3Q (WillWWWINIWWIWWHIWIWIWNW THS THESIS a I‘ are“ I) L) IliumlMilliIllilll‘t‘lllllflunull 3 1293 01812 9605 LIBRARY Michigan State University This is to certify that the thesis entitled Geographic Information Systems: An effective technological intervention for proactive policing presented by Tae-Jin Chung has been accepted towards fulfillment of the requirements for MS degree in Criminal Justice Major professor Date 5-1&—99 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution .- _ A . -- -..—" r..__< ‘ PLACE IN REFURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE IF to {0 :0 05920054295 mm 34730930 ma animus-nu GEOGRAPHIC INFORMATION SYSTEMS: AN EFFECTIVE TECHNOLOGICAL INTERVENTION FOR PROACTIVE POLICING By Tae-Jin Chung A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE School of Criminal Justice 1 999 ABSTRACT GEOGRAPHIC INFORMATION SYSTEMS: AN EFFECTIVE TECHNOLOGICAL INTERVENTION FOR PROACTIVE POLICING By Tae-Jin Chung Geographic Information Systems (GIS) use technology to predict areas of high crime. GIS has been shown to reduce crime in large cities, but at a substantial cost. Demographic factors such as, emplpywmeptflrflatg, 9.9.9.953?" dingy. flights! lexs'. andsemmynityinxglxemsm influence 6'9 effectiveness. Since so many factors influence Gls, it is necessary to have guidelines to determine whether measures other than GlS might be more cost effective. GIS are effective when law enforcement agencies have exhausted traditional crime fighting methods and are seeking innovative solutions to proactively reduce crime. One way GIS allows law enforcement agencies to more effectively use existing resources is by pinpointing statistically high crime areas, thus creating more time for improving community service. Overall, GIS will result in a lower crime rate, while developing a stronger community relationship with local law enforcement. GIS is an effective crime prevention method that eventually must be implemented in cities worldwide. Copyright by Tae-Jin Chung 1999 DECICATION To my family, especially my parents, whose loving support and encouragement are the sole reason this work was completed. Without Grand mother’s endless love (Veronica Kim), I could not achieve my academic goals. Grand mother (Teresa Kim), Aunt (Agnes Chung), are perpetually in my thoughts and prayers. ACKNOWLEDGEMENTS I would like to express the special appreciation to Dr. David L. Carter. Without his support, both academically and personally, this work would not have been possible. An appreciation is also extended to my committee members, Dr. Dennis Payne and Dr.Frank Horvath, for their interest and advice. I would also like to extend my sincere gratitude to Dr. Gill-Chin Lim, Dean of KDI School of lntemational Policy and Management. He was the first person who suggested me to study this research topic. His interests and concerns in my work have been a strong motivation for attaining my goals. I am also deeply grateful to Susan Trowjanowiz, whose organization, guidance, and dedication for my academic life at Michigan State University. Hazel Harden, she always provided me with the most valuable information of the school. Acknowledgements are made to Dan Pum' and Lt. Steven Person at Lansing Police Department. They spent a lot of time for helping my research. Amy Higgins at Michigan State Police helped me to collect the UCR data. Finally, to my friends and fellow students at Michigan State University who have supported me to finish this research without difficulty. Especially, Young- Rae Kim (Ph. D. candidate in Park & Recreation) and John F. Kennedy (Ph.D. candidate at Sam Houston). TABLE OF CONTENTS LIST OF TABLES ............................................................................................... Vll LIST OF ABBREVIATIONS .............................................................................. Vlll CHAPTER 1 Introduction ................................................................................................ 1 Statement of the Problem .......................................................................... 2 The Purpose of Study ................................................................................ 3 Limitation of the Study ................................................................................ 3 Hypothesis ................................................................................................. 4 CHAPTER 2 r V W ’ Police Technologies ................................................................................... 5 History of GIS ............................................................................................. 6 GIS in Law Enforcement Agencies ............................................................. 7 Benefits of Using GIS ................................................................................. 9 Barriers of Implementing GIS ................................................................... 10 Crime and Place ....................................................................................... 11 Hot Spot..................-. ................................................................................ 12 CHAPTER 3 V? "r“; " Data Collection ..... '.' ................................................................................... 14 Information Sources ................................................................................. 15 Reliability and Validity .............................................................................. 16 Data Analysis ........................................................................................... 17 CHAPTER 4 T - Hypothesis 1 ............................................................................................ 19 Hypothesis 2 ............................................................................................ 21 CHAPTER 5 - r‘ Section A— GIS in Proactive Policing ..................................................... 25 Section 8- Limitations of Findings .......................................................... 26 Section C- Implications .......................................................................... 28 APPENDIX A Total number of Using GIS in Police Departments .......... 31' APPENDIX B 1996 UCR Reporting Agencies in Michigan ......................... 32 APPENDIX C GIS in Lansing Police Department ....................................... 33 BIBLIOGRAPHY ................................................................................................. 35 vi TABLE 1. TABLE 2. TABLE 3. TABLE 4. TABLE 5. TABLE 6. LIST OF TABLES INDEPENDENT SAMPLE T TEST OF TOTAL NUMBER OF .......... ARRESTS IN LANSING FROM 1991-1998 ................................... 20 VARIABLE CODING AND PRELIMINARY DESCRIPTIVE .............. STATISTICS FOR TOTAL NUMBER OF ARRESTS IN LANSING .. FROM 1991-1998 .......................................................................... 20 INDEPENDENT SAMPLE T TEST OF TOTAL NUMBER OF .......... ARRESTS IN LANSING FROM 1991- 1998 ................................... 21 INDEPENDENT SAMPLE T TEST OF TOTAL NUMBER OF .......... REPORTED INDEX CRIMES IN LANSING FROM 1991-1998 ..... 22 APPENDIX E VARIABLE CODING AND PRELIMINARY DESCRIPTIVE .............. STATISTICS FOR TOTAL NUMBER OF REPORTED INDEX ........ CRIMES FROM 1991-1998 ........................................................... 23 INDEPENDENT SAMPLE T TEST OF TOTAL NUMBER OF .......... REPORTED INDEX CRIMES IN LANSING FROM 1991-1998 ..... 24 vii LIST OF ABBREVIATIONS GIS ........................................................... Geographic Information Systems viii Chapter 1 INTRODUCTION Geographic Information Systems (GIS) is a computerized mapping system, which allows police departments to identify criminal hot spots in their jurisdiction occur. Based on collected data, police departments can deploy more officers to these hot spots while still providing other services to the community. GIS has been used by law enforcement agencies for more than 10 years. However, many other police departments have not implemented GIS to date. There are several possible barriers encountered when implementing GIS, some of which are substantial cost, training. Currently, GIS is becoming important within the law enforcement agency. GIS allows law enforcement agencies to analyze and correlate data sources to make detailed crime incidents and relate factors within a community or other geographic area. As technology advances, criminals are getting smarter and crimes are getting more complicate. Some criminals use more sophisticated technology than the police do. In order to maintain public safety in the 21"t century, the information age, police should be equipped with high technology such as GIS. GIS is an important technology that police departments should implement to provide a higher quality of service to their clients, the community. Statement of the Prgblem GIS has been identified as a very useful system in many different areas such as military, forestry and other sciences. However, there has been no systematic data existed to prove the effectiveness of using GIS in Policing (Lavign, 199). Moreover, it is not widely employed by police departments and the national crime rate has been dropping continuously. Thus, it is difficult to prove the effectiveness of GIS in the law enforcement agency. Generally, it may be assumed that advanced technology provides some advantages when utilized with the older methods of gathering information. However, science always requires sufficient evidence to prove any claims of fact. Although overall society has been changed tremendously as a result of technology, police departments are still far behind when it comes to such technology. I The author will test whether the use of GIS for targeting crime hot spots relabs to the total number of arrests. Particularly, the author will test to determine whether a police department that utilizes GIS has attributed to increase arrest rate. The author also will test whether the use of GIS for targeting crime hotspots has attributed to the increase in the total number of reported index crimes. If GIS is found to be effective in reducing crime and increasing community relationships, it should be implemented by more police departments in the US. and worldwide. The Purpgse of Study The purpose of this study is to determine the relationship between the use of GIS for targeting crime hot spots and the total number of arrests in order to help police department assure that GIS is a very useful method for proactive policing. This will be achieved by examining the Uniform Crime Report (UCR) data from Michigan State Police. Specially, the author will look at the UCR arrest data and index crime data from 1991 to 1997 in Lansing, Michigan and the State of Michigan. It is believed that this data will show whether GIS has an influence on the arrest rate and index crime rate or not. The main focus of this research is to explain the effectiveness of using the GIS technology in policing. To provide the better service to community, police should be equipped with the best technology in order to their job effectively and efficiently. Both crime data and current academic literature will be presented to show if GIS effectively works in community policing. The research hypotheses will be tested, and policy implication will be made for using GIS in law enforcement agencies. Limitation of the Study This study contains arrest data and index crime data obtained from Michigan State Police and literature information collected from many different academic resources. However, these data and literatures are relatively few to support author’s hypotheses in terms of generalization. There are many limitations to be involved with this study. These limitations are including: —L . No current crime data locally available (only up to 1997). 2. No systematic data available for the use of GIS in policing. 3. Lack of research regarding the use of GIS in policing. 4. No other countries use GIS for policing other than the United States of America. In addition, contemporary administrative and technical concerns were identified by current available literature. Analysis and evaluation of academic journal articles focused on factors related to GIS. This study was concerned only with these issues in the United States. Therefore, it is very difficult to apply this research result to other countries. Hypothesis 1. H0: There is no significant relationship between the use of GIS for targeting crime hot spots and the total number of arrests in Lansing. 2. H1: There is a significant relationship between the use of GIS for targeting crime hot spots and the total number of arrests in Lansing. 1. H0: There is no significant relationship between the use of GIS for targeting crime hot spots and the total number of reported index crimes in Lansing. 2. H1: There is a significant relationship between the use of GIS for targeting crime hot spots and the total number of reported index crimes in Lansing. Chapter 2 REVIEW OF LITERATURE Today, GIS is becoming more popular among law enforcement agencies. This comes at a time when the cost for computer hardware and software is declining, the technological efficiency continues to improve, access to digital calls for service and incident data within police departments is increasing. Therefore, many criminal justice researchers are currently focusing GIS on more than ever. Although there are not many articles regarding GIS, the few that exist illustrate some very good example of GIS usage. Before examining the current literature on Geographic lnforrnation Systems, the author will discuss some of the findings from studies of other policing approaches. These approaches have been used and are also considered effective policing interventions. Police Technolggies American police departments largely converted from foot to automobile patrol between the 1930s and the present. Then they have patrol cars in conjunction with telephones and two-way radios. The telephone made it possible for the ordinary citizen to summon the police, and the combination of two-way radios and patrol cars allowed the police to respond quickly (Gay, 1977). Later, many police technologies have developed to Computer Aided Dispatch, Mobile Data Terminals, and Automatic Vehicle Location have become commonplace in many of the larger field agencies (Sparrow, 1991). However, these technologies M““ "'- - .- .—;., . er * ’ é w ‘ m .. ‘ V: “"H K‘N“ .2" , , I \\( are not considered é ‘proactive policing’ methods (Hart, 31990). These systems are not usedentire’ly' to prevent crime before it happens. The new terminology of a“ “technological proactive policing” emphasizes the use of technology, such as u H,» 5‘ ~ M , .. “w . , M - .,_. ‘V _,.,,-w---'- H GIS: to gather and monitor information on criminal activities. GIS Is not only NICO-n..— -.—->-0""' "" ' 7‘ “in. rs“ W targeted to prevent crimes, but also to increase community relationships with local law enforcement agencies. Technological proactive policing Is based on r_...——o- #4 Mun-Hav- Q'" community policing (Chung, 1998). As Sparrow (1991) reported, “Community policing seeks to revitalize that partnership for two major reasons: First, to produce a cooperative process of identifying police priorities and Second, to provide a more effective methods of achieving those jointly nominated goals” (p.26). Histog of GIS GIS has been around in manual form for a long time. For example, the French cartographer Louis Alexandre Berthier drew up a map of the 1781 Battle of Yorktown in the American Revolution which contained hinged overlays showing troop movements (Star & Estes, 1990). Such manual geographic information systems were useful in their day, but the people who used them had to contend with a number of problems. There were no standardized scales and the information was not always up-to-date (Rogers & Craig, 1996). Graphical crime pattern analysis in the form of pin maps has been used in police departments for many years. Typically, an analyst puts a pinwr‘epresenting a crime on a map to note how crimes cluster around other crimes, the land use I characteristics of the location, and so on. The mechanical task of plotting the ”WW: --I -‘ crimes was independent of any other data management or analytic task and must be kept up to date to be used. As the characteristics of the crime or the W" location multiply, or as the crimes proliferate, the density and complexrty of the m... a». _.—¢- mpg—«v—u- w" ”"" \W‘M.‘ map can become unmanageable Severely limited choices have to be made about which dimensions of data to display with different pins, or the data have to 'u—n‘.“ Hm.“ “ be culled frequently However, culling undated or indistinguishable pins cap be a burdensome task. The analyst must either remember other charactenstIcs of I “In. . crimes or search manually through a separate file to tie locations to cnmes ”Under these condItIons the maps frequently fall into disuse (Maltz,1991). 1-.“ m Before the day of computer, the map was the spatial database. In order to produce new information from maps, humans had to read and study them with a great deal of effort. There are also limitation of memory. However,_GlS can hold as much data as the computer hardware will allow. GIS can combine, add, flhr" hhhh M... subtract, multiply and divide data, and perform many other such operations. With _-.4 s-- GIS, users can answer extremely complex question (Parker, 1988). Initially, '._‘ “a: “,1..." .J a”. ”M computerized mapping started In the military, but new police are using it to map crime. The system combines geographic information with crime statistics (Pilant, 1997). Recently, GIS combines its technology mm the Global PosItIonIng Systems. GIS in Law Enforcement Agencies In 1996, the United States, had about 20.000(18769 in 1996) police, special police, sheriff departments (BIJ, 1998). Among those departments, less than 5% of the departments used GIS to its full potential (see Appendix A). Because there was a lack of knowledge and personnel that can operate the system effectively (Lavign, 1999). However, GIS interest among law enforcement agency executives and planners appears to be growing (Mamalian & Lavigne, 1999). Today computer technology has expanded so quickly that the market for computers has brought the price down substantially making GIS less expensive to acquire for law enforcement agencies. The cost of the system can also be “.4 -. .m“. ‘ ‘w- I...“ “ ‘fim...’ distributed among all agencres using the GIS such as police, fire, emergency mm. .. ”M M. ro- medical service, and public transportation. All of these departments utilize a .___,,__,r....._._t..-- «m, M. B..- -H...‘ w__.,‘v common street file and possibly a Parcel file. This assists in sharing the cost of _,'-w-r""" M MM- developing the base map and the cost of purchasing a computer server that will . WA..- house the data (Rogers & Craig, 1996). Efficiency deals with the cost of what Mam-m - W.” - police do In relation to what they achieve. The question is not what the police m.--— .n—Ifl'" accomplish but Whether the cost of What they do Is minimized. Efficiency and .‘N ”liq-Iv“... effectiveness represent the cost and benefits of police activity (Bayley, 1994). Q . --.I,.'-r’-‘ In general, when the crime rate increases, police departments seek to _.——‘— ..-..-_. increase their number of human resources allocated to a given problem To show that GIS re more effective than just hiring more officers, it needs to show the cost of hiring more officers is more expensive than implementing GIS. lg addition, GIS Is not used only by patrol units, but also crime investigation which fwd-“AM »_ ‘1."- -"\-_- flrfi.-n . traditionally has involved hundreds of man-hours gathering and evaluating fi/v", ‘AFM'. .-_ information (Pilant, 1993). My» 0" 1." r Benefits of Using GIS In comparison to using a wall rInap, GIS affords police officers the abIIIty to w t—A‘CI-q’“ “.1” ask more focused questions about crimes on their specific beats (Witkin,1997). MW: nw...~n—‘~-o F—‘n- Thus patrol officers know where they have to go and how to spend their b'me more effectively fighting crimes in their area. It also helps to identify emerging Wall-«mm. .~, a». Wane—(J ”‘4“ " ”WA... crime problems (hot spots) sooner by downloadrng crime deta more often, by MA” ‘ ‘Ls rm. «4t 9‘} simplifying the mapping software so more employees can utilize it, by creating MN... .- wag—7v... . 5-va “4.; I‘H‘ 39 visual maps of crime data Instead of utIIIzrng computer print oIIIts. It also allows va—Mr M .._..... - .. .. . NM. researchers better prediction capabIIIfies to foresee future crime patterns (Block M'— WWW.- «w "W m. m.- m ~‘fiWM“~..-wm-»~» N, & Green, 1997). Serious crime in New York City dropped 27% between 1993 and 1995, while a 2 % drop was recorded nationwide over the same period after using GIS. It indicated that GIS might significantly reduce crime. GIS allows citizens in the commuan to participate in the prevention of .4 , 4- ‘v . M M‘f'u ru' KM crime in their own neIghborhoods Both police and community members are “ “duo-1‘1“- , use; aware of where the problems are; this Increases the poIIce/communIty 1M ~_W'mm "v ‘- partnership In reducing crime (Radelat & Carter, 1994). It also Increases safety I ~., “*4 -thw 'D‘Nb' "q" ‘—-I-—d- ‘ ____ W and decreases victimization In neIghborhoods by monItonng ongoing problems a Wm'fizfi 1"” " M‘.‘ -‘ ‘M WWWq F. v I . such as a hot drug areas abandoned buildings, liquor store, prostItutIon etc u—Uw' "w" ""“~"flh~~. M ‘ ” v A”; rm ‘4 (Block & Green 1994). The maps were very useful as a method of transmIttIng I'I‘QVH- ‘n‘ 1" 1" M‘rM“ N L. ‘M is L‘s-Hid wH-m.‘ \ complicated information to community organizations. For the community groups, .Nm—ns‘nwfi ”W“ Ir.wm~mmmm M” ‘Wt no “r!- they are not only a graphic way of communicating, but they werealsowa Irnpte powerful tool because the maps made community residents’ concerns tangible 4M4-w' and concrete to the police and are a professional and polished means of W‘s-r _ .‘ transmitting their concerns. Thus, it Is believed that GIS Is one of the most . m._ ”w. Mar-n -""."‘"‘“ W- powerful tools available to connect police and community. Barriers of Implementing GIS Many articles published on GIS indicate that economic issues are the main factors when deciding to implement GIS (Rogers & Craig, 1996). There are also other barriers such as cultural and political considerations. Some police cultures are resistant to the idea of technology being used in their work. Because of this resistance and a general lack of knowledge in technical matters, they prefer to be dependent more on personal expen'ence than relying on statistically based information (Block & Green, 1994). Inprder to change the perception thatgls is difficult to operate, software training can be PFOYide, M‘ show how large quantities of police information can be displayed Visually on a” “mg-mo; ‘Imw v " W” map. In addition, we can demonstrate how simple it Is to see where crime hot qua—“p..- _H_ w. spots are located. For political reasons many politicians do not want police to become more .4”- technologically advanced fearing that the police will have too much power. www.- MW“ Showing the effectiveness of reducing the cn'me rate by new systems as well as using the police department's human resources can eliminate political barriers more effectively. In addition, other organizations and departments can share L.‘t.. \Wm-.~h. b _’V\_.p.,.2t GIS. For example, mapping hydrants underground hazards and emergency H.“ facilities would facilitate faster reaction in times of disaster (Binkley, 1991). This ”-‘s‘fi‘d' I would go a long way towards making GIS a more public system and not M.’ 'v‘OD—‘hg “Moi—IE.“ ,m .—-¢ 10 exclusively In the domain of the police alone. It will eventually become a true [W _ public-share system. Crime and Place Researchers and practItIoners in Criminal Justice recently began to shift their focus from people to places (Taylor, 1997). They realized that the specific ,,-- M”-M"’ "’ J... ' ' . . , , .- u..-“ ..- .-,._\W Humans,“ place where offenses occur is more important than the people who committed Mw-«..-m.nw~ ‘ ~v- * -.~»~ - ~. .-.-. » 2- - - « v---.~~,.. the offenses are (Weisburd, 1997). It Is more difficult to predict human behavior (Wt-u ”was...” " than place where crime will occur in tIheIfuture. Not every crime, but mOSt, crimes .,.—- ¢' occur in some specific locations. For example, domestic violence occurs mostly " -‘vh. at the same place as long as offender and victim are living together (Sherman, 'wwn MM- 1992). Burglary occurs in the same place if there Is a source of vulnerability. The ‘- ,_._..-_.—.~n - ' ‘ same house Is to be reburglarized after the initial burglary (Polvi et al., 1991). W-fl‘a JAN-f ‘»'.- Some crimes are most likely to occur at the same place (hot spot) such as rp-r’ -m 'H‘ flaw- _-___,,. V” assault drug and car-theft, etc. Those crimes are linked together with certain M environmental conditions. For example assaults that occur near nightclubs or »- bars, car-theft at unlighted parking lots, and drug dealing takes place at” demolished buildings or place Without management However, some kind of """“"‘.' n “n “W.“ IW m-‘ general street crimes, such as robbery, where neither victim nor offender Is fixed in place, makes this relatIonshIp weaker (Eck & Wartell, 1998). It Is more difficult «a... 1. m 5.51:“ am“) to predict crimes that do not link With certain environmental conditions. However, 'q-‘fln."fi‘kn. some fixed location crimes have been improved by twigs“ intervention) (Taylor, 1997). In addition, the place-centered view has also lead to police “*‘I-Jum success and shows potential for other areas of criminal justice practice, such as f0"-.- ‘7“ 11 parole and probatipIrIII (Taylor, 1997). Shifting from an offender-based to a place- M based criminological theory or more accurately, developing the place-based theory as a complement to the individual focus requires thinking in new ways (Weiburd, 1997). The new concept needs to be supported hy the new crime control and prevention methods such as 9'91; The use of GIS In proactive /wmwm ‘ ' M—J ». 4* u... -. ..-. v.“ policing will result In IrIeduciInIg crime rate ahd Increasing the better communIty ,Mwwww . . -.... relationship. The fundamental concept of place-based theory and the use of ISIS W W in policing aII'IeI based on the philosophy of communityIpolicing. Hot Spot Hot spot is a location of extremely high crime. The term is borrowed from geology. A hot 595’th may be a singlIIeIII address, a cluster of addresses close to one l.‘ .m“‘ another, a segment of street block, an entire streetIIbloIck or two, or an intersection n-rv -...,.., 4...... 4"“ (Taylor, 1997). According to Sherman (19II9I5), many service calls are from a ___,.,__.t... (W. h. w- - w ’- g n I 1...; — 4...? W- relatively small number of addresses. The hot spot concept has been moved from the calls for service to the location of drug market, uncivil or disorderly behaviors such as prostitution (Maltz et al., 1991). The underlying WIZ‘PT Perm-Wed and pla9e-9§§e.d.ta9t9r§.ma..x. contribute to VIctImIzatIon state deflencefl andII III'sk hetegoge enity. High -fl.u,‘vw OW“ victimization risk may be dependent on the state the person is in at the time of victimization. For example, if a burglary victim has had his OII‘ her front door D’M "-my‘H-v'“ " kW‘w—cmffmmu iimmied, he 9* She is m9“!,Sueesptibletoanotherpvrglary anti! 39.919.25.18... repaired; that Is called state dependence. By contrast, some people are more Lug-an- 12 likely to be victinmjged‘gecause of their PEPJtS- routines, or occupations; this is "v--‘ INJW ' I'M-LN“ K". mfih‘sfl W. L n... "4..."...JAL ~ 1‘ called risk heterogeneity (Lauritsenfi. Quidnve‘thlggm; ”M— Sometimes. police officers would find that residents’ concerns centered on site that were neither crime based nor 911 call based hot spots (Maltz et al., 1991). Therefore, it is important to understand how activities in, attitudes toward, and crime in the hot spot itself are shaped by the surrounding context (Rosenbaum 8. Lavrakas, 1995). 13 Chapter 3 METHODOLOGY Data Collection The purpose of this research is to show the effectiveness of using GIS for targeting crime hot spots. Currently, there are not many police departments using GIS. The author will try to prove the effectiveness of using GIS in policing by comparing the total number of arrests before and after. The author believes that GIS has influence on the total number of arrests and the — ,.__.. ....___ ..._-- -. . __ _ “Au—.— -—‘—— total number of reported“ index crimes in Lansing. -flbip" u__-—.__—-—-- -'—"' The author will begin this study by collecting arrest records from Lansing police department. The author began to collect the data from the spring of 1998 to spring of 1999. The data contains all arrest record from 1991 to 1998. The author will record the number of arrest, types and numbers of change. This data will enable the author to test the hypothesis. This data will be used to analyze the relationship between the use of GIS for targeting crime hot spots and the total number of arrests in Lansing. For the second hypothesis, the author will use information obtained from the Michigan State Police. The data was collected during the spring of 1999. The data is compiled from all of Michigan (see Appendix B), and create the UCR for the State of Michigan. The goal of this research Isto‘determme how police departments use GIS effectively In terms of arrest rate. GIS technology Is relatively new and ,--.-—ur—- -——r—-—- H..- w- _._-_—.- ’-.—-a-.._-—--—--—-"'-‘-‘—‘—- 14 rare in policing; there is not much research on the topic. Generalization of other findings has yet to be established, so explanatory research is still needed in this area. Previous research has reported that GIS is an effective method in policing. However, there is no data that indicates the actual outcome with the number. The author believes that GIS dedicates itself to reduce the crime rates, but it is very difficult to prove the fact while overall national crime rate has been dropping. This study will be limited by the use of arr_e_s_t_reqo;rd_s,andee f total number oanesfiomLansing and Michigan State / Police, however, these sources seem to provide the most accurate and complete information for the author’s research proposal. Informa_tion Sources An analysis was conducted on the data maintained by Lansing and Michigan State Police department. The primary reason that the data collect from Lansing and Michigan is not only convenience of collecting data but also qualification of source of this research. There are not many police departments that use GIS for meir policing nationally. Lansing police department began to use GIS since 1997 (see Appendix C). It seems that it has been very effective in their jurisdiction. Lansing police department has 254 full time police officers, and their jurisdiction covers the area that has a population of total 127,321. Lansing State Journal (February, 1999) reported that the crime rate has decreased significantly from July 1998 to December 15 1998, murders decreased by 33.3 percent, burglaries by 28.5 percent, larcenies by 12.5 percent and auto thefts by 10.4 percent. War, Hollister, attributed the decrease in the reported crime statistics to an aggressive (community policing str'ategngccording to Lansing police Lt. Ray Hall, " afose community and an increaseninttechnology both playedflmajgr roles In imptoving the‘department' ‘ h The second reason to use the data from Michigan State Police is to compare the effectiveness of using GIS for targetingcrime hot §P.9.§.§..i_'.‘ Lansing in the larger context. The total number of UCR reporting agencies in r' Cal-.‘tluo‘ Michigan is 603 (Michigan Uniform Crime Report, 1996). This data also consists of Wber of arrests and 991.9,”??? of reported indexmes. By examining this data the author hope to identify the pattern of crime rate and arrest rate in Michigan from 1991 to 1998. Reliability and Validity The reliability of this research is determined by the measurement technique that the author will apply in this particular research. In other words, measurement reliability Is roughly the same as measurement consistency or M” Hid. uh-n—Mv"-" ' stability (Babbie 8. Maxfield, 1998). The author believes that there Is reliability __..._—__..—- .— —-4--—-I-‘-—-~—-|r+-..* I ____._-..----— M” in this research since the author collected and used secondaty data from the “M local and state agencies. The state government has compiled the data, so it could be more reliable than any other data. 16 , V! S 2. A”... 779% I: ((3.5%; /? flat *1.-- / [$(J’A ‘\ Cl" / The validity of this research is determined by the operational definition that reflects the concept under consideration (Babbie 8. Maxfield, 1998). In this research, the author measure the effectiveness of using GIS for targeting crime hot spots with arrest; crime rate in Lansing. The author believes that it has a validity of research since arrest and came. rate has been considered as _._,.....- ...-.-r.. "*"m ‘wun'm WM”--. indicators of crime actIvrty )The arrest and crime rates reflect the concept of author’s conceptual and operational definitions in the research. Generally, criminal justice agencies collect data for their own use, not for the use of researchers. So it is difficult to adapt agency records for specific research purposes. However, this thesis research shows that concepts are identical to measure the actual measures maintained by criminal justice agencies. Data Analysis To analyze this data, the author will use a T test for statistical significance to determine a relationship between my independent and dependent variables. Author’s variables (GIS, Type of Crime, Location) are nominal and (Arrest rate) is the ratio descriptions of traits (Lurigio et al., 1997). The T test is used for determining levels of significance for those types of variables. The independent-samples t test is used when two unrelated samples from normal distribution, or the sample size, must be large enough to compensate for non-nonnality (Norusis, 1997). This will tell author whether any average difference between the after and before intervention value (Norusis, 1997). 17 In order for the first hypothesis to be supported, collected data analysis would provide a relationship that is statistically significant between the use of GIS for targeting crime hotspots and the total number of arrests. Statistical significance would indicate that observed frequencies of these factors are not occurring by chance. In other words, GIS intervention will be more likely to increase arrest rates. For second hypothesis to be supported, the author would expect a statistically significant relationship between the use of GIS for targeting crime hot spots and the total number of reported index crimes. If these relationships are statistically significant, GIS will be related to an increase in total number of arrests and total number of reported index crimes in a manner that is not through chance or by coincidence. It is believed that this result will indicate that GIS does work for arrest and crime rates. If the hypothesis is supported, the author expects his results to be similar to the expectations made by other findings 18 Chapter 4 FINDINGS Hypothesis 1 The independent sample t-test is used to examine the relation between the use of GIS for targeting crime hot spots and total number of arrests in Lansing. The test result indicates that there is no significant relationship between those variables. In the table 1, consider the column labeled Equal variances assumed. The observed difference of 400 arrests, the statistics is .648 (T o calculate the t statistics, divide the observed difference of 400 by 617 the standard error of the difference estimate when the two population variances are assumed to be equal). The degrees of freedom for the t statistics are 6. The observed two-tailed significance level is 0.541. This indicates that only 54% of the time would author expect to see a sample difference of 54 arrests or larger, when two population means are equal. Since 54% is greater than 5%, author fail to reject the null hypothesis that there is no relation between the use of GIS for targeting crime hot spots and total number of arrests in Lansing. The 95% confidence interval for the true difference is from -1,111 arrestees to 1,911 arrests. This tells author it’s likely that the true mean difference is anywhere from -1,000 from +2,000 arrests. Since author’s observed significance level for the test that the two population’s means are equal was greater than 5%, the 95% confidence interval will contain the value of 0. 19 Table 1. Indegndent sample t test of total number of arrests in Lansing from 1991-1998 t-test of Equality of Means 95%confidence interval t df Sig. MD S.E. Lower Upper Equal variance .648 6 .541 400.16 617.79 -1111 1911 assumed Table 2. contains the descriptive statistics for the two groups. Before using GIS and after using GIS for targeting crime hot spots. The difference is 400 persons arrested. The average arrests from 1991 to 1996 are 7,868. The average arrests from 1997 to 1998 are 7,468. This difference indicates that there is no statistical change of arrests after using GIS for targeting crime hot spots in Lansing. Table 2. Variable coding and preliminary descriptive statistics for total number of arrests in Lansing from 1991-1998 Variable Coding Value N Mean 8D. GIS Yes 2 2 7468.5000 195.8686 No 1 6 7868.667 824.2239 Note. GIS= Geographic lnforrnation Systems 20 Table3. lndegndent sample ttest of total number of arrests in Lansing from 1991-1998 Levene’s Test for Equality of Variances F Sig. Equal variances assumed 1.620 .250 Based on the Levene test in Table3. there is no reason to doubt that the population variances are equal, so you can use the t value in the row labeled Equal to test the null hypothesis that in the population. There is no relation between the use of GIS for targeting crime hot spots and total number of arrests in Lansing. The two-tailed significance level is 0.541 (see Table 1), so don’t reject the null hypothesis. As expected, the 95% confidence interval for the mean difference includes the value of 0. Hypothesis 2 The independent sample t-test is used to examine the relation between the use of GIS for targeting crime hot spots and total number of reported index crimes in Lansing. The test result indicates that there is no significant relationship MI! W between those variables. In the table 4, consider the column label Equal variances assumed. The observed difference of 251 crime offenses, the statistics is .583 (To calculate the t statistics, divide the observed difference of 251 by 430, the standard error of the difference estimate when the two population variances are assumed to be equal). The degrees of freedom for the t statistics are 6. 21 The observed two-tailed significance level is 0.581. This indicates that only 58% of the time would author expect to see a sample difference of 58 crime offenses or larger, when two population means are equal. Since 84% is greater than 5%, the author fails to reject the null hypothesis that there is no relation between the use of GIS for targeting crime hot spots and total number of reported index crimes in Lansing. The 95% confidence interval for the true difference is from -802 crime offenses to 1,304 crime offenses. This indicates that the true mean difference is anywhere from -802 from +1.304 crime offenses. Since the observed significance level for the test that the two population’s means are equal was greater than 5%, the 95% confidence interval will contain the value of 0. Table 4. Independent sample t test of total number of reported index crimes in Lansing from 1991-1998 t-test of Equality of Means 95%confidence interval t df Sig. MD S.E. Lower Upper Equal variance .583 6 .581 251.00 430.39 —802 1304 assumed Table 5. contains the descriptive statistics for the two groups. Before using GIS and after using GIS for targeting crime hot spots. The difference is 251 crime offenses. The average arrests from 1991 to 1996 are 9650 offenses. The average arrests from 1997 to 1998 are 9399 offenses. This difference indicates that there is no statistical change of arrests after using GIS for targeting crime hot spots in Lansing. Table 5. Variable coding and preliminagy descriptive statistics for total number of reported in_dex crimes from 1991-1998 Variable Coding Value N Mean 8D GIS Yes 2 2 9399.000 197.9899 No 1 6 9650.000 570.6014 Note. GIS= Geographic Information Systems Based on the Levene test in Table6, there is no reason to doubt that the population variances are equal, so you can use the t value in the row labeled Equal to test the null hypothesis that in the population. There is no relation between the use of GIS for targeting crime hot spots and total number of arrests in Lansing. The two-tailed significance level is 0.581 (see Table 4), so don’t reject the null hypothesis. As expected, the 95% confidence interval for the mean difference includes the value of 0. 23 Table6. Independent sample t test of total number of reported index crimes in Lansing from 1991-1998 Levene’s Test for Equality of Variances F Sig. Equal variances assumed 3.213 .123 24 Chapter 5 DISCUSSION AND CONCLUSION This chapter is divided into three sections. It begins with a discussion of the study results and how these findings may relate to the contemporary policing. Limitations of the study are discussed in the second section. The third and final section addresses the direction of future research. Section A— GIS in Proactive Policing In this study, the use of GIS for targeting crime hot spots, and arrest rate and crime rate, resulted in that there is no relationship between these variables. Initially, it was expected that the use of GIS for targeting crime hot spots significantly affect both the arrest and crime rates in Lansing. However, statistical results for this thesis research indicate that there is no relationship between independent and dependent variables. The research results are also different from those expectations from other literatures. In spite of these statistical results, it is believed that some relationships exist between the use of GIS for targeting crime hot spots, and arrest and crime rate. Because the total number of arrests and total number of reported index crimes in Lansing have not been changed significantly before and after using GIS. Although the national crime rate has been dropped confinuously, the crime rate and arrest rate in Lansing do not appear to have been influenced by the national trend. 25 Section B- Limitations of Findings The major limitation In interpreting the above information concerns ..... “M M... A.“ M ,..........I generalizabmlt is highly likely that the findings are not generalizable to all my.” times and places. The primary reason for this concern steams from the use of GIS in Lansing is not representative of any other cities or countries. Since there is less than 5% of police departments use that GIS for policing, Lansing police do .flmgw» -—»- ‘w. - ,I, ,n’wwwwn’fm W not represent the overall law enforcement agencies. The author also found that. _‘ “mg. vxrs W‘ my“ h. nut-l ‘- "' n‘ (a. “.4 there are many factors to influence arrest and crime rate other than GIS. These 17“...“ '\"s . - In. ”M ‘W‘Aw’gwxfl “-.‘\ factors are not considered when the author began to investigate the relationship between GIS, and arrest and crime rate in Lansing. W"""‘ of residents are not considered In this research. Without understanding this information, it is not useful to test the relationship between the use of GIS for targeting crime hot spots, and arrest and crime rate. Since GIS is not the only factor to reduce the crime rate, the effectiveness of using GIS can not be p- .An... n..- a“. .s‘wmhrJ‘ wt. v~I-Wn. 1’ I I I explained by itself alone. For example, low employment rate of residents Is an M important factor to accelerate the criminal activity In their area. Insufficient Mffl' W'J’uw'ue economic resources of human society create the competItIon for survivals often .rv f_‘b\"wfl w‘ H" V ‘ tw 0- u‘ \N‘Ih' HL“ ld‘ahfliw“ expressed by cnmInal actIvrtIes The average age of residents Is also an “WHM .in ' “—m "0,, . "hr ’ gal-Cub .r- .I you! km hm 1W .wmw- '7‘" important factor to influence the crime rate. Younger generation in the population ran—ow is more likely to cause the crime activity than elder generation. The education W -~ . W level of residents Is also an important factor to Influence the crime rate: Since ,fil-JWM"" , ., -‘-.~»« M ' ’P*I’ lower educatronalbackgmundpeeplemaehkely to involve the criminal activity 9"- ‘r-‘z ~-.w 6,» (m than higher eduIcIaItiIenal baekgroundIIpeeeIIe. Therefore, it is difficult to establish the effectiveness of using GIS in policing with the only arrest or' crime offense data. Depending on the demgraphic factors, crime rate cguld. be affected wan-- J... ”A M significantly. In order to establish the effectiveness of using GIS in policing, m Wu. “I. '1" demographic factors must be studied in advance. Second, legal aspects are not considered In this research. For example, it ---. ‘_ -._; - ‘-.-- . - -‘.,. _..-..W-- depends on percentage of prosecutIon and sentencIng gUIdeIIne In the V» fly .“ w m- jurisdiction, both arrest and crime rate will be different. The higher level of WW‘W rub—t \W prosecution sends the more people to the court. Conservative sentence guidelines cause more people to receive guilty decisions. However, there is some intervention influence on the research such as plea-bargaining, fine and other alternatives. Third, types of crimes fer GIS are not considered In this research, because ‘ m ’1'”me ~M_.—....__..- v GIS works for certain crimes much better than others do. For example, cer theft, M‘m'v ' “' r burglary and domestic violence are more likely to be influenced by IIGIIS. w” wr However, murder, assault, and kidnap are not likely to be Influenced since those 5“ _____. Nu.— N ”,4 my ,‘ankvfla-J' Q" 14" -"‘_' "T'N" “v «Irv-H"... crimes are not place-based crimes. Therefore, this thesis research has a .- ,. ., .a-I... W “-4. '_L.-- - .IH-fl-w._.._.—H—_— -- r” " -Ir-—-..._,_,m limitation to prove the effectiveness of using GIS without considering these kinds of problems. The author used the total number of arrests; total number of reported index crimes in Lansing. Fourth, operational level of GIS by police officers Is also not considered' In JI‘Au-MIO» u. avg-u "1"“... L/ *m -o—u I‘m. flwnnfiwanwfl this thesis research. If there Is no trarned officers for GIS ILcould be a useless MFVM _— . h'lw“ *""--W"+o~q J machine. Depending on operator’ s capability of using GIS, the outcome will be NW”! “'WWMms 27 varied. In order to get the full performance of using GIS In policing, there should ,4 1H)“ \‘ be adequately trained officers who could utilize its full function Thus GIS should W not be consIdered as the only factor that influences arrest and crime rate. Section C- Implications The findings presented in this research have implications to the over all crime control and prevention strategies developed by components of law enforcement agencies. First, GIS technology should be implemented by police departments in the United States and worldwide. As technology advances, crimes are getting more complicate and criminals are getting smarter than ever. To protect and serve for the community, police departments need to be equipped with the higher technology that satisfy the safety of community. These findings show that there are no relationship between the use of GIS for targeting crime hot spots, and total number of arrests and total number of reported index crimes in Lansing. The null hypothesis was accepted since the significance level was more than 0.05. Therefore, the effectiveness of using GIS in proactive policing statistically could not be established. However, these findings are not conclusive since there is little, if any, data and research available to support these statistical results. Second, GIS should be given more research attention. Researchers have found that there are few data available for GIS regarding policing. Criminal justice. researchers have not studied GIS compare to other topics. However, GIS technology is no longer considered as a new technology. Recently, GIS is getting more attentions by many police departments. In order to promote the use of GIS 28 to police departments, there should be more studies mandated. Without academic support, police may experience difficulty in implementing GIS in their departments. Not only for police departments but also for researchers and practitioners in criminal justice field, GIS should be studied more, and without delaying. Time and criminals are not waiting for police departments’ technological evolution. 29 APPENDICES 3O APPENDIX A TOTAL NUMBER OF GIS IN POLICE DEPARTMENTS Percentage of Law Enforcement Using GIS 0 US Law Enforcement I Approx. Using GIS I Approx. Using GIS Successfnll " 31 APPENDIX B 1996 UCR REPORTING AGENCIES IN MICHIGAN Wimhfinhfi“ Ibwyo "immuwdkflflund mmfimdflfidwddmd * hm—‘Ihnflbufi. TUMO- “nWIOM‘~Whfi‘ Hfldhupfiiuhhwflt wanna-inhuman: 32 APPENDIX C GIS IN LANSING POLICE DEPARTMENT Lansing, Michigan Police Department Offenses 8&E Non B&E Res Larc fm Veh E I SIOIen Vd'l. ' é MultiCrime I - IE /\/ Streets ' r _ . - E] Police Teams ,. I _ , [:] Lansing Boundary b .4... 0.1 0 0.10.2 0.3 0.4 mm 5253—25: 33 BIBLIOGRAPHY Bibliography Bayley, DH. (1994). Police for the future. New York: Oxford University Press. Binkley, LL. (1991, April). Futuristic system helps Long Beach comer career criminals. Police Chief 142-145. Block, C.R., & Green, LA. (1994). The Geoarchive handbook: A guide for developing a geographic database as an information foundation for community policing. Illinois Criminal Justice Information Authority. Chung, T.J. (1998). Technological proactive policing. The 12‘" lntemational Congress on Criminolpgy. Seoul, Korea : The Korean Institute of Criminology. Duyn, V. (1991). Automated crime information systems. Philadelphia: TAB Professional and Reference books. Eck, J.E., & Spelman, W. (1987, January). Problem-oriented policing in Newport News. Research in Brief. Washington, DC: National Institute of Justice. Eck, J.E., and Weisburd, D. (1998). Improving the management of rental properties with drug problems: A randomized experiment. In civil remedies, ed.L. Mazerolle and J. Roehl. Monsey, New York: Criminal Justice Press. Gay, W. (1977). Improving patrol productivityI Volume I, Routine patrol. Washington, DC: Government Printing Office. Lauritsen, J.L., and Quinet, K.F.D. (1995). Repeat victimization among adolescents and young adults. Journal of Quantitative Criminolpgy. 11. 143-166. Lurigio, A.J., Dantzker, M.L., Seng, M.J., Sinacore, JM. (1997). Criminal iustice statistics: A practical approach. Boston: Butterworth-Heinemann. 35 Main, D. (1999, January 15). Crime drops, patrols increase. The State News pp.9. Maltz, M., Gordon, AC, and Friedman, W. (1991). Mapping crime in its community setting. New York, New York: Springer-Veriag. Norusis, Marija, J. (1997). SPSS 7.5 : Guide to data analysis. Simon & Schuster Company. Parker, HD. (1988). The unique qualities of a Geographic lnfonnation System: A commentary. Photogrammetric Engineering and Remote Sensing, 54, 1547-1549. Pilant, L.(1993, January). Computerizing criminal investigation. Police Chief 29- 41. Polvi, N., Looman, T., Humphries, C., and Pease, K. (1991). The time-course of repeat burglary victimization. British Journal of Criminolpgy, 31, 411-414. Pilant, L. (1997, December). Computerized crime mapping. Police Chief 60-69. Radelet, L. A. & Cater, D.L. (1994). The police and the community. New York: Macmillian College Publishing Company. Rogers, R., & Craig, DE. (1996). Geographic Information Systems: Computers in law enforcement. Journal of Crime and Justice 16. 61-73. Rosenbaum, DP, and Lavrakas, P.J. (1995). Self-reports about place: The application of survey and Interview methods to the study of small areas. J.E. Eck and D. Weiburd (Eds), In crime and place (pp.285-314). Monsey, New York: Criminal Justice Press. Sherman, L.W. (1992). Policing domestic violence. New York, New York: Free Press. Spparow, MK. (1991, April). Information systems: A help or hindrance in the evolution of policing? Police Chief 26-29. Taylor, RB. (1997). Social order and disorder of streetblocks and neighborhoods: Ecology, microecology and the systemic model of social disorganization. Journal of Research in Crime and Delinguency, 33, 113- 155. Weisburd, D. (1997). Reorienting crime prevention research and mlicy: From the causes of crimina_litv to th_§ context of crime, Research Report. Washington, DC: US. Department of Jusfice, National Institute of Justice, NCJ 165041 WItkin, G. (1996, December). Making means streets nice. Computer maps that take the ‘random’ out of violence. US. News 8. World Report, 63. 37 "IIIIIIIIIIIIIIII'LIIIIIII“