Mike McGuire - DOC by keara


									Mike McGuire Quantitative Data Analysis Final Paper Prof. Wieting 21 December 2004

Comparative Effects of Drugs on Social Breakdown In U.S. Metropolitan Cities

There has been a great deal of upheaval brought about concerning drug use among members of society. The federal and state governments of the U.S. have strict laws making drug use a crime. They proceed in handing out intense punishments and fines for these offenses, as well as creating huge budgets to handle these crimes. It is apparent from various legislations that the U.S. government considers drug use by its citizens to be a very bad thing, citing some inherent danger. My project is designed to examine the statistical relationship between drug use and social breakdown; and to see just how strong the relationship is or if one even exists. I take homicide (the intentional murder of another person) to be the most extreme case of social wrongness and moral depravity. In my mind is the essence of dangerous behavior to society. By examining the relationship between drug use and homicide, we can see to what degree drugs influence this most dangerous behavior, and then use the knowledge to implement accurate, just laws pertaining to drug use. I also use unemployment rates and density rates to see what effect they have on homicide, and compare those relationships to that of drugs and homicide. My sample for the study consists of fifty metropolitan cities in the U.S. The cities used are generally the largest cities in any particular state. For each of these fifty cities the drug arrest rates, unemployment rates, density, and homicide rates were collected for study. “Drug arrest rates” is the main independent variable and it consists of the number of illegal drug arrests per 100,000 people in the year 2001. It is difficult to come to an accurate measure of drug use in a


city since most illegal drug users are not likely to come forward and admit to their behavior. I believe that the drug arrest rate for a city is the most workable measure of the amount of drug use going on in each city. “Unemployment rates” is the second independent variable and it consists of the percentage of unemployed individuals currently in the civilian labor force of each city in the year 2000. “Density” is the third independent variable (taken from 2000 data) which is arrived at by taking the total population of a given city divided by the number of square miles in that city. “Homicide rates” is the dependent variable in the study and it consists of the total number of homicides per 100,000 people in each city during 2001. One of the main goals of this project is to see which of the independent variables has the strongest association with the dependent variable (homicide). They study will give us some indication of which variable poses the greatest threat to civilized society. I hypothesize that drug rates will not have the strongest relationship with homicide. Drugs have often times been the scapegoat for explanations of social breakdown, but I believe that high rates of unemployment and high rates of density create a conflict situation among people that is much more harmful and dangerous to society than drug use. In this study I will employ a series of statistical measures and correlation methods to derive solutions to these hypotheses. My two sources for this project come strictly from hard copy U.S. government documents to ensure accuracy. One source is the County and City Data Book: 2000, published by the U.S. Census Bureau. The other source is The Profile of Drug Indicators: City, published in 2003 by Congressional Information Service, Inc. Both sources were found in the Government Publications Department of the University of Iowa Main Library, using the search engine Lexis Nexis.


Assumptions For this study I assume the assumption basic that all samples are random and independent of each other. I assume that the data of the dependent variable is on the interval-ratio level, and that population variances are equal. Hypotheses H0 (null hypothesis): All independent variables will have the same association with the dependent variable. H1 (research hypothesis): Independent variables will have differing associations with the dependent variable. Also, the association of either unemployment or density to homicide will be greater than the association of drug arrests to homicide. An important point to note about my study is that my essential analysis does not directly examine population means. It is a two-tailed study that focuses on the correlations between independent and dependent variables and comparing those associations. My study is not the type that will adhere to the testing of a hypothesis that relies on a specified population mean statistic for the basis of acceptance or rejection. Sampling Distribution and Test Statistic To test the significance of the relationships I will use a Pearson correlation statistic. The correlation statistic will be compared to relative p values to test for significance in associations. In my study, a p value must be .05 or lower to be considered significant. A p value of .05 is not perfect by any means, but it is a moderately strong basis on which to determine significance. The Pearson correlation values of all independent-dependent variable associations will be compared with each other to determine which association is the strongest. A higher Pearson correlation statistic indicates a stronger relationship.


Summary Table Variable Pairings (correlation) Unemployment- Homicide Drug Arrests- Homicide Density- Homicide Association (Pearson Correlation) .559 .357 -.038 Significance Significant at .01 Not Significant Not Significant

Interpreting Results From the summary table, we can see that unemployment has the strongest association to homicide as well as having the only significant association. Drug arrests have the second strongest association to homicide, but the relationship is not significant. Density has almost no effect on homicide in my study. The lack of a density effect on homicide is contradictory to Durkheimian theory, and is a rather curious finding. My density data could very well have come from an aberrant sample, whose statistics deviate from the norm. The various regions of the country my data were taken from could have caused this aberration of density data. After looking at these results, there is enough evidence to reject the null hypothesis. The data shows that the independent variables have varying associations with homicide. It also shows that drug arrest rates have a weaker association to homicide than do unemployment rates. The drug arresthomicide association was also not significant. These findings support the research hypothesis stated previously. Using control variables to analyze the associations can give us an even better picture of them. When I controlled for density and tested the unemployment-homicide association, the


correlation statistic rose from .559 to .600. The association was still significant independently of density, which further strengthens the claim that unemployment has the strongest and most significant effect on homicide. Since the correlation value rose when controlling for density, we can say that there is some suppression going on when we are not controlling for density. But it still remains that unemployment and homicide have a significant association, whether density is being controlled for or not. Next I controlled for unemployment to test the association between density and homicide. When unemployment was controlled for, the correlation between density and homicide rose from -.038 and became positive at .081. This tells us that unemployment may be suppressing the density-homicide relationship. Controlling for unemployment did not give us a significant density-homicide association, however. The association is still non-significant and almost nonexistent. Conclusions Examining the results shows that drug use has a small to moderate effect on homicide, which I believe to be the greatest signifier of social breakdown. Unemployment has the strongest association with homicide, which tells us that cities with high unemployment rates are likely to have high homicide rates. Density shows to have a very weak effect on homicide. Looking at the interplay between the independent variables shows that there is an interaction relationship among them. This means that the presence of one independent variable does not clearly explain changes in other variables or the situation in general. In this case it is up to social reasoning to determine the real interaction that is going on in the U.S. today. My analysis of the situation coincides with Durkheimian theory of breakdown. I believe (with the support of my findings) that unemployment is the greatest danger to society. It could very well be the cause


that in cities where unemployment is rampant, people become involved in various crimes as a result of the poverty that is a consequence of unemployment. Unemployed individuals (especially those who were recently employed) experience a breakdown in the normal patterns of everyday life (an instance of Durkheimian anomie.) They come to view themselves as alienated, marginalized members of society and therefore are likely to drop out of society’s productive sector. They are then more likely to become involved in criminal behavior such as murder (and incidentally drug use) in order to survive since they find themselves in economic and social conflict with individuals around them. In sum, I believe that unemployment leads to attitudes and behaviors that facilitate social breakdown. Implications Coming to a concrete solution as to the question of what causes social breakdown would probably win me the Nobel Prize. Since there are so many psychological and social factors related to breakdown that I did not study for, it is almost impossible to be conclusive about the subject. Adding variables to this study would allow for more interaction and therefore give a broader solution to homicide and social breakdown. Using samples from rural U.S. cities as opposed to urban ones would probably yield different results, but I felt that drug arrests and homicide rates would be so small in these cities that they would almost be immeasurable. My feeling (with the support of my study) is that in alleviating social breakdown the U.S. government should take the intense spending and focus off of drug crimes and spend more time and money implementing policies that relieve the unemployment problems in the country, since they are most closely linked with social breakdown.


To top