The U. S. Railroad Retirement Board's Bureau of the Actuary publishes valuations of the retirement plan for railroad workers every three years. The most current report, the Twenty-Fifth Actuarial Valuation, released in August 2012 covers the years 2008-10. We use data on mortality, disability retirements, age retirements, and other final withdrawals contained in the Technical Supplement of this valuation to estimate worklife expectancies (WLE) of railroad workers. There are substantial changes in WLE when compared to WLE based on the Twenty-Third Actuarial Valuation and the Twenty-Fourth Valuation. We identify the sources of these changes. In addition to WLE, we estimate entire probability mass functions and provide summary measures of distributional characteristics of time spent in railroad work. Our main results use a competing risks/multiple decrement model, but we also provide WLE based on the railroad Markov process model and show that the former model is also a Markov model, albeit a different Markov model from the railroad Markov model.Abstract
This paper provides empirical evidence on the behavior of various Mexican Net Discount Rates (NDRs) by examining the time series properties of several yields on Mexican government securities and annual wages for nine industrial sectors. Results indicate that NDRs are characterized as a stationary series around a shift in its mean. The forecast for the mean of the NDRs should be based on the period after the time of the shift, which took place in 1994. The discussion addresses relevant case law and the application of the NDR approach to the calculation of damages for Mexican undocumented workers as an effort to equip forensic economists with the tools to conduct such estimates.Abstract
The study explores time series properties of three wage net discount rate series derived using interest rates based on (alternatively) 1-year, 6-month and 3-month Treasury securities coupled with wage growth rates, initially for the period from 1981:01 to 2012:06, then subsequently through 2012:12. Stationarity tests are run on the full series and various sub-series to identify any portion of the series on which reliable forecasting can be based. Initially no support is found for the total offset hypothesis based on the full-time series (but support is subsequently found for total offset when exploring various data sub-series). Positive findings include that the three trended wage net growth rate series for the entire period under study are stationary, implying that reliable short-term forecasting of wage net discount rates is possible based on each of the trended series. Short-run forecasts based on the trended series are presented for 2012:07–2012:12 and compared to actual data in the period. Finally, the three trended wage net discount rate series are re-examined for 1981:01–2012:12, with short-term forecasting equations presented. Various sub-series of WNDRs are then explored in hopes of finding one or more that may be stationary about a constant term. Additional testing identifies three sub-series all ending 2012:12 with varying start dates that have desirable stationarity properties. The sub-series starting 2007:11 is highly stationary, but peculiar, with an associated constant WNDR that is negative and statistically different from zero; and the two sub-series beginning 1990:12 and 1994:05 have stationary attributes, yet possess constant terms that are slightly positive but not statistically different from zero, thus providing modest support for the total offset method.Abstract
The National Association of Forensic Economics (NAFE) has approximately 650 members across the United States and in other countries. While the association has been an active group at the annual meetings of the Allied Social Science Association (ASSA) and at other regional meetings of economists, the growth of NAFE in terms of longevity and finances has allowed the organization to develop a more professional presence for its academic and practitioner members. This paper will update the original history of NAFE authored by Michael L. Brookshire in the Litigation Economics Review in 2003 which covered the period from NAFE's inception in 1986 through 2001.Abstract
Using Consumer Expenditure Survey data, this paper presents a unified approach in estimating adult personal consumption across five family types: married wage earners living together by themselves, married wage earners living together with their minor children, married retirees living together by themselves, single wage earning parents living together with their minor children, and single wage earners living alone. In addition to advancing the literature with a unified analysis of family traits, this paper uses over 700 microdata expenditure classifications instead of the approximate 30 summary expenditure classifications which have been relied on in other personal consumption studies. Microdata expenditure classification reduces problems of allocating expenses among family members. The paper begins with an overview of the Consumer Expenditure Survey, the federal data source that is most appropriate for estimating personal consumption. Next presented are the allocation rules to separate personal from survivor and household expenditures. The data are then identified followed by empirical estimation.Abstract
This note rearranges the data from Census Bureau Personal Income (PINC) tables 32–35 showing earnings of full-time, year-round workers from 1974 to 2012 to display the 39-year time series for average real earnings by education, age group, and sex. Aggregated data show strong upward trends for all males and all females combined, for males and females with a bachelor's degree or more and for those with less than a bachelor's degree. However, all trends have flattened or become negative since 2000. Shifts in the composition of the work force over time mean that trends in aggregated earnings incorporate those shifts, so may not be useful for projecting earnings growth for any individual plaintiff of a specific age and educational background. The note addresses that problem in two ways: First it re-weights the aggregate statistics to remove some effects of shifts in composition of the workforce. Second it gives summary statistics for log-linear regressions of real earnings on time for all education-age-sex combinations for the period 1974–1999 and for 2000–2012. Growth for the latter period is negative for most of those combinations. Supplemental material for a Microsoft Excel workbook contains the underlying data and several figures and is accessible from the Journal of Forensic Economics website.Abstract