name: inverse layout: true class: center, middle, inverse --- #Union Veterans, ##African American Suffrage, ##and Reconstruction Michael Weaver University of Chicago September 1, 2016 --- ##Iowa in the 19th Century -- ### Constitution included "black" codes -- ### Passed exclusion law (1851) -- ### Resoundingly rejected black suffrage (1857) --- ##1868 ###Iowa passed referendum extending suffrage to African Americans --- ##How did this transformation take place? ###Racism deeply embedded in American Society ###Opposition to slavery ≠ Support for equal rights --- ##How did many Americans come to support a revolutionary party that -- #### Maintained a state of war until 1871 #### Imposed military rule in the South -- #### Ended slavery with the 13th Amendment (and the Army) #### Extended civil and political rights by law and by force --- -- ##The Civil War --- ##Veterans --- layout:false .left-column[##Veterans ###Postwar ] .right-column[ ###Veterans a large share of the electorate More than two million Northern men served Restrictions on suffrage magnified their influence ###Civil War veterans and pensions Veterans active in early politics of welfare state (Skocpol 1995) ] --- layout:false .left-column[##Veterans ###Postwar ###Evidence ] .right-column[ ### Veterans shape politics after returning home (Jha and Wilkinson 2012, Bellows and Miguel 2009) ### Veterans are more politically active (Teigen 2006, Blattman 2009) ### Combat veterans adopt "hawkish" positions (Grossman et al 2015, Koenin 2015) ] --- layout:false .left-column[##Veterans ###Postwar ###Evidence ###Expectations ] .right-column[ ###Oppose Suffrage ####Racism Contemporaries expected veterans to return *more* racist ###Support Suffrage ####Combat experience * Made veterans opposed to pre-war status quo in the South * Suffrage makes "loyal" constituency in the South. ####Inter-group contact * Service exposed soldiers to many more African Americans ] --- template:inverse ## Data and Design --- layout:false .left-column[##Data ###Why Iowa] .right-column[ ### Good starting point 1. Veterans listed by residence 2. Good data on elections, eligible voters 3. Suffrage votes before and after election ] --- layout:false .left-column[##Data ###Why Iowa ###Veterans ] .right-column[ ### Civil War Data Base * Relational database of soldiers, units, battles * 4 million+ soldiers, varying levels of personal data * Iowa units recorded pre-war residence * Possible to match to census by name, age, and other identifiers ### Measurement * County-level * Number of Veterans / Number of military-age males in 1860 ] --- layout:false .left-column[##Design ###Panel ] .right-column[ ### Differenced `$$Suffrage1868_i - Suffrage1857_i = \alpha + \beta * VeteranPercent_i + \epsilon_i$$` ### Lagged DV `$$Suffrage1868_i = \alpha + \beta * VeteranPercent_i + \gamma * Suffrage1857_i + \epsilon_i$$` ] --- template:inverse ##Results: --- layout:false .left-column[##Analysis ###Results ] .right-column[ ####1. More veterans is associated with *more* support for suffrage ] --- class: center, middle
Suffrage Voteshare
Δ Suffrage Voteshare
(1)
(2)
(3)
(4)
Suffrage 1857
0.674
***
0.675
***
(0.109)
(0.113)
Veterans
0.242
0.303
*
(0.138)
(0.141)
Surviving Veterans
0.302
*
0.389
**
(0.142)
(0.148)
Constant
0.401
***
0.401
***
0.342
***
0.340
***
(0.047)
(0.040)
(0.045)
(0.039)
Lagged DV
yes
yes
no
no
Differenced
no
no
yes
yes
N
80
80
80
80
R
2
0.486
0.479
0.141
0.132
Adjusted R
2
0.473
0.466
0.130
0.121
*
p < .05;
**
p < .01;
***
p < .001
--- layout:false .left-column[##Analysis ###Results ] .right-column[ 1. More veterans is associated with *more* support for suffrage 2. This effect is conditional on combat experience ] --- class: center, middle Marginal Effects, Linear interaction
--- class: center, middle Marginal Effects, Kernel estimator
--- layout:false .left-column[##Analysis ###Results ] .right-column[ 1. More veterans is associated with postwar *increase* in votes for suffrage 2. This effect is conditional on combat experience 3. More veterans associated with postwar *increase* in votes for GOP ] --- class:center, middle
Republican Voteshare
(1)
(2)
(3)
Lagged Republican Voteshare
0.803
***
(0.018)
Post War x Veterans (%)
0.269
***
0.409
*
0.124
***
(0.039)
(0.172)
(0.033)
Lagged DV
no
no
yes
County FEs
yes
yes
no
Year FEs
no
yes
yes
N
2,370
2,370
2,278
R
2
0.634
0.718
0.741
Adjusted R
2
0.619
0.703
0.738
*
p < .05;
**
p < .01;
***
p < .001
--- layout:false .left-column[##Analysis ###Results ] .right-column[ 1. More veterans is associated with postwar *increase* in votes for suffrage 2. This effect is conditional on combat experience 3. More veterans associated with postwar *increase* in votes for GOP 4. Parallel trends assumption holds: identified? ] --- class: center, middle Marginal effect of veterans by election year
--- layout:false .left-column[##Analysis ###Results ] .right-column[ 1. More veterans is associated with postwar *increase* in votes for suffrage 2. This effect is conditional on combat experience 3. More veterans associated with postwar *increase* in votes for GOP 4. Parallel trends assumption holds: identified? ] --- template:inverse ##What's next? --- .left-column[ ##Next step ###Measurement ] .right-column[ 1. Match veterans to pre- and post-war counties * By residence * Using Census schedules 2. Instrument for enlistment rates? * Draft quotas? ] --- .left-column[ ##Next step ###Measurement ###Analysis ] .right-column[ 1. Replicate analyses for suffrage votes in WI, NY 2. Extend analysis of GOP voteshare to all Northern States 3. Effects on composition of Reconstruction Congress? ] --- template:inverse #Thank you