Gun Violence辅导、辅导Java，c/c++编程语言、讲解Python设计
- 首页 >> 其他 Gun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
Assignment # 5 & 6 Descriptive patterns
Table 1 shows the descriptive statistics of key variables in analysis at the individual level.
In particular, we focus on gaps in trust between four groups of Americans, namely, those who
reported no experience of gun threat and gun-shot during their course of life so far (reference
group), those who experienced gun threat or gun-shot as an adult (adult group), those who
experienced gun threat or gun-shot as a child (child group), and those who experienced gun
threat or gun-shot during both their childhood and adulthood (both group).
--- Table 1 about here ---
To visualize the gaps, Figure 1A-1C compare trust levels (the probability of saying most
people can be trusted, people are helpful, or people are fair) across the four groups of
Americans. Consistently, they show that compared to the reference group, those who had
experience of being threatened by a gun or receiving a gunshot wound are significantly less
likely to say that most people can be trusted (Figure 1A), to say that people are helpful (Figure
1B), and to say that people are fair (Figure 1C). Within the three groups that reported gun
victimization experiences, while the trust gap between the adult group and the child group does
not seem to be substantial, those who had the experience of a gun threat and had been shot by a
gun during both their childhood and their adulthood, show much lower trust than all other
groups. Specifically, while about 42 percent of Americans who reported no experience of gun
victimization would say that most people can be trusted, 36 percent say that most people can be
trusted if they experienced gun victimization as an adult, 35 percent if they experienced gun
victimization as a child, and only 24 percent for those who has repeated gun victimization during
both their childhood and adulthood. The pattern remains consistent regardless how trust is
measured (see Table 1 and Figure 1B&1C). The overall results show that individuals’ past
experiences of gun victimization significantly and negatively affect their trust and that repeated
gun victimization (in both childhood and adulthood) has the strongest effect on trust.
--- Figure 1A-1C about here ---
Figure 1: Gun Victimization and Trust in the United States, 1973-1994
No (n=13,386) Yes, as an adult (n=2,738) Yes, as a child (n=444) Yes, both (n=153)
Generalized Trust
Can people be trusted? (0=no, 1=yes) 43.07 35.68 34.68 23.53 41.46
People helpful or looking out for selves (0=not helpful, 1=helpful) 53.97 43.21 41.98 29.38 51.69
People fair or try to take advantage (0=not fair, 1=fair) 63.13 48.86 49.34 33.54 60.18
Demographics
Female (0=male, 1=female) 62.50 29.68 27.61 11.88 55.81
Age (in years, 18-89 or older) 45.03 42.70 35.94 37.51 45.99
Education (in years, 0-20) 12.35 12.25 12.44 12.44 12.83
White (0=no, 1=yes) 85.7 82.93 78.48 71.88 85.30
Black (0=no, 1=yes) 11.53 14.95 18.91 26.88 12.41
Other (0=no, 1=yes) 2.32 2.12 2.61 1.25 2.28
Never married (0=no, 1=yes) 17.01 15.48 26.74 21.25 17.05
Married (0=no, 1=yes) 60.4 58.13 59.35 50.63 59.91
Widowed (0=no, 1=yes) 10.79 5.87 1.94 1.88 9.68
Divorced/separated (0=no, 1=yes) 11.8 20.53 11.96 26.25 13.35
Family income in constant dollars (inflation-adjusted) 41267 41716 42486 41142 41371
Happiness and Financial Satisfaction
Very happy (0=no, 1=yes) 34.39 28.8 28.26 26.25 33.25
Pretty happy (0=no, 1=yes) 55.32 54.66 56.74 53.75 55.23
Not too happy (0=no, 1=yes) 10.29 16.54 15 20.00 11.52
Satisfied (0=no, 1=yes) 31.1 23.36 22.39 18.13 29.50
More or less (0=no, 1=yes) 45.27 39.79 40 31.87 44.12
Not at all satisfied (0=no, 1=yes) 23.63 36.86 37.61 50.00 26.38
Have you erver been threatened with gun or shot at? TotalGun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
At the aggregate level, we explore whether higher levels of gun victimization lead to
lower levels of trust across the U.S. Census Bureau subdivisions and over time. The unit of
analysis in this place association is subdivision-year (9 subdivisions*15 years from 1973 to
1994=135 data observations). Using the GSS data, Figure 2A-2C show the scatter plot between
the probability of people who were threatened with gun or shot at (x-axis) and the probability of
people who trust (y-axis) by Midwest, Northeast, South, and West regions. Each circle, or plus,
triangle, and square, represents one subdivision in a particular year (e.g., New England-1975,
Mountain-1994). Clearly, they show that a higher probability for a subdivision in a particular
year leads to a lower level of trust for that particular subdivision and year. The pattern is
consistent across the four higher levels of regions irrespective of how we measure trust.
Similarly, using the data from CDC’s Multiple Cause of Death Files from 1999-2016 and
combing with the GSS trust data, Figure 2D-2F show the scatter plot between the crude rate of
homicide due to firearms (x-axis) and the probability of people who trust (y-axis) by Midwest,
Northeast, South, and West regions. Similarly, the unit of analysis in this association is also
subdivision-year (9 subdivisions*22 years from 1980 to 2016=198 data observations). Overall,
these figures demonstrate that, regardless how we measure trust, higher crude rates of homicide
due to firearms lead to lower levels of trust.
--- Figure 2A-2C about here ---
Figure 3A-3C: Gun Victimization and Trust across Regions, 1973-1994 Gun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
--- Figure 2D-2F about here ---
Figure 3D-3F: Gun Homicides and Trust across Regions, 1980-2016
--- Figure 3A about here ---
Figure 3A: Gun Violence and Trust within the U.S. Census Divisions, 1972-1994.
.2 .3 .4 .5 .6
A: People can be trusted
0 .1 .2 .3 .4
Ever threatened with gun or shot at?
Midwest Northeast
South West
.3 .4 .5 .6 .7
B: People are helpful
0 .1 .2 .3 .4
Ever threatened with gun or shot at?
Midwest Northeast
South West
.3 .4 .5 .6 .7 .8
C: People are fair
0 .1 .2 .3 .4
Ever threatened with gun or shot at?
Midwest Northeast
South WestGun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
Finally, to demonstrate that gun crime is also associated with trust over time, Figure 3A
shows the gun crime and trust pattern within each of the U.S. Census Bureau subdivisions across
15 years using the aggregated GSS data. Clearly, the general pattern is that within each
subdivision, years with higher percentages of gun victimization are also years with lower levels
of trust. In Appendix 4, Figure 3B-3F replicate the figure using different trust measures and data
from the CDC (1980-2016). Overall, they suggest that gun crime and trust is highly related over
time at the place level.
(Paper under review, please do not share with others outside class)
Assignment # 5 & 6 Descriptive patterns
Table 1 shows the descriptive statistics of key variables in analysis at the individual level.
In particular, we focus on gaps in trust between four groups of Americans, namely, those who
reported no experience of gun threat and gun-shot during their course of life so far (reference
group), those who experienced gun threat or gun-shot as an adult (adult group), those who
experienced gun threat or gun-shot as a child (child group), and those who experienced gun
threat or gun-shot during both their childhood and adulthood (both group).
--- Table 1 about here ---
To visualize the gaps, Figure 1A-1C compare trust levels (the probability of saying most
people can be trusted, people are helpful, or people are fair) across the four groups of
Americans. Consistently, they show that compared to the reference group, those who had
experience of being threatened by a gun or receiving a gunshot wound are significantly less
likely to say that most people can be trusted (Figure 1A), to say that people are helpful (Figure
1B), and to say that people are fair (Figure 1C). Within the three groups that reported gun
victimization experiences, while the trust gap between the adult group and the child group does
not seem to be substantial, those who had the experience of a gun threat and had been shot by a
gun during both their childhood and their adulthood, show much lower trust than all other
groups. Specifically, while about 42 percent of Americans who reported no experience of gun
victimization would say that most people can be trusted, 36 percent say that most people can be
trusted if they experienced gun victimization as an adult, 35 percent if they experienced gun
victimization as a child, and only 24 percent for those who has repeated gun victimization during
both their childhood and adulthood. The pattern remains consistent regardless how trust is
measured (see Table 1 and Figure 1B&1C). The overall results show that individuals’ past
experiences of gun victimization significantly and negatively affect their trust and that repeated
gun victimization (in both childhood and adulthood) has the strongest effect on trust.
--- Figure 1A-1C about here ---
Figure 1: Gun Victimization and Trust in the United States, 1973-1994
No (n=13,386) Yes, as an adult (n=2,738) Yes, as a child (n=444) Yes, both (n=153)
Generalized Trust
Can people be trusted? (0=no, 1=yes) 43.07 35.68 34.68 23.53 41.46
People helpful or looking out for selves (0=not helpful, 1=helpful) 53.97 43.21 41.98 29.38 51.69
People fair or try to take advantage (0=not fair, 1=fair) 63.13 48.86 49.34 33.54 60.18
Demographics
Female (0=male, 1=female) 62.50 29.68 27.61 11.88 55.81
Age (in years, 18-89 or older) 45.03 42.70 35.94 37.51 45.99
Education (in years, 0-20) 12.35 12.25 12.44 12.44 12.83
White (0=no, 1=yes) 85.7 82.93 78.48 71.88 85.30
Black (0=no, 1=yes) 11.53 14.95 18.91 26.88 12.41
Other (0=no, 1=yes) 2.32 2.12 2.61 1.25 2.28
Never married (0=no, 1=yes) 17.01 15.48 26.74 21.25 17.05
Married (0=no, 1=yes) 60.4 58.13 59.35 50.63 59.91
Widowed (0=no, 1=yes) 10.79 5.87 1.94 1.88 9.68
Divorced/separated (0=no, 1=yes) 11.8 20.53 11.96 26.25 13.35
Family income in constant dollars (inflation-adjusted) 41267 41716 42486 41142 41371
Happiness and Financial Satisfaction
Very happy (0=no, 1=yes) 34.39 28.8 28.26 26.25 33.25
Pretty happy (0=no, 1=yes) 55.32 54.66 56.74 53.75 55.23
Not too happy (0=no, 1=yes) 10.29 16.54 15 20.00 11.52
Satisfied (0=no, 1=yes) 31.1 23.36 22.39 18.13 29.50
More or less (0=no, 1=yes) 45.27 39.79 40 31.87 44.12
Not at all satisfied (0=no, 1=yes) 23.63 36.86 37.61 50.00 26.38
Have you erver been threatened with gun or shot at? TotalGun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
At the aggregate level, we explore whether higher levels of gun victimization lead to
lower levels of trust across the U.S. Census Bureau subdivisions and over time. The unit of
analysis in this place association is subdivision-year (9 subdivisions*15 years from 1973 to
1994=135 data observations). Using the GSS data, Figure 2A-2C show the scatter plot between
the probability of people who were threatened with gun or shot at (x-axis) and the probability of
people who trust (y-axis) by Midwest, Northeast, South, and West regions. Each circle, or plus,
triangle, and square, represents one subdivision in a particular year (e.g., New England-1975,
Mountain-1994). Clearly, they show that a higher probability for a subdivision in a particular
year leads to a lower level of trust for that particular subdivision and year. The pattern is
consistent across the four higher levels of regions irrespective of how we measure trust.
Similarly, using the data from CDC’s Multiple Cause of Death Files from 1999-2016 and
combing with the GSS trust data, Figure 2D-2F show the scatter plot between the crude rate of
homicide due to firearms (x-axis) and the probability of people who trust (y-axis) by Midwest,
Northeast, South, and West regions. Similarly, the unit of analysis in this association is also
subdivision-year (9 subdivisions*22 years from 1980 to 2016=198 data observations). Overall,
these figures demonstrate that, regardless how we measure trust, higher crude rates of homicide
due to firearms lead to lower levels of trust.
--- Figure 2A-2C about here ---
Figure 3A-3C: Gun Victimization and Trust across Regions, 1973-1994 Gun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
--- Figure 2D-2F about here ---
Figure 3D-3F: Gun Homicides and Trust across Regions, 1980-2016
--- Figure 3A about here ---
Figure 3A: Gun Violence and Trust within the U.S. Census Divisions, 1972-1994.
.2 .3 .4 .5 .6
A: People can be trusted
0 .1 .2 .3 .4
Ever threatened with gun or shot at?
Midwest Northeast
South West
.3 .4 .5 .6 .7
B: People are helpful
0 .1 .2 .3 .4
Ever threatened with gun or shot at?
Midwest Northeast
South West
.3 .4 .5 .6 .7 .8
C: People are fair
0 .1 .2 .3 .4
Ever threatened with gun or shot at?
Midwest Northeast
South WestGun Violence and Trust (Wu 2019)
(Paper under review, please do not share with others outside class)
Finally, to demonstrate that gun crime is also associated with trust over time, Figure 3A
shows the gun crime and trust pattern within each of the U.S. Census Bureau subdivisions across
15 years using the aggregated GSS data. Clearly, the general pattern is that within each
subdivision, years with higher percentages of gun victimization are also years with lower levels
of trust. In Appendix 4, Figure 3B-3F replicate the figure using different trust measures and data
from the CDC (1980-2016). Overall, they suggest that gun crime and trust is highly related over
time at the place level.