PETER CRAMTON, University of Maryland - Department of Economics
Email: peter@cramton.umd.edu
HAMID MEHRAN, Federal Reserve Bank of New York
Email: Hamid.Mehran@ny.frb.org
JOSEPH S. TRACY, Federal Reserve Bank of New York, National Bureau of Economic Research (NBER)
Email: joseph.tracy@ny.frb.org

By the early 1990s, employee stock ownership plans (ESOPs) had become as prevalent in unionized firms as in nonunionized firms. However, little research has been devoted to examining the implications of ESOPs for collective bargaining or, more generally, for cross ownership. In this paper, we extend the signaling model of Cramton and Tracy (1992) to allow partial ownership by the union. We demonstrate that ESOPs create incentives for unions to become weaker bargainers. As a result, the model predicts that ESOPs will lead to a reduction in strike incidence and in the fraction of labor disputes that involve a strike. We examine these predictions using U.S. bargaining data from 1970 to 1995. The data suggest that ESOPs do increase the efficiency of labor negotiations by shifting the composition of disputes away from costly strikes. Consistent with improved bargaining efficiency, we find that the announcement of a union ESOP leads to a 50 percent larger stock market reaction when compared with the announcement of a nonunion ESOP.

PAUL M. SECUNDA, Marquette University - Law School
Email: paul.secunda@marquette.edu

Although all legal regimes result in some inevitable inequities occurring to some parties, the inequities resulting from the application of ERISA are systematic as a result of how preemption and/or remedial provisions are applied in these employee benefit cases. Although these claim have a systematic quality, they are not inevitable. Rather they result from two interrelated phenomenon caused by the U.S. Supreme Court interpretation of the pertinent ERISA provisions: (1) a decision to broadly construe the preemption provisions of ERISA so that most state laws do not survive ERISA preemption; and (2) a separate, but connected, decision to narrowly construe the remedies available to participants and beneficiaries available under ERISA's civil enforcement scheme.

The dual operation of preemption and remedy provisions on employees' benefit claims is that many employees are finding that they have no meaningful remedy for their claims under the statute. It is the central argument of this paper that this state of affairs exists because the Court, in the voice of the strict constructionist Justices, has consistently favored employer interests under the statute rather than those of employees. They have accomplished this feat by elevating a secondary purpose of ERISA, to make sure employers voluntarily adopt employee benefit plans, over the primary purpose of ERISA, to ensure employees and their beneficiaries are protected in their pension and welfare benefits.

To right the ERISA ship and stop permitting employers to use this statute as a shield against employee benefit-related claims, I argue for the adoption of the remedialist approach championed by another group of Justices over the years. These Justices rightly believe that analysis under ERISA should start with the common law of trusts, upon which ERISA is primarily based. To the extent that there are deviations from that common law, only then should the unique characteristics of the ERISA statute be considered in applying the preemption and remedial provisions.

CRAIG COPELAND, Employee Benefit Research Institute (EBRI)
Email: COPELAND@EBRI.ORG

This paper examines the level of -- and trends in -- IRA assets. In addition, it includes the most recent Internal Revenue Service (IRS) data on the distribution of assets and contributions to IRAs by IRA type, thereby permitting analysis of the assets and contribution levels of traditional (deductible and nondeductible) IRAs, Roth IRAs (nondeductible contributions and tax-free withdrawals), and other IRAs (employment-based SEPs and SIMPLEs).

Total IRA assets are larger than those accumulated in either private-sector defined benefit (pension) plans or defined contribution (401(k)-type) plans. At year-end 2007, IRAs held $4.75 trillion, private-sector defined contribution (401(k)-type) plans held $3.49 trillion, and private-sector defined benefit plans held $2.33 trillion. IRA growth continues to be fueled by rollovers from other types of retirement plans, not new contributions. There has been a big shift in market share in IRA assets over the past quarter-century, with mutual funds and brokerage accounts now dominant. Mutual funds held 47 percent of IRA assets at year-end 2007, followed by brokerage accounts (38 percent). Life insurance companies (8 percent) and banks/thrifts (7 percent) split the remainder. About 90 percent of all IRA assets are held in traditional (taxable on withdrawal) IRAs, but most contributions are going into Roth (untaxed at withdrawal) and other types of IRAs.

DAVID P. BERNSTEIN, affiliation not provided to SSRN
Email: spstat@yahoo.com

Current multi-pillar pension plans have adopted the same retirement age for the defined benefit (DB) and defined contribution (DC) components of the plan. This paper considers potential benefits obtainable from coordinating DB and DC benefits. The plan proposed here involves establishing an earlier standard retirement age for the DC component of the multi-pillar pension plan than for the DB component of the multi-pillar pension. Individuals who receive low returns on the DC pension plan are eligible for earlier non-standard disbursements from their DB pension plan. The lower standard retirement age on the DC component of the multi-pillar pension plan than on the DB component of the multi-pillar pension plan reduces costs of funding the DB plan while maintaining a lower overall potential retirement age for individuals. The possibility of non-standard early DB disbursements for retirees who realize low investment income reduces the instability of retiree income due to market fluctuations. A lower standard retirement age on the DC component of the multi-pillar pension plan than on the DB component of the multi-pillar pension plan provides an incentive for workers to delay retirement until reaching the standard DB retirement age. This work incentive does not increase costs for the DB plan.

VILLE-PEKKA SORSA, School of Geography, University of Oxford, University of Helsinki - Department of Political Science
Email: ville-pekka.sorsa@spc.ox.ac.uk

Pension funds form an own domain within the general domain of finance as their institutional environments and arrangements of investment differ from other sub-domains. This working paper studies the institutional organization of the field of US occupational pension funds with new institutional theory including discussions on cultural economy approach to social studies of finance. US occupational pension fund investment actions are constituted and affected by various relevant institutions that are usually very diffused. The funds operate in heterogeneous regulative and normative environment in terms of form and sources, but the environment is somewhat homogeneous in substance. They face numerous constraints in their institutional environment, but most of these constraints simultaneously enable and give mandate and legitimization to various kinds of actions thus limiting the effectiveness of most constraints. The most restricting feature in the field level is the incentive to mimetic replication of conventional investment practice, which suggests that occupational pension fund investments are dependent on conventional models of investment whereas they are among the only actors that are capable of changing the very same models in the domain of finance. The main result of the paper in case of the organization field observed is that organization and governance methods can break conventional patterns and broaden the horizon of all pension fund actions enabled by their institutional environment. This may require, however, very strong institutional entrepreneurship in order to change investment behavior and especially corporate engagement practices in the occupational fund sector due to its cultural-cognitive restrictions.

TIM V. EATON, affiliation not provided to SSRN
JOHN R. NOFSINGER, Washington State University - Department of Finance
Email: john_nofsinger@wsu.edu

Using a comprehensive sample of 2002 and 2005 U.S. public retirement systems, we found that public pension plan underfunding grew dramatically in these years despite a good economy, increasing state tax revenues, and strong stock market returns on average, plans were only 83% funded. Teacher plans and plans with the most retirees were more underfunded. We found no significant differences related to asset allocations or actuarial assumptions about inflation and rate of return. A primary factor associated with significantly lower underfunding was more female active participants in the plan, suggesting another risk to women's retirement income.