Skip to main content
Log in

Multistate Transfer Rate Estimation from Adjacent Populations

Population Research and Policy Review Aims and scope Submit manuscript

Abstract

In multistate populations, the rates of interstate transfer cannot generally be determined from the size and composition of the populations at the beginning and end of a time interval. With N living states, the population data give only N equations to determine the N 2 possible rates. Here, the QERT (quadratic estimation of rates of transfer) approach is advanced that allows the transfer rates to be estimated when the products of selected pairs of rates can be assumed constant. The solution can be written in closed form and, for N living states, involves no more than N−1 quadratic equations. Compared to the leading alternative approaches, QERT provides very similar numerical estimates, while yielding the underlying behavioral rates, having flexible input requirements, accommodating all structural zeros, and reproducing the exact solution when interstate transfers are strictly hierarchical. The QERT approach is applied to construct labor force life tables for U.S. men and women for 2005–2010. The results show that labor force participation differences between men and women have continued to narrow, and that the QERT approach can generate robust worklife estimates. QERT thus provides new opportunities for demographic analysis in the absence of direct data on behavioral rates.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

References

  • Abel, G.J. 2013. Estimating global migration flow tables using place of birth data. Demographic Research 28: 505–546 (15 March, Article 18).

  • Arias, E. 2012. United States life tables, 2008. National vital statistics reports, Vol 61, No 3. Hyattsville MD: National Center for Health Statistics.

  • Bishop, Y. M., Fienberg, S. E., & Holland, P. W. (1975). Discrete multivariate analysis: Theory and Practice. Cambridge, MA: MIT Press.

    Google Scholar 

  • Bureau of Labor Statistics. 2005, 2010. Household Data. Downloaded July 20, 2013 from www.bls.gov/cps/cpsaat03.htm

  • Bureau of Labor Statistics. 2014. Current Population Survey; Frequently Asked Questions. Downloaded January 21, 2014 from www.bls.gov/cps/faq.htm

  • Chilton, R., & Poet, R. (1973). An entropy maximizing approach to the recovery of detailed migration patterns from aggregate census data. Environment and Planning, 5, 135–146.

    Article  Google Scholar 

  • Eisinga, R. (2008). Recovering transitions from repeated cross-sectional samples. Methodology: European Journal of Research Methods for the Behavioral and Social Sciences, 4, 139–151.

    Article  Google Scholar 

  • Guillot, M., & Yu, Y. (2009). Estimating health expectancies from two cross-sectional surveys: The intercensal method. Demographic Research, 21, 503–534.

    Article  Google Scholar 

  • Kruithof, J. (1937). Calculation of telephone traffic. De Ingenieur, 52, E15–E25.

    Google Scholar 

  • Nair, P. S. (1985). Estimation of period-specific gross migration flows from limited data: Bi- proportional adjustment approach. Demography, 22, 133–142.

    Article  Google Scholar 

  • Philipov, D. (1978). Migration and settlement in Bulgaria. Environment and Planning A, 10, 593–617.

    Article  Google Scholar 

  • Rogers, A., & Raymer, J. (2005). Origin dependence, secondary migration, and the indirect estimation of migration flows from population stocks. Journal of Population Research, 22, 1–19.

    Article  Google Scholar 

  • Rogers, A., & von Rabenau, B. (1971). Estimation of interregional migration streams from place- of-birth-by-residence data. Demography, 8, 185–194.

    Article  Google Scholar 

  • Schmertmann, C. P. (2002). A simple method for estimating age-specific rates from sequential cross sections. Demography, 39, 287–310.

    Article  Google Scholar 

  • Schoen, R. (1988). Modeling multigroup populations. New York: Plenum.

    Book  Google Scholar 

  • Schoen, R. and S.H. Jonsson. 2003. Estimating multistate transition rates from population distributions. Demographic Research 9, 1–24 (29 August, Article 1).

  • Schoen, R., & Standish, N. (2001). The retrenchment of marriage: Results from marital status life tables for the United States, 1995. Population and Development Review, 27, 553–563.

    Article  Google Scholar 

  • Schoen, R., & Weinick, R. (1993). The slowing metabolism of marriage: Figures from 1988 U.S. marital status life tables. Demography, 30, 737–746.

    Article  Google Scholar 

  • Schoen, R., & Woodrow, K. (1980). Labor force status life tables for the United States, 1972. Demography, 17, 297–322.

    Article  Google Scholar 

  • Smith, S.J. (1982). New worklife estimates reflect changing profile of labor force. Monthly Labor Review March, pp. 15–20.

  • Smith, S.J. (1985). Revised worklife tables reflect 1979-80 experience. Monthly Labor Review August, pp. 23–30.

  • U.S. Census Bureau. (2007). Current Population Survey, Annual Social and Economic Supplement, 2005.

  • U.S. Bureau of Labor Statistics. (1950). Tables of working life: Length of working life for men. Bulletin No. 1001. Washington DC: U.S. Government Printing Office.

  • U.S. Census Bureau. (2010). Census Summary File 1. Population by Age and Sex: 2000 and 2010.

  • Willekens, F. J. (1980). Multistate analysis: tables of working life. Environment and Planning A, 12, 563–588.

    Article  Google Scholar 

  • Willekens, F. J. (1982). Multidimensional population analysis with incomplete data. In K. C. Land & A. Rogers (Eds.), Multidimensional Mathematical Demography (pp. 43–111). New York: Academic Press.

    Chapter  Google Scholar 

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Robert Schoen.

Appendix: Constructing Labor Force Life Tables for U.S. Men and Women, 2005–2010

Appendix: Constructing Labor Force Life Tables for U.S. Men and Women, 2005–2010

The 2005 and 2010 U.S. census populations, by gender, were tabulated by age groups \(0{-} 4,{ 5}{-} 9, \ldots ,{ 75}{-} 7 9,{\text{ and 8}}0 +\). At every step, separate calculations, using the same procedures, were done for the male and female data. Because the census populations were not closed to migration, the 2010 population, at all ages over 5, was taken to be the 2005 population survived 5 years by the person-year values in the 2008 U.S. Life Tables. The Bureau of Labor Statistics (BLS) data on the employment status of the civilian noninstitutional population were then applied, age by age, to divide the populations into the statuses In the Labor Force (A) and Not in the Labor Force (N). Since the BLS data had no one in the labor force below age 16, interpolation based on the 2008 U.S. Life Tables was used to allocate persons in the 15–19 year age interval. At every age, the same mortality was assumed to apply to both labor force statuses.

The BLS considers persons to be in the labor force if they are employed or unemployed. Employed persons are those who (i) did any work for pay in the survey reference week, (ii) did unpaid work for 15 or more hours in a family business, or (iii) were temporarily absent from their regular job. Unemployed persons are those without a job who have actively looked for work in the prior 4 weeks and are currently available to work. Persons are classified as not in the labor force if they are neither employed nor unemployed during the reference week (Bureau of Labor Statistics 2014).

Flow Eqs. (13) provide the structure for the life table calculations. First, linear interpolation using 2008 U.S. Life Table values was used to shift the age categories by half an interval, so that the intervals begin at ages \(0,{ 2}. 5,{ 7}. 5, \ldots ,{ 72}. 5,{\text{ and 77}}. 5\). Population (P) values for ages x to x + n were taken to be at age x + n/2. The person-years lived at each age by the age cohort during the 5 year interval from 2005 to 2010 were found from the linear relationship in Eq. (3). The k values were also linearly adjusted to the new age intervals. QERT Eqs. (15) and (16) were then used to determine the age-specific rates of entry to and exit from the labor force.

The life table construction procedures followed Schoen (1988, Chap. 4). Because of the magnitude of the rates at young ages, the exponential (or constant forces) method was used to calculate the state-specific survivorship () and person-year (L) values. The equations in Schoen (1988, p. 86) were used to determine the person-years lived, by state, at ages over 77.5. Once the person-year values, by age and state, were found, the number of transfers between states (d ij ) were calculated using the relationship \(d_{ij} = m_{ij} L_{i}\). Summary measures were found in a straightforward manner. The expected number of years in the labor force, \({\text{e}}_{\text{A}} \left( 0 \right)\), is \(\Sigma_{x} L_{A} \left( x \right)/\ell \left( 0 \right)\), i.e., the number of years lived in the labor force by the life table cohort at all ages, divided by the initial size of the life table cohort. The proportion of life lived in state A between ages 17.5 and 67.5 is the sum of the L A values for those ages divided by the sum of the total L values at those ages. The average number of entries to the labor force per person is \(\Sigma_{x} {\text{d}}_{\text{NA}} \left( x \right)/\ell \left( 0 \right)\). The average duration in the labor force per entry is the total number of years lived in the labor force, \(\Sigma_{x} L_{\text{A}}\), divided by \(\Sigma_{x} d_{\text{NA}}\), the total number of labor force entries. Consistency checks verified the flow equations at every age, and confirmed that the total number of entries to each state equaled the total number of exits.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Schoen, R. Multistate Transfer Rate Estimation from Adjacent Populations. Popul Res Policy Rev 35, 217–240 (2016). https://doi.org/10.1007/s11113-015-9376-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11113-015-9376-7

Keywords

Navigation