Women and Retirement Issues
Olivia Mitchell, University of Pennsylvania

The four papers in this session explored labor market and demographic determinants of older women s wellbeing, focusing on behavioral patterns of women in their 50 s and 60 s. Two of the studies used to good advantage the interesting new Health and Retirement Study (HRS), a longitudinal survey of Americans age 51-61 in 1992 and their spouses. In her research, Marjorie Honig from Hunter College evaluated whether and how earnings, pensions, social security, and health insurance affect older married women s expectations of working after age 62. This paper, entitled "Married Women's Retirement Expectations: Do Pensions and Social Security Matter?", concluded that married women who work full time plan on remaining employed longer when their pay and health benefits are better. Conversely, factors inducing them to retire earlier include having a pension and higher levels of family income. Tastes also play a role, not surprisingly: wives whose husbands expect to keep working also plan to remain on the job longer, while women valuing joint leisure anticipate earlier retirement. The second HRS analysis, by Melinda Pitts and Michelle Trawick from Salisbury State University, asked how wives work patterns are influenced by their husbands poor health. In their paper, entitled "Married Women's Retirement Decision: Does Husband's Health Matter? , the authors evaluated which factors might induce working women to reduce their work hours over time. The richness of the health measures captured in the HRS permitted the authors to assess whether having a husband with a life-threatening health problem has a differential effect than would having a spouse with chronic and/or emotional health problems. In addition they explored responses to health practices including smoking, exercise, and drinking.

Understanding how government pensions influence women s retirement was the focus of Nabanita Gupta and Bent Christensen from the University of Aarhus, in their paper "The Effects of Pension Systems Reform on Female Retirement: Evidence from Denmark." Ostensibly to reduce unemployment, the Danish government permits workers as young as age 60 to take early retirement. This early retirement option is quite attractive to many, inasmuch as government pensions are normally available only at age 67. The authors estimated a hazard model using data on older married couples, and use the estimated parameters to simulate benefit rule changes. The evidence suggested that reducing benefit amounts saves the underfunded pension system more money than does raising the age of eligibility, and keeps women in the labor market longer. Government pensions also were the topic of the final paper in this session, an analysis by Sharmila Choudhury and Michael Leonesio from the Social Security Administration. These authors showed that American women s eventual social security benefits depend not only on the women s work histories, but also on their spouses earnings patterns and how long they were married. Therefore dramatic changes in women s employment, pay, and marriage/divorce patterns in the last forty years will have long-lasting and still not fully understood effects on women s eventual social security retirement payments. Interim results were summarized in a paper entitled "Marital Status and Lifetime Earnings: Outlook for American Women". The authors suggested that younger women now have spent many more years in the paid labor market than did their precursors, but these additional work years do not necessarily translate into benefits that exceed what they could have received as spouses.

Excellent discussion of the papers was afforded by Pamela Loprest of the Urban Institute; Liliana Pezzin of the Agency for Health Care Policy and Research at HHS; Rebecca Luzadis from Miami University, and Richard Johnson from Rutgers University.

"Developments in the Labor Market for Women"
Joyce P. Jacobsen, Wesleyan University

Donna Ginther (Washington University) and Chinhui Juhn (University of Houston) presented "Employment of Women and Demand-Side Forces," which examines the relationship between female employment growth and changes in labor demand. Using the 1964-95 March Current Population Surveys and the 1940-1990 Censuses, they find that female employment growth accelerated in the 1970s and 1980s and that growth was biased in favor of more skilled women. They are able to account for the pattern of biased growth using labor demand proxies, but are unable to account for the overall accelerating growth trend with their demand proxies.

Elizabeth Brainerd's (Williams College) paper, "Women in Transition: Changes in Gender Wage Differentials in Eastern Europe and the Former Soviet Union," uses household survey data from three former Soviet republics and six East European countries to consider whether women have maintained their relative positions in the wage distribution since the reforms of the early 1990s. She reported a substantial increase in female relative wages in the Eastern European countries, but a substantial decline in female relative wages in Russia and Ukraine. In all the studied countries, women have been negatively affected by the tremendous widening of the wage distribution, but in Eastern Europe the losses were offset by gains in returns to observed skills and an apparent decline in discrimination against women.

Dean Dudley (US Military Academy), Deborah Viola (New York Medical College), and Casey Wardynski (RAND) gave the third paper, "The Incidence of Military Service on Married Household Earnings." They reported preliminary results from 1990 Census data and Department of Defense data contrasting earnings of spouses in civilian and military households. The data supported the expected finding of an earnings penalty related to being a spouse in a military household.

The final paper, by Karen Mumford and Peter N. Smith (both University of York), was titled "Men, Women and the Hiring Function." They use Australian quarterly data from 1980:4 through 1991:4 to characterize worker flows between the states of unemployment, employment, and not in the labor force. Using log-linear regression equations to relate the flow rates, they find evidence supporting the ranking of men over women in the hiring process. Among men, employed job seekers are ranked above those unemployed, and in turn, those not in the labor force. Among women, the unemployed and employed did not appear to be competing with each other, while those not in the labor force are ranked below the unemployed.

Discussants Theresa Devine (Congressional Budget Office), Katherine Terrell (University of Michigan), Leslie Stratton (Virginia Commonwealth University), and Amy Aukstikalnis (Welch Consulting) provided thoughtful contributions.

"Gender and Risk Aversion"
Kathryn Shaw, Carnegie Mellon University

The general question raised by this session was, are women more risk averse than men? If so, this could imply, for example, that women will have lower levels of wealth in the long run if they allocate their wealth to assets that are too conservative. This is particularly important as Congress debates whether to create retirement accounts in Social Security that would give people the right to allocate their wealth across alternative assets. This session was remarkably cohesive in content and very well attended. Catherine Eckel (National Science Foundation), Philip Grossman (University of Texas at Arlington), and Nancy Lutz (Virginia Polytechnic) presented "Playing It Safe: Men, Women, and Risk Aversion" (joint with V. Padmanabhan of Stanford University). Using data on mens' and womens' decisions to purchase extended warranties for computers, the authors find that women are more risk averse than men-they are more likely to purchase warranties, all else constant (including their perceptions of risk). The remaining three papers look at the allocation of pension wealth. In "Women, Risk Taking, and Pensions Decisions," Nancy Jianakopolos (Colorado State University) uses data from the 1989 Survey of Consumer Finances to examine womens allocation of wealth between defined contribution plans and other assets (in joint research with Vickie Bajtelsmit and Alexandra Bernasek also of Colorado State). They find that women allocate a smaller proportion of their wealth to DCP plans, and that the difference arises from a number of factors-older women, single women, and women with kids allocate less to these plans than do men. Using more recent data from the SCF (for 1992 and 1995), Annika Sunden (Federal Reserve Board of Governors) presented results on how women and men allocate their pension wealth, given that they have a defined contribution plan (from "Gender Differences in the Allocation of Assets in Retirement Savings Accounts," with Brian Surette). They find mixed results: single women are more likely than all others to allocate wealth to stocks and married men and women are least likely to do so. Thus, there is no clear gender effect in risk aversion. Leslie Papke (Michigan State University) also looks at how people allocate the wealth that is in their defined contribution plans, but uses data from the NLS mature women sample for 1992 in "How are Participants Investing Their Accounts in Participant-Directed Individual Account Pension Plans?." Assuming that those who invest in "mostly stocks" are taking greater risk, she finds no difference in gender for this sample of older women. Thus, overall, the effects are a bit mixed, though previous evidence has tended to find some greater risk aversion among women. The discussants were Kip Viscusi (Harvard Law School), Andrea Kusko (Federal Reserve Board), John Turner (International Labour Organization) and Annamaria Lusardi (Dartmouth College). One issue that was noted is that using data on wealth allocation and warranty purchase, it is difficult to separate out the effects of the husband's risk preferences, thus authors need to turn to data for single-headed households, which are smaller in number.

Annamaria Lusardi (Dartmouth College) estimates a precautionary saving model using an improved measure for income risk in "On the Importance of the Precautionary Saving Motive." Using data from the Health and Retirement Survey she finds that financial and total net worth are increased when individuals perceive an increase in the variance of earnings. Nancy Jianakopolos (Colorado State University) offered suggestions for extensions of the paper; separating out bequests and precautionary saving motives, getting at early retirement issues through Social Security and pension wealth, and controlling for risk aversion and health status.

Joanne Doyle (James Madison University) and Toni M. Whited (University of Maryland and University of Delaware) present evidence against representative-agent models in explaining industry-level investment in, "Fixed Costs of Adjustment, Coordination, and Industry Investment." They find that deviations of the actual and optimal capital stock are more persistent for low idiosyncratic (relative to aggregate uncertainty) industries. This finding lends support to the proposition that smooth industry-level investment results from lumpy, asynchronous firm-level investment. Jane Ihrig (University of Virginia) noted that the results are very strong given the use of more aggregated and lower frequency data than called for in this situation. She suggested looking into the newer and better capital stock data, employ subperiod tests, and reconsider whether idiosyncratic shocks are i.i.d.

Rather than estimate the average correlation of wages and employment, Marcello Estevao and Beth Anne Wilson (Board of Governors of the Federal Reserve System) depart from common practice and attempt to identify both the labor supply and labor demand curves to differentiate among macroeconomics models. In "Nominal Wage Rigidity and Real Wage Cyclicality," nominal wage rigidity is tested by identifying movement along the labor supply and labor demand curves using appropriate instruments. They conclude that nominal wage rigidity is an important channel for transmission of aggregate demand shocks to the real side of the economy. Janice Eberly (University of Pennsylvania) suggested reconsidering whether innovations in the real federal funds rate is the appropriate instrument to use in identifying the labor demand equation; close the model by specifying price determination; rule out alternative interpretations of the results.

Erica L. Groshen (Federal Reserve Bank of New York and Mark Schweitzer (Federal Reserve Bank of Cleveland) presented "Inflation's Grease and Sand Effects in the Labor Market." Using a 40-year panel of wage changes they sort out inflation-induced deviations among employers' mean wage-changes (disruptive or sand effects) from inflation-induced, inner- occupational wage-changes (beneficial or grease effects). They find that the grease and sand effects tend to coexist, but cancel at low levels of inflation. At high levels of inflation, the sand effect dominates. Shulamit Khan (Boston University) was convinced that the authors have indeed identified the two effects with the grease being limited and the sand effect unlimited. She offered several technical suggestions.

"Banking and Financial Crises"
Hali J. Edison, Federal Reserve Board

Hali Edison (Federal Reserve Board) chaired the CSWEP session on "Banking and Financial Crises." Victoria Miller (University of Montreal) presented "The Double Drain with Cross- Border Twist: More on the Relationship Between Banking and Currency Crises." The paper discusses the possible links between currency and banking crises and looks at the transmission of financial crises. It gives a cross-border twist by providing examples of how a banking (currency) crisis in one country can give rise to a currency (banking) crisis in another. In an empirical paper -- "Financial Crises in Asia and Latin America: Then and Now" -- Graciela Kaminsky (Federal Reserve Board) and Carmen Reinhart (University of Maryland) examine the behavior of fifteen indicators to identify behavior of economies prior to a crisis. In general, they find that financial crises are more severe in Latin America than elsewhere, but that these differences, especially compared to the current Asian crisis, are narrowing.

Nancy P. Marion (Dartmouth College) and Robert P. Flood (international Monetary Fund) delivered a paper that provided a more general overview of the literature on currency crises. Their paper entitled "Some Perspective on the Present Currency Crisis Literature," provides a review of the new theoretical and empirical literature. It goes on to extend the existing models by deriving a model that incorporates the optimal commitment to a fixed exchange rate. Finally Liliana Rojas-Suarez (Inter-American Development Bank) gave the paper "Early Warning Indicators of Banking Problems: What Works for Latin America?" This paper shows that bank-based early warning systems developed for industrial countries are not effective for detecting banking problems in Latin America. The paper proposes and tests and alternative set of indicators, which is quite robust.

The discussants were Linda Goldberg (Federal Reserve Bank), Angela Redish (University of British Columbia), Caroline Betts (University of Souther California), and Sara Calvo (World Bank).

"Human Capital Accumulation and Growth"
Daniel H. Newlon, National Science Foundation

Four papers were presented at this session. Anne Owen of Hamilton College and Murat F. Iyigun of the Federal Reserve explored the implications for growth of the existence of more than one type of human capital, showing how the choice between entrepreneurship and professional employment evolves as an economy develops and examining how individuals' decisions to accumulate different types of human capital affect the economy's long-run potential.

Ruth Judson of the Federal Reserve Board measured human capital in value rather than person-years and then used this new measure in cross-country panel regressions to show that human capital accumulation yields a positive and significant, but relatively small (about ten percent) elasticity with per-capita GDP and that the stock of human capital as a share of GDP increases with GDP.

Mary Lovely and Douglas Holtz-Eakin of Syracuse University found that an aging population could increase government spending on education despite shifts toward voters with a lower preference for educational spending because elderly voters also favor policies that increase growth rates and a higher stead-state level of output per person provides a larger tax base for government investment in education.

Jane Ihrig of the University of Virginia and Karine S. Moe of Macalester College showed that government policies of taxation and enforcement significantly affect the size of the informal sector across regions of the world as well as in individual Asian miracles and that the size of the informal sector significantly and negatively affects real GNP per worker.

Lant Pritchart of the World Bank, Patricia Pollard of the Federal Reserve Bank of St. Louis, Gerhard Glomm of Michigan State University and Shohreh Majim of Western Michigan University provided extensive and very helpful comments on the papers. There were also questions and comments on each paper from the audience in this productive session.

CSWEP Roundtable on "Social Security Reform: How Will Women Fare?"
Cordelia W. Reimers, Hunter College and the Graduate School of CUNY

Teresa Devine of the Congressional Budget Office led off the CSWEP Roundtable on "Social Security Reform: How Will Women Fare?" with an overview of the financing problem facing Social Security and key features of the existing system and major reform proposals, some of which involve partial privatization. She pointed out that Social Security currently is the most important source of income for older women, but still leaves many of the oldest old who live alone on a one-person benefit -- most of whom are women -- in poverty. In addition, because the current tax-benefit structure favors one-earner couples, married women may not see a direct return on their payroll taxes. Because most women marry and earn less than their husbands, the future economic security of most older women will depend on what reform does to their Social Security benefits based on their spouse's earnings. Decisions about how much to allocate to the "social" defined-benefit income security program versus the "private" individual pension accounts, the size of spouse and survivor benefits in the social program, and the rules governing distribution of individual accounts (e.g., mandatory joint and survivor annuities, spouses' and ex- spouses' ownership and inheritance rights) will affect women significantly. Teresa was applauded at the end of her brief but information-packed summary of the issues.

Because Edith Fierst, Washington attorney and member of the 1994-96 Advisory Council on Social Security, could not be present in person, excerpts from an article she had written for the May 5, 1997 issue of Tax Notes, "Privatization of Social Security -- A Threat to Women," were read by Cordelia Reimers of Hunter College-CUNY, who chaired the roundtable. Ms. Fierst emphasized the importance to women of a legal right to widows' and spouses' benefits that do not reduce their husbands' own benefits. She predicted that under privatization many wives would lose their rights to retirement income, due to divorce or to unilateral decisions by their husbands, and many people would outlive their incomes because private annuities are too expensive and lack inflation protection. She advocated increasing the survivor benefit, with money being saved by reducing the spouse benefit and increasing the benefit computation period from 35 to 38 years.

Karen Holden of the University of Wisconsin-Madison described the income consequences of widowhood among women in the U.S. and the protection provided by the current Social Security program and employer-provided pension plans. She presented data on the correlates of election of survivor benefits by husbands and emphasized the importance of federal laws governing pension distributions in increasing the percentage of women receiving survivor pension benefits. She concluded by urging that any proposals for individual accounts include mandates for default annuitization, a default survivor pension offer, and spousal consent if the survivor pension is rejected.

Steven Sandell of the Social Security Administration reported the results of research by Howard Iams and himself concerning the effects of certain proposed reforms on poverty among the older population. They examined the effect on poverty rates of reducing the spouse benefit to one third of the retiree's benefit, and increasing the survivor's benefit to 75 percent of the couple's total benefit.

Marjorie Honig of Hunter College-CUNY argued that a shift away from the redistributive and insurance aspects of the present Social Security system to individual accounts is not an issue particular to women, because women's labor force participation, earnings, private pensions, savings, and investment behavior will become more like men's over the next 30 years. She observed that due to rising divorce and falling remarriage rates, women know they must take care of their own financial futures, and that therefore their labor force participation and financial expertise will continue to increase. Individual accounts need not be more risky for women than for men, provided the default distribution is a joint and survivor annuity and any alternative requires spousal consent. As women gain experience with defined-contribution employer pension plans and as investors generally, their investments will become less conservative and therefore earn higher returns, as well.

Marilyn Moon of the Urban Institute, who is a public trustee of the Social Security and Medicare Trust Funds, emphasized that privatization and rates of return are not the only issues. In making changes to Social Security, it is essential to consider what is likely to happen to Medicare and Medicaid, since retirement income needs depend in part on medical expenses and all three programs must be financed. She observed that women have a greater stake in a more progressive, redistributive system because they are likely to continue to have lower benefits than men.