The American welfare state is as embedded in private employment as it is embodied in public programs. Paid work is not an exit from the welfare state, it is a point of entry. For the fortunate, work is a link to a vast array of public and private benefits that pay for our health care, support us in old age, tide us over between jobs, care for us if we are injured or disabled, and support our dependents in the event of our untimely death. Grafted onto the work force, this rich set of benefits forms the work-based welfare state.
Societies need productive workers, and rewarding productivity with security might make sense if the rewards were open to all. But the work-based welfare state has neer been an equal opportunity employer. It reserves its most generous protections for the well-paid, longer term employee whose work involves uninterrupted attachment to a single firm. This is "the good worker," and only he and his dependents reap the full measure of the work-based welfare state.
"Good workers" have always been an elite class dominated by white men. Those who lack good worker status might be protected as dependents, but in an era of rampant divorce this avenue is risky. Most of those denied good worker protections are in the work force. They labor at jobs that offer them few or no protections, either because their employers choose not to provide them, or, more subtly, because arcane regulations deny them the benefits their employers ostensibly offer. Lately, however, even the good worker has seen his security undermined. The class of good workers is contracting, and many of the white men who have long been its chief beneficiaries find themselves excluded from a system they rarely thought about but long depended on. Painful though the process might be, the contraction of the good worker model may at last focus our attention on it and compel us to rethink our work-based welfare state.
LINKING SECURITY AND EMPLOYMENT
Private, work-based benefits have a long history. Before the turn of the century, American entrepreneurs established employee loan funds, initiated free medical care, began recreation programs, and provided their employees with insurance. These private benefits persisted even after public systems were initiated, and their growth was spurred by the odd synergy of World War II tax laws and wartime wage controls. The Revenue Act of 1942, enacted to finance the war effort, imposed an 80 percent excess-profits tax on American businesses. At the same time, in an effort to control inflation, the National War Labor Board (NWLB) adopted a wage freeze. Neither measure was popular with industry. As sociologist Beth Stevens tells the story, the NWLB sought to mollify employers chafing under the wage freeze with a 1943 ruling that excluded "fringe adjustments" from the category of wages. This decision led the way out of both traps. Pretax profits could be lowered and employee loyalty rewarded by the institution and expansion of fringe benefits.
After the war, unions, which had previously supported most public benefit programs, shifted. Faced with the need to deliver tangible results to the rank-and-file, unions actively and successfully bargained for ever-increasing packages of employer-provided benefits. Continuing high rates in the personal income tax made these tax-free benefits attractive even to the average worker who, before the war, had been relatively unaffected by tax policy.
America's public benefit system also focuses on the labor force, often to the exclusion of the public at large. In assembling the Committee on Economic Security, the body that designed and drafted the 1935 Social Security Act, President Roosevelt and Labor Secretary Frances Perkins passed over the proponents of more inclusive proposals to choose a group that would adapt private benefit structures for use in a public, work-linked system. The statute they offered to Congress reflects their deference to the private sector and the conviction that the Social Security system must reinforce, not undermine, the capitalist virtues of initiative and competition.
The Social Security Act is still the basic source of public benefits in the United States. It creates both work-based and non-work-based benefits. The latter, collectively labeled "the categorical programs," were designed to assist categories of citizens considered incapable of work: the blind, the elderly, and children. The balance of the act was grafted onto the workplace, producing some ironic results. For example, Old Age Insurance (OAI), the program most people mean when they use the generic term "Social Security," is America's public retirement system--but it is not open to the general public. Only workers and their dependents qualify for OAI, via a complicated formula loosely tied to the worker's wage. Unlike most European systems, OAI has no "first tier" or basic benefit paid to all citizens as a matter of right. Instead, as historians Edward Berkowitz and Kim McQuaid note, America's public pension system relies entirely on private wages to judge a person's worth.
Throughout this century, work-based and non-work-based, public and private benefit systems have proliferated. The Social Security Act has been expanded to add disability insurance benefits and two immense medical programs, Medicare and Medicaid. In dollar terms, the growth of the work-based benefit has been stunning. From less than 5 percent of overall compensation in 1950, fringe benefits tripled by 1983, and nearly doubled again by 1990, reaching 27.3 percent of compensation. One estimate by the Social Security Administration suggests that noncash benefits will account for 40 percent of the total value of employee compensation by the middle of the twenty-first century.
We have, then, a history in which private employers invented the fringe benefit, government adopted their models for the public sector, and labor, representing the good worker, ultimately championed the result. The product is the work-based welfare state, an odd mix of public and private provision, including contractual benefits (like health insurance), statutory benefits (like unemployment insurance), and benefits that are both (OAI and private pensions). This structure fits well with an ideology that glorifies individual initiative in the private sector and mistrusts government provision. It is, as a Ford Foundation blue ribbon panel once described it, "the peculiarly American system of employment-related social welfare protection."
A welfare system tied to the work force by definition excludes those who have no jobs and those who are not recognized dependents of workers. But the current system also leaves millions of productive workers without critical benefits, such as health insurance. Between 30 and 40 million Americans have no health insurance, and non-working adults account for only 18 percent of this uninsured population. Nearly 50 percent of the uninsured are employed adults, and another 33 percent are under the age of 18. Two-thirds of those who are employed but uninsured work for an employer who does not offer health insurance. An additional 22 percent are ineligible, for a variety of reasons, for the coverage that their employer ostensibly provides.
Similar gaps occur in old age benefits. In 1979, 50 percent of full-time private sector workers were covered by pension plans; by 1992, the figure had fallen to 46 percent. Two phenomena are likely to make matters worse rather than better. First, the sectors that have traditionally provided pension coverage, such as manufacturing, are shrinking, while those where job growth is likely--smaller employers in the service industries--are notorious for their resistance to adopting pension plans. Second, many of those employers who do offer pensions are changing the form of their plan from one that does not require employees to contribute to one that does. The once dominant "defined benefit plan" is increasingly being replaced by so-called 401(k) plans. If an employer offers a 401(k) plan, an employee must take money out of her paycheck and put it in her pension fund in order to "join" the pension plan. If she cannot afford to do this, her employer will contribute nothing, and she has no pension. This shift to 401(k) plans effectively forecloses low-wage earners from pension plan participation.
Our work-based welfare scheme lets private employers decide whether their employees will receive certain benefits. The check on this decision is, presumably, employee bargaining power. In a free market, employees can refuse to work for an employer who does not offer benefits, forcing that employer to change its policies in order to compete. But such a check will only work, if at all, in a tight labor market, or in a world of empowered employees--hardly the conditions of contemporary America. Instead, unskilled, non-union workers find that jobs are scarce, and employers know they can fill them with employees who will work without health or pension coverage.
Ardent free market advocates might view this state of affairs with equanimity: if being without benefits is undesirable, workers will be motivated to organize or to improve their skills. But even if these efforts could increase the employees' stock of private benefits, some public benefits would remain out of reach. For some workers, bars to good worker status are built in to the statutes, and access will be denied until the law changes. An example is workers' compensation.
Workers' compensation is a collection of state programs providing benefits to workers who are injured on the job. Because they operate under state law, the programs vary, but all of them exclude certain workers from coverage simply because of the type of work they do. In a majority of states, all domestic service workers and all "casual" workers are ineligible for workers' compensation benefits. Many states also deny coverage to all agricultural workers and to employees in very small firms. Some of these workers are excluded from other statutory systems as well. Unemployment compensation, for example, categorically excludes workers on small farms and domestic workers who perform services for several employers.
The point, of course, is that neither statutory nor contractual benefits offer coverage to all workers. At the bottom of the wage scale is a cluster of occupations whose workers are categorically excluded from benefit statutes and who lack the bargaining power to negotiate contractual benefits. For these workers, and for their children, the work-based welfare state has never worked.
THE DEVIL IN THE DETAILS
Statutory exclusions and contractual omissions are, however, only the system's glaring flaws. Less visible, but equally treacherous, are the byzantine rules governing waiting periods, vesting, and work interruptions that deny many workers the benefits their employers theoretically provide.
An employee beginning a new job frequently learns that some of her benefits will be withheld until she has worked a certain number of weeks or months. If she gets sick, is disabled, or dies before completing her waiting period, she and her survivors have no benefit claim. Waiting periods effectively mean no coverage, and they are extremely common. A 1990 study of midsize and large firms by the Bureau of Labor Statistics (BLS) found that 49 percent of employee health plans and 51 percent of life insurance plans impose waiting periods. The most common waiting period is three months, though a one-month wait is nearly as prevalent. An occasional plan requires a longer wait, such as six months (8 percent of health plans and 9 percent of life insurance) or, rarely, one year (1 percent of health plans and 2 percent of life insurance). Not all workers have their benefits withheld during waiting periods; lower status workers are more likely to be affected. The BLS study found that 58 percent of production workers but only 38 percent of professional and administrative workers have waiting periods for health benefits. For life insurance, 59 percent of production workers but only 41 percent of professionals and administrators are subject to waiting periods.
Since waiting periods go with the job, not the person, each job change begins the clock anew. For the good worker, whose work life involves a stable association with a single employer, waiting periods don't matter very much. They are endured once, and full benefit entitlement follows. For the frequent job changer, however, mobility means repeated stays in the limbo of waiting periods, with no benefit coverage.
To know how many employees are caught in waiting periods in any year, we would need a way to identify job changes. For 45 years, the BLS measured job turnover, but the practice was stopped early in the Reagan administration because of budget cuts. Recent data on job turnover is thus unavailable. We do, however, have data on the more dramatic episode, the "job interruption," which involves a temporary departure from the labor force rather than a mere job shift. There the evidence clearly shows that women and blacks experience many more job interruptions than white men. In 1984, the Bureau of the Census reported that 72 percent of the women but only 26 percent of the men in its survey had experienced a work interruption. Among professionals, the gender difference was equally pronounced: 61 percent of women and only 15 percent of men had experienced a job interruption. Sorting the data by race showed that 35 percent of black men, but only 15 percent of white men, had experienced a work interruption caused by an inability to find work.
Job turnover and job interruptions are not the same thing, and knowing that women and blacks experience more interruptions than white men does not necessarily mean they also experience more job turnover. However, if the last hired is indeed the first fired, we should expect to see more turnover among newer employees. This newer group will include many women and minorities since, by virtue of their multiple labor force entries, they are "new" more often than white men.
Whatever the composition of the affected group, the result of job turnover and interruption is clear. Benefits tied to jobs are only as secure as the jobs they are tied to, and new benefits may not be immediately forthcoming, even if a new job is.
PATCHING THE GAPS
Two major pieces of federal legislation are a direct response to the economic vulnerability of the mobile worker. The Consolidated Omnibus Budget Reconciliation Act of 1985 (COBRA) gives employees the right to remain in their employer's health insurance group for up to 18 months after leaving a job. This provision partially disentangles work and benefits, allowing an employee who loses one to keep the other. The continued coverage that the statute provides is, however, entirely at the employee's expense. Many former employees cannot afford to maintain their coverage while they search for a new job or wait for new coverage to kick in. In the interim, they and their families will probably have no health insurance at all. Moreover, those who try to buy insurance on the open market, and even those who find a new job that offers health coverage, are at risk of being denied insurance because of a pre-existing condition once their 18-month COBRA entitlement has lapsed.
Eleven years before COBRA, Congress grappled with a different ramification of job turnover and interruption--loss of pension benefits. The closing of the Studebaker plant in South Bend, Indiana left thousands of long-term workers without pensions and launched a congressional investigation of pension plans that quickly uncovered a rat's nest. Not only were plans generally underfunded, but the events that would disqualify an employee were so numerous that few workers would ever see a pension check. Even in a group of particularly stable plans, the Senate Labor Committee found that only 4 percent of participants ever qualified for benefits.
The product of this investigation was one of the most important and complex pieces of legislation in this half century: ERISA, the Employee Retirement Income Security Act of 1974. ERISA does not uncouple work and benefits in quite the manner of COBRA, but it is similar in important ways. ERISA imposes a stringent set of pension "vesting" rules on employers. The purpose of a vesting rule is to give employees an absolute right to their pensions after they have worked for the employer for some set period. Since 1986, the most common vesting period is five years. Once employees complete five years' service, they can leave the company at any time and still receive a pension from that company at retirement. Once vesting occurs, the pension and the job are unlinked.
For employees who leave before vesting, however, the news is not so good. Unless they return to the same employer within five years (and, in certain cases, less) they forfeit everything. Any contributions standing in their names go back to the employer. Unlike the good worker, whose stable association with a single employer will insure that his pension vests, mobile workers see theirs go up in smoke.
Job stability also enhances the dollar value of pensions. Many pension plans base benefits upon the wage employees earn in their best three or five years with the company. Since the good worker works for one company all his life, his benefit will be based on the highest wages he earns in his lifetime. For the job changer, however, there is a problem. If she worked for an employer long enough for her pension to vest but left the employer some time ago, her benefits will be based on an older--and presumably lower--wage. The costs of her job change will thus follow her into retirement and stay with her for life.
But this assumes that the job changer's pension vests at all. If she is a woman, chances are that it will not. In 1983, the median job tenure for all males in the work force was 5.1 years, just over the ERISA vesting threshold. For women, the median was 3.7 years, well shy of ERISA minimums.
Exclusionary regulations are not limited to employer-provided benefits; they pervade statutory schemes as well. Social Security Disability Insurance (SSDI), for example, is available only to workers who have been employed for some minimal period prior to their disability. Standing alone, such a requirement might seem reasonable, particularly since SSDI is a contributory program. But a welter of regulations requires not only minimal contributions but also recent contributions. For an employee who has a gap in her employment history, these recency rules can produce bizarre and inequitable results. (See "Social Insecurity," opposite page.)
The losers in all of these stories are guilty of the same sin: they deviate from the good worker model. They work for low wages, they change jobs frequently, or they leave and re-enter the labor force. By contrast, the good worker earns a relatively high wage. This increases the likelihood that his employer will offer him health insurance and a pension, and it reinforces his ability to bargain for those benefits, or to go elsewhere if his employer does not offer them. The good worker's good wage directly enhances benefits like pensions (both Social Security and private) and unemployment compensation, which are computed using the wage as a base. His long tenure means that he has fulfilled all waiting period and vesting requirements. His lack of interruption secures him against "recency" rules and also maximizes his pension benefits. And our good worker will provide his dependents with a link to the most generous provisions of the work-based welfare state.
GOOD WORKERS, GOOD WIVES, AND GOOD JOBS
Some workers, like those in agriculture and domestic service, have always been denied good worker status. In recent years, however, structural changes in both the family and the work force have swelled the ranks of those on the outside. High divorce rates and the rapid expansion of the "contingent" labor force have together fueled the growing insecurity of American workers.
In 1940, most women married for life, and once married, they found their access to paid work barred, sometimes by law, and frequently by employer practice. Although black women, married and single, continued to toil at the bottom tier of the work force, white married women were largely excluded. Obviously, these women could not hope to attain economic security as good workers. Instead, they were to seek it as good wives. Most good worker benefits cover dependents, and it was through this status that white married women were linked to the work-based welfare state.
Just as the good worker needs a lifelong attachment to his job, however, the good wife needs a lifelong attachment to a man. By 1980, she was nearly three times as likely to lose that link as her counterpart of 1940. When the institution of marriage is fragile, dependency-based benefits are risky.
Certainly, as marriage has gotten shakier, women's links to the labor force have grown stronger. Sex discrimination in the work force is now illegal and most married women work for pay. Many, however, never attain good worker status. Women remain concentrated in low-wage fields where employers offer few or no benefits. And it is still women, not men, who leave the labor force to care for children or for elderly parents--their spouses' as well as their own. These departures and re-entries disqualify many women from such benefits as their employers do offer.
Like other groups caught in the benefit squeeze, divorced women have brought their issues to Congress. COBRA, the law permitting ex-workers to buy health insurance as part of their former employer's group, also applies to ex-spouses of workers. And the Social Security Act has been amended five times since 1950 to give divorced women better access to its benefits. Despite this backing and filling, however, many women find that divorce expels them from the protected class of "good worker's dependent" to join the ranks of the benefit poor.
While divorce rates spiraled upward from the mid-1960s into the 1980s, a second trend was building: a stunning growth in the ranks of part-time, temporary and contract workers. (See Virginia DuRivage, "Flexibility Trap: The Proliferation of Marginal Jobs," TAP, Spring 1992.) During the 1970s and 1980s, many businesses tried to increase efficiency by making their work forces more elastic. Layoffs are a crude tool for this purpose, since they undermine employee morale and boost unemployment compensation costs. Shifting to temporary or contract workers, however, enables employers to trim the work force without layoffs.
Few temporary or contract workers receive health insurance, pension coverage, or other contractual benefits, and their short-term employment is often peppered with gaps. These gaps exacerbate the workers' vulnerability by disqualifying them from statutory benefit coverage as well. Unemployment insurance, for example, covers only those employees who have worked for a sufficient period (varying by state) in the recent past. In the 1970s about half the unemployed qualified; by 1986, the number had dropped to one-third.
If the 1960s pushed divorced women out of the good worker fold, and the 1970s added contingent workers, the newest casualties may be former good workers themselves. As once invulnerable giants like GM, IBM, and Boeing "restructure," many men are losing the only job they ever knew. Theirs was the classic good worker tale, now lacking the expected happy ending. Their plight, the focus of much attention in the recent presidential campaign, may finally bring the frailties of the work-based welfare state into focus.
RETHINKING THE WORK-BASED WELFARE STATE
It is hard to dispute President Clinton's assertion that America needs more good jobs. With national unemployment over 7 percent and with areas where the rate exceeds 13 percent, it is painfully clear that the economy is too small for the work force, at least cyclically and perhaps chronically. But if we focus on jobs alone, we will miss a golden opportunity to make more fundamental--and badly needed--repairs to the work-based welfare state. The problem is not simply lack of jobs, though that problem is real and must be addressed. The problem is that many people work for a lifetime and never have a good job. It is difficult to believe that any job creation strategy will change that fact. What we can change, however, are the consequences.
In the heady days of the postwar economic boom, we compounded Social Security's fundamental error. At the time it seemed feasible--even profitable--to have basic social contracts run not from government to citizen, but from employer to employee. So long as the employee's dependents were covered, it seemed that all worthy citizens would be secure. But the bright world this system envisioned always left many workers--particularly racial minorities--by the wayside, and as boom has turned to bust, it has pushed out many others. We can continue to ignore these gaps, of course, and concentrate instead on good jobs for the workers cast off by IBM. But we have another choice as well. We can take the bold step of giving all workers a link to the protections the good worker has long enjoyed.
Opening up the good worker's world will require at least two kinds of efforts. First, we must identify and abolish the mechanisms that now shut out many mobile workers and that make others afraid to move for fear of losing benefits. Second, we must insure that all workers have access to the network of protections that good workers now enjoy.
Denying economic security to new employees, re-entrants, and workers with less than five years on the job is fundamentally unfair. It assumes that employers can and should demand that employees show a commitment to their jobs before they are offered basic social protections. Yet no concomitant commitment is required of employers who are, in fact, rapidly moving in the opposite direction. As they cut "core" workers and expand contingent work, employers are reneging on the commitments they formerly made to employees. Limiting social protections to longer term workers is also anachronistic. In his State of the Union address, President Clinton noted that the 18-year-old of today will probably retrain eight times during his work life to keep pace with job demands. But our benefit system functions like a 1940s Ford plant--it assumes that the good worker is the one who stands still.
Knocking down these technical hurdles will, however, solve only part of the problem. We must also face the fact that when we locate vital social provisions in the employment contract, we give employers the power to just say no. So long as they have this power, many American workers will remain economically insecure, and attempts to regulate fringe benefits that are not mandated by law will be futile. Fortunately, this issue is already on the table. Any plan for universal access to health care must address the fact that some employers offer health insurance and some do not. Whether the route to universal coverage is via employer mandates, direct government provision, or some other plan, the solution to the gaps in health coverage can serve as a model for expanding the scope of other good worker provisions as well.
Admittedly, the task is daunting, but there is reason for great hope. When we talk about fairness for those who work, we tap a large reservoir of fundamental agreement. Most Americans believe that people who work should get by--not in luxury, perhaps, but with some modicum of peace of mind. In addition, many of the needed changes seem almost like technical adjustments. Waiting periods, vesting rules, and recency requirements don't entangle us in constitutional combat or in deep moral quandaries. The idea of eliminating them is not shocking. What is shocking is that some hard-working people are well cared for by a welfare state they access through their work, while others, equally hard-working but less powerful, are left out.
Our new president tells us that he came to Washington to represent those who work hard and play by the rules. He will need to begin by making sure that the rules themselves play fair.