Multiemployer Plans and the Chronic Lack of

Retirement and Health Security in

Low Wage Labor Markets

 

 

Teresa Ghilarducci

 

Associate Professor of Economics, 510 Flanner Hall, University of Notre Dame, Notre Dame, Indiana 46556. 574-631-7581.

 

 

DRAFT- PLEASE DO NOT QUOTE WITHOUT PERMISSION FROM AUTHOR

 

Prepared for the Twenty-Fifth Annual Association for Public Policy
Analysis and Management
Research Conference, held
November 6-8, 2003 in Washington, DC. 

 

Note for conference: "The 'M' in APPAM" I am aware that the APPAM conference organizers, for APPAM's Silver Anniversary, wanted presenters to focus on “the critical intervening variable” - Who gets these outcomes to happen and how do they do it? I think I have found that unions in a few areas have played the role of “administrative entrepreneurs” as “outers" to coordinate employers to provided pension coverage to often low income workers, or workers who would ordinarily not be included in a health or pension plan.

 

 

 

Title of panel: Building a Better Pension System: How Effective Are New Proposals to Improve Retirement Security?

 

 

 

 

Documenting the persistent scarcity of employer–provided pension and health coverage for low-wage workers is not hard to assemble – the rolling strikes of thousands of workers in the fall of 2003 trying to keep their health and pensions coverage is just a dramatic expression of daily worries facing almost every American household, especially those towards the bottom of the wage distribution.

The first part of the paper reviews the studies on employee benefit coverage for workers earning below the median and for employees in industries with wages lower than average (Munnell and Halperin 1999, Hinz and Turner 1998, Fronstein 2000, Calabrese and Medoff 2001, General Accounting Office 2002, Commonwealth 2003,). There are three reasons a worker may not be covered by a pension plan, they do not work for a firm that provides one, they aren’t ever or yet eligible to be covered, or they are eligible bur opt out (generally only an option for DC or 401(k) type plans). Low income workers are especially vulnerable to all three sources of non-coverage. Workers who earned more than $60,000 in 2002 were three times more likely to be in a workplace retirement plan than those earning $20,000 or less per year. Overall, most groups in the labor force (except higher paid women) are less likely to be participating in a workplace retirement plan – retirement plan participation has fallen for a third year, from 58% in 1999 to 53.5% in 2000 (Congressional Research Service 2003).

Trends in employee benefits contribute to the ever-growing inequality in the distribution of compensation. Such inequality affects the distribution of well-being among the families and the elderly. The unequal distribution of workplace pensions follows into the retired population. Only social security and the home ownership counteracts the un-equalizing effects of business assets, stocks, bonds, and employer pensions in income and wealth inequality among elderly Americans (Slottje, Woodbury, and Anderson 2000, Wolff 2003). This paper will refer mostly to pensions, though health insurance and other types of fringe benefits are discussed.

The second part explores often-contradictory views of researchers about what causes chronic gaps in coverage between higher and lower income workers. We examine four theories explaining the low rates of coverage: 1.) the competitive labor market theory, 2.) the noncompetitive market, segmented market hypothesis, 3.) the public policy theory and 4.) the bureaucratic-institutional theory.

The third part of the paper analyzes who among low wage workers actually obtain coverage with a keen eye towards whether public policy could facilitate the institutions and conditions that give and extend coverage in middle class employee benefit plans to low income workers. The multi-employer plan is described in great detail as a mechanism that closes the coverage gap by filling the cracks in coverage that plague low income workers: working for small firms (who face high fixed administration costs); working for firms that favor high income workers in providing employee benefits; working in part time, unstable, casual, temporary jobs that don’t provide pensions or eligibility; earning such low incomes that needs, more urgent than voluntary savings for retirement, come first.

The brief fourth section revisits arguments against public policy relying on employment-based, social insurance programs. Last, we review various policy recommendations for improving low wage coverage. We take on the task mindful of the late Sherwin Rosen’s observation that:

 

Fringe benefits are of importance to such fundamental labor market problems as the social organization of work and production, as well as to social and moral obligations of citizens. These issues cut deeply into core issues in labor economics and, indeed of economics systems more generally. They deserve more attention than they have generally received from the economic research community. (2000: 29)

 

I. The Problem

We know health and pension coverage rates are falling fastest for low wage male workers – in 1979 a large percentage, 32.8%, of male workers at the bottom 10% of the income distribution had pension coverage, in 1998 only 17.8% did. For women at the bottom decile, the coverage rate stayed between 14.4 and 14.5 percent. Coverage rates at the top 10 percent for men also dropped, from 80.4% to 76.7%, but by a much smaller amount. In contrast, women’s coverage rates at the top actually increased from 41.3% to 42.7%. (Calabrase and Medoff 2001)  See Table 1.

 

 

 

 

Table 1:

Employee Benefits by Income Distribution and Occupation

 

 

Blue Collar

White Collar

 

1979

1998

1979

1998

 

Health Coverage

62.9%

49

69.3

60.3

Pension Coverage

46.7

33.6

49

49

 

 

 

Employer Spending 1979 – 1998 change

Health: Employer

Spending 1979 – 1998 change

Pension: Employer

Spending 1979 – 1998 change

Bottom third

-6 %

-35

Middle third

0

-27

Top third

+4%

-4

Source: Calabrese and Medoff 2000

 

Smaller firms are less likely than larger firms to have health and pension coverage, manufacturing more likely than service, white collar more likely than blue.  Workers aged 45-55 have the highest rate of coverage rate at 65 as do those with more education -- 79 percent for those with graduate degrees versus the 58 percent coverage rate for those with high school educations. The union effect on coverage is stark. 90 percent of those in union contracts compared to 54 percent of workers not in unions. White collar workers have higher rates of coverage than men blue collar workers, 68 compared to 42 percent (Fronstein 2001). 

Industry matters, a lot. Retail and service industries are the lowest paid and fastest growing industries and are the most likely to have secondary labor markets and part time and contingent workers. These are also the industries which are experiencing the fastest growth in cash as a percent of compensation. The slowest growing industries are associated with primary market jobs are the industries where growth in employee benefits as a share of total pay is increasing. Also, large firms are more likely to provide higher pay, a higher share of employee benefit shares, training and some degree of security. Yet It used to be a rule of thumb you could count on – which is the nature of rules of thumb – if you want employee benefits work for a large company. However, between 1987 and 2001, the rate of uninsured workers in large companies climbed by 57% (Commonwealth Fund 2003).

There are more studies about the pension coverage gap by income than why the gap exists. The idea advanced in this paper is the gap in coverage is due to the growing divisions in a dual or segmented labor market. The alternative hypothesis is the competitive market hypothesis, which concludes that competitive market forces produces the level of coverage that workers want to “pay” for with reduced wages.

 

II. The Reasons for the Pension and Health Insurance Gap

The motivation for the study is to evaluate how government policy helps (or hinders) pension and health coverage among low-income workers and whether more can be done to expand it. It is important to note that it is far from unanimous among economists that low coverage deserves a policy response, any more than the price of paper or type of car people own do. This argument, that coverage is not a problem policy makers should address, is based on the conviction that the distribution of health insurance and employer pensions are the result of efficient choices and agreements workers and firms undertaking economic exchanges.

The United States is unique among developed nations in depending heavily on the provision of social insurance through the employment relationship. The unique phenomenon is aptly referred to as “American exceptionalism:” the U.S. is an exception because in the U.S. the employment nexus is where health and life insurance, disability coverage, retirement benefits, and numerous other programs are provided although, in other developed OECD countries these are provided at the government level.  As a consequence of this country’s unique approach, most analysis of the employee benefit environment has focused on how and why a particular employer offers work-based health, pension, and other programs. 

Social scientists, including economists, have four major explanations for the development of a highly evolved system of work place “welfare” plans, often called “welfare capitalism,” in the United States. These are worth exploring if solving the problem and puzzle of low coverage among low-income workers. The four major reasons explaining America’s peculiar situation are: labor- economic; radical political economy or class struggle; public policy unintended effects; and bureaucratic entrenchment. Each provides insight into the puzzle at hand: why do low-wage workers have small amounts of pension and health care insurance coverage?

 

Competitive Market Explanations for Employee Benefits

Since employers have sound reasons for offering employee benefits and many workers have sound reasons for preferring pensions and health insurance, and would accept lower wages to obtain coverage, then “firms that do not offer the wage/benefit packages that workers desire can experience higher turnover rates as well as difficulties in recruiting” (Fronstein 2000, 92). The market explanation for low coverage rates among low-income workers is that workers and employer’s objective functions are being maximized in the exchange between them, which achieve the observed level of pension and health insurance participation. This view has been elaborated within a framework called the “economics of personnel” which view aspects of the employment relationship as adaptive responses to the needs of workers and employers (Lazear 1995). The largest and most chronic problem faced by firms since the beginning of the industrial revolution is the shortage of workers and the concomitant expense in recruiting and hiring. Pensions and health insurance were offered, the explanation details, because they are “tenure related” employee benefits -- they increase in value as the worker ages. This means that from the point of view of the worker, job leaving, turnover, is less tempting.

One solution was to arrange pay and the labor process to minimize turnover or minimize the costs of turnover (Jacoby 1997).[1]Health insurance and pensions help solve the chronic labor shortage. Therefore, the unusually large role of employee benefits in the American labor relationship, what is called “American exceptionalism,” compared to other OECD nations is explained by the American employers’ response to solve a persistent problem, turnover.        

From this “market economy” point of view the low-income workers don’t have “tenure-related” benefits because their loyalty or lack of turnover is not the firm’s concern.

There are four other related or sub-arguments to this overarching competitive economic explanation for low coverage for low-income workers. The first two focus on workers choosing not to have coverage and the second two reasons are why firms employing low wage workers may not want to offer pensions and health insurance.

First, the demand for employee benefits might be small for low-income workers, because these forms of compensation are income elastic. If wages grow more slowly for lower income workers (the trend towards income disparity was notable in 1980s and 1990s) then the demand for these benefits among these groups would be low (Hammermesh 1999).

Second, tax subsidies – employee benefits are not taxed -- provide significant savings for higher income workers because of the progressivity of the federal income tax schedule (Woodbury and Bettinger 1981, Reagan and Turner 2000). One reason low-income workers obviously have low coverage because they have low tax rates. Next are the reasons that focus on the firm’s motivations.

Third, some firms, perhaps the ones that face costly tenure, would grow adept at providing fringe benefits cheaply (this coincides with the bureaucratic view explained below) (Rosen 2000) and low income workers may not be employed in firms that can provide employee benefits as cheaply as higher paid workers’ firms. We do know that large firms pay higher wages than smaller firms cet. par. Therefore, low-income workers may not have pensions because their firms aren’t efficient in providing them (lack of experience and economies of scale perhaps).

Fourth, a fringe benefit may elicit more productivity, as well as reduce turnover, and in that way pay for itself (Gustman 1990) and low paid occupations may be such that even if workers were motivated by employee benefits to provide more effort and loyalty such effort and loyalty do not affect productivity in such a way that improves profits.

 

Noncompetitive Market Hypothesis

The second economic explanation of “American exceptionalism” focuses on the distribution of rents and other forms of economic power that are the result of noncompetitive markets and strategic decisions by firms and other. One argument explains employee benefits as stemming from the need for employers to “divide” the labor force, prevent unionization, and to maximize effort for the least work. This is a particularly important project for American employers because they face a potentially powerful workforce because of the persistent shortage of skilled labor. In this view, employers took on a mantle of patriarchy and paternalism in order to curry loyalty and other forms of allegiance among workers. Indeed, the firms most successful at staving off unions and being paternalistic were well liked, in general by their workers and were privately held (Jacoby 1997).

Employee benefits segmented the “elites” from the “non elites.” The elites had jobs in firms that were core in the production process and their employers could afford to pay well and the income was more stable. Welfare capitalism exaggerated the difference between the haves and have-nots even more than the distribution of cash did. Evidence provided that ‘welfare capitalism’ was used to block unions and divide workers is in the case histories of various long-standing companies such as Sears and Roebuck (Jacoby 1997). Unionization and unionization density in an industry may be important explanatory variables. If low-income workers are not unionized or in industries that are their low rates of coverage could be explained by that.

Despite the differences explores above there are many similar features between the two genres of thought. The economics of personnel and the class struggle theories are somewhat compatible. In the case of the economics of personnel though the ‘technology’ and demand conditions faced by the firm are not chosen or constructed by the firm, dual economies can emerge whereby large core firms adopt production processes that depend on low turnover and peripheral firms thrive on low paid, “weakly attached” employees who are inexpensive to hire and train. Likewise, in the “radical political economy” view the firm adopts strategies, including forms of compensation that emphasize divisions between workers, raises the cost of dissent and unionization. Both result in structures of labor markets that separate, segmented, so that workers, who have equal skill, effect similar levels of effort, etc., are paid differently depending on the sector they are in. The segmented hypothesis is that workers’ pay depends more on what segment they are employed in rather than their productivity. Since dualism is generated around whether employees are costly to replace (or difficult to monitor) pay strategies that reward longevity with the firm would likely appear in the primary or core sector. Thus, when health, pensions and pay are included as earnings we would expect the gap between the two markets would be larger than if only cash wages were considered, which is exactly the current situation in the U.S. Instead of helping equalize well being by subsidizing insurance, the employee benefits system makes compensation more unequal. We explain the difference between one market and dual labor markets in the appendix. [i]

 

Political Explanations for Fringe Benefits

The third view is that employee benefits are an accident of a history of public policy decisions that focused on managing the politics and economics of war.  This is the public policy explanation that employee benefits were the unintended effect of war. In war, the nation has to raise revenue, prevent inflation, and avoid political problems inherent in the tremendous amounts of profits some firms inevitably find themselves making when government demand for war related products soar. In addition, in the U.S. tax code had become a vehicle by which the government motivates certain kinds of behavior. Firms have many reasons besides a tax break to provide pensions for their elite, management employees. During wartime, without intervention, tight labor markets can cause higher wages that cause inflation. These factors, taken together explain why the U.S. government in World War II and in the Korean War taxed “excess” profits, profits significantly above historical trends, at 90%. Those profits would be tax exempt if they paid for to pensions, health, vacation, in other words, “noninflationary” pay to the rank and file as well as management. These rules stemmed inflation and avoided populist resentment by encouraging elite fringe benefits to be extended to the lower paid employees.

Thus, this view emphasizes workplace welfare as an accident of history even though union and companies cooperated in employee benefit provision and consolidated their cooperation around developing and modeling that type of employer contract after WW11. Path dependency would explain that industries with high coverage were those subject to war profits and who had faced unions early in the formation of their pay practices. Those industries with low coverage might have industries not directly subject to the excess profits taxes, like retail trade and that might be new industries.

 

Bureaucratic Explanations

This brings us to the last explanation of American exceptionalism and that is that interest groups have, over time, consolidated to a promote an employer based system because it is profitable for these groups and not necessarily because firms or workers particularly want them. Evidence for an entrenched bureaucracy explaining the predominance of employer benefits is that the American Medical Association and the insurance companies’ efforts continually explain the failure of Presidents FDR, Truman, Ford, Bush I and Clinton’s bids to nationalize health insurance. Low-income workers may not, for some reason, meet these bureaucracies’ needs. Perhaps, their pension amounts are too low to bother administering, etc.  

This bureaucratic view is interesting for our purposes because an institution called a multiemployer plan, which, as the name implies is a health insurance or pension plan that spans many employers, such as a construction trades pension plan for carpenters, or TIAA CREFF which is a single administrator of many universities’ pension plans for faculty. Many of firms in multiemployer structures would be small and ill-equipped to sponsor a pension plan if the multiemployer structure did not lower costs. Therefore, we expect low-income workers who are in multiemployer structures, to be covered because the cost of providing plans for the employer is lower than the firm size would suggest.

To summarize: I list below four reasons why low-wage workers may actually have pensions and four reasons why they would not based on the four views described above.

 

Why Low Income Workers Have Pensions

Low-income workers could have pensions and health insurance coverage if they want them and will trade cash wages for them. This is the neoclassical competitive market view that employers, in providing or not providing are merely accommodating employee preferences. Compensating wage differentials explain why some workers have pensions and others don’t.

Distinct from the “employee chooses” view, described above, is that high paid employees want coverage but employers constrained and motivated by the IRS anti discrimination rules coverage trickles down to lower income workers. This could cause low- income workers to have too much employee benefits.

The third reason low wage workers could have employee benefit coverage is if somehow employers of low wage workers face lower than average costs of including low wage workers in a employee benefit plan. An institution that lowers the cost for small employers, such as multiemployer plans do reduce costs of employee benefit plans for small employers especially.

Therefore, I hypothesize that firms that have access to an institution that pools resources and obtains economies of scale in administration will be more likely to provide pensions and health insurance. Multiemployer plans are such poolers of resources. Because multiemployer plans have to be maintained by a group of cooperating employers and the law recognizes collective bargaining agreements as a document that defines cooperation, only unionized workers in a few industries characterized by mostly small firms and mobile workers are in multiemployer plans that provide defined benefit (and DC) pension plans and health insurance. Using pension plan data from firms and plans (IRS Form 5500) there is a 28% correlation between low wage occupations and employees having their defined benefit coverage come from a defined benefit plan. Unions are the major coordinator of multiemployer plans. Making coverage cheaper by collective action means the employer is not indifferent between providing health insurance and pensions or cash. The table in endnote (ii)[ii] displays the extent to which defined pension coverage is derived from a multiemployer pension plan in certain industries.

Fourth, some low-wage workers may have skills or other attributes that make long job tenure, and loyalty, advantageous to the employer. If so the employers prefer to pay health insurance and pensions because of “the economics of personnel and these low wage workers are in some ways hard to replace. These low-wage workers might also be at some “risk” of unionizing and the firm wants to raise the cost of being fired for union organizing

 

 

 

Why Low Income Workers Do Not Have Pensions

The next set of reasons is about why low-income workers wouldn’t be expected to have pensions and health insurance coverage.

First, low-income workers may not have employee benefits because they don’t want them because they are younger or married females (it is a matter of preferences and no problem for public policy). This of course is also the “worker chooses” view, which implies there is very little for public policy makers to do.

Second, low income workers may want employee benefits and are willing to pay for them; but their employers do not know it and they lack mobility to move to find and employer who does offer employee benefits. A 2002 survey by the Transamerica Center for Retirement Studies (www.ta-retirement.com) found a substantial gap between employers’ perceptions of workers’ priorities and their actual attitudes regarding pension coverage.[2]  Low-income workers can’t afford health insurance and pensions, but low income workers would buy health insurance and pensions if low income workers could afford health insurance and pensions (therefore workers need subsidies like tax deferred medical accounts.)[iii]  The public policy implications are more education and subsidies.

Third, low-income workers work for industries and firms that face high costs of providing pensions, for instance small firms. Small employers (typically defined as businesses with fewer than 100 employees) give several reasons for not providing health and pension coverage: administrative burden, fiduciary and legal issues, an absence of in-house benefits expertise, cost, high employee turnover and/or a large proportion of part-time or part-year employees, the company’s uncertain future, and pensions falling relatively low on employees’ priorities. [iv]

Fourth, employers prefer to pay cash not employee because the employers’ attachment to low-income workers is low. Therefore, the probability of coverage may not be related to the how loyalty and effort are important in determining profitability. This would depend, in part, in how well the production process is protected from adverse effects of turnover. Low income workers stuck in the secondary part of the low wage labor market often have employers who prefer high turnover to keep wages low. Being in the secondary labor market explains low coverage.

 

III. Success Story: the Multiemployer Plan

First, I describe the historical development of multiemployer plans in the U.S.; second, the scope and special features of these multiemployer plans indicate adaptability and an ingrained bias toward the lowest paid worker. Last, we ask whether a multiemployer format might be general enough to be able to provide coverage to the substantial fraction of workers lacking employment-based pension plans in this country at present, and what role multiemployer plans might play in years to come.

 

 A Brief History of U.S. Multiemployer Pension Plans 

The Brotherhood of Electrical Workers and Electrical Contractors (Local 3) of New York was probably the first union to establish a multiemployer pension plan, in 1929 (EBRI 1997). However this was not an auspicious time since many pension plans disappeared during the Depression due to poor funding records. Other negotiated multiemployer plans were established in the needle trades and in coal mining, and in these industries, unions alone controlled their administration during this early period.

An important legislative innovation in the multiemployer pension business was put in place in 1947, with the Taft-Hartley amendments to the National Labor Relations Act (sec. 302c.5.). Under this regulation, employers sponsoring a multiemployer plan were required to have at least the same number of trustees on the pension boards, as did the union. This bill was initially opposed by organized labor, and to effect passage, Congress was forced to override President Truman’s veto.[v]

While the United Mine Workers of America (UMWA) was not the first group to negotiate pensions in the U.S., it greatly helped define the role that unions would have in establishing workplace pensions in years following the Great Depression.[vi]  Declining coal demand and automation of these dangerous debilitating jobs caused substantial worker displacement during this time. In 1945, the UMWA demanded a fully employer-paid, $100 per month, pension for old and retiring miners, characterizing the demand as “payment for past service.”  Instead of arguing that pensions were a type of deferred wage (payment for services rendered) that had to be accumulated before paying out a benefit, the UMWA (and many other industrial CIO unions afterward) argued that pensions were depreciation payments owed to labor and were analogous to employers' accounting for capital depreciation (Sass 1997). In this light, pensions could be seen as complements to, rather than substitutes for, higher wages. 

The UMWA demanded employer-paid pensions, but plan origins dictated that the coal miners’ plan be jointly-trusteed with the employers of these constituencies.  In the 1940s President Truman delegated his Secretary of Interior to mediate the negotiations in the key coal industry, and he encouraged the relatively weak employers in this vital industry to settle with the powerful union. 

Joint pension investment administration was initially not a hotly contested issue because the plans were operated on a pay-as-you go basis – that is, contributions collected were immediately paid out in benefits rather than investing the monies in a trust fund. Coal plan liabilities grew quickly over time as miners lost jobs during the Great Depression.  The union negotiated for and won a doubling of the royalties per ton that financed the pension during the period 1948-1952. President John L. Lewis argued that employers should share the benefits of automation since workers paid the costs. In an under-appreciated development, the UMWA pension plan negotiated a new funding base where contributions were structured as a function of tonnage mined per hour rather than work hours. This formula explicitly shared the benefits of greater productivity with the workers.

The industrial unions in the rubber, steel, and auto industries were also concerned with bargaining for employer-paid pensions to be paid immediately to current retirees and older workers. As a result, they concentrated on immediate payouts, rather than on building up a stock of assets in a funded plan. These unions did not argue for payments based on productivity nor did they bargain for control of the pension fund investments; in fact weaker industrial unions in a sense gave up control of the fund in exchange for past service credits.

During this period, many employers agreed to provide employee benefits to attract scarce workers and sometimes to accede to union demands. But social expectations led to the belief that employer benefits were needed to supplement the incipient social security program.  Pensions also helped employers manage skill supply and quality in dynamic industries such as construction.

 

Multiemployer Unique Framework Solves Low Coverage Due to Temporary Jobs and Small Firms

Most multiemployer pension plans in the United States are defined benefit plans and [3] generally cover many occupations in one industry, one craft in many industries, or many occupations in one industry.

A full 20 percent of active DB plan participants in the private sector had a multiemployer plan in 1996 (EBRI 1997). This represents an impressive. The figures are 59 percent of apparel employees, 73 percent of retail food store employees, and 39 percent of furniture industry participants).  The table in endnote ii. reports wage rates and shows that earnings in the service and retail area, where multiemployer plans are important delivery systems, are lower than average rate of $13.24 per hour.  However, multiemployer plans also deliver pension benefits in some highly-skilled labor markets as well.

Collectively bargained Taft-Hartley funds sponsored by the same union in a given craft or industry allow participants to build pension vesting and benefit service while holding jobs with different signatory employers who contribute to the same plan.  This is highly effective in facilitating labor supply in decentralized industries with migratory work forces like construction, trucking, retail food, the garment trades, and some of the service industries.  This level of local labor market portability in the Taft-Hartley world can be extended geographically through formal reciprocity agreements between different pension plans nationwide in the same industry sponsored by the same union.  Several unions actually administer national reciprocity agreements between large numbers of local and regional pension plans with formal dispute resolution procedures.

Examples of industry-wide, geographically based pension plans in the U.S. include the United Food and Commercial Workers fund in Northern California, which negotiates pensions and health insurance across a number of large employers including Safeway and other grocery stores. The International Garment Ladies Union and Amalgamated Clothing and Textile Workers Funds cover production workers across a range of employers in the needle trades.  Taking a broader perspective, the Sheet Metal Workers, Bricklayers, Carpenters, and other building trades funds cover particular trades operating across a wide range of diverse industries. The Western Conference of Teamsters pension plan covers many occupations in several industries in the 13 western states – grocery delivery drivers, warehouse workers, and long haul freight truckers.[vii]

 Multiemployer pensions are found in sectors other than the private sector: in addition, there are also multiemployer systems in the public and not-for-profit arenas including those established for churches, the Red Cross, charities, and, of course, university and college teachers. There are also hundreds of pension plans covering public sector workers including state, local, school, police, and firefighter employees (Mitchell, McCarthy, Wisniewski, and Zorn 1999).  These plans cover many hundreds or thousands of small townships, counties, state agencies, and school districts.  In total, there are over 15 million participants in these plans constituting just under 11 percent of all U.S. employees but more than 20 percent of all DB participants. Legislation enabling public section pension provision varies across state and locality, yet the structure is generally similar. Typically these plans are defined benefit in nature, though all 50 states also offer some form of defined contribution plan in addition. 

Some of the largest pension plans in the nation are the long-established church and educational institutions.  The Presbyterian pension fund began to cover church employees in 1717. [4]This plan now covers over 8,000 active ministers.[viii]  The Episcopal plan is joined by plans from several other religious denominations including the Baptists, Presbyterians, and Methodists. The Church Benefits Association in the U.S. has 45 Protestant plans and the Association also includes Jewish and Roman Catholic plans (Goldman, 1996). The Teachers Insurance Annuity Association and College Retirement Equities Fund, now known as TIAA-CREF, had its roots in a system endowed by Andrew Carnegie in 1918 for universities and college teachers. It is technically an association of single employer plans; but it is clearly one that functions as a multi-employer plan.

When establishing these pension systems, each of the plans had to confront a nettlesome “past service” issue during the plan’s startup phase. This refers to the political and practical difficulties of requiring young workers to pay for the pensions of current retirees who never contributed to the plan.  In practice, the problem was solved by eleemosynary action in the teachers and church plans (the rich donated enough to cover past service liabilities).  However, philanthropists were hiring and not retiring blue-collar workers and their employers. Workers and employers in industries characterized by competition and fluctuating demand either paid labor on a spot market or overcame collective action problems by creating funded pension plans that would deliver meaningful benefits and would be as trusted as banks to handle money well.

 

Multiemployer Plans Boost Training and Productivity

The following, quite long, and quite detailed, description of the governance structure is meant to impress upon the reader that workers are highly involved in the structure of their pensions and even though the workers may be low paid they have an employment based pension at the end of their working lives. The very nature of the joint governance structure may cause workers to more likely participate in these plans.

Management and labor trustees have a common purpose in the pension arena, though they may be opponents on other issues, in a multiemployer setting. The plans’ unique joint governance structure leads to many special outcomes.[ix] The legal trust arrangement coupled with reporting laws makes decisions public and transparent. The equal representation requirements help solve conflict of interest problems, because neither labor nor management can act to benefit their own sides alone.  In particular, trustee actions must be transparent and thus the funds must be participant-focused. As a consequence, the multiemployer fund may not forgo contributions in order to benefit an employer or preserve an agreement.  In addition, the plan governance structure ensures a participant focus by limiting means for contributions to be altered even in the face of potential actuarial gains. Ordinary definitions of prudent investment also apply; there is little difference in the asset allocations of public, private, and union funds. 

Multiemployer plans also help close the gap between high and low-income workers in terms of pension coverage because they offer an incentive for employers of low income workers to provide pensions[x]. Employers and workers in a multiemployer pension system pool their contributions so the same provisions, no matter where the job may be, as long as the employer is participating, cover workers. This will typically be in the same industry and, often, in the same geographic region. Employers in these sectors require workers with similar skills but because of the nature of their product demand, may not be able to hire workers on a long-term basis. In this context a collective action problem exists, such that no single employer has the incentive to provide pensions, health care, or training if the returns on these investments cannot be returned to the sponsoring firm. 

Multiemployer plans can be a mechanism by which firms and workers can share costs, thus solving the public good problem and creating economies of scale. Particularly when labor markets are fluid and jobs contingent, valuable skills derived from experience may be lost in turnover. In such a case, occupational loyalty can become increasingly important: the mobile workers obtain portable pensions and are able to save for retirement, while employers obtain skilled workers and a way to share in the cost of training. Thus, the multiemployer plan satisfies these needs by sharing the costs and benefits of increased mobility.

To describe how multiemployer pensions provide security in a dynamic market environment, it is instructive to consider the circumstances of a group of nurses employed in New Jersey healthcare providers. These nurses wanted to be included in a multiemployer plan operated by the International Union of Operating Engineers (nurses as operating engineers!) rather than their hospital’s single employer plan. Why? Hospital ownership has changed many times through mergers and restructurings, changing their pension along with the change in ownership. One nurse said she had been covered by six separate single employer DB plans, but she has no idea of what she’ll obtain from any of them. The pension plan administrator observed, “When an industry is poorly managed and unstable the multiemployer plan is the only option for meaningful retirement coverage” (Fanning 2001).   Here the multiemployer plan, by virtue of the diversity of its industry base and connection with a longstanding union, is seen as providing a more secure pension promise than the traditional single employer model.

In a multiemployer model, participating firms pay a specified contribution into the plan according to a collective-bargaining agreement that varies according to bargaining power. In other words, contributions depend on the company’s relative ability to pay and the union’s ability to bargain for benefits (Chamberlain and Kuhn 1986). Thus in the typical multiemployer plan, the employer’s contribution is defined but the worker’s benefit is also specified – the latter is what makes it formally a defined benefit pension. This is another hybrid aspect that makes them flexible in the face of changing conditions.[xi]

 During the 1970s recession, for instance, the Sheet Metal Workers Fund lengthened its rule regarding how long a worker could be out of employment before credited service was lost. This produces costs to the pension plan, of course, so other benefits had to be implicitly weakened in reaction. Nevertheless, this flexibility reveals multiemployer pension fund sensitivity to industry needs and worker concern with security. 

Multiemployer plans tend to have generous early retirement eligibility and disability rules, reflective of the blue-collar and low-pay nature of the jobs.  In addition, multiemployer plans tend to tailor disability and return-to-work rules to the needs of the particular group and accommodate employers’ differing abilities to pay.  The earliest age for collection of retiree benefits is 62, though employers may offer early retirement bonuses on an ad hoc basis. 

Multiemployer plans also have special early retirement provisions that are effective responses to changes in specific industry employment patterns. Recently some Taft-Hartley plans have also liberalized “suspension of benefit” rules prohibiting retirees from returning to work in their career industry, recognizing increases in demand for experienced employees.

Another way multiemployer plans have been responsive to the economic environment has to do with when the fund will cut off pensions to a pensioner who has come back to work. If labor is in short supply, the rules are liberal; if there is high unemployment the rules are restrictive. [xii]

Of course the strict suspension benefit rules can hurt union employment if there are no union employers for which the older individual can work, particularly if labor markets are tight. In such cases, it is likely that the suspension of benefit rules would be liberalized so that pensioners can work without losing benefits. Further having retired union members working for a non-union employer can improve the chances of successfully organizing. In some cases having a union employee working for a large nonunion company offers “salts” in places the union is seeking to organize.

Most multiemployer plans in the United States developed reciprocity agreements with each other in the 1970s and 1980s.  Reciprocity agreements are arrangements between local unions with different plans permitting continued benefits coverage when each others’ members work across jurisdictions. While reciprocity methods vary, the result is that workers continue to be covered when they move between member employers.

 

Multiemployer Plans Solve High Administrative Costs   

The average collectively-bargained multiemployer plan is larger than the average single employer plan, enabling this multiemployer format to take advantage of scale economies. Twenty percent of single employer defined benefit participants are in plans with more than 50,000 participants, while the fraction is 42 percent among multiemployer participants (USDOL 2000).  Many researchers, who find that larger pension plans are less expensive to administer than small plans, confirm scale economies. For example Hustead (1996) found that small funds (with 15 participants or fewer) spent about 58 percent of normal cost on administrative expenses, while large plans (with 10,000+ participants) spent about 2.5 percent of normal cost. Of course these costs vary with participant mix, since service levels differ depending on the fraction of actives versus retirees (Mitchell and Andrews 1981; Ghilarducci and Terry 1999).  On the other hand multiemployer plans are often regional and the fact that they are sometimes small explains why they were formed in the first place.

   All defined benefit plans entail cross-subsidies across members, a hallmark of insurance. The clearest one involves transfers between retirees who die earlier than their life expectancy versus those who live longer than their life expectancy.  In addition there have been periodic discussions about other types of cross-subsidies in the multiemployer context, such as between large and small employers, or wealthier and less well-off union locals. One case highlighting this issue occurred when the United Parcel Service proposed in 1997 that its employees leave the Teamsters multiemployer pension plan and instead have their own single employer plan. The employer argued that it was subsidizing smaller employers in that multiemployer framework, a point that the union partially conceded by recognizing that UPS membership helped achieve scale economies. On the other hand the union contended that without an actuarial study, it was impossible to know whether the resulting single employer plan would provide benefits that were as least as good, for less money.  In addition, the UPS workers would lose their trustee representatives in a single employer plan, which they believed would affect the chances of their receiving ad hoc COLA increases in their pension benefits, and portability rights to other participating employers would also be lost. 

In another context, some members of the Central Pension Fund (CPF) of the Operating Engineers have experienced higher levels of contributions and growth than others, leading to some concern that they are “carrying” the poorer and shrinking locals. In this case too, some 80 actuarial studies would have to be conducted to determine whether each local’s past and projected experience would yield better benefits than under the CPF.  Thus far no study has been conducted of the cross-subsidies and how they would be unwound if the parent plan were to be broken up.

In the Episcopal Church, there is an explicit policy of redistribution between higher- and lower-paid employees. This is evident in the pension formula, where the basic benefit consists of a higher percentage (1.75 percent) of the average of the highest seven years for the first $10,000 of salary, but 1.5 percent for pay beyond that.  The benefit is not indexed to inflation formally, but the Episcopal fund has provided an ad hoc cost of living adjustment since 1980 by issuing a “13th check” that is based on years of service. Within every level of service, the 13th check is the same however (e.g. a minister with a $7,000 benefit will get the same amount as one with a $30,000 benefit; Blanchard 2001).


 

IV. Challenges to the Employer-Based Benefit Paradigm and Implications for Multiemployer Plans

As noted above, pension coverage in the United States remains voluntary, and has leveled off at around 50 percent of the workforce.  The fact that coverage has not risen despite a booming economy and labor shortages in some sectors suggest that the labor market is segmented and that the growing secondary sector are not interested in providing fringe benefits. Wal-Mart, once a follower of the union sector employee benefits structure, has grown so large it now sets the pattern on the composition of compensation and that is to provide fewer benefits (Fuhrmans 2003).[xiii]  The largest private sector employer, Wal-Mart, with over 1,000,000 employees, provides health insurance to less than 40 percent of its workforce, and instead the company sponsors individual account savings plans dominated by investments in its own corporate stock.  Also, private sector trends toward “individual responsibility” have been manifested in the outright elimination of DB pensions and retiree health coverage (Weller 2003 and Wolff 2003).

Some argue that the decline of employee benefits is caused by the “social contract” between workers and employers having collapsed (Osterman 1999). The reasons given for the collapse and the rise of individuals having to take on the risks and responsibilities for their own health insurance and pension security range from worker choice to workers’ losing bargaining power. Worker mobility may be driving the trend to DC plans, because pensions now need to be de-linked from a single employer while weaknesses of the DB system eroded worker confidence that employers are committed to pensions. The 1990s bull market may have led some to discount the investment risk inherent in DC plans (Schiller 2000).  In general, these explanations point to workers rejecting traditional benefits like health insurance and pensions and paternalistic social contracts. Yet, employer norms, needs, and increased market power relative to workers may be causing diminished employer willingness to sponsor work-based insurance programs. To illustrate this point, when workers express their choices via unions and collective bargaining, they tend favor benefits: have over 30 percent more compensation in employee benefits.[xiv]  Additionally, they select DB plans with supplemental DC programs. This is key point. It seems if workers choose in a social forum and not in the intimate confines of their home and financial abilities individuals make different choices. The decline in unionism explains a good deal of the decline in coverage and the change in structure of employee benefit plans.

Policymakers and analysts from both the conservative and liberal spectrum contend that employee benefits should be severed from the employment relationship, especially to solve the problem for low wage employees. Market oriented reformers call for more personal tax credits for individual employee benefits, medical savings accounts, and E-commerce individual health plans.

Calabrese (2001) and Munnell and Halprin (1999) and Weller (003) argue for a reduction in the workplace model. Calabrese (2001) calls for putting the coverage of and “enfranchisement” of low-income workers as a major goal of any policy shift. He argues that federal retirement policy should be recast to be “citizen-based” rather than employer based and that federal tax policy should subsidize the cost of coverage by using matching tax credits and not tax deductions. He argues that integrating matching credits with the existing system and with social security can achieve more portable, universal, and equitable asset-building.

There is convincing evidence that it is the lack of economic power rather than employee preferences that are causing the decline in employee benefits. Labor economists tentatively agree that job instability is increasing,[xv] which could mean either that workers are voluntarily leaving their jobs after shorter periods of employment, or that firms are offering low-commitment relationships.  If employees were choosing to be more mobile why would the group that has most to lose in terms of pensions be the one more likely to move? The jobs where instability has increased the most are those for whom pension accruals are the most crucial: for men age 45–54 and 55–64, and particularly for African-American men. Workers with more than nine years of service are those on the brink of vesting, but this group is experiencing the largest declines in job stability. Women and workers with less than two years of service have been least affected. Occupational data shows managerial and clerical job security falling significantly but less than for service and blue-collar workers. (Neumark et. al. 2000). In sum, the increase in cash-intensive pay and job instability suggests that employer commitment to workers over the long term is waning. Yet, individuals may not tolerate taking on retirement risk as the stock market gyrates and employers may rediscover the benefits of “strategic paternalism” to reduce turnover.

            Multiemployer plans, I suggest, are a potentially useful vehicle to serve the needs of employer and workers in secondary markets. Employers, particularly smaller ones, can find substantial advantages to joining a multiemployer plan. This is because the multiemployer structure allows them to provide pensions to themselves, office staff, and their workers efficiently, and help solve labor supply problems.

One way in which pensions are beneficial to employers is that having a pension plan tends to reduce worker mobility. As a consequence, employers can invest in employee training in the expectation that these workers will be there in years to come. Such “win-win” trades help make the economy more productive (Ghilarducci and Reich 2001).

Another way multiemployer plans help employers manage the supply of labor is through reciprocity, where workers from one geographic region can work in another geographic region without losing benefit coverage. [5]Such pension reciprocity in regional-multiemployer plans keeps employers and workers from overspending on training, and helps protect workers from fluctuations in earnings.

 

V. Solutions:  The Future of Multiemployer Plans 

Because growth in U.S. pension coverage has apparently stalled, several groups have called for a new institutional framework that can serve to increase coverage among the noninsured. As ERISA attorney Michael Gordon has noted, this requires devoting attention to “the twin areas” of small business and the non-traditional workforce (Gordon 2000).  He notes that in single-employer plans, benefit provision is often motivated by corporate tax breaks for pension contributions. By contrast, high turnover rates among small businesses render such incentives rather valueless. Instead, Gordon posits that the multiemployer model could be extended by offering tax incentives to workers and small employers for participation in a new form of statutorily-approved multiemployer plan. Gordon claims that some representatives of collectively bargained multiemployer plans would welcome employers not in a collectively bargained agreement. This would permit employees to have complete portability between member employers, akin to the voluntary plans in Europe.

One disadvantage to pooling is that firms have to overcome competing with one another in order to coordinate the establishment of plans. Unfortunately these barriers are so high that most multiemployer plans are coordinated by a union or are in a nonprofit setting. The exceptions are public sector multiemployer plans where governing bodies initiate the coordination. Another disadvantage, related to the first, is that the arrangements are voluntary. When one firm views its liability as much less than average they will have incentive to leave. (The adverse selection problem is one reason that social security, unemployment insurance, workers’ compensation etc. have been structured as mandatory programs.)

From one perspective it would seem that individual-oriented plans would be chief competitors and substitutes for multiemployer plans. After all, single-employer pensions cover the worker population relevant to a given firm. But perceived barriers to expansion of group-based multiemployer pensions may be turned around to help create them. The popularity of 401(k) and 403(b)s stress their portability advantages while stock market volatility highlights their disadvantages. Multiemployer hybrids, with probability and defined benefit features, may look better in contrast.

Moreover, much of the focus on these plans is mainly because of the technologies that can them more visible. Because of this narrow focus, workers have a difficult time seeing how their 401(k) plan fits in with their entire package of employer benefits. It is possible that instead of being a barrier to the creation and maintenance of multiemployer plans, multiemployer plan sponsors can facilitate employee education and provision of a range of benefits. There are at least 11 different types of multiemployer plans including pensions. Employers would gain from using the technology to help workers recognize and appreciate all the non-cash compensation provided.

            Perhaps, the technology allowing individuals and employers to find medical plans on the Internet do what group-based employer plans did which is to get economies of scales. Perhaps, the technology allows individual arrangements to be cheaper than employer coordinated plans. The fundamental impetus for employers to combine and provide pension is not economies of scale but to restrain competition and the “a race to the bottom.” The most profitable short-term strategy is to provide no benefits at all and get market share with low prices. But employers from the beginning to of the industrial revolution knew competition through cost cutting was cannibalism. Taking wages and service and skill-related compensation out of competition the industry can invest in more capital and training.

Multiemployer plans may become a key labor-organizing tool in the Silicon Valley. Amy Dean, President of the South Bay Central Labor Council, AFL-CIO, promotes multiemployer plans because of the chronic lack of coverage by small employers. “There is potential for taking advantage of economies of scale in administrative costs by pooling small employers in the same industry together thereby increasing the portability of pensions” (Brenner et. al. 1999:67). The union-based organization, Working Partnerships USA, envisions employer-training networks to be connected to the health and pension consortiums.

 

VI. Conclusion

In my view, the increasing prevalence of cash compensation and the decline of employee benefits as a fraction of payroll are attributable to a sea-change in the nature of the employer-employee relationship in the U.S. labor market. In contrast to years gone by, a lifelong mutual commitment between firms and workers is no longer the norm. The multiemployer benefit framework, in our view, offers an important exception to this trend. 

Although multiemployer plans in both the private and public sectors cover very different types of workers, from janitors to university presidents, they are similar in that they solve three key problems. First, but perhaps least important, multiemployer plans can take advantage of savings on large group annuities and professional management fees. Second, multiemployer plans recognize the skill management and insurance needs of a heterogeneous workplace, where patterns differ by unique industry/occupation.  Third, multiemployer plans can help solve the collective action problem: that is, no one employer may have an incentive to provide benefits or training without its competitors doing so, so joint action can be efficient for all. 

The GAO writes that the long-standing proposals to “expand pooled employer arrangements and mandate private pensions unattractive because they may increase labor costs (GAO 2002, 36). Others feel that expanding social security, providing a defined contribution supplement that is incented by a tax credit may boost retirement wealth for low-income workers.

I maintain, in this paper, that these reforms can’t be assessed without a framework describing how and why U.S. labor markets produce employee benefits, pensions in particular. Identifying the framework is important because each imply radically different policy responses. Human capital theorists argue for a set of policies that enhance workers’ productivity and provide more information and education, reduce the effect of unions, and reduce the reliance on inaccurate demographic signals of labor quality in order for total compensation to be “equal between equals.”  Labor market segmentation inspired policies would give up on the secondary labor markets chances of every providing fringe benefits for those who want to buy it without significant interventions such as unionization or other kinds of regulation.

The continuing importance of negotiated multiemployer plans attests to how a central agent, such as a union, can offer a key benefit function in the service of employees and also of assistance to employers. Compared to the one-size-fits all structure of social security, multiemployer plans can adapt to the idiosyncrasies of particular industries and occupations. In this sense, de-linking pensions from the single employer via multiemployer plans may hold out hope to resolving portability and income security problems. Multiemployer plans may also serve as a model for expanding social insurance across employers. These plans do what human resource experts and industrial innovation experts say must be done: the plans can lower the cost of training by reducing the chance of workers leaving the industry or occupation. 

The continuing importance of negotiated multiemployer plans attests to how a central agent, such as a union, can offer a key benefit function in the service of employees and also of assistance to employers. The existence of multiemployer plans suggests the ability of employers and unions to segment markets. I am sympathetic to the view that expanding social security may be the only effective way to provide retirement income security to low income workers.

Yet, I am impressed by how, compared to the one-size-fits all structure of social security, multiemployer plans can adapt to the idiosyncrasies of particular industries and occupations. In this sense, de-linking pensions from the single employer via multiemployer plans may hold out hope to resolving portability and income security problems. Multiemployer plans may also serve as a model for expanding social insurance across employers. The Department of Labor can hold down the cost of forming these multiemployer plans by providing safe harbors and model legal documents that groups of employers or employee organizations and their legal counsel can use to establish a multiemployer plan. These plans do what human resource experts and industrial innovation experts say must be done: the plans can lower the cost of training by reducing the chance of workers leaving the industry or occupation.  These plans also do what experts say is necessary for human development; they allow workers to have a meaningful say in the decisions that affect their daily lives. In this way, unlike social security they fulfill the goal of subsidiarity, where decisions are made at the smallest unit of community as possible.


 

Abowd, John M., Francis Kramarz, David N. Margolis, Kenneth R. Troske. 2001. "The Relative Importance of Employer and Employee Effects on Compensation: A Comparison of France and the United States." Journal of the Japanese and International Economies: 15(4), December, 419-36.

Agency for Health Care Research and Quality. “1996 Medical Expenditures Panel Survey – Insurance Component.” Table 1 – c- 1. http:// meps.ahrq.gov

Blanchard Alan. 2001. Administrator for the Episcopal Pension Fund. Personal interview with Ghilarducci, Feb. 23.

Brenner, Chris, Bob Brownstein, and Amy Dean. 1999. Walking The Lifelong Tightrope. San Jose, Ca: Working Partnership USA and the Economic Policy Institute.

Cain, Glen. 1976. “Segmented Labour Markets.” The New Palgrave Social Economics. Edited by Jonathan Eatwell, Murray Milgate and Peter Newman. Encyclopedia London The Macmillan Press, LTD, 225–229.

Cairns, J. E. 1874. Some Leading Principles of Political Economy.  New York: Harper and Row Brothers.

Calabrese, Michael, and James Medoff.  2001.  “The Impact of Labor Market Trends on Health and Benefit coverage and Inequality”.  For the Pension & Welfare Benefit Agency.  U.S. Dept. of Labor, Feb. 28.

Calabrese, Michael. 2001. Presentation to the ERISA Advisory Council, U.S. Department of Labor, Working Group on Pension Coverage, Participation and Benefits, June 11.

Chamber of Commerce. 1999. Employee Benefits Survey 1999 and Annual Benefits Report.  U.S. Chamber Research Center: Washington D.C.

Chamberlain, Neil, and James W. Kuhn. 1986. Collective Bargaining. Third Edition New York: McGraw Hill.

Commonwealth Fund 2003. “The Growing Share of Uninsured Workers Employed by Large Firms” by Sherry Glied, Ph.D. Sarah Little,  Jeanne Lambrew, http://www.cmwf.org/programs/insurance/glied_largefirms_2ndpg.asp

Congressional Research Service. 2003. “Annual Report on Pension Sponsorship and Participation.”

Dickens, William T., and Kevin Lang.  1985.  “A Test of Dual Labor Market Theory,” American Economic Review 75: September, 792-805.

Dickens, William T., and Kevin Lang.  1992.  “Labor Market Segmentation Theory: Reconsidering the Evidence,” Working Paper No. 4087.  Cambridge, MA: National Bureau of Economic Research.

E.C.I., Bureau of Labor Statistics. Various years. Employment Cost Index. Washington, D.C.: United States Department of Labor.

EBRI, Employee Benefit Research Institute 1997. Fundamentals of Employee Benefit Programs. Fifth Edition. Employee Benefit Research Institute: 149 – 162.

ERISA Advisory Council, Working Group Report on Small Business: How to enhance and encourage the establishment of pension plans, November 13, 1998

Fanning, Mike, Administrator for the Central Pension Fund of the International Union of Operating Engineers. Personal interview with Ghilarducci, Feb. 20, 2001.

Friedberg, Rachel, Kevin Lang, and William T. Dickens.  1988.  “The Changing Structure of the Female Labor Market: 1976-1984,” Industrial Relations Research Association 41st Annual Proceedings, 41st Annual proceedings.

Fronstein, Paul. 2001 “An Economic Model of Employee Benefits and Labor Supply” In Employee Benefits and Labor Markets in Canada and the United States. Eds. William Albert and Stephen Woodbury.  Kalamazoo, Michigan: W.E. Upjohn for Employment Research: 

Fuhrmans, Vanessa. 2003.  “Even at Giant Companies,Many Lack Health Benefits: Cost of Employee Contributions, Stricter Eligibility Criteria Keep,  Low-Wage Workers Out of Plans Wall Street Journal Oct. 22, 2003 p. 1

General Accounting Office. 2002.  Private Pensions: Improving Worker Coverage and Benefits. GAO-02-225 Washington D.C.  Government Printing Office.  April 9.

Ghilarducci, Teresa. 2002. ““De-linking Benefits from a Single Employer: Alternative Multiemployer Models.” Benefits for the Workplace of the Future. Olivia Mitchell, Editor. Wharton School Pension Research Council, Philadelphia: University of Pennsylvania. 2003.

Ghilarducci, Teresa, and Kevin Terry. 1999. “Scale Economies in Union Pension Plan Administration.” Industrial Relations. 38:1 (January):  11- 17.

Ghilarducci, Teresa, and Michael Reich. 2001. “Multiemployer Plans as Solutions to Pension and Training Collective Action Problems.” Journal of Labor Research. XXll:3 (Summer) 615-634. 

Goldman, T.R.1996.  “God Squad; Church Alliance; Acting on Behalf of the Church Benefits Programs.” Legal Times. American Lawyer Newspapers Group Inc. February 19.

Goodfellow, Gordon, and Sylvester Schieber. 1993.  “Death and Taxes: Can We Fund for Retirement Between Them?” In The Future of Pensions in the United States.  Ed. Ray Schmitt. Pension Research Council. Philadelphia, PA: University of Pennsylvania Press. 126-187.

Gordon, Michael S. 2000. Testimony on “ERISA at 25; has the law Kept Pace with the Evolving Pension and Investment World?” United States Congress, 106 Congress, Subcommittee on Employer-Employee Relations, Committee on Education and the Workforce, U.S. House of Representatives, February 15.

Greenough, William C. 1959. Retirement and Insurance Plans in American Colleges, New York: Columbia University Press.

Gustman, Alan L. 1990. “Pension Portability and Labor Mobility: Evidence for the Survey on Income and Program Participation” SIPP Working Paper Series, No. 9015. U.S. Department of Commerce, Bureau of the Census.

Hammermesh, Daniel. 1999. “Changing Inequality in the Markets for Workplace Amenities” NBER Working Paper #6515. Cambridge, Mass. NBER.

Hustead, Edward. 1996. Pension Plan Expense Study. Completed for the Pension Benefit Guaranty Corporation. Hay Huggins Consulting Group. Washington DC.

Jacoby, Sanford M. 1997. Modern Manors: Welfare Capitalism since the New Deal Princeton NJ: Princeton University Press.  

Lazear, Edward. 1995. Personnel Economics. Cambridge, Mass. MIT Press.

Mitchell, Olivia S., and Emily Andrews. 1981. “Scale Economies in Private Multi-Employer Pension Systems.”  Industrial and Labor Relations Review 34: 522-530.

Mitchell, Olivia S., and Sylvester Schieber, eds. 1998. Living with Defined Contribution Plans  Pension Research Council. Philadelphia: University of Pennsylvania.

Mitchell, Olivia S., David McCarthy, Stanley C. Wisniewski, and Paul Zorn. 2000. “Developments in State and Local Pension Plans.” In Pensions for the Public Sector. Eds. Olivia S. Mitchell and Edwin Hustead. Pension Research Council.  Philadelphia, PA: University of Pennsylvania Press. 11-40.

Munnell, Alicia, and Daniel Halperin. 1999. “How Should the Pension System y Reformed” Brookings Institution Conference on ERISA after 25 years: A Framework for Evaluating Pension Reform. September 17.

Nationwide Financial Services Inc. (Columbus Ohio) 2001 survey of 700+ owners and decision makers with 25-200 employees.  Conducted by Greenwald and Associates, Washington D.C.

Neumark, David, Ed. 2000. On The Job: Is Long-Term Employment a Thing of the Past? New York: Russell Sage Foundation Books.

Osterman, Paul. 1999. Securing Prosperity. Princeton, NJ: Princeton University Press.

Purcell, Patrick, and Paul J. Graney. 2002. “Retirement Plan Sponsorship by Small Employers.”  Benefits Quarterly (First Quarter), p.25-

Reagan, Patricia B., and John A. Turner. 2000.“Did the Decline in Marginal Tax Rates in the 1980s Reduce Pension Coverage?” In Employee Benefits and Labor Markets in Canada and the United States. Eds. William Albert and Stephen Woodbury.  Kalamazoo, Michigan: W.E. Upjohn for Employment Research:  475 – 496.

Rosen, Sherwin. 2000. “Does the Composition of Pay Matter?”  In Employee Benefits and Labor Markets in Canada and the United States. Eds. William Albert and Stephen Woodbury.  Kalamazoo, Michigan: W.E. Upjohn for Employment Research: 13-30

Sass, Steven A. 1997. The Promise of Private Pensions: The First Hundred Years. Cambridge, Mass.: Harvard University Press: 139-159.

Saunders, Mike. 2001. Personal interview with Ghilarducci, Feb. 27.

Shiller, Robert J. 2000. Irrational Exuberance. Princeton University Press.

Slottje, Daniel, Stephen A. Woodbury, and Rod W. Anderson. 2000. “Employee Benefits and the Distribution of Income and Wealth.” In Employee Benefits and Labor Markets in Canada and the United States. Eds William Albert and Stephen Woodbury.  Kalamazoo, Michigan: W.E. Upjohn for Employment Research:  349-378.

Transamerica Small Business Retirement Survey, November 2002. Survey of 300 small businesses and 765 workers “in a variety of industries” by the Transamerica Center for Retirement Studies (www.ta-retirement.com)

Turner, John A., and Noriyasu Watanabe. 1995. Private Pension Policies in Industrialized Countries: A Comparative Analysis Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.

USDOL, United States Department of Labor: 2000. Private Pension Plan Bulletin. No.9 Winter. 1999-2000 Final 1996. Pension Benefits and Welfare Administration, Washington DC: Government Printing Office: P. 5 (Table A2. )

Woodbury, Stephen, and Douglas R. Bettinger 1997. “The Decline of Fringe Benefits in the 1980s in Structural Changes in the Labor Market Structure in the 1980s: Causes and Consequences In Randall Eberts and Erica Groshen editors. Armonk, NY ME Sharpe. Pp. 101-134.


ENDNOTES

 



[1]  Unfortunately for the employer, efforts to do the later through scientific management, automation, control of the skill work, aggravate the later because it causes worker alienation, discontent and more turnover.

[2] Among their sample of 765 workers they found that half would leave their jobs if they were offered a position with a better retirement package.  Only 29 percent of workers admitted that they had not paid much attention to retirement planning.  In contrast, 72 percent of the 300 small business owners and employers believed that their workers preferred not to think about retirement planning until they were close to retirement age.

[3] In a DB plan, the benefit formula is based on service times a percentage of final salary, much like the final pay-based single employer defined benefit plans.  The replacement rate for a 25-year employee depends very much on the industry that the worker was in. A short freight driver in the trucking sector could receive $1,700 per month from his pension, just under his social security benefit and hence equal to almost one-half of his retirement income.

[4] The Episcopal Church plan originated in 1917 with an initial endowment from a wealthy Episcopalian, J.P. Morgan.

[5] This has proven useful in industries experiencing substantial volatility such as in construction.  For instance, the University of Notre Dame is adding extension to its football field, something the University does only once or twice a century. As a result, the handful of bricklayers and masonry workers in South Bend cannot meet the demand and workers had to be imported from other regions. During this time, work in South Bend still counts towards service credit in their home pension plan.



[i]  Appendix 1: Labor Market Segmentation Theory

 

Labor market segmentation theory helps explain coverage patterns.  In this section we explain the development of labor market theories and how it helps us understand the distribution of employee benefits.

In the 1970s and 80s, labor economics renewed interest in Cairns’ (1874) observation of noncompeting groups in labor markets. (One focus was on African American and women’s poor labor market outcomes: e.g., low wages, job insecurity, and receipt of lower returns to skill acquisition and education due to discrimination.) The import of these explanations is that the markets diverge overtime, the barriers deepen as institutions, like employee benefits, training, loyalty, reinforces itself.  

But modern human capital theory de-emphasizes the existence or inherent persistence of barriers. Cain (1987: 226) argues, in fact, that modern neoclassical economics successfully explains the five sources of wage inequality identified as far back as Adam Smith. (These sources are: compensating wage differentials; human capital investment differences; transitory differences due to lags and demand changes; mobility barriers, such as geography, (Heckman and Hotz, 1986), unions, regulation, class and group deprivations (due in part to cultural skin color and sex biases in the culture), and real differences in productivity that are difficult to measure.  The important, but hard to measure, effects are firm-specific effects (Abowd et. al. 2001) and individual productivity characteristics that may make total compensation seem unequal between equally qualified individuals.) The argument is if these differentials in human capital were really measured then we would see compensation depending only on human capital. The distribution of wages and benefits would be a matter of individual taste. Human capital theorists presume that, over time, worker’s choices about investments and preferences will determine returns to work and changes in jobs, the wage structure, and the composition of pay. Alternatively, the “segmentationist” or “non-competing groups” hypothesis is that “the existence of sectors with distinct wage-setting mechanisms (Dickens and Lang 1985: 796), are rooted in the dynamics of market economies.”

Friedberg, Lang, and Dickens (1988) find little evidence that, in the 1980s, women are significantly present in any primary market.  But clearly, twenty years later, many women are obtaining employee benefits and are doing so a faster rate than men. We want to determine if markets are also becoming segmented for women as well.



 

 

 

 

 

 

 

 

 

 

[ii] There is a fairly high, 28%, correlation between low wages and presence of multiemployer plans

 

Table 1: Industries where DB pension-covered workers

have their pensions through multi-employer pension plans

 

Multiemployer pension participants as % of total DB plan participants

Average hourly wages for nonsupervisory and production workers ($)

  Drinking places

100

6.62

  Women's apparel

94

8.41

  Laundry garment cleaning

91

8.76

  Hotels

78

9.22

  Leather products/goods, commercial/other printing

58

9.69

  Nurses (RN and practical)

94

10.18

  General auto repair, bowling alleys,

100

11.48

  Producers, orchestras, entertainers

99

11.48

  Trucking

81

13.95

  Motion pictures

78

15.69

  Heavy Construction

73

16.74

Construction

94

17.13

  Electrical work, masonry, stone, carpeting, flooring, etc

98

17.43

  Coal Mining

10

19.28

  Pipeline

56

21.79

  Highway/street, general building contractors

94

N/a

  Hunting and trapping

58

na

  Fruit and nuts

55

na

  Water Transportation

80

na

  Public warehouse, terminals

65

na

  Apparel, piece goods, notions

91

na

  Meat and meat products

80

na

  Alcoholic beverages

55

na

  Meat and fish markets, clothing

97

na

  Liquor, new car dealers, grocery, hobby

69-78

na

  Shoe repair and hat cleaning

100

na

 

[iii] Munnell and Halperin (1999) and Kijakuzi (2002) argue for new federal spending on retirement supplements for Americans at the bottom of the wage distribution because these workers will not be able to retire despite savings education, tax-favored retirement programs, social security, and employer-provided pension and health plans.  (

 

[iv] The 1998 ERISA Advisory Council report found that initial pension plan startup costs ranged from $500 to $3000 and were often fixed according to business size.  They estimated administrative costs for large firms to be around 7 percent of total plan costs, while this figure rose to 14 percent for smaller employers.  A 1996 study estimated that a small employer with 15 workers would pay an average of $619 in administrative costs per employee to operate their own defined benefit plan.  (This dropped to $287 for a defined contribution plan). The equivalent administrative cost for defined benefit plan for a 75-person firm was around $345. EBRI’s 2001 Small Employer Retirement Survey (SERS) reported that fear of liability for employees’ investment decisions was a factor in the decision not to offer a plan for 33 percent of respondents and a major factor for 12 percent. 41 percent of workers at firms with less than 10 employees were employed on a part-time or part-year basis.  The equivalent figure was 33 percent for firms with 10-24 employees (Graney and Purcell 2002).

 

[v] Many of the regulations passed under the Taft-Hartley amendments were intended to limit union control over the pension assets. These included prohibitions restricting employers from paying pension contributions directly to union leaders, rules holding pension trustees liable for pension decisions, and requirements that rules be applied equitably regarding both employers and workers. Though these regulations were often decried by the labor movement, they provided legitimacy and ultimately substantial strength to multiemployer plans.

 

[vi] In the 1930s and 1940s the UMWA covered a majority (80 percent) of mineworkers. The union was so wealthy that it helped organize emerging unions in the Congress of Industrial Organization (CIO) which covered workers in the rubber, steel, and auto industries. In the mid 1940s, the UMWA directed its bargaining power toward pensions.

 

[vii] During the historical development of union-initiated and jointly-agreed upon multiemployer pension plans in the U.S., several issues became salient. In the 1880s, labor unions began as “mutual aid” societies, with their major function the collection of funds from members in order to provide collective goods such as funeral benefits. The concept of linking contributions directly to payroll was an extension of this concept of self-help.

 

 

[viii] The Episcopal Church plan idea came from a patron who noticed “that pastors and priests were working way past their age of effectiveness and they wanted to help people retire at a reasonable age. The age was set at 68.” Human resource management needs therefore shaped the development of this clergy plan. 

 

[ix] The assets of large public plans dominate the top 200 list, followed by corporate plans, union multiemployer plans, and church plans.  Over $4 billion in assets is controlled by the largest 200 pension funds, and most, $2.4 billion, are held by public employee plans. A comparatively smaller portion, $ 0.1 billion, is in private or church multiemployer plans and the rest, approximately $1.6 billion, is in single employer plans. Thus from the perspective of total private pension assets, the bulk is no longer held by single employer plans in the U.S. 

 

[x]

[xi] The Sheet Metal Workers’ pension plan is typical in many ways of the multiemployer model.  Here the retirement benefit depends on a formula that includes years of service, hours worked per year (hours over 1,400 per year are credited at lower rates), an adjustment rate determined by an actuary, and an hourly contribution that varies by local. Since contribution rates vary by local (as do wages), plan members with the same career profile but covered under different contracts will get different eventual retirement benefits. The Central Pension Fund of the Operating Engineers (CPF) and the Western Conference of Teamsters plans tie employers’ contributions directly to actual benefits paid. The benefit formula is then based on a rate determined by fund trustees and the balance of a final “account,” making this approach analogous to a cash balance plan.  For example, the monthly annuity is worth 3.5 percent of a retiree’s account balance; if a member works 30 years and has $50,000 of contributions credited from various employers, the retirement benefit would be $1,750 per month (in the 2001 CPF plan). Like cash balance plans, these benefits accrue on a career basis rather than a final average pay basis, and different rates prevail in different contracts, reflecting employers’ differential abilities to pay and levels of bargaining power.  In the CPF, contribution rates vary from $.050 to $5.60 per hour; in the Western conference of Teamsters the range is similar, from $0.60 to over $6.00 per hour. The range is similar in the Sheet Metal Worker’s plan.

 

[xii] These “suspension of pension” rules which requires funds to continue paying pension benefits if a retiree returns to work in his same industry but is employed fewer than 40 hours a month or 480 hours in a year. If he should work more than this, the fund can legally restrict his pension benefits until he leaves that employ.

In some sectors the particular threshold chosen to determine whether the pension is suspended, 40 or 480 hours, matters a great deal. For example, in construction, employers often need workers more than 40 hours per month but require them only a few months per year. As a result in that industry, a 40-hour per month rule proves more restrictive than is the 480-hour rule. In practice, since employers and unions manage multiemployer plans, the rules are often varied according to labor market conditions. For example, the Central Pension Fund of the Operating Engineers switched from the liberal to the restrictive rule in the 1970s to retard the growth in nonunion construction (Fanning 2001).  Likewise during the 1990s the severe labor shortages in most of the major crafts put pressure on the many of the funds to switch back to the more liberal 480-hour rule restriction. The Sheet Metal Workers fund and CPF have resisted the more liberal rules arguing that collectively bargained pensions might subsidize nonunion employers and erode pension contributions, if “return to work rules” are too permissive. On the other hand many unions have hailed the Congressional rule change lifting the earnings test for older workers under social security. Indeed the Western Conference of Teamsters plan only restricts employment eligibility for members under the age of 65. After age 65, pensions are no longer suspended for delayed retirement (Saunders 2001).

 

[xiii] Further evidence of changed “employee benefit paradigm” can be seen in the pension arena where worker participation in DC plans has risen dramatically and fallen in DB plans (Mitchell and Schieber 1998; Turner 1993, Turner and Watanabe 1995, Wolff 2003). The shift toward DC plans and the popularity of 401(k)s also represents a move away from group-based solutions to insurance to individual oriented solutions. This is reflected in the fact that the overall fraction of pay going to benefits has declined over time: the share of total compensation going to benefits survey declined from 42 percent to 37 percent between 1989 and 1999 (U.S. Chamber of Commerce various years).  Conversely, cash compensation dominates in casual, short-term, spot labor markets, and this benefit decline is not merely an artifact of the fastest growing industries and occupations.  At the same time, the business community has been re-thinking its role and its social obligation in sponsoring employee benefits.

 

[xiv] The difference in benefits is substantial if one compares them across union and nonunion sectors. For example union workers had 37 percent of compensation devoted to benefits versus 29 percent for nonunion manufacturing employees in 1999. This gap persists in non-manufacturing where benefits make up 33 percent of union workers’ compensation and 25 percent of workers’ remuneration (ECI 1999). The positive union effect on benefits may result from the workings of group processes enabling workers to overcome myopia and overoptimism regarding risks due to poor health, disability, and retirement.  Economies of scale may also explain the relative growth in benefits in multiemployer settings.  In addition, unions provide job protection and “voice,” helping form training and deferred compensation agreements.

 

[xv] Job tenure among men has fallen over time in the U.S. For example, average seniority on the job for men age 45-54 fell from almost 13 years to over 9 years (**time period?).   Among men age 55-64, average seniority was over 15 years in 1983, but had dropped to 11 by 1998. Among women, by contrast, tenure has increased by 5 percent over this same period, with the largest increases seen among older women workers (Bureau of Labor Statistics 1998).