How will demographic trends, in conjunction with the enormous changes in private retirement saving, affect market rates of return in the United States? The spread of 401(k) plans in particular is likely to change by an order of magnitude the flow into and out of financial assets. Thus to understand the implications of demographic change, the implications of evolving private personal retirement accounts must be understood.
The effects of demographic change on rates of return within a country can be moderated by capital flows between countries. This multi-phase project involves the development and application of an empirical model of how demographic variation across countries affects international financial flows and market rates of return.
Project NB06-11: Effects of Cross-Country Demographic Changes on International Capital
Flows and Financial Markets within and across Countries: Realistic Age-Saving Profiles Should be Countries.
The financial balances of the Social Security system depend to a significant extent on what happens to longevity in the future. This project examines the impact of current trends in health behaviors and medical technology on likely changes in future death rates in the United States, and thus on the financial implications for the Social Security system. First, we will examine the relationship between risk factors and death rates, including analyses of smoking, weight, hypertension, cholesterol, alcohol consumption, and diet. We will then examine how potential changes in risk behaviors in the future are likely to affect health and death rates.
The level of fertility in a society is a key determinant of its age distribution and of its growth rate. Among countries of Europe and North America, the United Status has the highest level of fertility - just below the replacement level. What accounts for the exceptional level of American fertility, how might it change in the future, and what does this suggest for the age-distribution of the US population in the future?
The debate between those supporting traditional pay-as-you-go Social Security and a funded Personal Retirement Account (PRA) approach is often characterized as the choice between a safe Defined Benefit (DB) plan and a riskier Defined Contribution (DC) one. This, of course, is an over-simplified characterization. A PRA system is not riskless, but is subject to market risk. A pay-as-you-go DB system is not riskless either, but is subject to political risk. The project compares these two forms of risk.
Close to 20 countries have converted at least part of their pay-as-you-go defined-benefit public pension systems to systems based on funded, defined-contribution accounts. This project considers the intergenerational implications of policy change, its relationship to "trust" or "distrust" in government, and the systematic differences in trust between developed and developing countries.
This project has two parts. The first part adds a number of enhancements to our existing models for forecasting model for Social Security: a) the use of structural time series models for all inputs; b) the effects of parameter uncertainty; c) treatment of immigration as a stochastic input; d) differences between the approach to modeling mortality taken by CBO and OASS, versus alternative methods. The second part of the project develops stochastic forecasts of infinite horizon measures of fiscal imbalance.
The goal of this project is to estimate how financial markets would value outstanding Social Security liabilities, accounting for future uncertainty in a way that investors would do, if they regarded future Social Security payments as dividends on assets, or liabilities of their own businesses. The project explores both the "tradable" value of future benefits, and the creation of innovative financial products that could enable such trades to take place.
Executive Summary Policy Abstract   Working Paper
This project examines nine approaches to reducing the amount of financial market risk borne by future retirees and future taxpayers in an investment-based Social Security system. Using stochastic simulation techniques the project examines both the magnitude of the risk and the distribution of the risk between future retirees and future taxpayers for each approach.
A "Notional Defined Contribution" (NDC) plan is a special kind of pay-as-you-go pension plan, in which the benefit is defined in a way that mimics some features of a defined contribution system. This project will consider how an NDC-type plan - as currently used in Sweden - might be structured in the United States. The project will also analyze the sustainability of Social Security finances under such a system, based on uncertain demographic and economic outcomes going forward.
Any reform to the social security system has implications for how the costs and benefits of the system are distributed across generations. For example, a shift from a public pay-as-you-go system to a system of private, defined-contribution accounts imposes an incremental burden on transition generations. In addition, however, social security reforms alter the sharing of risk among generations, both during and after the transition. This project evaluates both the intergenerational redistribution and the intergenerational risk-sharing properties of alternative reforms.
Project NB06-09: Notional Defined Contibution Pension Systems in a Stochastic Context: Design and StabilityExecutive Summary Policy Abstract Working Paper
This project considers how alternative asset allocation strategies would affect the risks faced by individual investors in a PRA system. Variants of three investment options are considered. The first is a broad market "lifecycle fund" that rebalances automatically towards less risky investments as the retirement age nears. The second is a "no-loss" strategy such as the one proposed by Martin Feldstein. The third is a simple US equity market index fund combined in various fixed proportions with US government bonds. The analysis considers the effects of such investment strategies on the distribution of PRA retirement wealth. It also considers the administrative costs of the lifecycle and no-loss strategies relative to simpler strategies and examines whether the risk reduction is likely to be worth the cost.
Some Social Security reform plans with personal accounts would guarantee that participants would receive an annuitized level of benefits above some threshold, such as currently scheduled benefits. The purpose of this project is to develop the tools for pricing, or assessing the market value, of such guarantees.
In this project, we will provide a comprehensive framework for analyzing the progressivity of the Social Security retirement program across different definitions of progressivity, alternative measures of individual and household well-being, and changes across cohorts.
The goal of this project is to develop Social Security policy options that could eliminate elderly poverty and reduce the share of elderly who are near poor. Initially, the project will identify the features in the current Social Security system that lead to people having income below the poverty level in old age. We will then analyze various approaches to reducing elderly poverty, such as changes in the replacement rate factors in the PIA formula, special provisions for widows and divorcees, and minimum benefit provisions.
Project NB06-08: Could Social Security Eliminate Poverty among the Elderly?
This Project examines changes in poverty among individuals 65 and over during the last three decades, comparing various income and consumption-based measures of poverty, as well as measures of extreme poverty and near poverty. We will also develop and apply to our research several methodological advances in poverty measurement.
This project considers the distributional and efficiency impacts of taxing Social Security benefits. It evaluates the extent of means-testing that can be achieved through tax policy. Taxation of benefits is also compared with other approaches to achieving distributional objectives in Social Security, such as altering the Primary Insurance Amount (PIA) formula.