Changes in the Economic Well-Being of Widows Following the Death of Their Husband: A Four Country Comparison

Burkhauser, Richard V.; Giles, Philip; Lillard, Dean R.; Schwarze, Johannes Article Changes in the Economic Well-Being of Widows Following the Death of Their Husband: A Four Country Comparison Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Burkhauser, Richard V.; Giles, Philip; Lillard, Dean R.; Schwarze, Johannes (2003) : Changes in the Economic Well-Being of Widows Following the Death of Their Husband: A Four Country Comparison, Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 123, Iss. 1, pp. 151-161, https://doi.org/10.3790/schm.123.1.151 This Version is available at: https://hdl.handle.net/10419/292045 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Changes in the Economic Well-Being of Widows Following the Death of Their Husband: A Four Country Comparison* By Richard V. Burkhauser, Philip Giles, Dean R. Lillard, Johannes Schwarze Abstract Using Cross-National Equivalent File (CNEF) data for the United States, Germany, Great Britain, and Canada, we investigate the economic loss widows experience. We find that, unadjusted for changes in household size, the loss measured by the household income replacement rate is smaller than the loss measured by the social security replacement rate. The household income replacement rate also varies less than does the social security replacement rate across countries and across groups of women whose husbands died at different ages. The median widow’s household size-adjusted income in the year after her husband’s death relative to the year before is about 0.9 when his death occurs at older ages and about 0.8 at younger ages in all four countries. JEL Classification: I 31, J 14, J 26, J 32 1. Introduction A mixture of private and public institutions offset the risk of economic loss to women following the death of their husband. Yet, most cross-national studies focus on the income protection against this loss provided by a single proSchmollers Jahrbuch 123 (2003) 1 Schmollers Jahrbuch 123 (2003), 151–162 Duncker & Humblot, Berlin * Multinational projects that attempt to document economic outcomes using complex data sets must rely on help from scholars familiar with the institutions and data of their home countries. This paper is no exception. We are indebted to Stephen Jenkins, Elena Bardasi, John Rigg, Nick Buck of Essex University, and Andrew Henley of the University of Wales Aberystwyth for their work in preparing the data needed to create the BHPS equivalent file. We thank Markus Grabka, John Haisken-DeNew, and Joachim Frick of the DIW Berlin for their continuing help with the GSOEP files. We thank Tecla Loup and Yeong-Song Kim for their help with the PSID files. We also thank Paola Valenti and Nigar Nargis who provided able research assistance and Florence Allen who prepared the manuscript. Funding for this project came from the United States Social Security Administration through the University of Michigan Retirement Research Center. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 152 Richard V. Burkhauser, Phil Giles, Dean R. Lillard, Johannes Schwarze gram: social security. 1 Lack of data has required these comparisons be made either with cross-sectional data or even worse, using a hypothetical average of a worker’s earnings history and his widow’s subsequent social security benefit across various countries. Here we use a newly expanded source of cross-national panel data, the Cross-National Equivalent File (CNEF), to trace the economic well-being of women as they transition from wife to widow. (For a fuller discussion of these data see Burkhauser et al. 2001). We take full advantage of the CNEF by using an event history based longitudinal sample design to examine the economic well-being of a women’s household prior to and following the death of her spouse. Our sample consists of the households of 450 German, 244 British, 627 Canadian, and 591 United States women whose husband died sometime during the life of the panel. 2 To measure changes in her household’s economic well-being, we track all sources of household income. 3 Because the members of her household will change over time (e.g. her husband dies, she moves in with relatives, etc.) we follow the women from the year just before to the year just after her husband dies. Since a husband’s death is a relatively rare event in each of our country data sets, in each of our country samples we pool our women by their husband’s age at death across all years available in each of the country panels. To do so, we realign our calendar year data into an event history framework where we label the year of her husband’s death as year (t). Our data include income years 1976 through 1997 for the United States Panel of Study Income Dynamics (PSID), 1984 through 2000 for the German Socio-Economic Panel (GSOEP), including respondents from the eastern states, 1991 through 1999 for the British Household Panel Study (BHPS), and 1993 through 1999 for the Canadian Survey of Labour and Income Dynamics (SLID). 2. How the Economic Well-Being of Women Change Following Their Husband’s Death Table 1 shows how the mean household income of women and its sources change from the year before to the year after the death of their husband within four age categories across four countries. 4 The sign (+, 0, –) in each cell indiSchmollers Jahrbuch 123 (2003) 1 1In this paper we use the term social security programs to refer to public, industrywide, insurance-based retirement, disability, and survivors programs where benefits are based on the worker’s earnings record. In some countries, social security programs could also include unemployment insurance, child benefits, etc. See Appendix Table 1A in Burkhauser et al. (2002) for a full description of the programs we include in our analysis. 2See Burkhauser et al. (2002) for a full discussion of the data used in this paper. 3The sources of income in each of these categories are described in more detail in Appendix Table 1A of Burkhauser et al. (2002). OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 Changes in the Economic Well-Being of Widows 153 Schmollers Jahrbuch 123 (2003) 1 cates whether income from that source increases, stays the same, or declines after their husband’s death. Asterisks indicate the income categories that account for the largest fraction of the total of the income increases and income decreases across all income categories. 5 As can be seen in Table 1, when a husband dies between ages 25 and 49, total household income declines and his lost labor earnings has the largest negative effect on household income in all four countries. But the importance of offsetting increases varies across the countries. The widow’s labor earnings increase in Germany and Great Britain, remains at about the same level in Canada, and falls in the United States. Income from the labor earnings of other household members increases in all countries. But the primary source of increased income comes from social security benefits in the United States and Great Britain, from the widow’s labor earnings in Germany and from private transfers in Canada. A husband’s death between ages 50 and 61 also reduces household income and the major loss in household income is due to his lost labor earnings. Now, however, the widow’s labor earnings either remain the same or fall. Others’ labor income falls in Germany, and rises in the other three countries. Increases in social security benefits offset lost labor earnings in all countries, but only in Great Britain is it most important. In the United States, increases in asset income (e.g. life insurance) are most important, while in Germany and Canada, it is a reduction in tax payments. While household income also declines after a husband’s death between ages 62 and 69, the husband’s lost labor earnings are primary only in the United States and Great Britain. In Germany and Canada the husband’s labor earnings have already declined substantially prior to his death and were not a factor. In Canada, declines in private pension benefits are most important. In Germany, declines in social security benefits are most important. The most important source of offsetting increases in household income also varies across the four countries. In the United States, Great Britain, and Canada, reduced tax payments are most important. In Germany, it is the increased labor earnings of the widow. 4Appendix Tables 6A, 7A, 8A, and 9A, in Burkhauser et al. (2002) provide the widow’s mean household income and its sources for three year before and three years after the death of her husband for the United States (Table 6A), Germany (Table 7A), Great Britain (Table 8A), and Canada (Table 9A). We use mean values unadjusted for household size to focus on the changes in the relative importance of various sources of income following her husband’s death. 5To decide whether to assign a (+, 0, -) to each income source, we first separately sum increases and decreases in mean household income across all sources after death. Each increase (decrease) in mean household income is calculated as a fraction of the total increase (decrease). For increases (decreases) greater than 10 percent a “+” (”-”) was assigned. A “0” was entered in all other cells. A decrease in tax obligations is treated as an increase in income. OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 154 Richard V. Burkhauser, Phil Giles, Dean R. Lillard, Johannes Schwarze Table 1 Direction of Change in Widows’ Mean Post-Government Household Income by Source After the Death of Their Husband by Country and Age of Husband at Death Income Source Aged 25 through 49 Aged 50 through 61 United States Germany Great Britain Canada United States Germany Great Britain Canada Total Post-Government -------- Private Sources Husband’s Labor Income -* -* -* -* -* -* -* -* Widow’s Labor Income - +* + 0 0 - 0 - Others’ Labor Income +++++-++ Private Transfers 0 0 0 +* 0 + 0 - Private Pensions ++++0+0Assets + + 0 0 +* 0 + + Public Source Transfers 0 - - + 0 0 - 0 Social Security +* + +* + + + +* + Taxes 0 + + + + +* + +* Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 Changes in the Economic Well-Being of Widows 155 Income Source Aged 62 through 69 Aged 70 and over United States Germany Great Britain Canada United States Germany Great Britain Canada Total Post-Government -------- Private Sources Husband’s Labor Income -* 0 -* 0 - - - 0 Widow’s Labor Income - +* - 0 0 + 0 0 Others’ Labor Income + + - + +* 0 0 +* Private Transfers 0000-000 Private Pensions - 0 - -* - - - - Assets 0 + 0 - - + - 0 Public Source Transfers 0 0 - 0 0 0 - + Social Security - -* + - -* -* -* -* Taxes +* + +* +* + +* +* 0 Source: Based on yearly information from the year prior to the year following the death of a woman’s husband. Authors’ calculations from the Panel Study of Income Dynamics 1976–1997, German Socio-Economic Panel 1984–2000, British Household Panel Study 1991–1999, and Survey of Labour and Income Dynamics 1993–1999. Notes: Sign is negative if change in sources accounted for at least a 10 percent decline in post-government income. Sign is positive if change in source accounted for at least 10 percent increase in post-government income. Sign is zero if change is less than 10 percent. Asterisks indicate the income categories that account for the largest fraction of the total of the income increases and income decreases across all income categories. This is an unbalanced panel. Sample size varies across years. See Appendix Tables 10A and 11A of Burkhauser, Giles, Lillard and Schwarze (2002). A detailed list of the income types included in each category and all mean values can be found in of Burkhauser, Giles, Lillard, and Schwarze (2002). No adjustments are made for household size. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 156 Richard V. Burkhauser, Phil Giles, Dean R. Lillard, Johannes Schwarze A husband’s death at ages 70 and above also reduces household income but the loss of his labor earnings is even less important. In all four countries, declines in social security income are most important. In Germany and Great Britain, reduced tax payments are the most important offset of this decline. In the United States and Canada, the increased labor earnings of other household members are most important. Note, however, that this could relate to a change in household composition rather than an increase in the labor earnings of those who were in the household prior to the husband’s death. Table 1 shows that total household income unadjusted for change in household size falls in all four countries following a husband’s death. Moreover, the pattern of loses and gains are closer across countries than across age groups within each country. None-the-less, there are some important differences in the primary source of offsetting income gains across countries. However, such differences are critical from a policy perspective only if they lead to important differences in the relative change in economic well-being of surviving household members across these countries. Before determining if this is the case, we discuss how adjustments to household income that account for differences in household size will affect measures of the change in the economic wellbeing of women following the death their husband in all countries. 2.1 Evaluating the economic well-being of individuals in households of different size Table 1 reports changes in the mean post-government income of widow’s households before and after the death of their husband. These values were not adjusted for household size. A large literature exists detailing the problems associated with measuring the economic well-being of individuals who live in households of different size. (See Moon/Smolensky, 1977; Burkhauser/ Smeeding/Merz, 1996 for examples of this literature). Simply comparing a women’s net-of-tax total household income, unadjusted for household size, before and after the death of her husband, as we did in Table 1, effectively assumes perfect returns to scale in household production. Alternatively, assigning each survivor a per capita share of net-of-tax household income, effectively assumes no returns to scale. Buhmann et al. (1988) propose a formula that accommodates these two extreme assumptions. Their formula is: E  D = Se  1  where an individual’s equivalent income (E) equals total household income (D) divided by household size (S) raised to the power (e). Assumptions about economies of scale in household production or consumption are captured in Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 Changes in the Economic Well-Being of Widows 157 the value one adopts for (e). At one extreme, when (e) equals 1, no economies of scale exist. Operationally, per capita income is assigned to each person in the household. At the other extreme, when (e) equals zero, economies of scale are perfect. Operationally, each person is assigned equivalent income exactly equal to household income. Burkhauser/Smeeding/Merz (1996) show the sensitivity of income inequality and poverty measures to variations in the value of (e) but recognize that economic theory does not suggest a particular value. They point out, however, that a common value used in the literature is (e) equals 0.5. In Table 2, we use the above formula to adjust post-government household income values in the year prior to the husband’s death (t– 1) and the year following the husband’s death (t+ 1) using alternative values of (e). As can be seen in Table 2, higher values of (e) reduce the post-government household income of each individual in the household. More importantly for our purpose, the ratio of mean household size-adjusted post-government income in (t+1) to mean household size-adjusted post-government income in (t– 1) varies dramatically with the choice of (e). When (e) equals zero the widow’s household size-adjusted income falls in all countries following her husband’s death. This result is found in Table 1. At the other extreme, (e) equals 1, the widow’s household size-adjusted income rises for most ages of their husband’s death in all countries. Table 2 shows that differences in the household size-adjusted income ratios across values of (e) are in general greater than the differences across age groups within a country or within an age group across countries. Burkhauser/ Smeeding/Merz (1996) have shown that the choice of (e) does not have a substantial effect when one compares income distributions or poverty rates across countries. However, they note that the choice of (e) can have dramatic effects on the demographic characteristics of households that are found in the lower end of the income distribution. For instance, because older persons live in smaller households, the smaller the value of (e) (i.e. the higher the assumed returns to scale) used to estimate equivalized income, the older will be the poverty population. Tables 2 provide evidence of a corollary to this rule. The smaller the value of (e) used to estimate equivalized income, following the death of her husband, the greater is the drop in measured economic well-being of the widow’s household. 2.2. Comparing social security and household size-adjusted replacement rates across countries We follow Burkhauser/Smeeding/Merz (1996) and use an (e) value of 0.5 in our analysis of the change in widows’ household income following the death of their husband. Most cross-national comparisons of how household Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15 158 Richard V. Burkhauser, Phil Giles, Dean R. Lillard, Johannes Schwarze Table 2: Widows’ Mean Post-Government Household Size-Adjusted Income Before and After the Death of Their Husband by Country and Various Returns-to-Scale Values Country Scale Value Aged 25 through 49 Aged 50 through 61 Aged 62 through 69 Aged 70 and over t ÿ 1t  1 Ratio (t  1 = t ÿ 1) t ÿ 1t  1 Ratio (t  1 = t ÿ 1) t ÿ 1t  1 Ratio (t  1 = t ÿ 1) t ÿ 1t  1 Ratio (t  1 = t ÿ 1) United States a) e= 0.0 42021 12711 .30 41570 21744 .52 29176 14998 .51 25973 14111 .54 e= 0.5 22565 11664 .52 24823 17923 .72 19372 14523 .75 17578 14798 .84 e= 1.0 12499 7865 .63 15192 14908 .98 13164 13596 1.03 12034 14271 1.19 Germany b) e= 0.0 54847 41813 .76 57811 41587 .72 49395 36240 .73 41767 31382 .75 e= 0.5 30053 26924 .90 35361 32612 .92 32959 32782 .99 28383 29484 1.04 e= 1.0 16921 18743 1.11 21985 26593 1.21 22221 30472 1.37 19457 28280 1.45 Great Britain c) e= 0.0 19776 17699 .89 20435 16668 .82 18397 10403 .57 15605 9684 .62 e= 0.5 11073 11844 1.07 12436 12448 1.00 12189 9122 .75 10714 8726 .81 e= 1.0 6362 8362 1.31 7824 10049 1.28 8170 8268 1.01 7428 8128 1.09 Canada d) e= 0.0 50648 35228 .70 48768 31198 .64 32796 24258 .74 30245 22240 .74 e= 0.5 27344 23498 .86 29313 22930 .78 21727 20674 .95 20760 19990 .96 e= 1.0 15285 16758 1.10 18028 18298 1.01 14647 18674 1.27 14378 18787 1.31 Source: Authors’ calculations from the Panel Study of Income Dynamics 1976–1997, German Socio-Economic Panel 1984–2000, British Household Panel Study 1991–1999, and Survey of Labour and Income Dynamics 1993–1999. Notes: This is an unbalanced panel. Sample size varies across years. See Appendix Tables 10A and 11A of Burkhauser, Giles, Lillard, and Schwarze (2002). a) Constant 1996 United States dollars. b) Constant 1996 German marks. c) Constant 1996 British pounds. d) Constant 1996 Canadian dollars. Schmollers Jahrbuch 123 (2003) 1 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.123.1.151 | Generated on 2023-04-04 12:31:15