DEMOGRAPHICS AND HIGH EXECUTION RATES
IN THE EARLY TWENTIETH CENTURY
Peggy Dianne Ray
B.S., University of Colorado, 1972
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Arts
Ray, Peggy Dianne (M.A., Sociology)
Demographics and High Execution Rates in the Early Twentieth Century
Thesis directed by Professor Karl Flaming
Between 1900 and 1940 the execution rate in this country was higher than
during any other period since executions were recorded beginning in 1608. This
study examined the effect of demographic changes on execution rates during this
period and for a sample of 10 states. The sample slates included all states from the
Northeast and Middle Atlantic regions and one southern state. These states were
selected because of a history of high immigration, growing population, and high
executions. Data was collected from a data set including variables measuring all
executions authorized by civil authority between 1608 and 1991. This data was added
to data measuring population characteristics from historical population data to create
a secondary' data set. The effect of the independent variables, native-born while
population and migration, foreign-bom white (immigrant) population and migration,
and black population and migration were tested against the dependent variable
execution rates. Using multiple regression, black migration was the only independent
variable having statistical significance and predicting an increase in the execution
This abstract accurately represents the content of the candidates thesis. I recommend
In memory of my father, Floyd James Ray, who died July 16, 1996.
From him I learned the importance of an inquisitive mind,
the value of integrity, and the necessity of tenacity.
I wish to thank Dr. Karl Flaming for his assistance during graduate school and the
writing of my thesis. I also warmly thank my friends and mentors without whose
support this would have beenan even more difficult task: Lloyd, Margaret, Roy, and
my sister Jennifer.
2. REVIEW OF THE LITERATURE......................................4
An Overview of Macro Sociology Theory......................4
Social Solidarity and Social Order...................4
Social Density, Punishment, and Ritual...............7
Anomie and Deviance..................................8
Thesis Statement and Hypotheses.....................16
Results of Pearsonian Correlation, r, of Independent
Variables and Dependent Variable Execution Rates............22
Results of Multiple Regression of Independent Variables
and Dependent Variable Execution Rates......................24
Charts Illustrating Demographics and Executions 1900-1940.25
Limitations and Future Studies....................................39
4.1 Intercensal migration....................................................26
4.2 Total population, 1900-1940............................................. 27
4.3 Composition of population, by race.......................................28
4.4 Execution by race........................................................29
4.5 Comparison of blacks and whiles executed.................................30
4.6 Total migration vs. total executions.....................................31
4.7 Total migration vs. total executions, log-log chart .....................32
4.8 Total executions vs. migration, by subgroup..............................33
4.9 Migration vs. executions, log-log chart .................................33
4.10 Number of executions compared to population size.........................34
4.11 Number of executions compared to population size, log-log chart..........35
4.12 Execution vs. population, by subgroup...................................35
4.13 Total executions vs. total population, log-log chart ....................36
3.1. Sample states by region...........................................18
The death penalty has been legally available for the punishment of crime since
before the United States became an independent country. However, the number of
executions and the list of crimes punishable by death have varied widely over time.
In the early part of the twentieth century, between the years 1900 and 1940, more
people were executed than in any other period in United Stales history. About 377r
(5,467 people) of all executions authorized by civil authority1 in this country
(between 1608 and 1991) occurred in these 40 years. In 1935, the number of
executions peaked w hen 197 people were put to death. Afterwards, both the per
capita and absolute execution numbers gradually declined to zero in the year 1968.
Four years later, in 1972, the United States Supreme Court imposed a temporary
death penalty ban.2
Once the Supreme Courts ban ended in 1976,3 the number of executions
started to rise, a trend opposite to that in other industrialized countries (Zimring and
Hawkins 1986). Between the 1976 reinstatement and 1996,313 executions occurred
in the United States. More than 3,100 inmates are currently on death row.
Executions have been increasing steadily, rising from two executions between 1977
and 1982 to fifty-six in 1995 (Sandokan 1996).
To what extent can historical demographic and execution statistics explain the
high number of executions between the years 1900 and 1940?
This is a quantitative study using historical data to examine relationships
between demographic change and execution rates during the early 1900s. This period
had historically high rates of internal migration, immigration, and executions. I will
determine to what extent the high concurrent rate of demographic change explains the
high execution rate between 1900 and 1940. I expect to find that population shifts,
such as immigration or internal migration, created social disorder which led to the
imposition of stricter criminal penalties including the death penalty.
It may be possible to draw some comparisons between demographics and use
of capital punishment at the beginning of the century with what is happening now, but
this is not a principal goal of this paper.
Capital punishment is controversial. The use of capital punishment is on the
upswing. Modem day proponents of capital punishment claim that its use is an
appropriate reaction to the problems of violent crime. Data exist on the
disproportionate high number of minorities who are on death row and are executed
(Death Penalty Information Center 1996). Moral, economic, and discrimination
issues surround the on-going capital punishment debate which receives almost daily
attention from newspapers, books, and movies. Politicians talk about deterrence,
retribution, alternate costs of execution and life imprisonment, and even public
hangings. Victims and offenders rights are concerns of communities, policy
makers, civil rights organizations, and the courts.
There have been other periods in American history when there have been
similar kinds of social disruptions and increased implementation of the death penalty.
I thought it would be of interest to see if there is a statistical relationship between
rapid demographic change and the use of the death penalty. It is easier to be
dispassionate and unbiased about historical statistics. An additional advantage of
historical data is that the cycle has played out. We can investigate what was
happening between 1900 and 1940. We do not know the ramifications of what is
happening today or what the future will be.
REVIEW OF THE LITERATURE
In this study, I am looking at possible effects of demographic changes on the
number of executions performed in this country during a specific period, 1900-1940.
I have chosen to study this subject from a macro sociological position. I do not intend
to minimize the importance of individuals, politics or history. I am including some
history of the United States from the late nineteenth and early twentieth centuries and
history of immigration in America. However, since I am doing a quantitative analysis
of hard census data, these are soft facts useful only as background.
An Overv iew' of Macro Sociology Theory
I open my literature review' with an overview of macro sociology theory that
is associated w'ith social systems, social order, and solidarity.
Social Solidarity and Social Order
Emile Durkheim was concerned with social solidarity and social order. He
believed that in primitive societies, legal codes are repressive and sanctions are
punitive: In complex societies, legal codes should become less punitive and more
restitutive (Durkheim  1984). It follows that the number of executions should
drop as societies become less primitive and more complicated. The rush of
immigrants produced disorder in society. One example of this disorder was the
rioting between organized labor and immigrant labor (Stephenson 1964). The
newcomers, while they all looked alike to the natives, were assuredly different from
native-bom Americans; they became the others. Perhaps society experienced a
collective yearning for simpler or more primitive times (the good old days)
expressed in part by social sanctions and an adherence to stricter criminal
punishment. In the decades between 1950 and 1970, immigration radically decreased
in the United States (U.S. Immigration and Naturalization Service 1996) and in the
1960s the number of executions declined to zero (Espy and Smykla 1994). Second
and third generation immigrants had become more familiar by assimilating the
characteristics of the host country. It appeared that this source of disorder had
disappeared. Immigration numbers were low and executions were not being
performed. By the 1990s, immigration was once again increasing, legislation
restricting benefits to immigrants once again was being passed, and executions once
again were increasing. American societys response to the disorder resulting from
change was to use legal sanctions to control the immigrant once admitted to the
country. Legal codes and societys response to them are significant According to
...Durkheim [legal codes] are the best empirical indicator of solidarity.... (Turner,
Beeghley and Powers 1989, p. 315).
Talcott Parsons grand theory about social control and equilibrium within
social systems also pertains to my thesis. Parsons states that when ...from whatever
source, a disturbance [a change or strain] is introduced in the [social]
system...disequilibrium may arise... (Wallace and Wolf 1991, p. 43). Parsons does
not specify the source of strainit could come from natural disasters, wars (e.g.,
World War I and II), economic disruptions, shifting population, internal migration, or
crime. Strain could come from the new presence of others in the social system. He
says that when society is disturbed, socialization and social control emerge to restore
equilibrium. Parsons believes that courts and legal institutions in the United States
function as social controls to restore equilibrium and social order, w hile Durkheim
views legal codes as the best indicators of social solidarity (Wallace and Wolf 1991).
When others enter a new community, society may be disrupted for a number
of reasons, including communication difficulties as a result of language limitations,
cultural differences, and different values and norms. The newcomers or others arc
viewed as intruders because they are different. Stephenson (1964) believ es these
newcomers displace or lower the market wage for native-born white laborers, require
additional social service spending by local governments, and may need bilingual
education for their children. This situation creates conflict and a state of disorder,
perceived or otherwise, which part of the existing citizenry blames on the newcomers.
In the context of this study I am defining others as a group of immigrants or
aliens whose presence is noticed in a community or society. The others include
groups w ith distinct ethnic and racial characteristics including foreign-bom w'hiles
(immigrants), native-born whites and native-born blacks, groups from rural areas
whether foreign-bom or native-born, and any group whose presence is noticeable by
its size. A definitive statement of what constitutes the other is missing in sociological
literature. However, Erving Goffman does describes a norm in American society (a
male) by which all others compares themselves. He is ...young, married, white,
urban, northern, heterosexual Protestant father of college education, fully employed,
of good complexion, weight, and height, and [has] a recent record in sports
(Goffman 1963, p.128). This interpretation, although noteworthy, is limiting for the
scope of this study. Historically, America is a diverse society. Arturo Madrid (1988)
defines others as those who are different, marginalized, alienated or excluded.
Names, pigmentation, physiognomy, and accents are all attributes of otherness.
Conflict theory' provides another line of reasoning for my thesis. Lewis Coser
incorporates conflict theory' within the functionalist perspective as a support of social
order (Collins 1994). He believes external conflict strengthens boundaries of the
existing group ...introducing a strong negative reference group to which the...
[existing group] contrasts] themselves... (Wallace and Wolf 1991, p.153). As new
groups moved into communities, they were defined as the out-group by the citizen
group already in the community, the in-group. The new groups constituted an
external source of conflict that put a strain on the social system. This conflict
intensified and strengthened the in-group and determined behavior toward the new
groups. Norms and sanctions of the host culture dominated society (Taylor and
Moghaddam 1994). In the early 1900s, the in-groups were the native-bom white
English-speaking Americans; the out-groups were blacks, migrants from rural or
urban areas, and the second major immigrant stream of foreign-bom whites.
Social Density. Punishment, and Ritual
Variations in social density are caused by changes in society: growth of
population, urban immigration, technology, and communication (Turner, Beeghley
and Powers 1989). Durkheim saw migration, population growth, and ecological
concentration as causing increased material density, which in turned caused
increased moral...density...that...escalated social contact and interaction (p.318).
Increases in urban crowding may be reflected in incidences of moral density.
Individuals or groups seek enactment of ritual as a means of managing external stress.
In western, industrialized societies with twentieth century institutions Durkheims
collective consciousness is expressed...[through]...death penalty legislation [and] its
symbolic character (Zimring and Hawkins 1986, p. 10-11). Capital punishment
legislation, the ritualistic nature of the murder trial, and the cathartic power of the
death sentence are long-established institutions in the United States serv ing a function
for society (p. 11). Thus, capital punishment can be understood as a ritual contributing
to solidarity in society (Collins 1994).
Anomie and Deviance
Anomie is a Durkheimian concept that describes a state of normlessness in a
society that exists when individuals either no longer recognize societys norms or
have lost their belief in them (Stark 1994). In an anomic state, individuals look to
illegitimate or illegal means to obtain worldly success or to even just surv ive.
Between 1900 and 1940, society was anomic and crime was high. Large
numbers of people entered the United States, having in common the need for finding
and holding a job (Barkan 1996). At the turn of the century these immigrants were
welcomed to fill industry and other service jobs. However, according to Barkan, by
the 1920s, this situation was changing. He claims that the 1920s were one of the
worst decades in American history for outsiders.... Immigrants and their children
confronted racism, nativism, and exclusion... (p. 18). Society was crime-plagued in
the 1920s and 1930s, largely in response to Prohibition and the Great Depression.
Although only a tiny percentage of immigrants participated in this crime wave, the
media exploited their role (p. 30). At the time, new immigrants were among
Americans requiring assistance to survive the 1930s economic crisis (p. 53).
The others were not integrated. The sense that they were part of the larger
collectivity was missing. Often social distances remained] unbridged (Bogardus
[ 1928] 1971, p. 6). Durkheim reasons ...that crime is normal because it is bound to
the fundamental conditions of all social life (Durkheim [ 1895] 1978, p.184). He
believes that in every society some individuals will diverge from the collective type;
therefore, it is natural for there to be deviant behavior or crime.
I have found little empirical literature directly relating to the relationship
between social order, others, and execution rates. However, there is relevant research
that supports my thesis.
The phrase, The United States is a nation of immigrants, is a popular and
almost entirely true statement. Migration has alw ays been a distinguishing behavioral
characteristic of the human race. In the long-term evolutionary sense, all nations are
nations of immigrants.
Until the modem era, however, there was little concern about how migrants
might be received wherever they arrived. If the land was unoccupied, the
migrants settled it. If it was occupied, the newcomers might be absorbed...if
they came as individuals but, if in numbers, they often fought those already
there...(Briggs Jr. 1996).
Immigration has had a significant role in the development of America. Over
the years, this role changed as the needs of the country' changed. During its first
hundred years (after 1776) this country needed settlers and immigrants were
welcomed to satisfy this need. With industrialization, in the late 1800s and early
1900s, the country needed cheap unskilled labor and immigrants filled this need. At
this time, immigrants were entering the United States in large numbers congregating
primarily in congested, stratified urban areas, such as New' York City and
Philadelphia, because of the availability of work in these urban centers (Stephenson
1964). By 1900, 13.7% of the country was foreign-bom. This number continued to
grow, and in the early 1900s the highest number of immigrants ever entering in a one
year period (1,285349) entered in 1907, followed in 1914 with 1,218,480
immigrants. In 1910, 14.8% of the population was foreign-bom, the highest number
to date (Census Bureau 1910). Immigration was of critical importance to the heavy
industrial growth experienced by the United States in the late 1800s and early 1900s.
Briggs (1996) states that many jobs available in the expanding urban labor market did
not require skills, education, literacy, or fluency in the English language. Immigrants
easily fit these criteria.
However, immigrants w ere no longer unconditionally w elcome. Their
presence was causing problems for society, and limitations began to appear,
culminating in the restrictive Immigration Law of 1924 (also known as the National
Origins Act). This law' put a low ceiling on the number of immigrants who could
enter the United States and included an ethnic screening system. According to Briggs
(1996), these legal restrictions were a reflection of two areas of societal concern
regarding mass immigration:
1. Economic concerns of depressed wages, repressed unionization, and
2. Nativist social reactions to the ethnic, racial, and religious diversity of the
Briggs (1996) claims that native-born white Americans wanted the
immigrants jobs for themselves. Even w'hen native-born w'hites w'ere given the jobs
over an immigrant, w'ages were lowered because the supply of immigrant labor
lowered the market wage. Native-born whites often were unemployed while jobs
were given to the immigrants. Economic competition was great in the cities, and
there were riots and outbreaks between the native-bom and foreign-bom laborers
(Stephenson 1964). According to Briggs (1996), tolerance for racial, ethnic, and
religious differences was decreasing and discrimination against the immigrant was
becoming a social problem. In states such as Massachusetts, Pennsylvania, and New
York the feeling against foreigners (the others) was strong and social order was
disrupted. Stephenson (1964) states that the eastern states were hostile toward
immigrants. This hostility was fueled by societies supporting nativism1 and a state
of social disorder resulted. Immigrants had become the new others in the community
despite their presence in America for over 300 years.
The immigrant arrives....At first he see things as different, and then he is made
to seem different. He is different. This awareness is isolating. He left the old
and is now...new, different, peculiar, an object of mirth....Confidence...built up
in a peasant primary group, has been jarred; contacts with blasd strangers have
been made. The mother tongue has failed, and exploitation has stalked across
the immigrants pathways....The immigrant, now a stranger, does not receive
so extensive a welcome as he anticipated. His ways are viewed as inferior,
and he falls within the shadow. Another culture is on the throne...[and] his
own culture [is] unjustly berated...(Bogardus  1971, pp. 4-5).
According to Bogardus ( 1971), immigrants were generally tolerated by
the natives as long as the they ...stay[ed] in their place, amenable to control... (p.
8). He states that social conflicts occurred u'hen immigrants became socially mobile
(e.g., moving into a higher class neighborhood), improved their economic status, or
heightened their profile in the community by such acts as involvement in labor
strikes. Bogardus also says that the presence of large numbers of immigrants in the
community, due to their numbers alone, caused competition and rivalry. As the new
immigrants worsened an already severe housing shortage, conflict and social disorder
occurred. Bureau of Census figures lor 1920 showed that about 13% (about one out
of eight) of the total United States population was foreign-bom. If individuals bom
of foreign and mixed parentage are included, the number increases to about 34.4% (or
about one out of three). Parsons would define this massive number of new other
people in the community as a major strain on the social system which would require
the use of social controls to restore equilibrium to the system (Wallace and Wolf
A wide variety of social controls are associated with deviance and crime,
including imprisonment and capital punishment. What social controls were applied to
the immigrant situation around the turn of the twentieth century ? The Reports of the
Immigration Commission, Immigration and Crime, Volume 36, (1970), stales that
immigrants are not involved in crime disproportionate to the increase in adult
population. Statistics obtained by the U.S. Immigration Commission indicate that
foreign-bom immigrants have been less likely to commit crime than native-born
Americans (p.l). However, an analysis of the influence of immigration on
development of penitentiaries around the 1900s (Pisciotta 1978) found that
immigrants were disproportionately imprisoned in relation to native-born white
Americans. Pisciotta believes this situation reflected a justice system that was biased
against immigrants and discrimination arose from a social need to control the
Smith and Guyon (1993) concluded in a study that outsiders were more likely
to be executed. They maintain that the United States is a society with a tradition of
executing our least valued members, outsiders not necessarily [being] defined in
racial terms... [but] those at or beyond societal boundaries... (1993, p. 43). A case
Smith and Guyon use to illustrate this describes the trial and execution of Johannes
De Boer in late nineteenth century rural Illinois. De Boer was a German-bom,
German-speaking, 19-year-old male immigrant who murdered a young native-born
white woman in the community. He was hanged lor the crime he admitted
committing. De Boer ...was on the fringes of a community that twice threatened to
lynch him (1993, p. 48).
Social status is an indicator of the extent to which one is located on the
margins of societal boundaries. Radelet (1989) found social status of the offender and
victim to be determining factors in the application of criminal justice including capital
punishment. In a case study of 30 executions involving white-on-black murders,
Radelet concluded that race alone is crucial in death penalty outcome and that
executions primarily involve lower status defendants convicted of crimes against
higher status victims. However, other status attributes were contributory when
1. Occupational status of victim surpassed that of defendant.
2. Offenders standing in community was marginal before the homicide.
3. Defendants were hated, and had low status or threatened the white power
structure (p. 536).
Organizations of native-born white Americans emerged to oppose
immigration and the problems the immigrants brought, such as the Know-Nothing
Party which organized in the 1840s and 1850s to oppose the massive Irish
immigration spurred by the potato famine. In the late 1890s the Know-Nothings re-
emerged as the American Protective Association (APA) in a renewed effort to stop
immigration (Stephenson 1964). This time their targets were the Catholic immigrants
from Italy and Poland. Their rhetoric w'as volatile and they were instrumental in
causing conflict between Protestant and Catholic groups. Historically, the Know-
Nothings were important because like the Ku Klux Klan, [they were] a symptom of
hostility to certain elements in our population ( p. 146). Other groups such as the
Immigration Restriction League and the Immigration Commission of 1911 presented
evidence that immigrants w ere racially inferior and could not be assimilated (Pavalko
1980). Pavalko states that an Americanization movement redefined the definition
and treatment of the immigrants. Restrictive immigration laws decreased the
immigrant threat to American racial, religious, and ethnic purity by decreasing the
number of immigrants. The immigrants already in the country- were now considered a
new threatto political and economic orderby certain segments of the native bom.
One theme of Durkheims theory is population density'. There has been
empirical research regarding the consequences of population density. One study,
Population Density and Pathology: What Are the Relations for Man? (Galle, Gove,
and McPherson 1972), looked at the relationship between population density and
pathological conditions in Chicago. Five variables w'ere considered in relationship to
population density: mortality, fertility public assistance, juvenile delinquency, and
mental hospital admissions. They found that all variables, except fertility, were
correlated w ith the overcrowding associated with the population density created by
the urban concentration of the industrial labor market and immigrant housing. The
overriding contributing factor, causing the most harm due to overcrowding, was the
number of people living in a room; a situation not dissimilar to the conditions of
immigrants living in tenements. Galle concluded that the escalation of both social
demands and the need to inhibit desires.. .become[s] particularly problematic w'hen
people are crowded together in a dwelling w ith a high ratio of persons per room
David Ward (1968) in his article, The Emergence of Central Immigrant
Ghettoes in American Cities: 1840-1920, reports on a study of immigrants and
urbanization in Boston. He discusses immigration and urban conditions of the times
and concludes that one of the main characteristics of the sustained and heavy
immigration to the United States was congested ghettoes where the foreign
immigrants lived. Most immigrants settled near the sources of jobs for unskilled
laborers at the edges of emerging central business districts. Many of the adjacent
residential areas had been profitably converted into crow'ded tenements. For most
new immigrants, these living conditions were the best they could afford. Tenements
could be rented by the room (crowding many into one room) at a very cheap rate for
the tenants, but adding up to considerable profits for the landlords. Tenements,
however, had social attraction in addition to their price, because tenants could find
individuals from their own cultures residing in them who provided a ready-made
social network of familiarity and support. Ward states immigrants were overwhelmed
by the living environment they encountered in urban tenement districts such as those
in New' York City. They came from impoverished and oppressed lands, but this was
a new type of living condition to manage (p. 462). The settlement of immigrants in
central business areas has been associated with the blighting effect of commercial
encroachment in nearby residential areas. According to Ward, through association
with the term, blight, used to describe the urban industrial conditions of the times,
immigrants were often viewed as blight by the existing community (p. 465).
This review of theory and literature leads me to the following thesis and
Thesis Statement and Hypotheses
Populations shifts create a state of social disorder and conflict. As new people
enter the community, they are defined by the existing group as the others. The
existing citizenry' elects governments which try to restore order and equilibrium by
imposing stricter criminal penalties, including the death penalty. Immigration and
migration, or population flow, of new people into new areas creates social disruption.
High use of capital punishment is one of several sanctions used to restore order. It is
the most extreme sanction and, therefore, useful for present purposes.
Hypothesis 1: An increase in the immigration rate will result in an increase
in the execution rate.
Hypothesis 2: An increase of new people in the population is associated with
an increase in the execution rate.
Corollary 1. An increase in the foreign-bom white (immigrant) population
will have an effect on the execution rate.
Corollary 2. An increase in the black population will have an effect on the
Hypothesis 3: A states urban population will have a greater effect on
execution rate than will a states rural population.
The methodology used for this study included data collection, sampling, and a
procedure for statistical analysis of the data, as described in the following sections.
To locate relevant data to examine the thesis posited regarding the relationship
between demographics and executions between 1900 and 1940,1 combined data from
two primary sources to create a secondary data set. The first source of data was from
a data set titled Executions in the United States, 1608-1991: The Espy File, issued on
CD-ROM by the Inter-University Consortium for Political and Social Research at the
University of Michigan. This dataset includes 21 variables on 14,634 records. Each
record represents an individual case of an execution performed under civil authority
in the United States between 1608 and April 1991. From this data set, I extracted
variables such as total executions by state, total executions by year, race of offender,
crime committed by offender, and occupation of offender. Population data was
collected from various census sources including individual census volumes for 1900,
1910, 1920, 1930 and 1940. Historical population data was also obtained from a
volume of Historical Statistics of the United States, Colonial Times to 1970,
Bicentennial Edition, Part 1. This volume is a collection of census tables reporting
data that has been collected for long historical periods.
Data on immigration, executions, and the death penalty were obtained from
statistics available on the Internet from the Bureau of Census, Immigration
Naturalization Service, and Bureau of Justice Statistics.
My study involves population shifts and immigration in the years 1900 and
1940. I selected a sample of states based on criteria of geographic proximity,
immigration and migration characteristics, population composition and growth, and
high numbers of executions in the states. These states were impacted by late
nineteenth and early twentieth century migrations of white ethnic groups from
southern and eastern Europe and blacks from the American south. As categorized by
historical census data, I included all states in the New England and Middle Atlantic
Regions, and one southern state from the South Atlantic Region for a total of 10
Table 3.1. Sample states by region
New England Region Middle Atlantic Region
Maine New York
New Hampshire New Jersey
South Atlantic Region
Since the Espy file has one record per execution, I selected a subset of Espy
data in the selected states and years, and then used SPSS to aggregate up to the state
level. I then merged the state-level Espy data with the already stale-level census data.
From the historical statistics, I aggregated population data on the states and merged
that data w ith the data from the Espy file to create a secondary data set. I now had a
data set including population variables measuring the racial and ethnic composition of
the population, migration of the population by racial and ethnic groups, and rural and
I created new' variables to standardize the population and execution data into
rates. My data has a moderate N and proportion (p) that is extremely small (number
of executions) (Blalock 1979). I found I had a Poisson distribution rather than a
normal distribution w hich violates an assumption of the multiple linear regression
modelthat the dependent variable is normally distributed within the independent
variables. To reduce error caused by the Poisson distribution, I standardized the
variables by computing rates which would help normalize the sample distribution.
All of the variables in my data appeared to meet the following assumptions for
using a multiple regression model (Loether and McTavish 1993): (a) observations are
statistically independent, (b) there appears to be a linear relationship betw een
variables, (c) variables are interval measure, (d) independent variables are not highly
correlated, (e) effects of independent variables can be added together to yield a
prediction of the dependent variable, and (0 dependent variable is normally
distributed. Therefore, I chose multiple linear regression as the appropriate statistical
analysis for testing my hypotheses.
This chapter includes a discussion of my findings as they pertain to the
hypotheses, a discussion of the statistical results, and charts illustrating population
Hypothesis 1: An increase in the immigrant population will result in an
increase in the execution rate.
This hypothesis is not supported. Immigrant population is statistically
significant (.0253, b = -.052093) but an increase in population predicts a
corresponding decrease in the execution rate.
Hypothesis 2: An increase of new people in the population is associated with
an increase in the execution rate.
Corollary' 1. An increase in the foreign-bom white (immigrant) population
will have an effect on the execution rate.
An increase in the foreign-bom white population is statistically significant
(.0253, b = -.052093), but is negatively correlated. It predicts for an increase in
foreign-bom white population there will be a corresponding decrease in the execution
rate. This corollary does not support hypothesis 2 which predicts an increase in the
execution rate for an increase of new people in the population.
Corollary 2. An increase in the black population will have an effect on the
Black migration is statistically significant (.0400; b = .007808). As the black
population increases through migration from other states, the execution rate increases.
This corollary' supports Hypothesis 2 which predicts an increase in the rate of
execution with an increase of new people in the population.
Hypothesis 3: A states urban population will have a greater effect on the
execution rate than will a states rural population.
This hypothesis is not supported. Urban and rural populations are not
I calculated the same linear regression model using raw data (not transformed
into rates). Results showed a larger Multiple R (= .97094) and R square (= .94184)
than when using data transformed into rates (Multiple R = .79953 and R square =
.63925). Also rural population was significant at .0000 and it had not been significant
using rates. These results lead me to conclude that the raw data model is being
dramatically pulled by population growth.
The following sections discuss the results of a Pearsonian correlation, r,
of independent variables and dependent variable execution rates and the results of a
multiple regression of those same variables.
Results of Pearsonian Correlation, r.
Of Independent Variables and Dependent Variable Execution Rates
1 calculated a correlation matrix to look for association, strength, and direction
of relationship between variables, and to determine influencing variables for the
multiple regression model I used. I found interesting relationships between the
following three subsets of independent variables and the dependent variable: (1)
variables measuring migration in the population; ( 2) variables measuring the
composition of the population; and (3) variables measuring rural and urban
Migration Variables and Total Execution Rate. Black migration is not
significant at. 135 (p < .05) and therefore is not correlated w ith the execution rate.
However, both foreign-bom white (immigrant) migration and native white migration
are positively correlated with execution rates with native white migration having
greater strength. Foreign-bom white (immigrant) migration is significant at .000 with
r = .469 and r square = .219, w hich means that about 21.9% of the variation in the
dependent variable execution rate is explained by association with the independent
variable foreign-born white (immigrant) migration. Native white migration is
significant at .000 with r = .705 and r square = .497, which means that about 49.7% of
the variation in the dependent variable execution rate is explained by association with
the independent variable native white migration.
Population Composition Variables and Total Execution Rates. Foreign-bom
white (immigrant) population, native w'hite population, and black population are
correlated with the execution rate with black population the greatest strength. Black
population is significant at .000 (p < .05) with r = .768 and r square = .589. This
means that about 58.9% of the variation in the dependent variable execution rate is
explained by variation in the independent variable black population. Native while
population is significant at .000 (p < .05) w'ith r = -.566 and r square = .320. This
means that about 32% of the variation in the dependent variable execution rates is
explained by the variation in the independent variable native white population.
Foreign-bom white (immigrant) population is significant at .003 (p < .05) with r = -
.387 and r square = .149. Therefore, about 14.9% variation in the dependent variable
execution rates is explained by variation in the independent variable foreign-born
white (immigration) population. It is important to note that both native white and
foreign-bom white populations are negatively correlated.
Rural and Urban Populations and Execution Rates. Rural population and
urban population are not significant at .101 and .102 (p <.05), respectively.
To summarize, Pearsonian correlation, r, indicates that the racial and ethnic
composition of the population is more highly correlated with execution rates than
migration or rural or urban groups, per se. For population composition, black
population has the strongest correlation, followed by native w hite population,
follow ed by foreign-bom white (immigrant) population. For migration groups, native
white migration has the strongest correlation to execution rates, followed by foreign-
bom w hite (immigrant) migration; and black migration is not correlated.
Results of Multiple Regression of
Independent Variables and Dependent Variable Execution Rates
I next used a hierarchical Ordinary' Least Squares model to look for a linear
relationship between independent variables and execution rates. I initially included
the eight independent variables in the regression equation using the enter method:
black migration, foreign-bom w'hite (immigrant) migration, native w'hite migration,
black population, foreign-bom white (immigrant) population, native white
population, rural population, and urban population (all standardized as rates). I set
the alpha level at .05.
The multiple linear regression calculation produced the following results:
black migration, foreign-bom white population, and native white population variables
are significant and have a linear relationship to execution rates. The constant is also
significant. Black population has a significance level of .0404, but SPSS did not enter
it into the model because the minimum tolerance = 5.286E-05. SPSS does not enter a
variable into the model if it results in a very small tolerance for any of the
independent variables because of computational problems (Norusis 1995). Tolerance
measures collinearity, and it is likely that the black population and black migration
variables possess this characteristic.
Native white population is significant at .0014 (p < .05) with the b coefficient
= -.063146. This predicts that for every increase in native white population of
100.000, executions will decrease by .06 per 100,00 population. Foreign-bom white
(immigrant) population is significant at .0253 (p < .05) with the b coefficient =
-.052093. Thus, for every increase \n foreign-born white population of 100,000,
executions will decrease by .05 per 100,000 population. Black migration is
significant at .0400 (p > .05) with b coefficient = .007808. This predicts that for
every increase in black migration of 100,000, executions will increase by .008 per
100.000. The constant is also significant at .0003 (p < .05) with b coefficient = 6.649,
which means there was a rate of 6.6 executions per 100,000 population before any of
the independent variables were entered into the regression equation.
Multiple R = .79953 and represents the total amount of variation in execution
rates explained by variation in the independent variables included in the equation. R
square = .63925 and represents the total proportion of the variation in executions rates
explained by the regression equation.
To summarize, although these three independent variables are significant, they
are not highly so. Native white and foreign bom white populations arc both
significant and are negatively correlated with execution rates. Therefore, as their
populations increase, the execution rate decreases. On the other hand, black
migration is significant and is positively correlated with the execution rate which
predicts that as black migration increases, the execution rate also increases.
Therefore, the migration or flow of blacks into a new area has a greater effect on the
execution rate than does the existing racial and ethnic population composition. It is
noteworthy that urban and rural population are not significant.
Charts Illustrating Demographics and Executions 1900-1940
This study looks at the effect of rapid demographic change on the execution
rate between 1900 and 1940. The following series of charts illustrates population and
execution numbers in the ten sample states during this period. First, I display raw
figures on immigration, internal migration, population, and executions during this
period. Second, 1 display a number of simple scatterplots and regression lines
looking at the independent variables separately. Third, I evaluate a regression model
that ties together all of these factors.
As can be seen in Figure 4.1, there were literally millions of immigrants
moving into the United States prior to World War II, mostly from Italy and Eastern
Europe. There was also a considerable internal migration of blacks to the northern
Atlantic seaboard states. Finally, there was a modest outmigration of native bom
Intercensal Migration (in 000s)
Figure 4.1. Intercensal migration
This migration led to considerable increases in population with total
population almost doubling in this 50-year period, as shown in Figure 4.2.
a Native White
Figure 4.2. Total population, 1900-1940
The actual composition of the population stayed relatively stable, as can be
seen in Figure 4.3.
Figure 4.3. Composition of population, by race
Executions were high during this period, higher than during any other period
in the United States between 1608 and 1991.
Figure 4.4 shows execution numbers broken down by race. The category
others represents Hispanics, Asians, and American Indians.
Blacks show up in disproportionate numbers, an effect that is emphasized by
the following series of doughnut charts (Figure 4.5).
(Pop is outer ring)
Figure 4.5. Comparison of blacks and whites executed
Figure 4.6 shows total immigration versus total population. In each of m\
records there is a population number and an execution number. The scatterplot shows
the plotting of the migration population and execution pairs.
Tots r-460-| al Mig ration vs Tc >tal E) tec utii 3n
o> o o o 0 1 6 3.00-20 D.00 .00 20( 3.00 40C .00 60C O 00 o o 3.00 10C 30.0 1 200.0 0
Figure 4.6. Total migration vs. total executions
To reduce the impact of the great differences in scale between these two
variables, I also prepared a chart of total migration versus total executions as a log-
log chart (Figure 4.7.)
10.00-1 1.001 1. Total M igration vs Total E ixecul tions
l I .
* % t
Figure 4.7. Total migration vs. total executions, log-log chart
Now, lets break this out by type of migration. Figure 4.8 shows execution
versus migration broken out into subgroups of native white, foreign-bom white, and
black. (In historical population data, Hispanics were included in the white category,
except in 1930 when they were put in the other category; Asians were a sufficiently
small group that I am not including them.)
1200.00 n 1 nnn nn. Tota 1 Exec utions vs Mic jration
Ann nn _
Ann nn -
40n nn - 4 Native white Foreign White Black
pnn nn _ 1 i .
.00 J *w\ D 4 1 im L. '
3 6 D f8 3 1 )0 1 >0 *1 I \0 160
-4no 00- A *
-600.00- -- 4 -
Figure 4.8. Total executions vs. migration, by subgroup
Once again, I used the same data and prepared a log-log chart (Figure 4.9)
showing migration versus total executions.
Migration vs Execute
Figure 4.9. Migration vs. executions, log-log chart
From these charts and the regressions using independent variables of
population and migration, we see that the strongest relationship exists between the
number of migrating blacks in the population and the number of executions.
14000 -i 12000 --- Ex( o c o ns vs. Popu lation " '
OUUU ennn .
4000 onnn . if A
n! 5 J w
0 2 0 4 0 60 0 0 1C X) 1 20 1- 40 160
Figure 4.10. Number of executions compared to population size
The following chart (Figure 4.12) compares executions to population size
broken out by subgroups of native white population, black population, and foreign-
bom white population.
12000-1 1 noon. - - E xecutic ms vs Pop
onnn - t
uwu * *NWPOP BKPOP FBWHTPOP
2000 J A 4 add m i B ,
0 20 40 6 0 0 0 1 DO 1 20 V 10 160
Figure 4.12. Execution vs. population, by subgroup
Figure 4.13 is a log-log chan making the same companson of executions to
population size broken out by subgroups.
Total Executions vs Total Pop
Figure 4.13. Total executions vs. total population, log-log chart
I began this study with the question, To what extent can historical
demographic and execution statistics explain the high number of executions between
the years 1900 and 1940? The methodology and statistical procedures I followed
have provided a small insight into this question. Existing theory led me to predict
that a shifting population defined as the movement of, or increases in, new others in
the community would lead to social disorder which in tum would lead to increased
use of capital punishment.
The data showed that increases in immigrant population did not lead to
increases in the execution rate, as I had hypothesized. Native white migration also
did not lead to an increase in the execution rate. However, migration of the black
population did lead to an increase in the execution rate. It appears then that internal
migration of the black population race was more influential than immigration of
foreign-bom whites in increasing the number of executions. Other people arriving in
the community were noticed; they may even have disrupted social order; but the only
group showing a statistically significant increase in the execution rate was the black
It is important in discussing the results of this study to take into consideration
the growth in population. This study includes the years 1900 through 1940, a time of
rapid growth in the United States. The total population of the United States in 1900
was 75,994,575 and in 1940 it was 131,669,275 (a 73.3% increase). The total
population in the ten sample states in 1900 was 21,576,000 and in 1940 it was
37,873,000 (a 75.5% increase). The total number of executions in the United States
fluctuated between 1900 (99 executions) and 1940 (124 executions) with an overall
increase of 64.4%. Population and executions increased at almost the same rate. This
population effect is also visible at the state level. Population growth was an important
factor even before considering other variables and their relationship to the execution
Historically, population shifts and executions were not happening in a
vacuum. Here are some high points of these decades (Zinn 1992).
1900s. During the first two decades of this century, there were World War
I, womens suffrage, and the highest number of immigrants ever entering
the United States in one year periods (over one million for six different
1920s. This decade saw the Teapot Dome scandal, A1 Capone, and
Prohibition. In 1929, 60% of all Americans had an annual income below
the poverty level of $2,000. Child labor was legal. Scopes was arrested
and tried for discussing evolution in the classroom. Leopold and Loeb,
two white men from wealthy Chicago families, were tried for murder and
sentenced to life imprisonment. Saccho and Vanzetti, two white
immigrants, were arrested for murder and executed after a seven year trial.
Their trial became a lightning rod for the frustrations and anti-foreign
anger of Americans. At one point, the judge himself referred to the
defendants as anarchist bastards (Barkan 1996, p.19). The stock market
crashed in October 1929, beginning the Great Depression.
1930s. The Great Depression was in full sw ing. Wages dropped to 60%
below the 1929 levels to about $400 annually. A presidential commission
was formed to deal with lawlessness. A national commission found that
about one-third of American school children were receiving inadequate
education. Average life expectancy was 59. Richard Bruno Hauptman, a
German immigrant, was executed for the kidnapping and death of the
Lindbergh baby. United States entry in World War II was imminent.
All of these events of history, especially those that w ere economic and
criminal, would fall into Parsons definition of social strains (Wallace and Wolf
1991) that cause disruption in the social order of society.
Limitations and Future Studies
I looked specifically at flow and composition characteristics of population in
less than one-fifth of the states between 1900 and 1940 (a time of unusually high
executions). It is not possible to generalize to the entire population, although it
should be noted that the Northeast was at least the initial destination of European
immigrants during the period studied. This data also gives us no information about
why the United States has death penalty legislation when other western countries do
Further research studies should include the follow ing:
1. Studies looking at the effect of status on execution rates by measuring
education, occupation, and income of both offender and victim
2. Studies looking at impact of economic conditions
3. Studies looking at political traditions of the executing jurisdictions
4. Studies including all 50 states and comparing executions in years prior to
the Supreme Court ban in 1972 and after the Supreme Court ban in 1976
5. Studies including and controlling for crime rates.
It is a common belief that capital punishment is not a deterrent in the United
States (Radelet and Akers 1996). There are statistics that indicate an increase in the
murder rate following executions in a state. Researchers (Dunne 1996) have found
the following: (1) Floridas murder rate increased 28% after the State resumed
executions after a 15-year moratorium. (2) Georgia experienced a 20% increase in
the homicide rate in 1984, while nationwide the homicide rate fell by 5%. (3) In a
New York study of homicide rates from 1907 to 1963, when New York led the nation
in executions, it was found that there had been an average of two additional
homicides in each month following an execution.
The reasons execution rates fluctuate (from zero on up), even when death
penally legislation is stationary, are complex. The ritualistic and symbolic nature of
murder trials, executions, and even death penalty legislation cannot be minimized
(Zimring and Haw'kins 1986). This may explain why de facto abolition1 of the death
penalty may exist in a contemporary society without need for executions, as in the
United States in the late 1960s and early 1970s when executions were not occurring.
The execution rate has been slowly increasing since the repeal of the ban on capital
punishment in 1976. Are demographic changes influencing the use of the death
penalty today? The United States is experiencing another w'ave of high immigration.
In 1990,7.9% of the population was foreign-bom and executions were increasing; in
1930, the decade when executions averaged 150 a year, 11.6% of the population was
foreign-bom; in 1970,4.7% of the population was foreign-bom and executions had
virtually stopped by an unofficial, self-imposed, country-wide moratorium. Data from
the Immigration Naturalization Service shows that betw een 1900 and 1940 the
majority of immigrants came from southern and eastern Europe. In the 1990s
immigrants are primarily Hispanics from Mexico and the Americas and Asians.
The results of my study identified one social indicator of the high use of
executions. The internal shift of the black population had a more disturbing effect on
the community than did the immigration of southern and eastern Europeans. Is there
a relationship between increasing immigration and increasing execution rates in
1996? If so, how would it compare to the early twentieth century? Without
additional research, it is not possible to generalize the findings of my study to 1996.
A study including all 50 states and additional population and immigration data might
provide a clearer understanding of the relationship of these variables.
My findings allow me to make the following statements. Between 1900 and
1940 execution rates were higher than at in other time in the United States since 1608.
Arrival of others in the community is associated with the execution rate during this
period in the 10 sample states. Migration of the black population in the 10 sample
states is associated w'ith an increase in the execution rate between 1900 and 1940.
1. Civil authority refers to any government entity that has the legal authority to
impose a death penalty. Although this is usually the state government, local and
Federal governments have also authorized executions.
2. Furman v. Georgia, 408 U.S. 238 (1972).
3. Gregg v. Georgia, 428 U.S. 153 (1976).
1. Nativism refers to the belief that native-born Americans and their rights were
superior to the rights of the foreign-bom or immigrants. Groups supporting nativism
appeared in response to the large number of immigrants present in the United States
in the late nineteenth and early twentieth centuries.
1. De facto abolition of the death penalty refers to voluntary moratorium on
Barkan, Elliott Robert. 1996. And Still They Come. Wheeling, IL: Harlan Davidson.
Briggs Jr., Vernon. 1996. Immigration Policy and the U.S. Economy: An
Institutional Perspective. Journal of Economic Issues. 33:2:371-392.
Blalock, Hubert M., Jr. 1979. Social Statistics. New York, NY: McGraw-Hill.
Bogardus, Emory S.  1971. Immigration and Race Attitudes. New York, NY:
D.C. Heath and Company.
Bureau of the Census. 1975. Historical Statistics of the United States, Colonial Times
to 1970, Bicentennial Edition, Part 1. Washington, DC: U. S. Government
Bureau of the Census. 1900-1940. Population Census. Volumes 1900, 1910, 1920,
1930, and 1940. Washington, DC: U.S. Government Printing Office.
Collins, Randall. 1994. Four Sociological Traditions. New York, NY: Oxford
Death Row Penalty Information Center, Washington, DC. 1996. Twenty Years of
Capital Punishment: A Re-evaluation.
http://www.essenlial.Org/dpic/dpic.r01.html#sxnl. (14 Oct. 1996).
Dunne, Jimmy. 1996. How Executions Boost The Murder Rale.
http://www.abolition-now.com/executions_boost_murder_rate.html (15 Aug.
Durkheim, Emile.  1984. The Division of Labor In Society. Translated by W.D.
Halls. New York, NY: Free Press.
________1978. Crime et Sante Sociale. Pp. 181-188 in On Institutional Analysis.
Translated by Mark Traugott. Chicago, IL: University of Chicago Press.
(Originally published in Revue Philosophique 39 (1895):518-23.
________ 1982. The Rules of Sociological Method and Selected Texts on
Sociology and its Method. Translated by W. D. Halls. New York, NY: Free Press.
Espy, M. Watt and John Ortiz Smykla. 1994. Executions in the United States, 1608-
1991: The Espy File [Computer file], 3rd ICPSR ed., compiled by John Ortiz
Smykla. Ann Arbor, MI: Inter-University Consortium for Political and Social
Research [producer and distributor].
Galle, Omer R., Walter R. Gove, and J. Miller McPherson.  1974. Population
Density and Pathology: What are the Relationships for Man? Pp. 198-213 in
Comparative Urban Structure, Studies in the Ecology of Cities, edited by K. P.
Schwirian. Lexington, MA: D.C. Heath and Company.
Goffman, Erving. 1963. Stigma. New York, NY: Simon & Schuster.
Loether, Herman J. and Donald G. McTavish. 1993. Descriptive and Inferential
Statistics. Boston, MA: Allyn and Bacon.
Madrid, Arturo. 1992. Missing People and Others: Joining Together to Expand the
Circle. Pp. 6-11 in Race, Class, and Gender, edited by M. L. Anderson and P. H.
Collins. Belmont, CA: Wadsworth Publishing Company.
Norusis, Marija J. 1995. SPSS 6.1 Guide to Data Analysis. Englewood Cliffs, NJ:
Pavlako, Ronald M. 1980. Racism and the New Immigration: A Reinterpretation of
the Assimilation of White Ethnics in American Society. Sociology and Social
Pisciotta, Alexander W. 1978. Immigration, Social Control, and Development of the
Penitentiary in the United States. Journal of Social Problems. 33:2:165-178.
Radelet, Michael L. 1989. Executions of Whites for Crimes Against Blacks:
Exceptions to the Rule? The Sociological Quarterly. 30:4:529-544.
Radelet, Michael L. and Ronald L. Akers. 1996. Deterrence and the Death Penalty:
The Views of the Experts. http://www.sun.soci.niu.edu/~critcrim/dp/dppapers/
Sandokan. 1996.New Peak In Executions In The USA.
http:www.derechos.org/amnesty/dp/ (14 Oct. 1996).
Smith, Beverly A. and Lois A. Guyon. 1993. Two Executions in 19th Century' Rural
Illinois. Federal Probation. 57:2:43-53.
Stephenson, George M.  1964. A History of American Immigration, 1820-
1924. New York, NY: Russell & Russell.
Taylor, Donald M. and Fathali M. Moghaddam. 1994. Theories of Intergroup
Relations, International Social Psychological Perspectives. Westport, CT:
Turner, Jonathan H., Leonard Beeghley, and Charles H. Pow ers. 1989. The
Emergence of Sociological Theory. Belmont, CA: Wadsworth Publishing
U.S. Immigration and Naturalization Sen ice (INS), Department of Commerce,
Bureau of Census, Ethnic and Hispanic Branch. 1996. Immigration Fact Sheet.
http://ww'w7usdoj.gov/ins/public/299-310.html (20 Aug.).
U.S. Immigration Commission.  1970. Reports of the Immigration
Commission, Immigration and Crime. New York, NY: Amo and The New York
Wallace, Ruth A. and Alison Wolf. 1991. Contemporary Sociological Theory:
Continuing the Classical Tradition. Englewood Cliffs, NJ: Prentice Hall.
Ward, David.  1974. The Emergence of Central Immigrant Ghettoes in
American Cities: 1840-1920. Pp. 457-475 in Comparative Urban Structure,
Studies in the Ecology of Cities, edited by K. P. Schwirian. Lexington, MA: D.C.
Heath and Company.
Winsborough, Hal H.  1974. The Social Consequences of High Population
Density. Pp. 193-198 in Comparative Urban Structure, Studies in the Ecology of
Cities, edited by K. P. Schwirian. Lexington, MA: D.C. Heath and Company.
Zinn, Howard. 1995. A People's History of the United States, 1492-Present. New
York, NY: Harper Perennial.
Zimring, Franklin E. and Gordon Hawkins. 1986. Capital Punishment and the
American Agenda. Cambridge, MA: Cambridge University Press.