tiểu luận kinh tế lượng factors affecting criminal rate in countries between 2007 and 2016

FOREIGN TRADE UNIVERSITY FACULTY OF INTERNATIONAL ECONOMICS -🙞🙜🕮🙞🙜 - ECONOMETRICS REPORT Factors affecting criminal rate in countries between 2007 and 2016 Class: KTEE218.1 GROUP Name Students’ ID Class Tao Thị Ha My 1814450055 Anh 02 CLC KTQT Nguyen Thi Thuy Duong 1814450023 Anh 01 CLC KTQT Vu Doan Hai Anh 1814450015 Anh 01 CLC KTQT Mai Thanh Phuong 1814450063 Anh 02 CLC KTQT Pham Thanh Hai 1814450034 Anh 02 CLC KTQT Hà Nội, tháng năm 2019 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com MỤC LỤC ABSTRACT INTRODUCTION I) SECTION 1: Overview I.1) Theoretical framework I.1.1) Definition of crime and crime rate I.1.2) Overview of crime cases over the world during 2007-2016 I.1.3) Factors affecting crime rate I.2) Literature review I.2.1) Previous methods of crime statistics: I.2.2) Related published reseaches: I.2.3) Recent Vietnam studies about crime: I.2.4) The aim of our team research: 10 I.2.5) Differences in our research: 10 II) SECTION 2: MODEL SPECIFICATION 11 II.1) Methodology 11 II.1.1) Methods 11 II.1.2) Theoritical model specification: 11 II.2) Descriptive Statistics and Data Description 12 II.2.1) Descriptive Statistics 12 II.2.2) Data Description: 13 II.3) Correlation between variables 14 III) SECTION 3: ESTIMATED MODEL AND STATISTICAL INFERENCE 15 III.1) Estimation model 15 III.1.1) Estimate result 15 III.1.2) Sample regression model 16 III.1.3) Result analysis 16 III.1.4) Meanings of estimated coefficients 16 III.2) Hypothesis testing 17 III.3) Resolution and Recomendation 19 III.4) Limits of report 20 IV) CONCLUSION 20 V) APPENDIX 22 VI) REFERENCES 27 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com ABSTRACT The report aims to illustrate the influence of population, GDP growth rate, unemployment rate and HDI on criminal rate, and to give a solution to reduce criminal rate base on the result of the model INTRODUCTION Crimes are always one of the most persistent problem with whom each nation must face In consequences, many methods have been applied to reduce criminal rate like increasing the number of policeman, or developing people’s education and awareness, e.t.c However, our group is always curious about the factors affecting criminal rate, about the impact of the countries’ macroeconomics index such as GDP or Unemployment rate on crimes We are always wondering what drives people to such extreme As a result, our group choose the topic “Factors affecting criminal rate in countries from 2007 to 2016” to satisfy our curiosity After analyzing the data of 10 countries over the years, specifically from 2007 to 2016, by the linear regression model using OLS method, our group has finally found the correlation between population, GDP growth rate, unemployment rate, HDI and criminal rate Our report consists of section: I.Section 1: Overview of the topic II.Section 2: Model specification III.Section 3: Estimation model and Statistical references IV.Section 4: Recommendation and Resolution LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com I) Section 1: Overview I.1) Theoretical framework I.1.1) Definition of crime and crime rate Whether a given act or omission constitutes a crime does not depend on the nature of that act or omission It depends on the nature of the legal consequences that may follow it An act or omission is a crime if it is capable of being followed by what are called criminal proceedings Different countries define crime differently In England, the definition of "crime" was provided by the Prevention of Crimes Act 1871, and applied for the purposes of section 10 of the Prevention of Crime Act 1908: “The expression "crime" means, in England and Ireland, any felony or the offence of uttering false or counterfeit coin, or of possessing counterfeit gold or silver coin, or the offence of obtaining goods or money by false pretences, or the offence of conspiracy to defraud, or any misdemeanour under the fifty-eighth section of the Larceny Act, 1861.” In ordinary language, a crime is an unlawful act punishable by a state or other authority The term "crime" does not, in modern criminal law, have any simple and universally accepted definition, though statutory definitions have been provided for certain purposes The most popular view is that crime is a category created by law; in other words, something is a crime if declared as such by the relevant and applicable law One proposed definition is that a crime or offence (or criminal offence) is an act harmful not only to some individual but also to a community, society, or the state ("a public wrong") Such acts are forbidden and punishable by law Usually, to be classified as a crime, the "act of doing something criminal" must – with certain exceptions – be accompanied by the "intention to something criminal" While every crime violates the law, not every violation of the law counts as a crime Breaches of private law (torts and breaches of contract) are not automatically punished by the state, but can be enforced through civil procedure Crime rate is a count of crimes complied to assess the effectiveness of a crime control policy, and the impact of the policy on the risk of crime victimization For example, burglaries/total population is the standard "crime rate" reported by the FBI and used by social scientists Crime rates differ widely across different times and different places, for a variety of reasons that are studied by social scientists Violent crime rates may define perceptions regarding community safety I.1.2) Overview of crime cases over the world during 2007-2016 Crime definitions differ between the countries because of different penal codes, and dissimilar reporting behaviour and recording practices; consequently the differences of crime levels in different countries may be based on different definitions, reporting behavior and recording practices rather than differences in actual crime Therefore trend analysis is a more fruitful approach as it shows how crime has developed Crime cases over the world were LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com insufficiently recorded However, the cases can be indicated by some representative countries and regions In America, crime cases had a downward trend in most of the time surveyed, from 2007 to 2014 However, in two years end of the period, the cases increased sharply Figure 1: Crime offense figure in America from 2007 to 2016 - Source: FBI’s Uniform Crime Reporting (UCR) Program     In 2011, an estimated 1,203,564 violent crimes occurred nationwide, a decrease of 3.8 percent from the 2010 estimate When considering 5- and 10-year trends, the 2011 estimated violent crime total was 15.4 percent below the 2007 level and 15.5 percent below the 2002 level There were an estimated 386.3 violent crimes per 100,000 inhabitants in 2011 Aggravated assaults accounted for the highest number of violent crimes reported to law enforcement at 62.4 percent Robbery comprised 29.4 percent of violent crimes, forcible rape accounted for 6.9 percent, and murder accounted for 1.2 percent of estimated violent crimes in 2011 In the European Union, The crime statistics presented in this article cover offences recorded by police in EU Member States and some other European countries These data not purport to describe all crime in Europe: some crime goes unreported and changes in rates of particular offences may be affected by new policing strategies or methodological changes National data have been aggregated to provide estimates for the EU as a whole for the purpose of identifying overall trends Any inferences for the EU or Member States should be based on trends over time There was a general tendency for levels of recorded crime to decrease in recent years: the number of most types of crimes recorded by the police in the EU-28 fell between 2007 and 2012 as can be seen from Figure LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com Figure 2: Crimes recorded by the police, 2002–2012 - Source: Eurostat I.1.3) Factors affecting crime rate There are certain social and economic characteristics that affect the formation of crime rates, either positively or negatively researchers have identified a number of factors that influence crime rates and typically are present in jurisdictions where crime rates are high  Poverty and socio-economic conditions Statistically, poverty goes hand-in-hand with crime Where poverty is prevalent in a community, that community will experience higher levels of crime Generally, it's not the poverty itself that leads to higher crime rates but the factors associated with poverty, such as chronic joblessness, less access to quality schools, employment, role models and the real or perceived lack of opportunity  Social Level of Morality People's upbringing and social environment can shape their view of the world and directly affect their decisions in the future For example, research shows that people who have been physically, sexually or emotionally abused as children are three times more likely than LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com non-abused adults to commit acts of violence In communities where crime is tolerated, a person may commit a crime simply to fit in with his peers  Police Policy A well-resourced police force coupled with tough sentences for perpetrators may help to reduce the crime rate For example, the U.S crime rate was much higher in the 1960s and 1970s, before lawmakers responded by enacting tough-on-crime policies and building prisons Mass incarceration removes people from the streets who would otherwise be committing crimes Visible policing may also have a deterrent effect When New York Mayor Rudy Giuliani implemented a policy of "broken-windows policing" – cracking down on minor crimes to make neighborhoods feel safer – crime declined at a precipitous rate Some argue against credit given to Giuliani's policies, suggesting that a drop in unemployment rates may have been the reason for the crime reduction  Age of the Population There's a correlation between the crime rate and the age of the population Specifically, most crimes are committed by people in their teens, 20s and 30s, especially in areas where the population is both young and transient Apart from these factors, there are other elements affecting crime rate, which will be clarified by this report I.2) Literature review Crime rate differ from different time and different place, for a variety of reasons that are studied by social scientists There have been many methods established aiming at measuring crime rate in a country However, each one has their benefits and also drawbacks Public surveys are occasionally conducted to the amount of crime which are not reported to the police, which is much more reliable than assessing trends But they also have their limitations while reporting such as ignorance of offences against children, not useful enough information for the local crime prevention and also uncounting number of offenders before the criminal justice system I.2.1) Previous methods of crime statistics: The major methods for collecting crime data is law enforcement reports, which only reflect crimes which are collected, reported, recorded and not subsequently cancelled as well as victim study (victimization statistical surveys, relying on individual’s memory and honesty) Laws and practices vary between jurisdictions, therefore comparing crime statistics can be difficult Some countries have established a new method to measure almost absolute result in crime rate statistics, especially the USA In the USA, there are two major data collection programs First, the Uniform Crime Reports, which is considered most reliable, was established in the early 1930s It is administered by Federal Bureau of Investigation, complies data of the crimes reported to local police Because the data is based on local information, the URC permits statistical analysis for local areas LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com The accuracy of the UCR varies whether people are likely to report crimes to the report However, depending on types of crime, it has different accuracy rate Long – term analyses of UCR data is still questionable because computer storage is better than police keeping in the past 20 years Still, the UCR is the standard and most widely used in measuring crime in the USA Secondly, US Census Bureau administered the National Crime Victimization Survey, which begun in 1973 This measured is a representative telephone sampling of 40,000 households to determine how many people was the victim by one of the seven crime in last year such as rape, robbery, assault, personal theft, household theft, burglary and motor vehicles theft, but not murder rate because the victims cannot be surveyed The NCVS was a national survey so it cannot break down by state or locality Most criminologists consider the NCVS more accurate than the UCR because people can acknowledge that they have been victim of a crime to an anonymous survey, even they did not report the crimes to the police In International Encyclopedia of the Social & Behavioral Sciences of Martin A Andersen, it is stated that to measure crime rate, we have to considered two necessary pieces of data: crime data and population at risked data (which is denoted in the equation ∗ 𝑆𝑐𝑎𝑙𝑎𝑟) This count is most often based on Uniform Crimes Report mentioned above However, it also meets some limitations:    Crime reporting can be varied from different police detachment Not all criminal events are reported to police Lack of population at risk of crimes data In spatial criminology, the most common source of population at risk data is the census – resident population, but it also varies in different regions I.2.2) Related published reseaches: Some contributions have theoretically tried to establish a relationship between crime, growth and development (e.g., Bourguignon, 2000, 2001; Fajnzylber et al., 2002a, Mauro and Carmeci 2007) and some studies quantify economic and social cost of crime for different countries [Australia (Mayhew 2003), France (Palle and Godefroy 2000), the United Kingdom (Brand and Price 2000), New Zealand (Roper and Thompson 2006), the United States (Miller et al 1996), Italy (Detotto and Pulina 2012), for some Latin America States (United Nations 2007) and Colombia (Cotte Poveda 2012)] The econometric results show that crime leads to a negative effect on real per capita output and employment For example, Peri (2004) using panel data of Italian provinces for 1951-1999 observes crime has a statistically significant negative effect on economic output and employment, indicating the possibility of nonlinearities in the crime-growth relationship In particular, while she finds a statistically significant adverse violent-crime effect on growth, the impact of property crime is weak and in some specifications perverse A World Bank study (World Bank, 2006), using a sample of 43 countries for the period of 1975-2000, shows a strong negative relationship between crime and growth even after controlling for human-capital accumulation and income inequality Similarly, Càrdenas (2007) finds a significantly negative association between LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com crime and per-capita output growth in a panel of 65 countries using homicides data for 19711999 Time-series studies (e.g., Dettoto and Pulina 2009, Dettoto and Otranto 2010) using single country data also find a negative association between crime and income levels However, Chatterjee and Ray (2009) based on a large cross-country sample for the period 1991-2005 and controlling for human capital and institutional quality, find no strong evidence of a uniformly negative association between crime and growth Burnham et al (2004) using US county level data are not able to establish a clear connection between central city crime and per capita income growth In 2014, Colin Webster and Sarah Kingston from Leads Metropolitans University, have gathered and reviewed 173 of the most cited articles and important monographs published mostly between 1980 and 2013 that either directly or indirectly tested the poverty and crime link in the United State, United Kingdom and Europe It is concluded that there is no direct relationship between unemployment rate and crime For example, in the findings of the effects of family disruption resulted from jobless has an important impact on crime (Sampson, 1987) In one of a series of important founding studies of the relationship between unemployment rate and crime rat, Cantor and Land (1985) using time series data for the United States, from 1946 to 1982, found a small but significant influence of the unemployment rate on property crime A similarly influential study by Chiricos (1987) reviewing data from when unemployment rose dramatically in the 1970s, concluded that indeed total unemployment and the rate of increase of unemployment have strong influences increasing property crimes In 2016, a group of student in University of Kebangsaan Malaysia conducted a research about the relationship between economic growth and crime rate in the period from 1980 to 2013 while they examine in the long – run and short – run relationship The conclusion was that in the long – run as well as short – run relationship, the impact of economic growth toward criminal rate was considered to be positive and statistically significant, against the economist arguments when good economies tends to create more crime and the opposite occurs during bad economies In 2017, an MPRA (Munich Personal RePEC Archive) research, the authors (students of Goa Institute Management in India), include 13 factors: poverty level, literacy rate, net enrollment ratio, per capital ratio (upper primary level), per capital schools, rural population, sex ratio, household availing bank services, population density, per capita GSDP (constant price), capita income (constant price), no of jails and minority population) The conclusion showed the analysis reflects a fairly strong relationship between population density, sex ratio and literacy rate and per capita crime under IPC in India and these factors play an utmost important role in determining the crime rate Factors like Per capita gross state domestic product and per capita schools form a relationship with per capita crime but don't effect it by a large margin Rural population share and minority population play a very negligible role in determining the criminal cases registered in India I.2.3) Recent Vietnam studies about crime: In Vietnam, there is a report about criminal rate of the country from 1986 to 2008, dividing in two periods and differs from different types of crimes based on the data collected by Supreme People’s Court of Vietnam LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com To clarify the trend of crime situation, they use the three – year period to examine and the results showed that the level of crime rate in Vietnam from the period of 1986 – 2008 tend to increase in different levels and two dramatically increasing periods:     First level, measured from 1986 to 1994, the crime rate increased from 1,93% to 15,86% Second level, including two periods: 1995 – 1997 and 1998 – 2000, the maximum increasing level was 25,71% Third level, from 2001 to 2008, the crime rate increased from 8,54% to 18,52% Two extreme changes in crime rate were in 1995- 1997, extreme increase (64.85%) and 2001 – 2003, extreme decrease (6,34%) Nowadays, crime situation is much more complicated than ever before, there are more people commit to crime, more criminal behaviors, more variable types of crimes, and also the intention to commit crime also differ from benefit or accidentally caused Although there have been several reports about criminal rate in Vietnam, the government still not acknowledge about the reasons as well as the consequences and impacts of the density of population, GDP growth of the country, the unemployment rate and also HDI of the citizens on the growth of criminal rate Therefore, our team choose the topic of “Factors affecting criminal rate in countries from 2007 to 2016” Though the method of using econometrics model, our team will give remarks and forecast about components affecting on crime rate in 10 countries during period 2007 – 2016 I.2.4) The aim of our team research: The general objectives of our topic is to identify and analyze the impacts of the four factors on the criminal rate of ten countries across the world  Systematize theoretical framework and empirical research about the elements affecting crime rate  Estimate regression model and analyze impact of variables on crime rate  Recommend about decreasing crime rate due to changing policy in the four factors  I.2.5) Differences in our research: Firstly, our research collect data of GDP growth rate, population, HDI and also unemployment rate of ten representative countries all over the world to identify the overall view of impacts of these four factors affecting crime rate in the period of 2007 – 2016 Secondly, we use these factors because it is commonly more socialized than just number statistical and reports of cases Therefore, it is easier to give recommendations of each fields to the government and local authority 10 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com Standard Error 4.51*10-15 0.1133747 0.0519548 0.2147701 Median 7.13*1011 0.029512 0.86791 0.86791 Standard Deviation 9.07*1011 0.038183 0829002 0198024 Obs 100 100 100 100 II.2.2) Data Description: (1) POPULATION (Population of countries from 2007 to 2016) (a) Mean: 7.13*1011 (b) Standard Deviation: 9.07*1011 (c) Minimum Value: 1.23*1010 (d) Maximum Value: 3.23*1012 (2) GGR (GDP Growth Rate of countries from 2007 to 2016) (a) Mean: 0.029512 (b) Standard Deviation: 0.0606718 (c) Minimum Value: -0.056 (d) Maximum Value: 0.256 (3) HDI (Human Development Index of countries from 2007 to 2016) (a) Mean: 0.86791 (b) Standard Deviation: 0.0829002 (c) Minimum Value: 0.632 (d) Maximum Value: 0.934 (4) UNEMPLOYMENT (Unemployment rate of countries from 2007 to 2016) (a) Mean: 0.056023 (b) Standard Deviation: 0.0198024 (c) Minimum Value: 0.0196 (d) Maximum Value: 0.096 13 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com II.3) Correlation between variables  After running corr criminalrate population gdpgrowthrate hdi uneploymentrate, we have correlative table as below: CRIMINAL RATE POPULATION GGR HDI UNEMPLOYMENT CRIMINAL RATE 1.0000 -0.3823 -0.0847 -0.0526 0.6310 POPULATION -0.3823 1.0000 -0.1896 0.0950 -0.0045 GGR -0.0847 -0.1896 1.0000 -0.2910 -0.2454 HDI -0.0526 0.0950 -0.2910 1.0000 0.2799 UNEMPLOYMENT 0.6310 -0.0045 -0.2454 0.2799 1.0000 Correlation coefficient between UNEMPLOYMENT and CRIMINAL RATE: 0,6310 Correlation coefficient between GGR and CRIMINAL RATE: -0,0847 Correlation coefficient between HDI and CRIMINAL RATE: -0.0526 Correlation coefficient between POPULATION and CRIMINAL RATE: -0,3823 Conclusion: GGR, HDI, POPULATION have an negative effect on CRIMINAL RATE, whereas UNEMPLOYMENT has a positive effect on CRIMINAL RATE UNEMPLOYMENT and POPULATION have the highest correlation coefficients Therefore, these two variables affect CRIMINAL RATE strongly II.4) Expectation in the parameters of the model 14 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com In fact, when POPULATION increase, ceteris paribus, CRIMINAL RATE will go down So 𝛽1 has the (-) sign As for GGR and HDI, if the country is more developed and the people have a healthy life, the CRIMINAL RATE will also decrease Therefore 𝛽2 and 𝛽3 both have the (-) sign Finally, increase UNEMPLOYMENT RATE will raise CRIMINAL RATE Thus 𝛽4 have the (+) sign III) SECTION 3: ESTIMATED MODEL AND STATISTICAL INFERENCE III.1) Estimation model III.1.1) Estimate result - Based on the chosen variable, we has sample regression model: CRIMINAL RATE = 𝛽0 + 𝛽1 POPULATION+ 𝛽2 GGR +𝛽3 HDI + 𝛽4 UNEMPLOYMENT In which: - - 𝛽0 : the estimator of β0 𝛽1 : the estimator of β1 𝛽2 : the estimator of β2 𝛽3 : the estimator of β3 𝛽4 : the estimator of β4 Based on the result after running the multiple regression on Stata, we have the following table: Coefficient 𝛽0 𝛽1 𝛽2 Value P>|𝑡| 95% Confident Interval 0.1085017 0.018 0.0187459 0.1982575 -2.46*10-14 0.000 -3.36*10-14 -1.57*10-14 -0.0838357 0.461 -0.3089129 0.1412416 15 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com 𝛽3 𝛽4 -0.1631367 0.002 -0.26628 -0.0599934 2.079775 0.000 1.653402 2.506148 III.1.2) Sample regression model - Based on the chosen variable and from the result above, we have sample regression model: CRIMINAL RATE = 0.1085017 - 2.50*1010 POPULATION - 0.0838357GGR 0.1631367 HDI + 2.079775 UNEMPLOYMENT + 𝑈 III.1.3) Result analysis        Number of Observations = 100 Total Sum of Squares = 0.364425863 Explained Sum of Squares = 0.213271187 Residual Sum of Squares = 0.151154676 Model degrees of freedom Dfm= Residual degrees of freedom Dfr = 95 Coefficient of determination R2 =58.52% shows that the suitability of the regression model is at a very medium high degree The value R2 also let us know that the proportion of variance in the dependent variable which can be explained by the independent variables is 58.52%  Adjusted R-squares= 56.78% III.1.4) Meanings of estimated coefficients  𝛽0 = 0.1085017:  This mean that the expected value of CRIMINAL RATE is 0.1085017 when all dependent variables equal to zero  𝛽1 = - 2.46*10-14 :  When the population increases by 1%, other factors are constant, the criminal rate will decrease by 2.46*10-14 % This support the conclusion that population and criminal rate have negative relationship  𝛽2 = -0.0838357 :  When the GDP growth rate increase by 1%, other factors are constant, the criminal rate will decrease by 0.0838357% This supports the conclusion that GDP growth rate and criminal rate have negative relationship 16 LUAN VAN CHAT LUONG download : add luanvanchat@agmail.com  𝛽3 = -0.1631367 :  When the HDI is increase by 1% other factors are constant, the criminal rate will decrease by 0.1631367% This also supports the conclusion that GDP growth rate and criminal rate have negative relationship  𝛽4 = 2.079775 : When the unemployment rate increase by 1%, the criminal rate will also increase by 2.079775% This supports the conclusion that the unemployment rate and criminal rate have positive relationship III.2) Hypothesis testing In general, all factors are agree to our expectation and the correlation coefficients we mentioned above However, we still have to check variables’ significance There are three ways to decide whether to reject or not reject the null hypothesis (in which the Ho hypothesis is a hypothesis stated that coefficient is not statistically significant) but below we just show ways which is easier to see:  Method 1: The p-value method : If the p-value of an independent variable is less than the significance level (there are levels of significance: 1%, 5% and 10%), then rejecting Ho, accepting H1, which means independent variable is statistically significant for the CRIMINAL RATE  The coefficient β0 has p-value = 0.018