Mark J. Warshawsky

IN SICKNESS AND IN HEALTH: AN ANNUITY APPROACH TO FINANCING LONG-TERM CARE AND RETIREMENT INCOME

This article examines the implications of the positive correlation of mortality and disability for the benefits of combining an immediate income annuity with long-term care disability coverage at retirement ages. It also investigates the value of such a combined benefit to various subgroups of prospective purchasers and the implications of possible errors and moral hazard in the reporting of disability status and making claims. (The Journal of Risk and Insurance, 2001, Vol. 68, No. 2, 225-254)

Specification of the Joy of Giving: Insights from Altruism

This paper analyzes the joy of giving bequest motive in which the utility obtained from leaving a bequest depends only on the size of the bequest. It exploits the fact that this formulation can be interpreted as a reduced form of an altruistic bequest motive to derive a relation between the value of the altruism parameter and the value of the joy of giving parameter. Using previous discussions of an a priori range of plausible values for the altruism parameter we then derive plausible restrictions on the joy of giving parameter. We demonstrate that this parameter may well be orders of magnitude larger than assumed in the existing literature.

Private Annuity Markets in the United States: 1919-1984

Load factors on life annuities issued to 65-year old males and females over the period 1919 through 1984 have ranged from 10 cents to 29 cents per dollar of actuarial present value. From 8 cents to 16 cents of these loads represents the cost of adverse selection, and approximately 7.5 cents represents transaction costs. The cost of adverse selection increased during the middle period of study, 1941-1962, on annuities sold to males, while the cost of adverse selection continually declined for females. Annuities were at the height of their popularity in the early 1930s when policies were credited with interest rates higher than those available in the financial market; implicitly 7.5 cents of the load factor was returned to policyholders. The use of interest rates lower than market yields in the last period of study, 1963-1984, however, increased load factors by 6.6 cents. This study examines the pricing of individual life annuities over an extended historical period. It attempts to identify and examine the factors relevant to annuity pricing: the roles that adverse selection of mortality risks, the use of portfolio yields, and transaction costs play in the determination of load factors on individual annuity contracts. The claim by Abel (1986) that the cost of adverse selection increased with the introduction of Social Security and pension plans is also briefly examined. Empirical evidence on private annuity markets may help to broaden and inform current discussions in the economics literature concerning the life cycle hypothesis of savings and the impact of the introduction of Social Security on steady-state wealth and social welfare. Relying on simulation work, Davies (1981) attempts to reconcile the life cycle hypothesis with the observed slow dissaving of marketable assets by the retired. He claims that it is unnecessary to resort to a bequest motive for explanation because plausible parameters of Mark Warshawsky is an Economist with the Board of Governors of the Federal Reserve System. The author thanks Joyce Payne for help in assembling the data, Amy Bassan for pointing out the data source on annuity prices, and Andrew Abel, Michael G. Bradley, Benjamin Friedman, Frank Nothaft, Francis Schott, seminar participants at the University of Pennsylvania, and an Associate Editor and anonymous referee of this Journal for helpful comments. The opinions expressed in the paper are those of the author and not those of the Board of Governors of the Federal Reserve System or its staff. This content downloaded from 157.55.39.114 on Fri, 14 Jul 2017 17:35:18 UTC All use subject to http://about.jstor.org/terms Private Annuity Markets in the United States 519 the utility function and the absence of private annuity markets can adequately explain the slow dissaving of assets by the aged in the face of an uncertain lifetime. Abel (1985) employs this scenario to demonstrate that the introduction of a Social Security system alleviates the need for precautionary saving and therefore leads to a reduction in steady-state wealth. Eckstein, Eichenbaum, and Peled (1985) demonstrate how particular schemes for the introduction of a Social Security system lead unambiguously to a Pareto improvement in social welfare in the absence of private annuity markets. The results of these theoretical analyses, however, are suspect because of the strong and incorrect assumption that private annuity markets are nonexistent. The more relevant questions which researchers should instead pose are: (a) whether observed load factors on annuity contracts are sufficient to account for the lack of participation in private annuity markets and the slow dissaving of assets by the aged; and (b) whether a private annuity market with load factors would increase or decrease social welfare when compared to a social annuity program like the Social Security system. It is hoped that the empirical evidence on private annuity markets produced in this article will enable researchers to pose these questions.

The Impact of Liabilities for Retiree Health Benefits on Share Prices

Health insurance provided by corporations to retired employees represents an important benefit to retirees and a significant and growing expense to employers. If liabilities for retiree health benefits have already reduced share prices dollar-for-dollar or at least in a similar manner to other liabilities, there would be little cause for concern that the introduction of new, more revealing, accounting standards for these benefits would lead to financial pressure on companies to redi.ce or cancel benefits. This study identifies firms as sponsoring or not sponsoring retiree health plans, and liabilities for retiree health benefits are estimated using the framework of the new accounting pronouncement on nonpension postretirement benefits. Stock value is regressed on several independent variables, including estimated retiree health liability. The results clearly suggest that retiree health liabilities impact stock prices. Furthermore, some evidence indicates that the impact may be less than balance sheet liabilities. This latter finding is consistent with market expectations that the firms or the federal government will take actions to reduce future payouts of health benefits to retirees by corporate sponsors, although it also may be due in part to measurement error associated with the variable for retiree health liabilities.

Does Freezing a Defined-Benefit Pension Plan Increase Company Value? Empirical Evidence

This study empirically tests whether freezing or closing a defined-benefit (DB) pension plan increases the sponsoring company’s market value. The database used for this study consists of 82 publicly traded U.S. companies that announced freezes/closes in 2003–2007. On the basis of this extensive sample and through a set of parametric and nonparametric tests under the event study methodology, the study finds generally negative or insignificant abnormal returns in stock prices that can be associated with the freeze/close events. Little evidence supports the hypothesis that freezing or closing a DB plan increases company value. In recent years, pension coverage in the private sector has been shifting from defined-benefit (DB) plans to defined-contribution (DC) plans. Some employers have closed their DB pension plans to new hires (“close”) or frozen the plan benefit accrual for some or all existing plan participants (“freeze”). The commonly cited motivations for such changes include cost and volatility reduction, consistency with industry-competitive practice, and employee desires and satisfaction. Many have wondered whether and to what extent a DB freeze/close alters the market value of the corporate plan sponsor. Research on this topic, however, is scanty. This empirical analysis tests the hypothesis that freezing or closing DB pension plans increases the sponsoring companies’ market values. The premise for this hypothesis is that DB plan freezes/closes depress the growth of pension liabilities and thus allow more funds to be directed to profit-generating corporate businesses or to other forms of compensation that are less risky or less costly to the company. Our tests, however, found generally insignificant, often negative, abnormal returns in stock prices associated with the freeze/close events and, therefore, yield little evidence to support the hypothesis. Our study contributes to the literature in several ways. First, we constructed a large database comprising 82 publicly traded U.S. companies for the 2003–07 time period in various sectors. Identified simply by the availability of specific freeze/close announcement dates, these companies can be considered random draws and are thus fairly representative of the population of corporations that have frozen or closed their DB plans. Second, on the basis of this extensive sample and through a set of parametric and nonparametric tests under the event study methodology, our analysis provides general and robust evidence on the tested hypothesis. In the benchmark test, stocks of 46 companies exhibited negative market-adjusted returns in the wake of the announced DB plan changes. The median value of price change is −0.41 percent, which can be attributed to the announcement. The majority of the DB events generate a statistically insignificant impact on stock price. Similar results are obtained by using alternative assumptions that consider sector-specific portfolios, time variations in events, and possible information leaks or delayed market responses. Third, we explored what factors might help explain the market price reactions following the DB freeze/close announcements. The regression results (e.g., negative coefficients on company size and plan-funding risk) seem to suggest a “signaling” interpretation. That is, companies’ decisions to freeze/close their DB plans might have induced investors to question the financial health of the whole corporate entities beyond the pension plans. This reaction dominated the prospect of potential reductions in pension cost or volatility (if applicable), for which the market might cheer. Our empirical tests generally reject the expectation that freezing a pension plan would deliver an immediate boost to the company’s market value. Several factors may play a role. First, whether the DB plan freeze/close would substantially cut corporate costs is unclear because employers often need to enhance the existing 401(k) plans in the benefits package in order to facilitate the transition and workforce management. Second, any positive financial impact of the plan freeze/close may be outweighed by negative effects on employee morale, productivity, attraction, retention, and optimal retirement patterns. Third, the freeze/close events are often partial and gradual. Many companies sponsor multiple pension plans, have frozen/closed some plans while keeping others open, and have left the retirement benefits intact for many workers covered by the frozen/closed plans (e.g., union members or those fulfilling requirements of age and/or service years). Finally, company management may have viewed DB freezes/closes as useful responses to short-term financial challenges, but the market appears to have been more cautious about the effects and implications of such DB plan events. Note: The views expressed in this article are the authors’ alone and do not necessarily reflect the views of Watson Wyatt Worldwide.

Recognizing Retiree Health Benefits: The Effect of SFAS 106

The Financial Accounting Standards Board recently issued Statement No. 106 to apply to 1993 financial statements of companies providing health benefits to retired employees. The FASB expressed its view that these benefits are a type of deferred compensation, and as such, should be accounted for on the accrual basis rather than on the cash basis required under the interim disclosure standard, Statement No. 81.

Life Insurance Savings and the After-Tax Life Insurance Rate of Return

This paper presents a calculation of the time series of the after-tax rate of return to whole life insurancy. When compared to the after-tax return on an alternative portfolio of similar risk, more than 60%of the decline in life insurance savings (suitably defined) in the past two decades can be attributed to a widening after-tax rate of return differen-tial.Both the existence and importance of this result depend on the characteristics of life insurance savings. Life insurance saving is intimately connected to life insurance coverage and therefore is long-term and quasi-contractual in nature. Furthermore (and, in part, because of the above characteristics), the interest earned on the fixed income portfolio of life insurance intermediaries has been taxed under a special set of rules. From 1958 to 1981,these rules have taken the rather complicated form of the Menge formula. This formula is very sensitive to changes in nominal interest rate levels and in particular, during inflationary periods it acts so as to dramatically increase the tax burden of life insurancesavings.Life insurance savings is therefore an example of the non-neutrality of monetary policy. This is important for studies of flow of funds and capital accumulation using the historical record.

Illustrating Retirement Income for Defined Contribution Plan Participants: A Critical Analysis of the Department of Labor Proposal

The design solution for retirement income from individual accounts is the next major challenge to the evolving retirement system in the United States. In that regard, the Department of Labor has put forward a proposal to require sponsors of defined contribution plans to provide their participants with an illustration of the retirement income that might be expected from the assets held in their individual accounts through the use of an immediate fixed life annuity at retirement. Using an empirical analysis based on historical simulations, I show how the Department of Labor’s proposal, as currently formulated and with various improvements I recommend, will operate under realistic market conditions. I find that the proposal is a good start, but it can be improved by—among other changes—making the income flows gender-distinct and reflecting an inflation-indexed life annuity (with a load) rather than a nominal fixed annuity.

Retirement Income for Participants in DefinedContribution Plans and Holders of IRAs:Recent Developments in Immediate LifeAnnuities and Other Income Products

As more and more workers in the U.S. are covered only by defined contribution plans, and as more and more of these workers approach retirement, the use of life annuities and other newly developed retirement income products should increase. This article reviews historical and recent payout experiences with immediate life annuities, both nominal and inflation-indexed. It also explains new income product types, ranging from those that are just distributions of investments to those that contain complex guarantees from insurance companies. Finally, it provides some considerations for plan sponsors in this still understudied area of the decumulation phase of the individual account retirement plan.

Practical Applications of Improving the System of Financing Long-Term Services and Supports for Older Americans

Practical Applications Summary In Improving the System of Financing Long-Term Services and Supports for Older Americans, from the Summer 2017 issue of The Journal of Retirement, authors Mark Warshawsky and Ross Marchand (then both with the Mercatus Center at George Mason University) explore possible weaknesses of the existing framework under which seniors can qualify for long-term services and supports (“LTSS”) under Medicaid. They explain that aspects of the eligibility criteria in certain states allow seniors to draw on the Medicaid resources even if they have sufficient resources to pay for their care. The authors also discuss how states with mandates to recover assets from deceased Medicaid recipients have a poor track record of doing so. The authors assert that the market for long-term care insurance (“LTCI”) is being depressed by the ability of seniors to qualify for Medicaid LTSS benefits even when they can afford private LTCI. The authors offer several policy recommendations, including tightening Medicaid eligibility standards, narrowing the “primary residence” exclusion, and implementing a more robust estate recovery scheme.