VI. UNION AND MANAGEMENT COOPERATION AND APPROACHES TO MULTI-EMPLOYER PLANS

Multi-Employer Pension Plans and the Pension Coverage Problem

TERESA GHILARDUCCI
University of Notre Dame

Abstract

      Unlike other O.E.C.D. nations, the United States depends heavily on the employment relationship to provide social insurance. Yet, academic employee benefit research almost exclusively focuses on the contract between individual employers and workers. Virtually no researchers study group-based employee benefit plans, although worker, union, firm, and public needs are met by multi-employer pension (and, by extension, health and apprenticeship) plans because they solve collective action problems. I argue that because firms, by themselves, will not pay for training and benefits unless their competitors are forced to, multi-employer plans serve the presumed public interest in raising the share of the labor bill devoted to employee benefits and social insurance.

      Although private and public sector multi-employer plans cover different types of workers, from janitors to university presidents, they similarly solve four key problems. First, multi-employer plans cover workers who would otherwise not have benefits. Second, multi-employer plans adapt to the skill and insurance needs of heterogeneous workplaces (unlike the uniform social security system). Third, multi-employer plans solve the coordination problem that no one employer has much incentive to provide benefits or training without competitors also being forced to pay. Fourth, multi-employer plans get scale economies and, thus, lower professional fees.¹

Too Much Cash and Not Enough Social Insurance

      Pension coverage has stalled at about 50 percent for all workers, the rate is higher for men than women, and less than a third of nonwhite workers are covered (Employee Benefits Research Institute [EBRI] 2001) and employers are paying less for benefits. The average industry share of total labor compensation going to noncash pay (i.e., benefits) decreased from '.3 to 26.5 percent between 1991 and 2000 (Employment Cost Index various years). Unfortunately, the benefit share dropped in the largest industries: in service from 24.6 to 24.3 percent and from 22.8 to 20.6 percent in retail between 1988 and 2000. However, the few workers covered by pensions in these industries obtain them through collectively bargained multi-employer plans. Regression analysis (available from the author) indicates that more unionization explains higher levels of benefits shares, as do increases in health insurance costs and the level of total compensation. The latter means the higher-paid jobs also have higher employee benefit shares.²

How Economists Debate Why Employee Benefits Shrunk

      What we think causes the decline in benefits affects our policy choices. Neoclassical pension-determination theories rely on “compensating wage differential” theories that argue workers choose combinations of wages and earnings, desirable job characteristics, and benefits. This implies that older men (especially) who are experiencing declining job tenure and are more mobile, would be expected to choose less tenure-related benefits (Goodfellow and Schieber 1993; Ippolito 1998). However, the finding that pay increases are associated with increasing levels of good working conditions weakens the “trade-off” theory. (Hammermesh [1999] found that jobs with declining injury rates, in the late 1970s to early 1990s, also had the highest earnings growth.) Therefore, the total compensation--wages, benefits, and nonmonetary desirable job attributes--gap is wider than conventionally measured.

      Alternatively, institutional economists argue that unions, employers, government, as well as workers, influence workers’ cash and benefit preferences. For example, the 1980s defined benefit (DB) pension plan termination and reversion trend may have motivated workers to opt for second best–defined contribution (DC) plans. Osterman (1999) and others argue the “social contract” between workers and firms collapsed in the 1980s. Evidence includes job instability sharply increasing for those whom pension accruals are most crucial--men age 45–64 (particularly African Americans) and, those at any age with more than 9 years of service (Neumark et al. 2000). The eroding social contract and expansion of secondary labor markets explains the eroding employee benefit coverage. (Workers in primary sectors--for example, 90 percent of public-sector workers--have high coverage rates; but, coverage rates in the private sector go as low as 30 percent and 35 percent in personal and business services and are close to zero in nonunion construction and trade; EBRI 2001).

The Role of Unions and Multi-Employer Plans in Benefit Coverage

      Buried in a critique of the stock market, Yale economist Robert Shiller (2000:23) argues that union decline causes a decline in group-based, DB-type pensions because without solidarity, or a desire to share risks, demand for social insurance is replaced by demand for immediate and individualized forms of pay. Indeed, the union benefit share is 30 percent higher than the nonunion share. ³ Unions initiate multi-employer plans, which provide benefits in casual labor markets where benefits are scarce. Ninety four percent of heavy construction workers, retirees, and dependents covered by DB plans are in multi-employer plans, as well as a full 55 percent of DC participants. Likewise, a whopping 73 percent of retail food store employees, 59 percent of apparel employees, and 39 percent of furniture industry participants have pensions only through multi-employer plans (details available from author; source: IRS Form 5500 various years).

      Multi-employer plans, though ignored in the literature, are key pension delivery systems. Twenty percent of active private DB plan participants had multi-employer plans in 1996 (EBRI 1997), having grown from 10 percent in 1950 to 18 percent in 1960. In all, 11 of the 92 million participants in all employer plan types had multi-employer coverage (US Department of Labor, 2000). Nonbargained multi-employer plans are also in the public and not-for-profit sectors, including the state and local sector, in churches, the Red Cross, charities, and, of course, university and college teachers.

The Scope and Special Features of Multi-Employer Plans

      Multi-employer plans may cover many occupations in one industry, or one craft in many industries, or many occupations in many industries, or are industry, occupation, and region-based. An example of the latter includes the United Food and Commercial Workers fund in Northern California, which covers many jobs in grocery stores, including Safeway. The older ladies’ garment union and the clothing and textile workers’ funds cover production workers across a range of needle trades employers. The building trades cover particular trades operating across diverse industries and regions. And, some funds like the Western Conference of Teamsters pension plan covers many occupations--grocery delivery drivers, warehouse workers, and long haul freight truckers--in several industries in 13 western states (Saunders 2000).

      The unions and firms both want to expand the product’s market share and improve training. They also view the nonunion contractor and economic downturns as common enemies. Therefore, multi-employer plans are embedded in long-term complex employment relationships. Key to their success is that multi-employer plans adapt well to employer needs and, as DC and DB hybrids, they combine the best features of each plan type. Participating firms contributions are collectively bargained, so they vary with relative bargaining power. For example, the Sheet Metal Workers’ plan bases benefits on service and hourly contributions that vary by local (as do wages) so that plan members with the same career profile, but covered by different contracts, will get different (but defined) retirement benefits. Thus, each employers’ financial circumstances are incorporated, and workers get a DB pension.

Labor Market Stabilizers

      Multi-employer plans help stabilize labor market cycles with “breaks in service” and “suspension of benefit” rules, and reciprocity agreements. “Breaks in service” (or “loss in service”) rules specify how long a participant may not work before losing the right to return to the same fund and resume accruing benefits. Multi-employer plans exhibit substantial business cycle sensitivity by altering these rules. For example, during the 1970s recession, the Sheet Metal Workers Fund liberalized loss-of-credit rules to help unemployed members keep coverage. Consequently, members stayed connected to their craft or “skill-set” and were available to union contractors in the upturn because they knew they could continue accruing benefits. The liberalization was costly, but it exhibits sensitivity to industry and workers’ needs.

      Another example of how multi-employer plans accommodate labor market conditions is in their response to the rapid 1990s expansion, when many multi-employer pension plans liberalized “suspension of benefit” rules, which prohibit retirees from returning to work in their career industry after collecting a pension. Though the prohibition is designed to avoid subsidizing low wages of nonunion competitors, the severe labor shortages of the time pressured funds to switch to liberal standards--requiring pension stoppage when annual hours exceeded 480 hours and not 40 weekly hours. (Examples are available from the author.) One fund revealed, ingeniously, they use retired union members as “salts” at nonunion sites during organizing drives, which helps the union and, though not emphasized, the unionized employer. Last, in one of many ways, these plans respond to the larger context, “reciprocity” features promote dependable labor supplies in decentralized and unstable industries by allowing participants, who are loyal to their skills, accrue pension benefits while working for different signatory employers.

Economies of Scale

      Multi-employer plans are larger than single employer plans and, thus, can obtain significant scale economies (Hustead 1996). There are more than double the fraction of multi-employer plans compared to single DB plans in huge plans, those with more than 50,000 members (42 percent and 20 percent; US DOL 2000). Despite the criticism that multi-employer plans should merge to save costs, the fact that some are small and decentralized suggests they succeed by adapting to local conditions. Their parochial nature may be their reason for existing.

Cross Subsidies

      All defined benefit plans entail cross-subsidies. The obvious transfer is from retirees who die earlier and to those living longer than average. A less obvious concern is that well-off employers subsidize marginal employers. The United Parcel Service proposed in 1997 that its employees leave the Teamsters multi-employer pension plan to form their own single employer plan because, UPS argued, it was subsidizing smaller employers. The union-recognized UPS membership helped achieve scale economies but contended that only a detailed actuarial study would reveal whether a single employer plan could provide the same or more benefits for less.

      The Central Pension Fund (CPF) of the Operating Engineers also faced internal dissension when one local experiencing higher level of growth argued they were “carrying” the poorer and shrinking locals. The fund responded that only over 80 actuarial studies, which the administrator implied was impractical, would determine whether each local’s past and projected experience would yield better benefits alone than with the CPF (Fanning 2000).

      The Episcopal Church Fund formula self-consciously has the rich subsidize the poor in two ways: by attributing a 1.75 percent factor for the first $10,000 of salary and a 1.5 percent for levels above that; thus, lower-paid workers have a higher replacement rate. Second, since 1980, it has historically provided inflation ad hoc adjustments with a flat amount in a “13th check,” which varies according to service but not pay (Blanchard 2001). (For instance, ministers earning say, $10,000 and $50,000 annually will get the same, say $1,000, extra payment in December.)

Joint Governance of Trusts Reduce Conflict of Interest and Principal Agent Problems

      The Taft–Hartley amendments to the National Labor Relations Act of 1947 and the 1974 Employee Retirement Income Security Act (ERISA) require that union trustee representatives cannot outnumber employer representatives and that they must adhere to the “loyalty” principle--that trustees act for the sole benefit of the plan participants. The legal structure prevents labor and management from using the plan to further their own goals: For example, unions may be tempted, but cannot allow, strapped employers to delay contributions, and employers can’t adjust their contributions until the contracts end. This means workers and employers share investment gains and losses. Between 1984 and 1996, single employer DB plan contributions per participant fell 29 percent, while multi-employer plan contributions fell by 37 percent. Despite this, benefits in multi-employer plans grew 26 percent versus only 6 percent in corporate DB plans. During the same period, multi-employer DC plan contributions rose 8 percent, while (contrary to popular belief) corporate employers cut back on DC contributions by 20 percent (Ghilarducci 2000). In practice, multi-employer plans tend to increase benefits when fund levels reach nontax–favored limits; in contrast, corporate employers tended to reduce contributions.

Multi-Employer Plans Advantages for Workers and Employers

      Multi-employer plans’ contribution, governance, transparency, and fiduciary framework minimize many risks faced by workers and firms. Workers are tempted to spend their retirement accounts and lose their pension accrual if they leave their employer. This “employment risk” is mitigated because coverage extends to all contributing employers and the DB structure ensures the funds are used for retirement. Furthermore, Pension Benefit Guaranty Corporation coverage and strict ERISA regulation minimizes “investment risk.” In addition, group plans minimize “consumer risk” with economies of scale and monitoring by eliminating the high professional fees charged to self-directed individual plans. DB plans minimize “longevity risk” that retirees outlive their accounts and “inflation risk” because multi-employer plans raise benefits more often than corporate DB plans. Finally, how multi-employer DB’s shrink “heuristic (choice) risk” is more subtle. The behavioral financial literature suggests self-directed participants make wrong choices, trade and borrow too much, engage in market timing, and experience high costs of trading (Bureau of National Affairs 2001), only because such tendencies are endemic to human behavior. Unlike DC plans, DB plan members do not face “heuristic risk.”

      Likewise, employers, particularly smaller ones, and those that chronically face skilled labor shortages obtain substantial advantages from multi-employer plans. Small business owners can provide good pensions for themselves and staff. In addition, occupational pensions reduce occupational mobility so that employers and workers have more incentives to invest in employee training because they can reasonably expect to recoup some of the investments costs. Such “win-win” trades help make the economy more productive (Ghilarducci and Reich 2001).

The Future of Multi-Employer Plans

      US pension coverage rates are stuck at 50 percent; therefore, we need to attend to pension delivery systems to expand coverage. Tax incentives for participating in new forms of statutorily approved multi-employer plans, other than collectively bargained plans, and allowing employers not party to collectively bargained agreements to join a collectively bargained plan, may work (Gordon 2000). Most experts (see Tim Lynch, this volume) believe employers balk at joining trusteed plans because of a withdrawal liability in an underfunded plan. Because there are ways to avoid this blatant cross subsidy, employers’ unwillingness to join a collectively bargained contract and trust fund may reflect preferences to remain nonunion and in sole control of their pension fund.

      It has been suggested that individual-oriented plans are chief competitors and substitutes for multi-employer plans because they too avoid dependence on a single employer and, in addition, that increasingly available plans garnered from Internet searches enable individuals to obtain scale economies without joining a group. Yet, I argue, the portability of DC plans adds to their popularity, but exposure to volatility risk makes them undesirable. From workers’ point of view, DC and DB hybrids have the best features of both types. Perhaps, a more important factor sustaining multi-employer forms is that employers and workers want and need more than scale economies; they need to take labor costs and training out of competition and avoid “a race to the bottom,” though the most profitable short-term strategy for each firm is to provide no benefits and training and compete on the basis of low prices. In the long term, quality erodes, demand falls, and labor shortages create chronic problems.

      Multi-employer plans may expand to uncovered groups in conventional ways. Amy Dean, President of the South Bay Central Labor Council, AFLCIO, and head of the organization Working Partnerships, USA, envisions employer–training networks to be connected to the health and pension consortiums for low-wage workers, and eventually to collectively bargained contracts in the Silicon Valley (Brenner et al. 1999:67).

Conclusion

      The rise of cash and the decline of employee benefit as shares of total payroll is caused by a sea change in the US employer–employee relationship; but, the multi-employer plan may serve as a framework for the fast growing supply of casual labor market jobs and “occupation-identified” workers. The continuing importance of negotiated multi-employer plans shows the power of coordination. These plans do what human resource experts and industrial innovation experts say must be done: they adapt to the idiosyncrasies of particular industries and occupations, induce training by increasing workers’ attachment to a industry or occupation, and provide desperately needed supplemental social insurance on the job.

Endnotes

1. Many of these arguments are covered in a longer paper (Ghilarducci 2001).

2. The unionization level in 2000 and the benefit share growth rate between 1989 and 2001 in various industries ranked by the size of the industry’s employment (in millions) is as follows (source ECI):

3. The benefit difference between union and nonunion sectors is significant: 37 percent of compensation devoted to benefits versus 29 percent in 1999, respectively. This gap persists in nonmanufacturing, where benefits make up 33 percent of union workers’ compensation and 25 percent of nonunion workers’ remuneration (ECI 1999). The positive union effect on benefits may result from the workings of group processes enabling workers to overcome myopia and over-optimism regarding risks due to poor health, disability, and retirement. Economies of scale may also explain the relative growth in benefits in multi-employer settings. In addition, unions provide job protection and “voice,” helping from training and deferred compensation agreements.

References

Blanchard, Alan. 2001. Administrator for the Episcopal Pension Fund. Telephone interview with Ghilarducci, Feburary 23.

Bureau of National Affairs. 2001. “Offering Internet Access to 401(k) Plans Can Bring Out the Day Trader in Employees.” Pension and Benefits, Vol. 1, no. 38 (February), p. 6.

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

Employee Benefits Research Institute (EBRI). 2001. “Pension Participation: February 2001.” Notes Executive Summary. Washington, DC: Employee Benefits Research Institute.

Employee Benefits Research Institute (EBRI). 1997. Fundamentals of Employee Benefit Programs, 5th ed. Washington, DC: Employee Benefit Research Institute, pp. 149– 62.

Employment Cost Index, Bureau of Labor Statistics, US Department of Labor, <http://stats/bls/specialrequest/owcw/ocit/ect/ecb10013.pdf>.

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

Ghilarducci, Teresa. 2001 “Delinking Employee Benefits From A Single Employer: Alternative Multi-employer Models.” Working paper for the Pension Research Council, Wharton School of Business, University of Pennsylvania, December.

Ghilarducci, Teresa, and Michael Reich. 2001. “Multi-employer Plans As Solutions to Pension and Training Collective Action Problems.” Journal of Labor Research, Vol. 22, no. 3 (Summer), pp. 615–34.

Ghilarducci, Teresa. 2000. “Small Benefits Large Funds: How Governance Affects Pension Plans.” In Tessa Hebb et al., eds., Working Capital: The Power of Labor’s Pensions. Ithaca, NY: Cornell University Press, pp. 44–66.

Goodfellow, Gordon, and Sylvester Schieber. 1993. “Death and Taxes: Can We Fund for Retirement Between Them?” In Ray Schmitt, ed., The Future of Pensions in the United States. Philadelphia, PA: Pension Research Council, University of Pennsylvania Press, pp. 126–87.

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, US House of Representatives, February 15.

Hamermesh, Daniel. 1990. “Changing Inequality in Markets for Workplace Amenities.” Quarterly Journal of Economics, Vol. 114, no. 4 (November), pp. 1085–123.

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

Ippolito, Richard A. 1998. Pension Plans and Employee Performance. Chicago, IL: University of Chicago Press.

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

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

Saunders, Mike. 2001. Telephone interview with Ghilarducci, February '.

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

United States Department of Labor: 2000. Private Pension Plan Bulletin. No.9 Winter. 1999–2000 Final 1996. Pension Benefits and Welfare Administration, Washington DC: GPO, p. 5 (Table A2.)