The Advantages and Disadvantages of Different Social Welfare Strategies

Transcription

1 The Advantages and Disadvantages of Different Socia Wefare Strategies by Lawrence H. Thompson* The foowing was deivered by the author to the High Leve American Meeting of Experts on The Chaenges of Socia Reform and New Administrative and Financia Management Techniques. The meeting, which took pace September 5-7, 1994, in Mar de1 Pata, Argentina, was sponsored by the Internationa Socia Security Association at the invitation of the Argentine Secretariat for Socia Security in coaboration with the ISSA Member Organizations of that country. * Deputy Commissioner of the Socia Security Administration. The views expressed in this artice are the author s own and do not necessariy refect those of the U.S. Government or the Socia Security Administration. Throughout the word, societies are reexamining, reforming, and restructuring their socia wefare systems. New ways are being sought to manage and finance these systems, and new approaches are being deveoped that ater the reative roes of government, private business, and individuas. t surprisingy, this activity has triggered spirited debate about the reative merits of the various ways of structuring socia wefare systems in genera and socia security programs in particuar. The current changes respond to a variety of forces. First, many societies are adjusting their institutions to refect changes in socia phiosophies about the reative responsibiities of government and the individua. These phiosophica changes are especiay dramatic in China, the former sociaist countries of Eastern Europe, and the former Soviet Union; but they are aso occurring in what has traditionay been thought of as the capitaist West. Second, some societies are strugging to adjust to the rising costs associated with aging popuations, a probem particuary acute in the OECD countries of Asia, Europe, and rth America. Third, some countries are adjusting their socia institutions to refect new deveopment strategies, a change particuary important in those countries in the Americas that seek economic growth through greater economic integration. And, finay, in many parts of the word, socia wefare reform is motivated by the need to adjust the costs of wefare systems to economies that are no onger growing as fast as they did in the first three or four decades after Word War II. Whatever their motivation, these changes are being discussed and debated widey. And the current discussions seem in severa ways to refect a new eve of sophistication about the compexities of socia wefare poicy. One exampe of this is the recognition of the important reationships between socia wefare systems and the economy in which they exist. The state of its economy wi often infuence a society s wiingness to support its socia wefare system. Heathier economies faciitate more generous socia wefare systems, whie economic difficuties frequenty ead, sooner or ater, to retrenchments. At the same time, the size and structure of Socia Security Buetin Vo. 57,. 3 Fa

2 socia wefare systems can themseves infuence the pace and rate of growth of economic activity. Socia wefare systems that inadvertenty discourage work activity --or encourage a shift from the forma sector to the informa sector---can reduce the aggregate amount of income avaiabe for distribution among society s members. Simiary. a system which discourages domestic savings can have the effect of sowing economic growth, whie. given the proper environment, one that increases domestic savings coud enhance economic growth. Another indication of the increased sophistication of current debates is the reaization that socia wefare poicies are deveoped to achieve a variety of different and often competing socia objectives. These socia objectives incude the effective protection of the popuation from various economic risks, the promotion of increased economic activity. the redistribution of economic resources, the faciitation of the smooth operation of a free abor market, and the efficient operation of socia institutions. singe poicy wi be best abe to achieve a of these various objectives, so the choices actuay made wi necessariy refect decisions, either impicit or expicit. about the reative importance of the various ob.jectives. Since assessments of the reative importance of competing socia objectives wi vary from country to country and from time to time, socia wefare poicies appropriate in one time and pace need not necessariy be appropriate in another. Moreover, the degree to which a particuar approach to the socia wefare system does, in fbct, advance a particuar objective wi aso vary from pace to pace and from time to time, Put simpy, the poicy that is most effective in achieving a given objective in Argentina may not achieve that objective in the United States, and the peope of the United States tnay wish to pursue a particuar objective with their socia wefare system that is not as important to the peope of Argentina. particuar structure is the most appropriate one at a times and in a societies. Phiosophica considerations wi aways he important to socia wefare decisions. But carefu and objective anayses of the many impications of each possibe socia wefare option are aso very important. That concusion emerged from a recent expert meeting hosted by the Internationa Labor Office (ILO) in Mexico City. The experts noted that a key eement of socia security reform was:...carefu and rigorous exporation of a of the options-in a particuar country s context--considering [such objectives as] administrative ef iiciency, cost containment, breadth of coverage. and transparency... My purpose in deveoping this paper is to contribute to the rationa debate and discourse referred to in the IL0 communique by deveoping further some of the thoughts 1 have just outined. In particuar. 1 wish to suggest a framework for thinking about the various goas, objectives, and structures and to offer some suggestions about the reative advantages and disadvantages of different approaches to achieving the various goas. 1 seek to further the diaogue about these topics by producing a document that wi provoke others to think and react. Some may suggest the addition of important socia objectives that they beieve I have overooked. Others may suggest additiona ways in w-hich particuar socia and economic structures may hep achieve a given socia objective. A such suggestions and eaborations wi hep us to increase our understanding of these important issues. Finay, in order to simpify (somewhat) an aready compicated discussion, my anaysis focuses amost excusiveiq on programs that suppy cash benefits; indeed the impicit focus is amost excusivey on pension programs. Much of the discussion contained here woud be appropriate for other major socia wefare programs, such as those concerned with acute heath care services. But heath programs aso introduce other issues that go beyond the scope of this paper, such as provider reactions to different institutiona arrangements. The Competing Approaches Most advanced societies rey on some combination of six approaches (or cose variations) to structuring income support programs. I begin by describing briefy each in turn. Socia Insurance Socia insurance is the argest singe eement in and the foundation for the socia wefare system of most countries. It grew out of vountary insurance arrangements of the medieva European craft guids, was institutionaized by European governments in the ate 19th and eary 20th centuries, and soon spread from there to the Americas. Since socia insurance is the foundation of so many socia wefare systems, it seems the reasonabe starting pace for this discussion. Athough detais vary from one country to another, socia insurance programs throughout the word share certain characteristics. The interaction of these characteristics gives socia insurance a unique set of attributes, and the differences between socia insurance and other socia wefare approaches can be traced primariy to differences in the mix of these characteristics. Some expert commentators cite seven characteristics as essentia eements of socia insurance. 1. Compusory participation. Most peope participating in socia insurance programs do so as a resut of a ega requirement. In some programs, a sma minority may be aowed to choose whether to participate. 2. Government sponsorship (and reguation). Governments create and supervise socia insurance programs, but do not necessariy manage them. The programs may actuay be operated entirey by private sector institutions (for exampe, the German heath and pension systems); by a combination of pubic agencies and private contractors (the mode used for Medicare. the heath insurance program for the aged in the United States); or directy by a pubic sector agency (the mode used by Ango- Saxon countries for operating their pubic pension programs). Where the private sector runs these programs, however, operations are tighty supervised by the pubic sector. 4 Socia Securit! Buetin Vo Fa 1994

3 3. Contributory finance. Mostsometimes virtuay a-of the resources needed to run the program are raised through expicit contributions coected from the empoyer or from both the empoyer and the empoyee. A worker s contribution is usuay a fixed percentage of his or her wage or income.3 4. Eigibiity derived from contributions. Eigibiity for benefits under socia insurance programs rests, in part, on current or previous contributions by the individua and/or the individua s empoyer. 5. Benefits prescribed in aw. Uniform sets of entiting events and schedues of benefits are deveoped, announced, and appied to a participants. Administrators of the program have itte discretion in determining who shoud get benefits or how much they shoud get. 6. Benefits not directy reated to contributions. Socia insurance programs usuay redistribute toward ower-wage workers or toward persons engaged in activities deemed to be sociay desirabe. Perhaps the most dramatic redistribution occurs in heath insurance and fat pension benefit programs, in which higherwage workers contribute more than ower-wage workers but everyone receives the same benefits. In most pubic pension programs, benefits are scaed to previous earnings. Even then, the owerwage worker tends to get back proportionatey more than the higher-wage worker, sometimes much more. Many socia insurance schemes aso subsidize benefits for nonworking members of the famiies of workers, students, members of the armed forces, homemakers caring for chidren, and others whose activities are deemed to be sociay beneficia. 7. Separate accounting and expicit ong-range financing pan. Socia insurance contributions are usuay earmarked to pay the socia insurance benefits. Governments typicay keep separate accounts that permit comparisons of program receipts and program benefits, though they may aso present financia information that integrates the socia insurance programs with the other government operations. Governments aso typicay deveop an expicit pan showing that projected revenues are sufficient to finance projected expenditures for severa years into the future (or, if revenues are not sufficient, expaining how the government proposes to baance projected receipts and schedued benefits). Some of these characteristics are necessariy inked. For exampe, if a socia insurance program is to provide adequate support to ower-income househods, it must redistribute from higher-income participants to ower-income participants (characteristic 6). But a program that redistributes expicity and significanty must aso be compusory (characteristic 1) or ese higher-income peope wi choose not to participate. And a program that is compusory must be sponsored by the government (characteristic 2). Other characteristics combine to give socia insurance some of its key attributes. For exampe, advocates of the socia insurance approach argue that it deivers benefits in a way that promotes a sense of individua worth and dignity among recipients. They argue that this resut fows from the fact that socia insurance is financed by expicit contributions (characteristic 3) and that entitement to benefits derives from the payment of these contributions (characteristic 4). The underying phiosophy of the program, then, is that beneficiaries have earned the right to receive their benefits by paying in their contributions. The inkage between contributions and benefits can aso serve as an incentive for compiance with socia security taxes, athough the strength of this incentive wi obviousy depend on how cosey benefit amounts are inked to prior contributions. Some advocates of the socia insurance approach aso argue that it introduces fisca discipine into the poitica process. Beneficiaries of socia wefare programs (whether pensioners or heath care providers) naturay favor raising the eve of benefits paid under the program. Because of the combination of contributory finance (characteristic 3) and the separate accounting of program receipts and benefits (characteristic 7), however, program expansions are ikey to require an increase in contribution rates. Thus, they argue, wage earners-who wi have to pay higher contributions to finance program expansions-have an incentive Characteristics - of competing socia wefare approaches - Item Participation compusory L Empoyee! Vountary tax ---- N/A N/A Government sponsors ~ Contributory finance Eigibiity based on contributions Benefits specified in aw , but, but, but Benefits reated to contributions Separate accounting Reserve financing required..,... i, but, but NA = not appicabe. Socia Security Buetin Vo. 57,. 3 Fa

4 to support restraints on the growth in socia insurance benefits.4 8. Reserve financing not required. Reserve financing is a fina attribute worth noting because it is not necessary-and, in fact, is reativey rare-in socia insurance, whereas it is more common in some of the aternative institutiona arrangements. Financia reserves are frequenty required in private sector pension pans in order to hep assure that pension promises are met. Socia insurance can be financed on a pay-as-you-go basis ony because it is backed utimatey by the taxing power of the State. Empoyer Mandates Another approach to providing socia wefare benefits invoves government mandates that a empoyers (or a arge empoyers) provide or finance specific socia wefare benefits to their empoyees. Empoyer mandates and socia insurance share many characteristics. In both cases: (1) the socia wefare program resuts from expicit government sponsorship (though, in this case, not government operation), (2) eigibiity for benefits is connected to empoyment, (3) participation is compusory, (4) benefits are financed primariy from empoyer (and, occasionay, empoyee) contributions, and (5) benefits are financed according to an expicit pan. Aso, empoyer mandates invariaby invove specification in aw of either a minimum eve of contribution or a minimum benefit package, athough empoyers are usuay free to offer a better package than the minimum.5 The two approaches differ in at east two important ways. First, programs resuting from empoyer mandates are ess ikey to redistribute; they focus amost excusivey on individua equity at the expense of socia soidarity. Second, empoyer-sponsored pension pans are more frequenty (though not universay) advance funded. That is, financia reserves are accumuated in advance to pay caims. Empoyee Mandates Empoyee mandates have ong been a part of the heath insurance systems of severa countries and have been used Means- Tested Programs more recenty in other countries as a Means-tested programs, often referred means of providing pensions. Empoyee to as socia assistance, pay benefits to mandates share many of the attributes of those who first demonstrate imited ecoempoyer mandates: programs are spon- nomic resources. In such programs, sored and reguated by the government, entitement has nothing to do with whetheigibiity for benefits is inked cosey to er caimants have had prior earnings or the payment of contributions, and bene- have ever paid taxes. A major advantage fits are financed to a significant degree of means-tested programs is that in confrom contributions. Perhaps two differ- trast to socia insurance, or empoyer and ences between empoyer and empoyee empoyee mandates, they do not ink mandates are worth noting. First, eigibiity for benefits to prior economic empoyee-mandated pensions must be behavior. As a resut, benefits can be advance funded; as far as I know, there is taiored to current individua circumno other way to organize them. Aso, in stances and assistance can be concenprincipe, a substantia amount of redis- trated on those with the fewest resources, tribution coud be buit into an empoyee who need them the most. Means-tested mandate program through the use of programs are financed from government targeted, government subsidies.6 Design- genera revenues so that program costs ing an empoyer mandate program with a are not separatey identified (except by substantia amount of redistribution anaysts who pore over budget docuwoud be more difficut. ments). Vountary Arrangements (Tax Expenditures) Governments often encourage empoyers to provide (and/or individuas to make) private arrangements which either repace or suppement pubic sector socia wefare programs. The encouragement usuay takes the form of a tax expenditure -an arrangement whereby someone s (usuay the individua s) tax iabiity is ower than it otherwise woud be as a resut of the undertaking of the activity. Government encouragement of private arrangements is probaby most common in the area of retirement savings, but is aso associated with some heath care benefits in many paces. Athough vountary programs are by definition not compusory, their shape is frequenty infuenced by government because programs quaifying for the specia tax treatment must meet certain minimum conditions. As with the empoyer and empoyee mandates, these programs rarey redistribute expicity. Vountary programs targeted at individuas do not have to be connected in any way to particuar empoyment patterns or situations, athough they can ony be effective with individuas who have a reasonabe income from which to make the vountary contributions. Universa Programs Some programs provide socia wefare benefits to a ega residents. These universa programs share severa characteristics with socia insurance: benefits are prescribed in aw and participation is compusory (to the extent that participation is a meaningfu concept for these programs). They differ in other important ways. Financing for universa programs usuay comes from genera revenues; eigibiity is tied to residence rather than previous earnings or contributions; and financing for the program is not panned for separatey from the panning for other types of government expenditures. Since there is no separate financing arrangement, there is no possibiity of advance funding through accumuation of reserves. Combinations and Hybrids Athough presented bet e as separate and distinct, the differences among these approaches sometimes become burred in actua practice. Societies often construct socia wefare systems that rey on a combination of the approaches; for instance, pubic pensions in Canada are provided through the combination of a fat rate, universa program and an earnings-reated socia insurance program. Occasionay, a new approach is 6 Socia Security Buetin Vo. 57,. 3 Fa 1994

5 deveoped by combining features of two or more of the traditiona approaches; for instance, medica insurance for the aged in the United States is provided through a program that shares many of the attributes of universa programs (at east for those age 65 and over), but which, technicay, is vountary and invoves the payment of a partia premium. Finay, hybrids aso arise as different ways evove to administer these programs; for instance, at the option of empoyers, the earnings-reated pensions in the United Kingdom are either operated by the State in the form of socia insurance or by empoyers in the form of an empoyer mandate program. And, under certain circumstances, empoyees may contract out directy. Socia Objectives I sha focus on attributes that can be custered around two broad socia objectives for the purposes of this discussion, recognizing that other categorizations woud be equay vaid. The first invoves attributes reated to providing effective socia protection to the popuation. The socia protection custer incudes: Treating peope with dignity and respect. Assuring the most compete coverage possibe for the system. A socia wefare system cannot provide effective protection to peope that it does not cover. Distributing equitaby the costs and benefits of the system; in particuar, assuring that those with more imited economic resources are protected adequatey. Efficient operation of socia wefare institutions so that, to the extent possibe, the resources devoted to them go to the improvement of beneficiary wefare rather than administrative overhead. My second custer invoves attributes that can hep promote a heathy economic environment. It recognizes that an effective socia wefare system can rest ony on the foundation of a heathy economy. It incudes: Encouraging individua thrift and not discouraging, unintentionay, individua work effort8 Fostering responsibe government fisca poicies by discouraging the tendency to overpromise socia wefare benefits and assuring that private saving is avaiabe to finance capita formation. Faciitating the smooth operation of markets, especiay abor markets, particuary by constructing institutions that minimize the socia costs of economic change. Interaction Approaches of and Objectives Having sketched the aternative institutiona approaches for constructing socia wefare systems and the aternative socia objectives for such systems, I now turn to the interaction of the two. In particuar, the foowing discussion compares the strengths and weaknesses of the different approaches and expores how effective each is ikey to be in achieving the different socia objectives. Individua dignity-one objective is simpy to organize the system in a way that treats each member of society with dignity and respect. As noted earier, socia insurance promotes individua respect and dignity through the phiosophy that those who make contributions have earned the right to the benefits. Other empoyer-provided benefits, as we as vountary and empoyeemandated approaches shoud be equay effective in promoting this objective. In contrast, those who receive means-tested benefits are often stigmatized. Universa programs shoud be just as effective at promoting individua respect and dignity as are the empoyment-based programs, as ong as they remain universay avaiabe. Some fear, however, that universa programs are poiticay unstabe-that they wi be vunerabe to the introduction of means-testing as a way of deaing with some future budget crisis. Those who vaue highy the objective of assuring the dignity of recipients and fear that universa programs may eventuay be means-tested usuay advocate one of the empoyment-based approaches, socia insurance, or either empoyer or empoyee mandates. They see the expicit contributions associated with socia insurance as buiding a poitica buwark against future means-testing: Comparative ratings of competing socia wefare approaches I-~---- I Socia 1 -Empoyer] -~-Goyee Tax; Means-tested 1 Universa Item insurance 1 mandates] maud~a~penditures! programs ~ programs Dignity..... Coverage... Costs and benefits scaed to abiity to pay... Administrative costs..... Encourage individua thrift..... Fisca discipine..... Faciitate abor market adjustment..... Budget cost..... N/A = not appicabe. A A A A F B C C F N/A C F F F A B D D D F D B A A F A F A F C A F A F A D A A A B B+ A C A D C A F Socia Security Buetin Vo. 57,. 3 * Fa

6 those who have been paying the contributions wi resist program changes that woud deprive them of the benefits that they fee they have earned. Deveopments in the Canadian pubic pension system raise concerns about the stabiity of universa programs (and suggest that socia insurance is somewhat more stabe), but they fa short of confirming the worst-case scenario just outined. Canadian pubic pensions consist of two separate programs: (1) a universa fat pension financed from genera revenues, suppemented by (2) an eamingsreated pension paid under a contributory socia insurance program. In 1988, in response to budget difficuties, Canada amended its persona income tax aw to caw back at east part of the universa pension payment from the higher-income edery and to tax it away entirey from those whose incomes are above a certain higher eve. Canada does not caw back socia insurance benefits9 Coverage.-Socia wefare programs cannot be effective if they do not reach the popuation in need. Broad coverage is achieved easiy under universa programs for which, by definition, everyone is eigibe. Broad coverage is more difficut under both socia insurance and empoyer mandates because eigibiity requires attachment to empoyment. Some peope simpy do not have sufficient attachment to gain eigibiity; others are empoyed in informa and casua abor markets, in which many empoyers do not compy with the aw. Empoyee mandates are ikey to be even harder to enforce than are empoyer mandates and vountary programs are ikey to have an even ower rate of participation. Finay, means-tested programs may fai to reach a substantia fraction of the popuation they are intended to serve, in part because not a of the target popuation knows the program is avaiabe and in part because some of them are discouraged by the hasse of appying for benefits or the stigma associated with accepting them. Distribution of benefits and costs.- The distribution of benefits and costs under a system of socia insurance is ikey to differ from that under a universa program ony to the extent that the revenue bases of the two programs differ. Socia insurance contributions tend to be proportiona to wage income, except among the highest earners. Furthermore, earnings tend to be a arger fraction of tota income for househods in the midde of the income distribution and a smaer fraction for househods at the owest and highest income eves. As a resut, wagereated financing is more regressive than persona income taxes; but is not necessariy more so than corporate taxes or saes, excise, and vaue-added taxes. A universa program financed by higher persona income taxes woud probaby be more progressive than socia insurance; one financed by increases in vaue-added taxes may we be ess progressive. In genera, socia insurance approaches wi be more redistributive than private sector arrangements, whether these arrangements are mandated or vountary. In heath insurance, the greater progressivity derives argey from the difference between wage-reated contributions and individua premiums. In pensions, it tends to derive from the provisions that afford ower wage workers proportionatey higher benefits under socia insurance programs. In principe, some greater redistribution can be buit into private sector programs through the use of targeted government subsidies. This woud appear to be more feasibe for empoyee mandates and vountary individua programs where a subsidy can be based directy on the househod income of the intended recipient than it is for empoyer-based programs where the subsidy cannot be targeted to appy ony to particuar participants. Peope with reativey ess taste for redistribution wi be more ikey to favor private sector approaches, and those who favor private sector approaches may do so precisey because they prefer ess redistribution. At east in the United States, higher-income workers occasionay argue that they shoud not be required to participate in Socia Security since they coud get a higher return on the amount they must contribute to the program if aowed to invest it privatey. They are probaby correct. But the ower return they receive under Socia Security is mainy the resut of the redistribution buit into the program. O Whether an entire birth cohort can receive a higher return under private arrangements depends on the reationship between the rate of growth of productivity and of popuation, on the one hand, and the rea interest rate on the other. Whatever the reationship may be, the gap between the two is not ikey to be arge and consistenty in one direction or the other for ong periods. Means-tested programs assure a greater redistribution because other transfer programs are ess effective at targeting benefits to those who are most in need. Paradoxicay, their more effective targeting does not mean that meanstested benefits are necessariy more effective in serving ow-income beneficiaries. These beneficiaries are ess apt to appy for benefits that carry the stigma associated with the means test than they woud be for benefits paid through a universa or socia insurance system. Administrative cost.-administrative expenses increase the tota cost of the socia wefare system without increasing benefits the system can pay, and can vary dramaticay from one system to another. Major administrative cost eements incude the cost of revenue generation, money management, benefit eigibiity determination, and ongoing benefit administration. Of the aternatives discussed here, universa programs probaby have potentiay the owest administrative cost. Revenues are raised as a part of genera tax coection procedures, there are no money management costs, and eigibiity usuay depends on meeting a few, reativey easiy verifiabe conditions. The administrative costs of socia insurance are aso potentiay quite modest since eigibiity tends to be based on information that can be coected and maintainer reativey easiy. On the other hand, means-tested programs can be reativey expensive to operate because of the neec to coect and verify extensive financia information in order to assure that individuas are (and remain) eigibe for benefits.12 Generaizations are more difficut about the reative administrative costs o pubic and private sector programs. Socia Security Buetin Vo. 57,. 3 Fa 1994

7 Pubic sector programs usuay enjoy ings. On baance, their resuts do not potentiay greater economies of scae in their operations and avoid a substantia amount of the saes expenses of private support the fear that such a system wi seriousy erode savings and capita formation. On the other hand, studies in the sector firms. At any given time and United States aso suggest that funded pace, however, inefficiencies inherent in pension pans do have a positive effect on many pubic sector enterprises-such as savings. Nationa savings may increase faciities or excess staff resuting from by 30 to 40 percent of any increase in the poitica pressures-may more than offset aggregate amount of assets being hed in the potentia advantages enjoyed by the pubic sector, especiay if the private sector firms are exposed to effective competition. I3 Within the private sector, pension and other retirement accounts. 5 Taken together, these two resuts suggest that, in the absence of offsetting changes in government fisca operations, shifting one woud expect that empoyer- from a pay-as-you-go socia security sponsored approaches woud invove system to an advance-funded pension fewer administrative expenses than system woud have a positive effect on woud empoyee-sponsored approaches nationa savings. since the former shoud more frequenty Fisca discipine.-another objective be associated with arger institutions that is to counter the natura tendency for are better abe to capture economies of poitica decision makers to overpromise. scae and that woud have ower costs of Granting benefit increases is popuar, but saes per individua served. entais the assumption of future iabii- Work and savings incentives.-for ties. If these iabiities become too arge, severa decades economists have ex- their costs can eventuay have undesirpored the effects on individua behavior abe economic and socia consequences, of various eements of the socia wefare such as: system. I4 In genera, their anayses sug- An increase in evasion of egisated gest that work effort woud be somewhat responsibiities or the growth of the higher in the absence of the payment of informa sector, either of which cash benefits. The magnitudes are not makes the socia wefare program overwheming, however, and the studies ess effective at protecting citizens; do not address such crucia issues as whether, on baance, socia wefare is A reduction in internationa competiincreased by aowing the edery to work tiveness of domestic businesses, ess. These studies do suggest that re- which wi eventuay undercut iving strictions on the amount that can be standards and reduce socia wefare; earned whie sti receiving retirement Infationary pressures, as govembenefits wi further reduce work effort ments seek to support a greater eve among the edery, at east somewhat. In of expenditures than can be financed genera, studies have not attempted to through current tax receipts, which anayze differences in the impact that wi eventuay interfere with ecodifferent approaches to constructing the nomic growth; or socia wefare system might have. Reducing benefits for current benefi- Studies of the effect of taxation on ciaries-who might have arranged work effort have, in genera, aso shown their economic affairs differenty had ony weak inkages. Higher tax rates they known that promises were not may have as much of an impact on com- going to be kept. (Benefit reductions piance as on actua work effort. In prin- can occur either directy through cipe, tax payments that are inked to egisated reductions or indirecty future benefit increases might introduce through inadequate adjustments for ess of a work disincentive, but the effect, infation.) if any, appears to be sight. A number of economists have exam- Of the various approaches, socia ined the impact that the pay-as- insurance and empoyee mandates are you-go socia security system in the unique in mobiizing countervaiing United States has had on individua sav- poitica pressures to constrain program costs. As discussed earier, the possibiity of an increase in socia insurance contribution rates to baance revenues and expenditures heps to create a poitica constituency for restraining benefits. The inkage between current contribution rates and future benefit eve is even more obvious under empoyee madates.16 In contrast, under other approaches, the cost of socia wefare programs is frequenty ess visibe. Cost increases in programs financed out of genera revenue generate pressures on poiticians to devise ways to finance the additions to the tota budget. But, they do not transate directy into tax increases that voters wi reaize are directy inked to deveopments in socia wefare programs. Fisca discipine can be preserved if appropriate budget procedures can be adopted and enforced. But, these procedures operate on tota spending, and do not prevent runaway costs in a socia wefare program from crowding out expenditures on other governmenta activities. Simiary, sooner or ater, increases in socia wefare costs imposed on empoyers wi show up as smaer increases in wages. In this way, workers wi utimatey pay for the socia wefare benefits that their empoyers are required to provide. But, the inkage is far ess direct and far ess obvious to workers. Thus, workers cannot be counted on to form an effective constituency for restraining the costs of empoyer mandates. Poiticians are not the ony members of society prone to over promising. The same phenomenon can affect private sector institutions. Left to themseves, empoyers may promise their workers that smaer wage increases today wi aow for improved pensions in the future and then never actuay set aside the resources necessary to assure payment of the pensions. Simiary, competition for business may cause investment companies to undertake risky investments in the hope that they can produce a better return than their competitors. In either case, the State may eventuay find that it is required to fufi inadvisabe promises made by private sector institutions. The irony, then, is that where the State wishes to rey on private sector institutions, it Socia Security Buetin Vo. 57,. 3 Fai

8 may be forced to reguate these institutions cosey to assure that they respect the fisca discipine of the market pace. The status of the nationa govemment s budget can aso infuence both the structure of socia wefare institutions and the economic effect of socia wefare poicies, For one thing, poitica choices may be infuenced by a desire to minimize the costs that appear in the budget. A major disadvantage of universa programs is that they invove substantia budget outays. Measured by budget outays, socia insurance programs are amost as expensive as universa programs; presumaby, they wi be equay expensive if coverage were as extensive. One potentia advantage that socia insurance has, however, is that by construction, it brings its own revenue stream with it. This doesn t change the tota amount that government must raise or wi spend, but it may make raising revenues easier. An advantage of means- tested programs is that they require far ess of pubic resources. The great attraction of empoyer and empoyee mandates is that their cost does not appear directy in the government budget (though the cost of empoyer mandates may be refected indirecty in ower business tax coections). Even when they must be suppemented by gapfier programs, mandated programs are ikey to appear cheaper. Another important impact of govemment budget poicy invoves the inkage, if any, between retirement savings and nationa savings. History gives many exampes of socia insurance programs in which arge surpus baances were diverted to other government uses, either directy through poiticay motivated investments or indirecty through govemment borrowing of the socia insurance surpus to cover deficits esewhere in the budget. In either event, government actions serve to offset the increased savings and capita formation that woud otherwise have been supported by retirement program surpuses. It may be that governments are somewhat more ikey to spend reserves in pubic sector programs than they are those in private sector programs. If this is the case, and if retirement program surpuses are to be used as a mechanism for increasing nationa saving, use of mandated private sector programs woud have a somewhat higher probabiity of actuay producing the desired economic resut. Faciitating market adjustments.- Though economic growth benefits society as a whoe, frequenty, growth disrupts the ives of particuar peope and institutions. Less productive firms and industries must be aowed to shrink to make room for more productive firms and industries. Such changes cause jobs to move from one firm to another and, perhaps, from one geographic area to another. Minimizing the impact of such economic shifts on the socia wefare of the popuation is worthwhie in itsef, and it wi increase the poitica acceptabiity of the economic growth and change process. The approaches to constructing the socia wefare system most vunerabe to disruption as a resut of economic growth and change are those tied cosey to par- ticuar empoyers-either the empoyer mandate or the vountary, empoyerbased systems. The other approaches break the ink between socia wefare benefits and any one particuar empoyer, thereby heping to insuate the system from the disruption of change. Concuding Comments Pubic socia wefare systems have become more compex and varied over the years. The eariest systems focused on what we now think of as means-tested assistance programs. Govemmentsponsored socia insurance emerged about a century ago. Athough vountary empoyer-provided socia wefare benefits predate socia insurance, the idea of reying systematicay on empoyer mandates is a ater deveopment. And empoyee mandates are even more recent. For many years, socia wefare debates were dominated by the set of concerns associated with providing effective protection, particuary coverage and cost distribution. More recenty, as the socia wefare systems have grown to account for substantia portions of government budgets, socia wefare debates have broadened to incude questions about the wise use of administrative resources and the impact that these systems are having on the wider economy. The growth in the set of options avaiabe for constructing these systems, aong with the broadening of objectives some think these systems ought to serve, has ed to debates that are often voca but not aways productive. The premise of this paper is that such debates wi be much more productive if their participants agree that there is no one correct way to organize a socia wefare system. The various aternative approaches have different strengths and weaknesses. Most options are we suited to achieving one or more particuar socia goas, but are not particuary we suited to achieving another. We shoud not be surprised if socia wefare systems continue to become both more diverse and more compex. Since societies seek to achieve a variety of socia objectives simutaneousy, we shoud not be surprised to find them turning increasingy to mixtures that utiize some combination of the various approaches outined here, rather than reying primariy on one approach. And, since different societies attach different importance to different goas, at any given time we shoud expect to find different structura arrangements favored at different paces. And, finay, since socia goas change over time, we can expect that the particuar mixture of approaches empoyed in any one pace wi aso evove and change over time. In this paper I have suggested a framework for discussing the various options. I have aso suggested some of the advantages and disadvantages of using each of the approaches to achieve each of the possibe socia objectives. In some cases, my comments have been supported by empirica evidence; in other cases, they have been more specuative. In a cases, they are more apt to refect experience in Western Europe and rth America than esewhere. 1 trust that others wi be abe to modify or eaborate on my anaysis based on their own studies and their own country s experiences, and that in doing so, they wi hep us a to understand better these important issues. in Socia Security Buetin Vo. 57,. 3 Fa 1994

9 tes I If we were speaking about socia wefare poicy more broady, we woud aso have to consider a seventh approach, direct provision of socia wefare services by the government. This is a common strategy for deivering heath services, but not reevant to a discussion of pensions. 2 This ist is adapted from the definition of socia insurance deveoped by the Committee on Socia Insurance Terminoogy of the Commission on Insurance Terminoogy of the American Risk and Insurance Association. This group aso specifies that socia insurance coverage must extend beyond government empoyees. See Robert J. Myers, Socia Security, 3d ed., Bryn Mawr, PA: McCahan Foundation, 1985, pp i In some countries, such as the United States, socia insurance contributions are indistinguishabe from taxes since they are specified in aw and coected by the taxing authority. The connection between taxes and contributions is not quite so cose in some other countries. For exampe, each German sickness fund (technicay, a private sector institution) estabishes its own contribution rate and coects a socia insurance contributions from its members. Aso, vountary suppementa contributions are aowed under severa countries pension programs. 4 Socia insurance deveopments around the word suggest that this is an argument more appicabe in some poitica and socia systems than in others. Whereas these arrangements appear to have heped assure fisca discipine in Germany or the United States, different institutiona arrangements may be required in other countries to prevent over promising. ( Socia insurance aso tends to offer a uniform package of benefits, with empoyers having the option to provide separate, suppementa programs. h As a practica matter, this may be more important for heath insurance than pensions. A good case can be made for subsidizing heath insurance costs for those with ow current period income. It is ess cear whether one woud want to subsidize pension contributions for those with ow current period income. Presumaby in pension programs redistribution is more appropriatey based on ifetime income rather than current period income. For exampe, see Nicoas Barr, Economic Theory and the Wefare State: A Survey and Interpretation, Journa ofeco- nomic Literature, Vo. 3,. 2, June 1992, pp Barr deveops the foowing objectives of the wefare state: macro effciency, micro efficiency, economic incentives, poverty reief, protection of accustomed iving standards, income smoothing, vertica equity, horizonta equity, dignity, socia soidarity, inteigibiity, and absence of abuse. Since the major socia purpose of retirement programs is to aow peope to retire, these programs wi necessariy produce some reduction in work effort reative to the situation that woud exist in their absence; if they did not produce any reduction in work effort, they woud have faied. Other things equa, however, one woud want them to not discourage work effort prior to retirement or to prevent those who decide to retire from suppementing their retirement incomes, shoud they desire to do so. 9 r does the caw back poicy appy to Canadian heath insurance, which aso foows the mode of a universa program. See U.S. Genera Accounting Office, Socia Security: Anaysis of a Proposa to Privatize Trust Fund Reserves, Washington, DC: U.S. Government Printing Office, See Henry J. Aaron, The Socia Insurance Paradox, Canadian Journa of Economics and Poitica Science, Vo. 32,. 3, August 1966, pp I2 Experience in the United States iustrates the reative magnitudes. The annua administrative costs for the U.S. Od-Age and Survivors socia insurance program average about 0.8 percent of annua benefit payments. By comparison, the administrative costs of the parae means-tested program for the aged and disabed operated by the same agency averages about 7.6 percent of benefit payments. I3 rth American heath insurance iustrates the potentia size of the gap between private and pubic sector institutions (and among pubic sector institutions empoying different approaches). The Canadian nationa heath insurance system foows the universa mode and experiences administrative costs of just under 1 percent of tota outays. The U.S. nationa heath insurance for the aged foows the socia insurance mode and empoys more compicated provider payment mechanisms; its administrative costs run about 2 percent of outays. U.S. private heath insurance companies administrative costs average 10 to 12 percent of outays. See U.S. Genera Accounting Office, Canadian Heath Insurance.. Lessons from the United States. (GAOiHRD , June 1991) Washington, DC: U.S. Government Printing Office, 199 I, p. 3 1, I4 For exampe, Henry J..4aron, Economic Effects of Socia Security, Washington, DC: The Brookings Institution, 1982; and Michae V. Leonesio, Socia Security and Oder Workers, Socia Security Buetin, Vo. 56,. 2 (Summer) 1993, pp Presumaby, the rest of the increase in pension assets is either offset by arger iabiities esewhere in the economy or repaces asset woud have occurred in the absence of private pensions. For exampe, Aicia H. Munne, The Economics of Private Pensions, Washington, DC: The Brookings Institution, One can see the importance of this effect in recent reforms in German heath care financing. Germany uses contributory socia insurance to hep finance persona heath care services, and in recent years, has enacted a series of reforms that have reduced physician incomes and pharmaceutica company revenues. These reforms coud be achieved even though important interest groups were being forced to absorb economic osses because they were advertised as necessary to keep worker contribution rates from rising. In other words, the poitica infuence of those paying the contributions offset the poitica infuence of those benefiting from the program. The same effect was iustrated in the United States in 1983, when the Congress adopted imited taxation of Socia Security benefits and a higher retirement age as part of a package to restore fisca baance to that program. Neither proposa woud have been adopted had not the aternative been another increase in the Socia Security contribution rate. Socia Security Buetin Vo. 57,. 3 Fa