SSRN Id3925085
SSRN Id3925085
SSRN Id3925085
OF ECONOMICS
VINCENT GELOSO
AND
LINAN PENG
Abstract
After the Civil War, the Democratic party carried an important electoral penalty from
being associated with the war. To deal with this penalty, the party took increasingly anti-
immigration positions to compete with Republicans. This led some Republican strongholds
such as California to become competitive and also forced Republicans to embrace stricter
immigration proposals. In this paper, we argue that adopting anti-immigration and raising
awareness against immigration made California increasingly competitive in electoral terms.
This electoral competitiveness can serve to explain the genesis of the 1882 Chinese Exclu-
sion Act.
* Assistant professor of economics at George Mason University Email: vgeloso@gmu.edu Address: Department
of Economics, George Mason University, MS 3G4, Fairfax, VA 22020, USA.
†
Assistant professor of economics at DePauw University Email: linanpeng@depauw.edu.
Economic historians and historians have had a longstanding interest for the topic of immigration
in postbellum America. There are two reasons for this interest. The first is that the volume of
migrants was historically unprecedented – more than 22 million foreigners arrived between 1865
and 1910 (Bohanon and Van Cott, 2005, p. 529). The second is that there was a strong nativist
backlash against immigrants that culminated in a series of legislative actions against Asian
immigrants first (in the 1880s) and later against eastern Europeans (from the 1890s onward)
(Timmer and Williams, 1998). To explain the backlash, economists and historians have tended
to emphasize how native workers perceived the effects that immigration had on their wages
(Goldin, 1994) who saw incentives to mobilize politically to find ways to reduce immigration
(Wong et al., 2018).
However, there is a problem with that narrative. Large numbers of immigrants were entering
the United States as early as the 1840s and the perceived effects on wages were present in the
1850s (Margo, 2000, p. 144). The nativist backlash was immediate (Cohn, 2000) but it never
translated into any form of anti-immigration policy at the federal level until the 1880s. What
policies were adopted were adopted at the state-level. And even the most extreme policies
adopted at that level were relatively modest in scope and effect. Why are there close to four
decades separating the begininning of a nativist backlash and the adoption of anti-immigration
rules in 1882, most notably the Chinese Exclusion Act (Fong and Markham, 1991, 2002; Lee,
2002; Seo, 2011; Peng, 2021) which was “the first law in American history to ban any group of
people from entering America solely based upon race" (Seo, 2011, p. 826). Our answer is that
the Civil War altered political incentives enough to make immigration a potential wedge-issue
in swing states.
Following the end of the Civil War, the Democrats were stuck with a major electoral disad-
vantage. States that had contributed large contingents of Union soldiers were unlikely to vote
for the Democrats – a party associated with the initiation of the war. As a result, a large block of
demographically important states were locked in for the Republicans. Michigan, Pennsylvania,
Immigration to America picked up noticeably during the 1840s and 1850s (Davis et al., 1972;
Cohn, 2000; Margo, 2000). In states like New York and Massachusetts, an almost immediate
nativist backlash was directed at Irish and German immigrants while states like California
directed its backlash at the Chinese (Fong and Markham, 1991, 2002; Fisher and Fisher, 2001).
Generally, this took the form of sporadic outbreaks of mob violence and harassment against
immigrants. However, translating this backlash into policy was much more difficult. Actions
prior to 1880 were contained at the state-level, were of limited scope and had limited effects. It
is only in the 1870s that federal legislation became politically possible and only in 1882 that a
strong anti-immigration piece of legislation – the Chinese Exclusion Act – is adopted.
The backlash of the 1840s and 1850s is quite illustrative in terms of understanding the lag
between the backlash and policy action. This period is marked by the rise of the Know-Nothing
Party in the 1850s which was able, by 1855, to control virtually every New England state and
secure large delegations in Congress (Cohn, 2000, pp. 361-362, 374). Unsurprisingly, such
electoral successes led to the adoption of state-level anti-immigration policies. However, such
laws always a limited scope and yielded equally limited effects.2 The most extreme example is
that of Massachusetts. The state deported foreigners who sought help from almshouses (Hirota,
2016, p. 1). This was mostly aimed at Irish immigrants. However, only 715 of the 33,436 Irish
passengers who arrived to Massachusetts between 1837 and 1845 were deported back to Ireland
(Hirota, 2016, p. 215) even though it also deported larger numbers to other American states.
In the 1850s, Massachusetts adopted even stricter deportation criteria so that, from 1851 to
1863, some 15,438 persons were deported back to England, Canada or other US states (Hirota,
2
One example of state laws of limited scope is that of state passenger laws which allowed inspectors commis-
sioned by the state government to refuse entry to people deemed to exhibit a mental or physical defect (Hirota,
2016, p.3).
Our argument above essentially predicts that Democrats moved first on anti-immigration poli-
cies and that, upon realizing the electoral threat, Republicans followed later. The problem
with this argument is that white immigrants could vote which limited the gains from advancing
nativist rhetoric. This is why we focus on California.
First, the state had a particularly strong nativist backlash against Asian immigrants (who
were labelled as Chinese). Second, the state was also increasingly important in terms of elec-
toral college votes. In fact, it handed its six electoral college votes over a 1.8% margin to the
Republican candidate in the 1876 election. The Republican candidate, Rutheford B. Hayes,
had won the election with 185 electoral college votes against 184 for his opponent. California
switching sides would have cost Hayes the victory.
Third, the state fits the pre-war narrative described in section 2. Peng (2021) notes that anti-
Tilden (presidential candidate in 1876) were all from New York. Winfield Hancock (the 1880 presidential candidate),
although his home state was Pennsylvania, was a resident of New York during the 1880 election.
(1)
0 0
|margin|i = Chinese/W hitei δ + xi β + i
10
For example, during the years when the transcontinental railroads were being completed, Chinese workers
were employed on railroad construction outside California. When they were completed in 1869, numerous workers
returned to California at the same time that the state’s manufacturing sector was exposed to greater competition
from east coast manufacturers. This renewed calls for action against the Chinese (Coolidge, 1909).
11
The Workingmen’s Party of California, the most influential labor organization that affects the politics of
California in the late 1870s, provides a good example of this. After seizing control of the legislature in the late
1870s, the Workingmen’s Party rewrote the state constitution. The 19th article of the revised constitution prohibited
municipal works and corporations from hiring Chinese and authorizes cities to remove Chinese from within the city
limits to specified areas. The U.S. Circuit Court declared the new state constitution unconstitutional in violation of
the 14th amendment and the Burlingame Treaty in 1880.
12
This can be best in a referendum in 1879 against more Chinese immigration which passed with the over-
whelming majority of 154,638 to 883 (United States Immigration Commission, 1911, p. 73). With such low counts
of pro-immigration votes in a state with 10% of its population being of Chinese origin, it is clear that the Chinese
did not influence electoral outcomes as well.
both the state and federal levels (Stanley, 1973, p. 213). This was in line with the post-war
political ascendancy of the Radical Republicans who had pushed for the Civil Rights Act of
1866 and the 14th Amendment to the United States Constitution in 1868 (Fisher and Fisher,
2001). In contrast, Democrats, such as Henry Huntly Haight who was elected governor in 1867
with a strong majority, had consistently promoted strong anti-Chinese policies (Stanley, 1973,
p. 214). According to Stanley (1973), the late 1860s – marked by a near Democratic victory
in the state during the presidential election of 1868 and a 11-point victory by Haight in the
1867 gubernatorial election – constitute the reversal point for Republicans. At both the federal
and state levels, Republicans began adopting increasingly anti-Chinese policies. By 1871, “racial
issues no longer distinguished Republicans from Democrats" in state politics (Stanley, 1973,
p. 215). By 1877, the same could be said to apply to federal politics as both parties adopted
highly similar policies (Fisher and Fisher, 2001, p. 59).13 Because Democrats were faster to jump
13
Quite tellingly, in the 1880 election (where both major candidates had adopted anti-Chinese planks), there
was a major scandal in California which probably threw to the state to the Democratic presidential candidate. In
October 1880, a forged letter purportedly sent by James Garfield (the Republican candidate) to an “Employers
4 Results
The results from the regressions confirm our intuition to some extent: the ratio of Chinese
to White people in a county makes that county more electorally competitive in presidential
election. This applies for all elections.
First, we ran multivariate regressions in OLS with each of the different elections. The results
are depicted in table 2 below. Each pair of column represents an election with two different
specifications. The difference in each pair is that one column relies on the log of manufacturing
output (capturing the marginal product of labor in manufacturing) and the other measures how
important was the agricultural sector relative to manufacturing sector (with the assumption that
the Chinese were disproportionately engaged in manufacturing). As can be seen, all columns
report that the Chinese to White ratio reduces the absolute value of the victory margin in a
county. The effect is always significant above the 10% level and is often significant above the 5%
level.
Most importantly, the coefficients are quite large. As the ratio is expressed in decimal points
(i.e. one Chinese to ten Whites is 0.1), our coefficient imply that a one extra Chinese per 100
whites reduced the margin of victory by between 0.135 and 0.352 percentage points. To provide
context to those proportions, the election of 1876 could have tipped to the Democrats had they
been able to get less than 1% of the California’s republican voters to switch sides.
Union" was reproduced in a Democrat newspaper. The letter showed Garfield essentially arguing that his party’s
plank was a decoy to win California even though he was secretly favorable to Chinese immigration (Arrington,
2020). The letter, which is widely believed to be a forgery, can be seen as an attempt to essentially distinguish
between two candidates who held the same view in the electorally crucial state of California.
10
Observations 52 52 52 52 48 48 48 48
R-squared 0.073 0.055 0.080 0.076 0.122 0.132 0.289 0.224
Figure 1: Effect of Chinese/White Ratio on Absolute Victory Margin
-15
-20
-25
-30
-35
The evolution of the coefficient for the ratio of Chinese to White is also consistent with our
explanation. Its importance diminishes monotonically from election to election. This can be
seen in figure 1. The line labelled “Effect (1)" refers to the coefficients for Chinese to White ratio
from the odd-numbered columns of table 2 while “Effect (2)" shows the even-numbered columns
of table 2. As we pointed out above, we expect that Democrats initially got important returns
from adopting anti-Chinese policies. However, as Republicans realized the electoral cost of
appearing pro-Chinese in California, they adopted increasingly hostile proposals. This blurred
the difference between them on this key topic so that it minimized the electoral advantage
conferred by endorsing anti-Chinese policies.
More importantly, the results depicted in figure 1 suggest that bipartisan support for anti-
Chinese policies in the late 1870s – which is well documented by historians such as Fong and
Markham (2002) and Tichenor (2009) – was born out of political needs. More precisely, the
need to negate the electoral advantage of the early mover (i.e. the Democrats) on anti-Chinese
12
5 Conclusion
In this paper, we used the state of California to explain why there were many decades separating
the onset of a nativist backlash against immigrants and the adoption of federal immigration
control policies. We argued that pre-war politics came with incentives that made politicians
unwilling to engage in national-level policies to deal with immigration. The Civil War altered
those incentives by pushing Democrats to attempt to create a West-South alliance to secure
presidential victories. California was seen as a key state to capture and adopting anti-Chinese
policy proposals was the way to do so (especially as Chinese immigrants could not participate
in elections). As Republicans saw the inroads of Democrats in California and realized that it
could cost them the presidency, they began adopting very similar anti-Chinese policies.
Using the county-level results, we found that counties with large Chinese populations were
more electorally competitive all else being equal. Moreover, and consistent with the fact that
Democrats were the first mover, the effect diminishes over time as Republicans began to adopt
anti-Chinese proposals. These results provide evidence for the narrative that we propose
whereby the Civil War altered incentives in ways that made possible the adoption of immi-
gration control policies.
Our results and arguments should be seen as complementary to those already advanced by
historians and economic historians rather than as substitutes. For example, Hirota (2016) argues
that the few northeastern states such as New York and Massachusetts that had strong nativist
13
14
15
16