Violent Crime Rates Essay
As a policy maker who wishes to promote rehabilitation (for nonviolent offenders) while effectively reducing the number of prisons and penitentiaries , it is essential to understand what possible factors influence violent crimes. Before you take steps to reduce prison funding and attempt to eliminate higher security facilities, it is absolutely necessary to ensure that the population within the surrounding city is not at risk for high violent crime rates. Before we proceed to statistical analysis, lets first take a brief look at exactly what we are dealing with.Violent crimes include offenses in which the offender uses or threatens the use of violence upon the victim. These crimes include homicide, rape, aggravated assault, robbery, and several others.
You have asked us to analyze relationships between population characteristics and their respective state violent crime rates. With this information you can effectively decide whether or not to aggressively pursue a cost cutting campaign targeting prisons throughout the United States. Questions: What factors will influence violent crime rates?Are these factors present in large concentrations throughout city? Based on statistical evidence, is the surrounding city likely to have a high violent crime rate? Should this prison be a possible target for cut backs? Model Several components influence violent crime rates across the United States. We hope to identify trends and relationships between crime rates and these contributing factors in SANDY CAMEL CONSULTING 2 VIOLENT CRIME FALL 2010 these rates and population characteristics in order to predict crime rates in similar areas.If we are able to define a correlation between crime rates and population characteristics, we can effectively determine whether or not a particular city will be an ideal candidate for cost cutting measures. We expect violent crime rates to be strongly influenced by such factors as poverty, population density and presence of minority groups.
Several other factors most definitely have an effect on violent crime rates, but for the purposes of this analysis we will only be able to examine a handful.In areas where poverty is abundant, individuals often find themselves with fewer options in life. These people are not only more likely to have adverse health issues, but are also more likely to lack education. These factors lead to a higher risk of delinquent behavior and ultimately higher violent crime rates.
Also, as population density increases we believe that violent crime rates will rise. As cities develop and expand, human interactions tend to increase. These interactions create more opportunities for violent crime.We also expect greater percentages of minority groups to influence violent crime rates. These groups do not always have access to all opportunities offered by the community. Although social programs are in place to provide guidance to these groups, they do not always take advantage of such programs and often resort to gang membership and lives of crime.
“One way to make sure crime doesn’t pay would be to let the government run it. ” Ronald Reagan SANDY CAMEL CONSULTING 3 VIOLENT CRIME FALL 2010 DataThe data used for the analysis comes from the United Sates Census. Every state is represented in the data set. Crime statistics are reported in the number of violent crimes committed per 1000 people. • As you can see from Table 1, throughout the United States, an average of about 4 violent crimes are committed per 1,000 people. The state with the highest levels of violent crime is Washington D.
C. (16. 08) and the state with the lowest levels is North Dakota (0. 78). • Unemployment rates did not have a lot of variance and had a mean value of 5.
64%.Average percentages of mobile housing units and of population over 65 living under the poverty level were similar (8. 59% and 9. 99% respectively).
• The average median age throughout the United Sates is 36 years old with a minimum of 27 found in Utah and maximum of 39 in West Virginia. • The average median income is $41,346 per year with a minimum of $29,696 seen in West Virginia and a maximum of $55,136 earned by residents of New Jersey. An average of 72% of each state has been developed. Noteworthy Statistics: Highest violent crime rate: (16. 8) Washington D. C.
Lowest median income: $29,696 West Virginia Highest unemployment rate: 8. 2% Oregon Highest percentage of development: 100% Washington D. C. Table 1: Descriptive Statistics Variable Name Violent Crime Unemployment Mobile Home % Old Poor % White % Median Age Median Income Urban % Mean 4. 25 5. 64 8.
59 9. 99 0. 78 35. 52 41,346.
75 0. 722 Std. Deviation 2. 43 1. 05 5. 28 2.
98 0. 15 1. 89 6,286 0. 15 Minimum 0. 78 3.
6 0. 1 5. 8 0. 24 27. 1 29,696 0.
38 Maximum 16. 08 8. 2 20. 3 18. 8 0.
97 38. 9 55,136 1 SANDY CAMEL CONSULTING VIOLENT CRIME FALL 2010 Results After performing statistical analysis on the violent crime data, we are able to provide a few recommendations that will help determine whether or not a particular population is a good candidate for penal reform. The variations within conditions that we chose to analyze (unemployment rate, mobile home percentage, percentage of population over 65 living in poverty, percent white, median age, percentage of development and location) account for about 60% of the variation that we see among violent crime rates.As seen below in Table 2, statistical evidence suggests that for every 0.
4% increase in the unemployment rate, we see an additional violent crime per 1,000 people. This is exactly what we expected to see from the results. Likewise, we see that for every 1% increase in population over 65 living below the poverty level, there is an increase of 4 violent crimes per 1,000 people. Table 2: Regression Statistics Variable Name Unemployment Mobile Home % Old Poor % White % Median Age Median Income Urban % North East Coef? cients 0. 403 0. 032 0.
261 -4. 734 0. 163 0. 00 7. 476 -0.
841 Also, the presence of minorities appears to influence violent crime rates. As seen in the table, the percentage of whites in the area has a negative relationship with violent crime rates; statistics suggest that for every 5% decrease in white population (5% increase in minorities), there is an additional violent crime per 1,000 people. We also see strong statistical evidence that percentage of urban development influences violent crime rates. As we expected, for every 7% increase in development, we see an additional violent crime per 1,000 people.
However, after analyzing the affect that percentage of mobile homes had on the violent crime rate, we found that there is no statistical evidence suggesting an influential relationship. Likewise, we observe that there is little to no statistical evidence suggesting that median age or median income has any influence on violent crime rates. t Stat 1. 601 0. 426 1. 663 -1.
737 1. 132 -0. 099 2. 901 -0. 949 P – value 0.
117 0. 672 0. 104 0. 090 0. 264 0.
921 0. 006 0. 348 SANDY CAMEL CONSULTING 5 VIOLENT CRIMETable 3: Quadratic Regression Variable Unemployment Mobile Home % Old Poor % White % Median Age Median Income Urban % North East Urban%^2 Table 4: Interactive Regression Variable Unemployment Mobile Home % Old Poor % White % Median Age Median Income Urban % North East Urban%*NE Coef? cients 0. 390 0.
033 0. 346 -3. 293 0. 174 0. 000 10.
502 3. 434 -6. 110 t Stat 1. 579 0. 452 2.
141 -1. 174 1. 232 -0. 048 3. 383 1.
273 -1. 674 P – value 0. 122 0. 653 0.
038 0. 247 0. 225 0. 962 0.
002 0. 210 0. 102 Coef? cients 0. 396 0. 040 0. 274 -4.
018 0. 154 0. 000 -4. 168 -1. 158 8. 464 t Stat 1.
561 0. 521 1. 726 -1. 81 1.
058 0. 039 -0. 258 -1. 168 0.
730 P – value 0. 126 0. 605 0. 092 0. 175 0. 296 0.
969 0. 798 0. 250 0. 469 FALL 2010 Extensions One possible relationship that we wanted to analyze was that of the affects of urban development.
Our thinking behind a quadratic relationship was that after an area reaches a certain percentage of development, human interactions will reach a point at which violent crime rates will begin to fall. Nobody wants to commit a crime when there’s a huge group of people around to witness it! After completing the analysis, we find that statistical evidence suggests a linear relationship etween development and violent crime rates is much more likely. We also decided to test if any of our variables had an interactive relationship with each other.Statistical evidence suggests that percentage of urban development and sates in the North East indeed have an interactive relationship just as we had predicted. We also performed a logarithmic transformation to overcome any violations to the equal-variance assumption. As you can see in Table 5 6 SANDY CAMEL CONSULTING VIOLENT CRIME Table 5: Logarithmic Transformation Variable LnUnemployment LnMobile Home % LnOld Poor % LnWhite % LnMedian Age LnMedian Income LnUrban % NE Coef? ients 0.
763 0. 147 0. 753 -0. 503 1. 435 0.
528 1. 511 -0. 257 t Stat 2. 712 1.
848 2. 436 -1. 552 1. 462 0. 835 4. 503 -1.
363 P – value 0. 010 0. 072 0. 019 0.
128 0. 151 0. 408 0. 000 0. 180 FALL 2010 statistical evidence supports an exponential influence of population above 65 years old living below the poverty level.
On average, for each additional 1% increase in population over 65 under the poverty level, we would expect a 0. 75% increase in violent crimes per 1,000 people. The percent of urban development appears to have the greatest impact on violent crime rates (1. %) and statistical evidence now suggests that the log of percentage of mobile homes in fact influences violent crime rates. Conclusion Through statistical analysis we have determined that unemployment rate, percentage of the population over 65 living under the poverty level, percent white, and percent of urban development all influence violent crime rates throughout the United States. As an elected official attempting to close several prisons around the country, this study should help you determine which areas are at high risk of high violent crime rates (poor targets for prison closures).
SANDY CAMEL CONSULTING 7 VIOLENT CRIME FALL 2010 However, as seen by our relatively low coefficient of determination (about 60%) there are numerous other components that add to the variation observed between violent crime rates (making up the other 40%). Perhaps local economies, drug use, alcoholism or family values influence violent crime rates. Unfortunately Sandy Camel Consulting does not have access to this data, so we are unable to consider such scenarios in our statistical analysis. We can, however, make a few recommendations regarding the data that we do have.First, you must determine exactly what level of violent crime you are willing to accept in an area in which you wish to apply prison budget cuts and/or closures.
Based on the data, we feel that any target area should ideally have violent crime rates below 2. 57 per 1,000 people. The best candidates for prison budget cuts/removals will have low unemployment rates (less than 4. 6%), percentage of population over 65 living under the poverty level below 7%, percent of the population that is white more than 90%, and percent of urban development below 67%.We recommend leaving prisons with intact budgets in areas with high unemployment rates (more than 6.
7%), high percentage of poor elderly (more than 13%), low percentage of white population (less than 63%) and high percentage of urban development (more than 87%). Although these suggestions are founded upon statistical evidence, there uncertainty remains. As we stated earlier, several other factors (that we haven’t taken into account) influence violent crime rates throughout the United States. SANDY CAMEL CONSULTING 8