C H A P T E R

6 Who Gets What, HowMuch, and Under What Conditions Analysis of Eligibility Rules

“That’s not a regular rule, you invented it just now,” said Alice. “Yes, and that is the oldest rule in the book,” said the King.

—Lewis Carroll, Alice in Wonderland

Introduction

Fifty years ago, textbooks on economics referred to air and water as examples of “free goods.” So far have we come from that more plentiful time that it is now difficult to cite any example of a free good: free in the sense that it is neither rationed, regulated, nor priced. Because no social welfare benefit is a free good, rules and regulations allocating such benefits abound. Such rules and regulations are not dispensable. As long as the de- mand exceeds the supply of benefits and services, some rule or principle must be used as a guide for deciding who gets the benefit or service and who does not.

Social workers and other human service practitioners need to understand eligibil- ity rules because they work daily within the context of these guidelines and use them at all levels of complexity. For example, the practitioner may need to seek exceptions from those rules to meet a client’s/consumer’s special need or need to understand the eligibil- ity rule to decide whether to advise a rejected client/consumer to seek an administrative hearing on the issue. As an agency representative, the practitioner needs to understand the details of the rule so the applicant has the same chance to receive a benefit as every other citizen.

Practitioners must also live with the fact that they are in the business of denying as well as qualifying clients/consumers for benefits—a hard fact of life that is a consequence of scarce resources. In an earlier chapter, the argument was made that finite resources are one reason social policies had to be invented; in an important sense social policies are the

109

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

vehicle by which social resources, services, and benefits are rationed when there isn’t enough for everybody under every condition. So, as long as there are insufficient re- sources for every conceivable social need, every time a benefit or service is given to one client/consumer, it takes away the opportunity to give it to another one in need.

Eligibility rules are the most important vehicle for rationing benefits and services. On the positive side, they seek to target resources on those who need them or those who need them most. If they are off-targeted and go to clients who don’t need them then, at some point in time, someone who does need them will go without.

It is a mistake for practitioners to think they can just “work harder” to deliver ser- vices to clients/consumers so that no one will go without. Practitioners’ time is also a scarce and expensive resource, every bit as scarce and expensive as cash. It is tempting to think that a way around this problem is to deliver services via a “first-come, first- served” eligibility rule. Although that rule has qualities of “rough justice” that are somehow appealing, the justice involved is probably illusory. Think of how a first- come, first-served rule gives a not necessarily merited advantage to those who by chance hear of the rule or the service first; there is no particular justice in that. At some point, those who implement that rule will run out of resources, so that denying clients/ consumers has only been postponed. First-come, first-served is not inherently a bad eligibility rule, but it has no great virtue either—why shouldn’t it be preferable to give preference to those who are, on some basis, most needy? Indeed, practitioners may even be responsible for constructing such rules at some time later in their career and at that moment there is at stake a professional responsibility for good service to clients/ consumers. When a policy does not meet the needs of clients/consumers adequately, neither the exercise of simplistic eligibility rules nor a large dose of moral indignation will suffice to discharge professional responsibility. Practitioners are responsible for ad- vocating their clients’/consumers’ needs even to their own administrative superiors as well as to their colleagues in other agencies who have resources that clients/consumers need. Furthermore, practitioners are responsible for joining with others in pursuing legislative or judicial advocacy as a remedy.

Types of Eligibility Rules

The decentralized disarray of the U.S. welfare system creates literally hundreds of pub- lic and private programs that offer welfare services and benefits. Each has a somewhat different set of rules for determining who gets what, how much, and under which condi- tions. Faced with this bewildering variety, we need to reduce its complexity by some kind of scheme that makes it more understandable. The purpose of the following scheme is to group together eligibility rules so that we can talk about types of eligibility rules without the trouble of weighty discussions about lightweight differences. Many schemes serve this purpose—none perfect—so we will borrow heavily from one that seems well suited to the purpose. It was devised by Richard Titmuss, a student of social policy in the British tra- dition of Beatrice Webb, Beveridge, and others.1 Titmuss was humble about this analytic scheme: “This represents little more than an elementary and partial structural map which can assist in the understanding of the welfare complex today.”2

110 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

� Prior contributions � Administrative rule � Private contracts � Professional discretion � Administrative discretion � Judicial decision � Means testing (needs minus assets and/or income) � Attachment to the workforce

Eligibility Rules Based on Prior Contributions Eligibility for many important social welfare benefits is established by rules about how much prior contributions have been made to the system that will pay the benefit later. A prominent example is benefits paid by the U.S. Social Security system: retirement in- come for workers and survivors (OASI), disability income for workers and dependents (SSDI), and payments for medical care services (Medicare) for both the disabled and the retired. The basic ideas behind the prior contribution method of establishing entitlement are the same principles that lie behind all private insurance schemes: (1) payment in ad- vance provides for the future and (2) protection against the economic consequences of personal disasters is best achieved by spreading the risk among a large group of people.

Exactly how much prior contribution is required varies with the age at which ben- efit is drawn and the type of benefit in question, but some prior contribution is always nec- essary. In general 40 quarters of coverage (minimally 10 years) is required, although for disability benefits it is 20 quarters in the 10 years prior to determination of disability, with special insured status for persons who are disabled before age 31. The prior contri- bution of which we are speaking comes from the worker and the employer and in match- ing amounts, calculated by a complicated formula and expressed as a percentage of workers’ wages written into legislation. As of 2004, 6.2 percent of wages (up to $87,900) were paid by both employee and employer (12.4 percent for self-employed individuals) as a contribution to the Social Security Trust Funds. An additional 1.45 percent of all wages were contributed by employee and employer (2.9 percent for self-employed indi- viduals) to the Medicare Trust Fund. It is from those trust funds that retirement, medicare, and disability benefits will later be paid. Note that some citizens receive bene- fits not because they made prior contributions but were dependents of those who did: spouses, children, and other legal dependents of contributing wage earners.

Unemployment Insurance (UI) represents another program in which eligibility rules are based on prior contributions. In order to be eligible for benefits, an unemployed worker must have worked in covered employment for a period of time specified in state law for eligibility (in most states, the first four out of the last five completed calendar quarters). The insurance is funded by a tax on wages paid by employers3 in covered em- ployment, and funds are credited to each state’s unemployment insurance trust fund (maintained by the federal government). Note that a person who has not yet worked (no prior contribution made on his or her behalf ) is not considered unemployed by UI eligi- bility standards and, thus, not eligible for unemployment cash benefits. A new work record must be established by an unemployed worker who has exhausted benefits (an av- erage of 26 weeks) before any additional unemployment cash benefits can be received.

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 111

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

Eligibility by Administrative Rule and Regulation Although eligibility rules for public social programs may be laid out in some detail in the law, seldom are they sufficiently detailed so that no administrative interpretations need be made. Thus, we will call these administrative rules made to clarify the law. This is an advantage to client/beneficiaries because it gives social workers and other human ser- vice staff members a means by which to administer the benefit or service program even- handedly and reliably, so that people similarly situated are given similar benefits. On the other hand, administrative rules restrict the freedom of staff members to use their dis- cretion, that is, to judge need for the benefit or service in individual circumstances. There are some eligibility rules that are almost fully spelled out in the law, and the food stamp program is probably the best example. Almost all of the details necessary to de- termine whether a citizen is entitled to food stamps are built into the law. The exact amount of assets, as well as income, is specified by family size in the text of the act, along with definitions of what constitutes a household. Consequently, no discretion is needed in determining whether (for example) a live-in friend of either sex should be included in determining household size. The administrative rules for the TANF (Temporary Assis- tance for Needy Families—the replacement for the old AFDC) program, on the other hand, are so numerous and concern so many different topics that they are bound into ponderous manuals. These administrative tomes not only include the state and federal statutes relevant to the program but also (mainly) they address how those laws are to be interpreted. One reason for the complexity of eligibility rules in the TANF program is they are means-tested, meaning that eligibility is established by a test of whether a per- son’s assets and income are greater than some official standard of need for a given fam- ily size. Apart from all the administrative rules that concern how to count assets and income, the TANF program has to be built on numerous administrative rules that tell the staff who sign the eligibility documents how to interpret the law. For example, should a child’s paper route income be counted as family income or should Aunt Lily’s inherited piano count as an asset? It is in the character of administrative rules that they can be modified over time; if they are devised by administrators, they can also be changed by administrators. Therefore, it is important to know whether a certain enti- tlement rule originates with judicial decision, administrative rule, or individual staff dis- cretion, for on that fact depends the probability for change—staff decisions certainly are changed more easily than are formal (“manualized”) rules or statutes. Furthermore, as you might imagine, the method, resources, and time used to effect changes differ for each rule source. Chapter 7 will discuss the details of administrative appeal hearings that are required by law for all social programs established under the Social Security Act and for many programs that receive federal funds.

Eligibility by Private Contract Strange as it may seem, it is possible to become entitled to a public benefit through the provisions of private contracts. The workers’ compensation system is constructed this way.4 In every state, employers are required to purchase insurance policies from private insurance companies (or a state insurance fund) to pay to workers for income and med-

112 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

ical costs to replace what is lost through work injury. There is nothing optional about the law, and employers are subject to substantial fines for noncompliance. In this case, the benefit form is a cash payment plus a voucher for medical expenses.

Another source of entitlement to public benefits in which private contracts are in- volved is purchase-of-service contracting (POSC). In the past decade, more and more welfare services—counseling, legal advocacy, special education, day care, and some trans- portation services (e.g., for the elderly and/or disabled)—are delivered by private con- tractors. In the case of purchased services of various kinds, the state actually pays the bill (or some of it) directly to the private purveyor of the contracted service. Because the state is the purchaser of services, the state can insert conditions into the contract concerning who can obtain the service, for how long, and under what circumstances.5 Not only state and federal governments subcontract for services, but private charitable organizations do so as well.6 For example, private hospitals contract out to private profit-making corpora- tions the operation of psychiatric units—Humana Inc. operates many such units nation- wide. Another example is a midwestern private social agency that operates a high-tech foster care program for its state. This program serves severely emotionally disturbed children who cannot be cared for in the ordinary family foster home setting. In both these examples, the entitlement rules embedded in the private contract determine who is eligible for services.

Eligibility by Professional Discretion One of the most widely used sources of entitlement is the professional discretion of indi- vidual practitioners. A common and concrete example is eligibility for medical benefits, which is always contingent on the discretion of the physician (or physician surrogate). Almost every licensed profession controls part of the entitlement to some social welfare benefit: dental care for TANF children is entitled in part by the judgment of dentists; legal advocacy for low-income people is entitled in part by the judgment of lawyers and judges; foster care for children is entitled in part by social workers. In each case, the en- titling professional whose judgment is necessary is presumed to have some expertise about the matter. Social work and human service practitioners must keep in mind that such discretion can be challenged in an administrative or judicial hearing, and when such discretion seems prejudicial to their clients, practitioners have a professional obli- gation to help their clients challenge it. Sometimes professional discretion is the leading evidence that severs children from parents, as in child physical abuse, sexual abuse, or neglect cases: physicians, clinical psychologists, and social workers are commonly used. No doubt those opinions are important and often accurate, but social workers and human service workers should be wary of a blanket assumption about the validity of those very difficult judgments.

Eligibility by Administrative Discretion Another kind of discretion that serves as a source of eligibility for social welfare benefits is administrative discretion. A common example of this is the policy in some states and

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 113

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

counties that allows a county welfare worker to distribute small amounts of cash and credits for food, housing, and utilities to poor people who apply. This kind of policy is characteristic of the General Assistance programs (less than half the states and U.S. ter- ritories have them). General Assistance is financed by the local government and is usually oriented toward short-term emergency budgets. The staff member must account for the funds only in the fiscal sense; administrative judgment is seldom called to account, and there is little systematic effort to document its accuracy. However, there are more im- portant examples of administrative discretion. It is indeed widespread and the source of such extensive power throughout modern public organizations that, according to Michael Lipsky and Michael Brown, there may be serious questions as to whether staff members at the lowest level or the chief executive actually controls the organizational op- erations.7,8 All general organizational policies and administrative rules must be inter- preted and applied to individual situations, so it is important to understand that such interpretation and applications necessarily involve significant personal judgment on the part of the staff member. For example, a state patrolman sees a person stopped at the side of a highway and diligently applying a newspaper to the bare rear end of a five-year-old child. In a hairbreadth, the patrolman is dutybound to make a serious decision about whether to stop and make inquiries. His decision is essentially administrative because it comes out of the role he fills as protector of persons. Later, he may have to make an even more serious decision. Was what he saw simply a child who had tried the parent’s pa- tience and was being disciplined within acceptable bounds? Or was it a cruel physical at- tack that will leave black-and-blue bruises or break the skin of a child too young to defend himself ? The discretion entailed here concerns interpretation of the state child abuse statute. Does this instance, and the data selected to report it, constitute an example of what the statute delineates? The statute will not reveal to the patrolman what rules he should use for its interpretation; it will not say how inflamed the bruises should be—or even whether black-and-blue rather than red bruises count. Nor will it always protect the officer from consequences if the parent claims illegal detainment or false arrest. The same situation is faced by the social worker and the physician while examining a hospital emergency room patient. Here, however, it is professional discretion that is being asked for. Their task is to render a professional opinion about the matter, and they are prepared by training and experience and specifically empowered by law to make that judgment. The difference between professional and administrative discretion is the source of au- thority of each: Professionals exercise discretion because of the authority of their profes- sional preparation and training, whereas administrators exercise discretion because they are appointed by their superiors to do so.

There are important examples of administrative discretion gone amok, so that social work and human service professionals should be aware that administrative dis- cretion—as important and humane as it can be—also can be used in ways that work to the detriment of their client’s/consumer’s welfare. Few cases are so flagrant as the mas- sive disentitlement of the chronically mentally ill from Social Security Disability ben- efits during the early 1980s’ Reagan administration. Although it is not a common case, it is useful to summarize briefly here to illustrate this point. At that time, the Social Se- curity Administration (SSA), ostensibly concerned about the rising costs of the Social Security Disability system, began a systematic effort to reduce approved benefit claims

114 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

and to terminate the benefits of the chronically mentally ill whom the SSA believed to be unable to prove their illness. Some 150,000 beneficiaries had their benefits canceled during those few years. Not only the mentally ill were disadvantaged; there are docu- mented cases of rejected applicants with severe, disabling cardiac conditions who died in the waiting rooms of Social Security offices.

The mechanisms by which this was accomplished included changes in the stan- dards used for mental disabilities and attempts to impose a quota on SSA hearing judges for benefit denials—judges were expected to hand down a constantly increasing number of benefit denials. If the quota was not met, they were subject to considerable harassment: reassignment or mandatory attendance at lengthy “educational seminars,” for example. The Association of Social Security Administrative Law judges appealed these measures on the grounds that they constituted unlawful interference in the fair-hearing appeals system established by the Social Security Act as independent of administrative authority. The association won on those grounds, after showing that indeed there were systematic, illegal, plainly political attempts to influence the outcomes of the fair-hearing system. Thousands of appeals of disability denials ensued, sizable proportions of which were suc- cessful. It is clear from the data available that disabled beneficiaries who persevered in challenging the administrative decisions of the SSA increased their chances of a favorable decision to nearly 80 percent. The higher the federal appeals court rendering the deci- sion, the more likely was a decision against the SSA.9 Only through the efforts of both legal advocates and social work and human service advocates were these reversals accom- plished. This example should give heart to practitioners that advocacy can succeed and, when conditions warrant, should be a part of their daily work.

Practitioners who advocate in these matters should understand that the most im- portant issue in these cases was whether the SSA had followed its own rules in denying dis- ability claims. That is the standard required by administrative law, and it is the most common ground for appeals. The law requires that when specific criteria are established for public benefits, the agency must adhere to them and apply them equitably among ap- plicants. If rules change, the changes must be made public and published in certain ways. In a very concrete way, administrative rules are the “rules of the road,” and justice re- quires that citizens know about them so that they can equitably pursue claims to which they may be entitled. On that account many (though not all) rules of due process apply: due notice, opportunity to know the grounds for denial, opportunity to present testi- mony in a fair hearing, and the like. (Some of these rules will be discussed in Chapter 7.)

Eligibility by Judicial Decision Judicial decisions are important sources of eligibility, virtually ruling applicants in or out of program benefits and services. (So important is it that all of Chapter 2 was devoted to the issue of the judiciary as a source of public policy.) After a program has been in oper- ation over a period of time, it is very likely that a contention will arise about whether the enabling legislation or whether an administrative rule or discretionary judgment was faithful to the spirit and intention of the law under which the program or policy was es- tablished. Appeals to the judiciary for clarification of the law are routine and in the end they can become as important as the legislation or administrative rules themselves.

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 115

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

Sometimes judicial rulings prevent administrative rules from excluding people from benefits. An example is the 1969 ruling of the U.S. Supreme Court on the constitution- ality of what were then called “residence requirements.” Under residency rules in effect at the time, citizens of a state had to establish permanent residency over a specified number of days, weeks, months, or years (usually one year for the old AFDC program) as a condition of entitlement. (Residency is an ancient eligibility requirement going as far back as the Elizabethan poor laws in England.) In 1969, the U.S. Supreme Court held that such requirements were unconstitutional infringements on citizens’ (labor’s) right of free movement between states.10 The same applies to TANF, the program that replaced AFDC.

Some judicial rulings operate not only to prevent exclusion from a program but also to positively assert eligibility where none existed before. In one of the most familiar, Brown v. Board of Education (1954),11 the U.S. Supreme Court ruled that all children are entitled to an equal opportunity for education, regardless of race. It was truly a landmark decision and marks the beginning of an era of efforts to establish civil rights, benefits, and services. Paternity determinations serve as a source of judicial entitlement to the child support payments by nonsupporting fathers. There is now available a clinical test, the human leukocyte antigen (HLA) test, which will rule out, with 97 percent accuracy, whether a given individual is the father of a particular child. Nearly two-thirds of all courts accept use of the HLA test in paternity cases, with full confidence in its results.12 And now even more accurate (99.9 percent) DNA genetic material matching tests are well accepted by the courts. When evidence is accepted and decisions are rendered about paternity, eligibility for child support payments from the adjudged father is established by court decision. Another kind of judicial ruling that represents a source of eligibility is somewhat unusual, but it occurs in the process of the judicial review of all children in temporary foster care and is now required by law in most states.13 Begun in New York State in 1971, that state’s court review statute provides that for eighteen months, the family court will review cases of all children in involuntary placement and determine whether they shall be discharged to their biological families, continued in foster care, freed for adoptive placement, or placed in an adoptive home. In essence, this is a decision by a judge as to whether a child is entitled to parental care, adoptive placement, or foster home services.

Eligibility by Means Testing One of the best-known and most widely used of all eligibility rules is a means test: In- come and assets are totaled to see whether they are less than some standard for what a person is believed to need. If the assets and/or income exceed this standard, then no benefit is given; if assets and/or income are less than that standard, the person is given a benefit in such an amount that total assets and income are equal to the standard. The central idea is that when assets and income are up to the standard, the person “ought” to have enough to meet his or her needs.

Despite its apparent simplicity, the whole idea turns out to have enormous com- plications. For example, there is the issue of what to consider as income. Some means tests concern both assets and wages (SSI and TANF), whereas others concern wages

116 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

alone (workers’ compensation). Benefit amounts and types of beneficiaries turn out to be very different depending on which version of the means test is used.14 Note the world of difference in terms of administrative complexity between income only versus income and assets with regard to the test. Income is almost always a matter of record (often public) and is often in the form of cash; therefore, it can be immediately valued. Assets are usually held privately and, because they are seldom a matter of record, de- termining their valuation is often problematic. Then there is the question of how to es- tablish a standard of need against which to cast income and/or assets. Even determining what minimum nutritional need is can be very controversial; the same goes for minimum need for housing, clothing, and on it goes. As a result of the diffi- culty of arriving at a consensus on these issues, a different standard of minimum need applies in almost every state. In 1991, the state standards of need for the old AFDC programs—by any standard never a princely sum—varied by as much as nearly 300 percent! For a family of three in the contiguous 48 states, the standards varied from the lowest, New Mexico ($317 per month), to the highest, Vermont ($1,160 per month).15 Means-tested eligibility procedures also vary with respect to whose income and assets are being considered when the means test is calculated. For some programs, focus is on individuals (e.g., SSI) or on families (e.g., the old AFDC now TANF); in others, it is on workers (workers’ compensation); in still others, it is on households “who purchase food together irrespective of blood relationship” (food stamps); and in still others, focus is on blood-related families (Title XX; Child-Support-Parent Location services). Table 6.1 summarizes some of the wide variability that can be found in concrete, se- lected social programs with respect to means-testing policy. Means are tested along three major dimensions: (1) type of resource counted (wages and/or assets), (2) concept underlying needs, and (3) beneficiary unit (child, household, worker, and such). These

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 117

TABLE 6.1 Variability in Means-Testing Procedures for Selected Social Programs

Type of $ Counted

Program

TANFa

Food Stamp

WIC

SSIa

Wages

Almost all

Almost all

Almost all

Wages above $65 a month and half the amount over $65

Assets

Almost all

Almost all

Almost all

Assets over $2,000 in general

Concept Underlying Idea of “Need”

Absolute minimum subsistence Nutritional adequacy; income less than 125% of poverty line Nutritional adequacy; Income at or less than 185% of poverty line Absolute minimum subsistence

Beneficiary Unit of Concern

Child

Household: those who buy food together

Pregnant or post- partum women, infants, and children up to age 5 Individual

aVaries by state, but these are the best general rules.

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

dimensions are useful in knowing what to look for in analyzing means tests as entitle- ment rules.

Establishing Attachment to the Workforce When social welfare programs are aimed at the primary workforce, that is, the working populace, it is of crucial importance to determine eligibility by a means that will qualify only those who are part of the workforce. This is done by setting a minimum to be con- tributed (via wage deduction at a workplace) that will entitle a person to benefit. Note that in many programs that require prior contribution for eligibility, not just any prior contribution will do; it must be a particular minimum amount, over a specific time, that counts for eligibility. The U.S. Unemployment Insurance (UI) program is a major ex- ample of this mechanism of entitlement. The goal of the UI program is to benefit those who have some significant work history—the program does not now intend, nor has it ever intended, to benefit those working part-time, only in casual employment, or only for insignificant wages. Although specific rules vary state by state, UI typically requires the worker to have received wages for at least six months and to have received at least $200 in wages during each prior three-month period. The purpose of such eligibility re- quirements is to establish that the worker has some significant attachment to the work- force. The Social Security Retirement and the Disability Insurance (DI) program has a similar policy built into prior contribution policies, but workers’ compensation is not concerned with limiting coverage to those with attachment to the workforce. If a worker is injured on the first day of the first job ever, and the employer carries workers’ compensation insurance, that worker is eligible for benefits appropriate to the injury as provided by law.

Criteria for Evaluating the Merit of Eligibility Rules

Fit with the Social Problem Analysis Correspondence between Eligibility Rules and the Target Specifications of the Social Problem Analysis. For a program or policy to be a coherent solution to a social prob- lem, those who receive the program’s benefits and/or services must be included within the group whom the social problem analysis identifies as having the problem. Recall from Chapter 1 on social problem analysis the necessity of social problem definition contain- ing “concrete observable signs by which the existence of the problem can be known.” Those concrete indicators, subtypes, and quantifications are main sources from which entitlement and eligibility rules must be drawn. Eligibility rules that don’t correspond to those indica- tors will off-target the program benefits and services. If poverty is defined as annual cash income less than $18,390 (threshold for 2002) for a family of four, then at least one of the eligibility rules must restrict the benefits of a cash assistance program to those with that level of income. If inability to attain a university-level education for their children is

118 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

defined as one problem for families with annual incomes less than $20,000, then the same stricture applies to eligibility for student aid. If the social problem of providing income support for the physically and mentally disabled is defined as applying to those with a ver- ified disability and proven inability to work for the next six months, then indeed the eli- gibility rules are about verifiable standards for determining disability and the inability to work for that period of time. The quantifications embedded in the social problem defi- nition are the basis for the target specifications that must be a part of well-formed goals and objectives and it is to those that eligibility rules must be relevant.

A central question is whether the entitlement rules expand or reduce the agency’s ability to bring its program to those who are affected by the social problem. And recall that definitions in the social problem analysis can always be changed to widen target spec- ifications and, thus, provide an improved answer to that question. Note, however, that doing so may have serious consequences. First, an enlarged target specification of those who have the problem might create large increases in program costs. Second, a narrower target specification might very well change the causal factors on which a program design should focus and, thus, the whole program design. The reader should not interpret this as advice against changing a social problem viewpoint; rather, it is to alert the unwary. With experience, viewpoints on social problems become more sophisticated and hope- fully better program designs emerge.

Correspondence between the Eligibility Rules and the Ideology of the Social Prob- lem Analysis. Eligibility rules do more than just reflect target specifications; they also reflect general ideological positions that underlie or are associated with the viewpoints from which a social problem is defined. An example is commitment to the work ethic, an idea that refers to the common belief that work is inherently virtuous and that the virtue of a citizen is related to work effort and work product. English poor laws required work tests as a condition for eligibility; that is, one way a person proved he or she was poor was to be willing to accept placement in a nineteenth-century workhouse. The modern U.S. equivalent is the requirement that unemployed food stamp recipients be registered for work referrals at their local state employment agency. That requirement is itself a condition of eligibility and, thus, an eligibility rule. It reflects an ideological commitment to the idea that citizens should expect to work for their own bread and that, if they don’t, they should have to show that no work is available or that they are unable to do the work that is available. The “relatives-responsibility” policy is another instance of eligibility ideology. England’s Elizabethan poor laws, as well as U.S. social policy through about the 1950s, provided relief for the poor but were constrained by the common ideological commitment to the idea that families were always primarily responsible for their members. Thus, the underlying practical understanding of the social problem of poverty was that three descending generations in the family group— grandparents, their adult children, and their children’s children—must be poor before any individual member was deemed poor. The eligibility rules for the expenditure of public funds for the poor reflected that ideology: Parents were financially responsible for the relief of the poverty of their children, and children were responsible for the relief of the poverty of their own parents.

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 119

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

Good entitlement rules for personal social services must have an ideological fit with the relevant social problem analysis. For example, in the field of mental health, there is an interesting split between ideological positions: One implies that severe and chronic psychosis is a problem of greatest concern, and the other implies that prevention of mental health problems is the premier priority. Thus, for the eligibility rules to be consistent with ideology here, the former would give priority to those with psychotic be- haviors, whereas the latter would give priority to those considered to have high potential for the development of mental health problems (however those problems are defined). Of course, that applies only under conditions when resources are insufficient to provide services for both, but that is almost always the case for social policy and program systems. For a different example, consider child protective services. If the ideological position im- plies that children should never be considered to be a cause of their sexual abuse by an adult, then the rule that entitles them to protection by state intervention should also en- title them to remain in their own homes while the adult perpetrator is required to leave. That is not universally followed as a matter of public policy on child protection.

Criteria Specific to Eligibility Rules Stigmatization. Side effects of some eligibility rules may have such serious conse- quences that they outweigh benefits received. Some argue that these side effects are intentional. Two of the most widely discussed side effects of eligibility rules are stigma- tization and alienation. To be stigmatized means to be marked as having lesser value, to bear the burden of public disapproval. There are many meanings of alienation, but here the term refers to the subjective sense of being estranged from the mainstream of the society in which one dwells. How is it that an eligibility rule or mechanism can produce such strong negative social effects? Both alienation and stigmatization are serious side effects that are believed to be associated with many consequences (suicide, social deviance, tendency toward serious crime, and chemical addiction).

To understand how eligibility rules can produce these strong side effects, think for a moment of what is entailed in an application for a means-tested public assistance pro- gram like SSI or TANF. (Recall that TANF offers both cash benefits for all citizens whose income is less than some “official” needs standard and personal social services de- signed to do such things as increase parenting effectiveness, offer children foster care, and help single parents get jobs.) Basically, the application requires a person to lay bare the details of his or her financial and work history in order to document income and as- sets. Thus, it requires a person to reveal all details about when jobs were left (for what- ever reason), when spouses or children were abandoned—without regard for unflattering details. The application requires a person to say some or all of the following: “I’m broke. I can’t keep a job. I left my last job because I had to go to jail (or the mental hospital). I couldn’t be enough of a success in school to get the credentials so people would hire me. My parents, relatives, and spouse have all left me and don’t care enough about me to help out.” Revealing such details to a stranger cannot help but make the strongest constitution quiver in the telling. The ordeal is self-stigmatizing because the teller can no longer hide what may be humiliating facts—at least one other person knows. The more a person be- lieves he or she is regarded negatively, the more likely he or she will accept the stigma as

120 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

real. The stigmatization that appears to result from eligibility rules associated with the TANF or the old AFDC program is widely discussed in the literature.16,17,18 The Petti- grew article contains an excellent survey of the studies of labeling and stigmatization of welfare recipients.

It is important to observe that not all entitlement rules are associated with this kind of stigmatization. Few elderly feel stigmatized by the application process guided by the eligibility rules of the Social Security Retirement or Disability program. Nor do people feel stigmatized by the means tests involved in the application for student loans (BEOG [Basic Economic Opportunity Grant] or NSDL [National Student Direct Loan]) or, for that matter, the means test inherent in payment of income tax. Two factors distinguish a means test involving an application for TANF from a means test involving a loan for at- tending college or university and illustrate how stigmatization occurs. (1) The reason for application for TANF (say) is most likely to be something that must be apologized for or explained. In contrast, the reason for applying for a BEOG or NSDL loan is almost never the occasion for an apology or explanation; to the contrary, it is likely to be occasion for congratulation or recognition that a person is about to embark on a path of high social regard—going to college. The same applies to the instance of the means test entailed in paying taxes: The very fact that a person struggles over filling out forms and takes a long time at it suggests a person who has considerable assets and income.

The worst consequence of the means tests involved in a BEOG loan application is that a person would have to look elsewhere for funds or delay going to school for a year. Contrariwise, the best consequence of the means test for TANF or the old AFDC is that a person will receive a poverty level income and medical card.

Some years ago, George Hoshino suggested that, in a phrase of Gilbert and Specht, the means test did not have to be mean-spirited.19 Hoshino suggests that one of the main reasons for the negative effects of the means test as a way of determining eli- gibility is that it places great stress on determining unique individual needs when all that is really required is to determine average need for categories of family size, age, and so on. Verification of assets is a process filled with arbitrary and specious judgments of the market value of mundane goods. As any experienced social worker knows, the adminis- trative cost of such determinations far outweighs the relatively small misplaced benefit that might be given were the verification of the value of highly personal assets simply ig- nored. It was a considerable step forward when the means test for the food stamp pro- gram cleverly avoided these pitfalls and determined need on exactly the basis Hoshino suggested in earlier years—that of some concept of average need (for food, in this case) and by “average” deductions for major items like cars, houses, and insurance policies.20 Although there may be an applicant who has a house full of expensive new furniture and who might not declare it when applying for food stamps, this would probably not char- acterize the majority of food stamp applicants. The administrative cost of tracking down that odd exception outweighs any saving that might result.

Off-Targeted Benefits. Another criterion for judging eligibility rules and their asso- ciated procedures is the extent to which benefits are directed to population groups who are not the main object of the program. One example from the early 1980s concerned the NSDL funds for college and university students in the United States. NSDL loan

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 121

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

funds were very attractive to students in those years because their interest rates were far below the existing market rate of around 7 percent. The difference between the interest rate on the loan (some as low as 3 percent) and prevailing high interest rates in 1981 on such things as long-term savings accounts (12 to 17 percent) was so much that some stu- dents who already had sufficient school funds took out an NSDL loan simply to make a little money by banking it at a higher interest rate. Every year the cost of the interest on the loan was $300 (3 percent of $10,000), whereas the long-term savings account yielded perhaps as much as $1,700 (17 percent of $10,000). The net profit on this no- effort enterprise would have been the difference between the dollar yield of the two interest rates, that is, $1,700 – $300, or $1,400 total. It would be hard to think of a way to earn more than $100 a month more easily.

There aren’t many examples of porous eligibility rules as outrageous as that and, in fact, that gaping wound in the design of this eligibility rule was closed by raising in- terest rates on loans to competitive levels. But eligibility rules such as this should be judged negatively since the off-targeting is significant and has no obvious impact on the social problem. In fact, it represents off-targeting of the worst kind in that it takes away money for income transfers from those who are most clearly in need of them, and the profit for those who took advantage of the opportunity was and probably still is being paid for by taxpayers like you and me. It is important, nevertheless, to notice that some instances of off-targeting are intentional and not always a bad thing.

In fact, some social policies are operationalized in ways that purposely produce “seepage” of benefits to nonmembers of the target group. Perhaps the best example of off-targeting intended to produce positive results is the Social Security Retirement pro- gram (OASI). As noted earlier, the program is nonstigmatizing because its designer en- sured that nearly the whole population would receive benefits (because nearly all were entitled since they contributed as workers). In the usual case, social programs that are universal cannot stigmatize or alienate since it joins citizens to their peers rather than identifying them as “apart” or “less worthy.” In fact, the actual cost of “destigmatizing” this (or any) program is precisely the cost of the off-targeting. Though the total cost is not relatively large, OASI does off-target benefits; those over normal retirement age (65 plus 4 months in 2004) can earn unlimited income and still receive full retirement benefits.

There are other ways to avoid stigmatization besides universalizing eligibility. Here is an example of one that failed becoming law in the 1970s by only a single vote. Its virtues are that it does avoid stigmatization, has simple eligibility rules, can be adminis- tered without constructing yet another bureaucracy, and is probably more fair than other alternatives because most administrative discretion is removed from the eligibility pro- cess. This proposal would abolish all existing cash and cash-equivalent programs (TANF, food stamp, SSI, and UI programs) and replace them with a cash benefit that will provide a minimum subsistence standard of living for those who, for whatever reason, do not have a minimum amount of income and/or assets. The program would use the regular IRS ad- ministrative procedure for collecting income tax as a means of distributing benefits to the poor, a system generally called a negative income tax (NIT). Originally called the Fam- ily Assistance Program (FAP), the program was first sponsored in Congress in 1974 by the conservative Republican Nixon administration (called “Nixon’s Good Deed” by

122 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

some). The basic idea is that every three months, people would file an income tax state- ment. If their total income and assets were less than some designated poverty line, they would receive a monthly amount over the next three months that, when added to their past three-months’ income, would equal the poverty line for their household size. When income exceeded the poverty line, that household would incur a tax liability and be re- quired to pay the government additional tax dollars. This scheme was neither clearly uni- versal nor clearly selective. In fact, the system carefully selects and benefits most those in greatest need, even though all citizens can potentially benefit and the system is nonstig- matizing in that there is almost no public revelation of benefit receipt. The Nixon ad- ministration scheme was automatic in that it was operated by the Internal Revenue Service and had a built-in work-incentive feature. Using 1997 figures, if there was no earned income, the family would get a standard base payment, a fixed-dollar amount. Families could work and still keep part of the base payment. For example, the first $8,000 of earned income was excluded from consideration and doesn’t affect the base payment at all. The next $8,000 of earned income, however, reduces the base payment by $1,000 because for each earned dollar above $8,000, the base payment is reduced by 50 cents. The base payment is reduced even more for earned income above $16,000—for each earned dollar in this range, the base payment is reduced by 75 cents. More and more of the base payment is taken away as earnings climb, but each dollar earned up to $18,000 will still continue to add something to the family coffers and thus continue a work incen- tive. A point is reached where finally the base payment is totally wiped out by the reduc- tions for earned income.

In contrast to the negative income tax program, which targets benefits heavily on those presumed to be most in need, there is a program called Children’s Allowances. It is semiuniversal since it benefits every household having children irrespective of their level of need. Some form of Children’s Allowances operates in nearly every country in the Western industrial world except the United States (including countries such as Germany, France, and Ireland). Canada has had a Children’s Allowance since the 1930s and Great Britain since 1945. Although benefits are usually small, they are a significant addition to family finances for poor people. Proponents often argue that it targets benefits directly on children and their needs in ways that other (more or less) universal programs don’t in that benefits are paid directly to mothers.21 The German Children’s Allowance type ben- efits in 1997 were DM 220 per month for the first child (and an exchange rate around U.S. $1.80 per German deutschmark). With respect to off-targeting of benefits, the Children’s Allowance strategy involves considerable seepage, depending on how one per- ceives the program objective. If the objective is to supplement incomes, the seepage is very large—more nonpoor than poor will receive the benefit. If the objective is to in- crease the standard of living of all families with children, whatever their present income level, then there is probably much less off-targeting. Finally, note that the NIT idea must always involve some kind of means and asset test. It is the presence of this feature in NIT and the lack of it in Children’s Allowance that always generates controversy over whether there is strong off-targeting in any Children’s Allowance scheme. An additional point of vulnerability for Children’s Allowance proposals is that the benefits must be very low per family or else the cost is overwhelming. Simple arithmetic will show that a payment of

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 123

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

$100 per week per child in a nation with 50 million children would cost $260 billion per year—more than the cost of the U.S. defense budget in peacetime. Though child advo- cates would not find that unseemly, no doubt it would be an unacceptable division of the pie to the advocates for other constituent groups, like the American Association of Re- tired Persons (AARP) or AIDS advocates. It is important to note that Children’s Al- lowance schemes are not inherently bad proposals, but they are neither cheap nor insignificant in that their redistributive qualities would require a radically different na- tional consensus in the United States about the importance of children and the justice of large-scale income redistribution programs.

Trade-Offs in Evaluating Eligibility Rules

So, if off-targeting has both good and bad effects, how is the practical public policy ana- lyst to judge between them? It is an important question and doesn’t yield to a simple an- swer. Let us use the concept of trade-off to characterize what we will be considering here. It is not an exotic idea; rather it is one we all use in working out our everyday lives. We all learn that getting one good thing sometimes means having to endure some bad things. Usually, we choose so that the good outweighs the bad—but not always: If I have only enough money to buy badly needed new household appliances—perhaps a refriger- ator, a washer-dryer, and a stove, but I also need a better used car, the choice is not so sim- ple. Here is the trade-off: If I buy the appliances, I buy freedom from having to go to the laundromat, enjoyment of a new stove, ability to store food longer and, therefore, shop less often. In return, I have to endure an unreliable car that spends weeks in the repair shop, which forces me to depend on friends or public transportation. So, how does the ordinary person living an ordinary life make that decision? The answer ultimately de- pends on the relationship to what one values and disvalues—in a word, preferences. Now let us consider what those value/preferences might be and how a person might go about making decisions based on them.

The most obvious decision rule rests on a preference for getting the best value for the money. Taking into consideration only the most obvious costs and savings, one might add up the costs and savings of choosing (in this example) to buy appliances: Suppose the total cost is about $2,000 and from that I can subtract the savings from avoiding laundromat costs (say, $200 a year). But I must also add in the expected cost for car repairs (about $800 a year) and the extra public transportation costs (about $600). If a better used car will cost about $6,000, then (using out-of-pocket costs as a standard) I would be about $2,400 better off to buy the appliances and forgo the better used car: $2,400 = $6,000 – $2,000 – $200 – $800 – $600. A notable feature of the best- value-for-money standard is that it can depend on whether I want to make it work for the long or short term (these figures only take into account the first year). With every passing year, I lose another $1,400 in transportation costs. Simple arithmetic shows that in three years, I am $1,800 the net loser. Furthermore, when my appliances begin to need repair, I will go deeper into the hole. Thus, in the long run, I would be better off, dollarwise, in choosing the better used car; but in the short run, I am better off choosing the appliances. Still that doesn’t take into account those preferences that are

124 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

more difficult on which to put a dollar value—my preference for saving time and trou- ble by having a dependable automobile. Best-value-for-money is an obvious standard for choice, but it won’t sort out whether I would prefer the convenience of a reliable car compared with the convenience of new appliances. Choosing among trade-offs that involve social programs is no different in principle—whereas costs are important, they are not always (and in all ways) the crucial issue.

When we think about public benefits, trade-offs are ultimately cost and value issues—is the public interest better served by exercising a preference for avoiding stigma and increasing costs (as in Children’s Allowance or Guaranteed Income programs) or by exercising a preference for lower costs (e.g., in which case, the monies saved can be spent on reducing other social problems) at the expense of creating stigma for beneficiaries (as in the means-tested TANF program)? There are many other examples of trade-offs; in fact, almost all policy and program choices involve trade-offs of one kind or another, and because they ultimately are settled on value/preference grounds, it is one additional rea- son a value-critical perspective is essential for the practitioner. Two concepts are used to examine some types of trade-offs: vertical equity and horizontal equity. Vertical equity refers to the extent to which resources are allocated to those with the most severe need— the kind of close target efficiency spoken of earlier in this and other chapters. Horizontal equity refers to the extent to which resources are allocated to all those in need.22 The point here is that, given scarce resources, there is almost always a trade-off between ver- tical equity and horizontal equity—the difficult (sometimes tragic) choice between meet- ing a little of the need of all those afflicted and adequately meeting the need of those in most serious difficulties. There is no consensus on the value/principles on which that de- cision can or should be made. Other important criteria for evaluating eligibility rules in- volve trade-offs and are discussed in what follows.

Overwhelming Costs, Overutilization, and Underutilization

Bad eligibility rules can create severe overutilization and, thus, serious cost overruns. Medicare is a leading (and interesting) example in that cost containment is a major problem for Medicare. The entitlement for Medicare is universal for U.S. citizens age 65 and over who are entitled to OASI benefits (40 quarters of insurance coverage, min- imum of 10 years) and for those with fewer quarters of coverage if they pay Part A pre- miums. Medical care for the aging U.S. population is an expensive business because they need more care and there is a rising proportion of the elderly in the population. Not only that, but both absolute and relative costs of medical care have risen exorbi- tantly over the past decade as technology improves and corporate pharmacy profit increases.

For example, such procedures as bypass operations for heart disease are now rou- tine. Kidney dialysis is at present included as an acceptable medical procedure for Medicaid beneficiaries. The problem is even more complex because the long-term health benefits for both are debatable—kidney dialysis will extend life about ten years and heart bypass procedures last on average about five years before death or before

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 125

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

having to be repeated. The debate is about whether adding zero to ten years onto the life of a post–65-year-old citizen is the best expenditure in view of the pressing health needs of children and working adults in the United States. Recall that all policy systems operate under a condition of finite resources, so that every dollar spent for kidney dial- ysis and heart bypass procedures is a dollar that cannot be spent on disease prevention for children: The United States still does not make routine immunizations for small- pox, diphtheria, and typhoid available to its children, even though many Third World countries do so. The value-critical policy analyst must search for the value stance from which this policy choice is made. Universal entitlement to medical procedures is filled with great ethical issues, vexing and ambiguous in the extreme. As a nation, we seem unable to face these issues squarely. The consequence is that when it comes to the choice of which medical procedures will be universally provided for the people, it may be determined by which drug or medical supply or hospital corporation lobbies Con- gress most persuasively. There is nothing inherently wrong with profit making but profit making is an imprecise tool, often blunt and cruel, when it comes to determin- ing choices of who lives and dies. Note that priorities for scarce research and care money are now embedded in the Medicare/Medicaid policy system. But where should the first priority lie: AIDS, Alzheimer’s, developmental disabilities, chronic mental ill- ness, or neonatal intensive care—newborns born at less than one pound, ten ounces? Here are some quotes about the nature of such neonatal care:

� About half will live but three-quarters of those will have serious neurological damage.

� All stops are pulled out, . . . we are doing virtually everything that can be done to keep these children alive.

� One national study in 1988 revealed that a third of neonatologists said they had changed their medical practice and were treating babies they thought had nothing to gain and a lot to lose from aggressive medical care.

� After three months in the hospital, which cost close to a million dollars, . . . the triplets came home. All had grade 4 brain-bleeds making it virtually certain that their brains were damaged. (New York Times, September 30, 1991, p. A1 and April 8, 2003, p. D5 and D8)

These data are an example of an overwhelming and uncontrolled cost burden on the medical system in general. Because recent estimates suggest that such care is extremely expensive and that the babies’ chances for surviving into adulthood as fully functional adults are quite slim, on what value/preferences shall such choices be based?

The basic value problem is highlighted because public policy has avoided the basic value issue. It is not a scientific decision. Clearly, some public medical benefits can cre- ate their own unlimited demand. When private physicians control treatment and when physicians fear legal suit for not providing maximum care and treatment, neither the pa- tient nor the government is in a position to curb the use of modern technology (al- though health insurers seem to do so!). One of the reasons that Medicare and Medicaid costs have risen so rapidly is that nearly all the aged can qualify for benefits under one program or another.23 But it is quite clear that the issue is not simply who is entitled, but

126 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

also for what benefit. The basic question U.S. social policy has not yet answered is: “On what value premises shall medical care, indeed life and death, be rationed?”

Some eligibility rules create underutilization; that is, program benefits are not taken up by the people for whom they are intended. There are several important exam- ples of underutilization in the United States, some more serious than others. One that perhaps is less serious is the low take-up rate of the Low Income Energy Assistance Pro- gram (LIEAP), a federally financed program initiated by the Carter administration with the object of subsidizing increased energy costs among the low-income population. The entitlement program rules rested heavily on a reasonably flexible and nonstigmatizing income test, but the public was poorly informed about exactly how much benefit was possible. In fact, in many states the LIEAP benefit was certainly more than a trivial amount—sometimes covering an average of 150 percent of the total cold-weather energy costs of the average household.

One more serious example of underutilization of a public benefit program is the SSI program offering, which gives cash income maintenance benefits for which a means- tested entitlement rule is in place. The take-up rates for this program run between 55 and 60 percent.24 Although it is not entirely clear that this underutilization is totally an eligi- bility rule problem, there are suspicious signs: SSI is a program for which both the aged and the disabled qualify and, for complex reasons, much of the underutilization concerns the disabled. Note first that application cannot be made until a year after the disable- ment occurred. The eligibility rules have a very complicated procedure, which appears to qualify only those completely and totally disabled for long periods; also, it was origi- nally designed for physical, not mental, disabilities. For example, it ordinarily disqualifies those who can work only some of the time, which, of course, applies especially to those disabled for reasons of mental illnesses like psychosis, bipolar disorder, or schizophrenia. Reestablishing benefits takes as long and is as complex and demanding as the original ap- plication—seldom less than several months and often more than a year. Also, it is well known among the disabled population that the outcome of application is unpredictable at best. It is reasonable to expect that rational people will hesitate before committing themselves to pursuing such benefits, especially when they involve heavy expenses in time, long-term doggedness in documenting medical treatment and diagnosis and not trivial monetary sums for a population that has no discretionary income. The mentally ill are not the largest proportion of the homeless, but they are a significant group. Home- lessness, sometimes a consequence of long delays in gaining eligibility, creates public costs, an illustration of the point that underutilization doesn’t automatically create cost savings in tax dollars, for example, jail stays, emergency medical care, and street violence.

Overutilization and underutilization criteria have special applications in the per- sonal social services. A leading example of eligibility rules that create unintentional un- derutilization are programs for older Caucasian and minority children who, otherwise available for permanent adoption, nevertheless remained in foster care for lack of parent/ applicants. Until the 1980s, child-placing agencies that had adoptable minority children in their custody in fact contributed to their problems by holding to certain eligibility requirements: for example, requiring separate bedrooms for children, typical middle- class income levels, a nonworking adoptive mother for infants, and/or formal in-office

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 127

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

interviews held in a distinctively white middle-class office environment. Such eligibility rules actively disentitled working-class and minority and ethnic parent/applicants from consideration in two ways. First, working-class and minority status can mean absence of average incomes and many working mothers; therefore, if an eligibility rule is based on average income or the presence of a nonworking mother, it disentitles many minority and working-class applicants and single parents except those with incomes above the av- erage compared with their own racial group. Such a rule offends against the equity crite- rion because it systematically disentitles based on social class status that has nothing to do with any feature of the social problem the program is intended to solve.

Second, some people from minority and ethnic groups have limited experience in making formal applications—in fact, the whole idea of applying for children and having their parental and social competence judged is an experience outside the realm of their cultural expectations. For most such groups, not only is taking responsibility for others’ children not unusual—whether children from their own families or otherwise—but also it is usually negotiated in face-to-face encounters and in familiar surroundings with lit- tle or no expectation that motives are under scrutiny. Whereas there is good reason for adoption agencies to be concerned about applicants’ motives for parenting, any good el- igibility rule will take cultural practices into account and not run hard against them. Agencies have dealt with this by featuring initial contacts in the applicants’ home, church (or other religious site), or lodge; sometimes these contacts have been initiated by friends or acquaintances. In that way, the whole encounter in adopting a child occurs in the context of a familiar social network where the agency staff member, although a stranger, is at least vouched for by someone already trusted.

Here is another extreme example, this one from the eligibility rules apparently in use by some public Central American child-placing agencies: Part of the application pro- cess involves psychological testing via such measures as the Minnesota Multiphasic Per- sonality Inventory (MMPI). Firsthand interview data suggest that the results of such tests have serious implications for adoption placement decisions.25 In fact, a requirement for MMPI screening is listed in the administrative documents of one Central American pub- lic adoption agency. Screening is an issue because local Central American adoption agen- cies commonly have in their custody a number of local children of color (e.g., indigenous native people and Caribbean blacks), including infants and preschoolers, children whose only hope for kinship associations of their own are non-Indian (most frequently Latino) families. Those agencies report that adoption by local non-Indian citizens is uncommon. Psychological screening of this kind as an eligibility rule creates underutilization because it is so alien to the applicants’ experience (leaving aside the cogent argument about its cultural transferability to a Hispanic culture or the doubtfulness of its ability to predict good parents or even to screen out the mentally distressed). That alien nature of psycho- logical testing discourages scarce applicants. And news about agency experiences spreads widely in minority communities by word of mouth, especially among potential adoptive applicants, which further discourages applications.

These considerations probably apply equally well to other personal social ser- vices such as mental health and counseling services where their delivery takes place in formal clinics and office buildings. That is one reason why, years ago, street workers and

128 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

outreach programs were invented—to create access to services when they could be en- countered in the everyday and familiar lives of the people for whom they were intended, rather than limiting formal application to unfamiliar, hard-to-get-to office settings. In small communities, it can be stigmatizing to enter a building known to be the commu- nity mental health center. Certainly, that applies to more controversial birth control and/or abortion locations.

Clearly, eligibility rules for insurance-covered services can be sources of under- and overutilization. Insurance companies find eligibility rules for mental health services to be problematic, not least because for mental and emotional illnesses or problems, the need for treatment and what constitutes adequate treatment are debatable in the field— debatable in a way that appendicitis or diabetes or athlete’s foot for that matter is not. Insurance companies need to ask where is the clear and definable point at which a pa- tient is not helped by further office visits for the purpose of increasing self-awareness, self-concept, or personal insight. As might be expected, insurers don’t find underuti- lization a problem, but if eligibility rules for insurance coverage aren’t on target, under- utilization can lead to tragic results.

The general solution insurance companies have resorted to is to place arbitrary dollar or time limits on mental health and/or counseling service—$1,000 a year for out- patient services or fifteen days of inpatient services is not an unusual standard. Costs are an issue and insurance companies have a telling point, one that the psychotherapy in- dustry has yet to answer, coincidentally, because the insurance principle requires ability to forecast use (via some actuarial design) in constructing rate schedules for prepaid in- surance premiums for health coverage.

On the other hand, the health insurance industry is not noted for its leadership in this regard either; between the two, sizable underutilization and overutilization con- tinue because of the arbitrary nature of the caps placed on mental health and counseling services.

Work Disincentives, Incentives, and Eligibility Rules

Almost all agree that eligibility rules should be evaluated against their potential for work disincentives. The argument about whether cash benefits in social welfare income main- tenance programs cause people to choose benefits over work for wages is several hundred years old. A major concern during the Speenhamland experiment in England in 1795, it is presently a concern of U.S. economists and politicians as they attempt to reduce wel- fare costs.26 Both economic theory and common sense would seem to indicate that cash benefits from the public treasury could strongly reduce work effort on the part of the or- dinary citizen—why would people work if they didn’t have to? Both the question and the answer are complex issues that for years have eluded practical resolution and scientific experiment. Who, after all, would give money to someone just to see whether he or she would continue to work, work less, or not work at all?

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 129

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

In fact, that experiment has taken place. The economic theory behind the income guarantee experiments runs a little like this:

One person can view time as being divided among three activities: working for wages, working at home, and enjoying leisure time, depending on relative opportunities and rewards. The reward for market work is money income, which ultimately is used to buy goods and services. One of the goods that people may “purchase” is leisure, but each person pays a different price, one equal to his or her wage rate. Economists theorize that the amount of nonworking time “bought” by a person depends on two factors: (1) the wages that must be foregone and (2) the amount of nonwage income that is avail- able to the person. As a person’s wages rise, leisure (non-work) time becomes more expensive. So, besides the question of whether public benefits cause less work effort, two other questions arise: (1) whether if there is less work effort, it is due to the fact that leisure time becomes more expensive as income rises, causing people to regard increas- ing leisure costs as “expenditures” or (2) whether with more income, people value increased income less and are willing to substitute leisure for work.

These questions have vexed discussions of welfare reform for many years. The U.S. Of- fice of Economic Opportunity (OEO) undertook a series of large-scale experiments be- ginning in New Jersey in 1968 and extended in the 1970s in Iowa, North Carolina, Colorado, and Washington state. These experiments, all long term (five years for the most part), were carefully designed and instrumented, and strong attempts were made— not always successfully—to insulate them from external contaminating influences. It wasn’t a perfect experiment but then no experiment in the real world (outside a labora- tory) ever is. And real-world research has arguable advantages over small-scale simula- tions that, for example, might ask people to imagine their response to questions if they were low income and offered the choice between work or staying at home. We will focus here on the Seattle-Denver Income Maintenance Experiment (SIMDIME) because it was the last in this series and provides the best data. It had the largest sample among all the experiments, in that it included around 5,000 one- and two-parent families of black, white, and Hispanic ethnic origin. In SIMDIME, the families were assigned either to one of several experimental groups receiving cash assistance payments at various levels or to one of a control group of families who received no experimental payments but continued to receive whatever benefits they were eligible for under current governmental pro- grams. Hours of work of experimental families were compared with hours of work of the control-group families during the course of the experiment. First, the results showed no significant difference between responses by racial or ethnic background, holding all other characteristics constant. Next, some decrease in work effort was shown when people got an income guarantee, but the difference was small. The report has this to say about the results:27

The results for husbands show, for example, that if a family’s preprogram annual income was $4,000, a cash benefit that raised income by $1000 would cause the husband to work about an hour less per week . . . the effects on a wife in a family with the same income would cause her to work two hours per week less. . . . However, since wives usu- ally have lower wage rates than their husbands, a given benefit reduction rate usually would have a smaller dollar effect on the wife’s net wage than on the husband’s.

130 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

Table 6.2 presents the results of the effect of the income guarantee on work effort for all four work-incentive experiments. Although some of the wives’ reduction in work hours appears large, observe that the authors interpret this as a relatively small-scale response. “Since wives in poor families usually work relatively few hours to begin with, the large percentage change in their labor supply effort amounts to relatively small numbers of hours.”28 The net result, as stated before, is that a $1,000 increase in the family’s income “causes” the wife in a poor family to work only two hours less per week. So what should be our conclusion about the work disincentives of social welfare programs offering cash benefits like this one? A conservative conclusion, faithful to the facts the experiment re- veals, would be that the effect is there but is very slight, probably insignificant to most people. The experimenters believe that the results from all four of the experiments show a “striking similarity,” particularly considering that the experiments provided different sets of benefit levels and benefit reduction rates, that they took place in states with widely differing tax and transfer systems, and that different criteria were used to select the four samples.29

One result of the guaranteed income experiment should not go unnoticed: There was a marked increase in the proportion of marriage dissolution under the impact of an income guarantee. It was about the same for whites as for blacks but noticeably greater for Latinos.30 One important consequence here is that if a national income guarantee program were put in effect, the proportion of female-headed, single-parent families might increase substantially, particularly for whites and Latinos. Remarriage rates for blacks under conditions of income guarantee is sufficiently high, so it would not affect the proportion of single-parent families among that subpopulation.

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 131

TABLE 6.2 Estimated Percentage Reductions in Work Hours in Four Income Maintenance Experiments

Control/Experimental Group Differences as a Percent Control Meana

Husbands Wives Total Female heads

New Jersey (White Only)

6% 31 13 b

Rural Wage Earners

1% 27 13 b

Gary, Indiana

7% 17 8 2

Seattle- Denver

6% 17

9 12

aThese estimates are weighted averages of the response in hours worked by different study groups. Because of the technical problems in estimating the response of black and Spanish-speaking groups in the New Jer- sey experiment, estimates reported here for New Jersey are for whites only. Recent reanalysis of the New Jer- sey data provides evidence that the response of these groups is similar to that of whites. Total responses (and base hours) include only husbands and wives in the Gary and Seattle-Denver experiments; in the other exper- iments they include other family members as well. bNone included in the experiment.

Source: The Seattle-Denver Income Maintenance Experiment, Midexperiment Results and a Generalization to the Natural Population (Stanford, CA: Stanford Research Institute and Mathematical Policy Research, 1978), Table 2, p. 64. Reprinted by permission.

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

Subsequent to policy debates and the income guarantee experiment in the 1970s, interest in this major social policy shift for income maintenance programming dimin- ished. Instead, welfare reform has been focused more on adding work incentives (or simply requiring work as an eligibility requirement) for existing social programs. A prime example is the Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (PRWORA). A central theme of PRWORA is the goal of moving wel- fare client/consumers into the labor market. A popular slogan associated with this wel- fare reform policy is, in fact, “welfare to workfare.” PRWORA added a provision that working-age adult food stamp beneficiaries are required to sign up for work in order to receive benefits. Particularly clear is the message of the TANF program. TANF care- takers are given a time limit (five years or less) of cash assistance, and are encouraged or required to meet established work experience standards. A brochure from one state wel- fare department sends the message in the following:

You can earn your own money, choose what you want to do and take charge of your own life. You may never need to come in here again—but we’re here to help you if you do. We will provide day care for your children under age 13 while you are in approved TANF work activities. We may also pay you a Participation Allowance to help with transportation and other expenses. You may be able to get an Earned Income Tax Credit from the federal government if you are working or have worked recently. This money will not count against your TANF grant. After you have a job and are earning enough that your TANF case is closed, you may continue to receive help with child care, med- ical assistance and food stamps.31

It remains to be seen, however, just how successful so-called incentives to work will be for individuals who have been severely disadvantaged by lack of an adequate educational experience, insufficient social support systems, a history of other social and personal problems, and other barriers to employment. It is generally recognized that welfare leavers are concentrated in lower-paying service and clerical jobs and those who are working (estimated at around 50 percent) remain among the “working poor.” A re- search synthesis on the consequences of welfare reform prepared by the federal agency responsible for TANF, Administration for Children and Families, U.S. Department of Health and Human Services, summarized results from major research and evaluation projects. Most of the studies have concluded that welfare reform (TANF is the center- piece) has had substantial effects on reducing caseloads and “is responsible for a portion of the increase in work and earnings among single mothers during the last decade.” However, the report notes that some strategies for moving client/consumers quickly into jobs, such as short-term job search assistance, have not obtained positive results. It is also reported that some favorable effects attributed to welfare reform will not persist over time. Finally, the report indicates that financial work incentives inside the welfare system or earnings supplements outside the welfare system generate the strongest in- come gains and antipoverty efforts.32

Another measure has the manifest purpose of transitioning people with disabili- ties into the labor market. The Ticket to Work and Work Incentive Improvement Act of 1999 is designed to encourage individuals with severe disabilities and chronic con-

132 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

ditions who are receiving SSI and SSDI to enter the labor market. Historically, this population has been discouraged from seeking employment for fear of losing eligibil- ity for disability benefits and/or health care benefits. The Ticket to Work provision of this legislation was implemented in 2002 with mail-out tickets encouraging voluntary involvement of SSI and SSDI beneficiaries to contact employment networks that could assist in developing individual work plans and placement for appropriate employment. The major incentive provision is a feature allowing states and U.S. territories the op- tion to give SSI and SSDI beneficiaries the opportunity to earn more and keep Medic- aid coverage at little or no cost. This incentive feature, called the “Medicaid buy-in” option, allows states to extend Medicaid coverage to working disabled whose income would have prevented them from qualifying for the program. States can increase the Medicaid income and resource limits for these individuals and have the option to offer Medicaid to workers with disabilities whose medical conditions improve to the point where they no longer qualify for SSI. By February 2004, twenty-eight states had im- plemented Medicaid buy-in programs and another five states had authorized their plans for implementation.

Procreational Incentives, Marital Instability, and Generational Dependency

Other criteria for evaluating eligibility rules, especially cash-benefit programs, are the extent to which they provide incentives for procreation, marital breakup, and/or the de- pendency of the children of families who receive public benefits. The possibility that cit- izens conceive children in order to become eligible for, or to increase, welfare benefits surfaces regularly as a matter of public and political discussion. For some the issue is that when benefit eligibility is tied to the number of children in families it is possible that it serves as a significant childbearing incentive. The latter issue is usually argued from a so- cial problem viewpoint that is ideologically committed to the notion that work is a highly valued instrumental activity and that citizens have a predominant propensity not to choose work if there is an available alternative—no matter how grim. It is certainly pos- sible to conceive of a person who would endure the physical discomforts of bearing chil- dren as the preferred alternative to working, but even if the standard of living it afforded was considerably less than a poverty line existence, such a choice is neither economically nor socially rational. What are the costs of bearing and rearing a child, when measured against the welfare benefit gain? Where such calculation is made, only the person who could never expect to work at all would find it to her advantage to bear children just to ob- tain an increase in benefits. A “family cap” rule is used in most states to deny increased cash assistance to women who have another child while on TANF. Where additional as- sistance is given, it is less than the costs incurred in adding a child to the family.

Surely, there will always be a few people who make irrational choices that work against their own economic self-interest, but to rebut such an argument we only need assume that the ordinary person acts in ways that will be of most economic benefit to herself or himself. Furthermore, there is every reason to believe that the average poor

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 133

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

person, well acquainted with the realities of life at the poverty level, does that in serious matters of everyday life.

Persuasive evidence against the notion that financial incentives stimulate child- bearing is found in the results of programs in countries that need to increase population rapidly: Attempts to do so are made through social welfare programs that grant benefits, often sizable, to citizens who bear children. The most massive of such programs was the French attempt to raise their birthrate in a population decimated by World War I when France lost half its male population. Both Sweden, Russia, and more recently Canada have made similar attempts for similar purposes. All these programs have been entirely unsuccessful. It is worthwhile noting that in Third World countries as standards of living and wage rates rise, along with increased literacy and educational attainment of women, birthrates go down rather dramatically, irrespective of the availability of birth control.

In summary, does the fact of eligibility for welfare benefits serve as an incentive to procreation? Given the evidence reviewed before, it is very unlikely that there is any such effect in a population or even any of its subgroups, though there may be some marginal and individual instances. There is, of course, no wisdom in forming large-scale public policy around small marginal effects. We are left with the conclusion that increasing ben- efits with family size does not create an incentive to further childbearing.

Another widely discussed issue is that public benefits like the old AFDC (now TANF) create marital instability: Do families split in order to meet the condition that the major wage earner be absent from the home? Nancy Murdrick studied that issue directly through data on when AFDC applications occurred relative to the marital split and ob- served differences between high- and low-income families with respect to the same issue. Study results are clear. The data show that most AFDC applications occurred nearly two years after the split. Murdrick concludes that the AFDC application is a re- sponse to the consequences of the split, not a premeditated outcome. Nor does it matter whether the applicants had an above-average or below-average income prior to their split.33,34 Surely, some couples do split up just in order to qualify for welfare benefits, but the policy-relevant issue is what is the case for most people. Once again, the issue of policy trade-offs is relevant: Doing good policy for the general population may create some negatives for a smaller number of others.

Yet another problem said to be a consequence of eligibility for public welfare ben- efits is that, generally speaking, citizens who now receive public benefits were reared in families who depended on public benefits and that this current generation will produce children who also will live at the expense of public benefit. In its most rational form, this argument over generational dependency (as it is sometimes called) asserts that social and personal identity is crucial in determining the choices made about work and “get- ting by.” It assumes that a child who grows up in a family in which there are no models of working to make a living will simply follow the pattern set by adults, so he or she will search out the welfare option. In its more unsympathetic form, the argument asserts strong antisocial, deviant motives to both parents and children in poor economic cir- cumstances. In order to make this argument plausible, it would seem necessary to assume that generational dependency must involve primarily those children whose families spent long periods as welfare beneficiaries, since the learning of role models and the socializa-

134 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

tion process referred to is never a short-term matter. No current explanation or approach to socialization suggests otherwise. If that is the case, the data from the Michigan Panel Study of Income Dynamics bear strongly on the plausibility of the generational depen- dency argument.35 This study, which has few challenges to its methodology or conclu- sions, shows clearly that only 12 percent of all welfare beneficiaries had received benefits for as long as four years, cumulatively. The authors conclude that there is little support for the existence of a sizable welfare class, that the most characteristic welfare recipient receives public benefits for about two years in succession and then may move on and off benefits for two considerably shorter periods of time later in their lives.36 If there is no large number of persons who spend long years on welfare benefits, it seems unlikely that the necessary conditions are available in which the mechanisms that are said to create generational dependency can work. Of course, this only shows that if generational de- pendency exists at all, it is a small-scale problem. Other studies, based on less extensive data than the Rein and Rainwater study, support the preceding general conclusions.37 It is noteworthy that in both the Rein and Rainwater and the Podell studies, the definitions of welfare dependency are loose (“receiving over half the total income from public funds” in the latter case and simply “to have received any income from public funds” in the former).

Opportunities for Political Interference via Weak Eligibility Rules

At one level there is every reason to believe that political influence is one route to the en- titlement to public benefits for individuals and groups—of course, social programs are a vehicle by which political interests are (and should be) expressed. But once the program or policy is implemented, it becomes bad social policy for citizens, or groups of them, to be either eligible or not simply because of political influence that circumvents the legisla- tive or judicial processes that keep social policy as an expression of the will of the people in a democracy. Equity is the value issue here. Citizens in a democracy should have equal access to public benefits, and that access should not depend on whom one knows or doesn’t know. Nor should it depend on the desire of the executive branch of government to shape a social program in ways that it couldn’t achieve through the regular channels of the legislative or judicial process. There is an unusual modern example of the latter, which we will briefly review for its value in illustrating the great danger posed by eligi- bility rules that are vague and uncertain in administration. Well-formed eligibility rules are not valuable just for their tidiness, but also that they might avoid political interven- tion in the operations of social programs, an intervention of a particularly vicious sort for vulnerable people. This example, from the mid-1980s, concerns the Social Security Dis- ability Insurance (DI) program.

Probably the premier policy problem of a social program for people with disabil- ities is to construct a useful and stable definition of disablement, and the DI program is no exception. Robert Ball, chief actuary for the Social Security Administration for many

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 135

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

years, reports that the slippery DI definition of disability allowed opposing biases to be used within one rather short period of time.38 The reason for its “slipperiness” is that it leaves one part of the eligibility rule to medical and administrative discretion—the deter- mination of whether a disability exists in fact. Thus, administrators and physicians were left to liberal interpretation of medical facts. One has only to look at the sizable propor- tion of initial application decisions that were reversed and “re-reversed” at every stage of reconsideration and appeal to realize that what is technically called “interjudge relia- bility” was a hallmark lack in this eligibility process.39 Over a ten- to twelve-year period beginning in the 1970s, reversal rates on disability denial appeals rose to nearly two- thirds of all appeals; in regard to mental disabilities, reversal rates reached as high as 91 percent of all appealed denials of benefit applications.40

In explanation, Robert Ball noted that in the early years of the program, “I can as- sure you gentlemen, that the general attitude . . . [was] wanting to pay claims.”41 To the point, it is notable that in this climate, even though Congress expressly forbade the So- cial Security Administration (SSA) from reversing the findings of state disability deter- mination units, it did so regularly (to the advantage of applicants).42 However, under the prodding of a Congress worried about rising program costs and a presidential adminis- tration looking with disfavor on most welfare benefits, SSA began by a variety of means to administer a very different definition of the term disability. Clearly, SSA was able to turn the DI system around simply by the strength of its own ability to reinterpret the de- finition of disability and change some of its procedural mechanisms: In five years, DI benefit allowance rates were cut in half, terminations increased, and total costs slowed significantly. “[D]isability examiners have become more conservative in the way in which they interpret and apply standards [for DI awards].”43

Despite the significant changes that had already occurred, with the 1981 inaugura- tion of President Reagan, who had made explicit campaign promises to reduce the size of entitlement programs, not only were new applicants under fire, but people with disabil- ities who already received benefits were affected as well. Unprecedented terminations of thousands of DI beneficiaries took place between 1980 and 1985: 71,500 in 1980; 98,800 in 1981; and 121,400 in the first five months of 1982—with 360,000 expected to be ter- minated in 1984.44 “In the 1960s, the loose and ambiguous definition of disability could not constrain a [Democratic, neo–New Deal] political administration determined to ex- pand the program any more successfully than in the 1980s it could constrain a [Republi- can, conservative] political administration determined to reduce the size and costs of the DI program.”45 Now another highly placed Social Security administrator could say, mimicking Robert Ball’s earlier statement, “I can assure you gentlemen, the general atti- tude [in the Social Security Administration] is to deny, deny, deny.”46 Intelligent pro- grams cannot be administered under such conditions of radical political changes in programs. Beneficiaries with serious disabilities have had reason to expect that they could count on their benefit income in one year, only to learn a few years later that despite no change in their condition, benefits will be withdrawn. Worse, they learned a few more years later that many if not most terminations were illegal in the first place, so that if they reapply there is good chance that their benefits will be reinstated (Minnesota v. Schweiker, 1983; Social Security Regulations no. 83-15, 16, 17s, 1986). On such grounds as outlined before, it is clear that this policy system was in ragged disarray.

136 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

The definitional ambiguity of disability with which the DI has (and still does) op- erate has been used by parties of opposing political persuasions to expand and contract the program at will. Note that it is possible to increase substantially the clarity and relia- bility of medical disability determinations, as Mashaw (and Nagi before him) have clearly shown, by fairly simple attention to definitional clarity, plus well-known modern re- search findings on making clinical judgments. It is also clear that there is every reason to expect further political adventures into the Social Security system absent the correction of this policy problem. “If the same policy weaknesses that made possible the political in- trusions into this social program are still in place when the next liberal administration comes into office, it will simply use the very same weaknesses to restore the system to its former condition.”47 Such a political scenario would continue into infinity, a prospect that is not in the best interests of the country or its citizens with disabilities.

This example teaches two key lessons: First, it highlights for us from Mashaw the conclusion that there is nothing inherently wrong with using expert judgments as a basis for eligibility rules. Second, it shows that some conditions are necessary to keep the pro- cess on track and functioning. Social researchers have learned of those requirements: define very carefully the thing to be judged, train and orient judges to apply only that definition within a specific procedural context, and indoctrinate new judges into that system, a few at a time. This process is not inexpensive, but almost any trained re- searcher can achieve a 90 percent agreement with almost any set of judges making even complicated judgments. Costs will surely be less than the direct administrative costs of disentitling and reentitling beneficiaries with disabilities time and time again.

Summary

This chapter presented concepts to assist the practitioner in understanding the variabil- ity among common eligibility rules and procedures. The following types of eligibility rules were discussed:

� Prior contribution � Administrative rule � Private contracts � Professional discretion � Administrative discretion � Judicial decision � Means testing � Attachment to the workforce

Whereas the ultimate test of the merits of any particular eligibility rule is its fit with the social problem conception that underlies the program or policy under consideration, special problems are likely to be created by eligibility rules. The practical analyst should examine the available data and the general workings of the policy or program to search for evidence of the following special problems:

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 137

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

� Stigma and alienation � Off-targeting of benefits � Overwhelming costs � Overutilization and underutilization � Political interference � Negative incentives and disincentives (work, procreation, marriage, and so on)

The presence of any of these special problems works against the achievement of a func- tional policy and programs—against adequacy, equity, and efficiency.

E X E R C I S E S

1. What is the difference between the eligibility rule known as administrative discretion and the one known as administrative rule?

2. What are the consequences of basing eligibility for social welfare benefits solely on the type of entitlement called “attachment to the workforce”?

3. There are three branches of U.S. government: legislative, executive, and judicial. What role does each play in establishing the eligibility rules for TANF benefits? What may each branch do to affect eligibility rules once the TANF program is established? (Remember, no state is required to have a TANF program.)

4. What is the major difference between professional discretion and administrative dis- cretion as methods of determining eligibility for social welfare benefits or services?

N O T E S

1. The scheme concerns only “selective” eligibility rules. However, this book will not consider the traditional selective versus universal distinction in regard to (among other things) eligibility rules, sid- ing with Titmuss in his belief that its utility for the practical policy analyst is only marginal.

2. R. Titmuss, “Welfare State and Welfare Society,” in Commitment to Welfare (London: Allen and Unwin, 1968), pp. 130–134.

3. In three states minimal employee contributions are also required. 4. Only in the United States is the workers’ compensation system operated as a private enterprise. 5. K. R. Wedel, “Designing and Implementing Performance Contracting,” in R. L. Edwards and

J. A. Yankee (eds.), Skills for Effective Service Management (Silver Spring, MD: NASW Press, 1991), p. 86.

6. K. R. Wedel and S. W. Colston, “Performance Contracting for Human Services: Issues and Suggestions,” Administration in Social Work, 12(1) (1988): 73–87.

7. M. Brown, Working the Street: Police Discretion and Dilemmas of Reform (New York: Basic Books, 1981).

8. M. Lipsky, Street Level Bureaucracy: Dilemmas of the Individual in Public Services (New York: Rus- sell Sage Foundation, 1980).

9. D. E. Chambers, “The Reagan Administration Welfare Retrenchment Policy: Terminating So- cial Security Benefits for the Disabled,” Policy Study Review (1985): 207–215.

10. D. E. Chambers, “Residence Requirements for Welfare Benefits,” Social Work, 14(4) (1971): 29. 11. Brown v. Board of Education, 347 U.S. 483 (1954).

138 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

12. H. D. Krause, Child Support in America (Charlottesville, VA: The Mitchie Company, 1981), pp. 330–340.

13. T. B. Festinger, “The Impact of the New York Court Review of Children in Foster Care: A Follow-up Report,” Child Welfare, 55(8) (1976): 515–544.

14. AFDC beneficiaries live below the poverty line and seldom during their whole lives work at jobs that pay more than minimum wage, whereas workers’ compensation beneficiaries almost always earn average incomes and do at least semiskilled work.

15. Characteristics of State Plans for Aid to Families with Dependent Children (Washington, DC: U.S. Department of Health and Human Services, Social Security Administration, Office Administration for Children and Families of Family Assistance) (Part Two, Table 1, 1991), pp. 409, 455, 488.

16. J. Feagin, Subordinating the Poor (Englewood Cliffs, NJ: Prentice Hall, 1974), pp. 103–118. 17. T. F. Pettigrew, “Social Psychology’s Contribution to an Understanding of Poverty,” in V. T.

Covello (ed.), Poverty and Public Policy (Cambridge, MA: Shenkman, 1980), pp. 198–224. 18. Susan Sheehan, A Welfare Mother (New York: Signet Books, 1976). Social Security Benefit Rates,

Leaflet NI 196. Department of Health and Social Services (HMSO), 1990. 19. G. Hoshino, “Simplifying the Means Test,” Social Work (July 1965): 98–103. 20. N. Kotz, Hunger in America (New York: The Field Foundation, 1979), pp. 28–29. 21. Child Benefit, Leaflet CH1. Department of Health and Social Security (HMSO), 1990. 22. S. Danziger and K. Portnoy, The Distributional Impacts of Public Policies (New York: St. Martin’s

Press, 1988), p. 124. 23. That is because 85 to 95 percent of the U.S. workforce is covered by the Social Security system,

so they are automatically eligible for Medicare when they receive retirement and disability benefits. 24. J. Menefee, B. Edwards, and S. Scheiber, “Analysis of Non-participation in the SSI Program,”

Social Security Bulletin, 44 (1981): 3–21. 25. Data were gathered by author in personal interviews with public (and private) child-placing staff

and administrators while conducting research on exportation of Central American children to Europe and the United States for adoption in Honduras, El Salvador, Guatemala, and Costa Rica (1991–1992).

26. Speenhamland was an English town whose council solved its poverty problem by providing bread—not cash—to needy persons. There was a public outcry from those who believed it would de- stroy all incentive to work.

27. The Seattle-Denver Income Maintenance Experiment, Midexperiment Results and a Generalization to the National Population (Stanford, CA: Stanford Research Institute and Mathematics Policy Research, 1978), p. vii.

28. Ibid., 11–12. 29. Ibid., 12–14. 30. M. T. Hannan, N. B. Tuma, and L. P. Groeneveld, “Income and Marital Evidence from the In-

come Maintenance Experiment,” American Journal of Sociology (May 1977): 1200–1201. 31. Oklahoma Department of Human Services, Family Support Services Division—TANF Section.

(2000). TANF Work: The Future Is Yours. . . . (No. 93-10) [Brochure]. Oklahoma City, OK: Author. 32. Administration for Children and Families, U.S. Department of Health and Human Services,

“Temporary Assistance for Needy Families (TANF) Fifth Annual Report to Congress,” February 4, 2002, www.acf.dhhs.gov/programs/ofa/annualreport5/chap13.htm.

33. N. Murdrick, “Use of AFDC by Previously High and Low Income Households,” Social Service Review, 52(1) (1978): 110.

34. B. Bernstein and W. Meezan, The Impact of Welfare on Family Stability (New York: Center for New York City Affairs, 1975), p. 99.

35. M. Rein and L. Rainwater, “Patterns of Welfare Use,” Social Service Review, 52(4): 511–534. 36. J. N. Morgan et al., Five Thousand American Families: Patterns of Economic Progress, vol. 1 (Ann

Arbor, MI: Institute for Social Research, 1974), pp. 1–9. 37. L. Podell, Families on Welfare in New York City (New York: The Center for Study of Urban Prob-

lems, 1968), pp. 28–29. 38. M. Derthick, Policy Making for Social Security (Washington, DC: The Brookings Institution,

1979). 39. Chambers, “The Reagan Administration,” p. 4.

C H A P T E R 6 / Who Gets What, How Much, and Under What Conditions 139

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5

40. U.S. Congress, Senate Subcommittee on Oversight of Government Management of the Senate Committee on Governmental Affairs, “SSDI Reviews: The Role of the Administrative Law Judge,” Hearing Report, 98th Congress, 1st Session (June 8, 1983), Appendix, memo from Carl Fritz to Louis Hays, Chief of the Appeals Division of the Social Security Administration.

41. Derthick, Policy Making, p. 310. 42. D. Goldsborough et al., “The Social Security Administration: An Interdisciplinary Study of

Disability Evaluation” (Washington, DC: George Washington University Law Center, 1963), mimeographed, pp. 98–100.

43. M. Lando, A. Farley, and M. Brown, “Recent Trends in the SSDI Program,” Social Security Bul- letin, 5(2) (August 1982): 50.

44. J. Mashaw, Bureaucratic Justice (New Haven: Yale University Press, 1983). 45. Chambers, “The Reagan Administration,” p. 7. 46. Mashaw, Bureaucratic Justice, p. 37. 47. Chambers “The Reagan Administration,” p. 15.

140 P A R T T W O / A Style of Policy Analysis

Bacon. Copyright © 2005 by Pearson Education, Inc. Social Policy and Social Programs: A Method for the Practical Public Policy Analyst, Fourth Edition, by Donald E. Chambers and Kenneth R. Wedel. Published by Allyn and

IS B

N : 0 -5

3 6 -1

2 1 1 2 -5