PART TWO—
THE INVESTMENTS APPROACH TO SOCIAL PROGRAMS

Introduction to Case Studies

In the preceding chapters I have examined several sets of obstacles to the establishment of basic socioeconomic rights and to the use of public social programs to realize these rights. I now want to apply these analyses to case studies of three existing programs. In turn, in chapter 8 I will integrate the conclusions of these case studies and detail a new approach—the investments approach—to American social provision.

The case studies presented here employ the focused comparison technique of Alexander George.^[1] The following questions guide these studies:

Distributive Justice

What are the criteria for the distribution of benefits?

Do they reflect effort or results and thus appeal to American concerns for human dignity and vertical equity?

Do they reflect need? If so, are they categorical, thus raising concerns for horizontal equity? What mechanisms exist for certifying the need?

To what degree are benefits linked to the provision of basic goods?

To what degree does this program represent an attack on traditional property rights?

Negative Liberty and Individualism

Is the program compulsory or voluntary?

In what ways and roughly to what degree do the taxes supporting this program constrain liberty?

Aside from compulsory taxation, to what degree and in what ways are the liberties of benefit recipients constrained?

To what degree and in what ways is the liberty of various third-party providers constrained?

Economic Efficiency

To what degree and in what ways might this program stimulate or retard propensities to work in the paid labor force?

To what degree and in what ways might this program stimulate or retard household propensities to save?

To what degree and in what ways might this program influence the propensity for corporate managers or entrepreneurs to take risks?

To what degree and in what ways might this program influence the international competitiveness of American goods and services?

What are the direct monetary costs of the program and who pays them? Is redistribution horizontal, intergenerational, or vertical?

To what degree does this program target a population with particularly difficult problems? What, if any, disturbing consequences do these problems hold for program operations?

Administrative Complexity

How do the tasks the program requires relate to its administrative capacities? What is the program's administrative efficiency?

How effectively are limits or harmonization problems managed?

How does the public profile of the program relate to abstract American preferences for limited and local government?

Conflicts of Interest and Compliance

To what degree and in what ways does the program raise important conflicts of interest among explicitly targeted beneficiaries, national government officials, and third-party (public or private) providers?

To what degree and in what ways is the program structured to provide incentives for cooperation with program objectives from explicitly targeted beneficiaries, national government officials, and third-party (public or private) providers?

Five—
Social Security

As a social insurance program that provides income protection against the hazards of old age, survivorship, and disability, social security represents a limited but generally encouraging model for the investments approach. Medicare, the medical services component of social security, involves both social insurance and other characteristics and is therefore discussed separately, in chapter 7.

Historical Background

The Social Security Act of 1935 marked the first inclusive, long-term effort on the part of the American national government to cope with social hazards.^[1] Civil War pensions, for example, were both categorical and short-term. When these pensioners died off, the United States did not, as some progressives had hoped, follow the industrial nations of Western Europe in creating social insurance for the general population.^[2] Instead, national government efforts declined, and there was less federal public social provision in the early twentieth century than there had been in the late nineteenth. A few state governments, spurred by the American Association of Labor Legislation and other groups, launched their own programs, but these efforts were slow and piecemeal at best.^[3] The principal successes were in workers' compensation programs and, to a lesser degree, unemployment benefits and "widows' pensions."

The depression of the 1930s brought the national government back into the enterprise of helping victims of social hazards. Prior to 1940, however, this help came in the form of emergency pro-

grams offering work or public assistance in response to the exceptional difficulties posed by the worldwide economic crisis. The Social Security Act, meager as its initial provisions now seem, was the first effort on the part of the national government to help vulnerable citizens generally and on a continuing basis. The Social Security Board (SSB) portions of the 1935 act involved three programs: unemployment; public assistance for poor elderly and blind citizens (now SSI) as well as some children (now AFDC); and—of least significance at the time—social insurance for the elderly.

Payroll taxes, an equal and modest 1 percent of a $3,000 wage base on both employer and employee, were instituted in 1937; at that time only workers employed in good-sized private industrial and commercial operations were covered. Under the original plan, modest benefits, starting at $10 a month, were to begin in 1942. But in 1939 the initiation of benefits was moved up to 1940, and eligibility was extended to dependents of both retired and deceased covered workers. This had the effect of shifting social security from a reserve to a pay-as-we-go system.^[4] Additionally, SSB became part of the newly created Federal Security Agency and no longer routinely reported directly to the president. In 1946 SSB was renamed the Social Security Administration (SSA), and in 1949 the unemployment compensation program was moved to the Department of Labor.

Under a major expansion of social security in 1950, benefits were raised by about 75 percent, coverage was extended to self-employed persons and domestic and farm laborers, and state and local government employees were given the option of participating. In 1953 SSA became part of the new Department of Health, Education, and Welfare (HEW). The expansion of eligibility and increases in benefits, tax rates, and the wage base continued such that by 1954 benefits were about double their pre-1950 level, and the vast majority of the labor force was covered.

After a lengthy struggle, disability coverage was added in 1956. Other changes included the extension of coverage to professional groups and the liberalizing of rules, particularly with respect to disability claims. Administration of public assistance was transferred in 1963 to a newly created bureau within HEW. Health insurance for the elderly (Medicare), the focus of a gargantuan debate, was introduced in 1965; SSI, a public assistance program for

poor, blind, elderly, and disabled citizens and administered by social security, was inaugurated in 1974. Throughout this period, the tax rate and wage base continued to climb.

The new era of expansion begun in the early 1970s marked a break with SSA's tradition of careful growth shepherded by the program's senior executives and a pair of congressional committees (Ways and Means in the House, and Finance in the Senate). Some attribute the impetus for growth to the ambitious aspirations of Robert M. Ball, then the SSA commissioner, and his fellow executives. But a variety of interest groups and public officials were also eager to see the program expand. Generous increases in benefits proved to be politically popular, and benefits were raised by 15 percent in 1969, by 10 percent in 1971, by 20 percent in 1972, and by roughly 10 percent in 1973–74.

Two dramatic changes were introduced in 1972. First, benefits were to be subject to automatic cost-of-living increases. Some proponents argued that price-indexing was essential to ensuring steady purchasing power, thus maintaining the relationship between benefits and need. Others viewed the automatic increases as a politically expedient way of rationalizing election-year increases in benefits.

The second change concerned the economic assumptions to be used in calculating the program's future stream of revenues. From the early 1950s the program had calculated its fiscal soundness through a comparison of future liabilities and income. Under an assumption that the program would be continued indefinitely, social security, in contrast to private insurance programs, had not attempted to accrue the funds to cover all future liabilities. Rather, it sought balance through assuring the future revenues to cover these liabilities. One aspect of the calculation of future revenues involved the level wage assumption , under which future revenues were estimated on the basis of current wage levels. But since wages increased over time, the level wage assumption repeatedly produced short-term surpluses. This manner of calculation, with its built-in margin for error, typified the conservative approach of SSA executives during the program's formative years. The dropping of the level wage assumption in 1972, however, removed this actuarial cushion.

Unfortunately, these two changes were enacted just as the econ-

omy slipped into stagflation. The ensuing double-digit inflation meant hefty annual increases in benefits, while high levels of unemployment meant lower revenues for the program.^[5] Short-term deficits occurred and long-term deficits were predicted, based on the falling worker-beneficiary ratio forecast for the early twenty-first century. For the first time since the program's earliest years, serious concerns were voiced about social security's financial solvency.

These difficulties gave rise to a decade-long struggle over funding and benefit practices.^[6] Both the Carter and Reagan administrations took modest measures to cut costs and increase revenues. If these measures are maintained, the program will be solvent through the year 2025, but more fundamental changes are likely to be necessary to accommodate the long-term decline in the worker-beneficiary ratio.

Despite current efforts to slow program growth, social security has continued to expand, both in the numbers of labor-force participants and beneficiaries and in absolute and relative costs. In 1986, nearly 38 million Americans (slightly less than 20 percent of the population) received cash transfers from social security, and benefits are now, even by Western European standards, fairly generous. Although opposition to particular operational aspects has been growing, the program remains popular, and many would agree that "the social security system is among the most effective and successful institutions ever developed in the United States."^[7] As noted in earlier chapters, much of social security's success and popularity reflects the visions of its creators, who presented their innovative program in terms of cherished American political symbols. More importantly, the success of social security in helping citizens survive old age and disability in a dignified manner has indelibly influenced Americans' attitudes toward the general concept of public social provision.

Distributive Justice

Social security operationalizes a sense of vertical, or work, equity.^[8] Recipients have extensive work histories (or, minimally, work histories covering the bulk of their adult lives) and have made regular contributions through payroll taxes.^[9] The program posed some horizontal, or support, equity problems for workers who were in-

eligible because of their age at the time social security was initiated, and a few related problems occasionally arise. However, today roughly nine of every ten workers and retirees participate in the program.^[10] The largest single group of workers currently not covered are federal government employees, who have their own pension programs, but under current legislation, they will eventually be brought into social security. For employees of state and local governments, participation in the program is optional but widespread.

Eligibility for social security benefits has been based on a rigorous interpretation of the investments (effort) principle of distributive justice, to use Hochschild's terms.^[11] Benefits may be viewed as a reward for regular exertion in the paid labor force (or, minimally, for being a dependent of such a person). The scaling of benefits to wage histories fits Hochschild's criterion of results, although since the early 1970s the overall benefit structure has been redirected toward meeting basic needs.

This firm link to regular participation in the work force has facilitated SSA's argument that recipients have an earned right to their benefits, a point bolstered by the use of carefully selected terms: "contributions," rather than "taxes," for example. The most important rhetorical strategy in distinguishing these earned rights from other forms of social provision has been the concept of individual social security accounts. By calling the records of workers' payroll taxes "accounts," the creators of social security effectively evoked the image of bank accounts, sums of money that workers accumulate through their contributions and that are held in their names by the SSA.^[12] And while recipients of means-tested public assistance must exhaust their resources in order to receive benefits, social security recipients simply draw on their personal accounts of prior contributions regardless of current means or need. This concept of earned rights to public income-maintenance benefits rather nicely conforms to American attitudes about human dignity, self-reliance, and self-help, thus clearly distinguishing social security recipients from the "undeserving poor."

Because need is not a criterion for receiving benefits, the program cannot ensure that benefits are used for basic needs rather than for frivolous luxuries. But we have good reasons to assume that most benefits paid out through the mid-1970s were spent on

basic needs. For one, professionals and some other high-income workers were not covered by the initial act; physicians, for instance, were not added to the program until 1965. Additionally, because social security was designed to supplement individual initiative with respect to thrift and savings, the proportion of a retiree's former earnings that benefits provide, the replacement rate, is generally low—around 25 percent for high-income recipients, rising to 40 percent for medium incomes, and climbing to 55 percent or so for those with low incomes.^[13] Indeed, state means-tested benefits have frequently been more liberal than social security pensions. Even in the early 1980s the minimum annual benefit was only $1,464, and the average monthly pension for a couple was $640.^[14] Earned (but not unearned) income greater than a specified amount (currently $7,500) reduces—and once eliminated—benefits.

Since the mid-1970s the system has included increasing numbers of high-income citizens, and benefit levels have risen. But not until 1980 did the percentage of elderly households below the official poverty thresholds fall to levels similar to those among the population at large.^[15] To some degree the formula by which benefits are calculated offsets recent trends. While the dollar value of benefits is directly related to recipients' earnings histories, the replacement rate rises as income falls.^[16] In contrast, the peculiarities of the indexing system adopted in the early 1970s led for a time to gradually rising replacement rates for everyone.

From the standpoint of an individual the redistribution that social security entails can be described as life-cycle: a covered individual contributes to the system when he or she is working and draws benefits when covered social hazards interrupt or halt employment. In comparison to means-tested programs that characteristically redistribute from more affluent strata to the poor, social security thus cannot be viewed as an attack on traditional property rights or wealth. As we will see, even the moderately progressive replacement rates are offset to some degree by regressive taxation and differences in the contributor-recipient ratios of high- and low-income workers.

Given the increases in benefit levels, the rising proportion of affluent beneficiaries, and the concerns about the program's financial

solvency, it is no surprise that social security has in recent years attracted wider critical attention.

Negative Liberty

Involuntary participation was an issue in the legislative wrangling over the 1935 Social Security Act; some analysts and officials preferred a system of voluntary annuities. But since then, involuntary participation has spawned surprisingly little criticism, except from the libertarian position.^[17] This counterintuitive development may be an example of Americans' operational or pragmatic views overwhelming more Lockean abstract or ideological views. But social security has also represented an extraordinarily good investment. Recipients have to date received vastly more in benefits than they have contributed in taxes, and benefits have, particularly over the last fifteen years, offered protection against inflation that few private annuities could equal.^[18]

Social security taxes a fixed percentage of a given wage base—currently (1988–89) 7.51 percent on $45,000. Employers match the worker's contribution. This tax is minimally not progressive in that it takes an equal proportion of earnings from low- and medium-wage earners within the wage base, and it is regressive in that people who earn beyond the wage base pay a lower overall rate. SSA has argued over the years that the progressivity built into the benefit structure adequately compensates for this tax regressivity, but not all analysts are convinced on this point.^[19] One problem with SSA's interpretation is that low-income workers tend to spend a greater period of their lives under this tax structure than do high-income workers, who usually start working later in life. And low-income workers often die younger than do professionals, so the former spend a smaller portion of their lives under the progressive benefit structure.

Social security taxes clearly reduce discretionary income, and for many low-income workers social security taxes exceed their federal income taxes. But aside from the compulsory reduction of discretionary income, the program leaves recipients unconstrained. Because it is a national program, there are no residency restrictions; nor do workers forfeit any contribution credits when they change jobs—a penalty often levied by private retirement pensions.

One of the virtues of direct national government delivery of social programs without third-party intervention is that it obviates complaints about constraints on the liberty of providers. Whenever private physicians have to alter their customary and preferred practices in order to deal with Medicare or other public patients, they can complain that the state constrains their freedom. But social security does not divert providers from their normal tasks to deal with paperwork or transfer payments. Even in the disability program, where SSA relies on state-level vocational rehabilitation agencies, public restrictions on the liberty of private providers are inconsequential.

Economic Efficiency

Relations between social security and various aspects of economic efficiency are disputed.^[20] With respect to social security's effect on work incentives for working-aged adults, some economists see a substitution effect: As taxes reduce the return from work, individuals increasingly prefer leisure. But other economists see an income effect: Higher tax burdens increase the need to work in order to take home enough to meet fixed or rising household expenses.^[21] Under the latter assumption, however, off-the-books moonlighting may seem preferable to labor that is taxed—a phenomenon that undermines social security's revenue base.^[22]

Social security entails some work incentives, but their importance is disputed. The central incentive is that a steady work history will be rewarded by an economically secure retirement.^[23] However, a worker armed with "economic man" preferences and good information would not necessarily produce steadily over a working life of thirty or forty years on the basis of this incentive alone, since social security does not require more than ten years of effort for any category of benefit. Clearly, other incentives—eating in the interim—come into play, and the middle-class character of social security recipients probably reinforces these incentives. By and large social security beneficiaries believe, in part on the basis of their life experiences, that effort returns a variety of encouraging results. Also, in contrast to private or union pensions, social security facilitates labor mobility, easing for instance, the process of mid-career workers shifting from sunset to sunrise industries.

For recipients, the program provides modest work incentives. For example, the earnings test that defines retirement has risen steadily over the years, and the current $7,500 disregard provides an incentive for productive labor. But overall social security appears to act as a work disincentive for those eligible for retirement or disability.^[24] This disincentive is probably stronger for those whose jobs offer few if any intrinsic rewards and for those whose jobs entail hard physical labor. Whether these disincentives are bad is a separate question. During the Great Depression a program that enticed older workers into retirement and made room for unemployed younger workers in the labor force was desirable. As the worker-beneficiary ratio continues to decline, however, these same disincentives may be undesirable. To contain costs, Congress is likely to raise the age at which people become eligible for benefits. But such a move could threaten program work incentives if people, especially low-income workers, cannot reasonably expect to collect benefits for more than a few years.

Some aspects of the survivors and particularly the disability portions of social security have been criticized as sufficiently generous to discourage recipients from working, especially if their options for work are dreary, dirty, or dangerous jobs that offer no rewards other than money. In seeking to protect working-aged adults from episodic social hazards, this argument runs, social security may create "unintended rewards" for experiencing such hazards.

The question of how social security influences the risk-taking decisions of corporate managers or entrepreneurs involves subtle judgments about the relative strength of other factors—such as social security's consequences for work incentives and saving—that are themselves hotly disputed. One smaller issue is, however, clear: social security does increase the cost of doing business. The costs related to handling the paperwork afflict all employers roughly equally, but economists disagree on the effects of the payroll tax. To the degree that this tax—particularly the employers' portion—is considered an added expense necessitated by the program, social security does raise the cost of doing business and thus discourage some risk taking. But to the degree that the payroll tax is considered deferred wages that would go directly to the employee in the program's absence, this deleterious effect disappears. This latter interpretation seems less applicable on the fringes of the labor

market, among small-scale employers or relatively unskilled workers, but even in the mainstream economy social insurance costs may contribute to incentives to move jobs offshore, particularly to third-world countries.

Because payroll taxes contribute to the costs of production, one might also argue that social security hampers the international competitiveness of American goods and services. But cost is by no means the sole determinant of competitiveness. BMWs are not clogging American highways because they are less expensive than Mercurys or Buicks. More importantly, the people who manufacture BMWs pay higher social insurance costs than do American producers. Thus it does not seem appropriate to lay any significant portion of the blame for America's competitiveness problems on social security taxes.

Although the United States does not spend as large a percentage of GNP on public social provision as do most other advanced industrial societies, against the consciousness of American political elites and the population generally, social security commands impressive resources.^[25] And there is the temptation to think that somewhere in the $202 billion that SSA expended in 1987 there exists waste. In response, SSA has repeatedly boasted that administrative operations consume only 1 to 2 percent of revenues, a low rate by comparison with most other social programs or private insurance operations. SSA has also reminded the public that, since no general revenues are used, the costs of the system are borne entirely by the beneficiaries. While this is in a sense true, the pay-as-we-go funding means that current workers are paying for current beneficiaries, and in turn they hope to be supported by the next generation of workers.

During the program's early years, these intergenerational transfers did not attract great attention or create funding difficulties. The program was largely perceived as facilitating a life-cycle redistribution, with old-age benefits bearing a direct and sensible relationship to a recipient's prior contributions. But once benefits substantially exceed contributions—even when adjusted for interest, inflation, and greater productivity with higher standards of living—systematic disparities develop and the intergenerational redistribution cannot easily be recouped.

American social security, like the social insurance systems of

many other advanced societies, currently faces this difficulty. To the degree that Americans take their individual claims to benefits on an "I paid for my benefits!" criterion as opposed to a more general good-faith effort approach, their claims are dubious.^[26]

In order to reduce these intergenerational transfers to reasonable proportions, several changes regarding program taxes and benefits have already been legislated for introduction later. Further changes seem likely. Overall, the program will probably continue to shift further away from a defined-benefit approach and toward a defined-contribution approach.^[27]

Finally, on the criterion of target efficiency, the degree to which benefits are focused on the neediest, social security by design gets generally low marks. Unlike a public assistance program intended to help the impoverished, a system of social insurance serves a socioeconomic cross-section of the public and aims to keep those who confront various episodic hazards from falling into poverty. Thus target efficiency is not among the criteria relevant to the evaluation of social security.

Administrative Complexity

Social security is a federal program that for the most part does not rely on state or local officials. In centralized Baltimore offices, national officials rather quietly carry out an array of complex operations: actuarial calculations, rule development and interpretation, and record keeping. SSA's fairly clear, if not exceptionally simple, instructions do not allow for much decentralized discretion in its field offices. Thus while social security is not a simpler program than the notoriously complex AFDC, its complexities are centralized and hidden, and recipients and the public generally perceive its operations as relatively straightforward. Senior program administrators have also been efficient and effective in choreographing adroit political maneuvers to ensure the program's popular success.^[28] Until the early 1970s SSA executives confidently knew what needed to be done to fulfill the program's objectives, had the requisite expertise to plan the necessary tasks, and had the autonomy to execute and monitor their plans.

Since then, program executives have seen the size and complexity of their tasks grow while their autonomy has been eroded by

the executive and legislative branches.^[29] Given that social security now touches the vast majority of American households, it is not surprising that Congress and the White House have begun to take more interest in the program. In the face of increasing outside intervention in program decisions, even this gifted and devoted program bureaucracy can no longer assure a high degree of administrative efficiency.

Executive and congressional interventions designed to counter funding shortfalls have also focused attention on several distinct limits issues. The first, sometimes referred to as the adequacy-equity dilemma, concerns in part the adequacy of benefits for workers in the lower socioeconomic strata. By design, social security does not single out pockets of exceptional disadvantage and focus its resources on them. Indeed, the program's contributory prerequisites exclude the most disadvantaged, and retirees with histories of low earnings qualify for only modest pensions. At the program's founding, problems of social disadvantage were believed to lay in the bailiwick of short-term public assistance. Public-employment efforts begun during the Great Depression and worldwide economic recovery were expected to ease unemployment, and the addition of survivors benefits to social security would, it was hoped, eventually eliminate the need for public assistance. This view, however, overestimated the national commitment to full employment and overlooked social discrimination, structural unemployment, the inadequacy of the wage structure for household support, and the resulting underclass. Today it seems both necessary and reasonable to ask social insurance to do more than social security now does for the disadvantaged. Early SSA administrators were probably correct that "a program for the poor is a poor program."^[30] Precisely because this is so, the disadvantaged must be incorporated into broadly inclusive insurance programs to the maximum degree that the nature of their resource inadequacy problems allows.

The equity aspect of the dilemma emerges at the other end of the socioeconomic spectrum in the practice of granting relatively high benefits to persons with substantial unearned income. Social security currently serves some households that cannot reasonably be seen as experiencing a social hazard. But halting earned benefits in instances in which aging poses no financial threat carries the risk

of diluting the firm association social security currently has for Americans with principles of distributive justice of which they generally approve—effort and results.

An independent limits concern stems from the narrowness of the range of social hazards that social security covers. In addition to the deficiencies with respect to the three social hazards that are covered, social security offers no general security against catastrophic illness or against wage disruption due to childbirth and or child-rearing to name two basic gaps in protection.

Because social security focuses almost exclusively on people who are no longer expected to work due to age or disability, until recently the program had not prompted many of the harmonization concerns that frequently arise in discussions of social programs. One of the rationales used to sell social security to a hesitant Congress in the 1930s was the encouraging effect it was anticipated to have on aggregate demand, and there is little reason to suspect that these hopes have gone unfulfilled. Precise effects are difficult to estimate, but most social security benefits are spent rather than saved, so they contribute, minimally, to maintaining demand and employment levels. That these benefits therefore also contribute to inflation did not until recently attract a great deal of attention from economists, with the exception of caveats against the double indexing of the 1972 benefit increases (corrected in 1977). Today, however, the overall size of the program, particularly in conjunction with recent federal budget deficits, casts an increasingly long shadow over federal fiscal policy. In the words of Peter G. Peterson, a former secretary of commerce, "the prospects for social security and for general prosperity are now inseparable."^[31]

As noted earlier, social security has had an odd relationship with American concerns for limited and local government. The Roosevelt administration's initial plan for social security was assaulted by Townsendites, among others, as too little, too late for the elderly who faced immediate resource inadequacy problems. At the same time, the Chamber of Commerce and the American Medical Association (AMA) decried the compulsory program as an attack on individual freedom and discretionary income.

In anticipation of these latter concerns, the program's formulators made every effort to distinguish social security from the limited and rather disappointing record of previous American efforts

at public social provision. So, for example, the SSB portion of the 1935 Social Security Act did not use the term social insurance . Two separate bills, one authorizing the collection of payroll taxes by the Treasury and the other authorizing the appropriation of these revenues by SSB, were based, respectively, on Congress' power to tax and appropriate funds.^[32]

Surprisingly, amidst all this hostility, social security prospered. The Great Depression no doubt facilitated the acceptance of the three-rights conception of democracy and the contemporary conception of human dignity, but most appealing was the program's firm basis in the effort-oriented principle of distributive justice. As social security grew to benefit a larger group of people more thoroughly, it increasingly fulfilled its theoretical underpinnings, which in turn themselves became more popular despite American preferences for limited and local government.

Conflicts of Interest and Compliance

Since social security's inception, SSB/SSA program executives have been the public elites most actively involved with social security policymaking. If we examine their initiatives over time, it is hard to argue that they have not been concerned with the objective of helping vulnerable citizens.^[33] The disparities between SSA preferences and what we might regard as recipient interests reflect in part matters of timing and emphasis. The cautious, conservative approach adopted by SSB in the early days can be justified by the need to carve out a secure niche for the immature system in a hostile environment. Nonetheless, as Jerry Cates convincingly argues, several preeminent SSB executives were also swayed by their pre—New Deal experiences with the Wisconsin school of social policy.^[34] Their idea about the proper way to handle social hazards, which Cates terms "conservative social insurance," led them to limit the initial extent of social security benefits and to later reject various liberalizing alternatives. All in all, program executives were more concerned with safeguarding the philosophy and form of their new program than with addressing the needs of the nation's depressionera elderly.

From the mid-1950s on, as social security officials became less fearful of their opponents, the program's conservative orientation

became less obvious, and an effective incremental approach to program development was adopted.^[35] But program executives still insisted on extensive and limiting eligibility prerequisites, the supplemental and wage-based character of social security benefits, and financing through regressive taxation. Since program executives were far more involved in initiating policy than were others in the legislative or executive branches,^[36] their conceptions prevailed. In the 1950s, however, increases in social security benefits were a relatively common accoutrement of election years, and by the late 1960s political candidates enthusiastically associated themselves with program development. And while Congress exhibited considerable reluctance with respect to initiating disability benefits, once the disability program was in place, members of Congress tended to side with disgruntled disability applicants against SSA. To many beneficiaries' advantage, SSA executives are no longer the independent arbiters of their interests.

Employers represent another affected constituency, with the program dependent on employers' cooperation in completing the paperwork and paying their share of the payroll tax. Collectively, individual small-scale employers can limit the program's reach at the fringes of the labor market. Employers of domestic workers and restaurant help have at times circumvented social security by imposing exchanges of desperation on their employees. Other than that, private third-party providers are involved only marginally (disability determination) in the non-Medicare operations of social security.

In the area of compliance with program objectives, social security offers various incentives. For the working-age population, the program reinforces socially approved behavior, namely working regularly and earning as much as possible in covered employment.^[37] The program's capacity for relying on positive reinforcement is bolstered by two exceptional features of retirement as a social hazard. First, it happens to most people and, second, it happens at a fairly predictable time, late in life. A social program aimed at the elderly thus avoids the disincentive problems that plague such programs as unemployment insurance.

Up to now, participant noncompliance with the retirement aspects of social security has come only from small, fairly unrepresentative groups. For instance, some workers would like to opt out

of social security, arguing that they could do better financially with a private alternative. Other workers, particularly those in low-income jobs, are troubled by the tax bite and seek work in the unrecorded labor market.

Recipient noncompliance with social security's disability provisions is more problematic. Although the disability program remains a modest portion of social security ($23 billion as opposed to $171 billion in 1986),^[38] the number of recipients has grown much more rapidly than was anticipated. It is hard to ignore the possibility that people working in low-paying, dispiriting jobs may abuse the program, and the potential for such abuse cannot be eliminated entirely. During the lengthy struggle to add disability coverage to social security, Roswell B. Perkins, an assistant secretary of HEW, argued forcefully that a safeguard would be provided by having disability determined by state vocational rehabilitation agencies.^[39] But the recent growth of disability claims suggests that a revised approach may be needed.

Through it all, SSB and SSA program executives have perceived themselves to be acting in the national interest, and their actions represent a relatively rare example in which statist interpretations fit American politics.^[40] In the last fifteen years the executive branch has increasingly sought to slow the proportion of national resources consumed by social security. This effort will necessarily continue. But antagonistic third-parties such as the Chamber of Commerce and the AMA are unlikely to overturn the program. Even if a serious intergenerational struggle develops over social security in the early twenty-first century, the program might be altered but hardly scrapped.

Summary and Implications

Despite recent controversies, against the criteria applied here social security is a strong program that enjoys an exceptionally encouraging set of attributes from the perspective of the American political culture. Some of social security's strengths, however, cannot be transferred to programs directed at other social hazards. As a public pension program for the elderly, social security intrinsically enjoys greater popular support than any other category of public social provision.^[41] The program's inclusiveness is clearly one

attraction: though not everyone lives to be elderly, most people probably expect and hope to do so, whereas far fewer people expect or minimally hope to experience serious illness, unemployment, disability, or divorce. Additionally, by and large it is not difficult to make the case that elderly recipients have made some contribution to society. Further, the social hazard of retirement is clearly defined by age, although specific eligibility rules (age 62, 65, or 67) may be controversial. Also, the program has a relatively stable beneficiary population, a characteristic that facilitates efficient administration. Recipients die, but rarely do they move off and then back on the rolls.

In contrast, those strengths of social security that arise either from the inherent characteristics of social insurance or from certain politically acceptable design features could be extended to other social hazards. Among these strengths are the focus on distributive justice, the prerequisite of sustained effort, the indexing of benefits to prior wages, and the supplementary character of benefits. The theme of personal accounts, which gives workers a sense of having a personal stake in the program, is preferable to the unemployment insurance system's reliance on employer accounts. Additionally, social security, apart from the taxation it entails, poses few constrains on negative liberty or property. And since social security taxes are regressive, the wealthy cannot on the whole claim that they are paying for benefits accorded to others, even though the benefit formula is progressive. The economic features of the program include some work incentives and some support for aggregate demand and employment. While the cost of the program is high, this cost is in a sense borne by the beneficiaries, although potential intergenerational inequities exist. Finally, because the program targets a fairly representative cross-section of the population, recipients are not stigmatized.

This degree of fit between social security practice and the values of distributive justice, negative liberty, and economic efficiency is insufficient for strict libertarians and free-marketeers. But as social programs go, social security involves less conflict with these values than do some others.

The practical virtues of social security may also provide useful models for other programs. SSA executives have accrued an enviable record of administrative efficiency, and their strong com-

mitment to their vision of what the program should become has contributed importantly to improving American attitudes toward public social provision.

Additionally, social security relies heavily on a structure of incentives that allows for the positive reinforcement of interests people have for acting in productive, responsible ways. It does so by rewarding work and thrift.^[42] It enables people to meet some of life's most disturbing hazards in a dignified socially approved way.

Juxtaposed to these strengths are some weaknesses. Social security creates a pattern of distribution that deviates from what the market would produce. The intergenerational transfers occurring now are hardly modest, and whenever a larger generation is followed by a smaller one, the systematic granting of benefits that vastly exceed contributions are apt to produce serious funding difficulties and intergenerational resentment.^[43] Imbalances between benefits and contributions also become more troubling as a growing proportion of benefits flow to households for whom retirement does not produce much economic vulnerability or constitute a social hazard.

Another potentially disruptive facet of social security derives from its constraints on individualism and negative liberty. Thus far, however, the program's compulsory character has not been keenly contested. This is a surprise, given the position of negative liberty in American political culture. One explanation may be that social security has represented a good deal financially. And perhaps various features of the program's design offset the provocation that compulsory participation poses for individualism and negative liberty.

With respect to economic efficiency, social security clearly reduces the work incentives of people over sixty-five. In 1935 this disincentive was deemed desirable; today many argue the opposite. As a program financed through payroll deductions, social security contributes to production costs. So in both cost-push and demand-pull senses, the program contributes to inflation. As we have seen, however, we cannot attribute America's lack of competitiveness to social security.

In theory social security may be faulted for violating American preferences for limited and local government, but in practice this clash has not affected the program's popularity. Again, this coun-

terintuitive result most likely reflects the program's design: complex operations are centralized in low-profile, efficiently administered bureaucracies.

Despite social security's track record as a good deal, until recently the retirement benefits have been of questionable adequacy. The introduction of Medicare in 1965 and the cost-indexing of benefits have provided some remedies. But social security's targeting of both social hazards and population remains limited, and the program thus presents some horizontal or support equity problems.

Finally, it is reasonable to expect that current obligations to future beneficiaries in conjunction with the declining ratio of workers to beneficiaries portends increased conflict over social security. When the interests of elected officials, SSA administrators, workers, and beneficiaries differ, the burden is all too likely to be borne by the most vulnerable.

That the preceding list of weaknesses and potential problems has not seriously undermined social security's political popularity and programmatic success suggests that some of social security's strengths eclipse its weaknesses—such as its compulsory character, its size, and its central administration. Part of social security's aura derives from its focus on retirement and is not transferable to other social programs. But social security's three quintessential design features can be applied to other forms of social provision: (1) eligibility rests on doing what society deems appropriate for able-bodied adults—working regularly and building a private contribution to financial security; (2) efforts exerted in the labor force create a clear index by which rights to social insurance benefits are earned; and (3) the program's specific features and language evoke symbols richly laden with traditional American values. Such design features have been crucial to the success and political viability of social security, and they can be adapted and extended to protect Americans against some social hazards now under the aegis of public assistance, as my proposals in chapter 8 will show.

Six—
Aid to Families with Dependent Children

AFDC is a public assistance program designed to mitigate some consequences of destitution for a category of the poor. While AFDC is managed pursuant to national guidelines, states have considerable leeway with respect to specifics. This case study emphasizes the common features of program operation.

Historical Background

In the Anglo-American tradition, children in households that cannot support them have long been considered among the "deserving poor" who are thought to merit public support.^[1] In the United States the institutions for such support have changed over time. In the nineteenth century widows unable to support their children frequently had to give them up. In the twentieth century, at the prodding of reform groups, several states enacted guidelines for "widows' pensions." Local options were allowed, and the discretion of local administering officials was generally broad. This precedent of state and local control exerted a fateful influence on the organization of the public assistance and unemployment insurance programs established by the Social Security Act of 1935. Both the officials (many of whom were from Wisconsin) who dominated the executive initiatives and numerous state congressional delegations were anxious to shield peculiarities of existing state programs from federal intervention.^[2] So, unlike social security, both public assis-

tance and unemployment insurance were constituted as state programs under loose federal guidelines.^[3]

From 1935 through 1962 Aid to Dependent Children (ADC), as the program was then known, was managed by the Bureau of Public Assistance under the direction of SSB/SSA. The national government set standards—generally procedural rules—and provided a portion of program funds. In essence ADC made transfer payments, with the titular beneficiaries being children in households whose adults were unable to support them. By and large these households were headed by unmarried, divorced, or widowed women who had few, if any, vocational skills and little experience in the labor market.

SSB officials prefered social insurance to public assistance. They saw the latter as an exceptional response to the Great Depression, and they thought that the need for ADC would gradually wither away. But to officials' consternation, the use of ADC increased in the late 1940s and accelerated in the 1950s. Between 1950 and 1960 the number of children in the program increased from slightly less than 1 million to over 2.5 million; during the 1960s repeated efforts to slow program expansion were met by a nearly identical increase in the number of families served. Although the relative importance of various factors contributing to this growth was disputed, much of the growth during the sixties is attributable to an increase in the proportion of eligible families making use of the program.^[4]

SSA officials were severely pressed to account for these trends.^[5] The Depression was long over and yet use of ADC was higher than ever. Based on the experience of ADC's predecessors, particularly state-initiated widows' pensions, officials had at first assumed that ADC would serve primarily a population of children whose fathers had died. When survivors' benefits were added to social security in 1939, it seemed reasonable to project that ADC would wither away, given the economy's long-term capacity to provide new jobs and the continuation of public employment efforts. The withering thesis reflects as well the limited scope of earlier local and state-level relief efforts that the federal government was now beginning to coordinate and help fund. These local programs had made relatively little effort to plumb the depths of social disadvantage. Widows' pensions, for example, went almost exclusively to white wid-

ows, and even among this group a variety of explicit and tacit restrictions limited claims. SSA's view also reflected naivete with respect to social disadvantage. SSA executives understood episodic social hazards such as aging, disability, illness, and even unemployment, but they were far less aware of the size and problems of the nation's socially disadvantaged underclass. They did not contrue "the poor" as a focus for public policy; their primary concern was keeping the elderly and others with records of regular self-support from joining the poor.

But by the mid-1950s it was clear that divorce and desertion, rather than widowhood, were accounting for the growing numbers of women that ADC served. Some observers argued that ADC in effect encouraged desertion by restricting eligibility to households with no able-bodied adult male—even as the national economy hindered males with relatively modest job skills (thus a high proportion of minority males) from finding steady employment and wages sufficient to support a family. While most ADC recipients were (and still are today) white, the proportion of minority recipients was increasing, and in the large cities they were becoming a highly visible and concentrated group. Through the early 1960s ADC appears to have served to a considerable degree as a substitute for unemployment insurance for these families: as nonwhite male unemployment increased or decreased, so did ADC cases.^[6]

The problems of family instability among households on the economic fringes of society were far more complex and intractable than those the program's initiators had foreseen. In 1962 the program was renamed and transferred to a new bureau, roughly the coequal of SSA, within the Department of Health, Education, and Welfare. The new AFDC included a variety of social services—family planning, basic adult education, child-care, vocational rehabilitation, and employment assistance—designed to increase the likelihood of employment among adults in recipient households.^[7] After 1962 a series of new programs—AFDC-UP for unemployed fathers, Community Work and Training (CWT), Work Experience Training (WET), and Work Incentive (WIN)—combined AFDC benefits and job training with the ultimate objective of enabling families to become self-supporting. As we will see later in this chapter, these new efforts all foundered on a combination of inadequate labor market opportunities, design features that created weak work incentives, and problems particular to program recipi-

ents. Despite these failures, reformers clung to an implicit version of the withering thesis. They persevered in the objective of reducing the AFDC recipient population.

Both the population designated as impoverished by federal guidelines and the population that uses AFDC have fairly high rates of turnover.^[8] Thus while poverty and AFDC usage are transitory conditions for many, the pools of persons experiencing poverty and AFDC usage over the span of a decade are quite large, about a quarter of the American population in each case.^[9] Furthermore, AFDC serves recipients with two distinctive patterns of use. Most recipient households use the program episodically to cope with a given social hazard—the loss of a job or separation of parents—whose resolution generally leads to leaving the program.^[10] For the most part these households have modest resources in the best of times, and their need for public assistance can be prompted by relatively modest hazards. In addition, parents in these households are frequently among the nearly one-third of the American work force to whom unemployment benefits are not extended.

Although public perceptions of widespread persistent use plague AFDC, less than 3 percent of the U.S. population, disproportionately urban and black, persistently use AFDC, and persistent dependence (relying on AFDC for over half of family income) is less than one-half of this level.^[11] Over the course of a decade persistent users constitute about one-sixth of AFDC recipients; but at any given point in time persistent users typically constitute a bit over half of the program's beneficiaries, and some of these households depend on public support for lengthy periods of time.

Implicitly recognizing the inadequacy of the withering thesis, the Nixon administration sent a version of a guaranteed-income plan to Congress in 1969.^[12] This proposal, known as the Family Assistance Plan (FAP), would have reduced the national government's payments to AFDC recipients in high-benefit states (California, Massachusetts, and New York, for instance) and would have provided help to the working poor who did not qualify for cash-transfer programs.^[13] Revisions of this plan were eventually defeated in Congress.

President Carter came into office on a campaign pledge of welfare reform, but his administration proved similarly unable to elicit congressional enthusiasm and support. Carter's proposal, Program for Better Jobs and Income (PBJI), resembled FAP in

many respects, although it was more ambitious with regard to employment and coverage (childless couples and single adults were included).^[14]

The Reagan administration has been concerned largely with reducing the national government's financial support for AFDC.^[15] It has also encouraged states to experiment with various workfare options. From the recipients' perspective, the single most important change has probably been the four-month limitation on allowing beneficiaries to keep the first $30 of earnings and one-third of the remainder. After four months—once limited work-related (including child-care) expenses are deducted—most earnings are offset by reductions in benefits.

In sum, little improvement has been made in the policy or design of public assistance in the last two decades. By and large recipients, program personnel, political leaders, organized interests, and the general public have little good to say about AFDC, but as yet there has been no consensus about how to reform the program. The 1988 legislation that emerged from conflicting ideas in Congress unfortunately deviates only slightly and unimportantly from the practices of the recent past.

Distributive Justice

AFDC has continually run afoul of American conceptions of distributive justice, in part because benefits are based on need and not explicitly linked to responsible, constructive behavior by recipient households and in part because benefits to the target group—needy children—are supplemented by "caretaker" grants to their parents. While the public generally supports the policy of providing for needy children,^[16] public attitudes are skeptical at best toward the parents of these children. Most of these parents are not employed full time, and the attention they devote to childrearing no longer elicits the public approval it did a half century ago when predecessor widows' pension programs served predominantly white women. Today, adults who use AFDC are often censured for their inability to provide for their families; frequently they are the subject of racial slurs and accusations of sexual promiscuity as well.

As a public assistance program, AFDC involves the vertical redistribution of resources from better to less well off. At current funding levels (approximately $15 billion a year), the program is a

modest attack on taxpayers' property that regularly serves roughly 11 million persons (nearly 5 percent of the population). But while AFDC benefits are not linked to prior effort or current results, neither does the program systematically apply the criterion of need to all Americans. Rather, AFDC responds to the needs of one category of people—poor children primarily in single-parent households—and ignores other equally needy people. The remediation of this horizontal inequity was one aspect of both the Nixon and Carter proposals for guaranteed annual income plans.

The public's response, at both elite and mass levels, to AFDC stresses two distinct themes.^[17] Americans believe that legitimate needs should be met by public social provision, but there is a widespread feeling that some proportion of the adults who draw on AFDC do not really need to do so, but lazily choose to do so. Thus AFDC benefits are perceived as inconsistent with human dignity, as indicative of dependence, and as destructive of work incentives and other aspects of economic efficiency.

Questions of distributive justice became particularly vexing in the late 1960s and early 1970s. Prior to 1967, AFDC benefits were, with only occasional erroneous or fraudulent exceptions, so low as to extend to only food, shelter, and clothing. But a series of changes in 1967 allowed recipients to keep (in addition to allowances for work expenses) the first $30 of monthly earnings and one-third of the remainder. In conjunction with relaxed eligibility for Medicaid and food stamps, these changes enabled a small proportion of AFDC recipients, generally residents of urban areas in industrial states, to combine earnings and public program benefits into household incomes bordering on the national median income.^[18] Such income levels threatened the otherwise close association between program benefits and basic resources. And as recipients in the higher-income brackets stayed on AFDC, the proportion of program resources going to the neediest families declined.^[19] The changes that permitted these aberrations have largely been rescinded by the Reagan administration.

Negative Liberty

From the libertarian perspective, participation in AFDC as a beneficiary is voluntary. Public assistance, in contrast to social insurance, characteristically involves no compulsion in the form of

government pressure. The destitute are free to ignore public aid, although the drive to survive may compel many disadvantaged and vulnerable citizens to make use of the program. But this is, from the libertarian perspective, a choice freely made.

Taxpayers involuntarily participate in the program by funding it. Federal tax revenues cover over half of the national cost, with states covering the rest. The amount of taxpayers' discretionary income redirected to AFDC is far less than that allocated to social security, but AFDC transfers involve a higher degree of vertical redistribution.

Historically, AFDC has been associated with a number of constraints on recipient liberty. The legal basis for some of the more notorious of these practices was withdrawn in the late 1960s when the federal courts invalidated residency and "man-in-the-house" rules. Considerable infringement is still possible, although its actual imposition varies widely with region, time, and caseworker.^[20] Most states have also dropped the elaborate intrusive investigations of recipient need, and some use a self-report system similar to that used by the Internal Revenue Service for taxes.

At the level of the field office the extraordinary complexity of AFDC holds conflicting implications for the freedom of program providers. On one hand, case workers are free to exercise individual discretion within a set of complex rules that cover a variety of matters, such as the determination of work-related expenses. In practice, heavy caseloads and high turnover lead caseworkers to substitute intuition and discretion for laborious bureaucratic procedures.^[21] On the other hand, the diversity of complicated social services provided by AFDC requires service providers, such as the Department of Labor, to conform to overall program guidelines and desiderâtâ.^[22]

Economic Efficiency

AFDC executives and welfare reformers have continually struggled over the issue of work incentives for recipients. One overriding work incentive, stronger among some groups than others, is the stigma associated with AFDC status. The complexities of the application process and subsequent eligibility testing also provide some deterrent to program use and thus may serve as a work in-

centive. As a group, however, AFDC mothers tend to have low levels of education, job skills, and work experience. So even those who are eager to work, as most initially are,^[23] have difficulty finding jobs that provide either much job security or income sufficient to support their families. For many, earnings from even full-time employment fall below official poverty levels or, in some cases, below the value of AFDC grants in conjunction with the benefits of other programs—food stamps, Medicaid, and housing—for which AFDC recipients are generally eligible.

Despite these obstacles to household support, adult AFDC recipients generally do work, frequently part-time, using the program episodically when work runs out. Recipients thus do not constitute a perverted culture whose members are unwilling to contribute to the support of their households. Instead they generally do contribute, but their economic position is fragile and highly contingent on periodic changes in family composition and employment status.^[24]

This fragile situation is complicated by AFDC's design features. Since 1962 AFDC has experimented with different approaches intended to encourage recipient self-help. But in practice these measures have penalized recipients who are too diligent. For example, until 1967 recipients' earnings, less work-related expenses, were deducted from their program benefits, a practice that "taxed" earnings at about 70 percent. In 1967, in connection with the WIN program, Congress effectively reduced this rate to about 50 percent.^[25] Then in 1981 the Reagan administration limited the more generous treatment of earned income to the first four months of a claim. Thus recipients eventually encountered a "notch," a point at which additional earnings cause a drop in total household income because they render the household ineligible for AFDC benefits and programs such as Medicaid and food stamps.^[26]

Although the term work incentives arises repeatedly in discussions about AFDC, it is fair to say that there is no agreement as to what work incentives or disincentives are associated with the program. Some critics of the program argue that the very distribution of resources to working-aged adults reduces their incentive to work. Other critics cite the high rate of benefit reductions against earnings and the aforementioned notches as the major disincentives.^[27] Proponents and detractors of AFDC also disagree on the inter-

pretation of the program's results. For instance, attrition rates in job training programs are high, but some people do achieve independence from AFDC through WIN training and job placement.^[28] Yet the attrition rate is commonly interpreted as an index of the failure of work incentives, rather than as a measure of the difficulty of the circumstances faced by recipients.

The various social services introduced in 1962—child-care, basic adult education, family planning, work training—are frequently viewed as work incentives. The delivery of these services, however, has been spotty, largely because funding has been much more limited than the initial program announcements suggested. Additionally, the objectives of services such as basic adult education have never been clearly defined and are sometimes delivered in ways that serve the needs of providers more than those of program recipients. For instance, since the federal government picks up a higher portion of costs associated with social services, state welfare agencies have renamed offices that scrutinize recipients' budgets as budget-counseling services.^[29]

Regarding incentives for household savings, AFDC's means test is clearly counterproductive. State regulations vary, but in general means tests restrict program eligibility to households that have virtually no income or saleable assets.^[30] Means tests thus mesh poorly with the episodic character of most AFDC usage: families are not eligible for assistance until they have depleted their resources—a small savings account, marketable possessions—a practice that only reinforces a sense of hopelessness with respect to savings. Means tests also make leaving AFDC protection risky, since the recipient will have to fall back into destitution in order to reenter the program.

Guaranteed-income experiments conducted in Seattle and Denver suggest that means tests which ignore assets in determining eligibility may be a more constructive approach to aiding the working poor and encouraging savings.^[31] The abolition of means testing would reduce target efficiency but would bring AFDC in line with social security and unemployment insurance, both of which use current earnings as a measure of eligibility. (No one would ever suggest that laid-off workers covered by unemployment insurance sell their homes, cars, and other possessions before applying for benefits.) Since AFDC serves as a substitute for unemployment insurance for

a high proportion of workers on the fringes of the economy, parity in the type of eligibility rules does not seem unreasonable.

For businesses and corporations AFDC's consequences for risk-taking are mixed. Because the program is financed from general revenues, rather than payroll taxes, the association between levels of program support and the costs of doing business are small and indirect. The program's deleterious effects on work incentives may hurt recipients but should have little influence on business. The argument that AFDC curtails participation in the labor force by offering a limited and unrepresentative segment of the labor market roughly comparable levels of financial support for rearing children may have considerable face validity for businessmen, but this contention is of dubious empirical accuracy. Nor can AFDC be implicated in debates about the competitiveness of American goods and services in global markets. The taxes that sustain AFDC are paid largely by individuals, not by business enterprises, and they represent less than 0.5 percent of GNP (compared to 5 percent for social security), or about 1.5 percent of the federal budget (compared to 20 percent for social security).

The total costs of AFDC are modest: less than $15 billion to serve about 11 million persons in 1986. Herein lies one of the program's virtues from the standpoint of economic efficiency. It is inexpensive—far less expensive than any program attempting to achieve self-support among AFDC's adult recipients would be.

Slightly more than half of AFDC's costs are covered by the federal government from general revenues. The redistribution of resources is thus strongly vertical: since the incomes of AFDC recipients are low, few of them pay federal income tax, and higher-earning citizens shoulder the bulk of AFDC's costs.

States fund the balance of AFDC expenditures through a variety of taxes that are usually more regressive than the federal income tax. The tax burden that AFDC poses for residents of different states varies sharply. Nearly one-half of the benefits are disbursed in three states (California, New York, and Massachusetts) that provide high benefits, based on high costs of living, to large numbers of recipients.^[32] But other states do not share these exceptional fiscal problems.

Overall AFDC is a highly focused program that concentrates on a particular swath of urgent need. Consequently, AFDC has a

reputation for target efficiency; that is, AFDC benefits—in contrast to those of social security—go to extremely needy households. Prior to the late 1960s varying interpretations of what constituted a "suitable home" and other restrictions combined to deny aid to many needy families within the ostensible target population.^[33] Between 1967 and 1981 target efficiency was diluted somewhat by more lenient rules covering earnings retention, but only a small proportion of AFDC households were able to take advantage of these rules.^[34]

Specific aspects of AFDC's highly focused design, critics claim, exacerbate the plight of destitute households and undermine economic efficiency. One argument of this sort is that by supporting only single-mother households, AFDC encourages family dissolution.^[35] Fathers who cannot or who are reluctant to take responsibility for the support of their dependents either leave the household or never become a formal part of it. Empirical evidence for this argument is sparse: one analysis shows that the variation in the level of AFDC benefits across states has no effect on family structure and living arrangements, but other data suggest that the incentive structure to which this feature of AFDC contributes is not overlooked by some of the males associated with AFDC families.^[36] Nonetheless, experiments with the negative income tax, particularly the Seattle/Denver work, suggest that programs that aid households regardless of composition may facilitate family dissolution as well, through an independence effect—that is, women who have independent resources are apt to leave a bad marriage.^[37]

A second troublesome feature is the wide variation of benefits among states and their subdivisions. Economic efficiency is not well served if poor households migrate from the rural south, where benefits are lowest, to northern and western cities. Any concentration of AFDC-dependent households in cities already facing impressive structural unemployment problems produces a permanent urban underclass. A program of standardized national benefits, in contrast, might encourage poor households to move to areas in which the cost of living is lower and job opportunities are more plentiful. Social security, for instance, has probably facilitated a migration of elderly citizens from the frost to the sunbelt, and the Denver guaranteed-income experiment suggests a similar migration pattern.^[38]

In summary, little about AFDC contributes to economic efficiency, but programs that target narrow, highly vulnerable segments of the population are, by nature, not intended to realize economic efficiency. Nevertheless, extending help to the impoverished and economic efficiency may not be as incompatible as is generally thought. As we will see, the economic efficiency of a program aimed at helping the poor can be enhanced, but such improvements exact a price in terms of program budget.

Administrative Complexity

For a national program, AFDC allows an astonishing degree of regional variation,^[39] with each state having its own rules for eligibility, subsequent household budgets, and a variety of other matters. Typically, these rules are extraordinarily complex, and their application requires continual interpretive decisions by state-level and field-office personnel. The density of substantive field-office decision making, fostered through decentralization, is further complicated by high turnover among recipients and their concurrent use of other public assistance programs.

AFDC's original objective was to supervise transfer payments, but the addition of social services components in 1962 presented more demanding and less well-defined program objectives. So-called hard services, those clearly related to employment such as job training and placement, have had relatively modest success, and providers in other government agencies have complained about the constraints that serving AFDC recipients places on them. The so-called softer services—basic adult education—have less clear-cut objectives and have sometimes become twisted to serve the needs of state financing rather than those of recipient merging.^[40]

Compounding the difficulties posed by complex procedures and ambitious objectives, AFDC operates in a highly decentralized fashion. Decentralization opens the door to differences in objectives among the national, state, and local levels as well as among the various agencies involved at any given level. Additionally, decentralization places a great deal of decision-making responsibility at the bottom of the organizational pyramid, where a variety of personnel problems and high rates of employee turnover hamper administrative efficiency. To reduce some of these complexities, the guaranteed-income proposals of the 1970s sought to standardize

many rules on a national basis and to return the program's focus from social services to income transfers.

AFDC's most prominent limits issues involve the types and degrees of need covered by the program. To take but one prominent example, fewer than half of the states chose to expand eligibility to households with an unemployed male by instituting AFDC-UP.^[41] State-by-state guidelines about what constitutes need reinforce the notion that AFDC is, despite its formal legal standing, a type of gratuity rather than a socioeconomic right.^[42] The English poor laws linger on in contemporary AFDC practice.

AFDC's harmonization with other relevant public policy objectives is sharply uneven. Public revenues allocated to AFDC are usually spent quickly, so the program minimally helps to maintain consumer demand and therefore employment. But since most AFDC recipients are not full-time members of the labor force, one could argue that the program removes potential workers from labor-market participation. Unlike social security, AFDC is not considered as an important contributor to inflation.

AFDC's central harmonization problem lies in the insulation of the program from the labor market—an inherent aspect of program design that is highly artificial in a society that places great emphasis on work. Efforts by the Nixon and Carter administrations to reform public assistance by replacing AFDC, rather than by tinkering with the specifics of the existing program, were well intended, but their guaranteed-income proposals were also seriously at odds with American political values about work.

With respect to the values of limited and local government, AFDC offers a mixed but not generally encouraging fit. The program is relatively small, but it looms large in the public's perception. Regional and, less legitimately, case discretion provide some measure of local governance, and there have been several interesting state-level experiments in recent years. But local discretion can just as easily become a tool for institutionalized racism and other disturbing practices.^[43]

Conflicts of Interest and Compliance

A program such as AFDC is, by nature, a locus for conflicting interests, the most central of which is that between public officials'

desires to restrict spending for "the poor" and the needs of the program's clientele. Throughout AFDC's history, this conflict has taken different forms. At the program's inception the administrative officials at SSB sought to limit public assistance relative to social insurance, and this preference persisted for a number of years. Up through the early 1960s program executives viewed AFDC as a transfer program concerned with maintaining a limited number of households beset by highly exceptional circumstances. This view was gradually superseded by the alternative conception of AFDC as serving a more progressive task. In 1962 increasing costs prompted the Kennedy administration, on the advice of social work professionals in HEW, to add a social services emphasis to AFDC. Representing the social work view about the path to self-reliance, these services were intended to reform recipients in ways consistent with increased self-sufficiency, but they also enhanced the role of social workers, offering them more interesting tasks than handling caseload paperwork.^[44]

Successive initiatives by political leaders through WIN in 1967 were driven most immediately, though not exclusively, by a concern with controlling costs. Since the early 1970s a concern with limiting program costs has been joined, and occasionally even eclipsed, by concerns relating to vertical or work equity and permissiveness. Overall, however, interest among political elites in AFDC has been extremely uneven. The relatively small federal expenditures for AFDC and the concentration of costs in a handful of states means that most members of Congress need not pay much attention to the program. Representatives from affected states have often viewed fiscal relief for state government as more of an issue than program reform, and few House members have significant AFDC constituency interests.^[45] Nor have AFDC program executives typically exhibited the devotion and adroitness of their social security counterparts.

At the state level, conflicts of interest between public officials and program clientele vary over time. Some states have been relatively generous in extending eligibility and higher benefits; others have adopted practices designed to deter or discriminate against applicants and restrict benefit levels. The primary federal sanction for achieving state-level adherence to those aspects of the program covered by federal guidelines—the withdrawing of federal funds—

has not proved effective. A few states have been so bold as to call the federal government's bluff in such matters.

The incentives for compliance and cooperation among beneficiaries, federal officials, and third-party providers are no less problematic. With respect to the majority of episodic users, AFDC does provide invaluable assistance for bridging social hazards, but the means test, which generally requires program applicants to be indigent before qualifying for services, is clearly counterproductive. Rather than stimulating self-help efforts, the means test creates strong cross-pressures between self-sufficiency and dependence among persons whose views about their economic futures are understandably uncertain. Once a household qualifies for the program, these conflicting pressures continue, as increased earnings are penalized by reduced benefits. Episodic users work their way toward renewed self-sufficiency in spite of, rather than because of, the program.

For the less numerous persistent users of AFDC, the bonds of habit and dependency have thus far proved intractable. Whether such bonds can be broken humanely—without disruption, suffering, and controversy—is a troubling question.

Three factors mitigate against persistent users' achieving independence from AFDC. First, these are people who have limited or undeveloped capacities with respect to the existing labor market; they are caring for young children and have little in the way of job skills and experience. Second, in addition to information and transportation problems that complicate the matching of recipients and jobs, the existing labor market does not offer enough paid positions capable of sustaining the families that require support. Third, against this structural inadequacy, AFDC offers an enticing form of economic security, one that in extreme cases invites recipient abuse. But while New York City, for example, has been troubled by systematic patterns of abuse, the vast majority of these abuses concern extremely small sums of money. (The popular myth of AFDC recipients driving Cadillacs is precisely a myth.) Nonetheless, cases of abuse do illustrate how the program's design features encourage certain kinds of counterproductive behavior from recipients.

As a program that deals with a narrow and unrepresentative group of vulnerable and generally disadvantaged citizens with

whom upper-middle-class bureaucrats do not readily empathize, AFDC is at a disadvantage compared to programs serving a broad cross-section of the population. Between 1935 and 1962, when AFDC was administered by SSB/SSA, the program executives who were so dedicated to social security exhibited far less enthusiasm for the public assistance program.^[46] Even after AFDC became an organizationally coequal division of HEW, program officials did not display the resourcefulness and commitment that characterized their counterparts at social security. The joint national-state character of the program surely impedes administrators' efforts, but the crusade to keep the elderly and the disabled from joining the poor has never been extended to those who are poor.

National elected officials have similarly evaded a strong commitment to AFDC and related public assistance issues. With the exception of Ford, every president since Roosevelt has taken some important initiative with respect to welfare reform, but only Johnson found these programs compelling enough to make persistent efforts on their behalf in the face of congressional opposition. In Congress, as noted earlier, interest in AFDC is sparse and more concerned with formulas for federal cost-sharing than program design or success.

From the perspective of the federal government, state and local officials are third-party providers of various sorts. Characteristically these providers have priorities that differ from those of AFDC's federal program executives. A few states offer aid that exceeds federal guidelines, but other states are reluctant to adhere to the spirit of the law, particularly in the implementation of the social service components. Two key problems have been the use of private contractors whose agendas do not match program objectives and the misappropriation of federal cost-sharing available for social services to finance routine administrative operations.

Summary and Implications

On the whole, AFDC fares much worse than social security on the criteria used in this case study. One must bear in mind, however, that AFDC, confronts tasks that are inherently tougher than the providing of pensions for the elderly, a particularly favored task among social insurance programs.

Despite a poor fit with American notions of distributive justice, AFDC distributes resources that are important, if not essential, to the survival of a shifting recipient population that currently numbers about 11 million people a year, most of them children. The program's effects in reducing starvation, homelessness, and other forms of human misery are laudable, especially given its low cost.

Additional strengths include the noncompulsory nature of participation, minimal effects on economic efficiency, relatively high target efficiency, and—for better or worse—a considerable degree of regional and local discretion.

The list of AFDC's weaknesses is far longer. Because the distribution of benefits is not systematically related to either need or effort and does not require effort-based certifying activities from beneficiaries, the consequent public skepticism about AFDC beneficiaries' true neediness undermines the program's goal of enhancing human dignity. The financing of AFDC poses relatively few constraints on citizens' discretionary income, but the income transfers are highly vertical, and additional constraints are posed on the liberty of recipients and providers. The program's consequences for work, savings, and risk-taking, while uncertain and probably modest, do not look promising.

AFDC also receives low ratings on administrative efficiency. Its artificial insulation of program recipients from the labor market creates notable limits and harmonization problems. The program has not appealed to traditions and symbols respected in American political culture and has not aroused the interest of political elites on matters other than containing costs.

Finally, the program does not well serve the needs of either episodic or persistent users. In particular, efforts to promote greater self-sufficiency among AFDC recipients have foundered on three central issues: program features that create only weak work incentives, inadequate labor market opportunities for recipients, and the personal disadvantages and limitations of recipients.

The reforms initiated in 1962 focused on improving beneficiaries' capacities for using the labor market through job training and placement, related social services, and simple coercion. It is possible that this approach failed because we simply did not try hard or long enough, but the difficulties experienced along this path suggest that we try other avenues.

Nonetheless, today one sees a resurgence of support for the failed approach of the 1962 reforms. In Beyond Entitlement , Lawrence Mead, for example, thoughtfully and provocatively argues that the fundamental fault with the American welfare system is not its size but its permissiveness. Programs of public social provision, according to Mead, should require more rigorous effort at self-development and help among beneficiaries than AFDC or public assistance in general currently do. While I agree with this premise, there are two serious flaws in Mead's proposal that we develop intraprogram policing capacities for assuring that beneficiaries are working, studying hard, treating fellow family members reasonably, and avoiding crime.

The first flaw is the impracticality of such a proposal. For one, programs with administrative teeth are hard to develop. Although California's new Greater Avenues for Independence (GAIN) and Massachusetts's Employment and Training Choices (ET) have reputations as more constructive than workfare generally, both programs ultimately rest on the individual recipient's willingness to cooperate.^[47] Furthermore, intraprogram policing takes us back along a route we have already traversed and decided, through a series of court decisions, we did not like. And American political values are not likely to be well served by a set of larger, seriously more intrusive state bureaucracies that monitor public assistance.

The second general flaw in Mead's approach is the particular character of its optimism, which seems in part a reincarnation of the withering thesis. Mead does not imagine that all poverty will wither away through coercively imposed work, but he is optimistic that a good deal of it will. Yet to suppose that a considerable measure of poverty rests on personal actions alone is to ignore characteristics of the labor market that limit its universal usefulness in providing a living wage for many workers. Low wages and modest skills combine to keep the working poor hovering barely above poverty, and the slightest misfortune is apt to plunge them into destitution. Many households recover through their own efforts, but the pool of households likely to experience episodic poverty is so large and the persistence of their collective problems so great that eliminating or even dramatically reducing such poverty through work alone seems unlikely. The overall problem is less one of changing the attitudes of a small subculture, although this may be

a factor among persistent AFDC users, than of overcoming a variety of structural problems in the labor market that preclude adults from being able to support their families.

Moreover, Mead's optimism seems to ignore the effects of certain demographic trends, including the birth rates for single mothers and the divorce and desertion rates among couples with young children. To take but one example, for poor women the single most frequent escape route from poverty is marriage or remarriage.^[48] But marriage and remarriage rates are notably low among single urban black women who have children. One explanation of this trend holds that the limited economic opportunities of urban black men contribute to their reluctance to take on the obligations of a wife and children or stepchildren; black women, seeing little economic relief in such marriages, might be equally reluctant.^[49] Another explanation looks to sex ratios. In those urban areas where women greatly outnumber men, men—and not just American black males—tend to be less responsible toward women and children; in areas where sex ratios are more balanced, stable two-parent families predominate.^[50]

In contrast to Mead's vision, the AFDC reforms initiated in 1967 (and dismantled in 1981) represent a nascent example of an approach that incorporates a more adequate recognition of the limitations of the existing labor market. These reforms extended to a category of the poor a minimal level of economic support (a guarantee) that could be enhanced by work. Expanding upon this approach, the FAP and PBJI proposals offered both advantages and drawbacks in comparison to AFDC. They covered more people, thus reducing horizontal or support inequities. But they also brought a much larger population into prospective programs that offered their recipients economic support insulated from requirements of working. And the language of guarantees, as developed from WIN through PBJI, contradicted the core American concerns of work and self-sacrifice as essential to human dignity.

During the last quarter-century, then, public assistance programs have tried either to alter the capacities of people ill suited for the existing labor market or to insulate these people from the rigors of that market. In the renewed debate over workfare conservatives have tended to argue that in order to reduce the public assistance rolls we need to put people to work, and liberals have

tended to argue that we therefore need to create jobs that will allow people to work. We can, I believe, carve out a position between these views. Most Americans would agree that able-bodied working-aged adults ought to participate in the paid labor force, rather than be insulated from it. For single adults and childless couples, self-support through the labor market is a reasonable expectation. But for households with children, particularly single-parent households, regular full-time work at or near the minimum wage will not cover basic needs or cushion a household from poverty. And in single-parent households regular full-time employment is extremely difficult in the absence of adequate and affordable child-care. For these households, then, it seems appropriate and socially productive to establish a program of public social provision that would both facilitate labor market participation and supplement wages inadequate to support a family.

In chapter 8 I will propose such a program, one that will create a structure of opportunities and incentives far stronger than any that has characterized AFDC to date. Rather than enlarge existing public assistance bureaucracies for the policing of recipients, my measures would facilitate the socially responsible activities we want to encourage and, as necessary, supplement low-income workers' constructive efforts at self-support. But before I lay out my proposal, we need to examine the special problems associated with the delivery of social services. For this purpose, let us turn to a case study of Medicare.

Seven—
Medicare

Delivering a service such as medical care is inherently more problematic than distributing income-maintenance checks. But American efforts to develop public medical-care programs have encountered exceptional difficulties, in large part because such initiatives have generally been perceived by physicians and other medical-care providers as infringing on their professional autonomy. Whereas public income maintenance fills a void that intrigues few if any private practitioners, public medical-care policy challenges increasingly powerful private interests. Thus ideological positions concerning the boundary between public and private spheres become pretexts for entrenched private economic interests.

As we shift our attention to indispensable services such as medical care, questions about how recipients are to earn their benefits no longer monopolize our attention because providers' attitudes become a much more prominent issue. Nonetheless, there are two important reasons for including service provision in our analysis of the investments approach. First, by examining a basic service such as medical care we learn something about the limitations of the investments approach with respect to social services in general. Second, we are reminded that severe resource inadequacy problems cannot be resolved by income maintenance alone. Access to medical care is an integral aspect of comprehensive protection against severe social hazards.

Historical Background

In comparison to other advanced industrial societies the United States remains a "laggard" with respect to public medical-care

programs.^[1] Although the nation's preventive public health measures—water and sewage treatment, garbage collection, and inoculation against selective infectious diseases—are generally sound, public programs involving therapeutic medicine are distinctly more circumscribed in the United States than in comparable industrial societies. Most noticeably, the United States has no public program to assure universal or minimally widespread and comprehensive medical care. In the twentieth century four separate and lengthy efforts to achieve some form of national health insurance in the United States have failed.

The first of these crusades, in the years preceding World War I, was prompted by the American Association for Labor Legislation (AALL) and other progressive groups.^[2] AALL leaders saw public health insurance as a natural extension of workers' compensation insurance. Having had some success at the state level on the latter, the AALL began state campaigns for public health insurance programs. The AALL's proposals focused primarily on protecting low-income workers against poverty arising from illness. Provisions included sick pay and some coverage for the costs of medical services. Given the relatively low medical costs at the time, sick pay was as important as the insurance provisions.

The AALL argued strongly for the macroeconomic benefits of its plan: health insurance would improve economic efficiency by reducing various industrial costs—worker absenteeism—associated with existing working conditions. As the Progressive candidate for the presidency in 1912, Theodore Roosevelt put a political twist on this proposition, maintaining that no nation could be strong if its population was highly vulnerable to illness.

But business interests rejected these claims of economic efficiency. The AALL plan was also opposed by organized labor, particularly Samuel Gompers and the American Federation of Labor, primarily because they saw it as preempting the unions from winning medical benefits for their workers and thus undercutting the advantages of union membership. For a time the American Medical Association (AMA) cooperated with AALL, but members increasingly perceived the proposal as threatening.^[3] Having thus failed to establish broad-based support, the AALL campaign dissipated during World War I. The war effort distracted attention from public health insurance, and German social programs—from

which the AALL proposal borrowed—became the subject of a barrage of intense and highly negative propaganda.

A strikingly different campaign for public health insurance surfaced in the late 1920s. By this time medical care had become much more expensive, and now reformers focused on public health insurance as a vehicle for helping the middle class afford medical care. The Great Depression, however, posed more urgent social issues, and public health insurance did not get much attention from the New Deal until 1938. Physicians and hospitals had not been spared from economic disasters, and some welcomed talk of public health insurance as a means of expanding the demand for medical care. But by the time the Roosevelt administration sent its proposals to Congress in late 1938, Congress had become more conservative and resistant to New Deal efforts. In 1939 a related proposal, sponsored by New York Senator Robert F. Wagner, was also rejected. And again international turmoil and a world war turned everyone's attention away from domestic issues.^[4]

The third wave of reform began in 1943 and eventually led to the creation of Medicare in 1965. Unlike its two predecessors, this initiative was increasingly dominated by the federal government and garnered the support of new interest groups, particularly organized labor. These proposals looked to the federal government, rather than state governments, and more nearly embodied the concepts of universal and comprehensive coverage. The Wagner-Murray-Dingell bills, introduced in 1943, were the first such proposals. In 1945 President Truman wrested initiative from the Congress with a broader and in some respects more ambitious plan that called for federal funding of hospital expansion, medical research, and medical education and the expansion of the largely preventive public health program for mothers and children—the Maternal and Child Health Program.^[5] The most controversial point was Truman's proposal for a national health insurance scheme that would cover not only social security recipients but citizens generally.

Certainly Truman's proposals had several features that were attractive to medical providers: federal subsidies to enhance the profession's capital—hospitals, research, and education—and to finance demand for medical services. None of Truman's proposals

would have affected existing patterns of medical-care delivery; the dominance and autonomy of physicians were to remain intact. Nonetheless, the AMA, the American Hospital Association, and the Chamber of Commerce, among other groups, opposed aspects of the plan with varying degrees of ferocity. The AMA led the battle against a government presence in medical care with an unprecedented lobbying and public relations effort. As the Cold War intensified, opponents of national health insurance tarred the proposal as a socialist ploy.

In the aftermath of this defeat, proponents of national health insurance—officials in the Social Security Administration and a few other executive bureaucracies, specific members of Congress, some presidents, organized labor, and an assortment of liberal funders and intellectuals—generally adopted an incremental approach to achieving a legislative victory.^[6] A similar tactic proved effective in bringing disabled workers into the social security system, but the government's entry into the provision of medical care was more troublesome than adding disabled citizens to the social security rolls.^[7]

By the mid-1960s, however, the notion of public medical insurance for the elderly—a group generally held to be worthy, medically vulnerable, and of limited material means—had garnered fairly broad-based support. In 1965 innovative compromises within the House Ways and Means Committee resulted in a two-part Medicare program. Under Part A, Hospital Insurance (HI) became a standard social security benefit. An increase in social security taxes was authorized to finance this program. And under Part B, Supplementary Medical Insurance (SMI) for physicians' bills could be purchased by citizens sixty-five and older. This portion was financed in part through monthly premiums paid by beneficiaries and in part through general tax revenues. Finally, the legislative compromise expanded federal support of health care for the poorest Americans (Medicaid).

To reduce opposition from organized medicine, the architects of Medicare chose not to incorporate any provisions affecting the delivery of medical services.^[8] Most importantly, no effort was made to alter the prevailing high-technology, therapeutic orientation with its practices of hospital bills based on costs and physicians'

fees based on services rendered. Thus, Medicare bolstered the demand for medical care without creating any cost-control mechanisms, and the program exceeded cost projections from the outset. In the face of Medicare's rapidly rising costs, Congress has been unwilling to expand the program. Benefits were extended to disabled social security beneficiaries and end-state renal dialysis patients in 1972, but Medicare has not developed into the program of national health insurance that some of its early proponents had hoped for.

Indeed, the fourth campaign for national health insurance, during the 1970s, suggests that if national health insurance is ever developed, it is unlikely to follow the model of Medicare.^[9] For this fourth campaign resurrected a prominent theme of the AALL's pre-World War I crusade: economic efficiency. And in light of the federal budget deficits of the 1980s, any future proposal, even remotely universal and comprehensive in scope, would have to offer greater economic efficiency than Medicare.

From the standpoint of meeting the medical-care needs of the elderly, Medicare represents a distinct improvement over the conditions that prevailed prior to its enactment. However, Medicare's accommodation of existing practices in the medical-care professions limits both its effectiveness in assuring appropriate, efficient medical care and its value as a model for any expanded program of national health insurance.

Distributive Justice

In part Medicare represents an attempt to use public means to redistribute medical care in accordance with the criterion of effort rather than by the market criterion of ability to pay. HI benefits go largely to elderly people who have earned them by working for extended periods of time in jobs covered by the social security program.^[10] Medicare's attempt to distribute benefits in accordance with effort is imperfect, however, in two respects. First, it excludes those elderly people who have exerted sustained effort—in or outside the paid labor force—but not in jobs covered by social security. Second, with few exceptions, it excludes people who have exerted sustained effort in jobs covered by social security but who are under age sixty-five. Thus effort is not the sole criterion. Medi-

care relies as well on ascription—age—as a principle of distributive justice.

Certainly, the elderly are particularly vulnerable to health problems, and they characteristically have both limited incomes and limited access to employer-sponsored private group medical insurance. But other citizens are equally in need of medical coverage. By using ascription as a criterion for distributive justice, Medicare is by design unable to produce results consistent with the criteria of effort or need.

Even among elderly beneficiaries, Medicare coverage has its limits and is far from comprehensive.^[11] Beneficiaries pay a deductible and coinsurance that increases with the length of a hospital stay. SMI involves premiums, coinsurance, and limits on reimbursement for specific services. Outpatient prescriptions and nursing-home care are not covered at all. So while Medicare, particularly with the recent adoption of catastrophic-cost protection, does limit a hospitalized patient's financial responsibility for acute care, it all but neglects a common medical need of the elderly: custodial care for chronic ailments.^[12]

This focus was adopted by Medicare's designers to avoid antagonizing physicians and hospitals. But the emphasis on acute services has also provoked a debate about the relation between medical care and basic goods. In the sense that a wide variety of medical procedures, applied in varying contexts, may save lives, these procedures are as basic as food and water. Yet the dramatic developments in medical technology over the last several decades pose ethical and economic challenges to the simple equation of medical care with basic goods. The ethical question is whether the use of extraordinary life-maintenance techniques prolongs the suffering of patients and their families and constitutes an unnatural or unethical manipulation of life.^[13] The economic question is what proportion of its resources can or should a society allocate to expensive life-maintenance technology. The opportunity costs of systematically applying a variety of expensive treatments to lengthen minimally the lives of terminally ill patients are high.^[14] Indeed, data on cross-national medical expenditures show that these costs rise rapidly with increases in per capita gross national product—a pattern exemplified by luxury goods rather than by necessities such as food and shelter.^[15]

Our final consideration regarding distributive justice concerns vertical redistribution. The program's beneficiaries are primarily elderly people representing the middle ranges of the socioeconomic spectrum. Medicare does involve significant intergenerational transfers, but vertical redistribution from rich to poor is not prominent. But when we consider the market value of Medicare services, we confront a novel form of vertical redistribution. Payments for the medical services provided by Medicare flow to physicians, other medical professionals, and investors in health-care companies. This upward flow makes it difficult to construe Medicare as an attack on property.

Negative Liberty

The HI component of Medicare is financed, like social security, by payroll taxes on prospective beneficiaries. So while recipients have in a sense earned their benefits, their contributions are also compulsory. SMI is paid for in part by voluntary premiums from current recipients, and thus involves no constraint. Subsidies from general tax revenues slightly constrain the liberty of all federal taxpayers in order to extend benefits to recipients who elect SMI.

Apart from the taxes associated with the program, recipients are left unconstrained by Medicare. Medicare beneficiaries generally have as much freedom as privately insured patients in matters such as choice of physician and accepting care.

With respect to medical-care providers, only modest constraints have been introduced by Medicare. In order to facilitate provider cooperation, Medicare's architects explicitly rejected any proposals to intrude upon the organization of medical care. For a while, medical-care providers were important beneficiaries of the new legislation: Demand for medical care expanded slightly, but prices rose rapidly.^[16] Alarmed at the cost, the Nixon administration tried to introduce cost controls and other regulatory devices, and made efforts to stimulate the creation and use of health maintenance organizations (HMOs), which provide comprehensive care for a fixed annual fee.

But provider interests managed to hold the line against these efforts. The medical-care profession has retained the economic benefits bestowed by Medicare without surrendering its autonomy.

The greatest threats in this regard have been the limitations on hospital charges for various diagnostically related groups (DRGs), imposed in 1983, and the Reagan administration's effort since 1984 to encourage physicians to accept Medicare payment limits.^[17] But provider groups may have left their flank unguarded. While they were watching the government, medical care, buttressed by public financing, has grown immensely profitable, and private corporations are moving in to share the proceeds.^[18]

Economic Efficiency

Medicare has reinforced a variety of inefficiencies that characterized the medical-care market into which it was introduced. Among these inefficiencies are: a focus on therapeutic as opposed to preventive medicine, a reliance on expensive medical technologies, hospital charges based on costs, the prevalence of solo and small-scale fee-for-service practices among physicians, the litigious environment of contemporary American medicine, and third-party insurers who insulate both medical-care providers and many of their patients from the most direct, troubling consequences of spiraling medical-care costs. Some of these tendencies are interdependent; fee-for-service practice, for example, lends itself to a therapeutic approach.

This lengthy list of features contributing to economic inefficiency, however, cannot be blamed on Medicare or public policy. Rather these inefficiencies derive from the practices of the private market that the program was designed to leave unchanged. Before examining the provider-related problems for economic efficiency, let us quickly survey the recipient-related economic issues.

Since it serves a population that is already retired or disabled, Medicare has limited effects on labor-market participation. Recipients' rates of use for particular medical services increased somewhat following the introduction of Medicare, but not dramatically. Since use of medical services is not limited simply by economic concerns, public sponsorship of medical insurance seems unlikely to stimulate use any more than private medical insurance available to well-organized workers.^[19]

By providing partial payment for highly specified services tailored to a recipient's immediate needs, rather than income that can

be freely used, Medicare would appear to have little affect on savings incentives. It is unrealistic to imagine that people fail to save because Medicare will cover a portion of the medical expenses they might subsequently experience. Nor could most people ever save enough to cope with the costs of extensive medical care.

Like any social insurance program funded in part by payroll taxes, Medicare may retard corporate or entrepreneurial risk-taking. But since Medicare costs employers less than one-third of the income-maintenance components of social security, its effects on the competitiveness of American goods and services in the global market are minimal.

In 1967 Medicare benefits cost $4 billion; by 1986 benefits had increased more than fifteenfold, to $76 billion.^[20] Serving about 36 million persons in 1986, Medicare consumed approximately 7.5 percent of the federal budget (compared to 20 percent for social security and 1.5 percent for AFDC) and just under 2 percent of GNP (compared to 5 percent for social security and less than 0.5 percent for AFDC). More than two-thirds of Medicare expenditures were for HI, nearly all of that for inpatient hospital care. Of the slightly less than one-third devoted to SMI most was spent on physicians' fees.

The dramatic growth in Medicare expenditures—outrunning inflation generally— is attributable to many factors, but increases in hospital and physicians' fees account for the bulk of it.^[21] Shortfalls in Medicare's trust fund are not far down the road,^[22] and the United States is already spending a greater proportion of GNP on medical services than many advanced industrial societies that provide national health insurance.

Medicare involves several forms of redistribution. From the perspective of individual recipients, HI is an example of life-cycle redistribution, but current practices also involve intergenerational transfers that probably cannot be recouped. Since the incidence of poverty among the elderly has now fallen to the level prevailing among the population generally, the vertical redistribution effected by federal subsidies for SMI is probably less than that entailed by AFDC. But since social security payroll taxes are more regressive than the federal income tax, SMI may contribute somewhat more to vertical redistribution.

As mentioned earlier, Medicare's economic inefficiency reflects

the program's relatively good fit with a sector of the private economy riddled by inefficient practices. This is an unusual predicament in a culture that tends to saddle public programs with charges of inefficiency. Even more unusual have been the efforts of the federal government to stimulate greater market efficiency in the private medical sector. Until recently, in contrast, private insurers have done little in this regard.

Federal initiatives to promote economic efficiency in the private medical sector have been of two types: measures to regulate and efforts to encourage price competition. To date, regulation has predominated. With respect to providers, the Nixon administration applied price controls as part of a more general scheme of economic intervention. His administration also developed Professional Standards Review Organizations (PSROs), which are primarily attentive to hospitalization statistics, and Health Systems Agencies (HSAs), which focus on capital investment decisions. President Carter was unable to secure legislation that would have limited hospital price increases, but the Reagan administration did enforce limits on the federal government's liabilities for Medicare hospitalization and has made modest moves in a similar vein with respect to physicians. These efforts have all helped somewhat to control costs, but the price of medical care has continued to outpace inflation.^[23] Sterner regulatory procedures (bargaining, explicit rationing, and centralized planning) are apt to be required.

As part of its overall effort to control costs, the federal government has also repeatedly raised Medicare premiums and deductibles, and it has kept allowable fees for various services from rising as rapidly as physicians' charges. As a result, Medicare recipients' out-of-pocket expenses have increased, and some analysts recommend increased patient copayments as a cost-control device.^[24] While copayments do reduce federal expenditures, they will not reduce the nation's aggregate health-care bill unless they deter use of services. But there is not much evidence that suggests elderly citizens flock to physicians unnecessarily by contemporary standards. The most prominent increase in services has been in the area of short-term hospital stays, a service attributable more to physicians' choices than to patients' actions.^[25] So copayments appear to be less effective at reducing unnecessary services than at shifting costs to patients.

The limited success of federal regulatory efforts has prompted some analysts and policymakers to suggest that restructuring the medical-care market for better price competition is the strategy most likely to achieve greater economic efficiency. Throughout most of the 1970s, policymakers looked hopefully at group practices, such as HMOs, that provide medical care for a fixed annual fee, thus placing a premium on preventive and ambulatory care. This internal incentive structure seems an encouraging mechanism for cost containment.^[26] Indeed, the most successful and widely known HMOs—Kaiser Permanente, for instance—have kept costs and cost increases well below those of conventional hospitals and fee-for-service practices.^[27] In some metropolitan areas competition among HMOs and between HMOs and conventional medical-care delivery practices has slowed the growth of medical-care costs.^[28]

Despite the interest of federal government leaders and other elites, only modest restructuring has occurred. Many private insurance carriers have integrated specific changes into their operations, but the federal government's repeated efforts to stimulate the development of HMOs have had little success. HMOs are complex, relatively fragile organizations that face difficult circumstances in the current medical-care market,^[29] and some efforts to stimulate new HMOs have constrained existing ones.

Competition and cost-conscious incentives have also led to new problems. Cutting costs by cutting care and self-dealing group practices are abuses every bit as disconcerting as price increases.^[30] Regulation is apt to be necessary under any system of delivering medical services, not just Medicare.^[31] For example, the activities of PSROs in isolating underuse and overuse of hospitalization can be as important in detecting shortcuts in care by HMOs as in detecting cost padding by fee-for-service practitioners.

Today, reformers insist that some sort of restructuring is necessary to improve the economic efficiency of the medical-care market. And almost all analysts interested in reform agree that the only actor with both a sufficiently strong incentive and the necessary organizational capacity is the federal government. Providers have no economic incentive, at least in the short run. Some members of the public have the incentive, but no organization. Insurance companies also have some incentive, but until recently have found it easier to pass along increased costs to their customers. For

the moment, Medicare is not large enough to provide sufficient leverage for a general restructuring,^[32] but some future national health insurance program might.

Administrative Complexity

In part Medicare's complexity is dictated by its mandate. Medical and health services require extensive cooperation from both patients and providers. While Medicare is associated with improvements in health among the elderly,^[33] both patients and providers routinely follow counterproductive practices. Some patients persist in unhealthy habits (smoking, excessive drinking, and poor nutrition). Others cannot escape from environmental dangers at home or at work. Meanwhile, providers practice the episodic acute care for which they have been trained in expensive well-equipped institutions. Most show little interest in either preventive care or in less intensive forms of treatment for the nonacute chronic illnesses that tend to afflict the elderly.^[34]

If we use provision of medical care rather than health as the appropriate objective of Medicare, the inherent requirements for cooperation are still high, but they do appear to fall within the capacities of many providers and patients. Medicare and other federal medical-care initiatives have improved the availability, accessibility, and affordability of medical care.^[35]

Beyond requiring complex cooperation from providers and patients, Medicare requires a fairly complex bureaucracy to handle patients' claims and other aspects of program administration. The complicated rules for coverage, copayments, deductibles, and allowable fees place heavy demands on the Social Security Administration (SSA). Medicare recipients are reimbursed for only a portion of their costs, while covered amounts are paid directly to providers. HI claims are processed either by a recognized intermediary—frequently by Blue Cross—or directly by SSA, while SMI claims are handled by intermediaries. (In contrast, public programs that provide direct services, such as the British National Health Service, are generally less complex to administer but have other characteristics—direct employment of the majority of physicians—unlikely to ever be adopted in the United States.) Despite these complexities, SSA administers Medicare in a fashion that

compares reasonably well with the efficiency of many private insurance operations.^[36] Medicare does enjoy some advantages—large-scale and easy marketing—that help in this regard.

As was the case with AFDC, some of the most troubling questions about Medicare involve limits and harmonization issues. The first limits issue concerns eligibility. The HI component of Medicare is essentially a program for elderly social security recipients. In 1972 elderly non-social security recipients were allowed to enroll, but they must pay the full actuarial cost for coverage. Only fifteen thousand of the roughly one million elderly in this group have chosen to enroll,^[37] with cost the prohibitive factor. A portion of the elderly not covered by Medicare are eligible for Medicaid, and Medicaid has "spend-down" provisions for people who are medically needy but not otherwise recipients of public assistance. But Medicaid, a form of public assistance, is available only to people who have almost no assets.

SMI has from the start been available at the same subsidized rate to all persons sixty-five years of age or older. Over 300,000 non-social security recipients are currently enrolled as well as over 95 percent of the HI beneficiaries.

Medicare's categorical focus on the elderly, of course, excludes the participation of working-aged adults and their dependents, of whom as many as one-third have either no medical insurance coverage or inadequate coverage.^[38] Since the late 1970s the issue of cost control has so dominated discussions of public medical-care policy that questions of access have largely been ignored.

Concerns about the costs of Medicare are exacerbated by several harmonization issues. First, Medicare has contributed to the demand for medical services. While rising demand has generally sustained employment in this sector of the economy, it has no doubt contributed to extremely high levels of inflation in this sector as well. Second, policymakers have developed growing misgivings about the activities of recipients and providers. Access to medical care is only one aspect of health care. Healthy personal habits—diet, exercise, defensive driving—and preventive health care are far more cost-effective ways to promote health.^[39]

Finally, public respect, even reverence, for the medical profession has declined somewhat in the last decade. The industrial character of medical care in America has disillusioned the public and

policymakers alike. By the late 1970s expansion of Medicare was precluded as much by a reluctance among members of Congress to funnel more money into this segment of the economy as by the resistance of physicians and other medical-care providers.

Medicare's public image—cost issues aside—has been generally good. Even concerns about costs have increasingly been directed at providers rather than at the appropriateness of helping vulnerable elderly citizens defray medical expenses. One explanation for the near absence of controversy is that Medicare's public profile is relatively low. Functioning in a manner similar to a private insurance carrier, Medicare has required no dramatic changes in practice for most recipients or providers. While Medicare does not respect Americans' preference for limited and local government, it is not an intrusive manifestation of government. Additionally, most people think about Medicare in tandem with social security and their good will toward the latter rubs off on the former.

Conflicts of Interest and Compliance

Medicare was created to ease the financial concerns of the elderly with respect to medical care. On this score, the program may count several successes. Despite limitations in coverage and increases in recipients' copayments, the proportion of recipients' incomes devoted to medical expenditures has generally declined,^[40] even as the dollar amounts the elderly spend on medical care have increased. And Medicare patients generally pay an extremely modest portion of the extraordinary costs of extensive care. Further, fears that recipients would overwhelm the medical-care system have proved unfounded. We have seen no dramatic surge in the use of medical services by Medicare recipients. Compliance with program objectives by recipients has simply not been an issue.

In contrast, the conflicts between public officials and medical-care providers continue, even though public officials have tried to ensure both profits and relative autonomy to providers. The government has won the struggle over "first principles"—the entry of the state into financing medical care.^[41] But providers remain ambivalent: Medicare is clearly at odds with principles that organized medicine has not abandoned, yet Medicare has bestowed considerable short-term financial rewards on practitioners. Medicare's

limitations on public provision of medical services have turned out to be a mixed blessing in this regard. By enrolling a relatively high-risk population that private carriers were not interested in insuring at affordable rates, Medicare has raised the demand for services and contributed to a situation that has supported larger incomes and profits for medical-care providers. But it has also cracked open the door to public regulation of providers.

For public officials, the primary concern has been the cost of Medicare. Since the early 1970s officials have named rapid cost increases as the major problem afflicting contemporary American medical care.^[42] In their efforts to reduce the rate of cost increases, federal officials have increasingly pitted themselves against the interests of both Medicare's targeted beneficiaries and medical-care providers.

Finally, the decline in the "sovereignty" of physicians at the hands of corporate enterprise may hold a silver lining for advocates of national health insurance.^[43] The more obvious the industrial character of medical care, the less persuasive is the industry's claim to special status. Many service industries—certainly the most basic services—are regulated. If corporate managers continue to displace doctors from control of the medical-care industry, medicine's claims for exceptional treatment will lose credibility.

Summary and Implications

Given the lengthy period of organized medicine's resistance to any sort of public provision for medical care, one hears surprisingly little debate today about whether the elderly need the help Medicare provides or whether it is morally appropriate for the state to provide such help. Instead, the debates focus on costs and cost-benefits. Perhaps even more surprisingly, the culprits with respect to cost are generally held to be private providers and their lack of cost consciousness.

In part Medicare's relatively good image derives from its kinship and similarities to social security. Among social hazards, aging has a special character: its near certainty and fixed position in the life cycle, the widespread perceptions of the elderly as worthy citizens, and the stability of its beneficiary population. And Medicare shares some of social security's popular design features—individual accounts of earned rights, most prominently.

Among Medicare's specific strengths is its criterion of effort, which imparts consistency between the program and American conceptions of distributive justice in the economic domain. And Americans, more so than citizens of most other advanced industrial societies, view medical care as part of the economic domain.^[44] As we have seen, Medicare's infringement on negative liberty is modest; its administration is as efficient as that of its private counterparts; and recipient abuse is not a problem. The program targets a population that has particularly extensive medical-care needs and exceptionally limited capacity for meeting these needs, and has carved out an important area of complementary interests in limiting the financial obligations of program recipients for episodes of acute care.

Medicare also exhibits several important weaknesses. In terms of distributive justice, its egalitarian principles of need and effort mesh poorly with the differentiating criterion of ascription by age. Large numbers of people who need assistance and who have exerted effort in the paid labor market are excluded from Medicare. Even within the ranks of the elderly, Medicare's coverage is incomplete and fails to emphasize basic goods. The program's bias toward episodic acute care covers extraordinarily expensive treatments while needs for the most mundane sorts of care go wanting. This bias also promotes economic inefficiencies that were present in the private sector before Medicare was established. Finally, although Medicare's infringements on providers' professional autonomy are modest, it has perhaps contributed to the profession's concern with economic rather than health matters.

From the central focus of the investments approach—justifying the extension of socioeconomic rights to those confronting social hazards—little about Medicare is disconcerting. Medicare recipients earn their benefits and do not abuse their access to medical care. However, Medicare cannot simply be thrown open to serve all those who need medical care and have no private insurance. The two primary obstacles to extending Medicare into national health insurance are the opposition of medical-care providers and the aggregate cost of such a program. But since other advanced industrial nations do provide universal health coverage, perhaps these obstacles are not insurmountable.

To elicit the cooperation of medical-care providers, any program for the public provision of medical care must address pro-

viders' two principal concerns: profitability and professional control. If we look at the experiences of other nations in creating systems of public medical-care provision, we find no prevailing model for compromises on these issues. But the brief history of Medicare suggests the kind of compromise most likely to work in the United States: Maintain the lucrativeness of providing medical services in return for greater public policy control over the goals and means of medical care. Indeed, since the late 1920s proponents of national health insurance have tried to persuade provider groups that this compromise is in the national interest. Providers resisted this direction forcefully and successfully for a long time, but with Medicare's passage they lost a crucial battle in this struggle. In the aftermath of Medicare's enactment it became easier for providers to accept the payoffs of the program than to continue their earlier resistance. Corporate inroads in the medical-care fields have also been based on this pattern.

To make this compromise explicit and to expedite its implementation, the federal government should create a national bargaining council whose members would represent the federal government, providers, and perhaps other relevant constituencies. This council's central task would be to hammer out a series of guidelines that would assure profitability for providers in exchange for providers' cooperation on the goals and means of medical care. These guidelines should draw on existing institutions and practices. For example, private insurance companies should not be displaced; rather the federal government should expand, coordinate, and perhaps subsidize the efforts of these companies in a national drive to extend at least minimal insurance coverage to more Americans. This insurance could be similar in many respects to Medicare, but would specify affordable out-of-pocket annual maximum liabilities for households in different income ranges.

In return for these financial considerations, providers would be asked to increasingly resolve issues pertaining to the goals and means of medical care collectively with their public official counterparts on the council. These agreements, which would be subject to periodic renegotiation, would specify the sorts of care national health insurance would cover, how that care would be delivered, and the responsibilities of providers and patients. Such agreements might also limit the responsibilities of public policy in prolong-

ing the lives of the terminally ill, or include custodial care for chronic conditions, or make public financial help for certain health problems contingent on patient cooperation in giving up health-threatening personal habits.

The details of such agreements are less important than the initiation of a collaborative process between public officials and providers. In the absence of such a process, no one should expect any broad initiatives in the area of national health insurance.

Eight—
The Investments Approach

The investments approach to American public social provision is based on the five general principles introduced in chapter 2 and developed in subsequent chapters: reciprocity, supplementing efforts at self-help, inclusivity, social insurance, and social merging. Before detailing specific proposals to realize socioeconomic rights, I want to briefly review these principles.

Reciprocity is the foundation of the investments approach: Those who make constructive contributions to society may reasonably expect nurturing assistance from society with respect to basic needs when social hazards afflict their lives. The moral substance of this norm would probably not be questioned by most Americans. Popular reactions even to AFDC bespeak a willingness to help those who are really in need.^[1] However, applications of this principle are bound to create controversy and conflicts with important tenets of American political culture. To minimize such disharmony, I have narrowly defined constructive contributions as participation in the paid labor force; dependents of participants are also included. To recognize concurrent as well as prior participation, satisfactory performance in a training program is also deemed evidence of constructive contribution to the social product.

In developing the specifics of the basic level of support earned by citizens in exchange for their constructive contributions, I have tried to steer between the opposing reefs of classical liberalism and class-conflict theories. The contemporary libertarian interpretation of liberalism is hesitant about using large-scale public programs to provide nurturing assistance when hazards befall citizens

and resists the notion that these public benefits are rights.^[2] But most Americans are willing to relax classical liberalism and remediate some of the social hazards of contemporary advanced society.^[3] Mainstream political opinion recognizes the distressing limitations of both the labor market and community in a purely individualistic society.

Proponents of class-conflict theories, of course, will challenge the rationale of making social program benefits generally contingent on participation in the paid labor market, which they view as an instrument of exploitation. I have sympathy for this view. To be sure, there is little free consent on the labor side of the labor-management nexus.^[4] Nonetheless, this perspective will not be politically influential in America in the foreseeable future. Accordingly, it is only by partaking of certain socially approved activities that those on the fringes of society can hope to merge with the socioeconomic mainstream, thus achieving societal membership and earning personal dignity. And despite the inequities of the labor market, most Americans want the membership status and personal dignity that come from constructively taking part in society as others do.^[5] The investments approach would replace the stigmatizing aspects of public assistance with opportunities for dignified social merging.

My second principle is that social programs be designed as supplements to citizens' self-help efforts. Accordingly, benefits should be formulated to augment—rather than replace or discourage—a recipient's income. Two distinct implications follow from this principle. First, benefits should generally provide less than complete support. Recipients of income-maintenance benefits should be required to augment those benefits in order to live comfortably and should pay a share of the costs of medical services. Second, although benefits are earned, social programs should not be expected to readily supplement the incomes of well-to-do households that cannot reasonably be portrayed as suffering from a social hazard. A minority, but nevertheless a significant number, of social security households now pose problems in this regard. If retirement represents a period of luxurious leisure rather than a social hazard, the case for applying social program benefits is increasingly untenable. Rather than reducing or eliminating earned benefits through means or income tests, however, I recommend taxing

benefits in order to limit the public support that goes to wealthy households.

Inclusivity means that assistance in times of hazard will be accessible for all citizens who meet two criteria. First, the individual must have experienced a covered social hazard: disabling illness or injury; childbirth and child-rearing; aging; family dissolution due to death, divorce, or desertion; and unemployment. Second, the individual must have either made or be making the required contribution with respect to the labor force. In other words, a relevant need of assistance buttressed by the willingness to engage in constructive activity are the sole requirements to be certified. Broad measures of eligibility will make social programs useful to a wide portion of the population over time. Wide usage and the certification of recipients ought to prompt stronger support for these programs from the public and from powerful political figures.

These three principles—reciprocity, supplementation, and inclusivity—point us toward social insurance as a vehicle for resolving many resource inadequacy problems. The concept of social insurance, as we have seen, fits fairly well with American political culture, and the example of social security has set a successful precedent for social insurance using personal accounts. Thus my proposals rely heavily on social insurance programs.

Some resource inadequacy problems, however, do not lend themselves to programs requiring prior contributions. In these instances, social merging programs can extend the social insurance concept to allow for concurrent rather than prior efforts. Social merging programs that demand an effort-based quid pro quo rather than providing means-based public assistance also fit fairly well with American political culture. Their undeniable disadvantage is that they would inevitably force many people to accept undesirable jobs, thereby contributing to the exploitation of the economically weak by the economically powerful. But the advantages of this approach are nonetheless compelling: the effort requirement would allow recipients to maintain their dignity and societal membership; the provision of benefits to supplement recipients' self-help efforts would create incentives for recipients to act in socially approved ways; and the public perception that benefits were not handouts but earned assistance would reduce political attacks on welfare.

The heart of the investments approach resides in these principles, not in the specific proposals outlined in this chapter. Although the following proposals are indicative of my preferences for realizing socioeconomic rights, implementation of the investments approach could take many alternative forms.

Social Insurance

My central suggestion here is to expand the social hazard coverage of the social security program to provide short-term income-maintenance benefits in the event of disabling illness and injury, childbirth, absence of the primary household breadwinner due to desertion or divorce, and unemployment. In order to fund these new benefits with minimal increases in payroll taxes, I offer three proposals to reduce social security outlays for retirement benefits.

Retirement

The current income-maintenance provisions for retirees and their surviving spouses are adequate. My proposals are directed at controlling the cost of these benefits.

First, I propose selectively raising retirement ages so that people in relatively sedentary jobs will retire later than people in jobs that are physically demanding. The range for eligibility to receive full social security benefits might extend from sixty-two to seventy. Such a system of scaled retirement ages would offset to some degree the bias of the current system in favor of white-collar workers, who begin work later in life and generally live longer. As a safeguard, career tracks could be defined such that people could not qualify for early retirement by shifting occupations late in life.

Second, I propose linking benefit levels to increases in the consumer price index after retirement, rather than the current system of continuous benefit indexing during future beneficiaries' working years. This change would have the practical effect of making social security less like a defined-benefit plan and more like a defined-contribution plan.^[6] Increases in initial benefit levels for succeeding cohorts of retirees would be left to the discretion of the political process. As they do now, benefit cash values would vary directly with earnings, and replacement rates would vary inversely with earnings.

Third, I propose taxing a progressively larger portion of retirees' benefits as household income rises. For example, a portion of

benefits would become taxable when a household's total income reaches one-half of the national median income. The portion of benefits subject to tax would increase with income, and all benefits would be taxable for households at or above the median income. Such an approach would assure benefits for those who have paid into the social security system but also reduce the net cost of public social provision for households with relatively comfortable incomes.^[7] This approach also reflects the basic purpose of social insurance, to mitigate social hazards. If little hazard appears, only minimal insurance needs to be applied.

Other proposals might be made to address such issues as the equitable treatment of couples as opposed to single persons, but these are beyond the scope of my intentions here.

Disability

The social security system now provides income-maintenance benefits for qualified workers who suffer a long-term total disability. I propose extending this program to cover short-term disability due to illness or injury. This coverage would supplant the income-maintenance aspects of workers' compensation and would extend protection to non–job-related illness or injury. Benefit levels would be determined under formulas similar to those now used by social security as amended by my proposals above.

With respect to long-term disability, I propose that we reexamine the concept of total disability, particularly for younger workers. It seems reasonable to retrain younger workers for whatever occupations their disabilities allow and to assist them with job placement.^[8]

Childbirth

I propose providing women workers with income-maintenance benefits for a brief period (say, three months) before or after childbirth. Restrictions could be placed on the number of times (twice?) a woman could use these benefits. Again, benefit levels would be linked to earnings histories.

Survivorship

Social security now provides survivorship benefits for dependents of qualified workers. These benefits seem appropriate for retirement-aged surviving spouses, but I propose restricting survivorship benefits for working-aged adults. Working-aged spouses would be allowed to collect benefits for no more than, say, five years, and children could collect benefits until age eighteen (age twenty-two for full-time students). These provisions would restore college benefits for children but require that spouses adjust to their new circumstances within a fixed period.

Desertion and Divorce

I propose that working-aged adults and children separated from their family's primary breadwinner by divorce or desertion receive short-term (perhaps one month) income-maintenance benefits at levels comparable to other social insurance coverage. (Longer-term provision is discussed in the following section on social merging.) As with childbirth benefits, eligibility could be restricted to two or three times in an individual's life.

Unemployment

I propose that short-term (perhaps one month) unemployment insurance benefits be offered through social security. (Again, long-term provision is discussed in the following section on social merging.) Limits on eligibility could be set at once every two or three years. This proposal would replace the current system of unemployment insurance.

Administrative safeguards

Some safeguards will be needed to prevent workers who have achieved initial eligibility for extended benefits from abusing the system by creating a series of illnesses, injuries, childbirths, family dissolutions, or unemployment episodes to avoid the workplace. One such set of safeguards includes my proposed limits on the duration of benefits and the number of times a worker could collect certain benefits.

In addition, a blanket rule could restrict the overall use of benefits to a specified portion of a person's working life (from the end of his or her formal education to retirement). This proportion could be expressed on a sliding scale, with younger workers eligible to receive benefits for a fairly high proportion (say, 30 percent) of their working life thus far, and older workers limited to a much lower percentage (say, 5 percent) over a roughly forty-year working life. This scale would accommodate the high incidence of legitimate hazards appearing early in life, while clearly implying that sustained periods of participation in the labor force are required to earn benefits.

Costs

I have suggested several ways to cut the current costs of social security benefits for retirement, long-term disability, and survivorship as well as outlays for current workers' compensation and unemployment provisions. Though no one can accurately predict the results, I suspect that these savings will be more than offset by the costs of my proposals to extend social security coverage to short-term disability, childbirth, and household dissolution.

Should it be necessary to further increase revenues, my preference is for a progressive social security tax on all wages. The

rate for a worker's first $10,000 of annual income could be held at the current rate (7.51 percent) or even reduced a bit, with slightly higher rates for each tier of income, defined in increments of $10,000. In conjunction with the progressive taxation of social security benefits, this proposal would require ever closer coordination between the Social Security Administration and the Internal Revenue Service.

Social Merging

The heart of my proposals to facilitate social merging lies in expanding the incentives, opportunities, and capacities for self-support for those persons near or beyond the fringes of the labor market. Complementary measures are intended to enhance the capacity of the existing labor market to provide these people with full-time employment.

Incentives

In order to increase the incentives for self-support and decrease the "unintended rewards" that our current system sometimes provides for not working, I propose limiting public assistance to Supplementary Security Income (SSI) and including only persons with certifiable total physical or psychological disabilities and elderly persons not covered by social security. We should be fairly hard-nosed about admitting people to this program, but those who do meet the criteria should also be entitled to more intensive therapeutic and custodial care than is now typically available. Such care might be entrusted to private charities that accepted federal guidelines, perhaps with some public funding assistance.

Rather than terminating much of current public assistance in one motion, we could gradually phase out assistance to various categories of recipients, beginning with those who have the surest prospects for success in social merging programs.

Opportunities

First, to assist unemployed working-aged adults in increasing their opportunities for self-support, we can reinvigorate our system of public employment services so that they serve more adequately as labor exchanges.

Second, to facilitate parents in entering the labor force and contributing to the support of their households, I recommend establishing a social insurance program of universal supplementary chid

allowances. In contrast to the social insurance programs mentioned in the preceding sections, these allowances would be sharply income-variable and would involve substantial vertical redistribution. Since this is the least conventional of my proposals, I will describe it in greater detail.

Let us consider a program that would provide income-variable supplementary benefits for up to three children per household. This program would cover all American households, including the vast majority of the children now served by AFDC, without subsidizing undesirably high birthrates. Households with annual pretax incomes up to that resulting from full-time work at the minimum wage (currently about $6,700) would be eligible for the full benefits; there would be no work disincentive for these low-income families. Benefit levels would be sharply staged so there would be minimal incentives for new births. And benefits would be taxed as income, further reducing the income-variable benefits for middle-and upper-income families.

Full benefits would be set at $175 a month for the first child, $125 a month for the second child, and $75 a month for the third child—for a maximum of $375 a month, or $4,500 a year. Thus a household with wages equal to those derived from full-time employment at the minimum wage and three children would have a total pretax income of $11,200 ($6,700 + $4,500), placing it at approximately the current federal poverty level for households of this size.

Households with income above the level provided by full-time employment at the minimum wage would see their benefits reduced dollar-for-dollar down to a universal minimum benefit of $75 a month for three children ($35 for the first, $25 for the second, and $15 for the third).^[9] This scaling would keep benefits extremely modest for middle- and upper-income families, although it might reduce incentives for low-income families to increase wages modestly. But these graduated levels of benefits would concentrate public social provision on households with the greatest need for child allowance supplements; that is, this program would have high target efficiency.

This program would be financed by a progressive payroll deduction on all wage earners. In addition to involving a substantial degree of vertical redistribution, this program presents an unusual

case of life-cycle redistribution, since child-rearing characteristically comes fairly early in adulthood.

It is possible that the benefit levels I have suggested would allow some rural households to subsist with little or no labor-market participation or would encourage families to migrate to rural regions.^[10] Should these consequences become problematic, benefits could be adjusted to discourage such activities, although moderate levels of migration might well be preferable to current inner-city conditions.

There are other means of striving to assure minimum levels of support to children. Indexing the minimum wage to the consumer price index and enforcing child-support court orders to noncustodial parents are frequently discussed. While neither of these options would contradict the general principles of the investments approach, neither would as efficiently and thoroughly accomplish the ends served by the child allowance system I have described. Indexing the minimum wage is not likely to concentrate resources such that single parents with limited skills can support more than one child. Increasing the rigor of child-support payments, as Wisconsin and other states are doing, will not facilitate household support for children whose noncustodial parent is unidentified, incommunicado, unemployed, or destitute. Thus I see these two options less as alternatives to child allowances than as sensible additional measures serving related, but distinct objectives.

My third concern regarding opportunities is that we cannot reasonably insist that parents, particularly single parents, work full time unless affordable child-care is available to them. Here, we have several alternatives in addition to commercial child-care centers: subsidies to employers who create child-care facilities in their workplaces, public child-care extended through the public school system, and reimbursements to parents for the costs of paying relatives or neighbors to look after children. The crucial point is that, while child-care raises the costs of social merging, many parents—especially single parents—cannot contribute to the support of their households in the absence of some provision for child-care. Whatever specific measures we adopt, we must make child-care available at subsidized rates for low-income families.

Collectively, the proposals in this section are designed to enhance individual opportunities by: (1) improving information about work

opportunities, (2) supplementing the financial returns of work on the fringes of the labor market for households with children, and (3) facilitating employment through the provision of adequate, affordable child-care.

Capacities

My recommendations here comprise measures intended to improve unemployed and underemployed workers' capacities for greater labor market participation. First, we need public work programs for young unskilled workers. Such programs could take many forms, including the model of the Civilian Conservation Corps (CCC). These would not be expensive or extensive training programs designed to teach sophisticated skills, but rather programs intended to instill basic work habits (attendance, punctuality) and to provide participants with an employment history.^[11] In effect, these programs would certify participants as good bets for subsequent employment and training by private employers. Small-scale work programs would be particularly appropriate in remote areas; in that context programs would be minimally disruptive of local economies.

Second, we should offer sophisticated training, retraining, or education programs to adults with satisfactory records of employment (say, encouraging work histories of at least five years) who face sudden labor market problems—such as layoffs in declining industries—and to parents (generally mothers) in households that experience changes in membership that compel them to enter the labor market. Giving individuals in these two categories opportunities to upgrade their labor market capacities is a social investment more than likely to return its costs, as participants' subsequent income tax payments are apt to exceed training expenditures. Among all high-unemployment groups, for example, single-parent homemakers have been the most successful in taking advantage of training opportunities to move on into full-time employment.^[12]

By satisfactory progress in these training programs, participants would also earn temporary income-maintenance support, in effect a scholarship to subsidize their vocational education. The amount of this support would equal that provided for unemployment or family dissolution under social insurance.

Complementary measures

To further expand the demand for unskilled, semiskilled, and entry-level workers, the federal govern-

ment could create more low- and midlevel positions as well as provide subsidies for state and local governments that wish to do the same. These efforts would help to replace the decent-paying jobs lost in manufacturing and thus increase the capacity of the labor market to provide positions capable of supporting households, particularly for the residents of central cities.

Demand could also be enhanced by offering employers in the private sector tax incentives for creating new low- and midrange positions. For example, a sharp tax break could be available to employers who increased their payrolls by a small percentage. This percentage would deliberately be relatively small to minimize displacement and training problems.

Implications for Housing and Education

Although housing policy and education are not an integral part of my analyses and proposals, the investments approach holds several important implications for public services in these areas. Chief among these is the principle that the public provision of basic goods and services needs to be tailored as closely as possible to the peculiarities of our nation's political culture. In many other advanced industrial nations, for instance, direct public provision of housing units is widely practiced and accepted. But in the United States such programs are generally aspects of public assistance and, with the exception of projects categorically limited to the elderly, residents often feel stigmatized.

Housing subsidies

Consistent with current trends in American policy,^[13] approaches that help recipients of public assistance participate in the regular housing market are desirable. This preference is not so strong as to mandate massive and abrupt changes in housing policy. And in specific urban areas, and perhaps other locales, expanded direct provision of accessible public housing units remains essential.^[14] But overall I would suggest that housing is best regarded as an aspect of income maintenance and that we continue to reduce the direct public provision of housing units, as feasible.

The use of housing vouchers, rather than cash payments, is an option that has achieved some support in recent years.^[15] Vouchers involve more complicated administrative services than cash, but

they may be more politically acceptable to citizens who raise suspicions about how public program recipients use their benefits. Vouchers are compatible with an approach to housing as an income-maintenance matter as long as the vouchers are legal tender for housing, and recipients are allowed some flexibility with respect to rents. For instance, households eligible for income-maintenance payments under social merging programs might be given a housing voucher worth one-half of their monthly cash benefits. Thus benefits for housing would amount to one-third of the income-maintenance total. Recipients could choose whatever housing they preferred, paying higher rent or home payments out of their pocket. This feature would be particularly helpful for families who do not want to move or sell their homes during a short-term disruption of income. And we ought also to continue giving recipients cash rebates for rents that do not consume a stipulated portion of their income-maintenance benefits.

Education

My analyses and proposals also hold interesting implications and opportunities for expanding the public school system. First, the system could be expanded to provide more vocational education for adults. Such programs would facilitate my proposal to offer training for experienced workers and would also create good jobs in the public sector.

Second, as mentioned in my discussion of child-care, one mechanism for providing affordable child-care would be to extend the public school system to include preschool programs for younger children.^[16] This suggestion would also serve to create new jobs in the public sector.

Together these two programs would also link people's lives more thoroughly to the public school system. And public support for the schools might increase as schools expanded their service capacity and became a focus for family and community services.

Finally, the principle of using social policy to reinforce desirable activity could be applied to stem the tide of parents' transferring their children from public to private schools. A particularly troubling aspect of this trend is that many parents who have opted out of the public school system are precisely the concerned activists who, had they stayed, would have expressed their dissatisfactions to teachers and school administrators, and thereby would have instigated changes and improvements.^[17] One way to encourage par-

ents to keep their children in the public schools would be to give students who attend public schools some degree of preferential treatment with respect to public financial aid for postsecondary education.

Medical Care

Some crucial aspects of the provision of medical care—gaining the requisite cooperation from physicians and other medical-care providers—lie outside the primary focus—socioeconomic rights of citizens—of the investments approach. Unfortunately in light of this, other aspects of medical-care provision—assuring people financial access to medical care—remain integral concerns of the investments approach. It would be much simpler to follow the libertarians by limiting ourselves to an income-maintenance system;^[18] however, such an approach will not assure citizens protection from common social hazards. The need for medical care is highly intermittent for most people, but the costs—once care is needed—are frequently quite high. These features make it infeasible to cover the provision of medical care through income-maintenance means. Instead, health insurance that assures access to care at manageable cost—the technique adopted by the contemporary American upper middle class—is the appropriate tool.

The process of assuring access to medical care in such a fashion need not entail confronting the providers of medical services with intolerable changes. What I have in mind is a system similar in some respects to the decentralized, quasi-public national health insurance of the Federal Republic of Germany. This system should have the following general features. With respect to depth of coverage we need to develop a reasonably comprehensive national minimum. In most areas of provision this minimum could be similar to the current Medicare. But we should place some feasible, income-graduated maxima on out-of-pocket annual expenditures (for incomes below $10,000, $500; for incomes between $10,000 and $20,000, $1,000; and so on). With respect to breadth of coverage, this program should represent an integral aspect of employment. The carriers themselves would be the existing private insurance companies. And coverage would be extended throughout the working population by public subsidy for certain groups of employers

and employees. Households with children, even those with no employed adult, would be served through the deduction of the relevant insurance premiums from the universal child allowances. These broad features leave the practice of medicine relatively unchanged for most providers. The basic changes are the extension of access to medical care through the requirement that employers offer their employees insurance coverage meeting the minimum criteria and the use of public subsidies to help pay for coverage.

With this brief overview behind us, let us look with greater care at the specifics. My guess is that developing this sort of coverage requires as a prerequisite the establishment of a national bargaining council, as discussed at the end of chapter 7. This council, operating within general guidelines such as those outlined in the previous paragraph, would create an initial version of the specific rules about prices and limits. Existing providers could expect to continue their activities as relatively autonomous units as well as to receive their customary incomes. Public officials in return would gain assurances about future cost increases. I would anticipate that limitations on provider practices would initially be fairly modest and would grow with time, experience, and the gradually broadening recognition that the providers of medical care in America form an industry, and that, as is the case for other important service industries, some degree of public regulation is necessary.

The most important aspects of the initial measures involve extending access to medical care to working-aged Americans and their dependents. The basic mechanism here involves a program of subsidies to low-income employees, and as necessary to small-scale employers, that allows employees to join a system of compulsory medical-care insurance meeting minimum national criteria. The employees of small-scale employers might, for instance, be consolidated into several groups for which risks are shared by private carriers with federal subsidies.^[19] Employers could be required to offer a choice of plans varying in coverage, expense, and manner of care—HMO or traditional fee-for-service. And a variety of cost-control features should characterize minimum coverage as is increasingly the case for existing private group insurance. The central objective of this coverage would be to place feasible upper limits on the annual out-of-pocket medical expenses of families in varying income ranges. Conventional practices with respect to deductibles

and coinsurance would not be excluded—only limited—by this goal. Whether coverage would extend to all known treatments would have to be debated in the bargaining council. These aspects as well as other facets of coverage limitations would be subject to change and, as with other rights' limits issues, would represent public decisions about cultural conventions.

Households with children would be covered under this program, regardless of the employment status of their adults, by virtue of an automatic system of deducting the relevant insurance premiums from the child allowances. This practice is consistent with the conception developed in chapter 2, that children are not responsible for their predicaments and should have support extended to them on the basis of need. Collectively, these two proposals would cover employed households and households with children. Unemployed single adults and childless couples should be able to join this insurance scheme at subsidized rates similar to those available for employed persons. These high-risk individuals could be spread across insurance carriers so as to minimize the problems of specific companies.

Individuals for whom an episodic social hazard did not mean an end to employment would be covered by the system described above. Those whose hazards spelled an end to employment—retirement or long-term total disability—would be picked up by the Medicare program. This program would also serve those enrolled and making satisfactory progress in job training programs. People whose disabilities placed them in the residual SSI program would also come under Medicare, but hopefully under expanded provisions that allowed appropriate human-intensive care.

Existing Medicare provisions can generally stand. But I do propose some changes that would bring Medicare into line with my other medical-care proposals. First, I propose an out-of-pocket ceiling for covered care similar to that for the households of working-aged adults. Second, we should extend limited coverage of low-intensity care for chronic conditions. Third, I propose cutting back the degree to which public programs support prolonging the lives of terminally ill patients.

Realizing these provisions would assuredly add to the proportion of GNP that is funneled into the medical-care sector of the economy, and it would add as well to the costs of public programs

directed at medical-care problems. For these reasons it is crucial that the provisions follow the efforts described in chapter 7 to create cost-control structures by both public regulation and market means. I suspect that no one could predict with any accuracy the cost increases that these proposals would involve, but I will identify major components. Extending minimum national medical-care insurance coverage to working-aged adults and their dependents will involve the largest single cost increase. I have suggested a way of placing limits on these increases by explicitly raising the possibility of placing some new, expensive, and exotic treatments beyond the limits of this minimum coverage. Coverage for these treatments should be available for those who want and can pay for it, but it need not be included in the initial basic minimum public coverage. Within Medicare, expenses would rise as a consequence of realizing these proposals, primarily due to the addition of new beneficiaries—participants in job training programs and SSI beneficiaries. The three recommendations that I made with respect to the existing Medicare provisions offset one another. Adding annual out-of-pocket payment ceilings and limited low-intensity care for chronic conditions would both increase expenses, but these increases would be largely offset if substantial cuts were made in the use of Medicare to prolong the lives of the terminally ill. The overall increases in public medical-care expenses that these proposals entail would add to the case for building progressive tiers into social insurance payroll tax deductions.

The Investments Approach and American Values

Anyone can, of course, trot out a series of social policy measures that sound encouraging to sympathetic ears. That is, however, not the point of the foregoing. Rather these proposals not only address a broad range of the social hazards prominent in contemporary America, but they do so in a fashion consistent with the five general principles discussed at the outset of this chapter. These principles are, in turn, consistent with important values of the American political culture. I will argue the case for these claims by applying the questions used in the case studies of chapters 5–7 to these proposals. Since my proposals cover a broader gamut than

any of the individual programs examined earlier, I will only sketch out the highlights, both encouraging and discouraging.

Distributive Justice

As important as any characteristic that this package of proposals offers is a set of criteria for the distribution of benefits that is more rigorous than either contemporary public assistance or some aspects of American social insurance. The dominant distributive criterion realized through these proposals is the exertion of effort in the paid labor market. Benefits are earned through prior or concurrent contributions of effort, which are recorded in individual accounts. The investments approach thus affords vulnerable citizens a dignified means for coping with a variety of social hazards. The distributive criterion of results (vertical equity) characterizes these proposals as well, since benefit levels reflect past earnings, just as current social security benefits do. This effect, however, is diluted by the taxation of benefits paid to middle- and upper-income households. Taxation also reinforces the fairly strong linkage of benefits to the provision of basic resources, a tie established by the supplementary character of the benefit levels. That benefits are intended to supplement recipients' efforts at self-support both precludes total dependency and enhances dignity.

Considered collectively, these proposals involve enough vertical redistribution to represent an assault on property more extensive than that entailed by existing social programs. Both the child allowances and the medical-care proposals, for example, address distinct concerns by redistributing resources toward our most vulnerable and needy populations. But these proposals temper vertical redistribution with the constructive-efforts requirement for working-aged adults; only children and the severely disabled are excused from this rule.

In summary, these proposals do not assure every citizen of support; rather, benefits are a quid pro quo. In this regard, these proposals have more in common with the market than with a system of distributive justice based on need. But whereas the market represents a results-oriented struggle, the survival of the most capable, these proposals operate on the basis of effort. Under the invest-

ments approach, market forces are supplemented by a measure of public provision so that people who apply themselves to support their households will have a much better chance of being able to afford basic goods and services.

Negative Liberty

One drawback of relying more heavily on social insurance is that its compulsory character would pervade life more thoroughly. As noted in chapter 5, however, the compulsory nature of social security has not provoked even moderate opposition. Indeed, in chapter 7 we found an example of people jumping at the chance to participate in a voluntary program of publicly subsidized insurance, SMI (Part B of Medicare). We do need to keep in mind, of course, that most social security beneficiaries to date have received far more in benefits than they have paid into the program,^[20] and far more than they might have reasonably expected from committing similar amounts to private investments. But, for the programs I am proposing, the ratio between contributions and benefits will be less heavily weighed in favor of benefits, which may create more reluctance about compulsory participation. The breadth of social hazard coverage and its inclusivity regarding potential recipients may, in turn, offset this reluctance to some degree. Apart from compulsory taxation, these proposals do not restrict the liberty of recipients in formal ways.

The private professionals most obviously affected by these proposals are physicians and other providers of medical services. My proposals require that, in return for reasonable income guarantees, these providers will increasingly have to share decision making about the ends and means of medical care with public officials. Like other basic service industries, the medical-care industry will become increasingly regulated.

Additionally, these proposals have disconcerting consequences for social workers. Apart from the scaled-down program for psychological and physiological disability, my proposed programs do not call for caseworkers. The role of caseworkers in the public sector would be reduced to activities in the criminal justice system and medical institutions.

Economic Efficiency

These proposals embody strong work and savings incentives. By and large, contributions in the form of paid labor would be necessary for program eligibility, and benefit levels would be designed to supplement household efforts at self-help. The proposed limitations on the use of individual hazard provisions and the blanket maximum-use provision represent safeguards against abuse and malingering.

The starkness of the work incentives would appear to strengthen the hands of corporate managers and entrepreneurs, although program costs would offset to some degree whatever advantages accrue to managerial risk-takers. The higher payroll taxes would surely increase the cost of doing business in the United States, but they would not raise such costs beyond those now incurred by some of our important international competitors—the Federal Republic of Germany, for instance.

A less sanguine possibility is that these proposals might reinforce either of two disturbing trends in the labor market: the bifurcation of the labor market between highly skilled, well-paid workers and relatively unskilled, minimum-wage workers or the exportation of jobs offshore.

Without doubt, these proposals would cost more than we now spend for social security, Medicare, Medicaid, AFDC, unemployment benefits, workers' compensation, and SSI. One cannot predict how much more. While the social insurance components would dominate expenditures, the greatest proportional cost increases are likely to come in the social merging programs. Relying more heavily on social insurance would shift the costs of public social provision more thoroughly to those who enjoy its benefits; the increased costs of social merging would represent the price of dignity, the price of creating conditions that facilitate dignified self-help.

The heavy reliance on social insurance and the supplementary character of benefits would create a life-cycle redistribution pattern. Substantial vertical redistribution would be injected by the child allowances, the proposals to expand the labor market, and the benefits for adults enrolled in training programs. These elements are essential for social merging, however, and each requires

formally or informally some type of quid pro quo from program recipients.

Through their inclusivity these proposals generally avoid the problems associated with focusing public social provision on particularly disadvantaged groups. Even the child allowance program is universal, albeit income variable, thus affording higher target efficiency.

Administrative Complexity

The social insurance proposals would require a substantial escalation of administrative activities within SSA and related bureaucracies. For this reason, the programs should be phased in over time rather than implemented simultaneously.^[21] The income-maintenance programs call for precisely those activities we already accomplish fairly well; they would require an increase in the quantity of administrative activities, but not a change in their nature. Other proposals, such as job training and child-care in particular, would require a greater degree of decentralized complexity. In dealing with particular recipient groups, these programs are likely to confront the difficulties identified in chapter 4. But even here we have some encouraging experience in each of these areas,^[22] and my proposals are generally tailored to these experiences.

Overall, these proposals attempt to come to grips with the major limits and harmonization issues. With respect to limits, different mechanisms apply to income maintenance and medical care. For the former, the principle of supplementary benefits and the progressive taxation of benefits assure that most benefits will be used to purchase basic material resources and that households with high incomes from other sources will receive only limited benefits.

In the case of medical care the limits problems are more severe. Rapidly expanding technological capabilities and growing uncertainties about the cost-benefits of existing practices create systematic dilemmas about the appropriate limits of medical care, particularly the public provision of care. Social programs cannot resolve these problems, but my proposal for a bargaining council provides a mechanism for focusing national policy discussion of limits issues. More importantly, the bargaining council's initial

recommendations and decisions—controversial and inadequate as they might be—could prompt a much-needed ongoing national debate on these issues.

One harmonization benefit of the investments approach lies in the complementary nature of the distinct provisions. The independent programs mesh together to avoid gaps or duplication of coverage and to provide consistent incentives for self-support.

But other harmonization issues may be exacerbated by the investments approach. The cost will create new budgetary difficulties, and the denser public policy environment may provoke unforeseen problems. Additionally, these proposals would stimulate aggregate demand and employment, thereby raising inflationary pressures. Despite selective measures to constrain these pressures—the supplementary character of benefits, the focus on high-unemployment populations, and the regulation of medical costs—these proposals are inflationary.

Overall, however, the investments approach represents a practical realization of the five principles central to American values in the area of socioeconomic rights. As national programs, though, these proposals do not operationalize concepts of limited and local government.

Conflicts of Interest and Compliance

For beneficiaries the investments approach provides broad humane incentives for cooperation in that programs would respect and reinforce the dignity of individual beneficiaries and their membership in society. Inclusivity and reciprocity should also reduce conflicts of interest between public officials and beneficiaries, while enhancing popular perceptions of public social provision. All constituencies would realize that prospective beneficiaries would include a broad cross-section of the American voting population, not just small categories of the particularly disadvantaged, and that through social programs the nation's public policy would be encouraging responsible behavior.

Conflicts of interest between private providers of medical care and public officials over costs and professional autonomy are likely to be aggravated, however. The best that the investments approach offers physicians, hospital administrators, and insurance com-

panies is the tradeoff of fairly lucrative financial returns in exchange for a loss of autonomy with respect to the ends and means of medical care.

While social workers may protest their reduced role in the new programs, these programs would expand opportunities for third-party providers of employment, job training, and child-care services.

Prospects

On balance, the investments approach represents an improvement over existing social programs with respect to core American values. It is reasonable to ask whether this approach will work. Clearly, no set of proposals will meet each and every person's needs. But within the context of American political culture, the programs I have recommended would demonstrably facilitate and supplement socially desirable efforts at self-help.

It is also reasonable to ask whether the investments approach is a neat theory that bears little relation in practice to the day-to-day lives of taxpayers, program beneficiaries, program bureaucrats, or prominent public officials. The investments approach, it might be argued, is just labeling. But in political matters, labeling and perceptions are crucial. And from the standpoint of the personal dignity and societal membership of program recipients, an earned income supplement is far preferable to an imposed work-fare assignment or public assistance that is viewed by fellow citizens as a handout.

Compared to current practices in other advanced industrial societies, my proposals may be criticized as being too narrow. The child allowances, for example, assure only poverty-level incomes for parents who work full time. This degree of generosity is not exactly stunning. And American society would probably be the better for it if we could bring ourselves to be more generous, by including, for instance, rearing children and housework among the contributions to the social product that would merit support. However, our political culture would not accept the establishment of the broader programs that many other advanced industrial societies take for granted.

I conclude that we have the moral, political, and economic tools

necessary to reduce social problems through the improvement of social insurance and the development of social merging programs. I will now turn to the question as to whether we will choose to use these tools. Can we expect a set of proposals such as these to be adopted in the near future?

For several reasons, I think the best we can expect is a slow incremental reform of current policies and procedures. One source of opposition to the investments approach will come from people committed to a highly limited state. People for whom market distribution, negative liberty, and the economic efficiency of unadorned market operations are extremely important will be far less impressed by the proposals offered here than people for whom these values represent only a portion or a limited conception of "the good." Further, although the design features of the proposed programs fit important American values, experience has shown us that compatibility or incompatibility is not a good predictor of the ease in adopting new social programs. Despite its compatibility with American values, Medicare faced decades of opposition, while the notably incompatible AFDC was adopted far earlier. Value compatibility does appear to have eased problems for Medicare's operation once it was adopted, however.

Yet even if we set aside lingering value conflicts, my proposals would have to compete with a host of meritorious projects for scarce public resources. The dollar cost of the investments approach is high, and so, too, therefore is the opportunity cost of implementing it. Successful adoption of the investments approach would require a fairly broad cross-section of prominent American public officials—the president and key members of Congress—to place it high on their priorities and political agendas. But a constellation of factors makes a broad consensus on reform a relatively rare event in American politics. Some of these factors are cyclical. Samuel P. Huntington describes sixty-year cycles in American politics that involve impressive reform efforts at one juncture, but he conceives of these reform periods largely in terms of attacks on state power.^[23] (For that reason, he does not include the New Deal—an effort to use state power to solve social problems—among his reform periods.) Even if Huntington is incorrect about the duration or motive of reform movements, clearly the mood of American political elites during the last decade has not been sup-

portive of social program reform in any sense other than cutting costs.^[24]

Hugh Heclo detects another historical pattern, identifying four periods in the development of social programs in advanced industrial societies: experimentation (1870s–1920s), consolidation (1930s–1940s), expansion (1950s–1960s), and reformulation (1970s–?).^[25] The dates fit Western European history better than American, but the last two phases of this pattern are relevant to recent American experience. The United States expanded social provision sharply in the late 1960s and early 1970s, and it did so in ways that fit poorly with American political culture. As would be expected, these expanded activities—at odds with cherished values—have drawn increasing attention and raised a hue and cry for reformulation. But these dissatisfactions began precisely at a time when the postwar economic boom yielded to the stagflation of the mid-1970s. As a consequence, social programs have had to compete for resources in a more constant-sum economic environment.^[26]

Yet another set of obstacles to comprehensive social program reform is posed by the relatively meager capacities American national political institutions have for sorting out diverse claims and forging a coherent national agenda.^[27] As social programs have become more prominent in American politics, their development has become a focus for more numerous interest groups, none inclined to compromise.^[28] In this light the interlocking character of my various proposals probably represents a political liability. Failing an unexpected cataclysm,^[29] such as war or depression, the current political environment is much more likely to allow piecemeal changes rather than a systematic renovation of American social programs.

The relative impotence of American political parties further muddles the characteristically American situation of highly fragmented, competing priorities and agendas among prominent political officials. A determined president joined by key congressional leaders could, for instance, establish the bargaining council that is prerequisite to national medical-care insurance. But the presence of a private insurance system for the upper middle class and well-organized workers siphons off much of the voice that would help prompt such widespread determination among preeminent political officials.^[30] And while social programs sometimes draw unex-

pected benefits from the interaction of competing elites,^[31] it is unlikely that the side effects of contention will produce an integrated scheme of proposals.

For all these reasons, the best we can expect is that any new policy increments will follow principles more in accordance with important American values than was characteristic of the changes enacted in the mid-1960s and early 1970s. This forecast may frustrate proponents of public social provision, but it would certainly be preferable to the recent calls among political elites for a general retreat from social programs. The call to retreat is hardly worthy of political theories and practices that profess to place significant value on the individual. Weaknesses in social programs do not discredit the underlying needs that socioeconomic rights address, and we have seen that these rights are crucial to individuals living in an advanced industrial society. In a rush to rectify poorly designed programs, we must be careful not to abdicate our responsibility or abandon our principles about defending individuals. Rather than retreat from the difficult issues of public social provision, we need to make more careful determinations of what forms of defense are appropriate, supported by a clearer vision of the limits of state capacities.