The Deadest of Dead Letters? Reevaluating the Impact of The Civil Rights Act of 1875, A Harvey, EA West

Tags: public accommodations, African American, Civil Rights Cases, Civil Rights Act of 1875, body mass, discrimination, anti-discrimination statutes, USCT, slave states, statutes, weight gain, state-level, United States Colored Troops, variation, United States, Franklin, African American health, racial attitudes, Civil Rights, George Wright, accommodations, racial discrimination, American body, African American body, African Americans, weight gain/loss, relative weight gain, northern states, relative weight, American body mass, John Hope Franklin
Content: "The Deadest of Dead Letters"? Reevaluating the Impact of The civil rights Act of 1875 Anna Harvey and Emily A. West May 25, 2015 Abstract Recent work on the impact of anti-discrimination statutes has suggested that statutory prohibitions on race-based discrimination can have independent effects on African American labor market, health, and educational outcomes that are not simply endogenous to more fundamental attitudinal or economic trends (Chay 1998, Chay et al 2009). Yet historians generally dismiss the impact of the Civil Rights Act of 1875, prohibiting racial discrimination in public accommodations. This statute is widely assumed to have been ineffective in the face of postReconstruction racial hostility, and thus to have been "the deadest of dead letters" at the time of its demise in The Civil Rights Cases (1883) (Franklin 1974, Gillette 1979, Foner 1988). These claims have, however, never been subjected to rigorous empirical analysis. We report the first estimates of the impact of the CRA of 1875, leveraging ex ante variation in the presence of state public accommodations laws, and using the panel data available from the repeated medical exams of Union Army and United States Colored Troops (USCT) veterans collected as part of the pension application process. We find that, controlling for several measures of pre-ruling health, USCT veterans in states without state-level public accommodations statutes experienced significant relative weight loss over the 8-year window bracketing the Court's ruling in The Civil Rights Cases. This relative weight loss among African American veterans is consistent with reduced access to transportation services and increased stress during the postruling years, suggesting that the CRA of 1875 in fact had a positive effect on African American health prior to being struck in 1883. Our findings are robust to a series of discontinuity and placebo designs, suggesting that they are not simply the result of geographical variation in racial attitudes or pre-ruling sorting across our border of interest. These findings suggest that anti-discrimination statutes can reduce discriminatory behavior across a wider set of contexts than suggested by existing work. They also suggest a need to reevaluate the contribution of the Court's ruling in The Civil Rights Cases to the deepening of racial inequality in the South in the years following Reconstruction. Harvey (corresponding author) is Professor of Politics and West is Ph.D. Student, Department of Politics, New York University, 19 West 4th Street, New York, NY 10012; Email: (Harvey) [email protected] and (West) [email protected] 1
1 Introduction Until relatively recently, studies of the impact of anti-discrimination statutes had generally failed to demonstrate convincing causal effects (Donohue and Heckman 1991). Improvements in African American labor market outcomes after the passage of the Civil Rights Act of 1964, for example, were plausibly due to long term attitudinal and/or economic trends, rather than to the statute itself (Smith and Welch 1989). More recent designs have more successfully identified causal effects using statutory, geographical, and temporal discontinuities. For example, Chay et al (2009) found that the desegregation of Southern hospitals resulting from Title VI of the CRA of 1964 plausibly increased relative black educational outcomes during the 1980s. Chay (1998) found that the 1972 Equal Employment Opportunity Act, extending to small businesses the protections against employment discrimination in Title VII of the CRA of 1964, plausibly increased relative black employment and earnings during the 1970s. Collins (2003) and Neumark and Stock (2006) likewise found that state-level anti-discrimination statutes enacted during the 1940s and 1950s also plausibly increased relative black earnings during the period before the enactment of the CRA of 1964. These studies suggest that anti-discrimination statutes can in fact reduce discriminatory behavior, thereby improving economic, health, and educational outcomes among former targets of discrimination. Yet it remains an open question as to whether such statutes can reduce discriminatory behavior in the context of more deeply held discriminatory attitudes. The Civil Rights Act of 1875, for example, prohibiting racial discrimination in public accommodations, is generally thought by historians to have been an utter failure in the face of widespread racial hostility in the post-Reconstruction 2
years. Historian John Hope Franklin asserted, for example, that the Act "was never effectively enforced" (Franklin 1974, 235). William Gillette characterized the Act as "the most meaningless piece of postwar legislation...the deadest of dead letters," a characterization echoed by historians George Wright and Eric Foner (Gillette 1979, 271, 279; Wright 1985, 58; Foner 1988).1 We reassess this historiographical consensus by conducting the first tests of the impact of the CRA of 1875, the first two sections of which were ruled unconstitutional in The Civil Rights Cases, 109 U.S. 3 (1883). Our identification strategy leverages ex ante variation in the presence of state public accommodations laws, and uses panel data collected from the repeated medical exams of Union Army and United States Colored Troops (USCT) veterans collected as part of the pension application process. We find that, controlling for several measures of pre-ruling health, USCT veterans in states without state-level public accommodations statutes experienced significant relative weight loss over the 8-year window bracketing the Court's ruling in The Civil Rights Cases. This relative weight loss among African American veterans is consistent with reduced access to transportation services and increased stress during the post-ruling years, suggesting that the CRA of 1875 in fact had a positive effect on African American health prior to being struck in 1883. Our findings are robust to a series of discontinuity and placebo designs, suggesting that they are not simply the result of geographical variation in racial attitudes or pre-ruling sorting across our border of interest. These findings suggest that anti-discrimination statutes can reduce discriminatory behavior across a wider set of contexts than suggested by existing work. They also suggest a need to reevaluate the contribution of the Court's ruling in The Civil Rights Cases to the deepening of 1These assessments by professional historians echo in the public sphere. The PBS notes to accompany the documentary "Reconstruction," for example, assert that the Civil Rights Act of 1875 "was not enforced." 3
racial inequality in the South in the years following Reconstruction. 2 The Civil Rights Act of 1875 The Civil Rights Act of 1875, enacted on March 1, 1875, directed that "all persons within the jurisdiction of the United States shall be entitled to the full and equal enjoyment of the accommodations, advantages, facilities, and privileges of inns, public conveyances on land or water, theaters, and other places of public amusement," subject only to any limitations "established by law, and applicable alike to citizens of every race and color, regardless of any previous condition of servitude." The Act provided for both civil and criminal penalties for violations of its provisions; victims of discrimination could sue for damages of $500-$1000, or bring criminal charges with potential sentences ranging from 1-12 months in prison. The Act also provided for both civil and criminal penalties for district attorneys who failed to prosecute violations; victims of discrimination whose cases were not pursued could sue for $500 in damages, or bring misdemeanor charges with potential fines ranging from $1000-$5000. These provisions were all ruled unconstitutional in The Civil Rights Cases, issued on October 15, 1883, on the grounds that Congress lacked the power under either the Thirteenth or Fourteenth Amendments to enact these provisions.2 Historians of the period generally agree that the Act was broadly ineffective, and that therefore the Court's ruling in The Civil Rights Cases also had little effect (Franklin 1974, Gillette 1979, 2This ruling consolidated five separate cases brought under the Civil Rights Act of 1875, namely United States v. Stanley, United States v. Ryan, United States v. Nichols, United States v. Singleton, and Robinson v. Memphis & Charleston R.R. Co. Despite its earlier ruling that the interstate commerce clause requires that prohibitions on withinstate racial discrimination by common carriers operating between the states "must come from Congress and not from the States" (Hall v. De Cuir, 95 U. S. 485 (1877)), the Supreme Court did not evaluate the constitutionality of the CRA of 1875 under the interstate commerce clause. The statute itself did not specify a constitutional justification. 4
Wright 1985, Foner 1988). Yet there are several reasons to reconsider this consensus. First, the Act's provisions had teeth. In 2013 dollars, the damages available for victims of discrimination ranged from approximately $10,500 to $21,000, while the average African American family in- come of the time was only approximately $5,250 per year (Ng and Virts 1989). For purposes of comparison, Title II of the Civil Rights Act of 1964 provided for neither civil nor criminal penal- ties, offering only injunctive relief. A contemporaneous criticism of the CRA of 1875 was in fact that the size of its civil penalties would induce enterprising plaintiffs' attorneys to solicit African American clients to test the law's provisions, with the expectation that any awarded damages would be shared between plaintiff and attorney (The New York Times, March 14, 1875). The Act's provi- sions were at least sufficient to induce civil litigation even in the states presumably most hostile to its provisions (i.e., those without state-level public accommodations statutes). Within days of the Act's passage, civil suits were filed in many of these states, including Virginia, Louisiana, North Carolina, Texas, Maryland, Tennessee, Arkansas, Georgia, Mississippi, Kentucky, Florida, and Missouri (Franklin 1974).3 Likewise, in 2013 dollars district attorneys who failed to pursue criminal prosecutions under the Act could be sued for $10,500 in damages, and faced fines ranging from $21,000 to $105,000. These provisions, also absent from Title II of the 1964 CRA, appear to have made at least some kind 3Some, although not all, of these cases resulted in judgments in favor of the plaintiffs. See, for example, U.S. v. Dodge, 25 Fed. Cas. No. 14976 (W.D. TX 1877), finding for the plaintiff in a suit against a Texas railroad for refusing to accommodate a black woman in the "Ladies" car. Although questions were raised about the Act's constitutionality in some of these cases, many judges rejected these challenges. Judge Robert Dick of the Western District of North Carolina, for example, dismissed a grand jury's query about the Act's constitutionality in April 1875: "The constitutionality of the Civil Rights bill has been asserted by the deliberate action of Congress, composed of many able lawyers and wise and enlightened statesmen, and...[was] a question which that body carefully considered and acted upon under the solemn sanction of official obligation" (Franklin 1974, 231). Judge Amos Morrill of the Eastern District of Texas threatened the editor of the Galveston News with contempt for questioning the Act's constitutionality in May 1875. Judge Alexander Rives of the Western District of Virginia likewise explicitly upheld the Act's constitutionality in October 1875 (Ibid.). 5
of impression. The U.S. attorney in San Francisco, for example, informing the Attorney General of his appeal of an adverse judgment under the Act in 1876, wrote that the appeal was "one of a large number of cases which I am compelled under heavy penalties to bring" (Franklin 1974, 233). An assistant U.S. attorney in New York likewise told a federal court, "It is imperatively required by the law that the District Attorney shall diligently prosecute all cases arising under it, under penalty of civil and criminal prosecution, in the event of failure" (Franklin 1974, 229).4 Again, the Act's provisions were at least sufficient to induce U.S. attorneys in numerous states without state-level public accommodations statutes to pursue criminal indictments under its provisions, and to pursue appeals after adverse lower court judgments (Franklin 1974). Second, some anecdotal evidence suggests that, beyond the cases filed under the Act, the severity of its provisions may have induced at least partial voluntary compliance. In the wake of the Act's passage, at least some operators of public accommodations that had previously excluded African Americans began to admit and/or carry African American customers (Franklin 1974, 227, Rabinowitz 1978). Some reports suggest the continuation of at least some degree of access to public accommodations through the early 1880s. Abolitionist observers touring the South in 1878 and 1879 reported with surprise, for example, that access to public accommodations was not significantly dissimilar to levels of access in the New England states (Woodward 1957, 16-18, 23). Third, there are reports of successful invocations of the Act in order to compel access to public accommodations. In the late 1870s, for example, many African Americans sought to emigrate up THE MISSISSIPPI River in order to escape states that had been "redeemed" by the Democratic Party 4It was believed by some that district attorneys could even be held liable if grand juries failed to indict, as observed by a federal judge in Tennessee: "The severe penalties imposed by this law upon prosecuting attorneys and other officials will, we are advised, be attempted to be enforced, should the grand jury fail to indict, in the assumption that their action will be controlled by such officers" (Franklin 1974, 228). 6
(Painter 1977). Under pressure from Southern employers, several riverboat companies operating on Mississippi barred passage to African Americans. Some of those seeking passage threatened suit under the Civil Rights Act. USCT veteran John Solomon Lewis, for example, who sought to emigrate from Louisiana with his family, told a riverboat captain, "I am a man who was a United States soldier, and I know my rights, and if I and my family gets put off, I will go in the United States Court and sue for damages." Lewis and his family were allowed passage (Painter 1977, 3). The U. S. Attorney General later issued a public statement to the effect that the companies' refusal to carry African American passengers constituted a violation of the Civil Rights Act of 1875, which "makes such refusal an offense to which considerable penalty attaches, and provides the method by which a prosecution for the penalty or for damages to the party entitled can be pursued" (Franklin 1974, 229). In the wake of the Attorney General's statement the operators of the riverboat companies began to carry all African Americans seeking passage (Franklin 1974, 229; Jack 2007, 50, 91-92). Finally, some anecdotal evidence suggests that, after the Court's ruling in The Civil Rights Cases, at least some operators of public accommodations returned to excluding African American customers (Woodward 1957, 23, 31; Rabinowitz 1978, 187, 334-338; Cohen 1991, 269-273). In Louisville, Kentucky, after an extended period of access to streetcars, the railroad, theaters, Sporting Events, and annual fairs and expositions, by the mid-1880s African Americans were being excluded from all these public venues (Wright 1985, 54, 59-60, 63). In 1887 Nashville, a group of African American businessmen reported the existence of the kind of racial discrimination in public accommodations that had been prohibited by the Act: "We want public conveyances open 7
to us according to the fare we pay; we want the privilege to go to hotels and to theatres, operas and places of amusement." (Rabinowitz 1978, 336). This anecdotal evidence suggests that it may be worthwhile to reexamine the claim that the CRA of 1875 had no positive impact on African Americans. For reasons of data availability, we focus our analysis on the Act's demise, rather than its enactment. Following Fogel (2004) and Floud et al (2011), we look primarily for an effect of the Act on African American body mass. These studies found, in various contexts, that short-term changes in body mass may result from short-term changes in the quantity and quality of available nutrition, and in the energy expended per calorie consumed. Below we suggest two paths by which the Court's strike of the CRA of 1875 may have negatively impacted African American body mass. Following a number of studies finding that reductions in body mass lead to reductions in basal metabolic rates (see for example Leibel et al 1995), we look secondarily for a negative effect of the Act's strike on African American resting respiratory rates. 3 The Impact of the Act's Strike The first path by which the CRA of 1875 might have affected African American body mass is through access to transportation services. As already noted, there is anecdotal evidence that the Act helped to secure African American access to interstate transportation. For example, there are reports that some employers in more southern states, preferring to restrict the mobility of African American workers rather than compete on wages, at times pressured interstate railroad and steamboat operators not to carry African American passengers (Franklin 1974, Painter 1977, 8
Cohen 1991, 269-273, Jack 2007). During at least one such episode in the late 1870s, the invocation of the Civil Rights Act of 1875 was apparently sufficient to compel steamboat operators on the Mississippi River to resume carrying African American passengers, despite the entreaties of some southern employers. The demise of the Act in 1883 may have decreased African American access to interstate transportation, implying higher costs to leaving economically hostile environments and/or to pursuing more economically favorable environments elsewhere. These higher transportation costs in turn may have implied a relative downward pressure on African American wages, leading to decreases in the quantity and quality of nutrition consumed. A leader of the Nashville African American community wrote in 1895, for example, "Our people...go to the markets Saturday nights and buy the spoiled meats and vegetables on which the flies and smaller insects have preyed all day. In these vegetables are the seeds of indigestion disorders and death. The vegetables and meats which are left over and not sold in the market and grocery houses are put in wagons and driven to the colored settlements where they are sold cheap for cash" (Rabinowitz 1978, 71-72). There is also anecdotal evidence that the Act helped to secure access to intrastate and intracity transportation services (Rabinowitz 1978, Wright 1985). Access to streetcars and omnibuses in particular was especially important for African Americans, who typically lived further from places of employment. According to historian George Wright (1985), "Using the streetcar was a necessity to blacks who lived in outlying districts but worked in the center of the city" (p. 52). Prior to Pennsylvania's passage of a statute requiring equal accommodation on all transportation services, for example, most Philadelphia omnibuses and streetcar companies refused to carry African Amer- 9
icans. A catalyst for the passage of the state's public accommodations statute was apparently the daily sight of African American Civil War veterans being forced to walk considerable distances to their jobs, while white veterans rode the streetcars (Foner 1973a, 1973b, Rabinowitz 1978, 123). The demise of the CRA of 1875 may have reduced access to intracity transportation in states without state-level public accommodations statutes, again contributing to downward pressure on African American wages and decreases in the quantity and quality of consumed nutrition. As the anecdotal evidence from Philadelphia suggests, reduced access to intracity transportation could also have resulted in relatively greater energy expenditures, even if nutritional intake remained constant. A second path by which the Act's demise might have negatively affected African American body mass is via stress. Stress induced by the experience of racial discrimination has been associated with significant short-term physiological effects, including increased blood pressure and heart rates (Marmot 2005, Locher et al 2005, Williams and Mohammed 2009). Of perhaps particular relevance here, stress induced by perceived racial discrimination over a 6-month period has been linked to increased nutritional risk among African American men (Locher et al 2005). The stress induced by an environment of greater racial inequality after the strike of the CRA of 1875 thus plausibly could have had a negative impact on African American body mass in the years immediately following the Act's demise. The bulk of our analysis consists of estimating the effect of the strike of the Civil Rights Act of 1875 on African American body mass. However, given the finding that reductions in body mass lead to reductions in basal metabolic rates (see for example Leibel et al 1995), we also report 10
estimates of the effect of the Act's strike on African American resting respiratory rates. These estimates check the validity of our primary estimates of the effect of the Act's strike on African American body mass. 4 Estimating the Impact of the Act's Strike We primarily want to estimate the effect of the strike of the Civil Rights Act of 1875 on African American body mass. However, we face several obstacles to identification. The first such obstacle is the possible temporal collinearity between the Court's 1883 strike of the Act and any other nationwide events or trends that might have affected African American body mass. Declining public support for racial equality over our period of study, for example, might also have negatively affected African American body mass, independently of the Court's ruling. We address this obstacle by leveraging the ex ante variation in state-level public accommodations statutes. Many states, generally the more northern states, enacted state-level public accommodations statutes either before or shortly after the Court's ruling in The Civil Rights Cases (see Table 1). In these states, the strike of the Civil Rights Act of 1875 presumably would have had little impact on African American body mass, because these state-level statutes remained in place. We can thus use changes in the body mass of African Americans living in these "control" states to difference out the effects of any other national-level events or trends on African American body mass in our "treatment" states, namely those lacking state-level public accommodations statutes. However, state-level public accommodations statutes were not randomly assigned. Residents of states lacking public accommodations statutes may have preferred greater racial inequality, rel- 11
ative to residents of states with public accommodations statutes. Greater preferences for racial inequality in turn would likely have been associated with a greater incidence of racially motivated discrimination and violence, including on the part of local governmental officials charged with implementing racially neutral statutes (Novak 1978, Kousser 1999). These preferences and associated actions might have negatively affected African American body mass in the states lacking public accommodations statutes, independently of the Court's strike of the CRA of 1875. We address this threat to inference by treating the primary border between states with and without state-level public accommodations statutes as a discontinuity (see Figure 1). The identifying assumption is that variation in racial attitudes (and associated actions) is relatively continuous, while variation in statutes is discontinuous. Within small intervals just north and south of the border separating our control and treatment states, we expect little variation in racial attitudes.5 Any significant differences in differences in African American body mass at this border should then be the result of statutory differences between our control and treatment states, not attitudinal differences. 5There is some anecdotal evidence to support this assumption. In the postwar period African Americans reported significantly more progressive racial attitudes in the former slave states bordering former nonslave states, relative to the deeper South, while those living in cities just across this border (e.g. Philadelphia) reported attitudes less racially progressive than those found in the more northern cities (Foner 1973a, 1973b, Wright 1985). African Americans living in Ohio after the Civil War reported differences in racial attitudes between the more northern Cleveland and the more southern Cincinnati, located just across the Ohio River from Kentucky (Gerber 1976, 52-59). 12
Table 1: States Enacting Public Accommodations Statutes
State Massachusetts Rhode Island Connecticut New York New Jersey Pennsylvania District of Columbia Ohio Michigan Indiana Illinois Iowa Minnesota Nebraska Kansas Colorado
Date statute/s enacted 1865, 1866, 1885 1885 1884 1873, 1881 1884 1867, 1881 1863 1884 1885 1885 1885 1857, 1873, 1884 1885 1885 1874 1885
Source: Johnson (1919)
Figure 1: States Enacting Public Accommodations Statutes (in Blue) with North-South Boundary States that did not enact public accommodations statutes might also have been less likely to enact non-racially differentiated statutes providing for other public goods, such as schools, hospitals, sanitation, and roads. The variation in these statutes might generate variation in African American body mass at our border of interest, confounding estimates of the impact of the Court's strike of the CRA of 1875. Thus, we further difference our estimates using changes in European 13
American body mass on either side of our border of interest. The identifying assumption is that variation in non-racially differentiated state-level public goods provision should have affected the body mass of both African Americans and European Americans. By differencing out changes in European American body mass, we can isolate the effect of the variation in racially differentiated public accommodations statutes at our border of interest. States that did not enact public accommodations statutes might also have been more likely to enact racially differentiated restrictive statutes in other areas. If these statutes had a negative effect on African American body mass, then any effects we observe at our border of interest might be the result of the presence of these statutes, rather than of the Court's ruling in The Civil Rights Cases. Our reading of the secondary literature, however, suggests that this is likely not a significant obstacle. The appearance of postbellum statutes differentiated by race has been a subject of intense historical interest. Those who have studied this period report that, aside from the state-level public accommodations statutes used in our analysis, few other racially differentiated statutes existed between 1879 and 1887, our period of study. Redemption of southern states by the Democratic Party, and the consequent repeal of any Reconstruction-era statutes promoting racial equality, was largely complete by 1878, prior to the start of our study period. Yet the passage of segregating and disfranchising statutes did not begin until the late 1880s, after the end of our study period (Woodward 1957, Kousser 1974).6 6There appear to have been two exceptions to this overall pattern. Statutes prohibiting racial intermarriage and statutes requiring segregated schools were not uncommon during this period. However, the statewide variation in the presence of these statutes does not map well onto the statewide variation in the presence of public accommodations statutes. Many states that prohibited racial discrimination in public accommodations during our study period also prohibited interracial marriage and/or required segregated schools. For example, at the start of our study period these states included New York, Pennsylvania, and Kansas (mandatory school segregation), Rhode Island, Michigan, and Nebraska (prohibition on racial intermarriage), and Ohio and Indiana (both) (Johnson 1919). 14
It is also possible that there was systematic pre-ruling sorting across our border of interest in ways related to health status. This border correlates closely (albeit not perfectly) with the border differentiating states wherein slavery was legal in 1860, and those wherein it was not (the District of Columbia both permitted slavery in 1860 and enacted a public accommodations statute in 1863). Healthier African Americans might have been more likely to escape and/or emigrate from former slave states. These healthier African Americans living just north of our border of interest might also have been more able to secure adequate nutrition throughout the 1880s, relative to their less healthy counterparts who remained in the former slave states just south of this border. Less wealthy (and possibly less healthy) European Americans unable to compete in the slave economy may also have been more likely to emigrate from the former slave states, and might have been less able to secure adequate nutrition throughout the 1880s, relative to their healthier counterparts to the south. Sorting might also have continued throughout our study period of October 15, 1879 to October 15, 1887. To address the possibility of cross-border sorting related to health status we use panel data from which we can compute both individual-level measures of change in body mass during the years bracketing the Court's 1883 ruling, and a battery of pre-ruling measures of individual-level health. We look for relative balance on these measures of pre-ruling health within our African American and European American samples. We also restrict our samples to individuals who resided on the same side of our border of interest throughout our study period of October 15, 1879 to October 15, 1887. Finally, we identify those individuals in our sample who are recorded in the 1860 census as residing on the other side of our border of interest, relative to their location during our study 15
period, and replicate our analyses after dropping these individuals from our sample. Pre-ruling sorting may also have occurred on the basis of racial attitudes. Several of the non- public accommodations states did abolish slavery through state action prior to the ratification of the Thirteenth Amendment (Maryland, Missouri, Tennessee, and West Virginia). However, it is still possible that those European Americans more supportive of racial equality may have been more likely to leave former slave states, while those opposed to racial equality may have been more likely to move into the former slave states. This earlier sorting (coupled with interfamilial transmission of racial preferences) could have produced racial attitudes that varied relatively sharply at our border of interest. We address the possibility of sharply varying racial attitudes at our border of interest by implementing placebo tests replicating all models during the period when the Civil Rights Act remained in effect (1875-1883). If we find sharp relative decreases in African American body mass at the border between our treatment and control states in these "pre-treatment" placebo tests, then it is likely that any relative decreases in body mass we find in our actual period of interest are due not to the Court's ruling in The Civil Rights Cases, but rather to the more hostile racial attitudes found south of our border of interest. It is also possible, however, that the effects of racial attitudes varying sharply at our border of interest might only manifest after a longer period of time. Repeated exposure to the stress of racial discrimination over a period of long duration has been associated with an increased incidence of systemic disease, such as Cardiovascular Disease and a weakened immune system, the latter leading in turn to an increased susceptibility to a host of pathogenic diseases (Pascoe and Richman 2009). 16
It is possible that any relative weight loss we observe at our border of interest during our later study period, but not during our earlier placebo period, may be due to the emergence of these long-term effects from repeated exposure to more hostile racial attitudes south of this border, rather than to the Court's ruling in The Civil Rights Cases. To address this threat we construct a series of first differences in the average incidence of a variety of systemic diseases across the period bracketing the Court's ruling. These conditions are unlikely to have been affected by a relatively short period of exposure to greater racial discrimination after the Court's 1883 ruling, but may have been associated with a much longer period of sustained exposure to more hostile racial attitudes south of our border of interest. Any relative increases in the average incidence of these medical conditions for African Americans just south of our border of interest are likely due to sustained exposure to more hostile racial attitudes varying sharply at this border. The presence of any such relative increases in morbidity would suggest that any relative decreases in body mass over the same period are also likely due to sustained exposure to racial attitudes varying sharply at our border of interest, rather than to the Court's 1883 ruling. 5 Data and Results Between 1862 and 1890, Union Army and United States Colored Troops veterans could apply for military pensions by claiming that a current disability, interfering with their ability to perform manual labor, was directly or indirectly related to their wartime service (Glasson 1918).7 One com- 7On June 27, 1890, the pension law was amended to allow for disability claims unrelated to wartime service; in 1907 old age was recognized as a pensionable disability. Under the pre-1890 system of pension applications, a wartime gunshot wound was the modal basis for disability claims for both UA and USCT applicants (Logue and Blanck 2010). 17
ponent of the pension application process was a medical examination. Upon receiving a veteran's application, the Pension Bureau would direct the applicant to appear before a board of examining surgeons "at a location near his place of residence" (Logue and Blanck 2010, 30). Examining sur- geons were given a standardized form to complete directing them to report any medical evidence relevant to an applicant's claim of disability, and to provide a recommendation to the Pension Bu- reau as to whether a claim should be granted, and if so, at what level of support ("Instructions to Examining Surgeons, 1870-1926," Department of the Interior, Pension Office, Washington D.C.). The form also directed examining surgeons to record basic intake information on all applicants, including weight, height, and age. Our data are sourced from these medical exams, a random sample of which has been made available by the Early Indicators Project.8 The Early Indicators sample comprises those members of 332 Union Army companies and 52 United States Colored Troops companies whose military records could be linked with postbellum pension records.9 We construct our primary dependent variable from the weight measurements reported in this sample of medical exams.10 Many applicants were given multiple exams over time, either because they reapplied after their applications were initially denied, or because they applied for an increase to their current pension. These panel exams enable us to identify changes in weight at the individual level. Individual level 8 9The Union Army sample is comprised of nearly 40,000 European American soldiers; the USCT sample is comprised of a little over 6,000 African American soldiers. Slightly more than 2 million Union Army soldiers served during the Civil War, along with almost 200,000 United States Colored Troops (Costa and Kahn 2006, Logue and Blanck 2010). 10It is possible that racial bias affected physicians' examination reports. Between 1862 and 1890, examining physicians recommended 75.1% of white UA applicants for pensions, but only 67.2% of USCT applicants (Logue and Blanck 2010, p. 57). Racial bias may have incentivized examining physicians to overreport USCT applicants' health, so that they would be less likely to be awarded pensions. Greater racial bias south of our border of interest would then lead to an underestimate of the effect of The Civil Rights Cases on USCT veterans' body mass. 18
first differences in weight were constructed by subtracting a veteran's average preruling weight over the four year period prior to the day on which the Court's ruling in The Civil Rights Cases was announced (October 15, 1883) from his average postruling weight over the four year period following that announcement. This first difference in weight is our primary dependent variable of interest. A small percentage of the surgeons' reports record a patient's town and state of residence at the time of the exam. Many more record the town and state wherein the examination was conducted. Latitude and longitude coordinates for a veteran's location at the time of an exam were retrieved from the USGS list of populated places for identifiable combinations of town and state using first, place of residence, if available and identifiable, and second, place of exam. Arc GIS was then used to calculate the shortest distance from each identified set of latitude and longitude coordinates to the border separating those states with state-level public accommodations statutes and those without. Our largest sample consists of those UA and USCT veterans who were examined at least once in the four year period prior to the Court's ruling in The Civil Rights Cases, and at least once in the four year period following that announcement, where each exam reported the veteran's weight and could be linked to an identifiable geographical location, and where the veteran stayed on the same side of the border distinguishing states with and without public accommodations statutes for all exams. This sample consists of 2874 UA veterans examined a total of 8596 times, and 114 USCT veterans examined a total of 319 times. We also construct samples using windows of 200 miles, 100 miles, and 30 miles on either 19
side of our border of interest. Our samples for these estimates were constructed by averaging the distance to the border across all of a veteran's exams. A veteran included in our largest sample is included in a windowed sample if the average distance to the border across all his exams is equal to or less than the sample window. Our sample for the 200 mile window consists of 1747 Union Army veterans examined a total of 5283 times, and 84 USCT veterans examined a total of 238 times. Our sample for the 100 mile window consists of 1130 Union Army veterans examined a total of 3446 times, and 66 USCT veterans examined a total of 181 times. Our sample for the 30 mile window consists of 347 Union Army veterans examined a total of 1068 and 29 USCT veterans examined a total of 85 times. Figures 2 and 3 map these samples of exams. Figure 2: Map of UA Exams in Full, 200, 100 and 30 Mile Samples with North-South Boundary 20
Figure 3: Map of USCT Exams in Full, 200, 100 and 30 Mile Samples with North-South Boundary It is possible that the populations of UA and USCT pension applicants differed on preruling health measures. The pre-1890 pension application process required either literacy or sufficient affluence to hire a pension attorney. Educations and incomes were lower among USCT veterans, relative to UA veterans; consequently only 19.7% of USCT veterans in the Early Indicators sample applied for a pension between 1862 and 1890, relative to 43.9% of UA veterans. USCT pension applicants were thus likely to be unusually advantaged relative to other USCT veterans, and possibly even relative to UA veterans and applicants (Costa and Kahn 2006, Logue and Blanck 2010). USCT veterans were also less likely than their UA counterparts to have received adequate medical treatment, such as hospitalization, for diseases incurred during the war (Costa et al 2007). But because the Pension Bureau asked applicants to provide wartime hospitalization records as documentation for medical conditions occurring during service, USCT pension applicants were disproportionately drawn from those USCT veterans who had received unusually diligent wartime 21
medical care. Again, this would have distinguished USCT pension applicants not only from USCT veterans, but also possibly from UA veterans and applicants.11 However, our design requires not that UA and USCT veterans had similar levels of preruling health, but only that the preruling populations of UA and USCT veterans north of our border of interest were relatively similar to their counterpart populations south of the border. We use the sampled medical exams to construct individual level preruling averages for weight, height, and age and the incidence of a variety of medical conditions. The list of the medical conditions for which we were able to construct these preruling averages is reported in Table 10 in the Appendix, along with the full list of variables used in our analyses. descriptive statistics for our dependent variable as well as covariate balance tests for measures of preruling health in the full sample are reported in Table 2, and Tables 11 - 13 in the Appendix report similar covariate balance tests for our 200, 100 and 30 mile samples. The summary statistics for individual-level changes in weight report the basic story in our data, which remains essentially unchanged throughout the remainder of our analysis. First, all veterans in our samples were on average undernourished prior to the Court's ruling. The veterans in the samples analyzed here were on average about 5 feet 8 inches tall, with little variation across race or geographic location. At this average height, the weight that would have minimized mortality and morbidity risks and increased labor force participation is 158.5 pounds.12 All of our samples report lower average preruling weight than this optimal weight. 11Pension applicants missing wartime medical records were required to submit sworn affidavits from fellow soldiers documenting the time, place, and manner of wartime injuries or illnesses, a difficult undertaking even for UA veterans. 12Using the medical exams of Union Army pension applicants from 1870-1910, Robert Fogel estimated that the body mass index (BMI) minimizing mortality risk was 24.1 (Fogel 2004). Increases in body mass up to this optimum were also associated with increases in labor force participation and decreases in morbidity risk for a variety of systemic diseases. 22
Table 2: Covariate Balance, Full Sample
Dependent Variable Weight Gain/Loss
USCT North USCT South DIFF UA North UA South DIFF
3.13
-0.01 3.14
-0.08
1.81 -1.89**
Pre Ruling Covariates Weight Height Age Respiratory Diarrhea Infectious Disease Cardio Eye Endocrine Gastrointestinal Kidney Hernia Injury Liver/Spleen/Gallbladder Tumor Musculoskeletal Varicose Veins Nervous System Rectum N
153.14 68.15 47.21 0.12 0.07 0 0.11 0.04 0 0.06 0 0.06 0.36 0.03 0 0.31 0.01 0.05 0.08 56
155.75 67.88 46.12 0.16 0.02 0 0.06 0.02 0 0.07 0 0.03 0.44 0.02 0 0.21 0.01 0.09 0.05 58
-2.61 0.27 1.09 -0.04 0.05 0 0.05 0.02 0 -0.01 0 0.03 -0.08 0.01 0 0.10 0 -0.04 0.03
148.37 68.29 45.74 0.15 0.08 0.00 0.13 0.06 0.00 0.12 0.00 0.05 0.32 0.05 0.01 0.19 0.05 0.12 0.08 2673
147.48 68.48 45.77 0.18 0.02 0 0.07 0.07 0.00 0.08 0 0.04 0.28 0.04 0.01 0.20 0.03 0.11 0.10 201
0.89 -0.19 -0.03 -0.03 0.06*** 0.00 0.06*** -0.01 0 0.04** 0.00 0.01 0.04 0.01 -0.00 -0.01 0.02 0.01 -0.01
Ns for Pre Ruling Age are 55 USCT North, 57 USCT South, 2665 UA North, and 201 UA South. Summary statistics for 200 mile, 100 mile and 30 mile samples in Appendix.
23
However, some veterans were heavier before the Court's ruling and also gained weight over our study period. USCT veterans, both north and south of our border of interest, were on average heavier than UA veterans during the preruling period, perhaps because the pension process required USCT pension applicants to be unusually advantaged. USCT veterans living in states with public accommodations statutes then gained weight on average after the Court's ruling in The Civil Rights Cases, relative to Union Army veterans in these same states. In the full sample, USCT veterans living in states with public accommodations statutes gained an average 3.1 pounds after the Court's ruling in The Civil Rights Cases, while UA veterans in these same states lost on average .3 pounds. This average relative weight gain for USCT veterans persists as we move closer to the border separating states with and without public accommodations statutes. Within 200 miles of this border this average relative weight gain is 4.1 pounds, within 100 miles of the border is 4.9 pounds, and within 30 miles of the border is 2.8 pounds. This relative weight gain for USCT veterans north of our border of interest is perhaps not surprising, given that they started from a baseline of better preruling health, relative to UA veterans north of our border of interest. UA veterans in the states without public accommodations statutes also gained weight over the period bracketing the Court's ruling, relative to UA veterans in the states with such statutes. In the full sample this relative weight gain averages 1.9 pounds, within 200 miles of the border it averages 2.4 pounds, within 100 miles of the border it averages 2.8 pounds, and within 30 miles of the border the average relative weight gain for UA veterans is 3.6 pounds. Given that this weight gain persists relatively close to our border of interest, it is likely due to statutory variation across this border. It appears to be the case that states just south of our border of interest provided a 24
statutory context of non-racially differentiated public goods that better supported the veterans in our sample, relative to states just north of this border. These average relative increases in weight for USCT veterans in states with public accommodations statutes and UA veterans in states without such statutes condition our expectations for the changes in weight that should have been experienced by USCT veterans in states without public accommodations statutes. Prior to factoring in any changes in weight due to the Court's ruling in The Civil Rights Cases, USCT veterans in the full sample for our treatment states should have gained an average of 5 pounds, relative to UA veterans in our control states. Within 200 miles of the border, we expect an average relative weight gain for USCT veterans of 6.5 pounds; within 100 miles of the border, an average relative weight gain of 7.7 pounds, and within 30 miles of the border, an average relative weight gain of 6.4 pounds. However, as reported in Table 2, and Tables 11 - 13 in the Appendix, what we observe is in fact no significant weight gain by USCT veterans in our treatment states, in any of our samples. Despite sharing a preruling weight advantage equal to if not greater than that reported by USCT veterans in the states just to their north, USCT veterans in states without public accommodations statutes did not share in the weight gain of these northern USCT neighbors. And despite living under the same otherwise favorable statutory regimes as the UA veterans in states without public accommodations statutes, USCT veterans in these states also did not share in the weight gain of their UA neighbors. We interpret this as a relative weight loss of approximately 5-8 pounds by USCT veterans in states without public accommodations statutes, as a consequence of the Court's ruling in The Civil Rights Cases.13 13We do observe what appears to be some measurement error in our data, with clustering on weight observations 25
Our regressions estimating this relative weight loss control for preruling measures of individuallevel health that might independently affect weight gain/loss. As reported in Table 2, and Tables 11 - 13 in the Appendix, however, within our samples of United Army and USCT veterans there is in fact little variation in preruling characteristics across our border of interest. The presence of relative balance on observable health indicators suggests that veterans did not systematically sort across this border on dimensions related to health status. Table 3 reports regression estimates of individual-level weight gain/loss for the full sample and for the windows of 200 miles, 100 miles, and 30 miles on either side of the border. In Table 3 our estimating equation is Weight Gain/Lossi = 0 + 1Southi + 2USCTi + 3South X USCTi + 4Xi + i, (1) where 3 reports the estimated marginal effect of the Court's ruling on USCT veterans located south of the border distinguishing those states with and without state-level public accommodations statutes, relative to all other categories of veterans. Numbers of USCT veterans north and south of the border are reported for each model. Columns 1-4 report estimates for models with no covariates, while columns 5-8 report estimates for models that include preruling average weight, height, and age, as well as the preruling average incidence of several medical conditions (X i). Robust standard errors are clustered at 30-mile intervals from our border of interest, with the interval-identifiers increasing in value with distance south of the border. All models report relatively consistent results. For the sake of brevity we focus on the models ending in 0 or 5. However, the magnitudes of our estimated effects appear to be sufficiently large so as not to be subsumed within this measurement error.
Table 3: Individual-Level Weight Gain/Loss, 10/15/1879 to 10/15/1887
South of Border=1 US Colored Troops=1 South of Border=1 X US Colored Troops=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N USCT North of Border USCT South of Border
Model 1 1.89*** (0.50) 3.22** (1.21) -5.03** (2.23) -0.08 (0.23) 2988 56 58
Model 2 2.36*** (0.49) 4.11*** (1.25) -5.38** (2.31) -0.30 (0.23) 1831 39 45
Model 3 2.84*** (0.43) 4.90** (1.88) -4.82* (2.29) -0.68*** (0.12) 1196 33 33
Model 4 3.58** (1.47) 2.84 (2.12) -7.54* (4.22) -0.72 (0.60) 376 22 7
Model 5 2.33*** (0.73) 3.74** (1.40) -4.67** (2.07) -0.11*** (0.02) -0.07** (0.03) 0.10 (0.10) -0.10 (0.57) 0.49 (0.62) -3.29 (3.55) 1.27** (0.57) 1.25 (0.92) -9.15*** (2.07) -0.51 (0.46) 9.09*** (2.93) 1.41 (0.91) 1.29** (0.57) 1.81 (1.36) 1.02 (3.05) 0.45 (0.59) 0.50 (0.68) -0.55 (0.87) 0.37 (0.77) 7.16 (5.30) 2627 54 56
Model 6 2.95*** (0.71) 5.08*** (1.44) -6.22** (2.26) -0.14*** (0.02) -0.07* (0.04) 0.07 (0.16) -0.45 (0.70) 0.31 (0.76) -5.17 (5.16) 0.40 (0.76) 0.14 (0.92) -9.29*** (2.48) -0.03 (0.61) 2.11 (1.62) 0.35 (0.64) 0.08 (0.52) 2.96 (2.29) 1.21 (4.37) -0.18 (0.74) -0.59 (0.74) -0.97 (1.32) 0.36 (1.25) 11.69 (8.16) 1622 38 44
Model 7 3.49*** (0.75) 6.62** (2.18) -6.02** (2.37) -0.15*** (0.02) -0.11 (0.06) 0.27 (0.19) -1.07 (1.12) -0.08 (0.72) 7.25*** (1.51) 0.35 (1.09) 0.91 (0.97) -12.13** (3.47) -0.03 (0.92) 4.88*** (1.30) -0.11 (0.47) -0.38 (0.35) 1.14 (3.46) -0.01 (6.89) -0.52 (0.77) -0.86 (1.08) -0.04 (1.64) 0.92 (1.65) 3.49 (8.76) 1069 32 32
Model 8 4.34*** (1.49) 4.77** (2.23) -9.15** (4.40) -0.17** (0.07) 0.01 (0.03) 0.06 (0.19) -2.51 (1.61) 1.90 (1.46) 0.00 (.) 1.84 (1.45) 3.30 (2.13) -6.76 (5.69) -1.29 (1.44) 0.00 (.) 0.67 (2.34) -0.29 (1.48) 9.30*** (3.01) 4.06 (3.97) 1.27 (1.47) -1.69 (1.80) 3.39* (1.82) 4.10** (1.70) 0.85 (11.67) 334.00 22 7
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border (except models 4 & 8, which are not clustered but restricted to samples 30 miles from border). * p<.10, ** p<.05, *** p<.01 Models 2/6, 3/7 & 4/8 restricted by 200, 100 & 30 miles, respectively.
27
including all preruling covariates. Consistent with the descriptive statistics reported in Table 2, and Tables 11 - 13 in the Appendix, UA veterans in our control states experience no significant weight gain or loss over our study period. But both USCT veterans in the control states and UA veterans in the treatment states experience significant weight gain, relative to UA veterans in the control states. These significant weight gains persist as we approach our border of interest. But USCT veterans just south of the border are not sharing in the apparent weight gains experienced by their fellow UA veterans just south or their fellow USCT veterans just north of our border of interest. In fact, USCT veterans located within 30 miles south of the border experience a relative weight loss of 9.2 pounds, relative to their fellow veterans within 30 miles north or south of the border. Within 100 miles of the border this relative weight loss is estimated at 6 pounds, within 200 miles of the border is estimated at 6.2 pounds, and in the full sample is estimated at 4.7 pounds. All estimates are significant at .05. Our second strategy for estimating the effect of the Court's ruling in The Civil Rights Cases uses the full sample but estimates a parameter for 30 mile intervals from the border separating states with and without public accommodations statutes. The variable 30 Mile Distance Intervals codes intervals south of the border from 0 to 50, increasing in distance from the border, and intervals north of the border from 0 to -50, decreasing in distance from the border. We interact this distance variable with the variable denoting those states without public accommodations statutes (South of the Border), as well as include this latter variable separately. Finally, we interact these three measures with the indicator for USCT veterans. Our estimating equation is thus: 28
Weight Gain/Lossi = 0 + 130 Milei + 2Southi + 3South X 30 Milei + 4USCTi + 5USCT X 30 Milei + 6South X USCTi + 7USCT X 30 Mile X Southi + 8Xi + i, (2) 5 and 7 capture the continuous effect of a more southern location on the weight gain/loss of USCT veterans, relative to the effect for UA veterans. The quantity of interest is South of the Border = 1 X USCT = 1 (6), which captures the discontinuous marginal effect on USCT veterans of location within 30 miles south of the border, relative to location within 30 miles north of the border, again relative to the effect for UA veterans. We estimate this model both with and without a vector of preruling covariates (X i), clustering errors on the 30 mile distance intervals. As reported in Table 4, in this specification there continues to be no significant weight gain or loss just north of the border for UA veterans. We again find significant weight gain both for USCT veterans just north of the border, and for UA veterans just south of the border. But again USCT veterans just south of the border do not appear to share in this weight gain. Instead, after including all preruling covariates, we see a relative loss of approximately 5 pounds for USCT veterans just south of the border. This relative weight loss is significant at .05. Tables 14 and 15 in the Appendix replicate these analyses dropping those veterans whose place of residence in 1860 was located on the other side of the border from their location during our study period. Results are qualitatively similar, with some loss of significance due to reduced numbers of observations. Tables 5 and 6 report placebo tests that replicate our analyses during the preruling period of
Table 4: Individual-Level Weight Gain/Loss, 10/15/1879 to 10/15/1887
30 Mile Dist. Intervals South of Border=1 South of Border=1 X 30 Mile Dist. Intervals US Colored Troops=1 USCT=1 X 30 Mile Dist. Intervals South of Border=1 X USCT=1 USCT=1 X Dist. Intervals X South of Border=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N USCT North of Border USCT South of Border
Model 1 -0.08** (0.04) 2.96*** (0.37) -0.22* (0.11) 3.74*** (1.16) 0.09 (0.25) -4.13 (2.49) -0.40 (0.65) -0.56*** (0.20) 2988 56 58
Model 2 -0.10** (0.04) 3.69*** (0.55) -0.29** (0.13) 4.44*** (1.32) 0.13 (0.27) -5.05** (2.32) -0.08 (0.51) -0.11*** (0.02) -0.07** (0.03) 0.10 (0.11) -0.10 (0.58) 0.51 (0.61) -3.66 (3.51) 1.37** (0.57) 1.37 (0.92) -8.75*** (2.03) -0.57 (0.47) 8.78*** (2.78) 1.30 (0.91) 1.31** (0.57) 1.68 (1.38) 1.05 (3.07) 0.51 (0.60) 0.49 (0.69) -0.65 (0.86) 0.30 (0.79) 6.49 (5.63) 2627 54 56
All results are from OLS regressions. Robust standard errors clustered on 30 mile intervals from border. * p<.10, ** p<.05, *** p<.01 30
October 15, 1875 to October 15, 1883, with a hypothetical placebo ruling assumed to take place on October 15, 1879. During this period the Civil Rights Act of 1875 remained in effect in states both with and without public accommodations statutes, eliminating the legal distinction at the border between these two sets of states. We are constrained by very sparse data on USCT veterans during this time period. Our largest sample for the preruling placebo tests is composed of 1033 Union Army veterans examined a total of 3042 times and 35 USCT veterans examined a total of 97 times. Our 200 mile window sample is composed of 614 UA veterans examined a total of 1801 times and 28 USCT veterans examined a total of 77 times. Our 100 mile window sample is composed of 399 Union Army veterans examined a total of 1159 times and 23 USCT veterans examined a total of 62 times. Finally, our 30 mile window sample is composed of 123 Union Army veterans examined a total of 354 times and only 8 USCT veterans examined a total of 21 times. All variables are constructed as for our main analyses. With or without pre placebo covariates, and across all geographical windows, there is no evidence of any significant difference in the weight gain/loss of USCT veterans south of our border of interest, relative to other categories of veterans. While necessarily tentative due to the small numbers of USCT observations, these results suggest that our results for the later period are not simply a product of greater racial hostility in the states south of our border of interest. Table 7 reports estimates for our original study period of October 15, 1879 to October 15, 1887, implementing Equation 1 on a series of individual-level first differences in the incidence of a series of medical conditions that are unlikely to have been affected by a relatively short period
Table 5: Individual-Level Weight Gain/Loss, 10/15/1875 to 10/15/1883 (Pre Ruling Placebo Tests)
South of Border=1 US Colored Troops=1 South of Border=1 X US Colored Troops=1 Pre Placebo Age Pre Placebo Weight Pre Placebo Height Pre Placebo Abnormal Resp. Pre Placebo Diarrhea Pre Placebo Infectious Disease Pre Placebo Cardio Pre Placebo Eye Pre Placebo Endocrine Pre Placebo Gastrointestinal Pre Placebo Kidney Pre Placebo Hernia Pre Placebo Injury Pre Placebo Liver/Spleen/Gallbladder Pre Placebo Tumor Pre Placebo Rectum Pre Placebo Musculoskeletal Pre Placebo Varicose Veins Pre Placebo Nervous System Constant N USCT North of Border USCT South of Border
Model 1 -3.01*** (1.07) -0.33 (2.56) -1.76 (3.35) 0.42 (0.36) 1068 16 19
Model 2 -2.97** (1.20) 0.23 (3.20) -3.02 (3.76) 0.19 (0.50) 642 12 16
Model 3 -2.57** (0.70) -2.78 (1.66) -1.76 (2.25) 0.67* (0.28) 422 9 14
Model 4 -4.35** (1.87) -4.80** (2.24) 2.60 (3.43) 1.30 (0.82) 131 6 2
Model 5 -3.72*** (1.13) 0.23 (2.34) -0.32 (3.15) -0.02 (0.03) -0.13*** (0.03) 0.59*** (0.14) 0.96 (1.00) -3.11** (1.40) -6.31*** (0.71) 0.27 (1.61) 2.87*** (0.87) 0.00 (.) 0.44 (0.98) 2.80 (5.58) -1.62 (1.42) 2.02*** (0.45) 2.09 (1.57) 6.40*** (0.99) 1.50 (1.66) -0.60 (1.10) 3.20 (2.24) -0.07 (0.66) -21.57** (9.23) 1062 16 19
Model 6 -3.85*** (1.19) 0.54 (3.13) -0.83 (3.43) -0.01 (0.03) -0.12*** (0.03) 0.73*** (0.19) 2.10 (1.21) -3.29 (2.32) -7.65*** (1.16) 0.66 (1.57) 2.64** (1.02) 0.00 (.) 0.69 (2.04) -8.26*** (2.45) 0.56 (1.34) 2.04*** (0.52) 3.68* (1.72) 7.24*** (1.15) 1.68 (2.08) -0.81 (1.45) 2.65 (3.15) 0.43 (0.99) -33.72** (11.71) 641 12 16
Model 7 -3.31*** (0.66) -1.82 (2.44) -0.75 (2.88) 0.01 (0.03) -0.10* (0.04) 0.48*** (0.12) 3.24 (2.43) 0.00 (1.80) 0.00 (.) -0.24 (1.94) 1.75 (1.33) 0.00 (.) 0.35 (3.24) -5.32* (2.57) 0.82 (2.02) 2.45** (0.80) 3.74 (2.31) 7.82*** (1.09) -1.68 (1.62) 0.15 (1.84) 3.57 (2.43) 0.28 (1.44) -20.45 (11.49) 422 9 14
Model 8 -5.60*** (2.00) -3.42 (3.42) 4.87 (5.67) -0.04 (0.10) -0.03 (0.06) 0.48* (0.28) -0.87 (2.39) 4.64 (4.82) 0.00 (.) 1.90 (3.27) -0.24 (1.85) 0.00 (.) 7.73* (4.20) 0.00 (.) 1.42 (2.92) 1.91 (1.76) 3.69 (2.78) 10.97*** (2.39) -5.51 (10.32) -1.33 (2.47) 1.89 (3.92) 0.50 (2.47) -26.43 (18.36) 131 6 2
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border (except models 4 & 8, which are not clustered but restricted to samples 30 miles from border). * p<.10, ** p<.05, *** p<.01 Models 2/6, 3/7 & 4/8 restricted by 200, 100 & 30 miles, respectively. Pre Placebo controls are equivalent to the Pre Ruling controls in other regressions (but calculated for the years 1875-1879).
32
Table 6: Individual-Level Weight Gain/Loss, 10/15/1875 to 10/15/1883 (Pre Ruling Placebo Tests)
30 Mile Dist. Intervals South of Border=1 South of Border=1 X 30 Mile Dist. Intervals US Colored Troops=1 USCT=1 X 30 Mile Dist. Intervals South of Border=1 X USCT=1 USCT=1 X Dist. Intervals X South of Border=1 Pre Placebo Age Pre Placebo Weight Pre Placebo Height Pre Placebo Abnormal Resp. Pre Placebo Diarrhea Pre Placebo Infectious Disease Pre Placebo Cardio Pre Placebo Eye Pre Placebo Endocrine Pre Placebo Gastrointestinal Pre Placebo Kidney Pre Placebo Hernia Pre Placebo Injury Pre Placebo Liver/Spleen/Gallbladder Pre Placebo Tumor Pre Placebo Rectum Pre Placebo Musculoskeletal Pre Placebo Varicose Veins Pre Placebo Nervous System Constant N USCT North of Border USCT South of Border
Model 1 0.04 (0.05) -3.38*** (0.87) 0.02 (0.32) -0.09 (3.43) 0.06 (0.46) -4.33 (4.16) 0.85 (0.90) 0.67 (0.46) 1068 16 19
Model 2 0.02 (0.06) -3.95*** (1.07) 0.04 (0.28) 0.40 (3.22) 0.04 (0.41) -2.65 (4.01) 0.80 (0.79) -0.02 (0.03) -0.13*** (0.03) 0.59*** (0.14) 0.92 (1.00) -3.10** (1.40) -6.29*** (0.77) 0.26 (1.61) 2.74*** (0.89) 0.00 (.) 0.47 (1.01) 2.88 (5.66) -1.64 (1.43) 2.03*** (0.45) 2.16 (1.55) 6.41*** (1.00) 1.47 (1.68) -0.59 (1.12) 3.18 (2.24) -0.02 (0.68) -21.45** (9.29) 1062 16 19
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border. * p<.10, ** p<.05, *** p<.01 Pre Placebo controls are equivalent to the Pre Ruling controls in other regressions (but calculated for the3y3ears 1875-1879).
of increased racial discrimination in public accommodations. If we observe any increases in the incidence of these medical conditions for USCT veterans south of our border of interest, relative to other categories of veterans, it is likely that these increases (along with the previously observed decreases in body mass) were due not to the Court's ruling in The Civil Rights Cases, but rather to long-term exposure to more hostile racial attitudes south of this border. All first differences in morbidity were constructed using the same procedure used to construct first differences in weight, for all medical conditions reported in Table 10 in the Appendix. Table 7 reports estimates for the full sample, including all preruling covariates.14 With only one exception, there are no significant changes in the incidence of any of these medical conditions for USCT veterans south of our border of interest, relative to other categories of veterans. The one exception is in the wrong direction: USCT veterans in states without public accommodations statutes were less likely to have musculoskeletal disorders reported on their examination forms after the Court's ruling in The Civil Rights Cases, relative to other veterans. However, this finding weakens as we approach the border, disappearing entirely in the 30 mile sample. 14Regressions on first differences in morbidity for the 200, 100, and 30 mile samples available from authors. 34
Table 7: Individual-Level Change in Average Disease Incidence, 10/15/1879 to 10/15/1887, Full Sample
South of Border=1 US Colored Troops=1 South of Border=1 X US Colored Troops=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N
(1) Abn. Resp. 0.02 (0.01) -0.01 (0.03) -0.03 (0.03) 0.00 (0.00) -0.00* (0.00) 0.01** (0.00) -0.43*** (0.02) -0.02 (0.03) -0.26*** (0.08) 0.00 (0.02) -0.07*** (0.02) -0.01 (0.11) -0.06** (0.03) -0.07 (0.07) -0.09*** (0.02) -0.09*** (0.01) 0.02 (0.05) -0.06 (0.06) -0.00 (0.03) -0.04*** (0.02) 0.00 (0.04) -0.02 (0.02) -0.10 (0.16) 2978
(2) Diarrhea -0.04*** (0.01) -0.01 (0.03) 0.04 (0.04) -0.00 (0.00) -0.00* (0.00) 0.00* (0.00) -0.02* (0.01) -0.27*** (0.03) 0.08 (0.07) -0.03** (0.01) -0.07*** (0.02) 0.10 (0.11) -0.01 (0.01) -0.11* (0.06) -0.06*** (0.01) -0.05*** (0.01) 0.02 (0.03) 0.04 (0.05) 0.04* (0.02) -0.05*** (0.01) -0.03 (0.02) -0.04*** (0.01) -0.08 (0.13) 2978
(3) Inf. Dis. -0.00*** (0.00) 0.01 (0.02) -0.02 (0.02) -0.00 (0.00) 0.00 (0.00) 0.00* (0.00) 0.00 (0.00) 0.00 (0.00) -1.00*** (0.00) -0.00 (0.00) -0.00 (0.00) -0.01*** (0.00) 0.00 (0.00) -0.00 (0.00) -0.00** (0.00) 0.00 (0.00) 0.00 (0.00) -0.00** (0.00) -0.00** (0.00) 0.00 (0.00) 0.01 (0.01) -0.00 (0.00) -0.05* (0.02) 2978
(4) Cardio. -0.04* (0.02) -0.04 (0.04) -0.05 (0.05) 0.00 (0.00) -0.00 (0.00) 0.01** (0.00) 0.03 (0.02) 0.01 (0.03) 0.31* (0.16) -0.76*** (0.02) -0.12*** (0.02) -0.33*** (0.07) 0.01 (0.03) 0.07 (0.09) -0.06 (0.04) -0.10*** (0.02) -0.01 (0.03) -0.06 (0.10) -0.04* (0.02) 0.07*** (0.02) -0.06** (0.03) 0.06** (0.02) -0.27* (0.15) 2978
(5) Eye -0.01 (0.02) -0.05*** (0.01) 0.04 (0.03) 0.00*** (0.00) -0.00 (0.00) -0.00 (0.00) -0.02 (0.01) -0.01 (0.02) 0.05 (0.13) -0.01 (0.01) -0.34*** (0.04) 0.06 (0.15) 0.00 (0.02) -0.12* (0.06) -0.04 (0.02) -0.03** (0.01) 0.03 (0.03) -0.01 (0.04) 0.00 (0.01) -0.04*** (0.01) -0.01 (0.01) -0.01 (0.02) 0.11 (0.11) 2978
(6) Endo. 0.01* (0.00) -0.00** (0.00) 0.01 (0.02) 0.00 (0.00) -0.00* (0.00) 0.00** (0.00) -0.00*** (0.00) -0.00*** (0.00) -0.00 (0.00) 0.00 (0.00) -0.00*** (0.00) -0.86*** (0.15) -0.00* (0.00) -0.01** (0.00) -0.00*** (0.00) -0.00*** (0.00) 0.00 (0.00) -0.00** (0.00) 0.00 (0.01) 0.00 (0.00) 0.00 (0.01) 0.00 (0.00) -0.03* (0.02) 2978
(7) Gastro. -0.02* (0.01) -0.08** (0.03) 0.00 (0.04) 0.00 (0.00) -0.00*** (0.00) 0.01*** (0.00) -0.05*** (0.02) 0.23*** (0.04) 0.24* (0.12) -0.02 (0.02) -0.12*** (0.02) 0.14* (0.07) -0.67*** (0.03) -0.01 (0.10) -0.14*** (0.03) -0.14*** (0.01) 0.14*** (0.05) -0.06 (0.06) 0.12*** (0.04) -0.09*** (0.01) -0.08*** (0.03) -0.06*** (0.02) -0.30* (0.15) 2978
(8) Kidney -0.00 (0.00) 0.01 (0.01) -0.02 (0.01) 0.00** (0.00) -0.00 (0.00) -0.00* (0.00) -0.00* (0.00) -0.00 (0.01) -0.01 (0.00) 0.01 (0.00) -0.01*** (0.00) 0.04 (0.06) 0.00 (0.00) -0.70*** (0.09) -0.01*** (0.00) -0.01 (0.00) -0.00 (0.01) -0.01*** (0.00) 0.00 (0.01) -0.00 (0.00) -0.01*** (0.00) -0.01 (0.00) 0.03 (0.02) 2978
(9) Hernia -0.00 (0.01) -0.01 (0.01) 0.00 (0.01) 0.00* (0.00) -0.00 (0.00) 0.00 (0.00) -0.00 (0.01) -0.00 (0.01) -0.10 (0.07) -0.01 (0.01) -0.02 (0.01) 0.00 (0.04) 0.01* (0.01) -0.03*** (0.00) -0.16*** (0.03) 0.00 (0.00) -0.02 (0.01) 0.05 (0.05) -0.01 (0.01) 0.00 (0.01) 0.01 (0.01) -0.01 (0.01) -0.02 (0.05) 2978
(10) Injury 0.01 (0.01) 0.03 (0.04) -0.01 (0.05) -0.00 (0.00) 0.00*** (0.00) -0.00 (0.00) -0.03** (0.02) -0.01 (0.02) 0.02 (0.06) -0.03** (0.02) -0.08*** (0.02) -0.10 (0.07) -0.04** (0.02) -0.04 (0.03) -0.04** (0.02) -0.23*** (0.02) 0.00 (0.02) -0.07 (0.06) -0.04** (0.02) -0.02 (0.02) -0.03 (0.02) 0.00 (0.02) 0.11 (0.14) 2978
(11) Liv./Spl./Gall. -0.02 (0.02) -0.01 (0.02) -0.04 (0.03) 0.00* (0.00) -0.00** (0.00) 0.00** (0.00) -0.01 (0.02) 0.16*** (0.03) 0.23 (0.14) -0.00 (0.02) -0.08*** (0.02) 0.03 (0.09) 0.13*** (0.02) -0.18*** (0.06) -0.07*** (0.02) -0.09*** (0.01) -0.68*** (0.05) -0.02 (0.06) 0.03 (0.03) -0.04*** (0.01) -0.04** (0.02) -0.02 (0.02) -0.11 (0.13) 2978
(12) Tumor 0.00 (0.00) -0.01*** (0.00) -0.00 (0.00) 0.00 (0.00) -0.00 (0.00) -0.00 (0.00) -0.00 (0.00) -0.00 (0.00) -0.00 (0.00) -0.00 (0.00) -0.01*** (0.00) 0.01 (0.03) 0.01 (0.01) -0.01** (0.00) 0.00 (0.01) 0.01 (0.01) -0.01 (0.01) -0.30** (0.11) -0.00 (0.00) -0.00 (0.00) -0.00 (0.01) 0.00 (0.00) 0.07 (0.06) 2978
(13) Rectum -0.05*** (0.01) -0.06* (0.03) 0.05 (0.04) 0.00* (0.00) -0.00** (0.00) 0.01** (0.00) -0.06*** (0.02) 0.20*** (0.03) 0.09 (0.12) -0.04** (0.02) -0.09*** (0.02) 0.03 (0.12) 0.11*** (0.03) 0.13 (0.15) -0.10*** (0.03) -0.11*** (0.01) 0.03 (0.04) -0.07* (0.04) -0.38*** (0.03) -0.08*** (0.02) -0.08*** (0.02) -0.05* (0.03) -0.25 (0.17) 2978
(14) Musculo. -0.01 (0.01) 0.11** (0.05) -0.16*** (0.05) 0.00*** (0.00) 0.00 (0.00) 0.00 (0.00) -0.08*** (0.02) -0.10*** (0.02) 0.02 (0.07) 0.06* (0.03) -0.14*** (0.02) -0.13 (0.15) -0.07*** (0.02) 0.05 (0.09) -0.10*** (0.03) -0.11*** (0.01) -0.02 (0.03) 0.00 (0.08) -0.05*** (0.02) -0.42*** (0.02) -0.05 (0.03) 0.04* (0.02) 0.01 (0.16) 2978
(15) Varicose -0.00 (0.01) 0.00 (0.02) -0.01 (0.03) 0.00** (0.00) 0.00** (0.00) 0.00 (0.00) -0.01 (0.01) -0.00 (0.01) 0.04 (0.05) -0.01 (0.01) -0.02*** (0.00) 0.12 (0.15) -0.01 (0.01) 0.16* (0.08) -0.01 (0.01) 0.01 (0.01) 0.01 (0.01) -0.06** (0.03) -0.00 (0.01) -0.00 (0.01) -0.20*** (0.02) -0.00 (0.01) -0.09 (0.09) 2978
(16) Nerv. Sys. -0.01 (0.02) -0.04 (0.02) 0.03 (0.05) 0.00*** (0.00) -0.00 (0.00) 0.00 (0.00) -0.02** (0.01) -0.02 (0.02) -0.04 (0.08) 0.02 (0.02) -0.01 (0.03) -0.00 (0.10) -0.01 (0.01) -0.16** (0.07) -0.04* (0.02) -0.02 (0.02) 0.01 (0.03) -0.04 (0.05) -0.07*** (0.01) 0.01 (0.01) -0.04 (0.02) -0.68*** (0.03) -0.05 (0.13) 2978
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border. * p<.10, ** p<.05, *** p<.01
Finally, we replicate Tables 3 and 4 using first differences in respiratory rates, given the association between weight loss and decreases in basal metabolic rates (Leibel et al 1995). Greater relative decreases in the respiratory rates of USCT veterans just south of our border of interest would support the inference that we are in fact observing real weight losses among this population over our study period. As reported in Tables 8 and 9, we do in fact observe greater relative decreases in resting respiratory rates among USCT veterans south of our border of interest. The effect is somewhat weaker than that for weight loss, losing significance in Table 8 in the 100 mile sample, with or without covariates, and in the 30 mile sample with covariates. However, the discontinuous effect within 30 miles of the border is negative and significant at conventional levels in Table 9. 36
Table 8: Individual-Level Change in Respiratory Rates, 10/15/1879 to 10/15/1887
South of Border=1 US Colored Troops=1 South of Border=1 X US Colored Troops=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N USCT North of Border USCT South of Border
Model 1 0.01 (0.27) 0.03 (0.43) -1.25** (0.62) 0.68*** (0.08) 2825.00 50 55
Model 2 0.16 (0.29) 0.35 (0.57) -1.66** (0.74) 0.65*** (0.10) 1731.00 35 42
Model 3 -0.06 (0.31) -0.10 (0.63) -1.12 (0.99) 0.82*** (0.04) 1137.00 29 31
Model 4 0.09 (0.65) 0.35 (0.70) -3.23* (1.73) 0.85*** (0.23) 353.00 19 7
Model 5 -0.05 (0.36) 0.01 (0.44) -1.24** (0.57) 0.01 (0.01) 0.00 (0.00) 0.02 (0.03) -0.27 (0.18) 0.08 (0.29) -0.14 (1.25) -0.21 (0.35) 0.03 (0.18) -1.06 (1.58) 0.34 (0.30) -0.81 (0.90) -0.32 (0.36) -0.03 (0.16) 0.19 (0.34) -0.58 (0.56) -0.35 (0.30) 0.01 (0.21) -0.08 (0.32) -0.23 (0.28) -1.16 (2.07) 2464.00 48 52
Model 6 0.11 (0.39) 0.26 (0.59) -1.59** (0.68) 0.02 (0.01) 0.00 (0.00) 0.03 (0.03) -0.15 (0.20) 0.11 (0.37) 0.90 (1.79) -0.27 (0.52) 0.02 (0.20) -1.13 (1.50) 0.46 (0.35) 0.38 (1.52) -0.56 (0.46) 0.04 (0.17) 0.53 (0.34) -0.43 (0.50) -0.29 (0.38) 0.29 (0.27) -0.24 (0.36) -0.62** (0.27) -2.08 (2.41) 1523.00 34 40
Model 7 -0.08 (0.39) -0.31 (0.56) -0.97 (0.83) 0.02 (0.01) 0.00 (0.01) -0.01 (0.04) -0.21 (0.29) 0.22 (0.45) -0.86 (1.12) -0.51 (0.79) 0.17 (0.34) -1.02* (0.42) 0.49 (0.44) 0.00 (.) -0.86 (0.61) 0.16 (0.21) 0.52 (0.44) -0.21 (0.82) 0.02 (0.39) 0.59 (0.33) -0.36 (0.55) -0.53* (0.27) 0.00 (3.25) 1008.00 28 29
Model 8 -0.38 (0.72) -0.14 (0.81) -2.12 (1.80) 0.02 (0.03) -0.00 (0.01) -0.09 (0.09) -0.13 (1.05) -0.97 (0.82) 0.00 (.) -2.21*** (0.75) 0.01 (0.76) -2.16 (3.01) 1.61** (0.75) 0.00 (.) -1.59 (1.00) 0.46 (0.78) 0.13 (1.02) -1.57 (1.04) -0.25 (0.83) 0.74 (1.00) 0.18 (1.65) -0.79 (0.61) 6.48 (5.72) 313.00 19 7
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border (except models 4 & 8, which are not clustered but restricted to samples 30 miles from border). * p<.10, ** p<.05, *** p<.01 Models 2/6, 3/7 & 4/8 restricted by 200, 100 & 30 miles, respectively.
Table 9: Individual-Level Change in Respiratory Rates, 10/15/1879 to 10/15/1887
30 Mile Dist. Intervals South of Border South of Border=1 X 30 Mile Dist. Intervals USCT=1 USCT=1 X 30 Mile Dist. Intervals South of Border=1 X USCT=1 USCT=1 X Dist. Intervals X South of Border=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N USCT North of Border USCT South of Border
Model 1 0.01 (0.02) 0.24 (0.27) -0.14* (0.07) 0.21 (0.53) 0.04 (0.05) -1.75** (0.79) 0.10 (0.10) 0.71*** (0.10) 2825.00 50 55
Model 2 0.01 (0.02) 0.26 (0.39) -0.17*** (0.06) 0.22 (0.52) 0.05 (0.05) -1.85*** (0.68) 0.14 (0.09) 0.00 (0.01) 0.00 (0.00) 0.02 (0.03) -0.28 (0.18) 0.09 (0.29) -0.10 (1.26) -0.21 (0.36) 0.04 (0.18) -1.08 (1.59) 0.32 (0.30) -0.78 (0.89) -0.33 (0.36) -0.01 (0.16) 0.19 (0.34) -0.60 (0.56) -0.33 (0.30) 0.02 (0.21) -0.08 (0.32) -0.23 (0.27) -1.14 (2.04) 2464.00 48 52
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border. * p<.10, ** p<.05, *** p<.01 38
6 Discussion During the four years prior to the Supreme Court's 1883 strike of the Civil Rights Act of 1875, the Civil War veterans in our samples were, on average, undernourished (Fogel 2004). Some veterans moved in a healthier direction, however. After controlling for preruling covariates, USCT veterans living in states with public accommodations statutes appear to have gained on average somewhere between 4.5-6.5 pounds over the eight-year span bracketing the Court's ruling, relative to their UA neighbors. As reported in Table 2 (and Tables 11 - 13 in the Appendix), USCT veterans were on average heavier than UA veterans during the preruling period, perhaps because those USCT veterans applying for pensions were necessarily unusually advantaged. It is then perhaps not surprising that, starting from a baseline of better preruling health, they continued on this trajectory postruling. UA veterans in the states without public accommodations statutes also gained weight over the period bracketing the Court's ruling, relative to UA veterans in the states with such statutes, averaging a 3-4 pound relative weight gain after conditioning on covariates. Given that this weight gain persists relatively close to our border of interest, it is likely due to statutory variation across this border. It appears to be the case that states just south of this border provided a statutory context of non-racially differentiated public goods that better supported the veterans in our sample, relative to states just north of this border. However, despite sharing a weight advantage equal to if not greater than that held by USCT veterans in the states just to their north, USCT veterans in states without public accommodations statutes did not share in the weight gain of those veterans in the years spanning the Court's strike 39
of the CRA of 1875. And despite living under the same apparently favorable statutory regimes as their UA neighbors, USCT veterans in states without public accommodations statutes also did not share in the weight gain of these neighbor veterans over this same time span. Controlling for preruling covariates, USCT veterans in states without public accommodations statutes failed to gain approximately 6 pounds in body mass that we would have expected them to have gained between 1879 and 1887, had some other factor not intervened. We suggest that this intervening factor was the Supreme Court's strike of the Civil Rights Act of 1875. A final question is whether the post-ruling relative weight loss among USCT veterans south of our border of interest persisted over time. While we lack sufficient observations to follow the veterans in our primary study samples, we can look at measures of average body mass among all USCT veterans in the Early Indicators sample. As can be seen from Table 2 (and Appendix Tables 11 - 13), prior to the Court's ruling USCT veterans south of our border of interest were on average heavier than the USCT veterans north of this border, although the difference is not significant for any of our samples. The post-ruling weight gain among USCT veterans north of the border narrowed the cross-border gap in average USCT body weights, with still no significant differences in average post-ruling weights across any of our samples. Figure 4 reports the effect of living within 30 miles south of our border of interest on average USCT BMIs in the full Early Indicators sample, relative to living within 30 miles north of the border, for five-year intervals between 1875 and 1900.15 Between 1875 and 1890, there are no significant differences in the average BMIs of USCT veterans living just north and just south of 15BMI is calculated as (weight in lbs/(height in inches)2) X 703. All regressions use only those veterans located within 30 miles of our border of interest, and control for age of veteran. 90% confidence intervals are reported. 40
our border of interest. But between 1890 and 1895, a clear gap opens up, with USCT veterans living just south of our border of interest reporting a significantly lower average BMI than their USCT neighbors just north of this border. This gap remains negative and significant between 1895 and 1900. During this latter period, the average BMI for those USCT veterans living just north of the border was 24.0, close to the optimal BMI of 24.1 estimated by Fogel (2004). However, the average BMI of those USCT veterans living just south of the border was only 23.4. These estimates are consistent with the persistence of relative weight loss among USCT veterans just south of our border of interest after the demise of the Civil Rights Act of 1875.16 Figure 4: Plot of Coefficients from Regressions of BMI on South of Border=1 Among USCT Veterans Located Within 30 Miles of Border; 5 Year Intervals 16The emergence of a gap in average USCT BMIs across our border of interest in the 1890s is unlikely to have been due to the enactment of segregating and or disfranchising statutes during this period, as these statutes were largely absent from the upper South until after 1900 (Johnson 1919). 41
Our confidence in our findings is limited by a number of concerns, not least of which is the sparseness of our USCT data. If our findings are eventually bolstered with additional data, however, they will support earlier work indicating that anti-discrimination statutes can in fact have independent effects that are not simply endogenous to more fundamental attitudinal or economic trends (Chay 1998, Chay et al 2009). Our results further suggest that such statutes can cause significant reductions in discriminatory behavior even in the context of perhaps more widespread discriminatory attitudes than existed in the 1960s and early 1970s. Our findings also may suggest the need to reexamine the Supreme Court's role in the deepening of racial inequality in the South in the years after Reconstruction's end. Historians generally agree that the Court's strike of the Civil Rights Act of 1875 had little effect (the Act itself having had little effect). Our findings suggest instead that the Court's strike of the Act set in motion a progressive worsening of the conditions of life for African Americans living in states that did not enact statelevel public accommodations statutes, a category that included 84% of African Americans in 1900. In the words of AME Bishop Henry McNeal Turner, the Court's strike of the Civil Rights Act of 1875 appears to have been a "barbarous decision" that "literally untie[d] the devil" (Redkey 1971, 60). References Acemoglu, Daron and James A. Robinson. 2000. "Why Did Western Europe Extend the Franchise? Democracy, Inequality and Growth in Historical Perspective." Quarterly Journal of Economics 115:1167­1199. Acemoglu, Daron and James A. Robinson. 2006. Economic Origins of Dictatorship and Democracy. New York: Cambridge University Press. 42
Alesina, Alberto and Dani Rodrik. 1994. "Distributive Politics and economic growth." Quarterly Journal of Economics 109:465­490. Alesina, Alberto and Edward Glaeser. 2004. Fighting Poverty in the U.S. and Europe: A World of Difference. New York: Oxford University Press. Chay, Kenneth Y. 1998. "The Impact of Federal Civil Rights Policy on Black Economic Progress: Evidence from the Equal Employment Opportunity Act of 1972." Industrial and Labor Relations Review 51:608­632. Chay, K.Y., J. Guryan and B. Mazumder. N.d. "Birth Cohort and the Black-White Achievement Gap: The Roles of Access and Health Soon After Birth." NBER Working Paper. Forthcoming. Cohen, William. 1991. At Freedom's Edge: Black Mobility and the Southern White Quest for Racial Control 1861-1915. Baton Rouge: Louisiana State University Press. Collins, William J. 2003. "Labor Market Impact of State-Level Anti-discrimination Laws, 19401960." Industrial and Labor Relations Review 56:244­72. Costa, Dora L., A Helmchen and Sven Wilson. 2007. "Race, Infection, and Arteriosclerosis in the Past." Proceedings of the National Academy of Sciences of the United States of America 104(33):13219­13224. Costa, Dora L. and Matthew E. Kahn. 2006. "Forging a New Identity: The Costs and Benefits of Diversity in Civil War Combat Units for Black Slaves and Freemen." Journal of Economic History 66:936­962. Dobbie, Will and Roland G. Freyer. 2014. "The Impact of Attending a School with High-Achieving Peers: Evidence from New York City Exam Schools." American Journal: Applied Economics 6(3):58­75. Floud, Roderick, Robert W. Fogel, Bernard Harris and Sok Chul Hong. 2011. The Changing Body: Health, Nutrition and human development in the Western World Since 1700. New York: Cambridge University Press. Fogel, Robert William. 2004. The Escape from Hunger and Premature Death, 1700-2100: Europe, America and the Third World. New York: Cambridge University Press. Foner, Eric. 1988. Reconstruction: America's Unfinished Revolution 1863-1877. New York: Harper and Row. Foner, Philip S. 1973a. "The Battle to End Discrimination Against Negroes on Philadelphia Streetcars (Part 1): Background and Beginning of the Battle." Pennsylvania History 40(3):260­290. Foner, Philip S. 1973b. "The Battle to End Discrimination Against Negroes on Philadelphia Streetcars (Part II): The Victory." Pennsylvania History 40(4):354­379. Franklin, John Hope. 1974. "The Enforcement of the Civil Rights Act of 1875." Prologue Magazine 6(4):225­235. Gerber, David A. 1976. Black Ohio and the Color Line, 1860-1915. Urbana, IL: University of 43
Illinois Press. Gilette, William. 1979. Retreat from Reconstruction, 1869-1879. Baton Rouge: Louisiana State University Press. Glasson, William. 1918. Federal Military Pensions in the U.S. New York: Oxford University Press. III, John T. Donohue and James Heckman. 1991. "Continuous Versus Episodic Change: The Impact of Civil Rights Policy on the Economic Status of Blacks." Journal of Economic Literature 29(4):1603­1643. Jack, Bryan M. 2007. The St. Louis African American Community and the Exodusters. Columbia, MO: University of Missouri PresS. Johnson, Franklin. 1919. The Development of State Legislation Concerning the Free Negro. New York: Columbia University Press. Kousser, J. Morgan. 1974. The Shaping of Southern Politics: Suffrage Restriction and the Establishment of the One-Party South. New Haven: Yale University Press. Kousser, J. Morgan. 1999. Colorblind Injustice: Minority Voting Rights and the Undoing of the Second Reconstruction. Chapel Hill: University of North Carolina Press. Leibel, R., M. Rosembaum and J. Hirsch. 1995. "Changes in Energy Expenditure Resulting from Altered Body Weight." New England journal of medicine 332:621­628. Locher, Julie L., Christine S. Ritchie, David L. Roth, Patricia Sawyer Baker, Eric V. Bodner and Richard M. Allman. 2005. "Social Isolation, Support and Capital and Nutritional Risk in an Older Sample: Ethnic and Gender Differences." Social Science and Medicine 60(4):747­761. Logue, Larry M. and Peter David Blanck. 2010. Race, Ethnicity and Disability: Veterans and Benefits in Post-Civil War America. New York: Cambridge University Press. Marmot, Michael. 2005. The Status Syndrome: How Social Standing Affects our Health and Longevity. New York: Henry Holt and Co. Neumark, David and Wendy A. Stock. 2006. "The Labor Market Effects of Sex and Race Discrimination Laws." Economic Inquiry 44(3):385­419. Ng, Kenneth and Nancy Virts. 1989. "The Value of Freedom." Journal of Economic History 49:960­961. Novak, Daniel A. 1978. The Wheel of Servitude: Black forced labor After Slavery. Lexington: The University Press of Kentucky. Painter, Nell. 1977. Exodusters: Black Migration to Kansas After Reconstruction. New York: Knopf. Pascoe, Elizabeth A. and Laura Smart Richman. 2009. "Perceived Discrimination and Health: A Meta-Analytic Review." Psychological Bulletin 135(4):531­554. 44
Rabinowitz, Howard N. 1971. Race Relations in the Urban South 1865-1890. Athens: University of Georgia Press. Redkey, Edwin, ed. 1971. Respect Black: The Writings and Speeches of Henry McNeal Turner. New York: Arno Press. Saint-Paul, Gilles and Thierry Verdier. 1993. "Education, Democracy and Growth: Journal of Development Economics." Journal of Development Economics 42(2):399­407. Smith, James P. and Finis R. Welch. 1989. "Black Economic Progress After Myrdal." Journal of Economic Literature 27(2):519­564. Williams, David R. and Selina A. Mohammed. 2009. "Discrimination and Racial Disparities in Health: Evidence and Needed Research." Journal of Behavioral Medicine 32:20­47. Woodward, Vann C. 1957. The Strange Career of Jim Crow. New York: Oxford University Press. Wright, George C. 1985. Life Behind a Veil: Blacks in Louisville, Kentucky 1865-1930. Baton Rouge: Louisiana State University Press. Appendix 45
Table 10: List of Variables
Weight Gain/Loss US Colored Troops = 1 South of Border = 1 Pre Ruling Weight Pre Ruling Height Pre Ruling Age Pre Ruling Respiratory Preruling Diarrhea Preruling Infectious Disease Preruling Cardiovascular Preruling Eye Preruling Endocrine Preruling Gastrointestinal Preruling Kidney Preruling Hernia Preruling Injury Preruling Liver/Spleen/Gallbladder Preruling Tumor Preruling Musculoskeletal Preruling Varicose Veins Preruling Nervous System Preruling Rectum 30 Mile Dist. Intervals
Avg postruling weight (10/15/1883-10/15/1887) - avg preruling weight (10/15/1879-10/15/1883) 0 = Union Army veteran, 1 = United States Colored Troops veteran 0 = States with public accommodations statutes, 1 = States without public accommodations statutes Avg pre ruling weight in lbs Avg pre ruling height in inches Avg pre ruling age Avg pre ruling presence of respiratory disease,where 0 = asymptomatic, 1 = presence of upper or lower resp. disesase, dullness, abnormal resp. sounds, cough, abnormal breathing, pectus carinatum or excavatum Avg preruling presence of diarrhea or dysentery, where 0 = asymptomatic, 1 = presence of diarrhea or dysentery Avg preruling presence of infectious disesase, where 0 = asymptomatic, 1 = presence of current infectious disease Avg preruling presence of cardiovascular disease, where 0 = asymptomatic, 1 = presence of abnormal pulse, palpitations, murmur, abnormal valves, enlarged heart, edema, cyanosis, dyspnea, arteriosclerosis, or impaired circulation Avg preruling presence of eye disease, where 0 = asymptomatic, 1 = presence of conjunctivitis, other eye infection, abnormal cornea, or cataracts Avg preruling presence of endocrine disease, where 0 = asymptomatic, 1 = enlarged thyroid or diabetes Avg preruling presence of gastrointestinal disease, where 0 = asymptomatic, 1 = presence of gastrointestinal disorder, impaired digestion, restricted diet, nausea, vomiting, vomiting blood, enlarged or tender abdomen, or dysphagia Avg preruling presence of kidney disease, where 0 = asymptomatic, 1 = presence of kidney or bladder inflammation, bladder stones, or uremia Avg preruling presence of hernias, where 0 = asymptomatic, 1 = presence of hernia Avg preruling presence of bodily injury, where 0 = asymptomatic, 1 = presence of bodily injury Avg preruling presence of liver, spleen, or gallbladder disease, where 0 = asymptomatic, 1 = presence of enlarged, painful, or atrophied liver, ascites, jaunidce, dark urine, or enlarged or painful spleen or gallbladder Avg preruling presence of tumors, where 0 = asymptomatic, 1 = presence of tumor Avg preruling presence of musculoskeletal disease where 0 = asymptomatic, 1 = presence of rheumatism, sciatica, or spinal abnormality Avg preruling presence of varicose veins, where 0 = asymptomatic, 1 = presence of varicose veins Avg preruling presence of nervous system disease, where 0 = asymptomatic, 1 = presence of paralysis, abnormal reflexes, tremors, neuralgia, numbness, vertigo, headaches, anxiety, epilepsy, memory loss, mental illness, or dependence on others as a result of any of the former conditions Avg preruling presence of rectal disease, where 0 = asymptomatic, 1 = presence of hemorrhoids or abnormal rectum Average veteran distance from border in 30 mile intervals; intervals south of border coded 0 to 50, increasing in distance; intervals north of border coded 0 to -50
Table 11: Covariate Balance, 200 Mile Sample
Dependent Variable Weight Gain/Loss
USCT North USCT South DIFF UA North UA South DIFF
3.81
0.79 3.02
-0.30
2.06 -2.36***
Pre Ruling Covariates Weight Height Age Respiratory Diarrhea Infectious Disease Cardio Eye Endocrine Gastrointestinal Kidney Hernia Injury Liver/Spleen/Gallbladder Tumor Musculoskeletal Varicose Veins Nervous System Rectum N
151.12 67.47 47.57 0.09 0.07 0 0.09 0 0 0.08 0 0.06 0.34 0.01 0 0.33 0.01 0.45 0.08 39
156.14 68.06 43.86 0.14 0 0 0.02 0.02 0 0.07 0 0.02 0.42 0.02 0 0.23 0.01 0.11 0.07 45
-5.02 -0.59 3.71* -0.05 0.07** 0 0.07 -0.02 0 0.01 0 0.04 -0.08 -0.01 0 0.10 0 -0.07 0.01
148.33 68.42 45.34 0.15 0.08 0.00 0.14 0.07 0.00 0.12 0.00 0.04 0.31 0.04 0.01 0.18 0.04 0.10 0.08 1561
147.32 68.54 45.93 0.19 0.02 0 0.07 0.07 0.00 0.08 0 0.05 0.27 0.04 0.01 0.19 0.03 0.11 0.09 186
Ns for Preruling Age are 38 USCT North, 44 USCT South, 1558 UA North, and 186 UA South.
1.01 -0.12 -0.59 -0.04 0.06*** 0.00 0.07*** 0 0.00 0.04 0.00 -0.01 0.04 0 -0.00 -0.01 0.01 -0.01 -0.01
Table 12: Covariate Balance, 100 Mile Sample
Dependent Variable Weight Gain/Loss
USCT North USCT South
4.22
2.24
DIFF UA North UA South DIFF
1.98
-0.68
2.16 -2.84***
Pre Ruling Covariates Weight Height Age Respiratory Diarrhea Infectious Disease Cardio Eye Endocrine Gastrointestinal Kidney Hernia Injury Liver/Spleen/Gallbladder Tumor Musculoskeletal Varicose Veins Nervous System Rectum N
151.58 67.04 47.95 0.11 0.03 0 0.09 0 0 0.06 0 0.08 0.33 0 0 0.39 0 0 0.09 33
156.10 68.40 44.43 0.20 0 0 0.03 0.03 0 0.06 0 0.03 0.36 0 0 0.21 0.01 0.12 0.06 33
-4.52 -1.36** 3.52 -0.09 0.03 0 0.06 -0.03 0 0 0 0.05 -0.03 0 0 0.18 -0.01 -0.12** 0.03
147.82 68.47 45.70 0.15 0.09 0.00 0.14 0.07 0.001 0.12 0.00 0.05 0.32 0.04 0.01 0.17 0.04 0.10 0.08 959
147.93 68.53 46.12 0.21 0.02 0 0.07 0.08 0 0.09 0 0.05 0.25 0.05 0.01 0.18 0.03 0.09 0.08 171
Ns for Preruling Age are 32 USCT North, 32 USCT South, 957 UA North, and 171 UA South.
-0.11 -0.06 -0.42 -0.06* 0.07*** 0.00 0.07*** -0.01 0.001 0.03 0.00 0 0.07* -0.01 -0.00 -0.01 0.01 0.01 -0.00
Table 13: Covariate Balance, 30 Mile Sample
Dependent Variable Weight Gain/Loss
USCT North USCT South
2.11
-1.84
DIFF UA North UA South DIFF
3.95
-0.72
2.86 -3.58**
Pre Ruling Covariates Weight Height Age Respiratory Diarrhea Infectious Disease Cardio Eye Endocrine Gastrointestinal Kidney Hernia Injury Liver/Spleen/Gallbladder Tumor Musculoskeletal Varicose Veins Nervous System Rectum N
152.32 66.60 46.00 0.11 0.02 0 0.09 0 0 0.09 0 0.11 0.27 0 0 0.39 0 0 0.14 22
156.21 67.57 41.36 0.50 0 0 0.14 0 0 0 0 0 0.29 0 0 0.29 0 0.21 0 7
-3.89 -0.97 4.64 -0.39** 0.02 0 -0.05 0 0 0.09 0 0.11 -0.02 0 0 0.10 0 -0.21*** 0.14
148.17 68.47 46.01 0.15 0.08 0 0.14 0.08 0.00 0.10 0.00 0.06 0.32 0.05 0.01 0.19 0.02 0.12 0.10 290
150.10 68.37 46.64 0.19 0.02 0 0.10 0.11 0 0.17 0 0.04 0.22 0.06 0 0.13 0.03 0.09 0.06 57
Ns for Preruling Age are 22 USCT North, 7 USCT South, 288 UA North, and 57 UA South.
-0.93 0.10 -0.63 -0.04 0.06* 0 0.04 -0.03 0.00 -0.07 0.00 0.02 0.10 -0.01 0.01 0.06 -0.01 0.03 0.04
48
Table 14: Individual-Level Weight Gain/Loss, 10/15/1879 to 10/15/1887, Excluding Those Residing on Other Side of Border In 1860 Census
South of Border=1 US Colored Troops=1 South of Border=1 X US Colored Troops=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N USCT North of Border USCT South of Border
Model 1 1.43** (0.66) 3.23*** (1.20) -4.57* (2.44) -0.09 (0.23) 2936 56 58
Model 2 1.88** (0.73) 4.10*** (1.25) -4.90* (2.58) -0.29 (0.24) 1785 39 45
Model 3 2.33** (0.71) 4.90** (1.86) -4.31 (2.54) -0.68*** (0.14) 1157 33 33
Model 4 3.47** (1.69) 2.88 (2.12) -7.43* (4.31) -0.77 (0.61) 357 22 7
Model 5 2.03*** (0.66) 3.76*** (1.39) -4.37* (2.21) -0.11*** (0.02) -0.07** (0.03) 0.11 (0.10) -0.06 (0.56) 0.52 (0.62) -3.30 (3.52) 1.26** (0.58) 1.22 (0.95) -7.83*** (1.79) -0.65 (0.44) 9.13*** (2.93) 1.27 (0.94) 1.28** (0.56) 1.92 (1.38) 1.04 (3.01) 0.38 (0.60) 0.43 (0.66) -0.50 (0.88) 0.25 (0.76) 6.59 (5.44) 2580 54 56
Model 6 2.61*** (0.67) 5.09*** (1.41) -5.89** (2.48) -0.14*** (0.02) -0.08* (0.04) 0.08 (0.16) -0.42 (0.68) 0.36 (0.75) -5.18 (5.06) 0.24 (0.74) -0.01 (0.93) -7.75*** (2.50) -0.24 (0.45) 2.07 (1.68) -0.01 (0.80) 0.09 (0.51) 3.18 (2.28) 1.18 (4.33) -0.36 (0.72) -0.73 (0.65) -0.93 (1.36) 0.33 (1.26) 11.78 (8.23) 1581 38 44
Model 7 3.05*** (0.63) 6.69** (2.13) -5.65* (2.50) -0.14*** (0.02) -0.11* (0.06) 0.28 (0.19) -1.01 (1.11) -0.07 (0.67) 7.14*** (1.48) 0.08 (1.01) 0.76 (1.01) -12.16** (3.83) -0.41 (0.64) 4.95*** (1.28) -0.61 (0.87) -0.40 (0.37) 1.56 (3.44) -0.07 (6.85) -0.77 (0.67) -1.10 (0.88) 0.05 (1.70) 0.84 (1.65) 3.11 (8.92) 1035 32 32
Model 8 4.04** (1.71) 4.77** (2.24) -8.80* (4.49) -0.15** (0.07) 0.01 (0.04) 0.04 (0.20) -2.37 (1.69) 1.64 (1.47) 0.00 (.) 1.17 (1.41) 3.44 (2.26) -6.40 (5.89) -1.05 (1.49) 0.00 (.) 0.79 (2.50) -0.26 (1.56) 9.54*** (3.03) 3.94 (4.00) 0.76 (1.46) -1.62 (1.84) 3.36* (1.84) 4.02** (1.71) 1.46 (12.40) 317 22 7
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border (except models 4 & 8, which are not clustered but restricted to samples 30 miles from border). * p<.10, ** p<.05, *** p<.01 Models 2/6, 3/7 & 4/8 restricted by 200, 100 & 30 miles, respectively.
49
Table 15: Individual-Level Weight Gain/Loss, 10/15/1879 to 10/15/1887, Excluding Those Residing on Other Side of Border In 1860 Census
30 Mile Dist. Intervals South of Border =1 South of Border=1 X 30 Mile Dist. Intervals USCT=1 USCT=1 X 30 Mile Dist. Intervals South of Border=1 X USCT=1 USCT=1 X Dist. Intervals X South of Border=1 Pre Ruling Age Pre Ruling Weight Pre Ruling Height Pre Ruling Abnormal Resp. Pre Ruling Diarrhea Pre Ruling Infectious Disease Pre Ruling Cardio Pre Ruling Eye Pre Ruling Endocrine Pre Ruling Gastrointestinal Pre Ruling Kidney Pre Ruling Hernia Pre Ruling Injury Pre Ruling Liver/Spleen/Gallbladder Pre Ruling Tumor Pre Ruling Rectum Pre Ruling Musculoskeletal Pre Ruling. Varicose Veins Pre Ruling Nervous System Constant N USCT North of Border USCT South of Border
Model 1 -0.08* (0.04) 2.47*** (0.66) -0.23** (0.10) 3.75*** (1.15) 0.09 (0.25) -3.64 (2.76) -0.39 (0.66) -0.57*** (0.20) 2936 56 58
Model 2 -0.09** (0.04) 3.30*** (0.53) -0.26** (0.11) 4.47*** (1.30) 0.13 (0.27) -4.66* (2.50) -0.11 (0.52) -0.11*** (0.02) -0.07** (0.03) 0.12 (0.11) -0.05 (0.57) 0.54 (0.61) -3.66 (3.49) 1.34** (0.59) 1.36 (0.95) -7.41*** (1.78) -0.69 (0.44) 8.82*** (2.77) 1.17 (0.95) 1.29** (0.56) 1.80 (1.39) 1.08 (3.03) 0.43 (0.61) 0.41 (0.66) -0.60 (0.87) 0.18 (0.78) 5.92 (5.76) 2580 54 56
All results are from OLS regressions. Robust standard errors clustered at 30 mile intervals from border. * p<.10, ** p<.05, *** p<.01 50

A Harvey, EA West

File: the-deadest-of-dead-letters-reevaluating-the-impact-of-the-civil.pdf
Author: A Harvey, EA West
Published: Mon May 25 20:12:26 2015
Pages: 50
File size: 1.29 Mb


La fantasy, 72 pages, 0.47 Mb

Secret buy-backs of LDC debt, 19 pages, 0.56 Mb
Copyright © 2018 doc.uments.com