Women’s Income and Marriage Markets in the United States Tables

B.2 Variables Variable Date of first husband’s death Date of pension application Source Union Army database (Fogel et al 2000) Widows’ pension database (Salisbury) Date of pension receipt Widows’ pension database Date of remarriage Widows’ pension database Date of death Widows’ pension database Age at widowhood Widows’ pension database Number of children Union Army database Potential minor pension Union Army database No pension attorney Widows’ pension database Washignton pension attorney First husband: height First husband: log occupational wage Widows’ pension database First husband: age at death County of residence County male-to-female ratio County percent urban County population density Name homogeneity index Last name: mean occupational income Last mean: mean immigrant status Last name: mean literacy Union Army database Union Army database; Preston and Haines (1991); United States Census of Agriculture (1900) Union Army database Widows’ pension database Haines and ICPSR (2010) Haines and ICPSR (2010) Haines and ICPSR (2010) Ruggles et al (2010); Atack and Bateman (1992) Ruggles et al (2010); Preston and Haines (1991); United States Census of Agriculture (1900) Ruggles et al (2010) Ruggles et al (2010) Notes Based on dependents’ pension applications or military death records Date at which widow filled out pension declaration form; if missing, date at which pension application received by pension bureau Date of issuance on pension certificate; if missing, date of pension approval on pension brief Based on marriage certificates or affadavits rendered in support of minors’ pension application or application for widow to be restored to the pension rolls under a later act. Based on pension drop cards, or death records filed in support of minors’ pension application. Deduced from widow’s first pension declaration, in which age and date of application are both provided. Equal to number of children under the age of 16 when widow first filed for pension. Calculated as $8/mo until youngest child turns 16, or $8/mo plus $2/mo for each child under 16 if widowed after July 25, 1866. Equal to one if the widow did not hire an attorney at the time of filing her first claim Equal to one if the widow first hired an attorney from a Washington firm at the time of filing her first claim Soldier’s height at enlistment Based on soldier’s occupation at enlistment Based on implied birth year from age at enlistment County listed on first pension application form Weighted mean of male-to-female ratio in 1860, 1870 and/or 1880, depending on date of application. See above. See ablve. Herfindahl index of concentration of unique spellings within phonetic surname groups among household heads in 1 percent IPUMS sample from 1860-1880. Phonetic groups created using NYIIS algorithm. Mean occupation status of household head, calculated using 1900 wage distribution, by phonetic name group in IPUMS 1 percent sample from 1860-1880. Mean literacy of household head by phonetic name group in IPUMS 1 percent sample from 1860-1880. Mean immigrant status of household head by phonetic name group in IPUMS 1 percent sample from 1860-1880. Last name: residence mean farm Literacy Immigrant stats Ruggles et al (2010) Linked widow sample (Salisbury); ancestry.com Linked widow sample; ancestry.com Mean farm status of household head by phonetic name group in IPUMS 1 percent sample from 18601880. Literate in census of 1870 or 1880 Immigrant in census of 1870 or 1880 Tables and Figures Table 1. Source of Information on Marital Status in Pension File Data, by Pension Status Group: Source of Information: Without General Law Pension With General Law Pension Remarried Minor's application after remarriage Widow's application under late law Other communication N, remarried 64% 30% 6% 84% 11% 5% 77 141 Never Remarried Dropped from pension rolls Minor's application after death Widow's application under law of 1890 Other communication N, never remarried 0% 52% 48% 0% 81% 13% 0% 6% 26 153 Unknown N, unknown 14 87 N, total 117 381 This table summarizes the sources of information about widows' remarriage status, separately by pension status. Sample includes women widowed by 1880 and who applied for a pension within five years of widowhood. Table 2. Summary Statistics from Pension File Data Mean Median SD Min Max N 0.817 0.674 0.865 2.280 1.000 0.285 1.000 0.906 0.387 0.958 0.342 4.583 0.000 0.014 0.000 0.112 1.000 5.767 1.000 50.500 498 498 498 431 Age widowed Age at first marriage Age at remarriage Number of children (first marriage) Husband died during war years 31.867 20.838 32.080 2.566 0.721 30.000 20.000 31.000 2.000 1.000 9.410 5.025 7.641 2.240 0.449 15.000 9.000 18.000 0.000 0.000 73.000 48.000 65.000 13.000 1.000 487 474 213 498 498 Remarried Remarried without pension Time to Remarriage: All Remarried with pending claim Remarried after pension Time to remarriage following pension 0.549 0.181 1.000 0.000 0.498 0.386 0.000 0.000 1.000 1.000 397 425 4.305 2.392 5.351 3.911 3.348 1.838 4.096 2.573 3.708 1.866 4.038 3.956 0.230 0.230 0.915 0.047 26.036 8.778 26.036 25.463 215 76 139 134 1854.6 1865.6 1868.9 1866.2 1869.0 1856 1864 1867 1864 1866 7.820 4.526 5.124 5.044 9.145 1822 1862 1863 1862 1862 1879 1879 1889 1883 1928 489 497 215 498 451 0.129 0.303 0.410 0.090 0.023 0.041 0.002 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.336 0.460 0.492 0.287 0.149 0.198 0.045 0.000 0.045 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 0.000 1.000 488 488 488 488 488 488 488 488 488 Variable: Pension Variables Applied within 1 year Time to first application General law claim accepted Processing time of accepted gen law claim Age/Marriage Variables Calendar Years First marriage Widowhood Remarriage Pension application Pension certificate Region of Residence New England Mid Atlantic East North Central West North Central South Atlantic East South Central West South Central Mountain Pacific Sample includes women who were widowed before 1880 and who applied for a pension within five years of widowhood. Sample drawn from Union Army Database (Fogel et al 2000). Data collected from Civil War pension files at the National Archives in Washington, DC. Table 3. Linkage Rates to 1870 and 1880 Census. All Married Unmarried Linkage Rate: 1870 census N 0.576 408 0.686 159 0.628 164 0.271 85 0.377 53 1880 census N 0.606 467 0.681 204 0.764 165 0.184 98 0.262 61 0.711 235 0.523 111 0.879 124 1.000 23 1.000 20 0.629 283 0.435 131 0.805 149 1.000 18 1.000 16 Group: Fraction linkable through children 1870 census N 1880 census N Marital status unknown has children from all 1st marriage Sample includes all women widowed by relevant census year and who are not known to have died by this year. A woman is considered linkable through children if she is living with a child from her first marriage who has the same last name as her first husband. Table 4. Estimated Fraction of Widows Observed in Union Army Data Panel A: Distribution of Observations by Date of Death and Marital Status Category: N: Total 39,341 married Total unmarried 3,102 714 Died before 1880 (N=7,953) Dead during war (N=5,777) unknown married unmarried unknown 3,777 1,755 654 3,446 Date of death unknown (N=11,552) married unmarried unknown 572 46 10,934 Panel B: Lower-Bound Estimates of the Fraction of Widows Observed in UA Data Assumption Implied fraction of widows that appear in the UA data Reference group: Died before 1880 Died during war A1: Everyone with missing death date died before 1880, and everyone with missing marital status was married. 16.9% A2: Everyone with missing death date died before 1880, and 63.4% of men with missing marital status were married 23.9% A3: 16% casualty rate, and everyone with missing marital status was married. 27.9% A4: 16% casualy rate, and 52.6% of soldiers with missing marital status were married. 45.7% This table provides lower-boumd estimates of the fraction Union Army widows who filed pension applications, using different assumptions about missing data. Marriage rates in A2 and A4 are imputed marriage rates for the full sample and the sample killed in the war, respectively. These are based on marriage probabilities imputed from a regression of marital status on age, state, and occupational class using the 1860 1 percent IPUMS sample. Table 5. Characteristics of Wives Identified in 1860 Census Links. Soldiers who died during the war t test for equality of means Wife's age Soldier's age Wife literate Soldier literate Wife immigrant Soldier immigrant HH head personal property ($1,000) HH head real estate ($1,000) Soldier farmer Soldier professional or proprieter Solder skilled worker Soldier laborer Solder no occupation Urban county NE MW SO N Mean: Wife observed in pension data Mean: Wife not observed in pension data 29.4890 (8.5129) 33.0864 (8.4058) 0.9109 (0.285) 0.9153 (0.2786) 0.1359 (0.3429) 0.1542 (0.3614) 0.1708 (0.2903) 0.4900 (1.0522) 0.3051 (0.4608) 0.0336 (0.1803) 0.2088 (0.4067) 0.2146 (0.4108) 0.0073 (0.0852) 0.1441 (0.2455) 0.4131 (0.4928) 0.5153 (0.5001) 0.0657 (0.2479) 26.8571 (8.1493) 26.5940 (8.2747) 0.9323 (0.2521) 0.9699 (0.1714) 0.2045 (0.4049) 0.1805 (0.386) 0.6727 (2.6348) 1.4727 (4.3341) 0.3083 (0.4635) 0.0677 (0.2521) 0.0752 (0.2647) 0.2632 (0.442) 0.0075 (0.0867) 0.1870 (0.2938) 0.3534 (0.4798) 0.6165 (0.4881) 0.0301 (0.1714) 685 133 OLS regression (1) - (2) 2.6320*** 6.4920*** -0.0200 -0.0560** -0.0680** -0.0280 -0.5000*** -0.9840*** Dependent variable=1 if wife observed in pension data -0.0096*** (0.002) 0.0193*** (0.002) 0.0064 (0.048) -0.0845* (0.050) -0.0880* (0.051) 0.0107 (0.048) -0.0226 (0.014) -0.0205*** (0.008) -0.0040 -0.0360* 0.1320*** -0.0480 -0.0002 -0.0440* 0.0600 -0.1000** 0.0360 -0.0518 (0.067) 0.1106*** (0.035) 0.0064 (0.032) 0.1013 (0.142) -0.1707*** (0.057) 0.0000 (0.000) -0.0506* (0.028) 0.0649 (0.057) 801 Sample of soldiers in UA data who died during the war, are linked to the 1860 census, and who appear to be married based on the composition of their household in 1860. Regression model includes a constant, and R2=0.175 Table 6. Determinants of the Hazard Rate of Remarriage and Pension Receipt (1) Outcome: Effect of pension Remarriage (2) Pension -0.1923 (0.1602) Number of Children Year of widowhood Time to pension application Potential minor pension at widowhood No pension attorney Washington pension attorney First husband: age at death First husband: log occupational wage First husband: height (feet) County male-to-female ratio County percent urban County population density Mid Atlantic East North Central West North Central South [2,3) [3,4) [4,5) [5,6) [6,7) [7,8) [8. ") vlow (constant in columns 1-2) 1.5122 (0.4158) 1.5575 (0.4432) 1.5361*** (0.4538) 1.2324*** (0.3906) 0.9929*** (0.3388) 0.5071** (0.2172) 0.6036** (0.2489) 0.0808*** (0.0256) -2.4050*** (0.2192) (3) Pension -0.4867** (0.1935) Age at widowhood ! for years: [1,2) Remarriage 1.0519*** (0.1409) 0.6498*** (0.1191) 0.4793*** (0.1154) 0.3610*** (0.1091) 0.2130** (0.0890) 0.1742** (0.0885) 0.1464* (0.0855) 0.2443*** (0.0526) -0.4650*** (0.0891) Pension -0.5361** (0.2213) -0.0967*** (0.0157) -0.1327** (0.0605) -0.0463* (0.0255) -0.0346 (0.1206) 0.1149 (0.1613) 0.1884 (0.3196) 0.0358 (0.2125) 0.0278* (0.0153) 0.1280 (0.4251) -0.7049** (0.3543) 2.1664** (1.0528) 0.2710 (0.3611) -0.0310 (0.0208) 0.2804 (0.2674) 0.2525 (0.2608) 0.5323 (0.3622) -0.4368 (0.4183) 0.0065 (0.0103) -0.0359 (0.0364) -0.0694*** (0.0169) -0.0998 (0.0967) 0.1402 (0.1174) 0.2669 (0.2484) 0.0859 (0.1700) -0.0132 (0.0126) -0.1646 (0.2881) 0.1418 (0.2696) -0.0439 (1.1809) 0.3678 (0.2828) -0.0177 (0.0112) -0.7319*** (0.2078) -0.6505*** (0.2149) -0.3906 (0.2943) -0.7574** (0.2971) -0.1147*** (0.0213) -0.1600** (0.0784) -0.0679** (0.0312) -0.0504 (0.1460) 0.1184 (0.1902) 0.3276 (0.3516) -0.0642 (0.2986) 0.0248 (0.0173) 0.2955 (0.4562) -0.9165** (0.4221) 2.2800* (1.1672) 0.3360 (0.3927) -0.0322 (0.0229) 0.4838 (0.3866) 0.4748 (0.3659) 0.9626* (0.5185) -0.3212 (0.5355) 0.0065 (0.0105) -0.0359 (0.0371) -0.0697*** (0.0170) -0.1032 (0.0968) 0.1407 (0.1188) 0.2680 (0.2491) 0.0851 (0.1730) -0.0133 (0.0128) -0.1629 (0.2951) 0.1417 (0.3077) -0.0377 (0.9914) 0.3681 (0.2723) -0.0177 (0.0113) -0.7345*** (0.2088) -0.6539*** (0.2127) -0.3948 (0.2990) -0.7615** (0.3008) 1.8779 (0.5721) 2.3504* (0.7447) 2.7242*** (0.9048) 2.5961*** (0.9380) 1.9935** (0.8000) 1.3197** (0.6135) 1.2622** (0.6110) 0.2095*** (0.0781) 0.3372 (3.4815) 1.1925*** (0.1748) 0.8273*** (0.1708) 0.7398*** (0.1996) 0.7981*** (0.2683) 0.3428* (0.1837) 0.3158 (0.1923) 0.3824 (0.2329) 0.5012*** (0.1412) 0.6750 (2.6406) 3.4782 (1.8574) 5.9411 (4.3221) 7.7541 (6.2593) 7.8301 (6.6760) 6.2107 (5.5455) 4.2473 (4.0310) 4.1337 (3.9533) 0.7081 (0.6612) -0.2168 (3.7550) 3.3428 (3.4592) 1.1982*** (0.1767) 0.8340*** (0.1736) 0.7470*** (0.2030) 0.8094*** (0.2740) 0.3480* (0.1871) 0.3211 (0.1970) 0.3885 (0.2385) 0.5099*** (0.1456) 0.6105 (2.6362) 0.7443 (2.7011) vhigh #1 0.5817 (0.7635) 0.3364 (0.7946) 0.0589 (0.1364) 0.0229 (0.1043) #2 #3 #4 Log Likelihood Observations Remarriage -1377.486 482 -1141.420 429 -1140.0844 429 Hazard coefficients are reported. Sample: women who applied for a pension within five years of husband's death. Column (3) includes a correction for correlated unobserved heterogeneity, and does not include a constant as this is not identified separately from one of the mass points in the distribution of the unobserved heterogeneity terms; columns (1) and (2) make no such adjustment, and include a constant. Age at widowhood and all widows' pension variables (including county of residence) are taken from the pension file data collected by the author. First husband characteristics come from the UA data and are based on enlistment variables; occupational wages measured using 1900 occupational wage distribution assigned to 1950 occupational codes, with an imputed wage for farmers (Preston and Haines 1992; Abramitzky Boustan and Eriksson 2010; Olivetti and Paserman 2012). County-level variables are taken at the time of pension application; they are the weighted average of these variables at the decadal censuses preceding and following the date of pension application (Haines and ICPSR 2010). On the time interval [0,1), the hazard rate of both risks is normalized to one (this is necessary because I include a constant in the model). The variables vlow and vhigh are the two mass points in the distributions of vm and vp. The variables #1-#4 are the estimated probability of each unobserved heterogeneity event. Table 7. First Stage Results: Effect of Name Homogeneity Index on Pension VARIABLES Name homogeneity index Age at widowhood Number of children Year of widowhood Time to pension application Potential minor pension No pension attorney Washignton pension attorney First husband: height First husband: log occupational wage First husband: age at death County male-to-female ratio County percent urban County population density Last name: mean occupational income Last mean: mean immigrant status Last name: mean literacy Constant Observations R-squared First stage F statistic (3) pen1 (5) pen2 (7) pen3 (9) pen4 (11) pen5 0.1660* (0.091) -0.0081* (0.005) 0.0025 (0.015) -0.0297*** (0.007) -0.0777** (0.031) -0.0000 (0.000) 0.0628 (0.103) -0.1599** (0.069) -0.1080 (0.111) 0.0792 (0.141) 0.0063 (0.005) 0.4369 (0.455) 0.1287 (0.122) -0.0000** (0.000) 0.2845 (0.440) -0.0097 (0.150) -0.3253 (0.422) 54.2775*** (13.329) 0.2014** (0.087) 0.0029 (0.004) -0.0107 (0.014) -0.0345*** (0.007) -0.1107*** (0.029) 0.0001 (0.000) 0.0791 (0.097) -0.0647 (0.066) 0.0285 (0.108) 0.2034 (0.135) 0.0003 (0.005) 0.0476 (0.431) 0.1933* (0.117) -0.0000 (0.000) 0.3510 (0.421) -0.1688 (0.143) -0.3456 (0.404) 61.6002*** (12.627) 0.1670** (0.082) 0.0033 (0.004) -0.0095 (0.013) -0.0393*** (0.006) -0.0882*** (0.027) 0.0001* (0.000) -0.0060 (0.090) -0.0504 (0.062) 0.0704 (0.103) 0.2714** (0.127) -0.0002 (0.004) -0.1043 (0.401) 0.1307 (0.109) -0.0000 (0.000) 0.3417 (0.391) -0.0866 (0.134) -0.4106 (0.380) 70.0858*** (11.748) 0.1933** (0.077) 0.0009 (0.004) -0.0098 (0.012) -0.0393*** (0.006) -0.0653** (0.025) 0.0000 (0.000) -0.0056 (0.084) 0.0092 (0.059) 0.0238 (0.099) 0.0027 (0.120) 0.0016 (0.004) -0.1993 (0.374) -0.0260 (0.104) 0.0000 (0.000) 0.3826 (0.366) -0.0180 (0.127) -0.1187 (0.358) 71.6802*** (11.155) 0.2061*** (0.074) 0.0030 (0.004) -0.0145 (0.012) -0.0387*** (0.006) -0.0737*** (0.025) 0.0000 (0.000) -0.0229 (0.082) 0.0372 (0.057) -0.0288 (0.096) -0.0079 (0.116) 0.0010 (0.004) -0.1620 (0.363) -0.0529 (0.101) 0.0000 (0.000) 0.3658 (0.361) -0.0476 (0.123) -0.1211 (0.347) 71.0150*** (10.834) 368 0.219 3.18 362 0.245 5.26 356 0.270 4.35 348 0.269 6.73 347 0.281 7.16 Name homogeneity index is Herfindahl index of unique surname spellings within phonetic name groups, calculated using household heads from IPUMS 1 percent samples, 1860-1880. See notes to table 6 for description of sample and remaining variables. Table 8. OLS and 2SLS Estimates of Relationship between Pension and Remarriage (1) (2) (3) (4) (5) Dependent variable: married w/in time frame (OLS) 1 year 2 years 3 years 4 years 5 years 1 year Pension granted w/in time frame -0.0473 (0.032) -0.0810* (0.049) -0.2182*** (0.057) -0.2736*** (0.063) -0.2564*** (0.066) -0.3180 (0.365) -0.8348 (0.518) -1.0244 (0.656) -1.0627* (0.557) -1.1018** (0.534) Age at widowhood -0.0039 (0.003) -0.0131 (0.009) 0.0040 (0.004) -0.0228 (0.018) -0.0000 (0.000) 0.0665 (0.061) 0.0073 (0.041) -0.0666 (0.066) 0.0090 (0.084) 0.0023 (0.003) 0.6539** (0.267) 0.0453 (0.073) -0.0000 (0.000) 0.0823 (0.228) 0.0279 (0.085) 0.0665 (0.165) -8.2092 (7.861) -0.0115*** (0.004) -0.0175 (0.013) 0.0032 (0.006) -0.0315 (0.027) -0.0000 (0.000) 0.0242 (0.088) 0.0346 (0.060) -0.0666 (0.098) -0.0151 (0.124) 0.0050 (0.004) 0.2972 (0.385) -0.0520 (0.107) 0.0000 (0.000) -0.1783 (0.331) -0.1311 (0.123) 0.2034 (0.239) -4.2195 (11.471) -0.0156*** (0.004) -0.0287** (0.014) -0.0082 (0.007) -0.0233 (0.029) 0.0000 (0.000) -0.0015 (0.095) -0.0060 (0.065) -0.0886 (0.110) 0.0678 (0.136) 0.0071 (0.005) 0.4341 (0.420) 0.0151 (0.117) -0.0000 (0.000) -0.4119 (0.361) -0.1325 (0.135) 0.3486 (0.261) 18.1355 (12.635) -0.0186*** (0.004) -0.0287** (0.014) -0.0123* (0.007) -0.0248 (0.030) -0.0000 (0.000) -0.0403 (0.097) 0.0029 (0.068) -0.1981* (0.116) -0.1829 (0.140) 0.0058 (0.005) 0.6191 (0.429) 0.0245 (0.121) -0.0000 (0.000) 0.0008 (0.370) -0.2075 (0.140) 0.2987 (0.267) 25.5147* (13.207) -0.0211*** (0.005) -0.0313** (0.015) -0.0137* (0.007) -0.0372 (0.030) -0.0000 (0.000) -0.0472 (0.098) 0.0105 (0.069) -0.2849** (0.116) 0.0091 (0.140) 0.0076 (0.005) 0.9918** (0.431) 0.0529 (0.122) -0.0000 (0.000) 0.1302 (0.377) -0.1505 (0.140) 0.0060 (0.268) 26.4345** (13.331) -0.0068 (0.004) -0.0119 (0.010) -0.0028 (0.011) -0.0467 (0.036) -0.0000 (0.000) 0.0857 (0.073) -0.0382 (0.074) -0.0868 (0.082) 0.0247 (0.096) 0.0043 (0.004) 0.8111** (0.340) 0.0813 (0.094) -0.0000 (0.000) 0.2760 (0.307) 0.0110 (0.100) 0.0083 (0.304) 3.5142 (20.909) -0.0099* (0.005) -0.0256 (0.018) -0.0208 (0.019) -0.1218* (0.069) 0.0000 (0.000) 0.0954 (0.125) -0.0131 (0.086) -0.0377 (0.131) 0.1148 (0.188) 0.0059 (0.006) 0.3744 (0.517) 0.0940 (0.174) -0.0000 (0.000) 0.1466 (0.532) -0.2973 (0.208) -0.0237 (0.516) 38.0825 (33.982) -0.0139** (0.006) -0.0354* (0.019) -0.0409 (0.026) -0.0892 (0.070) 0.0001 (0.000) -0.0041 (0.121) -0.0686 (0.089) -0.0389 (0.150) 0.2681 (0.237) 0.0079 (0.006) 0.3307 (0.543) 0.1253 (0.171) -0.0000 (0.000) -0.0387 (0.566) -0.2040 (0.201) -0.0970 (0.571) 76.3273 (47.071) -0.0202*** (0.005) -0.0353* (0.018) -0.0439* (0.022) -0.0711 (0.052) -0.0000 (0.000) -0.0459 (0.118) -0.0210 (0.083) -0.1797 (0.141) -0.1999 (0.168) 0.0086 (0.006) 0.4141 (0.539) -0.0021 (0.146) -0.0000 (0.000) 0.6612 (0.552) -0.2099 (0.180) 0.0570 (0.505) 81.3946** (41.250) -0.0204*** (0.006) -0.0433** (0.019) -0.0466** (0.021) -0.0963* (0.055) -0.0000 (0.000) -0.0571 (0.121) 0.0184 (0.087) -0.2991** (0.142) -0.0329 (0.173) 0.0093 (0.006) 0.8311 (0.545) 0.0185 (0.151) -0.0000 (0.000) 0.8199 (0.563) -0.2446 (0.189) -0.0104 (0.516) 84.6270** (39.485) 375 369 363 355 354 368 (-!, +!) 0.065 0.094 0.173 0.235 0.256 Time Frame Number of children Year of widowhood Time to pension application Potential minor pension No pension attorney Washignton pension attorney First husband: height First husband: log occupational wage 55 First husband: age at death County male-to-female ratio County percent urban County population density Last name: mean occupational income Last mean: mean immigrant status Last name: mean literacy Constant Observations AR 95% Confidence Region for pension effect R-squared (6) (7) (8) (9) Dependent variable: married w/in time frame (2SLS) 2 years 3 years 4 years 362 356 348 [-5.57, -0.057] [-20.61, 0 .031] [-4.390, -0.148] (10) 5 years 347 [ -3.686, -0.247] Instrument used in 2SLS specification is name homogeneity index. "Last name" variables are means by phonetic name group among household heads in IPUMS 1 percent sample from 1860-1880. AR 95% confidence region for pension effect is 95 percent confidence interval for the effect of the pension based on the Anderson-Rubin statistic, which is robust to weak instruments. See notes to table 6 for description of sample and other variables. Table 9. Sensitivity of Estimates to Sample Restrictions Model: Simple Full Panel A. Baseline Effect of Pension on Marriage Rate Log-Likelihood Observations -0.4867** (0.1935) -1141.420 429 -0.5361** (0.2213) -1140.0844 429 Panel B. Limit to time with minor children Effect of Pension on Marriage Rate Log-Likelihood Observations Effect of Pension on Marriage Rate Log-Likelihood Observations -0.2782 -0.2967 (0.2111) (0.2406) -943.556 -941.6477 339 339 Panel C. Limit to Information from General Law Pension Applications -0.3980* -0.5024** (0.2100) (0.2427) -1023.468 -1021.626 429 429 Panel D. Linked Only Effect of Pension on Marriage Rate Log-Likelihood Observations -0.4777** (0.2181) -896.135 302 -0.4915** (0.2428) -893.9608 302 Panel E. Linked Only: immigrant and literacy controls Effect of Pension on Marriage Rate Log-Likelihood Observations -0.4675** (0.2206) -892.845 302 -0.4597* (0.2449) -890.5271 302 Panel G. Husband died during war Effect of Pension on Marriage Rate Log-Likelihood Observations -0.4635** (0.2191) -902.990 348 -0.4751** (0.2320) 903.314 348 All specifications include the full set of controls from table 6; see notes to this table for explanation. The top panel replicates the baseline results. Panel B restricts the analysis to years in which the widow has a child under the age of 16. Panel C discards information that comes from applications under the law of March 3, 1901. Panel D restricts the sample to women who are successfully linked to the census of 1870 or 1880. Panel E poses a similar restriction, but includes imigrant and literacy controls available in the census data. Table 10. Pensions and the Timing of Remarriage: Widows Linked to Census Through Children Model: Year: =1 if pensioned w/in 5 years Age at widowhood Number of children Year of widowhood Time to pension application Potential minor's pension No pension attorney Washington pension attorney First husband: height First husband: log occupational income First husband: age at death Coutny male-to-female ratio County percent urban County population density Literate Immigrant (1) (2) 1870 OLS 1880 -0.2304** (0.112) -0.0226*** (0.007) -0.0743*** (0.024) -0.0130 (0.030) -0.0574 (0.061) 0.0000 (0.000) 0.0465 (0.133) -0.0014 (0.108) -0.0745 (0.191) 0.2998 (0.215) 0.0075 (0.009) 0.5028 (1.016) 0.0597 (0.228) -0.0000 (0.000) -0.1888* (0.102) 0.0499 (0.134) -0.1561 (0.098) -0.0255*** (0.007) -0.0356* (0.019) -0.0097 (0.009) -0.0690* (0.040) 0.0000 (0.000) -0.1756 (0.153) 0.0672 (0.108) 0.0061 (0.184) 0.2280 (0.195) 0.0111 (0.008) 0.8729 (0.549) 0.1430 (0.196) -0.0000 (0.000) 0.1791 (0.124) -0.0900 (0.117) 23.7010 (56.548) Observations R-squared First stage F statistic 147 0.328 (5) (6) Pooled Pooled 1870 2SLS 1880 -3.2024 (5.137) -0.0156 (0.021) -0.1451 (0.134) -0.0422 (0.088) -0.3429 (0.524) -0.0005 (0.001) 0.2369 (0.460) 0.2398 (0.492) -0.4418 (0.777) 0.7901 (1.020) 0.0128 (0.022) -4.0906 (8.294) -1.2462 (2.319) 0.0000 (0.000) -0.3085 (0.321) 0.5253 (0.879) -0.7379* (0.433) -0.0219*** (0.008) -0.0377* (0.020) -0.0338* (0.019) -0.1072** (0.052) 0.0000 (0.000) -0.2592 (0.164) 0.1187 (0.130) 0.0254 (0.195) 0.2782 (0.211) 0.0116 (0.008) 1.0195* (0.596) 0.1867 (0.203) -0.0000 (0.000) 0.1809 (0.139) -0.1118 (0.127) 16.5803 (16.361) -0.1511* (0.079) -0.0244*** (0.004) -0.0516*** (0.016) -0.0087 (0.008) -0.0693** (0.028) 0.0000 (0.000) -0.0592 (0.113) 0.0329 (0.078) -0.0114 (0.155) 0.2371 (0.179) 0.0089 (0.006) 0.8098* (0.452) 0.1198 (0.179) -0.0000* (0.000) -0.0370 (0.080) -0.0522 (0.063) 0.0209 (0.041) 15.0851 (14.425) 85.0484 (170.098) 61.4020* (36.525) -1.1183 (0.776) -0.0203*** (0.007) -0.0646*** (0.021) -0.0469 (0.031) -0.1444* (0.082) -0.0000 (0.000) -0.0664 (0.177) 0.1132 (0.123) -0.0158 (0.179) 0.3190 (0.330) 0.0105 (0.008) 0.5259 (0.634) -0.0013 (0.243) -0.0000 (0.000) -0.0697 (0.089) -0.0137 (0.102) 0.0457 (0.053) 86.9513 (58.893) 158 0.352 305 0.306 144 153 297 0.333 7.917 4.008 Year=1880 Constant (3) (4) Dependent variable: Remarried in Census Sample includes women linked to the census of 1870 or 1880 who are residing with a child from their first marriage who has kept his or her deceased father's last name. In columns (3) and (6), both census years are pooled and standard erorrs clustered by widow. See notes to table 6 for a detailed description of the variables and the notes to table 8 for a description of the instrument used in columns 4-6. Table 11. Pensions and Match Quality VARIABLES Remarried after pension Number of children Number of children X pension Potential minor pension Potential minor pension X pension Age in census Literate Immigrant Age at widowhood Age at remarriage First husband: age at death First husband: height First husband: log occupational wage County male-to-female ratio County population density County percent urban Census year = 1870 Constant Observations R-squared (1) Husband log occupational wage (2) Husband literate (3) Husband-wife age difference^2 (4) Husband present in household 0.0581 (0.091) -0.0432 (0.033) 0.0097 (0.042) 0.0700 (0.091) -0.0930 (0.110) 0.0095 (0.010) 0.1302* (0.072) 0.1496 (0.109) -0.0053 (0.010) 0.0031 (0.011) -0.0054 (0.005) 0.0720 (0.099) 0.0323 (0.207) -0.0199 (0.023) -0.0156 (0.023) 0.0737 (0.142) 0.1009 (0.094) 5.4848*** (1.376) -0.0048 (0.058) 0.0290 (0.033) -0.0468 (0.038) -0.0935 (0.094) 0.0657 (0.103) 0.0079 (0.006) 0.4953*** (0.109) 0.0190 (0.056) -0.0072 (0.008) -0.0013 (0.005) 0.0055 (0.003) 0.0581 (0.068) 0.0557 (0.119) -0.0053 (0.011) 0.0090 (0.011) 0.0335 (0.066) 0.0905* (0.050) -0.4433 (0.720) -36.8906 (77.837) -23.6364 (16.468) 7.4168 (19.893) 11.3229 (54.726) 4.7589 (62.245) 4.5264 (3.590) -5.2923 (44.771) -33.0041 (38.211) -1.2926 (5.888) -0.2236 (4.296) -0.0419 (3.319) -57.2034 (53.081) 280.3327** (116.120) -16.5068 (11.165) -4.9571 (11.063) -89.6366 (56.799) 57.5048 (42.502) -1,434.0604** (679.590) 0.0010 (0.094) 0.0234 (0.029) -0.0485 (0.037) -0.0477 (0.089) 0.1079 (0.107) -0.0059 (0.009) 0.0731 (0.065) 0.1550** (0.066) -0.0014 (0.011) -0.0003 (0.009) -0.0005 (0.005) -0.0020 (0.081) 0.2091 (0.176) -0.0033 (0.033) 0.0313 (0.026) -0.1637 (0.173) 0.0452 (0.095) -0.1901 (1.187) 191 0.145 197 0.485 197 0.111 221 0.097 Sample consists of remarried widows who linked to the census of 1870 or 1880. These census years are pooled, and stanard errors are clustered by widow. County-level variables from Haines and ICPSR (2010). See table 6 for description of other variables. Figure 1. Possible Outcomes for Widows in Sample Remarriage Death 11 (2%)  141 (28%) *  Accepted H HH j None  6 Apply 153 (31%) - Pension Decision @ @ R @ 6 ? Remarriage 61 ( 12%) Remarriage *  16 (3%) Rejected  HH j None H 15 (3%) Figure 2. Empirical Hazard Rate of Remarriage and Pension Decisions 0 .05 .1 .15 .2 Panel A. Empirical Hazard Rate of Remarriage By Pension Status 0 2 4 6 8 10 analysis time Pension claim not granted Pension claim granted ® 0 .2 .4 .6 .8 Panel B. Empirical Hazard Rate of Pension Decision 0 2 4 6 analysis time (years) 8 10 ® Panel A plots the nonparametric empirical hazard rate of remarriage, separated by pension status, and etimated using kernal method (STS package in STATA). Panel B does the same for the hazard rate of pension receipt. The picture is cut off at t=10 because the rate of remarriage cannot be estimated for women without pensions at t>10. Figure 3. Empirical Hazard under Different Sample Restrictions .15 .1 .05 0 0 .05 .1 .15 .2 Panel B. Empirical Rate of Remarriage: Using Only Information from Gen. Law Pension App. .2 Panel A. Empirical Rate of Remarriage: Years with Minor Child Only 0 2 4 6 8 10 0 2 4 analysis time 6 8 10 analysis time pension = 0 pension = 1 pension = 0 pension = 1 ® ® Panel C. Empirical Rate of Remarriage: Linked Only 0 0 .05 .05 .1 .1 .15 .15 .2 .2 Panel D. Empirical Rate of Remarriage: Widowed During War 0 2 4 6 8 10 0 analysis time pension = 0 2 4 6 8 10 analysis time pension = 1 pension = 0 pension = 1 ® ® Nonparametric empirical hazard rate of remarriage, separated by pension status, and etimated using kernal method (STS package in STATA). Pictures are truncated at t=10 for ease of comparison; the rate of remarriage among women without pensions cannot be estimated at t>10. See notes to table 9. Table B1. Determinants of Linkage to 1870 and 1880 Census among Widows (1) (2) -0.0026 (0.003) 0.0497*** (0.012) -0.0027 (0.003) 0.0493*** (0.012) -0.0051 (0.018) -0.0248 (0.040) 0.0832 (0.102) 0.0995 (0.070) -0.0021 (0.075) -0.0027 (0.003) 0.0493*** (0.012) -0.0051 (0.018) -0.0248 (0.040) 0.0832 (0.102) 0.0995 (0.070) -0.0021 (0.075) -0.0001 (0.005) 0.0610*** (0.013) -0.0019 (0.019) -0.0034 (0.042) 0.0738 (0.108) 0.0845 (0.072) -0.0144 (0.077) -0.1218 (0.138) -0.0045 (0.005) 0.0823 (0.108) 0.5346*** (0.091) 9.9560 (34.445) 9.9560 (34.445) 401 0.042 401 0.049 401 0.049 Dependent variable: Age at Widowhood Number of Children Year of Widowhood Time to first application No pension attorney Washington pension attorney Pensioned w/in 5 yrs (3) (4) Linked to 1870 census First husband: log occupational wage First husband: age at death First husband: height (5) (6) (7) -0.0014 (0.002) 0.0613*** (0.010) -0.0042 (0.003) 0.0601*** (0.010) 0.0157** (0.006) 0.0169 (0.028) 0.0325 (0.093) 0.0047 (0.064) 0.0675 (0.064) -0.0042 (0.003) 0.0601*** (0.010) 0.0157** (0.006) 0.0169 (0.028) 0.0325 (0.093) 0.0047 (0.064) 0.0675 (0.064) -0.0021 (0.004) 0.0627*** (0.011) 0.0131** (0.007) 0.0309 (0.029) 0.0272 (0.098) 0.0065 (0.067) 0.0735 (0.066) -0.0202 (0.123) -0.0032 (0.005) 0.1745* (0.099) 4.4928 (35.956) -0.0000 (0.005) 0.0591*** (0.014) -0.0004 (0.019) -0.0122 (0.042) 0.0748 (0.108) 0.0864 (0.073) -0.0071 (0.077) -0.1332 (0.139) -0.0043 (0.005) 0.0518 (0.110) 0.0533 (0.055) 0.1918 (0.120) 0.0000 (0.000) 1.7777 (35.989) 0.5014*** (0.077) -28.7767** (11.379) -28.7767** (11.379) -24.8121** (12.310) -0.0018 (0.004) 0.0615*** (0.011) 0.0137** (0.007) 0.0282 (0.029) 0.0242 (0.098) 0.0125 (0.067) 0.0779 (0.066) -0.0328 (0.124) -0.0032 (0.005) 0.1420 (0.100) 0.0639 (0.049) 0.2060* (0.111) 0.0000 (0.000) -25.7085** (12.288) 376 0.067 376 0.074 458 0.074 458 0.093 458 0.093 426 0.102 426 0.111 Region of enlistment: Midwest Region of enlistment: South Region of enlistment: West Constant Observations R-squared (8) (9) Linked to 1880 census (10) Age at widowhod, time to first application, and all widows' pension variables are taken from the pension file data collected by the author. The remaining variables are from the Union Army database. First husband characteristics based on enlistment variables; occupational wages measured using 1900 occupational wage distribution assigned to 1950 occupational codes, with an imputed wage for farmers (Preston and Haines 1992; Abramitzky Boustan and Eriksson 2010; Olivetti and Paserman 2012). Sample includes women widowed by the relative census year and who are not known to have died by this year. Table B2. Sample Means: Linked Widows Data versus Nationally Representative IPUMS Data Linked widows data Married Women IPUMS IPUMS age adjusted Unmarried Women (16+) Linked widows data IPUMS IPUMS age adjusted Female Household Heads Linked widows data IPUMS IPUMS age adjusted 32.41 5.58 0.00 0.84 0.08 75.96 8.17 38.75 578.84 0.88 1204.18 517.04 0.18 0.06 1.11 0.36 0.62 0.02 0.01 36.67 5.47 0.01 0.86 0.26 61.73 26.64 41.45 595.54 0.90 2325.95 1048.54 1.44 0.27 1.03 0.38 0.38 0.22 0.02 32.82 5.90 0.01 0.86 0.29 60.43 19.79 37.93 603.37 0.90 2096.40 1016.49 1.61 0.29 1.03 0.37 0.39 0.21 0.02 39.71 5.18 0.16 0.88 0.11 540.58 166.08 32.26 6.19 0.26 0.86 0.19 342.49 178.48 40.96 5.51 0.27 0.81 0.26 717.47 330.22 40.76 4.75 0.14 0.84 0.12 739.80 220.66 48.39 4.01 0.15 0.77 0.24 1593.00 710.22 40.55 4.44 0.19 0.78 0.25 1430.58 534.71 0.78 0.06 1.02 0.40 0.50 0.10 0.00 1.90 0.31 1.01 0.42 0.31 0.26 0.02 2.37 0.34 1.00 0.44 0.26 0.27 0.02 0.96 0.06 1.02 0.40 0.51 0.09 0.00 2.10 0.30 1.01 0.39 0.28 0.31 0.02 2.25 0.31 1.01 0.36 0.28 0.33 0.03 104 57,217 34,247 139 40,544 16,669 99 6,951 5,018 41.22 6.12 0.02 0.96 0.04 46.64 590.60 0.94 0.33 0.07 1.07 0.29 0.65 0.04 0.02 37.41 5.31 0.02 0.89 0.24 42.21 610.90 0.91 1.44 0.30 1.05 0.34 0.39 0.23 0.03 40.95 6.08 0.02 0.89 0.30 45.75 620.69 0.91 1.56 0.32 1.04 0.36 0.39 0.21 0.03 48.41 4.30 0.11 0.85 0.14 33.13 5.91 0.29 0.90 0.18 48.96 4.87 0.24 0.85 0.28 47.90 3.88 0.11 0.82 0.16 50.81 3.81 0.20 0.83 0.30 48.25 4.08 0.22 0.83 0.33 0.82 0.09 1.03 0.38 0.51 0.09 0.01 2.09 0.36 1.02 0.40 0.34 0.24 0.03 2.30 0.38 1.01 0.42 0.29 0.27 0.03 0.63 0.07 1.02 0.35 0.55 0.09 0.01 2.10 0.37 1.02 0.40 0.31 0.26 0.04 2.21 0.37 1.02 0.37 0.32 0.27 0.04 112 74,698 45,068 171 54,697 25,264 128 9,869 7,428 Panel A. 1870 Age Family size Works Literate Immigrant Reas estate Personal property Husband's age Husband's occupational income (1900) Husband literate Husband's real estate Husband's personal property County population density (1000/mile^2) County percent urban County male-to-female ratio Northeast Midwest South West Observations Panel B. 1880 Age Family size Works Literate Immigrant Husband's age Husband's occupational income (1900) Husband literate County population density (1000/mile^2) County percent urban County male-to-female ratio Northeast Midwest South West Observations See appendix B for description of variables. The IPUMS age adjusted sample is weighted to have the same distribution of five-year birth cohorts as the analogous sample in the linked widows data. Table B3. Composition of Households in Linked Widows Data Married 1870 Unmarried Married 1880 Unmarried Resides with child having first husband's surname Resides with chlld of first husband (imputed) Resides with second husband's child from previous marriage Has children from second marriage 0.51 0.77 0.32 0.58 0.87 0.87 0.44 0.51 0.22 0.74 0.75 0.75 # Children from previous marriage (widow) # Children from previous marriage (husband) # Children from second marriage 1.49 0.63 1.05 2.78 0.94 0.33 2.33 2.10 Widow is household head 0.00 0.71 0.00 0.75 Observations 104 139 112 171