Risk and the Labor Market - Springer Link

2 Risk and the Labor Market Societal Past as Economic Prologue Sanford M. Jacoby

According to many pundits, we are living in a "high-risk society;' a kind of postmodern frontier economy (Mandel 1996). Workers are being advised to take care of themselves and their kin because government, unions, and corporations are unwilling to shoulder as much risk as in the past. Public programs such as Social Security and Medicare are in fiscal distress; at the very least, it is unlikely that government will assume new social risks, such as national health insurance. Meanwhile, unions are a shrinking portion of the labor force, representing less than 9 percent of private-sector employees; few believe that they will soon stage a major revival. And corporations-on whom workers once counted for stable career jobs and generous "fringe" benefits-insist that those days are over and never cOming back. In this "brave new world," the savvy postmodern worker is, we are told, one who adapts to the uncertainties of a globalized and downsized economy by acquiring multiple skill sets, saving money in IRAs and portable, defined contribution plans, and maintaining a full panoply of insurance policies. Personal responsibility is much in vogue; the risk-sharing institutions of the 20th century's "organizational revolution" are in decline. This chapter uses the concept of risk to examine the development of modern labor-market institutions and to assess how they are changing in the present period. Although individuals face many kinds of risks, among the most immediate and pressing are the risks attendant upon the cessation oflabor-market income due to layoff, sickness, accident, retirement, or death of a wage earner. Over the last one hundred years, modern societies have developed a diverse set of institutions for pooling labor-market risks and indemnifying against them. The precise mix of Sanford M. Jacoby • Anderson School of Management, University of California, Los Angeles, California 90095. Sourcebook of Labor Markets: Evolving Structures and Processes, edited by Ivar Berg and Arne 1. Kalleberg. Kluwer Academic/Plenum Publishers, New York, 2001.

31

32 Sanford M. Jacoby

institutions varies over time and across nations, making this area a rich lode for sociological mining. Most studies of risk have been conducted by economists and, more recently, by cognitive psychologists. Individuals are said to possess "preferences" for different types of risks; they will act rationally to reduce their exposure to risk through tactics such as diversification and risk pooling (insurance). Despite the power of the economic approach - even the boundedly rational version elaborated by Simon, Kahneman, Tversky, and others-there are important issues with which it cannot grapple. First and foremost is the question of how individuals select risks to consider and their willingness to accept risk. Niklas Luhmann (1993), one of the few sociologists to analyze risk, argues that these elemental decisions are not only economic and psychological but also social. Think, for example, of the cultural and class differences in the perception of cigarette-smoking risk or, to cite examples closer to the concerns of this chapter, how perceptions of life insurance (Zelizer 1979) and job loss (Keyssar 1986) have changed over time. Another dimension of risk in which social factors loom large is the process by which individuals select the group with whom to pool their risk. This gets at questions of social identity and collective action that lie at the heart of sociology. As Giddens (1990, 1991) has observed, the development of risk and risk sharing is an important chapter in the history of modernity, part of the movement from gemeinschaft to gesellschaft. From the 18th century to the present, one can see a shift from personalized forms of risk sharing to more anonymous systems such as commercial insurance. Preindustrial risk sharing was based on kinship and face-to-face relationships, rooted in trust. The rise of urban industrial society in the 19th century caused a proliferation of more impersonal but still mutualistic risk-sharing groups such as friendly societies, burial societies (especially among freed slaves), trade unions, and fraternal insurance groups. Toward the end of the century, welfare capitalism enlarged the "circle of 'we' " (Hollinger 1995) to include employees of a given enterprise. Finally, nations witnessed the emergence of welfare states that extended the risk-sharing umbrella to all citizens. While the welfare state might seem the epitome of a bureaucratic and impersonal gesellschaft, its creation required some nontrivial amount of social solidarity. More than most insurance schemes, the welfare state entails redistribution not only to the less affluent but also to veterans, the elderly, and other groups. Hence, support for social insurance has varied over time as individuals have seen themselves as being more or less a part of the national collectivity with whom they have cast their fate (Baldwin 1990). The most modern of all forms of risk sharing is the commercial insurance purchased by atomistic individuals. Willingness to carry such insurance requires as a precondition that people respect the expertise of actuaries and other insurance professionals and, more significantly, that they place their trust in anonymous others who are part of their risk pool. Yet the relationship to these others is tenuous; neither solidarity nor redistribution characterizes commercial insurance schemes. The economic effect is the same as that achieved by the extended family of the 18th century, but the social relations are completely different. Sociological analysis has an advantage over more individualistic methodologies because it recognizes that the range of risk-sharing choices open to individuals is not wide but rather is constrained by the set of risksharing institutions that exist

in a given society at a given time. While individuals can form new risk-sharing institutions, the collective action problem increases the probability that they will rely on sets of existing institutions. Individual choice models usually can supply a rational explanation for particular risk institutions. But the explanatory models are ahistorical; they have little to say about why some institutions but not others appear on the menu of choices. For example, economists developed implicit contract theory in the 1970s to "explain" the American phenomenon of real wage insurance, that is, wage rigidity combined with temporary layoffs and unemployment benefits. This was said to be a form of insurance that reflected the relative risk preferences of workers (riskaverse) and employers (risk-neutral) (Azariadis 1975; Stiglitz 1987). Yet the theory could not account for the limited incidence and scope of real wage insurance (e.g., cost-of-living escalators) or of income replacement for workers on layoff (e.g., supplemental unemployment benefits). Nor could it account for the substantial variation in wage rigidity over time (rising after the 1930s and on the wane in recent years) and across space (more marked in the United States than elsewhere). At best, the theory asserted that these phenomena were due to variations in risk preferences, an explanation that comes perilously close to tautology. On the other hand, a more sociological approach can tell us how the range of risk-sharing institutions is generated; from this range come effective (as opposed to notional) preferences. In this sense, institutional arrangements make preferences endogenous Gacoby 1990). Understanding risk requires analysis of the actual processes that generate innovation and change in a society's risk-sharing institutions. Once a particular constellation of risk institutions establishes itself, the likelihood of other institutional options is reduced (e.g., the early development of a welfare state in Scandinavia made it harder for private insurance companies to establish themselves, whereas in the United States, the early development of private insurance constrained the expansiveness of the welfare state). Economic historians term this phenomenon path dependence (Arthur 1994; David 1985). This mode of reasoning is not unknown to sociology. Sociologists have long paid attention to the constraints of institutional sequencing (Bendix 1964) and, more recently, to the role played by "organizational fields" (Fligstein 1990). In what follows, I trace the rise of modern institutions for sharing labor-market risk. The analysis first shows the play between economic and social factors in the rise of U.S. risk-sharing institutions and then it demonstrates how differences in initial conditions produced disparate outcomes in Europe and the United States. In the United States, a weak state, an individualistic ethos, and social heterogeneity combined to produce a set of institutions that put the weight of risk-sharing on private parties rather than government. Even after the emergence of a welfare state in the 1930s, private institutions played a significant role in risk mitigation, more so than in Europe. The chapter then shifts to the contemporary dual crises of welfare capitalism and the welfare state, brought on by economic globalization and demographic shifts. In Europe, the response has been various attempts to shore up the welfare state; in the United States, there is greater interest in privatization of risk-sharing arrangements. The chapter's intent is to provide a more sociological understanding of the contemporary labor-market situation in the United States. But it also has a methodological objective, namely, to demonstrate the value of a sociological (compara-

33 Risk and the Labor Market


tive and historical) approach to analyzing labor-market risk. Economic sociology faces stiff competition from the neoclassical economic approach to industrial and financial markets. But it can-and has-made significant contributions to the analysis of that most peculiar and noncommodified sector of the economy: the labor market. MUTUALISM The concept of risk is quintessentially modern because it requires a modern sense of time to make a clear distinction between past and present. Not until the 17th century did Western societies possess the armamentarium of strict time measurement, such as schedules, clocks, and calendars (Zerubavel 1981). Risk analysis also required the development of the mathematics of probability, which allows us to conceive the future as humanly created or humanly controlled "risk" rather than as Providential danger (Luhmann 1993; Knight 1921). As Giddens (1991:111) remarks, "The notion of risk becomes central in a society which is taking leave of its past." Prior to the 19th century, formal institutions for pooling labor-market risk were relatively uncommon. While one could obtain fire and marine insurance, these products were purchased primarily by wealthy businessmen. The bulk of the American population remained tied to an agricultural economy organized around the family unit. Sons accepted the responsibilities of caring for elderly parents. Those sons who became economically independent of their parents" still continued to live in a community in which their parents, siblings, grandparents, uncles and aunts, and cousins also lived, thereby involving them constantly with people who were related to them by birth and marriage" (Greven 1970:138). In maritime New England, fishing families headed by men past the age of 40 depended heavily on the earnings of teenage sons (Vickers 1994). These kinship networks provided aid in times of accident, sickness, or death. Local communities also pitched in to assist bereaved widows and children. In addition to pooling risks with family and neighbors, another strategy to protect against risk was diversification of income. Farmers and their wives often took up some light manufacturing, either at home or as seasonal employment. The same strategy was pursued by those who took jobs in the factories that began opening up in the 1820s and 1830s. During those years, factory workers in Lynn, Massachusetts, supplemented their seasonal manufacturing wages by turning to livestock, gardening, and fishing (Keyssar 1986). A different diversification strategy was to possess a multiplicity of skills, to be a "well-rounded" craftsmen. Such versatility made a worker attractive to a large number of employers, while simultaneously offering some protection against technological change that might render one's narrow skill-but not an entire ensemble-obsolete (Gutman 1976; Jacoby 1985). As industrialization and urbanization proceeded during the 19th century, people began moving in ever-larger numbers from rural to urban areas of the United States. Starting in the 1840s, the workforce was expanded by immigrants from the British Isles, Germany, and then, in the 1880s, from Eastern and Southern Europe. These changes forced millions of formerly rural people to seek new ways of dealing

with the uncertainties of modern life. City-dwelling workers could no longer rely on homegrown food to get them through a spell of joblessness. The elderly, who were an important part of rural family life, found industrial firms reluctant to employ persons past their prime. Young, unmarried women began to work outside the home, raising parental concern for their morals. Meanwhile, dangerous factories and crowded cities brought on occupational injuries and other health problems. For much of the late 19th and early 20th centuries, the family was the first line of defense for urban wage earners. Families adopted a "defensive" mode of economic cooperation, pooling their incomes, sharing homes, and hoarding resources for rainy days brought on by illness. Death of a man in his prime earning years was far more of a threat then than it is presently. While mortality rates for 55- to 64year-old white men in 1880 were about 1.9 times what they are today, the ratio at age 25 was 4.3 times higher (Modell 1979). The family-and sometimes the extended family-served to protect against loss of the principal earner. Unemployment, too, was far more of a threat than today. Between 1854 and 1914, recessions or depressions occurred every 3 - 4 years. Even during prosperous times, about one-fifth of the industrial workforce was unemployed for at least part of the year Oacoby 1985; Keyssar 1986). Industrial cities were harsh places-breeding grounds for illness and anomiebut their complex division of labor created new possibilities for social solidarity. Durkheim (1933) wrote of what he saw: an industrializing society in which group identity-gender, race, occupation, corporation, class, and nation-was becoming more salient and broadly inclusive. Those who had strong group identities and a need for help increasingly turned to other members of their group for aid in protecting against the risks of modern life. The first of these mutualistic groups were fraternal societies that emerged not only in the United States but also in other parts of the industrializing world. Starting in the 1840s, men from various social classes flocked to groups such as the Masons and the Odd Fellows. Among native-born Protestant men, antebellum fraternalism was "more popular than any other social organization except churches" (Laurie 1995:106). These organizations offered a combination of sociability, moral uplift, and mutual insurance. For a modest initiation fee and a small monthly contribution, members were eligible for death benefits sufficient to purchase a burial plot and, perhaps, additional funds to assist a family suffering from premature loss of its male breadwinner. Some societies offered sickness benefits. The early fraternals comprised a heterogeneous mix of skilled workers, selfemployed artisans, and proprietors. Ties of gender, race, and nativity transcended those of class in an era when the boundary between working-class and middle-class remained somewhat fluid and indistinct (Dumenil 1984). Those shunned by the WASP fraternals-the Irish, German Jews and African Americans-launched their own fraternal organizations such as the Royal Order of Hibernians, the B'nai Brith (started in 1843), and the Grand United Order of True Reformers, which spread across the South after the Civil War. Later in the century national fraternals opened separate branches for Jews and African Americans, such as the Prince Hall Masons and the Odd Fellows, to which some Jews and over 300,000 African Americans belonged in the 1890s (Muraskin 1975; Weare 1993). Hispanics of the Southwest also had their own fraternal organizations, the mutualistas (Zamora 1992; Sheridan 1992).

35 Risk and the

Labor Market


As Gould (1995) has argued, people have multiple identities; the identity that individuals choose to act on collectively depends, in part, on the social networks to which they belong. During the 19th century, skilled workers and lower-middle-class proprietors lived in the same neighborhood and shared a set of republican values based on hard work, sobriety, and antipathy to concentrated wealth (Thomas 1983). They disliked the new Irish immigrants, made the lodge an abstemious substitute for the tavern, and contested the notion that women were the moral guardians of the home (Clawson 1989; Roediger 1991). Thus, the fraternal lodge was the site not only for risk pooling but also for cross-class male bonding - and race bondingaround a set of shared values. And in recursive fashion, membership in a fraternal society reinforced shared values by creating salience and boundaries (Calhoun 1997). Fraternal organizations explicitly emphasized their cross-class identities. The Ancient Order of United Workmen (AOUW), founded in 1868 by a mechanicMason troubled by labor unrest on the railroads, presented itself as an alternative to militant trade unionism. By the end of the century, the AOUW had over 350,000 members, all white men, who came from a melange of working- and lower-middleclass occupations. 1 The Knights of Pythias said that they intended "to soften down asperities of life [and] bind in one harmonious brotherhood men of all classes and opinions" (in Cordery 1996:98). Fraternals sometimes acted as a barrier to the formation of class identity and trade unionism. But their secret gatherings and smallholders' antipathy to concentrated wealth also served as training grounds for future union leaders and for organizations such as the Knights of Labor (Voss 1993). As skilled workers increasingly found themselves engaged in wage labor, they turned to a different organization-the trade union-for risk pooling and other activities. Like fraternal societies, trade unions were based on common gender, ethnic, and racial identities. But whereas the fraternals usually were occupationally heterogeneous, the trade union brought together only those individuals who labored in the same craft. With the shift to factory employment, skilled workers faced greater risk from unemployment and injury than middle-class men but were less capable of dealing with these risks on their own. They had become what historian Peter Baldwin (1990) calls a "risk group;' actors "identified and given interests in common by their shared relations to the means of security" (p. 10). Risk groups are marked by the incidence of shared risk; coalescence occurs when the group's members are unable to shoulder these risks independently. While a risk group is similar in many respects to a social class, there is no presumption of fixed interests. Solidarity with other groups-as in the case of the Knights of Labor or heterogeneous fraternal societies - can wax and wane depending on the structure of risk institutions, the group's sense of solidarity with other groups, and its members' ability to fend for themselves. Given these facts, it is hardly surprising that the earliest trade unions were concerned with unemployment, a hazard on which fraternals did not focus because its incidence was so unevenly distributed across their membership. Trade unions were well equipped to administer assistance to jobless members, since they had information about whether a member was truly unemployed. Many of the early craft unions kept "out of work" lists; matching the unemployed with jobs was one of the primary functions of a union's business agent. Knowledge of conditions in the trade also permitted unions to administer tramping or traveling benefit systems,

which facilitated the movement of unemployed members to more prosperous regions by providing loans, information, and certificates of unemployment. Some British unions paid monetary benefits to their jobless members, but this was not as systematically practiced by American unions, who, at best, might maintain a small loan fund for the unemployed or provide special relief during depressions (although the Cigar Makers administered an elaborate out-of-work insurance program for its members) (Keyssar 1986; Ulman 1956; Henderson 1908). Trade unions also provided accident, illness, and death benefits. For example, the Carpenters offered a plan for paying lump sums in case of partial or total disability due to accidents, while the Cigar Makers paid a lump sum if a member became blind or lost both hands. Only a few unions provided sickness benefits, such as the Machinists and the Iron Molders, who each had modest plans. An even smaller group, notably, the Plumbers and the Engineers, provided old-age pensions to their members, in part to lessen the incentive for older members to work for lower-than-union wages. On the other hand, death benefits-what we would call life insurance - were quite common among trade unions that competed directly with fraternals and private companies in this area (Henderson 1908:85-99; Webb and Webb, 1920:147). Benefits programs mitigated the free rider problem associated with unionism and also helped the union retain its members over the course of a business depression (van der Linden 1996). But trade union benefits programs did not have a reputation for security or solvency. Many unions regarded their insurance responsibilities as secondary to other objectives. There was no legal bar to prevent a trade union from tapping its insurance funds to pay strike benefits or to defend against a lawsuit-such funds were not based on a contractual agreement with individual members-with the result that some unions consumed their funds. Because of these problems, trade unionists hedged their bets by purchasing additional insurance from other sources, either a fraternal society or an industrial insurance company.2 During the years of American industrialization-from 1880 to 1930-the American Federation of Labor's (AFL) craft unions excluded unskilled and semiskilled workers, many of them immigrants from Eastern and Southern Europe. Such exclusion was part of a larger pattern of keeping immigrants out of public as well as social and cultural life. Fraternal organizations, social halls-even the "better" saloons and movie theaters-were open only to "Americans" (Morawska 1987). Thus, as a matter of necessity - not only of choice - immigrant workers relied on their fellow countrymen to help them survive in the New World. Initially, the key risk-sharing unit was the extended family and/or village friendship network, which assisted immigrants in finding housing and jobs (Versteegh 1996). As the new communities reached a critical mass, the immigrants started-or restarted-their own ethnic fraternal and mutual benefit societies. 3 All immigrant groups had their own fraternal organizations, although they were especially prevalent among Bohemians, Slovaks, Italians, Jews, and Poles. Some of these fraternals were branches of organizations that had been started in the homeland. 4 For example, there were direct links between Slovakian medieval smithing guilds, 19th-century Slovakian fraternals in Europe, and Slovakian immigrant societies in the United States (Bodnar 1985). Other fraternals were based on communal practices that had existed throughout Europe since the Middle Ages. In



rural areas, it was an ancient custom to pool resources to pay for funeral expenses, since elaborate funerals were considered greatly important, as was the desire to avoid the indignity of a pauper's burial. The ethnic fraternals benefitted from the favorable climate for fraternalism and voluntarism that existed in the United States, and borrowed ideas and practices from native organizations such as the AOuw, and from fraternals started earlier in the century by immigrants who by now were solid members of the middle class. The landsmanshaftn of the post-1880 Jewish immigrants from Eastern Europe sometimes affiliated with existing Jewish organizations such as the B'nai Brith, hoping thereby to garner social prestige, better benefits, and more rapid acculturation of their members (Soyer 1997). Mutual assistance based on shared ethnicity was a common feature of immigrant fraternals, but they fractured along myriad other dimensions such as religion, occupation, and local or regional origin (Beik 1996). Italian immigrants formed over 6,000 fraternal societies by the 1890s; Chicago alone had about 160 such groups. While the fraternals usually included members from different economic classes, they were rife with invidious social distinctions. Fraternal officers usually came from the ranks of artisans and entrepreneurs, not the more proletarian social strata. Some Hungarian fraternals explicitly specified in their charters that only those who had been manual workers in Hungary-not peasants-could serve as officers (Bodnar 1985). Still, these were member-run organizations: The lodge was the basic unit, analogous to the union local and possessed of the same potential for direct control and democracy. Fraternal lodges organized the social lives of their members. There were athletic teams, parades, picnics, and other cultural events. In Chicago, the Italian paesani "met to talk politics, play bocce, and sponsor an annual celebration in honor of their village's patron saint" (Cohen 1990:69). Occasionally, the fraternals became embroiled in political struggles in the homeland or in labor disputes in the New World, as happened in the New York garment industry in the early 1900s (Soyer 1997). And, of course, the fraternals almost always concerned themselves with mutual insurance. Benefits were related to labor-market risks, chiefly the death of the breadwinningpaterjamilias, as well as his injury, illness, or disability. The fraternals were male organizations in a male-dominated milieu (most did not admit women as regular members until well into the 20th century). Hence, while spousal benefits were available, wives typically were insured through their husbands (Henderson 1908:114). Fraternal benefits were similar to insurance offered by private companies but were embedded in a vastly different social context, one that was opposedsometimes explicitly-to commercial principles. Fraternal insurance was based on membership, not contract, with members recruited by word of mouth rather than by agents working on commission (Zelizer 1979). This made administrative costs about one-third to one-sixth smaller than those incurred by commercial insurers. While most commercial insurers paid profits to shareholders (the exception were the mutual insurance companies, who distributed profits to shareholders), fraternals directed their surplus funds to the members they served, redistributing funds to the poorest strata; providing scholarships, camps, and tuition loans for members' children; and operating their own vocational training and recreational facilities. Large amounts were spent on real estate investments in the local community, either

via direct loans to members or by purchasing mortgages from banks and building associations (Renkiewicz 1980; Soyer 1997). By helping the immigrant get settled in the new land, the fraternals may unwittingly have undermined their own long-term vitality. But they also secured their base by constantly reinforcing ethnic identity and religion, and linking them. Many of the Polish fraternals had close ties to their local parishes, the Catholic clergy, and to national churches. Membership in some fraternals occurred through the affiliation of entire parishes and church communities. Tight inclusion implied strict, even fanatical, exclusion. The Polish clergy, for example, regularly attacked the more secular Polish fraternals for admitting atheists, infidels, Jews, anarchists, socialists, and Masons- "the names by which the enemy was known" (Renkiewicz 1980:119). Another problem for the fraternals arose from the fact that they were poorly managed. Dues were unrealistically low, life tables were ignored, and reserves were inadequate to cope with crises such as the influenza epidemic of 1918. As a result, serious financial problems were common among fraternals (Clough 1946). To cope with these difficulties, fraternals began rationalizing and bureaucratizing their operations, emulating their commercial competitors. They hired full-time agents, actuaries, and managers; raised dues; refused to issue new policies to the elderly; and ended the practice of insuring wives through their husbands. Small village societies merged with each other or joined national organizations. Also impelling these changes was regulation by state insurance departments and the steady growth of commercial life insurance in immigrant communities.

COMMERCIAL INSURANCE Today we take for granted the commercial insurance industry, although the industry'S existence in its present form dates back less than three hundred years. Marine and fire insurance were offered by the earliest concerns, such as the First American Company, established by Benjamin Franklin in 1752, and Uoyds of london, which opened in 1771. The rise of the insurance industry was closely related to developments in mathematics and statistics, and to the systematic gathering of quantitative data (Bernstein 1996). Thus, the industry's growth was part and parcel of other changes we associate with modernity: scientific empiricism, the rationalization of everyday life, and the control of time-what Giddens (1991:111) calls the "colonization of the future." Life insurance developed more slowly than other kinds of insurance, not appearing in the United States until the 1840s. Sluggish growth was not due to technical impediments-life tables and other actuarial technologies were readily available - but instead stemmed from a cultural aversion to commercialiZing human life. To place financial value on a life was thought to be immoral and sacrilegious; paying money to mothers and children was regarded as an immoral commodification of fatherhood (Zelizer 1979). Some feared that life insurance would encourage murder for money or that it would undermine nonmarket risk-sharing systems. Insurance offered by fraternal societies suffered from these attitudes far less than commercial insurance; fraternal principles of mutual aid and moral uplift cohered with the precepts of the premodern moral economy. And fraternal insur-



ance did not depend on abstract actuarial systems or trust in anonymous othersanother hallmark of modern society (Giddens 1991). Not surprisingly, social groups that were least attached to premodern values-the new, urban middle classeswere the first to purchase commercial life policies. To attract working-class consumers, commercial insurers in the 1870s began offering inexpensive "industrial insurance" so called because it was intended for sale to the industrial classes (Hoffman 1900:3). Whereas ordinary life insurance required hefty annual payments in the thousands of dollars, industrial insurance was available for small amounts, payable in cents per week. 5 Every member of the family could be insured, not just the household head, as with ordinary insurance. Yet premiums per dollar of industrial insurance were more expensive than ordinary insurance because the beneficiaries had higher death rates than the holders of ordinary policies Qames 1947:85). A key feature of industrial insurance was the agent, who came every week to a workers' home to collect the required payments. By the 1880s, companies such as Prudential, Metropolitan, and John Hancock each had thousands of agents in the field, collecting nickels from the insured and persuading others to sign up (Clough 1946:151; James 1947:78). Agents exhorted workers that it was virtuous, not immoral, to purchase life insurance, and persuaded them to trust profit-oriented companies. Prudential Insurance Company, the first in the industrial market, considered the agents' job "missionary work ... necessary to acquaint the masses with the facts." The Prudential agent was said to be "the true friend of those with whom he comes in contact. Few men more thoroughly understand the ways and means of the wage-earning population, and few have better opportunities to extend the teachings of the gospel of thrift" (Hoffman 1900:15, 148). Despite this enormous marketing effort, commercial life insurance periodically encountered resistance. In the late 1880s and 1890s, critics condemned juvenile life insurance because it allegedly created an incentive to commit infanticide. A number of states enacted bans on child insurance that stayed in effect until the 1920s (Hoffman 1900:46-48; James 1947:85, 122). The commercial industry also suffered from revelations of fraud involving smaller, fly-by-night insurers. Such problems-and the strength of ethnic and other group identities-ensured the continuing popularity of the fraternals, which in 1895 still had more insurance in force than commercial firms (Cordery 1996; Zelizer 1979). Some commercial companies tried to appear less avaricious by organizing as mutual insurance companies, which meant that policyholders "owned" the company and received profits in the form of reduced premiums. Others stressed their ties to fraternal and community organizations. The North Carolina Mutual Life Insurance Company, the largest black-owned insurance company in the United States, was started in 1881 by a former slave who had been active in a black mutual benefit society known as the True Reformers. North Carolina Mutual recruited its agents from the ranks of the True Reformers and created a positive image in the black community by sponsoring an all-Negro bank, a newspaper, a Colored State Fair, and a model cooperative farm (Weare 1993). Most active in community service activities was Metropolitan Life. Under Lee Frankel, a sociologist hired by Metropolitan in 1909, the company started a welfare division that provided industrial policyholders with health information and free services such as visiting nurses (Dublin 1943:421-441) Another way of establishing trust was the practice of pay-

ing claims on policies that had lapsed during a labor dispute, something most fraternals could not afford to do Oames 1947:121). Commercial insurers never missed an opportunity to sow doubts in the public mind about the solvency of fraternals, though they were careful to do this without appearing to attack the fraternals directly. This was one of the agent's prime tasks: "He must convince [fraternal members] of the unsoundness of a dozen wellmeant cooperative schemes ... which were his competitors; and he must not disparage these societies too much because, overlooking amateurish management of their insurance schemes, most of them filled a real need in the lives of working people" Oames 1947:74). Fraternal members responded with the now-familiar strategy of multiple policyholding, spurring the growth of industrial and commercial life insurance. In 1909, 2 out of 9 people in the United States were covered by industrial policies, and by 1918,7 out of 8 families had some kind oflife insurance (Keller 1963:14; Modell 1979). Now commercial insurers became less worried about competition from fraternals; hedging bets by holding multiple policies showed that fraternal and commercial insurance did not have to be competitive but could coexist. Some commercial companies even began to sell policies through the fraternals, usually under the fraternal's name. *

*

In Europe, commercial insurance ran up against a barrier that was of lesser consequence in America: the state. While the new American nation had its Hamiltons who wanted to tame the market and preserve rights via a strong state, it was the antistatist economic liberals whose ideas were dominant. The "American Creed", as Lipset (1996:28) calls it, promoted "personal responsibility, independent initiative, and voluntarism," along with "self-serving behavior, atomism, and a disrespect for the communal good." With the exception of Civil War veterans' pensions, the government before the New Deal offered little in the way of social insurance, whether for workers, mothers, or the elderly. In fact, the federal government possessed a relatively limited apparatus for administering any large-scale program (Skocpol 1993; Skowronek 1982). In continental Europe, liberalism was weak, certainly weaker than ]acobin and Hegelian tendencies to view the nation-state as the guardian of the public good. As inheritors of monarchial privileges, the European states stood on a moral plane superior to private interests. In the newly consolidated nation-states, the government played a major role in developing the economy and providing protection against the risks of industrial life. The motives for welfare legislation varied: from Bismarck's Bonapartist attempts to co-opt the working class to the Catholic corporatism seen in France and Italy. Political coalitions also varied: from the fusion of middle and agrarian classes in Scandinavia to the alliance between labor and the new liberals in Britain. But all looked to the state to provide citizens with insurance (redistributive in varying degrees) against risk. And once the state became active in this area, interest groups organized their risk strategies around the existence of public programs. Employer paternalism, mutual benefit societies, and commercial insurance never became so elaborate as in the United States; this was true even of Britain, which, with respect to public-private apportionment, was situated between the United States and the continent (Mann 1997; Dreyfus 1996; EspingAndersen 1990; Dobbin 1994).



Another factor causing social insurance to take hold earlier and with greater force in Europe was foreign trade. The European economies were more dependent on trade, which was subject to fluctuations that heightened the instability of employment and, therefore, agitation for unemployment insurance (Harrington 1998). That organized labor could press for such insurance-and for other social benefitsreflected the strength of the European labor movements and their state-focused social programs. With the exception of a brief period after World War I, American unions had neither national power nor interest in social insurance, which would have threatened their own voluntarist approach Gacoby 1991). Moreover, social solidarity in America was hampered by the great divides of race and ethnicity. The AFL's members were native-born or from Northern Europe and the British Isles, whereas the unskilled were Southern Europeans, Mexicans, and African Americans-groups perceived as unreliable and socially inferior. In fact, the strength of ethnic fraternalism in the United States was the reverse image of the nation's exclusionary unions and the inability of the middle classes to find common identity with the classes below them (except in ethnically homogeneous places such as the Upper Midwest). Conversely, the European nations had fewer immigrants and greater ethnic uniformity, which supported the kind of cross-class alliances that were vital to the creation of social insurance schemes. Some Europeans thOUght social insurance would strengthen their nation's racial characteristics, uplifting the poor and stemming population decline, an impulse that led to ugly eugenic programs to sterilize those with "inferior" characteristics. The left, including the Swedish Social Democrats, were as prone to these impulses as right-wing fascist parties such as the Nazis (Wolfe and Klausen 1997; "Sweden's Dirty Secret" 1997). By the early 20th century, U.S. commercial insurers began worrying that European social programs might spread to America. Insurance executives regularly sent expert delegations across the Atlantic to study the new social insurance schemes. (Lee Frankel was hired by Metropolitan after a European study trip that led to publication in 1910 of his book, Workingmen's Insurance in Europe.) At the state level, insurance companies fought heated battles to privatize the new workers' compensation systems, albeit with mixed success (Fishback and Kantor 1996). Men like Haley Fiske, vice president of Metropolitan Life, were unnerved at the prospect that government might someday preempt them. As Fiske told a gathering of Metropolitan managers in the early 1900s: Gentlemen, the influence of laws like those in Germany and England is bound to extend .... I do not want the State to take over this function; but it is going to take it over unless the people insure themselves.... Demagogues ... raising hopes that never can be met ... must be withstood. In my judgment, the only way to withstand them as far as insurance goes is to build up a great voluntary society.... If we had twenty million [insured] ... I would feel safe. (quoted in James 1947:180)

But it was unlikely that Metropolitan and other companies could ever reach 20 million workers using the one-on-one sales methods of industrial insurance. Another technique, one that took advantage of economies of scale, was needed. This turned out to be group insurance or what has come to be known as "welfare capitalism."

WELFARE CAPITALISM

Welfare capitalism consisted of a simple precept: that the business corporation, rather than government or mutual organizations, should be the primary source of security and stability in modern society. By the end of the 19th century, welfare capitalism could be found throughout the industrialized world, but it was especially popular in the United States. Not only did American employers favor welfare capitalism because they thought it would inhibit the growth of unions and government, but they also saw it as an efficient alternative to individualism: Training would be cheaper and productivity higher if employees spent their work lives with a single firm instead of seeking their fortunes on the open market; a pension would ease them out when they became superannuated. Also impelling welfare capitalism was a moral impulse; self-made business owners felt a sense of stewardship and paternal obligation to their employees. But virtue was conveniently conflated with strategic considerations, as when American employers convinced themselves that welfare capitalism constituted the best defense of freedom against laborism and statism. In short, welfare capitalism was a good fit for a distinctive American environment comprised of large firms, weak unions, and small government Gacoby 1997). Welfare capitalism was an influential movement for the first three decades of this century, embraced by employers as well as intellectuals, social reformers, and political leaders, all of whom shared the belief that industrial unrest and other problems could best be alleviated by this distinctively American approach: private, not governmental; managerial, not mutualist. To put its ideas into practice, employers cleaned up their factories, constructed elaborate recreational facilities, launched "company" unions, and even built housing for their employees. By the 1920s, welfare capitalism reached millions of workers at thousands of firms. It was an impressive, if imperfect, system, whose notions of order, community, and paternal responsibility recalled the preindustrial household economy. In its early years, welfare capitalism was a mixture of paternalistic policies aimed at improving a worker's moral character-these were often directed at recent immigrants-and more impersonal, pecuniary programs intended to stabilize employment and align the worker's financial interests with those of his employer. The former ran the gamut from Americanization programs to company doctors, while the latter comprised efforts to create career-type jobs that offered profit sharing, stock ownership, old-age penSions, health benefits, and sometimes even unemployment benefits. Those who worked for a given company came from different risk groups: Managers on average had more stable jobs, longer life spans, and fewer health problems than manual workers. Although treated unequally by the market and by society, this amalgam of white- and blue-collar employees was transformed by welfare capitalism into a new risk group possessing common interests. But this was vertical solidarity rather than the horizontal solidarity of the trade union or fraternal organization. Where once there had been a hierarchy of managers and workers, there now stood (at least in theory) a group. At Kodak-an early pioneer in welfare capitalism-the group was known as "the Kodakers" or the "Kodak clan;' an identity that legitimated the company's mildly redistributive welfare schemes. There were, however, two anomalies-nonwhites and women-that remained second-class citizens in these corporate communities. Welfare capitalism, then, was



primarily for white men; the industrial community was a brotherhood in which race and gender helped to bridge the gap between managers and workers. 6 Some companies, such as Kodak, went even further: Kodak hired manual workers from Anglo-Saxon backgrounds while avoiding recent immigrants, thus using ethnicity to solidify corporate ties and differentiate "insiders" from ethnic "outsiders." Labor-intensive companies usually did not have this option; they had to rely on immigrant labor. But industrial employers recognized early on that the immigrant's ethnic solidarity could potentially be a threat, one that might form the basis for organizing against the employer. Hence, welfare capitalism promoted Americanization programs and insurance plans that competed with the fraternals. As one manager said, "The Mexicans have [fraternal] societies, and of course if they organize one way, it is but a step to organize another way" (Cohen 1990:176). The earliest corporate insurance schemes were mutual benefit associations that started in the late 19th century to offer sickness and death benefits to corporate employees. These operated much like fraternals except that only employees were eligible to belong. Membership was voluntary and benefits were financed with member contributions, but employers provided administrative assistance and subsidies. Sears Roebuck had visiting doctors and nurses who provided aid to (and checked up on) employees collecting health benefits. Other pioneers included Solvay Process (chemicals), Sherwin-Williams (paint), McCormick (farm machinery), and Brown and Sharp (guns). In the early 1900s, before states passed workers' compensation laws, companies began providing accident insurance, usually paid for by the employer and administered by company medical examiners, who judged whether an injured employee deserved accident benefits and if so, how much. Some companies even offered their own self-insured pension plans, including Western Electric, Bausch & Lomb, and Swift Packing. And after 1910, a few paragons of welfare capitalism-including Dennison Manufacturing, Columbia Conserve, and Leeds & Northrup-offered unemployment insurance to their employees (Brandes 1976; Henderson 1908; Jacoby 1997; Nelson 1969). For commercial insurers, welfare capitalism offered tantalizing opportunities. Corporations were more likely than individuals or fraternals to make timely payments and remain solvent. And compared to industrial insurance, group insurance offered profitable economies of scale and a chance to realize Haley Fiske's goal of preempting the welfare state. But a major marketing effort was required. Insurers had to persuade laggard employers of the virtues of private welfare capitalism. And they had to convince progressive employers that they were better off buying group policies than self-insuring, while at the same time pushing them to adopt new programs such as major medical insurance (Keller 1963:290; Dublin 1943:421441; Carr 1975:213; Sass 1997). The first significant group insurance policy was written in 1912 by The Equitable for the employees of Montgomery Ward. Group insurance policies-death, health, and pensions-grew ninefold from 1918 to 1931. Rapid growth was spurred by a shift from noncontributory to contributory plans, in which the employee paid part of the premium. (By 1924, 90 percent of Metropolitan's group policies were contributory.) To encourage worker stability, insurance companies devised the "single premium" pension plan, whereby employees contributed initially but were gradually shifted to a noncontributory status as their tenure rose. At its peak in the 1920s, welfare capitalism covered about one-fifth of the industrial labor force, most

employed in mid- to large-sized companies in the manufacturing, retailing, utilities, and transportation industries. Around 45 percent of these companies offered health and accident insurance either directly or through a corporate mutual benefit association Oacoby 1985). Pensions were less common; only about 400 plans (covering around 2 percent of the industrial labor force) were in existence in the 1920s. Some of these were unfunded and discretionary; others were underwritten by large life insurance companies such as The Metropolitan (Sass 1997). Pension plans predominated on the railroads, in white-collar, career-oriented industries (e.g., utilities and banks), and in stable, consumer goods industries (soap, food). Least prevalent was private unemployment insurance; no more than approximately three dozen companies offered this benefit to their employees (Nelson 1969). The instability of employment reflected the uneven spread of welfare capitalism; in fact, employment stability actually decreased in the 1920s, causing workers in welfare capitalist companies to worry that they would lose their benefits if laid off. Nevertheless, welfare capitalism captured the public imagination and, at least among the largest and most progressive companies, appeared to be generating an inexorable trend toward wider and better coverage. When Eastman Kodak announced a lavish new pension plan in 1928, it was featured in The New York Times and analyzed at a special session of the American Management Association (Clough 1946:237; Dublin 1943:173; Hannah 1986; Jacoby 1993). The 1920s saw competition between the various providers of private life, health, and old-age insurance: the fraternals, employers, and ordinary private insurance (industrial and individual). Although a few fraternals now offered insurance in cooperation with private commercial companies, most of them were still selfinsured, and they began to lose out in the three-way competition. The total share of life insurance written by fraternals declined in the 1920s, and they were slowest to move into new areas such as health and retirement insurance (Cordery 1996:93). Part of the decline was due simply to the availability of employer-provided benefits. But it also was due to the steady weakening of ethnic identity. During the 1920s, immigrant families began moving into the cultural mainstream: purchasing national brands, watching Hollywood movies, listening to radio, and in other ways undergoing an assimilation that weakened identification with the ethnic group. Middle-class ethnics began to leave their enclaves and move to inlying suburbs, where they enjoyed more privatized and deracinated forms of consumption (Cohen 1990). Still, the loss experienced by the fraternals was relative, not absolute. The dollar value of their insurance in force held steady, while membership did not drop sharply, if at all (Cohen 1990). To retain members and attract new ones, the fraternals tried a variety of tactics. They offered new benefits to women and to juveniles, who were not likely to be covered by employer-provided insurance. And they emphasized their ties to the community by stepping up the level of scholarships and investments in the community (Renkiewicz 1980). Smaller societies continued to merge with each other, while adopting more professional management and marketing techniques. While not every immigrant worker wished to join a fraternal, some had no decent alternative. Nonunion miners and millworkers, for example, had neither union nor company benefits (their employers were uninterested in welfare capitalism), nor were they permitted to purchase industrial insurance because of their high accident and death rates (Bodnar 1985; Beik 1996).



As in earlier years, workers continued to purchase multiple policies that reflected their compound identities as workers, ethnics, and aspiring (or actual) middle-class consumers. This was also a matter of risk hedging, since employerprovided insurance was not entirely reliable due to the various restrictions that could make a worker ineligible in case of layoff or job loss. For all these reasons, the fraternals remained stable-albeit not growing-on into the 1930s, when there remained 200 major fraternal organizations with over 100,000 lodges (National Fraternal Congress of America 1938). Thus, on the eve of the Great Depression, the private market was competitive and complex, whereas government did little or nothing to insure Americans against the labor-market risks of the Industrial Age.

DEPRESSION, WAR, AND A NEW DEAL All forms of private insurance experienced difficulties during the Depression. Fraternal insurance was especially hard hit, since workers failed to make payments on their policies. The crash of the real estate market devastated fraternals that had invested in local communities. Some fraternals collapsed; others were left in shaky condition (Cohen 1990:229). Similar problems affected industrial insurance. Premium payments and coverage declined precipitously between 1931 and 1934; a few small companies canceled their insurance plans, which sent tremors through the market (Clough 1946:241). Private employers, too, found it difficult to keep their programs going. Somewhere between 10 and 15 percent of corporate pension plans in existence in 1929 were discontinued during the early 1930s, although a few new plans were started by a smaller group of companies that was able to dodge the Depression's worst effects Oacoby 1985:210, 211; Sass 1997:90). Similarly, while a handful of firms launched private unemployment insurance and relief plans in the early 1930s, including General Electric and Kodak, the number of new plans was dwarfed by failures of older plans, and even many of the new plans failed to pay benefits as promised Oacoby 1997). Simply counting plan births and deaths underestimates the Depression's effect, however, because unemployed workers usually were exempt from benefits coverage, even in companies that kept their programs intact, and many nominally intact programs were actually insolvent. In a harbinger of what was to come later in the decade, several railway unions turned to Congress in 1931 and asked it to take over the carriers' bankrupt pension plans, although the employers opposed the idea. (Sass 1997:93). Despite the inadequacies of the pre-Depression systems for insuring labormarket risk, they nevertheless legitimated the idea that a voluntary approach was a feasible substitute for social insurance. But the collapse of those private systems led industrial workers in two directions: first, to the Democratic Party, which promised a new approach to security organized by government; and second, to the labor movement, which made industry and occupation rather than ethnicity or employer the basis for risk sharing. The new industrial unions pursued a complicated approach to security: supporting a more active role for government while also pushing employers to supplement government programs by expanding welfare capitalism Oacoby 1997). Until the late 1940s, the focus of union attention-for unemployment, old-age, and health benefits-was on the state, as seen in the

various efforts to pass the Wagner-Dingell-Murray bill that would have beefed up Social Security and expanded it to include national health insurance. While working-class support was necessary to achieve a welfare state, it was not sufficient; an alliance with middle-class voters was required. The story of working-class mobilization in the 1930s is well known, and historians have also analyzed labor's efforts to expand the welfare state in the 1940s (Berkowitz and McQuaid 1988; Stevens 1990; lichtenstein 1995). On the other hand, relatively little is known about the middle class and how its attitudes to risk and risk pooling changed during and after the Depression. Yet it was middle-class voters who provided decisive support for the New Deal's array of social insurance programs. Prior to the 1930s, the American middle class was skeptical of social insurance, preferring to rely on private and voluntary approaches. This skepticism was part and parcel of American exceptionalism (and cultural pluralism). Even in Britainthe least corporatist of the European nations-the middle class would have found strange the American Creed's powerfully antistatist individualism (Wiener 1981). And most of Europe was more ethnically and racially homogeneous than the United States, thus creating stronger bonds between the middle class and the poor. This is not to say that the European middle classes automatically favored the redistributive risk pooling associated with the welfare state. Sometimes what was needed was a perceived threat to law and order, as in Bismarckian Germany, while at other times the balance was tipped by the profamily and antimarket ideologies of the Church, as in France and Italy (Esping-Andersen 1990). Once Europe and the United States started on their separate roads, institutionalization widened initial differences. In Europe, public provision undermined commercial insurance and mutual benefit societies, causing middle-class Europeans to orient their needs to the public sector; in the United States, a proliferation of private insurance led the middle class into thinking it was capable of handling risk through private channels (Dreyfus 1996). How, then, did the American middle class come to support a welfare state? The answer is not immediately obvious given the ease with which middle-class Americans could purchase insurance on their own and do so without the redistributive encumbrances of a welfare state. The answer lies in the Depression, which had two important effects on the middle class. First, it undermined middle-class trust in the private system for handling labor-market risks. The collapse of welfare capitalism, combined with mass layoffs and bank failures, created a crisis of confidence that caused middle-class Americans to mobilize in search of alternatives-such as the Townsend Clubs, Huey Long's Share Our Wealth plan, and Father Coughlin's National Union for Social Justice-all of which were co-opted by Roosevelt's promise of a New Deal (Brinkley 1982). Second, the depression caused the middle class to perceive its social status as precarious; actual or threatened downward mobility narrowed the social distance between the middle class and those below, making cross-class alliances more likely and weakening resistance to redistributive programs that could benefit the downwardly mobile. (This is the social affinity hypothesis.) But Roosevelt and his technicians were well aware that old habits die hard. They carefully designed Social Security pensions around the semifictive concept of individual accounts that were ostensibly (but in reality, only loosely) linked to individual contributions (Leff 1983; Karl 1983). World War II, though it brought the end of the Depression, reinforced social



cohesion between the middle and working classes. All citizens (except JapaneseAmericans) now were part of a concerted effort to defeat the enemy, one in which the health and welfare of the bottom strata were a matter of concern to all. Interdependence was fostered by official propaganda centered on inclusive concepts such as citizenship and "Americanism;' by a mass military draft, and by antidiscrimination programs started in war industries. Moreover, the widespread conviction that government was responsible for ending the depression and for winning the war bolstered confidence in public policy and in the belief that a strictly private approach to social and economic problems no longer was viable (except in health, which became a corporate fringe benefit during the war). Social Keynesianism-the notion that public spending ensured prosperity by raising purchasing power- became middle-class dogma (Gerstle 1989; Marshall 1950).7 What is especially interesting is the persistence of these attitudes for several decades. As Glen Elder (1974) has shown, the Depression created a lifelong obsession with security that was characteristic not only of adults in the 1930s but also their children. Postwar labor force entrants shared the economic anxieties of their parents, leading observers such as Daniel Bell (1960) to conclude that middle-class Americans had lost their antipathy to government. Throughout the prosperous postwar years, these "children of the Depression" (and of war) looked to government for basic insurance against labor-market risks (and for boosts to upward mobility such as G.I. loans for education and housing). Thus did Social Security become sacrosanct to middle-class Americans; even in the 1990s, it remained a hotbutton issue for the now-elderly cohort born before the war. In short, the experiences of the 1930s and 1940s inclined the American middle classes to favor social solidarity over rugged individualism, a shift that promoted the welfare state, public education, and other forms of government expenditure. Yet the American welfare state that developed after 1935 was a far cry from the egalitarian, generous, and broad-gauged programs that existed in Europe, even in Britain. Old-age pensions and unemployment insurance were relatively modest; coverage, at least initially, was far from universal; and health, with the exception of disabilities, was not insured. Moreover, the initial structure of Social Security preserved racial exclusions of the sort associated with fraternalism. Southern politicians and employers had no interest in assisting black agricultural and domesticservice workers, so these sectors initially were excluded from old-age insurance and, in Southern states, from unemployment insurance (Quadagno 1988). A penurious welfare state left considerable room for private alternatives-whether individual, fraternal, or corporate; indeed, the notion that public and private insurance were complements rather than substitutes was embedded in the very structure of the new system. Take welfare capitalism, for example. Most businessmen opposed the New Deal, including Social Security, as an excessive expansion of government power. Insurance executives and managers from progressive companies worried that the welfare state would choke off what remained of welfare capitalism. But a few farsighted business leaders, such as Marion Folsom of Eastman Kodak, had the ingenuity to see how their companies might profit from the situation. A welfare state financed by payroll taxes would narrow costs between welfare capitalist firms such as Kodak and those companies that spent little or nothing on welfare benefits. Intent on realizing this objective, Folsom served on key committees that helped

design the Social Security Act and he later lobbied the business community in support of it Oacoby 1993). But corporate liberals such as Folsom were determined not to permit Social Security from displacing welfare capitalism and other private efforts. Government, said Folsom, should provide only "basic minimum protection and it should not be intended to cover all the needs of everyone" (in Jacoby 1997:210). To ensure that workers still looked to employers for security, public benefits had to be kept low and private programs had to be given tax incentives. As a Washington insider, Folsom lobbied tirelessly - and successfully - for these ends. From the mid-1930s to the mid-1950s, Congress repeatedly refused to raise tax rates and benefits for Social Security. Meanwhile, Folsom-and organizations such as the Committee for Economic Development - tried to persuade other employers that Social Security was a Keynesian stabilizer, that it was preferable to radical alternatives, and that it could coexist with-and even subsidize-private efforts. The insurance companies took issue with these claims, fearing that social insurance eventually would shrink the lucrative private market. During the initial debates over Social Security, they sought a contracting-out provision that would have allowed welfare capitalist companies to opt out of the federal program if their pension plans met certain standards. Had this provision been enacted, it would have placed the private sector in direct competition with government to see who could offer better benefits. But Folsom opposed the provision, warning that it would bring undesired federal scrutiny of private pension standards. Instead, Folsom lobbied the business community in favor of a different approach - "the supplemental plan" -that restructured private pensions so that employers deducted from their premiums the amount paid to the government in Social Security taxes. Retirees would then receive two pensions, but the employer's cost would remain the same. The insurers opposed this kind of coordination, saying that only contracting out would prevent wholesale abandonment of private pension plans. But they were wrong. The supplemental approach not only preserved private plans but also provided incentives for employers to start new ones. After 1935, a majority of corporate pension plans were supplemental, that is, integrated with federal benefits. And many employers now started new pension plans, lured by tax incentives (fringe benefits were exempt from payroll taxes and, during the war, from wage controls and excess profit taxes) and pushed by a belief that restive workers would be less inclined to join unions if they were given better benefits. Indeed, for a small outlay, a company could earn its employees' gratitude by augmenting the inadequate pension paid by the government. Along with the supplemental approach, employers took other steps-such as stabilizing job structures and strengthening internal labor markets-as they recognized how much stability meant to their employees. Some liberals feared that the drop in mobility resulting from corporate paternalism was evidence of a "new industrial feudalism" (Elder 1974:186, 192; Jacoby 1985:275; Jacoby 1997). The new unions similarly recognized the centrality of security but were ambivalent about private benefit programs, fearing that these would weaken efforts to attain better social insurance. But the Congress of Industrial Organizations' (CIO) campaign to raise Social Security benefits repeatedly was defeated in the 1940s, along with proposals to expand Social Security to include health insurance. Unions



gradually came to the realization that they had more economic than political power, and after the National Labor Relations Board ruled in 1949 that fringe benefits were a mandatory bargaining issue, there was a surge of union-negotiated health and pension plans. In the postwar decades, welfare capitalism - whether unilateral or negotiated-coexisted easily with a "basic" welfare state. Fringe benefits, said Folsom, were premised on government's "taking care of the minimum requirements of the individual and not beyond that. If you keep [SOcial Security] on a minimum basic protection basis, the individual can supplement it from his own earnings and from any pension which he might derive from his employer" (in Jacoby 1997:215). Just as welfare capitalism prospered in the wake of the welfare state, so did other forms of private insurance. The policies provided by fraternal and ethnic organizations, though less popular than in the 1920s, kept a hold on middle-class males and ethnic communities (even as late as the 1970s, Polish fraternals still had over 700,000 members). With government paying survivors' benefits, one might have expected the private life insurance market to shrink. But in 1942, fraternals had $6 billion of insurance in force, industrial insurance stood at $23 billion, and group insurance (welfare capitalism) at $19 billion (Clough 1946:235). Given the meager scale and scope of public benefits, private insurance served as the individual analogue of the supplemental approach. It also was a way of hedging one's bets against problems with Social Security. These patterns persisted even with the scaling-up of Social Security benefits, which started in the mid-1950s, and with the addition of new health-oriented programs such as Medicare and Medicaid, launched in the mid-1960s. Come 1990, 44 percent of u.S. social welfare expenditures-not including education-still originated in the private sector, making the United States an outlier in the way it apportioned spending between the public and private sectors (Social Security Administration 1993:128-130). With the exception of Japan, no other advanced country had as much social welfare spending coming from the private sector; in France, the figure was about 5 percent, and in Scandinavia it was close to zero (Gordon 1988; Rein and Rainwater 1986). Health expenditures are an important reason for the United States being an outlier. Sixty percent of all health expenditures in the United States came from the private sector in 1987, whereas the Organization for Economic Cooperation and Development (OECD) average was only 23 percent (OECD 1990). Old-age insurance (pensions and death benefits) is less privatized than health insurance, both in the United States and in Europe, but even here the United States remains an outlier, with 25 - 30 percent of pensions coming from private sources versus an average of 10 percent in the 12 richest European countries (Social Security Administration 1993; Esping-Andersen 1990). While unemployment insurance is a public function in both countries, the Europeans have gone further in legislating forms of employment security that in the United States are the result of private arrangements, including those reSUlting from collective bargaining. Thus, the rise of a U.S. welfare state in the 1930s did not constitute a radical shift in risk institutions but instead was consistent with, and strongly influenced by, preceding developments. Not only did private insurance remain robust but it constrained the structure and scope of public programs.

THECONTEMPORARYSCENE In the current period, the institutions for handling labor - market risk - both in Europe and the United States-appear to be changing. Though the "shock" of increased global competition is not of the same magnitude as the Great Depression or World War II, it nevertheless has the capacity to affect institutional structures. The effects of globalization - growing income inequality, weaker nation-states, and more porous economic borders-have immediate consequences for risk institutions. Why is this so? First, as inequality widens, it is more difficult to sustain the cohesiveness that undergirds support for redistributive public spending (Moene and Wallerstein 1997). The position of the middle class relative to the poor is one of the strongest predictors of how much of its gross domestic product (GDP) a nation commits to social spending (Lindert 1996). Second, nations today have less economic autonomy due to the internationalization of capital markets and the voluntary secession of sovereignty to supranational bodies such as the European Union and the World Trade Organization. These developments undermine the logic of social Keynesianism by accelerating the "leakage" of domestic income (via the purchase of imports) and, more significantly, by restricting the ability of nations to initiate expansionary fiscal, monetary, and exchange rate policies Oacoby 1995). Also, the liberalization agreements reached during the Uruguay Round are bringing down national barriers to trade in financial services. Anglo-American insurance companies are poised to expand the sale of private pension and savings products in Western Europe as well as other parts of the world. Third, while labor remains less mobile than capital, recent years have seen an increase in labor migration from developing to developed nations. In many places-including Germany, Sweden, and California-the presence of substantial numbers of foreigners has fueled anti-immigrant passions. The resultant xenophobic atmosphere has the effect of undermining the ethnic and racial solidarity that contribute to public welfare spending (Handler 1997). Conventional wisdom has it that these effects of globalization will cause a "withering ofthe state." For Europe that would mean a convergence with U.S.-style risk programs (Le., greater private spending and less redistribution). But caution is advised; reports of the European welfare state's demise are, at this point, exaggerated. There remains a "governmental habit" in Europe; that is, European citizens still look to the state to shield them from risk (Hughes 1991). The welfare state stands as a symbol of patriotic unity; privatization is viewed with deep suspicion. Working- and middle-class citizens have fought against welfare spending cuts in France, Germany, Italy, and elsewhere. While some middle-class Europeans are hedging their bets with annuities and mutual funds, such activity remains limited by the absence of tax incentives for private supplementation. Only Britain and the Netherlands have taken substantial steps toward privatization; elsewhere, governments and citizens are leery of it. The data show a change in the second but not the first derivative of welfare spending; that is, spending continues to increase but at a slower pace ("A Pension Bonanza Beckons on the Continent" 1997; Garrett 1995; Huber and Stephens 1996).



Another factor likely to sustain the European welfare states is the link between welfare spending and political support for economic openness. For years the European economies have been oriented to trade; their openness is nothing new. But trade increases an economy's exposure to risk. Without some kind of compensation for that greater risk, citizens might oppose liberal trade policies. Expenditures on education, unemployment insurance, and other welfare programs help to sustain political support for an open economy. The empirical evidence suggests that welfare transfers are positively associated with the degree of an economy's openness. Hence not only is globalization compatible with government expenditures, but it may also actually increase them (Rodrik, 1996; International Labour Office 1997). Despite the Reagan - Thatcher antigovernment sentiment of the 1980s, there is no conclusive evidence that government spending is associated with poorer economic performance; in fact, recent evidence points in the opposite direction (Slemrod 1995). The "new growth theory" asserts that government expenditures can raise productivity by spurring human capital investments and reducing inequality (Romer 1990). Inequality hampers productivity growth, whereas welfare expenditures reduce inequality; therefore, the welfare state may be associated with higher productivity (persson and Tabbelini 1994; Atkinson 1995; Lindert 1996). While these findings are contested by some economists (e.g., see Tanzi and Schulknecht 1995), they do provide support for those Europeans who are unconvinced by neoclassical claims and wish to preserve their welfare states despite globalization. Indeed, many Europeans remain skeptical of the chief alternative to the welfare state: the relative privatized welfare system found in the United States. While Europeans are enamored of high U.S. job creation rates, they are not convinced that the American approach is a superior alternative. In the last decade, U.S. welfare capitalism has confronted its most critical test since the Great Depression. Heightened competition and rapid technological change have caused massive layoffs throughout American industry, often at companies (e.g., IBM and Kodak) that previously offered their employees an implicit job security guarantee. While not all career jobs are in peril-indeed, many such jobs remain-absolute security no longer exists, especially in blue-collar employment, and the number of career jobs is smaller than before (Neumark, Polsky, and Hansen 1998). Moreover, to reduce the cost of offering such jobs, employers have cut back on the fringe benefits associated with them. Instead of defined benefit pension plans, employees today are more likely to be covered by defined contribution plans such as the 401(k), which shifts much of the risk from employer to worker. Employee copayments for health insurance have increased sharply, as has the number of employees in health maintenance organizations. Some employers have gotten rid of fringe benefits entirely. To further cut costs, employers have increased their reliance on contingent employees (part-time, temporary, and contract workers), whose jobs serve as a flexible buffer for remaining core employees. Today, around 30 percent of the workforce does not hold standard, full-time jobs. The proportion of these workers receiving employer-provided health or pension benefits (12 percent) is much smaller than for full-time employees (60- 70 percent) (Kalleberg et al. 1997). The net result of these changes-fewer and less stable core jobs-has been a sharp drop in employer-provided insurance. (The drop in core jobs is itself a cut in risk insurance; employers are less willing to carry employees through the vicissi-

tudes of changing business conditions.) Overall pension plan coverage has fallen from 55 percent in 1979 to 51 percent in 1996, and the drop has been even sharper for blue-collar workers. A similar decline has occurred in health insurance coverage ("Unto Those That Have Shall Be Given" 1996; Kalleberg et al. 1997; Sass 1997). If these changes are a sobering portent for Europeans, why is it that Americans have not embraced the European approach and more loudly demanded an expansion of public responsibilities? Part of the answer is the miserably high unemployment rates in Europe, which many attribute to the welfare state, though the attribution may be incorrect. Another part of the answer is path dependence: Over time the United States has developed a set of institutions-and vested intereststhat hamper an expanded public role in benefit provision. The chief adversaries of the Clinton health plan were the insurance companies; these same companiesalong with the financial services industry -are also the main engine behind current efforts to privatize Social Security pensions (Feldstein 1997; ]udis 1995). Shifting to a less privatized system would require a political majority that is presently nonexistent. Somewhat more than half the U.S. workforce receives benefits from employers and a significant portion of this group supplements employer benefits with individually purchased pension plans (including individual retirement accounts, or lRAs) and health benefits. Hence, these workers are wary of trading their private coverage-however meager or faulty-for public benefits; that, too, became clear during the debates over the Clinton health plan. While one can imagine a social insurance coalition between full-time workers who lack any pension or health coverage (around 20 percent of the workforce) and contingent workers who lack coverage (10 to 20 percent of the workforce), there is a flaw in that reasoning: Not all contingent workers are unhappy with their present situation. About 40 percent of contingent jobs are of high quality, paying more, on average, than comparable full-time positions. These are the engineers, managers, and other educated specialists who populate Silicon Valley and fuel accounts of a new career model based on "employability" (Kalleberg et al. 1997; Kanter 1995). For them, the new career model offers the opportunity to move from job to job, constantly developing a more versatile set of skills. Their skills, rather than their relationship to a given employer, are conceived as the basis for security, just as was true of the "well-rounded" autonomous craftsmen of the 19th century. As mobile workers, they benefit from defined contribution pension plans that are portable across employers. The existence of a robust private insurance market allows them to use their hefty salaries to purchase Keogh pension plans, health insurance, and other benefits, sometimes even with tax benefits from government, as in the case of the new Roth lRAs. They also would be the chief beneficiaries of recent proposals to establish personal unemployment insurance accounts (Orszag and Snower 1997). Finally, because they are highly educated, these professionals feel better equipped to evaluate and ameliorate the risks they face than the experts who design corporate and government insurance programs. Giddens (1991:119) argues that this "reflexive mOnitoring of risk" is a signal feature of the postmodern world. But it may also be seen as the recrudescence of traditional American traits such as individualism and suspicion of large organizations. Such individualism faded during welfare capitalism's previous crisis in the 1930s. But today's crisis seems unlikely to produce the cross-class solidarity necessary for an expanded welfare state. Even if a feasible coalition did exist, the



commonality connoted by citizenship means far less to today's electorate, who not only do not vote but also "bowl alone" and are otherwise civically disengaged. Moreover, the recent growth of inequality makes solidarity more difficult to achieve. The modem middle class feels less threatened by economic change than in the 1930s; its members can retreat into gated communities and private schools. Unlike the children of the depression, they are "less engaged with one another outside the marketplace and thus less prepared to cooperate for shared goals" (Putnam 1993). While an expanded U.s. welfare state is unlikely, this does not rule out a revival of welfare capitalism, possibly with governmental assistance. Most Americans still look to their employers as the first line of defense against risk. As that line is pushed back, they question the fairness of today's leaner, meaner arrangements. For every IBM there are dozens of employers unconcerned with the niceties of employee commitment: places where the pay is low, jobs are temporary, and benefits are shrinking. Layoffs at firms such as IBM give less scrupulous companies a pretext for restructuring the social contract even when their economic situation does not warrant it. Seventy-five percent of u.s. workers think that companies are less loyal today than 10 years ago. These workers still look to employers to take care of them but lack the power to contest present arrangements (Avishai 1996). American unions might make a comeback as they did in the 1930s, by playing on employee resentment of broken promises and fears of job loss. While prospects for labor's revival are not bright, it would be a mistake to think that the labor movement is doomed. Recent data show that about one-third of U.S. workers would support a union if given the chance (Rogers 1996). Another possibility is embedded in recent proposals from Democratic legislators, including Edward Kennedy, Jeff Bingaman, and Richard Gephardt, to encourage companies to treat employees more like "stakeholders:' Employers who train their workers, give them decent health, family, and pension benefits, and have measures to cushion them from layoffs would receive preferential tax and regulatory treatment. In effect, such legislation would require government to subsidize the cost of private welfare capitalism, which is precisely how Marion Folsom and other welfare capitalists conceived of Social Security in the 1930s. Although identification with one's employer is probably weaker today than in the past, the interplay between public and corporate programs remains a feature of the American scene.

CONCLUSIONS How individuals respond to risk is a question that traditionally has been arrogated by economists. While economic analysis is necessary to answer that question, it is hardly sufficient. Choosing the group with whom to pool risk is mediated by both social identity and the institutions available for risk pooling. Those institutions affect individuals' attitudes toward risk, their propensity for solidarity, and their sense of entitlement. Although workers in every nation are exposed to similar labor-market risks, each nation has developed a unique constellation of risksharing institutions that reflects its singular history. In the United States, the system for handling labor-market risks has evolved

since the 19th century but still bears the stamp of its initial conditions: weak government, small unions, large firms, voluntarism, and ethnic heterogeneity. Even shocks such as depression and war did not eradicate an American proclivity for private insurance. While a redistributive welfare state arose after the Depression, it was predicated on the existence of supplemental private insurance. In the future, the United States is still likely to have a system in which the task of mitigating risk is shared by public and private institutions. Conversely, in Western Europe the early strength of the welfare state checked the proliferation of private alternatives and focused attention on the public arena. Postwar affluence and globalization have not eliminated the European inclination to inclusive, statist, and relatively redistributive welfare systems. In the future, governmental systems are likely to continue handling the bulk of Europe's labor-market risks. These facts constitute path dependence: Divergent starting points trace out cumulatively distinctive trajectories. As institutions proliferate along a particular path, they channel risk-mitigating behavior in particular directions. Path dependence is an evolutionary concept; it suggests that the future will resemble the past. It is, therefore, open to the charge that it promotes what David Fischer (1970) calls the "fallacy of continuity" by discounting discontinuous change: revolutions, punctuated equilibria, and the like. This is a potentially serious charge, although those making it often are unaware that there is an equally serious problem of what might be called the "fallacy of discontinuity": a fascination with new and seemingly novel arrangements. Social scientists and journalists are enamored with novelty because it builds careers and sells stories. But what is touted as "new" on closer examination often turns out to be more of the same. After all, even shocks such as war and depression did not eradicate an American proclivity for private insurance. While path dependence provides a valuable perspective on institutional development, it is a rather undersocialized and mechanical way of thinking about institutions, one that gives short shrift to issues such as identity, solidarity, trust, and social action. As we have seen, racialist ideas such as Ein Yolk played a role in creating and sustaining the European welfare state, while in the United States, ethnic fragmentation hampered the development of an inclusive approach. What is needed to understand risk systems is a blend of economic theory, historical institutionalism, and social analysis. This is precisely what I have sought to provide in this chapter. ACKNOWLEDGMENTS: Thanks to Roger Horowitz for helpful advice and to Kim Nguyen and Sorah Hong for research assistance.

NOTES 1. Note that some significant portion of these white men also were veterans of the Civil War and received military pensions from the federal government. According to Skocpol (1993), the graft associated with the military pension program dismayed the citizenry and delayed the emergence of universal old-age pensiOns until the 1930s. There are some problems, however, with her explanation. First, graft was commonly associated with other government programs (e.g., government contracting and hiring), yet this led to reform rather than shrinkage or elimination of functions. Why did the same not occur with pensions? Second, reformers who pushed for pensions also supported



2.

3.

4.

5. 6.

7.

unemployment insurance, a program that had no military analogue. Yet the reformers were no more successful in achieving unemployment insurance prior to the 1930s than in securing old-age pensions. A more nuanced and multicausal explanation is required to account for the delayed emergence of old-age pensions (and other aspects of the welfare state), one that would include the early development in the United States of private pension and life insurance systems, coupled with deep antistatist traditions that, while they fed revulsion at military pension graft, existed long before such graft ever occurred. In Gould's framework, multiple collective identities are assumed to be ascribed characteristics, yet they might also be analyzed as achieved or consciously chosen characteristics that individuals select for risk-mitigation (or other) purposes (Hollinger 1995). Because fraternals took on "instrumental" insurance functions and mutual benefit societies organized "expressive" leisure activities for members, there was no clear dividing line between the two types of organizations (Schmidt 1980). Like trade unionism, fraternalism was an international phenomenon of the 19th century, culminating in the first international congress of mutual benefit societies held at the Paris World Fair in 1900 (Dreyfus 1996). The average payment period shifted from weekly to monthly in the 1920s, a reflection of workers' increasing affluence (Dublin 1943:138). As at other companies, Kodak's welfare policies redistributed income from shareholders to workers (which is why many shareholders opposed the company's profitsharing scheme) and, at a very modest level, from better paid to less affluent workers. Redistribution also occurred from highturnover workers (predominantly female) to more stable employees. For a discussion of male solidarity at work, see Fine (1993) and Lewchuk (1993). In Europe, attitudes such as these led to a eugenicist concept of the welfare state.

REFERENCES "A Pension Bonanza Beckons on the Continent." 1997. Business Week (December 15):580. Arthur, W. Brian. 1994. Increasing Returns and Path Dependence. Ann Arbor: University of Michigan Press. Atkinson, A. B. 1995. "Is the Welfare State Necessarily an Obstacle to Economic Growth?" European Economic Review 39:723- 730. Avishai, Bernard. 1996. "Social Compact, Version 2.0." The American Prospect Ouly-August):28-34. Azariadis, Costas. 1975. "ImpliCit Contracts and Underemployment Equilibria." Journal of Political Economy 83: 1183 -1201. Baldwin, Peter. 1990. The Politics ofSocial Solidarity: Class Bases in the European Welfare State, 18751975. Cambridge, UK: Cambridge University Press. Beik, Mildred Allen. 1996. The Miners of Windber: The Struggles ofNew Immigrants for Unionization, 1890s-1930s. University Park: Pennsylvania State University Press. Bell, Daniel. 1960. The End ofIdeology: On the Exhaustion ofPolitical Ideas in the Fifties. Glencoe, IL: Free Press. Bendix, Reinhard. 1964. Nation Building and Citizenship: Studies of Our Changing Social Order. New York: Wiley. Berkowitz, Edward and Kim McQUaid. 1988. Creating the Welfare State, 2nd ed. New York: Praeger. Bernstein, Peter. 1996. Against the Gods: The Remarkable Story of Risk. New York: Wiley. Bodnar, John. 1985. The Transplanted: A History of Immigrants in Urban America. Bloomington: Indiana University Press. Brandes, Stuart D. 1976. American Welfare Capitalism, 1880-1940. Chicago: University of Chicago Press. Brinkley, Alan. 1982. Voices of Protest: Huey Long, Father Coughlin, and the Great Depression. New York: Knopf. Calhoun, Craig P. 1997. "The Cultural Politics of Global Economic Integration." Working Paper No.9. Seattle: Center for Labor Studies, University of Washington. Carr, William H.A. 1975. From Three Cents a Week. Englewood Cliffs, NJ: Prentice-Hall. Clawson, Mary Ann. 1989. Constructing Brotherhood: Class, Gender, and Fraternalism. Princeton, NJ: Princeton University Press.

Clough, Shephard B. 1946. A Century of American Life Insurance: A History of the Mutual Life Insurance Company of New York. New York: Columbia University Press. Cohen, Lizabeth. 1990. Making a New Deal: Industrial Workers in Chicago, 1919-1939. Cambridge, UK: Cambridge University Press. Cordery, Simon. 1996. "Fraternal Orders in the United States: A Quest for Protection and Identity." Pp. 83-109 in Social Security Mutualism, edited by Marcel van der Linden. New York: Peter Lang. David, Paul A. 1985. "Clio and the Economics of QWERTY" American Economic Review 75:332-337. Dobbin, Frank. 1992. "The Origins of Private Social Insurance: Public Policy and Fringe Benefits in America, 1920-1950." American Journal of Sociology 97:1416-1450. Dreyfus, Michel. 1996. "The Labour Movement and Mutual Benefit Societies." Pp. 673-683 in Social Security Mutualism, edited by Marcel van der Linden. New York: Peter Lang. Dublin, Louis I. 1943. A Family of Thirty Million: The Story of the Metropolitan Life Insurance Company. New York: Metropolitan Life Insurance Co. Dumenil, Lynn. 1984. Freemasonry and American Culture, 1880-1930. Princeton, NJ: Princeton University Press. Durkheim, Emile. 1933. The Division of Labor in Society. New York: Macmillan. (Original published in French in 1893.) Elder, Glen H., Jr. 1974. Children of the Depression: Social Change in Life Experience. Chicago: University of Chicago Press. Esping-Andersen, Gosta. 1990. The Three Worlds of Welfare Capitalism. Princeton, NJ: Princeton University Press. Feldstein, Martin. 1997. "The Case for Privatization." Foreign Affairs 76:24-38. Fine, Lisa. 1993. "Our Big Factory Family: Masculinity and Paternalism at the Reo Motor Car Company of Lansing, Michigan." Labor History 34:274-291. Fischer, David Hackett. 1970. Historians' Fallacies: Toward a Logic of Historical Thought. New York: Harper & Row. Fishback, Price V and Shawn Kantor. 1996. "The Durable Experiment: State Insurance of Workers' Compensation Risk in the Early 20th Century;' Journal of Economic History 56:809-836. Fligstein, Neil. 1990. The Transformation of Corporate Control. Cambridge, MA: Harvard University Press. Frankel, Lee K., and Miles Dawson. 1910. Workingmen's 1nsurance in Europe. New York: Charities Publication Committee. Garrett, Geoffrey. 1995. "Capital Mobility, Trade, and Domestic Politics of Economic Policy;' Interna-

tional Organization 49:657-695. Gerstle, Gary. 1989. Working-Class Americanism: The Politics of Labor in a Textile City, 1914-1960. Cambridge, UK: Cambridge University Press. Giddens, Anthony. 1990. The Consequences of Modernity. Stanford, CA: Stanford University Press. Giddens, Anthony. 1991. Modernity and Self-Identity: Self and Society in the Late Modern Age. Stanford, CA: Stanford University Press. Gordon, Margaret S. 1988. Social Security Policies in Industrial Countries. Cambridge, UK: Cambridge University Press. Gould, Roger V 1995. Insurgent Identities: Class, Community, and Protest in Paris from 1848 to the Commune. Chicago: University of Chicago Press. Greven, Philip J., Jr. 1970. Four Generations: Population, Land, and Family in Colonial Andover, Massachusetts. Ithaca, NY: Cornell University Press. Gutman, Herbert. 1976. Work, Culture, and Society in IndustrialiZing America. New York: Vintage Books. Handler, Joel. 1997. "Questions about Social Europe by an American Observer." Working paper. UCLA School of Law, Los Angeles, CA. Hannah, Leslie. 1986. Inventing Retirement: The Development of Occupational Pensions in Britain. Cambridge, UK: Cambridge University Press. Harrington, Michael. 1998. "Trade and Social Insurance: The Development of Unemployment Insurance in Advanced Industrial Democracies;' Ph.D. dissertation. University of California at Los Angeles, Los Angeles, CA. Henderson, Charles R. 1908. Industrial Insurance in the United States. Chicago: University of Chicago Press. Hoffman, Frederick L. 1900. History of the Prudential Insurance Company of America, 1875-1900. Newark, NJ: Prudential Press.



Hollinger, David A. 1995. Postethnic America: Beyond Multiculturalism. New York: Basic Books. Huber, Evelyne andJohn D. Stephens. 1996. "Reconsidering the Welfare State in Europe." Working paper, Department of Political Science. Chapel Hill: University of North Carolina. Hughes, Jonathan R.T. 1991. The Governmental Habit Redux: Economic Controls from Colonial Times to the Present. Princeton, r-ij": Princeton University Press. International Labour Office (ILO). 1997. World Labour Report: Industrial Relations, Democracy, and Social Stability. Geneva: Author. Jacoby, Sanford M. 1985. Employing Bureaucracy: Managers, Unions, and the Transformation of Work in American Industry, 1900-1945. New York: Columbia University Press. Jacoby, Sanford M. 1990. "The New Institutionalism: What Can It Learn From the Old?" Industrial Relations 29:316-340. Jacoby, Sanford M. 1991. "American Exceptionalism Revisited: The Importance of Management." pp. 173-200 in Masters to Managers: Historical and Comparative Perspectives on American Employers, edited by S.M. Jacoby. New York: Columbia University Press. Jacoby, Sanford M. 1993. "Employers and the Welfare State: The Role of Marion B. Folsom." Journal of American History 80:525-556. Jacoby, Sanford M. 1995. Workers of Nations: Industrial Relations in a Global Economy. New York: Oxford University Press. Jacoby, Sanford M. 1997. Modern Manors: Welfare Capitalism since the New Deal. Princeton, r-ij": Princeton University Press. James, Marquis. 1947. The Metropolitan Life. New York: Viking Press. Judis, John. 1995. "Abandoned Surgery: Business and the Failure of Health Care Reform." The American Prospect (Spring):65- 74. Kalleberg, Arne, Edith Rasell, Naomi Cassirer, Barbara Reskin, Ken Hudson, David Webster, Eileen Appelbaum, and Roberta Spalter-Roth. 1997. Nonstandard Work, SubstandardJobs: Flexible Worn Arrangements in the US. Washington, DC: Economic Policy Institute. Kanter, Rosabeth Moss. 1995. "Nice Work if You Can Get It: The Software Industry as a Model for Tomorrow's Jobs." The American Prospect (Fall):52-59. Karl, Barry D. 1983. The Uneasy State: The United States from 1915 to 1945. Chicago: University of Chicago Press. Keller, Morton. 1963. The Life Insurance Enterprise, 1885-1910: A Study in the Limits of Corporate Pawer. Cambridge, MA: Harvard University Press. Keyssar, Alexander. 1986. Out of Work: The First Century of Unemployment in Massachusetts. cambridge, UK: Cambridge University Press. Knight, Frank H. 1921. Risk, Uncertainty, and Profit. Boston: Houghton Mifflin. Laurie, Bruce. 1995. "'Spavined Ministers, Ling Toothpullers, and Buggering Priests': Third-Parytism and the Search for Security in the Antebellum North." Pp. 98-119 in American Artisans: Crafting Social Identity, 1750-1850, edited by Howard B. Rock, Paul Gilje, and Robert Asher. Baltimore: Johns Hopkins University Press. Leff, Mark H. 1983. "Taxing the 'Forgotten Man': The Politics of Social Security Finance in the New Deal." Journal of American History 70:372-374. Lewchuck, Wayne. 1993. "Men and Monotony: Fraternalism as a Managerial Strategy at the Ford Motor Company." Journal of Economic History 53:824-856. lichtenstein, Nelson. 1995. The Most Dangerous Man in Detroit: Walter Reuther and the Fate of American Labor. New York: Basic Books. Lindert, Peter. 1996. "What limits Social Spending?" Explorations in Economic History 33:1-34. lipset, Seymour Martin. 1996. American Exceptlonallsm: A Double-Edged Sword. New York: Norton. Luhmann, Niklas. 1993. Risk: A Sociological Theory. New York: A. de Gruyter. Mandel, Michael J. 1996. The High-Risk Society: Peril and Promise in the New Economy. New York: Times Business. Mann, Michael. 1997. "The Rise of Nation-Statism Across the First Half of the 20th Century." Los Angeles: Department of Sociology, UCLA. Marshall, T. H. 1950. "Citizenship and Social Class." pp. 1-85 in Citizenship and Social Class and Other Essays, edited by T. H. Marshall. Cambridge, UK: Cambridge University Press. Modell, John. 1979. "Changing Risks, Changing Adaptations: American Families in the Nineteenth and Twentieth Centuries." pp. 119-144 in Kin and Communities: Families in America, edited by Allan J. lichtman and Joan Challinor. Washington, DC: Smithsonian Institution Press.

Morawska, Ewa. 1987. "The Internal Status Hierarchy in the East European Immigrant Communities of Johnstown, PA, 189O-1930s." Journal of Social History (Fall):75-107. Moene, Karl O. and Michael Wallerstein. 1997. "Rising Inequality and Declining Support for Welfare Spending." Working paper. Evanston, IL: Northwestern University. Muraskin, William. 1975. Middle Class Blacks in a White SOciety: Prince Hall Masonry in America. Berkeley, CA: University of California Press. National Fraternal Congress of America (NFCA), Fraternal Life Insurance (Indianapolis, 1938). Nelson, Daniel. 1969. Unemployment Insurance: The American Experience, 1915-35. Madison: University of Wisconsin Press. Neumark, David, Daniel Polsky, and Daniel Hansen. 1998. "Has Job Stability Declined Yet? New Evidence for the 1990s." Paper prepared for Russell Sage Foundation Conference on Changes in Job Stability, February, New York. Organization for Economic Cooperation and Development (OECD). 1990. Health Care Systems in Transition. OECD Social Policy Studies NO.7. Paris: Author. Orszag, Michael and Dennis Snower. 1997. "From Unemployment Benefit to Unemployment Support Accounts." London: Economics Department, Birkbeck College. Persson, T. and G. Tabbelini. 1994. "Is Inequality Harmful for Growth?" American Economic Review 84:600-621. Putnam, Robert D. 1993. "The Prosperous Community: Social Capital and Public Life:' The American Prospect 13:35-42. Quadagno, Jill. 1988. The Transformation of Old Age Security: Class and Politics in the American Welfare State. Chicago: University of Chicago Press. Rein, Martin and Lee Rainwater. 1986. "The Public/Private Mix." Pp. 6-16 in Public/Private Interplay in Social Insurance, edited by M. Rein and L. Rainwater. Armonk, NY: M.E. Sharpe. Renkiewicz, Frank. 1980. "The Profits of Nonprofit Capitalism: Polish Fraternalism and Beneficial Insurance in America:' Pp. 113 -129 in Self-Help in Urban America: Patterns ofMinority Business Enterprise, edited by Scott Cummings. Port Washington, NY: Kennikot Press. Roediger, David. 1991. The Wages of Whiteness: Race and the Making of the American Working Class. London: Verso. Rodrik, Dani. 1996. "Why Do More Open Economies Have Bigger Governments?" National Bureau of Economic Research, Working Paper No. 5537. Cambridge, MA. Rogers, Joel. 1996. "What Do Workers Want? Findings of the Worker Participation and Representation Survey." pp. 279-287 in Proceedings of the New York University 48th Annual Conference on Labor. New York. Romer, Paul. 1990. "Are Nonconvexities Important for Understanding Growth?" American Economic Review 80:97-103. Sass, Steven A. 1997. The Promise of Private Pensions: The First Hundred Thars. Cambridge, MA: Harvard University Press. Schmidt, Alvin J. 1980. Fraternal Organizations. Westport, CT: Greenwood Press. Sheridan, Thomas E. 1992. Los Tucsonenses. Thcson: University of Arizona Press. Skocpol, Theda. 1993. Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States. Cambridge, MA: Harvard University Press. Skowronek, Stephen. 1982. Building a New American State: The Expansion of National Administrative Capacities, 1877-1920. Cambridge, UK: Cambridge University Press. Slemrod,Joel. 1995. "What Do Cross-Country Studies Teach about Government Involvement, Prosperity, and Economic Growth?" Brookings Papers on Economic Activity 2:373-415. Social Security Administration. 1993. Social Security Bulletin, Annual Statistical Supplement. Washington, DC: U.S. Government Printing Office. Soyer, Daniel. 1997.jewish Immigrant Associations and American Identity, 1880-1939. Cambridge, MA: Harvard University Press. Stevens, Beth. 1990. "Labor Unions, Employee Benefits, and the Privatization of the American Welfare State." Journal of Policy History 2:238-245. Stiglitz, Joseph. 1987. "The Design of Labor Contracts." pp. 47 -68 in Incentives, Cooperation, and Risk Sharing, edited by H. R. Nalbantian. Totowa, N]: Roman & Littlefield. "Sweden's Dirty Secret." 1997. Los Angeles Times (August 26). Tanzi, Vito and L. SchuJknecht. 1995. "The Growth of Government and the Reform of the State in Industrial Countries." International Monetary Fund Working Paper. Washington, D.C.



Thomas, John 1. 1983. Alternative America: Henry George, Edward BeUamy, H. D. Lloyd, and the Adversary Tradition. Cambridge, MA: Harvard University Press. Ulman, lloyd. 1956. The Rise of the National Trade Union. Cambridge, MA: Harvard University Press. "Unto Those that Have Shall Be Given." 1996. Economist 21:91. van der Unden, Marcel. 1996. Introduction in Social Security Mutualism: The Comparative History of Mutual Benefit Societies, edited by M. van der Unden. New York: Peter Lang. Versteegh, Pien. 1996. "The Role of Family and Ethnic Networks in the Settlement of Polish Migrants in the United States, 1890-1940." Paper delivered at the Social Science History Association Annual Meeting, New Orleans, September. Vickers, Daniel. 1994. Farmers and Fishermen: Two Centuries of Work in Essex County, Massachu· setts, 1630-1850. Chapel Hill: University of North Carolina Press. Voss, Kim. 1993. The Making ofAmerican Exceptionalism: The Knights ofLabor and Class Formation in the Nineteenth Century. Ithaca, NY: Cornell University Press. Weare, Walter B. 1993. Black Business in the New South: A Social History ofthe North Carolina Mutual Life Insurance Company. Durham, NC: Duke University Press. Webb, Sidney and Beatrice Webb. 1920. Industrial Democracy, 2nd ed. London: Longmans, Green. Wiener, Martin. 1981. English Culture and the Decline of the Industrial SPirit, 1850-1980. Cambridge, UK: Cambridge University Press. Wolfe, Alan and Jytte Klausen. 1997. "Identity Politics and the Welfare State." Social Philosophy and Policy 14:231-255. Zamora, Emilio. 1992. The World of the Mexican Worker in Texas. College Station: Texas A&M Univer· sity Press. Zelizer, Viviana A. Rotman. 1979. Morals and Markets: The Development of Life Insurance in the United States. New York: Columbia University Press. Zerubavel, Eviatar. 1985. Hidden Rhythyms: Schedules and Calendars in Social Life. Berkeley: Univer· sity of California Press.