive to optimize. To the contrary: if they decreased their cost, their per-diem was reduced; if they raised their cost, they documented it and tried to get an increase in their per-diem reimbursement.

As of 2004, remuneration is DRG43 based. The change was announced already three years before; therefore, hospitals had adequate time to prepare and have already slashed hospital stays before the change took effect in the beginning of 2004.

With empowerment reform, choice of suppliers means that hospitals will compete for the patient's business. Rather than banding together in hospital associations and trying to extract reimbursement increases from the sickness funds, they will address the patient (or his insurance company as his agent). They will form voluntary alliances with other hospitals, merge, and outsource. The strict separation of ambulatory care by feefor- service physicians from stationary care by hospitals and their doctors will be replaced by flexible cooperation dictated by cost advantages. In trying to bring costs down, hospitals and doctors will make better use of information services. Public hospitals will no longer be run by politicians and their appointees.

Auto liability provides a good analogy: It is mandatory as well, but there is free choice of insurers. Competition among them is stiff. Some of them would prefer a cartel-like setting where they add up their cost, present it to a regulatory body and negotiate rates. Free choice of liability insurers, diversity of plans, bonuses and rebates for safe drivers have kept rates at a competitive minimum.

Likewise, cost savings from opening hospital care to competition will be large, just because hospital care currently takes the largest slice of the cost pie and the bureaucratic system entails so much dead weight loss. None of these cost savings must come at the price of reduced quality. To the contrary, competition always provides the impetus to improve.

Unemployment insurance, as we know it in Europe, disregards this potential for premium savings. The unemployment payroll tax rate, being the same for all, implicitly assumes that the individual, through his own behaviour, cannot influence the risk of losing his job. True, we cannot fault an employee for every job loss; his employer may lay him off for reasons unconnected to individual work performance. Consequently, employers pay a share of the unemployment premium or payroll tax: in Italy, employers pay nearly all (4.41 of 4.71%); in Germany, they pay half (of 6.5%); in the Netherlands, they pay less than employees (1.55 of 7.35%) (note 44).

In June 2003, 34 per cent of unemployed Germans had a new job within three months45, suggesting that some of the unemployed became so voluntarily, having first given up their old job before looking for a new one. In Germany, unemployment compensation begins immediately, on day one and at a level of 60 per cent of net income for those without and 67 per cent for those with children; the unemployed do not pay the payroll tax for health or pension insurance, but continue to enjoy all benefits. Consequently, the difference between the take-home pay of a low-income worker is not much higher than unemployment compensation (note 46).

If the unemployment payroll tax were differentiated, i.e., lower for those that accept an unpaid grace period of six weeks, socially beneficial behaviour would be rewarded: Those that look for a new job would do so while still employed. Total unemployment would be lower, unemployment outlays would decline, making society- at-large better off; and, with a lower unemployment payroll tax, labour cost will decline and jobs will be more competitive.

Training and education have a distinct effect on unemployment. With its vocational training system, Germany is a good example. (The experiences of Austria, Denmark, or Switzerland, all having vocational training systems, are similar.) Few employees in Germany . 13.8 per cent of the labour force - are unskilled, having no vocational training or academic degree; yet they make up more than one third of the unemployed (note 47) and their unemployment is twice the average unemployment rate. On the other hand, those with vocational training, e.g., an industrial mechanic, have an unemployment risk 13 per cent lower than the average (note 48). (Figure 18).

Training is the best unemployment insurance (note 49). Why not reward those that undergo training with a lower unemployment premium? It is justified on individual actuarial considerations. And it is called for on social grounds, since society could be spared many negative effects that come with unemployment.

Training is also the best policy against youth unemployment. Countries with vocational training systems solve the school-towork transition problem. Such countries - Germany, Denmark, Austria, Switzerland - tend to have relatively low youth unemployment (Figure 19).

In turn, in countries without vocational training, youth unemployment rates may even exceed 20 per cent (Spain, France, Italy). As youth unemployment often lies at the root of criminal behaviour, youth training yields a win-win-situation for individual and society. Why not make youth training attractive with the prospect of a lower, actuarially warranted unemployment premium?

Furthermore, rebates and bonuses could be considered for those that have a good employment record, similar to auto insurance for drivers with an accident-free history. Whoever has been employed for a long time period, may not just have had plain luck, but often his attitude and efforts - such as further training - have made his job safe.

The empowerment model is ideal for incorporating these features:

• (1) Each worker receives - in the form of a deposit in his Social Savings Account - the money necessary to finance (insure) his current benefits. In the Bismarckian countries, the employer and employee shares of the unemployment payroll tax (in Germany, each side pays 50% of a payroll tax of 6.5%) would be put into the SSA. In the Beveridgean countries, the money would come from the government's current tax sources (Figure 20). In turn, the employee is required to finance his unemployment insurance. He is able to buy the current benefits.


• (2) The employee is, however, not required to buy unemployment insurance according to current benefits. Mandatory or minimum coverage would be at least as high as welfare support. Beyond that, coverage would be optional. Also, coverage would not have to start from day one, but, say, after six weeks in order to minimize search unemployment and entice people to start looking for a new job while still employed. The unemployment payroll tax would be differentiated according to training, i.e., risk categories.50 Likewise, bonuses reflecting the employment record would be offered.


• (3) In such a system, there is little reason for a monopoly unemployment agency. Unemployment coverage could be offered by any willing insurer. The beneficial effect would be that competing insurers will take an immense interest in job counselling. While that is currently done by government unemployment agencies, anecdotes about their inefficiency are over-abundant.
• (4) Any premium savings below the current payroll tax will be the employee's to keep and will augment his SSA balance. The employee can withdraw from it in case of unemployment or he can use it for other social security purposes. Premium savings from sick leave and unemployment will probably be the major sources for building up a fully funded pension plan.


• (5) Again, the fifth criterion is that the individual is free to choose his SSA fund manager.

As with health or sick leave, the empowerment model transforms the individual into a critical buyer. In the case of unemployment insurance, he will, especially, be critical of spending that is not insurance related. In many countries, governments have shifted make-work programs and other costly labour market policies onto their unemployment agencies. While these receive government funds in carrying out these mandates, they are usually underfunded, implying that the employees, without having been asked, must subsidise the remainder with their payroll taxes.

To the extent that only those that pay into unemployment insurance are the beneficiaries of these labour market policies, the subsidies may be justified. But since anybody in society may be a beneficiary, it is unjustified to impose the cost on the employees and their employers alone. The empowerment reform would lead to a demand by the insured employees that others also pay their share, i.e., that such labour market policies are fully paid out of general government revenues. This will lead to an honest discussion of the merits of such policies. In Germany, approximately 40 per cent of the unemployment agency's expenditures are for non-unemployment compensation, i.e., labor market policies. In 2003, the government paid about 11 per cent (6.2 billion Euro) of the unemployment agency's budget (57 billion Euro)(note 51). In a fair system, the unemployment payroll tax would be 4 rather than 6.5 per cent.

54. PENSION REFORM

a. PENSION REFORM: The Problem: The Generational Contract

Old age provision typically comes from three sources: mandatory government pension plans, company retirement plans, or voluntary, individual old age provisions (Figure 21).

In most Western countries the mandatory government pension plan dwarfs the two other provisions. The mandatory old age pension plans of the welfare states are financed in a pay-as-you-go system. It is based on the "generational contract", according to which the currently active employees pay for the pensions of the elderly and also raise the future generation, on the tacit understanding that the future generation will pay the pensions of the active generation, once it reaches retirement.

The generational contract is a euphemism of unrivalled cosmetic camouflage: There is no contract spelling out a binding formula with individual responsibilities firmly tied to benefits; one contracting party is not yet alive or has not yet grown up; and the actively employed may violate the contract with impunity - by not generating the offspring as the payers of the future.

We speak of social "security" and old age "insurance". This is a dangerous misnomer. Our language affects our thinking and we believe that the pension system is, indeed, insurance. But the pay-as-you-go system provides today's money only for today's elderly (yesterday's workers) in return for their previous efforts in raising the active generation. The active generation does not provide for its own future pension. Only that could be called insurance; it could take either of two forms,
- capital formation: with extra savings we can build up a fully funded pension plan which we then draw down in retirement; or
- human capital formation: raising children that will support us.

Financially, our mandatory pension systems have either no or insufficient funds to cover future pension claims.52 Demographically, we have broken the generational contract by no longer raising enough children as the alternative to a capital-funded plan.

The generational contract is a chain-letter, being sent by the pensioner to several actively employed, asking them to jointly share his pension payment. The letter concludes with the advice to generate a sufficient number of children, raise and educate them so that, when the parent generation is about to retire, it will have a sufficient number of addresses to mail similar letters to. If that advice were strictly adhered to, we could call this old age provision or insurance. It is but a pyramid scheme.

In a chain-letter, no letter writer can hold his addressees liable for his claim. Those that have raised children do not get a higher pension than those that did not, to the contrary: Women may get some pension credit for raising children, but in any case less than when they forego children for work. Consequently, the insurance notion of old age provision - raising an adequate future generation - is fully discredited.

Pyramids are built on sand. As the parent generation has failed to generate enough children, it realizes that it has cut off the branch on which it is sitting and has become a victim of insurance illusion. Nonetheless, as this generation retires, all members expect full benefits; looking for Ersatz support, they turn to the state and take it hostage as the presumed guarantor of the pyramid scheme. The ploy works, since with fewer active people, the pensioners will have sufficient voting power. Bismarck did not plan it this way, but neither was he a fan of universal suffrage.

In defence of the pay-as-you-go pension system, its supporters invoke solidarity. Nothing could be further from the truth. Solidarity asks society to stand shoulder to shoulder, with the strong supporting the weak, to the extent that the latter cannot help themselves. Solidarity also entails a quid pro quo: If the present generation expects the support of the future generation, it first has to do its best to raise the offspring to the necessary strength. But, having failed that, it usurps its voting power by taking out an onerous mortgage on a numerically weak future generation and assigning government as the trustee to this selfish scheme. Politicians find it to their gain to endorse and sanction this antithesis to solidarity. With such hypocritical abuse of the concept of solidarity, it is no wonder that the young respond with cynicism and even give the cold shoulder to those that legitimately need compassion.

The major fault is not that future generations are not being asked whether they want to "sign" the contract: Born into a society, we have to assume many responsibilities imposed and already assumed by generations ahead of us; in turn, we also receive, as a gift, the hard- and software, infrastructure and culture, that earlier generations have left behind. If we do not like the dowry, we can emigrate.

The major fault is that the active generation can uphold its claim of future pension receipts in full conscience of the fact that it has not sufficiently raised children and invested in their human capital. We do not want to hear that.

When Bismarck introduced pension insurance in 1889, the generational contract posed no imminent problem. There were many children, more than the actively employed, and there was a tiny number of elderly that could claim benefits, as life expectancy was about 45. Consequently, the age pyramid had a broad base and it was not high. Pension reform had the effect of lowering the birth rate, desirable at that time. Of course, a fully funded pension system would, even at that time, have been the better choice. We must correct for that now. In the 1950's, the demographic structure still resembled the pyramidical shape (with some war-related dents). Today, it resembles a pumpkin: few children, still many actively employed, and a rapidly growing number of retirees.

From 2001 to 2050, the share of those who are over sixty will rise from 24 to 37 per cent in the Germanys population. (Figure 22).

Changes within the pay-as-you-go system will not be able to reverse the trend. Furthermore, government policies have been partly offsetting: While child support payments and subsidies to families favour raising children, other policies are in support of women participating in the labour force which raises the payroll tax volume of the active generation, but also helps to raise the average age of women when they first give birth and reduces their desire to have (several) children. The retirement age also can be raised to balance the books, but that will now only partially offset the past trend towards early retirement (Figure 23).

All in all, such measures provide some relief and buy time. They do not solve the problem. The system is on auto-pilot towards insolvency (Figure 23)

After 2020, the demographic structure will come to resemble a mushroom: A weak stem of young and actively employed, many elderly. Around or soon after 2030, Germany will have one pensioner per one payroll-tax paying employee. Immigration, retirement age, labour force participation are variables that will determine whether this happens in 2028 or 2038, but they will not help us avoid the fateful outcome: That around that time, the average payroll tax would have to equal the average pension, unless, of course, pension benefits would be radically cut.

The average German pension payroll tax, standing at 19.5 per cent in 2004, would have to be raised to around 24 per cent, if the current structures of benefits and finances are to be preserved (Figure 25 and Figure 26) (note 53).

It is inconceivable that the actively employed of the year 2030 will be willing to pay that, as much as it is inconceivable that there will be enough jobs to support a payroll tax of that magnitude.

As a consequence various pension reforms aim to reduce the benefits while trying to keep the payroll tax from rising above the current level. The decline in benefits will be substantial, as can be seen from the projections in figure 26.

Whoever wants to assure the pensioners and those soon to retire that their pensions will be safe, must show how that can be achieved - not just for the year 2010, but for 2020 and 2030. Those entering the labour force today and beginning to pay into the system, will retire around 2030. They can expect an answer.

b. PENSION REFORM: Solution: Fully Funded Pensions and Family Values

With the generational contract defunct, we have only this alternative: either return to the pre-20th century intra-family "contract" under which the active generation cares for both its own parents and children, or take care of your own retirement by building up a capital stock for old age. We may yet see a partial return to the seemingly outdated bondage of three generations within a family. This will and need not take the form of three generations living under the same roof. In the 20th century, the societal, government-instigated intergenerational contract has contributed to the erosion of intra- family bonds, with the government posturing as a more reliable guarantor than family tradition. Why should this trend not be reversed in the 21st century, as we recognize that building on government was building on sand? Is it not cynical to deride compassion for those we know personally and whose needs we can judge ourselves, while demanding anonymous solidarity across society, with a bureaucratic entitlement system often giving too little to the needy and too much to those that can take care of themselves? The empowerment reform twins effective compassion and individual responsibility.

As un-guaranteed as the family contract has been (there never was and will be assurance that children will be willing and able to care for their parents), at least it builds on common values. Ethical norms are never guaranteed, but expected to be adhered to. While not air-tight and rock-solid, they are more stable than the current system's house of cards.

The second, complementary or alternative, pillar of the future old age provision will be a fully funded pension system as a complete or at least partial replacement of the pay-as-you-go system. Full replacement would be a gigantic task, especially in those countries where government retirement plans play the dominant role. It is only achievable if society is willing to shoulder an enormous dual burden: Those in the labour force would have to continue to finance the pensions of both the elderly and those soon to retire who no longer can build up a sufficiently sized fund; and the young would have to set aside further savings for building up their own retirement fund, so that, eventually, the next generation would only be responsible for itself. Even if the process would be stretched over an entire generation, the burden is enormous. Societal consensus on such a radical effort can probably only be reached in a situation of severe crisis, such as in Chile in 1980.

It is seemingly paradoxical: The less bankrupt the current system is, the less will a society be willing to change course to avoid catastrophe. But recognition of the right path should suffice to make at least partial provisions. Furthermore, the success of even a partially funded pension plan will unfold persuasive power. It is, therefore, not necessary to demand the full replacement of the current system, reasonable as that would be. Appetite increases with eating.

Even partial and gradual replacement of the current system by a fully funded plan will require the current generation to save considerably more and sacrifice consumption. The sooner we start, the smaller this sacrifice will be, and the sooner we can ease up. Asking for such sacrifice is not asking too much at this time. For the first time in this century, the young generation inherits assets that were not destroyed by a war or devalued by a hyper-inflation, especially so in Germany. Compared with the generation that lived through the first half of this century, this extra and only financial burden is moderate.

Furthermore, in an inter-temporal sense there really is no extra burden: If we keep the pay-as-you-go system, payroll taxes will skyrocket, imposing a tremendous extra mortgage on future generations. By building up a fully funded system, we discount the otherwise inevitable future burden and shoulder it, now. Moreover, there is a net gain: The eventual bankruptcy of the current pyramid scheme would come at the cost of social and political upheaval. By building up a fully funded plan, the current generations do have to shoulder an extra burden, but the future generations will be relieved of vast financial and social consequences. In an inter-temporal sense, this investment will have a high return.

The current generations' extra burden can be eased by the savings in the Social Savings Accounts. The more prudent employees become in drawing benefits from the other areas of sick leave, health and unemployment insurance, the less "fresh money" they will need in building up their retirement funds. While the empowerment model does ask the individual to get out from underneath the welfare state's mouldy but still warm blanket and test the breeze of individual responsibility, the SSA balances saved from other social security areas provide the charming comfort of a warm-up. Reforming all social security areas offers attractive synergies.

Apart from financing an extra burden now for the sake of preventing disaster down the road, pension reform takes the same steps as in the other areas:

• (1) Put all current pension payroll taxes, employer- and employee-paid, into the Social Savings Account of the employee. Initially, this money will only be in the SSA in transit, as it has to be paid out to the pension recipients that can no longer build up a retirement fund (Figure 27).
  At the same time, individuals will be asked to gradually build up a fully funded plan. This can be sped up, if we decide to allow individuals to use part of their current payroll tax for that purpose, and cover the shortfall in current pension outlays out of the government budget. If compensating budget cuts are unavailable and a tax increase is unavoidable, raising the value added tax would be least harmful, as this tax on consumption favours savings, savings which we need for additional voluntary contributions into the new funded pension system. In addition, exempting these voluntary contributions from income tax would favour saving still more (note 54).
(2) As the claims under the pay-as-you-go system decline with retirees passing away, more and more of the payroll tax can be steered towards the fully funded plan. Back to contents

VI. CONCLUSION

Individual Responsibility and Effective Compassion Reforming the welfare state is inevitable if we want to preserve social security: Without reform, the social systems would collapse. But we need not take an axe to benefits. A marketoriented reform whose cornerstone is individual responsibility need not be bloody.

With the empowerment reform, we get rid of the welfare state's top-down entitlement system. In the new social state, we let individual responsibility first start-up, then drive the system. Effective compassion takes the place of solidarity. The reform puts the individual where he belongs: at centre stage. He is both enabled and made responsible to make decisions. The call for efficiency and individual responsibility cannot be criticized as coming from a cold heart disregarding individual needs. By first financially empowering the individual to finance his choice, the no-loss rule is preserved. With the dual emphasis on individual empowerment and responsibility, the call for efficiency, often perceived as abstract, assumes a tangible, personal meaning: Individual purchasing power to buy current benefits, individual appropriation of savings resulting from voluntary restraint. Second, the consequences of poor choices are mitigated, since mandatory coverage levels provide a safety net. We provide a road system, traffic rules, guard rails, but do not dictate the choice of car vs. bus. We do prevent people from driving off the cliff.

The Social Savings Account can be applied to all of the four areas dealt with here - sick leave, health, unemployment, pension. That is a major advantage. Substitution within each area as well as between them - savings from unemployment insurance used for pension insurance - is not just possible, but 75 desirable, a feature that unfolds particular benefits for the gargantuan task of pension reform. Everyone can compose his social security menu according to his needs, with a minimum level being mandatory, but the provider being always the individual's to choose.

Inasmuch as the current beneficiaries, mostly employees, can only gain under this proposal, one might question why the noloss rule should be so strictly applied. If employees' sick leaves will decline, if they opt for lower unemployment compensation, would it not be better to share some of their premium savings with their employers? Labour cost would decline faster, job growth would be higher, the financial base for the social system would broaden - a win-win situation for employers and employees. We leave that decision to employees and employers and hope that such common sense will prevail. What we wanted to show is that, even when employees should refuse such concessions, the system can be successfully reformed. Employers, remember, would also gain, even when they fully transfer their current cost or payroll tax share into the SSA's of the employees. In the past, these components of indirect labour cost have risen much faster than direct wages. With their transfer into the SSA, they become part of direct wages and, in the future, rise only in step with them.

In regard to the no-loss rule it is also legitimate to question why all the current cross-subsidies should be sacrosanct: Why should it be written in stone that even well-off pensioners should have their health care expenses subsidized by the actively employed? Why should those that have foregone income while being trained carry part of the unemployment cost of the unskilled that have spurned training? Why should single and dual income households subsidize the health costs of middle-income families?

Today, all of these subsidies are being justified in the name of solidarity. Whoever questions them is reprimanded as lacking compassion. But does compassion demand even middle-class income transfers, which amount to a regressive tax? We remain mute on this question - not for lack of personal opinion, but because this judgment is society's to make. And it will do so. When enacted, this proposal will open the Pandora's box of debate on these subsidies. Once the currently hidden income transfers are made explicit and individually transparent on everybody's Social Savings Account, the curtain on this debate will open. Few people in Germany, for example, are currently aware that more than half of the health payroll tax goes towards such transfers. Information on who and how much is subsidized is an important side-effect of this reform. Full information is a pre-condition for a sound democratic debate. We must not keep the veil over these data.

Whatever the outcome of this debate, whether there is an outcry over these transfers or whether they are confirmed, the meaning of solidarity will no longer be abstract, unrelated to the knowledge of who receives what and whether he deserves it. Solidarity will regain the meaning that it has lost in the wake of the welfare state: people's intentioned and informed compassion for each other.

Simple, Radical, Social.

The proposed empowerment reform does not slash the social safety net. While it is radical in the sense of a root-change - abolishing the welfare state -, it neither axes benefits nor institutions. It allows the individual to evolve from passive beneficiary of entitlements into an active and sovereign consumer of social insurance products. A society of free and mature individuals does not treat its citizens as wards. And the empowerment reform also allows the current entitlement bureaucracies to evolve into efficient, client-driven providers.

The empowerment reform is social in the sense of adhering to the no-loss rule. Nobody loses except for those that abuse the system and contribute to efficiency losses. The quality of social security is, at the minimum, preserved. Its cost goes down. The empowerment reform is simple, because it can be financed, because it can build on current institutions and, above all, because it builds on people's ability to decide responsibly while also, as a back-stop, having a safeguard against individual errors.

When neither employees, employers nor society lose, and if each party senses more pleasure over its own than envy over the other's gain, then this reform should be a self-starter. Simple, unfortunately, is not easy. As Ronald Reagan remarked: "The experts tell us there are no simple answers to our difficulties. They are wrong. There are simple answers, just not easy ones."

Yes, empowerment reform is difficult. It faces the powerful opposition of the paternalistic guardians of the welfare state, whose status would melt like snow in spring once we are empowered to act with self-responsibility. We have accepted this custodial structure for too long; the welfare state has corrupted our minds, many of us having become insecure, doubting that we can decide for ourselves. Building self-confidence towards individual responsibility is the most difficult task. And it is worth the effort.

Back to contents

ENDNOTES

1: Want, disease, ignorance, squalor, and idleness.

2: The no-loss rule cannot be claimed by those that abuse the system.

3 Christoph Schroder: Industrielle Arbeitskosten im internationalen Vergleich. In: iw-trends 3/2003, Cologne, p. 10.

4 Institut der deutschen Wirtschaft (IW): Deutschland in Zahlen, various annual editions. For 2002, the Institut der deutschen Wirtschaft reports 77.9 percent, compared with 77.4 per cent in 2001. Differences in the data presented in the table are based on changes in the calculation method. Institut der deutschen Wirtschaft (IW): Personalzusatzkosten in der Deutschen Wirtschaft. In: iw-trends 3/2003.

5 OECD: Taxing Wages: Income Tax, Social Security Contributions and Cash Family Benefits 2001-2002. Paris, 2003, p. 16 and own calculations.

6 If the pension benefits would not be lowered (as they have been by several measures between 2000 and 2004).

7 The German public health system's mandatory sickness funds - Gesetzliche Krankenkassen - could be more aptly named "kranke Gesetzeskassen" - sick mandatory funds.

8 For an authoritative work on consumer-driven health care, see Regina Herzlinger: Market-Driven Health Care, Perseus Books, Cambridge, Mass. 1997.

9 In the case of unemployment insurance, similar subsidization takes place, where unemployment compensation is linked to family size and net income, while the payroll tax is based on gross income. And the pension system gives credit for time spent in the military or unemployment, among others, and does not make full actuarial deductions for early retirement. The pension system also gives credit for time spent in higher education or for rearing children. But these credits can be justified. In a pay-asyou- go system, children will pay the current generation's pensions, and higher income earners pay a higher payroll tax.

10 The mandatory minimum level can also be made dependent on income.

11 This restriction can, of course, be relaxed, if a substantial sum has been accumulated in the SSA. However, withdrawal for consumption purposes would require that the amount is first subjected to income taxation, since the deposits had been tax-free.

12 The range of investments that SSA balances can be used for can also be made dependent on the size of SSA balances. As the SSA is built up, more illiquid and higher risk investments can be considered.

13 If the balances would not go to the holder's estate, incentives to save them would be diminished, especially in old age; furthermore, while there should, in general, not be any inheritance tax, exempting at least SSA balances from inheritance taxes should be a given, since the future generation needs this capital for its own social security, especially in the long transition phase needed to build up a fully funded pension plan.

14 Bundesministerium fur Gesundheit und Soziale Sicherung: Gesetzliche Krankenversicherung . Personal und Verwaltungskosten 2002. Berlin 2003, p. 7 and 10.

15 To qualify for an HSA, the consumer must opt for a highdeductible plan with a minimum deductible of $1,000 for individuals and $2,000 for family coverage. For 2004, the consumer can then invest, tax-free, up to $2,600 (individuals) and $5,150 (families) into the HSA as long as the deductible is at least that high. Withdrawals from the HSA can be made tax-free for medical expenses. After age 65, money can also be withdrawn for any purpose and is subject to income tax. Therefore, the HSA extends beyond health and becomes a retirement account. Much credit for developing and promoting consumer-driven plans and savings account concepts is owed to market-oriented think tanks, in particular John C. Goodman of the National Center for Policy Analysis, www.ncpa.org, Greg Scandlen and Grace-Marie Turner of the Galen Institute, www.galen.org, Joseph Antos and Robert Helms of the American Enterprise Institute, www.aei.org, Stuart Butler and Robert Moffit of the Heritage Foundation, www.heritage.org.

16 Michael Rothschild, Why Health Care Reform Died, Wall Street Journal, September 22, 1994.

17 The rates pertain to 2003. Missoc . Mutual Information System on Social Protection in the EU Member States and the EEA, 2003.

18 Rigmar Osterkamp: Work Lost Due to Illness - An International Comparison. Cesifo Forum 4/2002, p. 37.

19 Christoph Schroder: Personalzusatzkosten in der deutschen Wirtschaft. iw-trends 2/2003, p. 11.

20 Sick leaves also fluctuate relative to job risks. They tend to be lower in recessions, further proof that health conditions are not the primary cause of variations in sick leaves.

21 In keeping with the no-loss rule, the maximum pay-out has to be the firm-specific cost. Companies with sick leave expenses lower than the average would be penalized if they paid out on the basis of the industry average.

22 Unemployment compensation, by definition, precludes a drop to welfare levels.

23 Rigmar Osterkamp: Work Lost Due to Illness - An International Comparison. Cesifo Forum 4/2002, p. 39

24 Nordiske Arbeidsgiverforeninger 2004.

25 In October 1996, sick leave payments, which have to be paid by employers, had been reduced to a legal minimum of 80 per cent . instead of 100 per cent . of income, although all labour contracts kept calling for 100 per cent. After the Social Democrats had won the 1998 elections, the mandatory sick leave payments were increased again to 100 per cent of income in 1999.

26 In Germany, total sick leave costs are estimated to be 71 billion Euro, or respectively, 3.44 per cent of GDP (Bundesanstalt fur Arbeitsschutz und Arbeitsmedizin. Kosten durch Arbeitsunfahigkeit 2001). According to the German employers' association, companies alone have to pay about 37 billion Euro for outright wage / salary compensation (BDA: Strukturen und Entwicklungen bei krankeitsbedingten Fehlzeiten. In: Soziale Selbstverwaltung, February 2004, p. 10 f.).

27 BDA: Strukturen und Entwicklungen bei krankheitsbedingten Fehlzeiten. In: Soziale Selbstverwaltung, February 2004, p. 11.

28 Statistisches Bundesamt: Index der tariflichen Monatsgehalter.

29 In public health insurance in West Germany, the share of administrative costs to total expenditures increased from 4.3 per cent in 1980 to 5.5 per cent in 2002 (Institut der deutschen Wirtschaft (IW): Deutschland in Zahlen 2003. Cologne 2003, p. 77.

30 A health maintenance organization (HMO) can be defined as a pre-paid service plan: Delivery of care and insurance are integrated. There is a fixed per-capita-fee per year/month irrespective of use of care. All use of the HMO's services are paid for by this capitated fee.

31 The rates pertain to 2003. Missoc . Mutual Information System on Social Protection in the EU Member States and the EEA, 2003.

32 In most Bismarckian countries (e.g. Germany, Greece, Luxembourg, Netherlands, Spain), there is an upper limit to the income subject to the payroll tax - in Germany's case and for 2004, the limit is at . 61,800 p.a. in case of pension and unemployment insurance, . 41.850 in case of health and nursing care insurance.

33 Of course, any single one of these characteristics does not suffice to make the payroll tax exceed the full actuarial value of benefits.

34 Until 1983, pensioners did not contribute at all; until 1989, they contributed at half of the payroll tax rate, and since 1989, at the full payroll tax rate. For pensioners, since they are no longer employed, there is no employer's share. As of April 2002, however, the full payroll tax is also assessed on pensions from company plans. This will increase the pensioners' share in the middle bar of Figure 17).

35 In contrast to U.S. Medicare, there is no fund for old age reserves from which at least part of the medical expenses of the elderly are paid.

36 In the transition from the old to the new system, an insurer or provider must not deny coverage to somebody with a preexisting condition or charge a higher premium. Correspondingly, the noloss rule dictates that a provider's losses from having to take on such bad risks would be recouped from slightly higher than actuarially calculated premiums charged to the vast majority. If, in the future, a good risk develops into a bad one, this poses no problem, as this is the nature of insurance. Also, free choice of providers must not be restricted to the first time of signing up, since otherwise a beneficiary would be locked in as he gets old. To allow for switching from one provider to another one at any time, providers must build up actuarially adequate old-age funds, since health expenses rise with age. The old-age fund must be vested in the beneficiary and thus be portable if he decides to switch to another provider. Since the current system does not have an old-age fund, building it up requires extra money to be spent now, i.e., a mark-up on the premium that reflects current actuarial benefits. In an intertemporal evaluation, however, this is not an added cost, since an old age fund would assure that the premium does not rise with age, in contrast to the age burden that we will have to expect under the current generational contract and our demographic trends.

37 As noted before, the distinction to medical savings accounts as applied in the U.S. is that unspent balances cannot, at year end, be withdrawn for just any spending, a vacation or other consumption spending. The reason is that the SSA encompasses all of social security.

38 This example underlines that reforming both sick-leave and health policies entails synergies.

39 In Germany, the administrative cost of the public health system amounts to 5,6 per cent of expenditures, or over 8 billion Euro in 2002, or 157 Euro per insured and year. (Bundesministerium fur Gesundheit und Soziale Sicherheit: Endgultige Rechnungsergebnisse der gesetzlichen Krankenversicherung nach der Statistik KJ 1 . 2002. Bonn 2003, p. 116 and 225.)

40 A hospital catering to patients of the sickness funds . which encompass 90 per cent of the population - has to obtain a state permit which imposes a ceiling on the number of beds etc., accompanied by a state commitment to pay for investment cost. With an excess supply of hospital beds, state consent to build or expand a hospital is difficult to obtain; and many communities are opposed to privatising their hospitals out of fear that they lose "control", although these hospitals typically require subsidies.

41 In physician practice-management companies, doctors do not operate independent practices, but belong, with their practice, to a tightly organized company. Economies of scale in purchasing supplies, effective use of equipment and savings of overhead can lower costs.

42 Bundesministerium fur Gesundheit und Soziale Sicherung: Gesetzliche Krankenversicherung . Personal und Verwaltungskosten 2002. Berlin 2003, p. 113.

43 DRG = Diagnosis Related Groups.

44 The rates pertain to 2003. Missoc . Mutual Information System on Social Protection in the EU Member States and the EEA, 2003.

45 Bundesanstalt fur Arbeit: Strukturanalyse . Zu- und Abgange an Arbeitslosen im Juni 2003 . Vorlaufige Ergebnisse. Nurnberg, 2003, p. 3.

46 This only holds for lower or medium incomes, since there is a cap on unemployment compensation as well as on the income that is subject to the payroll tax.

47 35.4 per cent in 2002.

48 Alexander Reinberg, Markus Hummel: Geringqualifizierte . In der Krise verdrangt, sogar im Boom vergessen. In: IAB Kurzbericht, Nr. 19 / 2003, p. 4.

49 Enthusiastic support for vocational training cannot be interpreted, as others might do, that training is a cure-all for unemployment. If it were, German unemployment would not be as high as it is. Vocational training can mitigate, but not outweigh the labour cost disadvantage of Germany.

50 A premium at full actuarial value would then mean that the unskilled would pay a higher premium than currently, if they sought current benefits; this would violate our no-loss rule. Therefore, the premium discount for those seeking training could be phased in gradually.

51 Bundesagentur fur Arbeit: Quartalsbericht der Bundesagentur fur Arbeit - Bericht uber das vierte Quartal und Kurzbericht zum Geschaftsjahr 2003. Nurnberg 2004, p. 5 and 12.

52 Germany's public pension system has no fund at all, and its liquidity reserves are hovering around zero, not even covering one month. Other countries�f pay-as-you-go-systems do not fare much better.

53 Nachhaltigkeit in der Finanzierung der Sozialen Sicherungssysteme - Bericht der Kommission unter dem Vorsitzenden Bert Rurup, Berlin 2003, p. 99.

54 Such tax exemption could be criticized as endangering efforts to balance the budget. An alternative would be to pay contributions out of after-tax income, but exempt the interest from taxation. This tax loss on future interest is outweighed or at least partially compensated by the growth-enhancing effect of these savings.