SOCIAL SECURITY ACT (1935)

Ideas and Institutional Change in Social Security: Conversion, Layering, and Policy Driftn

Daniel Béland, University of Calgary

Objectives. In recent years, social scientists such as Kathleen Thelen and Jacob Hacker have introduced new concepts to assist in the understanding of institutional change. Fostering some of these concepts, this article proceeds to augment the theoretical debate on institutional change in social science and policy research. A discussion of Social Security development in the United States advances the article’s main objective: to uncover the relationship between ideational processes and policy development. Methods. Qualitative and historical analysis is offered to examine three major policy episodes: the enactment of the 1939 amendments, the first mandate of the Nixon Administration (1969–1972), and the push for Social Se- curity privatization that emerged in the 1990s. Results. First, the analysis suggests that, through the process of institutional conversion, the 1939 amendments and the Nixon-era reforms altered the nature of Social Security. Second, the discussion on Social Security privatization stresses the impact of layering and policy drift on public and private pensions. Conclusions. The concepts of conversion, layering, and policy drift receive further empirical support through the presented analysis. Moreover, this article suggests that, for a full understanding of institutional change, a systematic analysis of ideational processes is necessary.

Scholars have long debated the nature of institutional change in policy processes. Although there is no consensus about what the main sources of policy change are, recent scholarship by researchers such as Kathleen Thelen and Jacob Hacker has provided social scientists working on public policy with useful analytical models to explore ‘‘how institutions evolve’’ (Thelen, 2004; Streeck and Thelen, 2005; Hacker, 2004). Yet, because the empirical evidence backing these models remains limited, there is ample room for empirical studies that could provide more ground to (or challenge) these models. Similarly, but back at the primary theoretical level, more scholarship is required on the effect that the relationship between institutional change and powerful economic, political, and ideological forces have on policy outcomes.

nDirect correspondence to Daniel Béland, Department of Sociology, University of Cal- gary, 2500 University Dr. NW, Calgary, Alberta, Canada, T2N 1N4 hdbeland@ucalgary. cai. The author is willing to share all data with those wishing to replicate the study. The author thanks Shauna Kadyschuk, Angela Kempf, Robert L. Lineberry, and three anonymous reviewers for their comments.

SOCIAL SCIENCE QUARTERLY, Volume 88, Number 1, March 2007 r2007 Southwestern Social Science Association

Contributing to the theoretical debate on institutional change, the following analysis will underline the relationship between ideational processes and policy development through a discussion of the political history of Social Security in the United States, with a specific focus on the 1939 amendments, the reforms enacted during the first mandate of the Nixon Administration, and the push for Social Security privatization that emerged in the 1990s. First, the analysis will suggest that the 1939 amendments and the Nixon-era reforms altered the nature of Social Security in what are two cases of institutional conversion. Second, looking at the debate over Social Security privatization, the discussion will underline the impact of layering and policy drift on pension institutions. Overall, this analysis will show how a systematic attention to ideational processes can shed more light on institutional change and, more precisely, on conversion, layering, and policy drift. Specifically, the analysis will ex- plain why one type of institutional change takes place rather than another, and discuss how ideas help in understanding the direction that such change may take.

The article is divided into four sections. The first will explore the concepts of conversion, layering, and policy drift at the center of the new theories of institutional change before showing how ideas can impact policy develop- ment. The last three sections will respectively explore the 1939 amendments, the Nixon-era reforms, and the debate over Social Security privatization. Because these episodes have already been documented in the existing his- torical literature, the analysis will remain concise, focusing mainly on the- oretically significant issues and processes. The conclusion will return to the literature on institutional change and underline the theoretical contribution of the article.

Institutional Change and Ideational Processes

Historical institutionalism is grounded in the assumption that a histor- ically constructed set of institutional constraints and opportunities affects the behavior of political actors and interest groups involved in the policy process (Immergut, 1998; Lecours, 2005; Orloff, 1993; Orren and Skowro- nek, 2004; Pierson, 1994; Steinmo, Thelen, and Longstreth, 1992; Weaver and Rockman, 1993). In contrast to sociological and rational choice insti- tutionalisms, historical institutionalism focuses on asymmetrical power re- lations as well as the impact of long-term institutional legacies on policy making (Hall and Taylor, 1996). Until recently, historical institutionalism had not paid much attention to institutional change (Clemens and Cook, 1999). This is especially true in the field of policy analysis, where the concept of path dependence has informed much of the historical institu- tionalist research (e.g., Pierson, 1994). In recent years, however, institu- tionalist scholars have begun to offer comprehensive theoretical accounts of

Ideas and Institutional Change in Social Security 21

how institutions change over time (Hacker, 2004; Orren and Skowronek, 2004; Streeck and Thelen, 2005; Thelen, 2003, 2004).1

As the title of her book suggests, Kathleen Thelen’s How Institutions Evolve sketches a systematic theory of institutional change (Thelen, 2004). One of the most powerful aspects of that book is the critique of the punc- tuated equilibrium model of institutional change. Punctuated equilibrium rests on the assumption that long episodes of institutional inertia follow rare critical junctures during which exogenous shocks provoke swift path-de- parting institutional transformations. Thelen does not reject the concepts of critical junctures and path dependence; rather, she convincingly argues that most forms of institutional change occur beyond such critical junctures and that in many contexts, endogenous mechanisms of change are more influ- ential than exogenous ones. This means that, in order to understand policy change better, social scientists should pay greater attention to incremental change occurring during long episodes of—relative—political stability.

In her book, Thelen identifies two main mechanisms of change: layering and conversion. On the one hand, layering involves ‘‘the grafting of new elements onto an otherwise stable institutional framework. Such amend- ments . . . can alter the overall trajectory of an institution’s development’’ (Thelen, 2004:35). The integration of private savings accounts into a tra- ditional pay-as-you-go pension system is a classic example of layering in the field of social policy (Thelen, 2003:277). On the other hand, conversion is about adopting new goals or bringing in new actors that alter the institu- tional role or the core objectives of an institution. The 1960s transformation of American social assistance programs into tools of racial equality is an example of social policy conversion (Thelen, 2003:229; see also Weir, 1992).

In a recent contribution on institutional change and social policy, Jacob Hacker borrows several concepts from Thelen in order to reinforce the argument that significant path-departing reforms may occur outside critical junctures and exogenous shocks (Hacker, 2004). Writing about social policy development in the United States, he argues that a series of low-profile processes have slowly transformed the nature of that country’s welfare re- gime. This argument contradicts the common wisdom that enduring policy legacies have favored strong institutional inertia in spite of powerful con- servative attacks (Pierson, 1994). Hacker argues that, in addition to layering and conversion, there is another crucial mechanism of change that can gradually transform the meaning and the role of existing institutional ar- rangements. Labeled policy drift, this third mechanism refers to the slow alteration of such arrangements due to changing socioeconomic circum- stances. For Hacker, new economic and social trends can make existing institutions become less and less adequate in the absence of significant

1To this list, one could add Paul Pierson’s Politics in Times (2004). Yet, this book focuses more on continuity than change, and only the fifth chapter is devoted to institutional change.

22 Social Science Quarterly

legislative reforms designed to adapt them to these changing circumstances. From this point of view, the inaction of policymakers who could take the necessary measures to sustain and enhance policy legacies in changing times constitutes a form of political behavior that can ironically lead to significant institutional transformations in a context of a rapid economic and social change. A major example of policy drift analyzed below is the current shift from pay-as-you-go to defined-contribution occupational pensions in the United States.

Thelen and Hacker have done much to improve our understanding of the politics of institutional change. Yet, they leave several questions unanswered. First, why does one type of institutional change take place rather than another? For example, although Hacker assumes that policy drift is likely to occur when obstacles to institutional revision are strong, his discussion about this crucial issue remains limited (Hacker, 2004). Second, authors such as Hacker and Thelen do not systematically address the following question: What factors explain the direction that policy drift and incremental change take? In his article on policy drift, Hacker essentially depicts policy drift as a conservative device that reinforces social inequality and increases economic insecurity. Yet, incremental change is not always conservative in nature, which means that one must explain the direction it takes. This article sug- gests that in order to answer these questions and understand institutional change fully, one must recognize the central role of ideational processes in politics and policy making. Because institutional change is generally related to the strategies of concrete social and political actors, understanding the effect of their ideas and assumptions on the social and economic world is essential for explaining the way in which these actors can bring about in- stitutional change in a particular policy area, and the form and orientation this change will take. As the analysis below suggests, the study of institu- tional change must take into account these actors’ beliefs and assumptions, which often take the form of a specific policy paradigm (Hall, 1993). For Peter Hall, a policy paradigm is ‘‘a framework of ideas and standards that specifies not only the goals of policy and kind of instruments that can be used to attain them, but also the very nature of the problems they are meant to be addressing’’ (Hall, 1993:279). Although policy paradigms are generally technical and low profile in nature, some of them can become known outside narrow expert and policy circles through the elaboration of widely mediated policy blueprints (Blyth, 2002).

Related to concrete economic and political conditions, paradigm shifts and more limited changes in social and economic assumptions can help explain the nature and the overall direction of policy change. Arguing that paying close attention to the ideas of those involved in institutional change is crucial, the following historical analysis provides more ground to the in- creasingly influential claim that ideas can shape the content of policy pro- posals and the perception of interests at the heart of political struggles (e.g., Blyth, 2002; Hansen and King, 2001; Lieberman, 2002; Weir, 1992). Yet,

Ideas and Institutional Change in Social Security 23

such a claim does not mean that ideational processes explain everything, and that traditional factors like electoral competition are irrelevant to the analysis of policy change. In fact, the analysis below will take electoral competition into account when exploring the sources of institutional change in American Social Security development. As evidenced below, the 1939 amendments and Nixon-era reforms were largely the product of the interaction between elect- oral competition and paradigm shifts affecting the socioeconomic or actuarial assumptions forming the background of policy debates and processes. As for the recent debate on Social Security privatization and related forms of layering and policy drift, they are unintelligible without a systematic understanding of the now mainstream conservative economic assumptions that shape the per- ception of interests at the center of the current struggles over public and private pension reforms in the United States. Overall, paying such close at- tention to changing policy ideas is necessary to explain the nature and the direction of institutional change across the three policy episodes.

Conversion, Phase 1: From Fiscal Conservatism to Family Protection

As enacted in 1935, Social Security constituted a federal social insurance scheme grounded in conservative actuarial assumptions. The policy para- digm that guided the elaboration of this program encouraged the emergence of a distinct set of principles regarding both general revenue financing and the contributory method. For President Franklin Delano Roosevelt, Social Security could exist only as a self-supporting program (Altmeyer, 1965; Leff, 1983; Witte, 1963). Rejecting general revenue financing, Roosevelt backed the idea of a strict actuarial relationship between the contributions of work- ers and employers, on the one hand, and the level of social benefits, on the other hand. For him, the contractual model tied to the contributory method would grant ‘‘the contributors a legal, moral, and political right to collect their pensions.’’ Because of those ‘‘earned rights,’’ he argued, ‘‘no damn politician can ever scrap my social security program’’ (Schlesinger, 1959:308–09; see also Perkins, 1946:281–83). Overall, the domination of the president’s conservative fiscal paradigm meant that benefits would re- main relatively modest.

A significant episode underlines the central role of President Roosevelt’s fiscal conservatism in the months preceding the enactment of the Social Security Act. Unsatisfied with the first draft of the bill, President Roosevelt ordered his Secretary of the Treasury, Henry Morgenthau Jr., to request that the Committee on Ways and Means modify the program’s tax schedule in order to avoid the possible use of general revenue financing after 1965 (Morgenthau, 1935:901–02). Following Morgenthau’s recommendations, the committee increased the initial tax rate and altered the tax schedule so that the rate would increase faster than anticipated, reaching a maximum of 6 percent in 1949 instead of 5 percent in 1957. As a result, the anticipated

24 Social Science Quarterly

size of the trust fund would increase by more than 300 percent (Béland, 2005:89).

Interestingly, a major political debate about the future size of this trust fund would only emerge more than a year after the signing of the Social Security Act in August 1935. This debate came about during the 1936 presidential campaign, as Republican candidate Alf Landon publicly at- tacked Social Security, which he labeled a ‘‘cruel hoax.’’ For Landon, the accumulation of a large reserve fund would lead to governmental mis- spending on the part of Democrats. ‘‘We have good spenders in Washington . . . . With this social security money alone running into billions of dollars, all restraint on Congress will be off’’ (Landon, 1936). As a result of this misspending, Landon claimed, money from the new federal payroll tax would not be used to pay Social Security benefits. Despite the defeat of Landon and the reelection of President Roosevelt in November 1936, the debate over Social Security funding did not dissipate.

Escalating the ongoing debate over Social Security funding was the August 1937 recession, which some economists argued was caused by the payroll tax that had taken effect in January of the same year (Berkowitz, 1983). To assess and contend with this controversial issue, the administration finally launched the Advisory Council on Social Security (1937). For the Chairman of the Social Security Board, Arthur Altmeyer, what became the 1937–1938 Advisory Council represented a major political opportunity for the presi- dent. ‘‘As a matter of fact, I think it is possible not only to offset these attacks on the Social Security Act, but really to utilize them to advance a socially desirable program, fully in accord with present fundamental prin- ciples underlying the Social Security Act and within our financial capacity’’ (Altmeyer, cited in McKinley and Frase, 1970:358).

In its 1938 report, the Advisory Council recommended a major reduction in the size of the trust fund through a change in the tax schedule and, more importantly, a major expansion of Social Security benefits (U.S. Advisory Council on Social Security, 1938). Shifting the focus of the program from fiscal conservatism to social adequacy and the protection of the family unit as a whole, the 1937–1938 Advisory Council advocated a more redistribu- tive vision of social insurance that implicitly contradicted the individualistic actuarial model adopted in 1935. J. Douglas Brown, who served as Chair of the Advisory Council, later summarized the nature of the paradigm shift that shaped the content of the 1939 amendments: ‘‘The Advisory Council of 1937–1938 shifted the whole concept of what became the OASDI program from a hybrid compromise between private savings and social insurance to a clear-cut concept of social insurance. The new focus became adequacy and the protection of the family unit’’ (Brown, 1972:136).2 The enactment of a

2The 1935 version of the program had a redistributive element in the form of a weighted benefit formula. According to this formula, federal old age insurance would offer higher replacement rates to lower-income workers.

Ideas and Institutional Change in Social Security 25

legislation grounded in this more redistributive yet low-profile policy paradigm became possible because the Roosevelt Administration accepted repudiating the strict social insurance model enacted in 1935—a decision that reflected the growing social and political opposition in the aftermath of the 1937 recession and the 1938 congressional elections (Polenberg, 1975).

To save his Social Security program, Roosevelt succumbed to pressures demanding the abandonment of fiscal conservatism. Reflecting this com- promise, Roosevelt finally asked Congress for reform in January 1939 (Roosevelt, 1939). Because the main political actors had already agreed to reduce the anticipated size of the trust fund in exchange for the increase in ‘‘family protection,’’ the legislative process proved to be unproblematic (Berkowitz, 1983). Against the financial imperatives imposed by the ad- ministration less than five years earlier, Social Security emerged as a much more redistributive program than the one established in 1935. No longer centered on actuarial equity and advance funding, the new version of the program included spousal and survivor benefits, two measures rooted in traditional gender roles (Berkowitz, 2001; Kessler-Harris, 1995; Mettler, 1998). These two measures favored income redistribution from singles to married couples (Berkowitz, 2001). From this perspective, it is not an ex- aggeration to talk about institutional conversion in the case of the 1939 amendments. Although the main actors involved in that ‘‘deal’’ remained quiet about the true scope of the reform they supported (Berko- witz, 1983), it represented a genuine paradigm shift in Social Security. This fact challenges the punctuated equilibrium model of policy development as the reform occurred well after the expansionist phase of the New Deal had ended. Furthermore, the 1939 reform underlines the role of ideas in institutional change, as there is clear evidence that the dominance of the family protection paradigm helped reshape some of the program’s core objectives. Finally, the intensification of electoral com- petition during President Roosevelt’s second mandate made the 1939 com- promise possible in the first place, as the weakened Democratic administration had little choice but to bargain with its opponents in order to preserve Social Security.

All in all, the emergence of a low-profile paradigm promoting economic redistribution and grounded in traditional gender roles paved the way to the 1939 amendments. In other words, a particular set of ideas largely explains the redistributive orientation of the institutional conversion that took place in 1939. Furthermore, these ideas, coupled with a changing electoral context and Republican pressures to shrink the trust fund, help explain why institutional conversion—and not layering or policy drift—occurred in that particular situation. Ultimately, new policy ideas enabled by a changing political environment made legislative action not only possible but unavoidable. This resulted in reform through formal institutional conversion.

26 Social Science Quarterly

Conversion, Phase 2: Social Security as a ‘‘Retirement Wage’’

In the United States, World War II favored a return to economic pros- perity and a strengthening of conservative forces in Congress. This situation resulted in a decade of relative legislative inaction in the field of Social Security reform, subsequently ushering in a gradual decline in the real value of benefits. Eventually, however, to offset the negative effects of inflation, the 1950s witnessed the enactment of several pieces of Social Security le- gislation that increased benefits (Achenbaum, 1986; Altmeyer, 1965; Béland, 2005; Tynes, 1996). Furthermore, the 1950s saw the enactment of a disability insurance scheme as part of Social Security (Berkowitz, 1987). Less than a decade later, in 1965, the enactment of Medicare, a program directly related to Social Security, represented another major step in the incremental expansion of the federal social insurance system (Corning, 1969; Marmor, 1973). Yet, as far as the old-age pension com- ponent of Social Security was concerned, the first mandate of the Nixon Administration constituted a major turning point that brought about a series of legislative actions gradually altering the very meaning of the program (Myles, 1988).

Similar to the context of reform in the late 1930s, intensified electoral competition in the 1970s also helped create conditions conducive to in- stitutional conversion. Seeking to reduce the capacity of Democrats in Congress to make electoral gains through ad hoc benefit increases, President Nixon pushed for the automatic indexation of Social Security (e.g., Nixon, 1971). In this context, Wilbur Mills (D-Arkansas) and other influential Democratic congressmen opposed this measure in order to maintain their capacity to claim credit for the ad hoc benefit increases (Weaver, 1988; Zelizer, 1998:312–46).3 To fight the apparent generosity of President Nix- on’s indexation proposal, Mills and his allies promoted the enactment of unprecedented benefit hikes that would significantly increase the real value of Social Security benefits. Three factors facilitated the adoption of these increases in the late 1960s and early 1970s. First, in the context of the Vietnam War, the Democratic leadership in Congress could include benefit hikes in major war-related bills, which President Nixon could not always afford to veto (e.g., Weaver, 1988:166). Second, the declining confidence in big business (Edsall, 1984:113) and occupational pensions (Hacker, 2002:145–53) increased support for more generous Social Security bene- fits while undermining possible conservative opposition to the expansion of a popular federal program that now included the middle class (Quadagno,

3Another factor undoubtedly motivating Wilbur Cohen to promote such potentially popular benefit increases was his plan to run for the presidency in 1972.

Ideas and Institutional Change in Social Security 27

1991).4 Third, a new ‘‘optimistic’’ actuarial paradigm repudiated the ac- tuarial conservatism that had prevailed since the New Deal.

Rather than assuming that wages would not increase in the future, deci- sionmakers embraced a ‘‘dynamic’’ actuarial paradigm that produced consid- erable projected windfalls instead of the after-the-fact windfalls traditionally generated by conservative assumptions (Derthick, 1979:349–57). What made the actuarial situation of the early 1970s unique was that ‘‘policymakers could use these projected windfalls in addition to the after-the-fact windfalls still available at the time’’ (Béland, 2005:134). Supported by economists from the Brookings Institution and Social Security Commissioner Robert Ball, these new assumptions made room for major benefit hikes without the enactment of potentially unpopular short-term tax increases. Consequently, the emer- gence of this low-profile actuarial paradigm facilitated the enactment of mas- sive benefit increases by reducing their perceived fiscal and political cost (Derthick, 1979; Weaver, 1988). This is true because the new actuarial as- sumptions helped Democratic members of Congress provide a rationale for their ambitious benefit increases, which appeared more reasonable in the mirror of such assumptions. All in all, the new actuarial paradigm largely explains the progressive direction of the institutional conversion that reshaped Social Security during President Nixon’s first mandate.

Considering these factors and the Democratic push for major benefit increases before the enactment of automatic indexation, Congress approved a series of benefit hikes that extended far beyond the inflation rate: 13 percent in 1967, 15 percent in 1969, 10 percent in 1971, and 20 percent in 1972 (Weaver, 1988). These ad hoc increases boosted the average replace- ment rate for Social Security before Congress finally enacted automatic indexation in 1972.5 According to John Myles, the changes adopted be- tween 1969 and 1974 transformed Social Security into a ‘‘retirement wage’’ program similar to the social insurance schemes adopted in western Europe during the postwar era (Myles, 1988).

As in 1939, intense electoral competition and a low-profile paradigm shift led to a genuine form of institutional conversion in Social Security, thus favoring a rapid expansion of the program. During President Nixon’s first mandate, optimistic actuarial ideas facilitated the triumph of the liberal expansionist agenda, thereby affecting the direction of change. The conver- gence of ideational and electoral factors created a unique opportunity for significant legislative revision, leading to a genuine form of institutional conversion, rather than a more indirect form of change such as policy drift.

4On the growing popularity of Social Security in the postwar era, see Brain (1991). 5For example, between only 1969 and 1971, the real value of benefits (i.e., net of inflation)

increased by 23 percent. After indexation became effective in 1974, however, high inflation rates coupled with cost-of-living adjustments (COLAs) further augmented the real value of benefits. As an illustration, in 1975 the average replacement rate for Social Security reached 55.9 percent. This figure was up from 33.5 percent in 1965 and 40.3 percent in 1967 (Myers, 1993:363).

28 Social Science Quarterly

Therefore, like the genesis of the 1939 amendments, this episode underlines the relationship between ideational processes, electoral competition, and institutional change in Social Security development.

Layering, Policy Drift, and Social Security Privatization

By the mid-1970s, the combined effects of inflation, high unemployment, and a flawed indexation system favored the emergence of a fiscal crisis in Social Security (Light, 1995; Snee and Ross, 1978). Considering the pop- ularity of this program and the powerful constituencies it created (Campbell, 2003; Pierson, 1994), major direct cutbacks were excluded from the legis- lative agenda and, in 1977, Congress enacted payroll tax hikes coupled with indirect benefit cuts such as the alteration of the indexation formula (‘‘de- coupling’’), which only affected future retirees—later known as the ‘‘notch babies’’ (Pierson and Weaver, 1993:117). Several years later, new legislative actions became necessary as enduringly high inflation and unemployment rates created a second fiscal crisis in Social Security (Kingson, 1984). After making several strategic mistakes (Stockman, 1986), President Reagan fi- nally put together the Greenspan Commission, which reached a last-minute agreement over the measures necessary to ‘‘save’’ Social Security (National Commission on Social Security Reform, 1983). A combination of modest tax hikes and benefit cuts, the 1983 amendments to the Social Security Act also included a gradual increase in retirement age that would take place between 2000 and 2022 (Light, 1995). Overall, the amendments did not alter the nature and the institutional goals of Social Security.

In the aftermath of the 1983 amendments, conservative actors rejecting the redistributive nature of Social Security began sketching a long-term political strategy that could undermine support for the program, leading to Social Security privatization, which refers to the shift from defined benefits and pay-as-you-go financing to defined contributions and advanced funding taking the shape of personal savings accounts (Quadagno, 1999; Teles, 1998). It is impossible to understand the core objectives of this strategy and the direction that the Social Security debate would take after 1990 without analyzing the financial paradigm that guides conservative efforts to privatize that federal program. Related to the old liberal support for individualism and market forces against economic redistribution, this financial paradigm seeks to increase personal responsibility and financial investment at the expense of ‘‘big government’’ security. More concretely, this policy para- digm states that investing payroll money in equity is beneficial to the econ- omy as well as to future retirees. This is true because privatization would create higher ‘‘return rates’’ than the current pay-as-you-go program (e.g., Ferrara and Tanner, 1998). Rejecting the redistributive logic of Social Se- curity, conservatives argue that this new federal savings scheme would stimulate economic growth by increasing national savings rates. As opposed

Ideas and Institutional Change in Social Security 29

to the family protection and the actuarial paradigms leading, respectively, to the 1939 and the early 1970s reforms, the new financial paradigm is as- sociated with a high-profile public discourse that is ever present in con- temporary media reports and political debates. More recently, the ideas at the center of this paradigm have been featured prominently in George W. Bush’s Ownership Society blueprint (Béland, 2005).

To transform Social Security in the sense of this individualistic and mar- ket-oriented financial paradigm, a number of conservative experts and pol- iticians argued that only incremental changes in private benefits coupled with a relentless ideological campaign against Social Security could lead to its partial or full privatization. Ironically described as a ‘‘Leninist Strategy’’ (Butler and Germanis, 1984), this long-term approach to Social Security reform includes institutional layering, as conservatives support the multi- plication of defined-contribution schemes like 401(k)s, which constitutes the explicit model for Social Security privatization (Teles, 1998).6 By pro- moting the expansion of these schemes through new tax provisions, con- servatives and their allies in Congress have helped create a massive retirement savings system at odds with the pay-as-you-go and the defined- benefit logics inherent to Social Security (Hacker, 2002).7

The development of tax-deferred 401(k) voluntary savings schemes began in the late 1970s, intensifying during the 1980s and 1990s. Conservatives hold the development of these schemes as proof that the existing Social Security program is obsolete and that privatization would bring efficiency to the program, as well as higher returns to the vast majority of workers (Teles, 1998). From this perspective, layering is an instrument of conservative pol- icy change that favors the promotion of specific policy ideas aimed at con- vincing citizens and interest groups that it is in their interest to support Social Security privatization and the related financial paradigm.8 The dom- ination of this paradigm coupled with the exceptional financial perform- ances of the mid-1990s contributed to the construction of the economic interests of the financial industry, which generally supported privatization.9

6Personal savings have had a long-standing presence as a major component of the Amer- ican pension system. Recent conservative efforts have only aimed at encouraging their growth in order to reduce the reliance on Social Security and defined-benefit pensions.

7The following quote underlines the rapid growth of 401(k)s during the 1980s and 1990s: ‘‘Although enabled by legislation in 1978, 401(k) plans effectively were not adopted until the Internal Revenue Service issued clarifying rules in 1981. Since then, they have grown re- markably and become the primary vehicle for retirement saving. In 1996, 33 percent of all private pension assets, 33 percent of all pension plans, and 45 percent of all active pension participants were in 401(k)s. The $104 billion in 401(k) contributions represented 61 per- cent of all pension contributions and 38 percent of National Income and Product Account (NIPA) personal saving that year. Benefits paid from 401(k)s represented 38 percent of total pension benefits disbursed’’ (Engelhardt, 2001:1).

8On the argument that ideas shape the perception of economic interests, see Blyth (2002). 9By the late 1990s, however, the support for Social Security privatization within the

financial industry had already started to decline as tangible legislative proposals showed the administrative problems related to this policy alternative (Rose and Celarier, 1999).

30 Social Science Quarterly

The push for Social Security privatization also broadly represents yet another facet of the conservative crusade against ‘‘big government,’’ which has be- come increasingly influential in the United States since the 1970s.

Although conservatives failed to convince members of Congress to enact Social Security privatization legislation in the 1990s (Derthick, 2001; Weaver, 2005), impressive stock market performances did help them push this policy alternative onto the agenda (Teles, 1998). Yet, popular support for Social Security remained strong, and conservatives failed to convince the majority of citizens that privatization would serve their interests (Cook, Barabas, and Page, 2002). In his 1999 State of the Union Address, President Clinton openly rejected privatization; but, considering the ideological weight of the financial paradigm, he felt compelled to propose the creation of new tax-sponsored savings accounts alongside Social Security (Clinton, 1999). Even though this initiative went nowhere in the context of the Lewinsky scandal (Weaver, 2005), it underlines the central role of layering in the politics of Social Security.

Nevertheless, the politics of Social Security in the United States is also about what Jacob Hacker calls policy drift (Hacker, 2004). As employers increasingly support the development of defined-contribution pensions in order to replace costly defined-benefit schemes, private pensions are drifting away from the logic inherent to Social Security. Related to cost-saving strategies on the part of employers, this form of policy drift is transforming private benefits despite the absence of path-breaking legislative actions: economic and managerial logics favor the transformation of the contract between workers and employers in the sense of a model of protection strik- ingly similar to policy alternatives labeled as Social Security privatization. Although the long-term impact of this form of policy drift on Social Se- curity is unknown, it exacerbates the growing clash between this program and tax-sponsored private pensions (Hacker, 2002; Klein, 2003).10 This tendency is in itself a significant form of institutional change through policy drift, despite the absence of legislation on Social Security privatization.11

Interestingly, the financial paradigm mentioned above is promoting the multiplication of defined-contribution pension schemes, as they implicitly serve as models for Social Security privatization. To a certain extent, this paradigm thus contributes to the individualistic and financial orientation of contemporary policy drift.

All in all, one could argue that although the powerful constituencies created by the postwar expansion of Social Security created major political

10In 2003, the Pension Benefit Guaranty Corporation (PBGC) ‘‘insured about 29,500 single-employer defined benefit plans, down from an all-time high of 112,000 plans in 1985. This decline primarily reflects a large number of terminations among small plans’’ (PBGC, 2003:11).

11Furthermore, in the context of growing income inequality, more income is earned above the Social Security earnings cap, which, in the long run, is weakening the program’s fiscal foundation.

Ideas and Institutional Change in Social Security 31

obstacles to privatization, the enduring influence of the financial paradigm and the incremental development of personal savings accounts in the private sector have transformed the American pension system in a profound way. Through layering and policy drift, the private, yet government-regulated, pension institutions have changed significantly since the 1980s. In turn, these trends strongly affect the perception of Social Security, a program whose postwar expansion was directly related to the development of tax- sponsored, defined-benefit pension schemes in the private sector (Hacker, 2002; Klein, 2003). With the decline of these schemes and the multipli- cation of defined-contribution plans and savings accounts, the defined-ben- efit model at the foundation of Social Security seems more and more marginal institutionally (Hacker, 2004). Only time will tell if layering, policy drift, and the ideological campaign of the right will succeed in mak- ing Social Security privatization appear unavoidable. As President Bush’s doomed 2005 privatization crusade demonstrates, institutional obstacles to privatization remain formidable (Béland, 2005). The Bush Administration’s failure to directly reform Social Security could encourage conservative pol- iticians and policy experts to ‘‘invest’’ even more in layering and policy drift, two major sources of incremental change in the contemporary American pension system. In short, the development of Social Security itself remains path dependent. Such path dependence subsequently encourages conserva- tives to support alternative forms of institutional change, such as layering and policy drift.

This remark points to the reason why layering and policy drift have dominated the contemporary pension reform scene in the United States. Because obstacles to formal institutional revision and conversion are strong, indirect change is the only option for political and economic actors seeking to implement conservative ideas. From that perspective, the tension between the conservative paradigm and blueprint on the one hand, and enduring institutional obstacles on the other hand, explains why conservatives have promoted layering and policy drift so extensively in order to reshape the American pension system. As for the conservative direction of contemporary policy change (i.e., individualizing protection and shifting economic risks onto workers), it reflects the imperatives of the financial paradigm that has become increasingly prominent in American policy debates since the early 1990s. High-profile conservative ideas prepare the ground for low-profile and indirect, yet conservative, change.

Discussion

This article has offered a dynamic understanding of Social Security de- velopment in the United States. Far from maintaining the same institutional path since its enactment in 1935, Social Security has changed over time in major ways. In 1939, the program became far more distributive than

32 Social Science Quarterly

originally conceived. During the Nixon era, increases in benefits trans- formed a relatively modest public pension scheme into a genuine income maintenance program. Since the 1990s, politically meaningful changes in government-regulated private benefits have not formally altered Social Se- curity; rather, they have reshaped the economic and institutional environ- ment of this federal program, thus altering its status and meaning in American society. As evidenced above, the first two episodes of institutional change described here correspond to what Kathleen Thelen labels institu- tional conversion, a process by which the major objectives of a policy are transformed. Concerning the third episode, it is clearly a case of layering, as conservatives supported the development of private savings schemes like 401(k)s alongside Social Security in order to weaken institutional support for that program and persuade the population that financial investment is the best way to ‘‘save’’ Social Security. Yet, consistent with the scholarship of Jacob Hacker, this article argues that policy drift is a core aspect of the new politics of Social Security. Namely, the absence of formal reform, coupled with the presence of social and economic change, combine to reshape oc- cupational pensions closely related to Social Security.

Beyond the more dynamic view of Social Security development and the new empirical evidence supporting the concepts of layering, conversion, and policy drift, this article contributes to our understanding of institutional change through a discussion about the role of ideas in public policy. What is crucial to understand here is that ideational processes can become a major source of institutional change, at least as far as Social Security development is concerned. More specifically, the analysis of ideational processes helps answer two questions raised in the theoretical section. The first question concerns the factors that explain why one type of policy change takes place rather than another. Regarding this issue, the discussion about the late 1930s and the early 1970s episodes shows that the convergence between the for- mulation of new policy ideas and the intensification of electoral competition are conducive to institutional revision and conversion. During both epi- sodes, intensifying electoral competition and the influence of ideas neces- sitating legislative revision reinforced one another to create the conditions for genuine institutional conversion. As for the more current situation, the growing influence of the financial paradigm in the context of strong in- stitutional obstacles to Social Security privatization this paradigm supports have transformed layering and policy drift into the main sources of con- servative policy change in American pension policy. Here, layering and policy drift become the main instruments of policy change for conservatives precisely because the legislative revision of Social Security that their financial paradigm supports remains elusive.

The second question deals with the direction that policy change takes. In 1939, as well as in the late 1960s and early 1970s, a paradigm shift con- cretely affected the overall direction of the institutional conversion that, on both occasions, affected Social Security. In 1939, the ‘‘family protection’’

Ideas and Institutional Change in Social Security 33

paradigm grounded in traditional gender roles helped reconstruct Social Security as a redistributive program. Three decades later, a change in the actuarial paradigm (i.e., the shift from conservative to ‘‘dynamic’’ assump- tions) made generous benefit hikes easier to implement for members of Congress. This resulted in a major expansion of Social Security. As for the current layering process involving the multiplication of private savings ac- counts, the financial paradigm dominant among conservatives provides a strong ideological rationale for this process and for the crusade against the redistributive logic of Social Security. Consequently, the analysis of the three episodes backs the claim that ideas play an instrumental role in determining the overall direction of policy change. Finally, such an analysis underlines the distinction between high- and low-profile policy ideas. As evidenced above, the ‘‘family protection’’ and the actuarial paradigms that respectively impacted the 1939 and the early 1970s reforms represented a set of low- profile policy ideas hardly debated outside expert and bureaucratic circles. In contrast, the contemporary financial paradigm has become a highly debated set of assumptions linked to a prominent policy blueprint: George W. Bush’s Ownership Society.

As this article suggests, recognizing the central role of ideas in processes of institutional change identified by Kathleen Thelen and Jacob Hacker does not mean that ideas are the only locus of policy development. In the first two episodes analyzed above, for example, electoral competition between the two main parties and/or between the presidency and Congress represented a major driving force for institutional change. Furthermore, in the case of the recent debate on Social Security privatization, the economic interests of employers and the financial industry constitute a key aspect of the political process. Yet, because these interests are constructed through ideational pro- cesses related to the financial paradigm, there is a close relationship between the politics of ideas and the politics of interests. In the future, scholars conducting research on institutional change should recognize the enduring weight of ideational processes in conversion, layering, and policy drift while paying attention to electoral struggles and the perception of economic in- terests that can weigh so heavily on policy making.

REFERENCES

Achenbaum, W. Andrew. 1986. Social Security: Visions and Revisions. Cambridge: Cambridge University Press.

Altmeyer, Arthur J. 1965. The Formative Years of Social Security. Madison, WI: Wisconsin University Press.

Béland, Daniel. 2005. Social Security: History and Politics from the New Deal to the Privat- ization Debate. Lawrence, KS: University Press of Kansas.

Berkowitz, Edward D. 1983. ‘‘The First Social Security Crisis.’’ Prologue 15(3):133–49.

34 Social Science Quarterly

———. 1987. Disabled Policy: America’s Programs for the Handicapped. New York: Cam- bridge University Press.

———. 2001. ‘‘Family Benefits in Social Security: A Historical Commentary.’’ Pp. 19–46 in Melissa M. Favreault, Frank J. Sammartino, and C. Eugene Steuerle, eds., Social Security and the Family: Addressing Unmet Needs in an Underfunded System. Washington, DC: Urban Institute.

Blyth, Mark. 2002. Great Transformations: Economic Ideas and Institutional Change in the Twentieth Century. Cambridge: Cambridge University Press.

Brain, Charles M. 1991. Social Security at the Crossroads: Public Opinion and Public Policy. New York: Garland Publishing.

Brown, J. Douglas. 1972. An American Philosophy of Social Security. Princeton, NJ: Princeton University Press.

Butler, Stuart, and Peter Germanis. 1984. ‘‘Achieving Social Security Reform: A Leninist Strategy.’’ Cato Journal 3:547–56.

Campbell, Andrea Louise. 2003. How Policies Make Citizens: Senior Citizen Activism and the American Welfare State. Princeton, NJ: Princeton University Press.

Clemens, Elisabeth S., and James M. Cook. 1999. ‘‘Politics and Institutionalism: Explaining Durability and Change.’’ Annual Review of Sociology 25:441–66.

Clinton, William. 1999. State of the Union Address. Washington, DC: The White House, Office of the Press Secretary.

Cook, Fay Lomax, Jason Barabas, and Benjamin I. Page. 2002. ‘‘Invoking Public Opinion: Policy Elites and Social Security.’’ Public Opinion Quarterly 66:235–64.

Corning, Peter A. 1969. The Evolution of Medicare . . . from Idea to Law. U.S. Department of Health, Education, and Welfare, Social Security Administration, Office of Research and Statistics, Research Report 29. Washington, DC: U.S. Government Printing Office.

Derthick, Martha. 1979. Policymaking for Social Security. Washington, DC: Brookings Institution.

———. 2001. ‘‘The Evolving Old Politics of Social Security.’’ Pp. 193–214 in Martin A. Levin, Marc K. Landy, and Martin Shapiro, eds., Seeking the Center: Politics and Policymaking at the New Century. Washington, DC: Georgetown University Press.

Edsall, Thosmas. 1984. The New Politics of Inequality. New York: W.W. Norton.

Engelhardt, Gary V. 2001. Have 401(k)s Raised Household Saving? Evidence from the Health and Retirement Study. Aging Studies Program Paper 24. Syracuse, NY: Center for Policy Research, Maxwell School of Citizenship and Public Affairs, Syracuse University.

Ferrara, Peter, J., and Michael D. Tanner. 1998. Common Cents, Common Dreams: A Lay- man’s Guide to Social Security Privatization. Washington, DC: Cato Institute.

Hacker, Jacob S. 2002. The Divided Welfare State: The Battle Over Public and Private Social Benefits in the United States. Cambridge: Cambridge University Press.

———. 2004. ‘‘Privatizing Risk Without Privatizing the Welfare State: The Hidden Politics of Social Policy Retrenchment in the United States.’’ American Political Science Review 98(2):243–60.

Hall, Peter A. 1993. ‘‘Policy Paradigms, Social Learning and the State: The Case of Eco- nomic Policymaking in Britain.’’ Comparative Politics 25(3):275–96.

Ideas and Institutional Change in Social Security 35

Hall, Peter A., and Rosemary C. R. Taylor. 1996. ‘‘Political Science and the Three Insti- tutionalisms.’’ Political Studies XLIV:936–57.

Hansen, Randall, and Desmond King. 2001. ‘‘Eugenic Ideas, Political Interests, and Policy Variance: Immigration and Sterilization Policy in Britain and the U.S.’’ World Politics 53:237–63.

Immergut, Ellen M. 1998. ‘‘The Theoretical Core of the New Institutionalism.’’ Politics and Society 26:5–34.

Kessler-Harris, Alice. 1995. ‘‘Designing Women and Old Fools: The Construction of the Social Security Amendments of 1939.’’ Pp. 87–106 in Linda K. Kerber, Alice Kessler-Harris, and Kathryn Kish Sklar, eds., U.S. History as Women’s History. Chapel Hill, NC: University of North Carolina.

Kingson, Eric R. 1984. ‘‘Financing Social Security: Agenda-Setting and the Enactment of the 1983 Amendments to the Social Security Act.’’ Policy Studies Journal 13(1): 131–55.

Klein, Jennifer. 2003. For All These Rights: Business, Labor, and the Shaping of America’s Public-Private Welfare State. Princeton, NJ: Princeton University Press.

Landon, Alf. 1936 ‘‘Text of Gov. Landon’s Milwaukee Address on Economic Security.’’ New York Times August 27:31.

Lecours, André, ed. 2005. New Institutionalism: Theory and Analysis. Toronto: University of Toronto Press.

Leff, Mark H. 1983. ‘‘Taxing the ‘Forgotten Man’: The Politics of Social Security Finance in the New Deal.’’ Journal of American History 70(2):359–79.

Lieberman, Robert C. 2002. ‘‘Ideas, Institutions, and Political Order: Explaining Political Change.’’ American Political Science Review 96(4):697–712.

Light, Paul C. 1995. Still Artful Work: The Continuing Politics of Social Security Reform. New York: McGraw-Hill.

Marmor, Theodore R. 1973. The Politics of Medicare. Chicago, IL: Aldine Publishing Co.

McKinley, Charles, and Robert W. Frase. 1970. Launching Social Security: A Capture-and- Record Account, 1935–1937. Madison, WI: University of Wisconsin Press.

Mettler, Suzanne. 1998. Dividing Citizens: Gender and Federalism in New Deal Public Policy. Cornell, NY: Cornell University Press.

Morgenthau Jr., Henry. 1935. Hearings Before the Committee on Ways and Means, House of Representatives on H.R. 4120. Washington, DC: U.S. Government Printing Office.

Myers, Robert J. 1993. Social Security, 4th ed. Philadelphia, PA: Pension Research Council/ University of Pennsylvania Press.

Myles, John. 1988. ‘‘Postwar Capitalism and the Extension of Social Security into a Re- tirement Wage.’’ Pp. 265–84 in Margaret Weir, Ann Schola Orloff, and Theda Skocpol, eds., The Politics of Social Policy in the United States. Princeton, NJ: Princeton University Press.

National Commission on Social Security Reform. 1983. Report of the National Commission on Social Security Reform. Washington, DC: National Commission on Social Security Reform.

Nixon, Richard. 1971. ‘‘Special Message to the Congress on Social Security (September 25, 1969).’’ Pp. 740–45 in Public Papers of the Presidents of the United States, Nixon, 1969. Washington, DC: U.S. Government Printing Office.

36 Social Science Quarterly

Orloff, Ann S. 1993. The Politics of Pensions: A Comparative Analysis of Canada, Great Britain and the United States, 1880–1940. Madison, WI: University of Wisconsin Press.

Orren, Karen, and Stephen Skowronek. 2004. The Search for American Political Development. Cambridge: Cambridge University Press.

PBGC. 2003. Pension Insurance Data Book 2003. Washington, DC: PBGC.

Perkins, Frances. 1946. The Roosevelt I Knew. New York: Vikings.

Pierson, Paul. 1994. Dismantling the Welfare State? Reagan, Thatcher, and the Politics of Retrenchment. New York: Cambridge University Press.

———. 2004. Politics in Time: History, Institutions, and Social Analysis. Princeton, NJ: Princeton University Press.

Pierson, Paul, and, R. Kent Weaver. 1993. ‘‘Imposing Losses in Pension Policy.’’ In R. Kent Weaver and Bert Rockman, eds., Do Institutions Matter? Government Capabilities in the U.S. and Abroad. Washington, DC: Brookings Institution.

Polenberg, Richard. 1975. ‘‘The Decline of the New Deal, 1937–1940.’’ Pp. 246–65 in John Braeman, Robert H. Bremner, and David Brody, eds., The New Deal: The National Level. Columbus, OH: Ohio State University Press.

Quadagno, Jill. 1991. ‘‘Interest-Group Politics and the Future of the U.S. Social Security.’’ Pp. 36–58 in John Myles and Jill Quadagno, eds., States, Labor Markets, and the Future of Old-Age Policy. Philadelphia, PA: Temple University Press.

———. 1999. ‘‘Creating a Capital Investment Welfare State.’’ American Sociological Review 64(1):1–10.

Roosevelt, Franklin D. 1939. ‘‘Annual Message to the Congress, January 4, 1939.’’ Vol 8, p. 7 in Samuel I. Rosenman, ed., Public Papers and Addresses of Franklin D. Roosevelt. New York, Random House.

Rose, Darby, and Michelle Celarier. 1999. ‘‘Where’s the Payoff? Wall Street Spent Millions to Promote the Privatization of Social Security—An Idea Going Increasingly Awry.’’ Investment Dealers Digest August 9. Available at hhttp://www.davidlanger.com/articles/misc/ iddigest.8.9.99.htmli. Schlesinger, Arthur. 1959. The Coming of the New Deal. Boston, MA: Houghton-Mifflin.

Snee, John, and Mary Ross. 1978. ‘‘Social Security Amendments of 1977: Legislative History and Summary of Provisions.’’ Social Security Bulletin 41(3):3–20.

Steinmo, Sven, Kathleen Thelen, and Frank Longstreth, eds. 1992. Structuring Politics: Historical Institutionalism in Comparative Analysis. Cambridge: Cambridge University Press.

Stockman, David A. 1986. The Triumph of Politics: How the Reagan Revolution Failed. New York: Harper and Row.

Streeck, Wolfgang, and Kathleen Thelen, eds. 2005. Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford: Oxford University Press.

Teles, Steven. 1998. ‘‘The Dialectics of Trust: Ideas, Finance, and Pension Privatization in the US and the UK.’’ Paper presented at the Annual Meeting of the Association for Public Policy Analysis and Management. New York.

Thelen, Kathleen. 2003. ‘‘How Institutionalism Evolves: Insights from Comparative Historical Analysis.’’ Pp. 208–40 in James Mahoney and James Rueschemeyer, eds., Comparative Historical Analysis in the Social Sciences. New York: Cambridge University Press.

Ideas and Institutional Change in Social Security 37

———. 2004. How Institutions Evolve: The Political Economy of Skills in Germany, Britain, the United States, and Japan. Cambridge: Cambridge University Press.

Tynes, Sheryl R. 1996. Turning Points in Social Security: From ‘‘Cruel Hoax’’ to ‘‘Sacred Entitlement.’’ Stanford, CA: Stanford University Press.

U.S. Advisory Council on Social Security. 1938. Final Report. Washington, DC: Social Security Board.

Weaver, R. Kent. 1988. Automatic Government: The Politics of Indexation. Washington, DC: Brookings Institution.

———. 2005. ‘‘Public Pension Reform in the United States.’’ Pp. 230–51 in Giuliano Bonoli and Toshimitsu Shinkawa, eds., Ageing and Pension Reform around the World: Ev- idence from Eleven Countries. Cheltenham: Edward Elgar.

Weaver, R. Kent, and Bert Rockman, eds. 1993. Do Institutions Matter? Government Cap- abilities in the U.S. and Abroad. Washington, DC: Brookings Institution.

Weir, Margaret. 1992. Politics and Jobs. Princeton, NJ: Princeton University Press.

Witte, Edwin E. 1963. The Development of the Social Security Act. Madison, WI: University of Wisconsin Press.

Zelizer, Julian E. 1998. Taxing America: Wilbur D. Mills, Congress, and the State, 1945–1975. New York: Cambridge University Press.

38 Social Science Quarterly