This action might not be possible to undo. Are you sure you want to continue?

A study presented to Ms. Angela D. Nalica Professor, Stat 136

In partial fulfilment of the requirements for STAT 136: Introduction to Regression Analysis University of the Philippines, Diliman, Quezon City

CRUZ, Clemence-Fatima MACARAIG, Miguel Rodrigo SANTOS, Marvin Allan

May 28, 2010

A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces

2

Abstract

This study focuses on the provincial crime rate in the Philippines for the year 2000. The aim is to ascertain the possible factors that affect the crime rate in the Philippines using multiple linear regression. The results present that, at a 0.05 level of significance, the following variables contribute to crime rate: population density, poverty incidence, number of policemen, and number of courts.

A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces

3

Introduction

Crime is a truth that exists for all, whether it is taken as a moral or legal construct. This is a truth that we would have to accept, no matter how appalling it may seem. Many efforts have been exerted in order to eradicate crime. Unfortunately, a world or a country without crime is strictly utopian. As such, an existence without crime is impossible to achieve. For us to be able to eradicate the concept of crime, we must first remove the concepts it violates which are morals and laws, which is, again, a utopian task. This would be a discourse meant for philosophical minds, and would therefore beyond our concerns. Thus, the best course of action would be to deter or lessen the crime incidence or lower the crime rate that prevails in our country.

Here in the Philippines, crime is one of the foremost problems present. However, in the midst of more urgent problems such as poverty, corruption, and hunger, crime loses most of its significance and is then relegated to the bottom of the Philippines‘s long list of problems. The solution to crime deterrence becomes limited to debates on revising punishments for crimes and reinstating the death penalty—a punishment which has no proven effect of deterring crime rate. What country officials fail to recognize is that bandaid solutions such as imposing severe punishments do not work on large-scale problems such as this. If so, what do we have to do in order to address this problem?

To any problem, the solution is to ascertain its true cause and attack it from its roots. This may sound easy and simple enough; however in our country where problems are more tangled than politics, this would be a complicated task. In most cases, people do not seem to know for certain which problem causes which. For example: is the Philippines poor because there is a high incidence of corruption? Or is the Philippines corrupt because there is poverty? The same goes for crime rate. Is there a high rate of crime because the Philippines is poor? Or is it that the Philippines is poor because there is a high crime rate? Here lies the dilemma.

This study will focus on crime rate in the Philippines using provincial data from the National Statistical Coordination Board (NSCB). and by how much do these factors affect crime rate? It is in hopes of answering these questions that this study was conducted. may be formulated. . by means of addressing or alleviating these factors.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 4 If the possible factors that affect crime rate are correctly or sufficiently identified. a more feasible solution. All data retrieved were from the year 2000. And thus. and to provide an estimate regarding the impact that these factors present through the use of multiple linear regression. This study aims to ascertain the possible factors that affect the crime rate here in the Philippines. we ask this question: what are the possible factors that affect crime rate here in the Philippines.

tips. • Other forms of compensation and net receipts derived from the operation of family-operated enterprises/activities and the practice of profession or trade. Cohort Survival Rate – the percentage of enrollees at the beginning grade or year in a given school year who reached the final grade or year of the elementary of secondary level. Consumer Price Index (CPI) – Indicator of the change in the average prices of a fixed basket of goods and services commonly purchased by households relative to a base year. bonuses.total number of pupils/students who register/enlist in a school year. • Receipts from family sustenance activities. purchase . • Commissions. family and clothing allowance. investment ventures. Family Income – includes primary income and receipts from other sources received by all family members during the calendar year as participants in any economic activity or as recipients of transfers. transportation and representation allowance and honoraria. They exclude all expenses in relation to farm or business operations. NSO) Primary income includes: • Salaries and wages from employment. etc. Enrolment . (2000 FIES. • Rentals including land owner‘s share of agricultural products • Pensions • Support and value of food and non-food items received as gifts by the family (as well as the imputed value of services rendered free of charge to the family).A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 5 Definition of Termsϯ Crime Rate** . • Interests. pensions. grants. Family Expenditures – refers to the expenses or disbursements made by the family purely for personal consumption during the reference period. Income from other sources include: • Imputed rental values of owner-occupied dwelling units. which are not considered as family operated enterprise.000 population.is the number of reported crimes per 100.

This measure is more meaningful if given as population per unit of arable land. family sustenance and entrepreneurial activities are also considered as family expenditures. Gross Regional Domestic Product . Gifts.represents a significantly higher level literacy which includes not only reading and writing skills but also numeracy skills. This skill must be sufficiently advanced to enable the individual to participate fully and effectively in activities commonly occurring in his life situation that require a reasonable capability of communicating by written language. Value consumed from net share of crops. Human Development Index .a measure of how well a country has performed. Notes: It is a measure of the extent to which the distribution of income/ expenditure among families/individuals deviates from a perfectly equal distribution. Simple/Basic – the percentage of the population 10 years old and over. etc. assistance or relief in goods and services received by the family from friends.aggregate of the gross value added or income from each industry or economic activity of the regional economy. and to have access to the resources needed to afford a decent standard of living.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 6 of real property and other disbursements which do not involve personal consumption. support. and consumed during the reference period are included in the family expenditures. but also in terms of social indicators of people‘s ability to lead a long and healthy life. with limits 0 for perfect equality and 1 for perfect inequality. not only in terms of real income growth. to acquire knowledge and skills. Functional Literacy . Literacy rate. Population Density – refers to the number of persons per unit of land area (usually in square kilometers). write and understand simple messages in any language or dialect. relatives. who can read. . fruits and vegetables produced or livestock raised by other households. Gini Ratio – the ratio of the area between the Lorenz curve and the diagonal (the line of perfect equality) to the area below the diagonal.

health.000 is a magnifier. It consists. (2009). Social Services . Usually. in some cases. .000. (2009 edition). 100.000 are used as magnifiers. Unemployment Rate – proportion in percent of the total number of unemployed persons to the total number of persons in the labor force. __________________ ϯ National Statistical Coordination Board. in varying numbers. of component cities.this covers expenditures for education. Its functions and duties in relation to its component cities and municipalities are generally coordinative and supervisory. labor and employment. social security. Philippines: Author. Province – the largest unit in the political structure of the Philippines.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 7 Poverty Incidence – the proportion of families/individuals with per capita income / expenditure less than the per capita poverty threshold to the total number of families/individuals. of municipalities and. housing and community development and other social activities. 1. Makati City. and as such any power of ten may be used.000 and 100. Philippine statistical yearbook. **Note that crime rate = (total crime incidence/population)*100.

According to Wadsworth (2001). These two. Rather. crime is not entirely psychological or social. At the forefront of this economic view on crime is Gary Becker‘s work Crime and Punishment: An Economic Approach published in 1968. Becker views crime as an economic construct—one that presents opportunity and economic costs and has a supply of offenses. Here.. in turn help determine labour distribution and personal income. it is the ―inequalities in the distribution of schooling‖ that is ―strongly related to the incidence of many crimes. according to Becker... employment is an important factor that affects crime rate.‖ Ehrlich echoes Becker‘s statement that the behaviour of crime is not merely psychological or social. There is a contextual influence of weak labor market opportunity that operates above and beyond influencing individual employment experiences. 1975) attempts to establish a link between education and crime.‖ As such.not because their basic motivation differs. crime had been viewed as a moral and social construct. however. These effects cannot be explained entirely by the fact that individuals who are unemployed commit more crimes.. again from an economic perspective. Ehrlich suggests training geared towards legitimate activities before convicts are released from prison. Specifically. this is due to the ―relative earnings‖ of offenders between legitimate and illegitimate activities. He further asserts that ―some persons become ‗criminals‘.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 8 Review of Related Literature Before. He asserts. but economic in which choices and utility are of importance. ―both industrial composition and labor force participation…have direct and indirect effects on violent and property crime rates. Ehrlich states that education may be viewed as an opportunity-maker. but also economic. Over the years. Ehrlich postulates.but because their benefits and costs differ.‖ Thus. it is not merely the . there has been a shift from a social point of view to an economic one. Education. As a form of rehabilitation. He further states that it is not educational attainment that is closely related to crimes. Ehrlich (On the Relation Between Education and Crime. is important for on-the-job training. In his work.

‘ Curtly said. It is in this work that they incorporated social. et al. 1998). As for Reynolds (2000). In general.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 9 fact that they are unemployed that turns them into criminals. economic factors such as mentioned above affect crime rate. 2010). This is especially true for the ―policy-sensitive variables. an implementation of more ‗get-tough policies‘ would be a helpful deterrent to crime rate since ‗federal programs to reduce the so-called root causes have done…more harm than good.‖ A possible solution to this is the ―combination of counter-cyclical redistributive policies…and increases in the resources of apprehending and convicting criminals…especially during economic recessions‖ (Fajnzylber. alcohol consumption. Reynolds believes that mere programs to alleviate the ‗root causes‘ of crime would not deter it. (2009) also state in their study of crime rate in Pakistan that poverty. Yasir. et al. Gillado and Cruz (2004) constructed a regression model for three different classifications of crime—against property. unemployment rate. Other factors may be the urbanization of a place (since urbanization opens new avenues for crimes) and police visibility (SanidadLeones. population density. against person.. demographic. Serious and strict policy-making is the key to deterring crime rate. and economic factors. and Wadsworth seem to share. corruption index. cohort survival rates in elementary and secondary education. It is mostly due to the individual experiences a person has of employment or lack thereof. considering only one . it can be observed that there is a shift from a social perspective to an economic one. However. and volatile policies may contribute to the rise of crime rate. Although crime may still be a social construct. and rape. average income of people in rural and urban areas. The following variables were considered for the three models: per capita regional domestic products. Gini coefficient. inflation. Becker. unemployment. police population. This is a perspective that Ehrlich. They further assert that a possible way to alleviate this would be the formulation of stable economic policies. and consumer price index.

In order to account for the possible effect of education on crime. human development index (HDI). . CPI. family income and expenditure. however. functional literacy. both economic and social. number of courts. police force population. are not available for provinces. The variables considered for this study are heavily based on the works abovementioned. For this particular study. Although there may be opposing views regarding the factors of crime. the variables that were taken into consideration are poverty incidence. Data on cohort survival rates. and population density. and geographical setting (based on archipelagic division). the researchers have included literacy. and enrolment as variables. but were included in this study are expenditures on social services.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 10 perspective would be insufficient. play their roles in affecting crime. it cannot be denied that these factors. unemployment rate. Other variables not included in the abovementioned works. since both perspectives are applicable to crime. These variables coincide with the variables mentioned in the study of Gillado and Cruz.

the F-test was used. For example. which is available both in print and in electronic format.000 population. cohort survival rates for elementary and secondary education. However. functional literacy rate. especially for the dependent variable.000 population.05 was set prior to any fitting or testing procedure. Each of these independent variables‘ significance was assessed through the t-test. whereas the rest of the observations would range from about 400 to 2000 only.gov. In order to check whether at least one of the independent variables would be able to explain the variability found in crime rate. Note that in the NSCB Philippine Statistical Yearbook. expenditures on social services. these observations were deleted from the data set. This level of significance was chosen since studies in crime rate do not present very severe consequences. Since SAS would omit observations with missing values. A level of significance of 0. family income. unemployment rate. The data set was also checked for any possible encoding errors. family expenditures. Here we check for any missing values for the variables. literacy rate. crime rate is defined per 1.nscb. poverty incidence. number of policemen.05 level of significance will lead to the removal of . the data is subjected to checking. the subject matter itself is of importance. deleted as it presents a possible encoding error. number of courts. then. crime rate is defined as the number of crimes per 100. This observation was.asp. enrolment rate.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 11 Methodological Sketch The data used in this study are obtained from the National Statistical Coordination Board‘s publication The Philippine Countryside in Figures.ph/countryside/default. and the two dummy variables for location. However. for this study. human development index. The variable crime rate was then regressed on the seventeen variables population density. consumer price index. Before fitting. the province of Camiguin recorded a number of 73549 policemen. The electronic format may be accessed via http://www. in which a pvalue of greater than the stipulated 0. since it is one of the foremost problems present in the Philippines.

Autocorrelation was checked using the Durbin-Watson statistic for which a value of d close to 2 is desired. Crime rate is then regressed on the remaining independent variables. Kolmogorov-Smirnov.05. and the same process is repeated until all the independent variables become significant. As for outliers. normality. As such. 0. These include tests on multicollinearity. Influence may be checked through Cook‘s D. an assessment of outliers is essential to ascertain whether or not any outliers would greatly influence the model. White‘s test and the spec option were utilized. for which pvalues should not be less than the level of significance. In order to assess whether or not the model is good. is not the only criterion in checking the soundness of the model. there are two areas for which they have to be detected: outliers in the dependent variable and outliers in the independent variables. the independent variable with the highest p-value is removed first. and Anderson-Darling tests. Thus. it is also necessary to check the influence of the outliers. The coefficient of multiple determination. Outliers among the dependent variable are detected using Studentized Residuals. R2. linearity. A leverage greater than the cut off (2p/n) implies that the observation corresponding to that leverage is an outlier. A funnelshaped plot would indicate a problem in heteroskedasticity. Outliers on the independent variables are detected through the leverage (the diagonals of the Hat matrix). heteroskedasticity. and autocorrelation. and DFBETAS.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 12 that corresponding independent variable. Multicollinearity among independent variables was checked with the use of condition indices and proportion of variation. Values for the Studentized residuals that exceed those corresponding to the t table imply that the observations corresponding to that Studentized residual value are considered outliers. Departures from linearity were checked using partial regression plots. Not all outliers are influential. Normality was checked using Wilk-Shapiro. In order to be more certain about problems with heteroskedasticity. several diagnostic tests have to be performed. Cramer-Von Mises. Furthermore. . DFFITS. Heteroskedasticity was checked through the shape of the residual plot (versus the predicted value of y).

A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 13 If any of these criteria is violated. . necessary actions such as transformations and removal of unimportant independent variables would have to be performed. then the model can reasonably predict crime rate. On the other hand. if the proposed model meets all the criteria.

σ2) . 0 if otherwise geog2 = 1 if the province is in Visayas.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 14 Results and Discussion Preliminary results and Diagnostic Checking In order to ascertain which factors influence crime rate. The initial model for crime rate is 𝐶𝑟𝑖𝑚𝑒 = 𝛽0 + 𝛽1 𝑝𝑜𝑝𝑑𝑒𝑛 + 𝛽2 𝑝𝑜𝑣𝑖𝑛𝑐 + 𝛽3 𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑟 + 𝛽4 𝑝𝑛𝑝 + 𝛽5 𝑓𝑎𝑚𝑖𝑛𝑐 + 𝛽6 𝑓𝑎𝑚𝑒𝑥𝑝 + 𝛽7 𝑙𝑖𝑡𝑟 + 𝛽8 𝑓𝑙𝑖𝑡 + 𝛽9 𝑒𝑛𝑟𝑜𝑙𝑚𝑒𝑛𝑡 + 𝛽10 𝑐𝑝𝑖 + 𝛽11 𝑑𝑖 + 𝛽12 𝑠𝑜𝑐𝑠𝑒𝑟𝑣 + 𝛽13 𝑐𝑜𝑢𝑟𝑡𝑠 + 𝛽14 𝑔𝑒𝑜𝑔1 + 𝛽15 𝑔𝑒𝑜𝑔2 + 𝛽16 𝑐𝑜𝑠𝑢𝑟𝑣𝑒 + 𝛽17 𝑐𝑜𝑠𝑢𝑟𝑣𝑠 + 𝜀 where crime = crime rate per province popden = population density per province povinc = poverty incidence per province unemployr = unemployment rate per province pnp = number of policemen per province faminc = average family income per province famexp = average family expenditure per province litr = literacy rate per province flit = functional literacy per province enrolment = enrolment rate per province cpi = consumer price index per province hdi = human development index per province socserv = expenditures on social services per province courts = number of courts per province geog1 = 1 if the province is in Luzon. a regression model was built. 0 if otherwise. cohsurve = cohort survival rate for elementary education cohsurvs = cohort survival rate for secondary education ε ~ N(0.

07913 -0.06577 0.20 -2. it is apparent that at least one of the independent variables can explain crime rate.52 Pr > |t| 0.5507 0.0333 0.13971 -0.70 0.7593.93 percent of the variability found in crime rate.4818 0.00550 -0.97 -2.39113 0.75 -0.53537 27.0001 Root MSE Dependent Mean Coeff Var R-Square Adj R-Sq 0.12483 -0.0351 0.00250 0.29 2.12593 0.10521 0. This means that the model formulated can explain 75.7695 0. And in checking the t-values.00007928 0.81 1.61629 Mean Square 64.6053 The coefficient of multiple determination has a value of 0.0001 0.0972 0. This means that the variability found in crime rate may be attributed to the regression model rather than the error.17788 2.18 0.3285 0.09856 0.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 15 Using the F-test under ANOVA.95606 4.0309 0. The mean sum of squares due to regression is also relatively large compared to the mean sum of squares due to error.42 Pr > F <.1340 0.11416 0. indeed.64607 349.00126 0.34424 0.86153 8.05934 1. .04865 2.0003 0.11380 0.17442 0.27471 0.21994 t Value 0.15076 -0.14712 -0.53 -4. Individual T-tests Parameter Estimates Variable Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT CPI GEOG1 GEOG2 HDI SOCSERV COURTS COHSURVE COHSURVS Label Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT CPI GEOG1 GEOG2 HDI SOCSERV COURTS COHSURVE COHSURVS DF 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Parameter Estimate 6.60 0. there are.00009885 0.09009 49.6306 0.00579 0.48 0.11458 Standard Error 22.22 -0.64 3.7593 0.2282 0.99 0.22 -1.72 -1.4564 0.24 -1.1082 0.06194 0.00532 -0.30810 68.73830 Source Model Error Corrected Total DF 17 40 57 F Value 7.00009779 0.00952 0.02358 1.50746 0.6570 Table 2. Table 1.4783 0. some independent variables that are significant.53181 1452.23942 1.0784 0. Analysis of Variance Results Sum of Squares 1102.00013465 0.71 1.45229 -0.

00009276 0.67 0.7579.51 0.18 -2.06 2.1165 0.00516 -0.00910 0.6120 0.98527 52. cohort survival rate for elementary became significant.00013025 0. The R2 dropped from 0.1027 0.4775 0.00125 0.99 -2.0003 0.2178 0.0331 0.09764 0.06398 0.0001 0.60 3.3464 0.15067 -0.65 1.25 -1.1305 0.72 0.10378 1.25 -0.0354 0.66 -1.95 0. and the number of courts.01635 0.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 16 Out of the seventeen independent variables in the model.00492 0.11303 0.00007802 0.21 -1.97 1. These are population density.12397 0.56 Pr > |t| 0.44333 -0.98843 1. . GEOG1 was then removed from the model because of its high p-value.0560 0.5166 0.50338 0.07440 1. poverty incidence.00531 -0.7593 to 0.16 0. Individual T-tests without CPI Parameter Estimates Variable Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT GEOG1 GEOG2 HDI SOCSERV COURTS COHSURVE COHSURVS Label Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT GEOG1 GEOG2 HDI SOCSERV COURTS COHSURVE COHSURVS DF 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Parameter Estimate 1.14110 -0.12947 0.26792 0.0366 0.00237 0.6306).11123 0.50581 26.14148 -0.00567 0. The first variable to be removed is CPI since it has the highest p-value at a value (0.9488 0.21731 t Value 0. only four turned out to be significant.54 -4.5137 0.11264 -0.17098 0. Table 3. number of policemen.22757 0.00009738 0.5768 After removing CPI.12224 Standard Error 19.

00013622 0.00007695 0.05650 0.19183 t Value -0.38564 0.05859 0.11183 0.0384 0.14972 -0.2893 0.07 0.00010185 0.46 2.5597 0.00495 -0.57 4.44495 0.08694 0. Individual T-tests without CPI.0327 0. Table 5.28893 0.20001 t Value -0.1556 <.00533 -0.16717 0.63 4.0400 0.16 -2.09 0.63 -1.00547 -0.0001 0.2829 0.28597 0. The variable GEOG2 is then deleted since its p-value is the highest among the remaining variables.00115 0.10 0.7563.4681 .12031 0.46781 -0.58 0.30 -0.0884 0.0344 0.2020 0. The number of significant variables did not change in any way.5626 0.46252 -0. and GEOG2 Parameter Estimates Variable Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT HDI SOCSERV COURTS COHSURVE COHSURVS Label Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT HDI SOCSERV COURTS COHSURVE COHSURVS DF 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Parameter Estimate -2.05250 53.55 -1.20356 0.0455 0.0366 0. Individual T-tests without CPI and GEOG1 Parameter Estimates Variable Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT GEOG2 HDI SOCSERV COURTS COHSURVE COHSURVS Label Intercept POPDEN POVINC UNEMPLOYR PNP FAMINC FAMEXP LITR FLIT ENROLMENT GEOG2 HDI SOCSERV COURTS COHSURVE COHSURVS DF 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Parameter Estimate -3.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 17 Table 4.74 1.95045 0.14 2.6486 0.0419 0.21 -1.50 -4.72 -2.11032 0.14496 -0.12 0. either.03 1.1240 <.09915 -0.16347 Standard Error 16.09181 0.09554 25.00013425 0.73 Pr > |t| 0.68016 0.00009402 0.09682 1.36 -0.14043 Standard Error 16.00898 0.1406 <.0001 0.4184 For this model the R2 is 0.0001 0.1107 <.60 -1.11823 0.16562 0.00010224 0.05913 0.09792 0.5855 0.14 -2. GEOG1.13270 -0.14410 -0.79 1.09593 0.00559 0.55 2.09829 -0.00877 0.0001 0.19 -1.12928 -0. which is still not different from the initial model‘s R2.27523 26.0806 0.8896 0.00230 0.00232 0.00007612 0.25504 0.59 -1.58537 53.0492 0.5525 0.8448 0.06 1.45 -4.20 2.1818 0.00495 -0.70 -2.09950 0.00552 0.00009490 0.00120 0.24798 0.12928 -0.82 Pr > |t| 0.

and number of courts. this model can only .07475 72.39710 Standard Error 1.07180 0.08870 t Value 4. deleted one at a time. it can be observed that there are six significant variables after removing geog2.05: population density.00182 0.11473 -0.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 18 Here.0001 Root MSE Dependent Mean Coeff Var R-Square Adj R-Sq 0. unemployment rate would have an effect on crime rate.00096282 0. From an initial value of 0. However. there are four significant variables at a level of significance of 0.0940 <.0011 0.7551. It can be observed that unemployment rate is still not significant.44 -1. Table 6. cohort survival rate for secondary education.89624 8. literacy rate. and expenditures on social services.68 -3.29 Pr > F <.7597.71317 F Value 23. Therefore. it is now only 0. it is retained since theoretically.44154 Mean Square 202.30 4. cohort survival rate for elementary education. poverty incidence. however. human development index.41891 2.0001 0.48120 487.48 Pr > |t| <. ANOVA Results and Individual T-tests for the Modified Model Analysis of Variance Source Model Error Corrected Total DF 5 56 61 Sum of Squares 1014.84835 0.7563 to0.0001 Note that in this modified model.00669 -0.6752 0.91 3.6752.95181 4. family expenditure. The F-test for this model implies that at least one of the independent variables will be able to explain the variability found in crime rate. The R2.93771 1502.0005 0. dropped. family income. The next variables to be deleted were enrolment.00414 0.12234 -0. The coefficient of multiple determination dropped from 0. number of policemen.70 -4. which is still not a far cry from the initial R2. The results are shown on the following table.03338 0. functional literacy.07796 0.6462 Parameter Estimates Variable Intercept POPDEN POVINC UNEMPLOYR PNP COURTS Label Intercept POPDEN POVINC UNEMPLOYR PNP COURTS DF 1 1 1 1 1 1 Parameter Estimate 9.0001 <.

90266 0. since 12 variables were removed from the model. we have the following result. though. In checking for normality.52 percent of the variability found in crime rate. It is necessary for remedial measures such as transformations to be performed. diagnostic checking is necessary.211265 1. These will be discussed later on in this section.0050 Note that for all the tests for normality.σ2) Before this model can be accepted as the best model.0050 <0.0222 <0.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 19 explain 67. The new model is given by: 𝐶𝑟𝑖𝑚𝑒 = 𝛽0 + 𝛽1 𝑝𝑜𝑝𝑑𝑒𝑛 + 𝛽2 𝑝𝑜𝑣𝑖𝑛𝑐 + 𝛽3 𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑟 + 𝛽4 𝑝𝑛𝑝 + 𝛽5 𝑐𝑜𝑢𝑟𝑡𝑠 + 𝜀 where crime = crime rate per province popden = population density per province povinc = poverty incidence per province unemployr = unemployment rate per province pnp = number of policemen per province courts = number of courts per province ε ~ N(0. 0.377216 -----p Value-----Pr Pr Pr Pr < > > > W D W-Sq A-Sq 0. The null hypothesis of normality of error terms is then rejected. . the p-values are less than the level of significance. This is understandable. Table 7. Tests for Normality Tests for Normality Test Shapiro-Wilk Kolmogorov-Smirnov Cramer-von Mises Anderson-Darling --Statistic--W D W-Sq A-Sq 0.0001 0.121649 0.05.

k‘ = 5. If d>du. Results for Spec Option. the values are n = 65. the test becomes inconclusive.754. Autocorrelation and multicollinearity were also checked. This value is compared to the tabulated values for the Durbin-Watson test.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 20 After normality. the model is free of problems on multicollinearity.24895 0.99517 6.11903 0.75062 4. However. For autocorrelation.438.246 with a first order autocorrelation of -0.767.00000 2.83022 0.02898 Condition Index 1. it can be observed that d lies in between dl and du. As for multicollinearity. Thus.37016 12. Durbin-Watson Test.13439 0. and Multicollinearity Indicators Collinearity Diagnostics* Number 1 2 3 4 5 6 Eigenvalue 4. Table 8. the test statistic under the Durbin-Watson test is 2. In order to ascertain this. The result indicates that the null hypothesis of constant variance should not be rejected. it can be observed that the condition indices do not exceed 30.63842 0. This will yield a value of 1. the spec option under the regression procedure was used. dl = 1. Using the table. the null hypothesis is not rejected. Results for these are shown on Table 8. The residual plot does not seem to exhibit any shape (funnel or diamond) that would imply heteroskedasticity. the null hypothesis is rejected. and du = 1.127 Durbin-Watson D Number of Observations 1st Order Autocorrelation *Proportion of variation is omitted. homoskedasticity is checked.40480 5. For a negative value of the first order autocorrelation.90959 Test of First and Second Moment Specification DF 20 Chi-Square 16.96 Pr > ChiSq 0. . the statistic 4 – d is used instead of d. if d<dl.127.6558 2. In this case.246 62 -0. A residual plot (versus the predicted value of crime rate) is utilized in order to check for homoskedasticity.

Cook‘s D.6633 19. Partial regression plots were obtained for crime rate versus each independent variable. The same observations may be influential as well.0912 Residual 8.0341 Residual 5.254 respectively.3173.407 and a Studentized deleted residual of 2.3599 1. Observations 12. its DFFITS is equal to 1.8476 0. These may be detected through Studentized residuals for dependent variables and leverages for independent variables.5191. Its leverage is equal to 0. As for the leverage. number of policemen. and DFBETAS have to be consulted. This observation may be considered influential since its Cook‘s D has the highest value relative to the other observations.1935.4596 0.6514 Observation 12 has a Studentized residual equal to 2. In checking for influence. The cut-off for Studentized residuals is equal to two. These values exceed the cutoffs computed.8913 Obs 12 18 Obs 58 RStudent 2.439 2.0296 4.7931 1.2853 Obs 58 RStudent 3.1896 Std Error Student Residual Residual 2. Moreover.3173 0. Table 9.622 and 0.4141 Std Error Student Residual Residual 2.5191 2. it can be observed that there are three possible outliers from the observations. and number of courts.6754 14.7934 Cook's D 0.3155 Hat Diag H 0. and 58 are possible outliers.7174 .066 DFFITS 1.5701 Hat Diag H 0. 18.382 Dependent Predicted Std Error Variable Value Mean Predict 22.407 2.7174 0. poverty incidence.814 1. the cut-off computed is 0.449 0.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 21 Linearity is also checked to ascertain whether any departures from it may be observed.8702 5.657 3. Outliers and Influential Observations Output Statistics Obs 12 18 Dependent Predicted Std Error Variable Value Mean Predict 25. population density.6627 10.979 Cov Ratio 0.6442 Cook's D 0. DFFITS. The plots show that no distinct departure from linearity may be observed (plots found in the appendices). For outliers.129 Cov Ratio 0.4339 DFFITS 1. The cut-offs for DFFITS and DFBETAS are 0.

129. Observation 18 has a Studentized deleted residual equal to 2. PNP. and the DFBETAS under POPDEN and POVINC exceed the cut-off. Its Cook‘s D is not relatively high. PNP. a Studentized deleted residual equal to 3. and may be candidates for influential observations. These support the supposition that observation 58 may be an outlier. and COURTS exceed the cut-off.6442.0912.1896. This implies that the value of crime rate may be an outlier for this observation while its independent variables are not. Observation 58 has a Studentized residual equal to 3. its DFFITS is equal to 0. . and COURTS exceed the cut-off.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 22 which is well beyond the cut-off. DFBETAS under the variables UNEMPLOYR. and leverage equal to 0.4141. however. and may also be candidates for influential observations. It may also be influential since its DFFITS is equal to 1.6514.0437 but its leverage is equal to 0. The DFBETAS under the variables POVINC.

poverty incidence. One possible corrective measure is transformation of variables. linearity. sqpovinc. autocorrelation. σ2) Again. it is necessary to perform corrective measures. Finally. the combination of the square roots of crime rate. this model is checked for normality.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 23 Corrective Measures Since there is a problem in normality. and outliers. heteroskedasticity. Some of these are the natural logarithms and square roots. For this particular study. Thus. the transformed model is 𝑠𝑞𝑐𝑟𝑖𝑚𝑒 = 𝛽0 + 𝛽1 𝑝𝑜𝑝𝑑𝑒𝑛 + 𝛽2 𝑠𝑞𝑝𝑜𝑣𝑖𝑛𝑐 + 𝛽3 𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑟 + 𝛽4 𝑝𝑛𝑝 + 𝛽5 𝑠𝑞𝑐𝑜𝑢𝑟𝑡𝑠 + 𝜀 where sqcrime = square root of the crime rate per province popden = population density per province sqpovinc = square root of the poverty incidence per province unemployr = unemployment rate per province pnp = number of policemen per province sqcourts = square root of the number of courts per province ε ~ N(0. multicollinearity. and sqcourts respectively) and population density helped in correcting the problem of normality. and number of courts (coded as sqcrime. Several combinations of transformed variables were also considered in order to correct the problem. several transformations of variables were made. .

49 5.00096767 Error 0.98 Pr > F <. Table 10. Results of which are found on the following table.55679 -0.32376 F Value 27. Recall that the R2 of the previous model was 0.57121 Mean Square 9. The four independent variables are still significant at a level of significance of 0. The results for which are shown on Table 11.05893 0.0001 0.6278 <.13 Pr > |t| <.29978 -0. The R2 of the transformed model is equal to 0.56900 1.13036 63.00018859 t Value 5.6752. . ANOVA Results and Parameter Estimates for the Transformed Model Analysis of Variance Source Model Error Corrected Total DF 5 56 61 Sum of Squares 45.55258 0. and autocorrelation. there was a significant increase in the R2 of the model.0291 0.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 24 After transformation.7141 0. multicollinearity.17 -0. This means that there is an improvement in the amount of variability the transformed model can explain.0001 The first tests to be performed after transformation are tests for heteroskedasticity.01379 0.29466 18.07181 0.74693 32.24 -4.0001 Root MSE Dependent Mean Coeff Var R-Square Adj R-Sq 0.7141.35 2.00036395 0.09904 0. while unemployment rate remains insignificant.42501 0.6886 Parameter Estimates Variable Intercept POPDEN SQPOVINC UNEMPLOYR SQCOURTS PNP Label Intercept POPDEN SQPOVINC UNEMPLOYR SQCOURTS PNP DF 1 1 1 1 1 1 Parameter Estimate 2.00672 0.05.95684 0.0001 0.0001 <.62 -5.00081505 -0.

90349 0.01086 Condition Index 1. normality. These observations were removed one at a time.02612 0.72823 0.00093940 0.00690 0. This will lead to the nonrejection of the null hypothesis of no autocorrelation.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 25 Table 11.01132 0. Next.00474 0. Recall that there are three outliers and influential observations.063.00249 0. and a first order autocorrelation equal to -0. outliers.19225 Test of First and Second Moment Specification DF 20 Chi-Square 16. Also. the outliers and influential observations were addressed.85197 8.04728 0.71655 0. Tests for Multicollinearity.7094 2. This yields a value of 1. Heteroskedasticity and Autocorrelation Collinearity Diagnostics Number 1 2 3 4 5 6 Eigenvalue 5.00060965 0.05286 0.00027899 0. For this. This implies that there is no multicollinearity after transforming the model. if this statistic is compared to the value on the Durbin-Watson table (recall that du is equal to 1.06375 0. and .00234 0. which is close to 2.037 Durbin-Watson D Number of Observations 1st Order Autocorrelation Note that the condition indices are less than 30.79524 Collinearity Diagnostics -----------------------------Proportion of Variation---------------------------Number Intercept POPDEN SQPOVINC UNEMPLOYR PNP SQCOURTS 1 2 3 4 5 6 0. At each stage. and at each removal of observation.36060 0.037. it can be observed that the value for 4 – d exceeds du.937. a diagnostic check is performed. it can be concluded that there is no problem of autocorrelation.02990 0. the model is checked for multicollinearity.45895 0.00305 0.10986 0.063 62 -0.97765 0.11067 21.02890 0.00000 3.00725 0.25731 0.29102 6.22684 0.15775 0. Thus.12 Pr > ChiSq 0.40328 0.35256 5.14538 0. The Durbin-Watson test statistic has a value equal to 2.00512 0.767). autocorrelation. consider instead the statistic 4 – d.45889 0.02230 0.07841 0.14420 0.00195 0.04345 0.08777 0.01717 0.18424 0.00123 0.01756 0.

43102 0.7246 Parameter Estimates Variable Intercept POPDEN SQPOVINC UNEMPLOYR PNP SQCOURTS Label Intercept POPDEN SQPOVINC UNEMPLOYR PNP SQCOURTS DF 1 1 1 1 1 1 Parameter Estimate 2.97826 0. Parameter Estimates.18696 .44434 1.46557 1.40 2. The results for these are found in Appendices A-30 onwards. Table 12.00261 0.07627 0.27 Pr > |t| <.75883 0.59768 0.0207 0.00031984 0.00076341 -0.00848 -0.79816 0.91652 0.00092325 0.10913 0. ANOVA.00000 3.01191 0. After removing these three observations.21676 F Value 30.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 26 heteroskedasticity.08455 5.7487 0.60067 -----------------------------Proportion of Variation----------------------------Number Intercept POPDEN SQPOVINC UNEMPLOYR PNP SQCOURTS 1 2 3 4 5 6 0.88793 8.10920 0.58547 11.18328 0.60853 1.42650 5.00171 0.02384 0.01183 0.0001 0.01606 0.00082873 0.00086302 0. and Multicollinearity Tests Analysis of Variance Source Model Error Corrected Total DF 5 52 57 Sum of Squares 33.17743 0.01014 Condition Index 1.31492 6.43714 0.4767 <.53004 Standard Error t Value 0.00465 0.61632 28.00010800 0.04533 0.80432 Mean Square 6. again checked for the previously mentioned criteria.01025 0.00239 0.85676 0.09268 22.00302 0.25736 -0.0001 Root MSE Dependent Mean Coeff Var R-Square Adj R-Sq 0.00501 0.00725 0.06349 0. the final model is.44097 0.0001 <.00015836 0.27129 44.02209 1.04222 0.48076 0.71709 0.99 Pr > F <.0001 Variance Inflation 0 1.54602 2.09 -0.72 -5.0001 0.47734 0.22789 0.00057169 0.00065690 0.06291 0.02305 0.07905 0.05667 0.00744 0.03426 0.39 -4.00378 0.23 6.23433 0.10035 Collinearity Diagnostics Number 1 2 3 4 5 6 Eigenvalue 5.

Thus. Under this model. Thus.095625 0.2244 After these corrective measures.081 Durbin-Watson D Number of Observations 1st Order Autocorrelation Tests for Normality Test Shapiro-Wilk Kolmogorov-Smirnov Cramer-von Mises Anderson-Darling --Statistic--W D W-Sq A-Sq 0. k‘ = 5).33 Pr > ChiSq 0. there seem to be no distinct departures from linearity.9372 2. This means that the model can explain 74. and again. and a value of 1.980824 0. If this is compared to the tabulated value (under the Durbin-Watson Table. In testing for heteroskedasticity. n = 60. d = 1. the null hypothesis of constant variance is not rejected.33.142 58 -0. the error terms follow a normal distribution.4877 >0.858 is greater than du = 1. This means that multicollinearity is not a problem with this model.05.767. Results for Spec Option.142 with a first order autocorrelation equal to -0. the final model has a coefficient of multiple determination equal to 0.87 percent of the variability found in crime rate. the statistic 4 – d is considered instead.486635 -----p Value-----Pr Pr Pr Pr < > > > W D W-Sq A-Sq 0. Also. Since the Chi-square value computed is equal to 11. there is no autocorrelation present.1500 0. the spec option in SAS is used. This implies that the null hypothesis that the error terms are normally distributed is not rejected.081.9372.1290 0. Linearity is checked using partial regression plots. . the p-values are well beyond the level of significance of 0. which is still close to 2. Again. Note that its VIF‘s do not greatly exceed 10 and its condition indices are all less than 30. The null hypothesis of no autocorrelation is not rejected. Durbin-Watson Test. the Durbin-Watson test statistic shows a value equal to 2. and its p-value is equal to 0.7487.096603 0. and Tests for Normality Test of First and Second Moment Specification DF 20 Chi-Square 11. Thus.858 is obtained.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 27 Table 13. there is no heteroskedasticity.

unemployment rate. Note that for the independent variables poverty incidence (povinc). there was political and economic instability due to the ouster of Former President Joseph Estrada. crime rate decreases.5014𝑐𝑜𝑢𝑟𝑡𝑠 ∗ where each parameter estimate after β0 represents an increase or decrease in the estimated mean of crime rate per unit increase in the corresponding independent variable holding all other variables constant. However. the signs of the coefficients are adverse to theoretical expectations. and unemployment rate and crime rate are inverted. a rise in poverty incidence would entail a rise in the crime rate. The estimated model for crime rate is then equal to 𝐶𝑟𝑖𝑚𝑒 𝑟𝑎𝑡𝑒 ∗ = 2. unemployment rate (unemployr). for every increase in poverty incidence.1856 + 0.2657𝑝𝑜𝑣𝑖𝑛𝑐 ∗ − 0. The same can be said for unemployment rate.0008𝑝𝑜𝑝𝑑𝑒𝑛 − 0. crime rate decreases. and crime rate for this year .0001𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑟 − 0. This is owed to the fact that during the year 2000. the relationships between poverty incidence and crime rate. equal to 𝐶𝑟𝑖𝑚𝑒 ∗= 𝛽0 + 𝛽1 𝑝𝑜𝑝𝑑𝑒𝑛 + 𝛽2 𝑝𝑜𝑣𝑖𝑛𝑐 ∗ +𝛽3 𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑟 + 𝛽4 𝑝𝑛𝑝 + 𝛽5 𝑐𝑜𝑢𝑟𝑡𝑠 ∗ +𝜀 where crime rate * = sqcrime povinc* = sqpovinc pnp* = sqpnp courts* = sqcourts. for this model. and number of courts (courts).A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 28 Conclusion The final model of crime rate is then. On the other hand. As common sense would dictate. For every increase in unemployment rate. a rise in the number of courts would mean a decrease in the crime rate. If the poverty incidence. That is.0008𝑝𝑛𝑝 + 0.

it can be observed that there is a direct relationship between this and crime rate. Thus. Although there is a difference between theory and empirical data in this study. the model may be a plausible predictor of crime rate in the Philippines. relative to other years. through this model. there would also be an increase in the number of reported crimes. and is still a plausible one. As for the number of courts. the opportunity for people to file cases would also increase. Since this model has satisfied the conditions and assumptions. poverty incidence. it can be observed that the poverty incidence and the unemployment rate for the year 2000 are high and the crime rate for the same year is low. . Thus. number of police per province.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 29 are compared to the others. which would lead to an increase in the crime rate. the model obtained is not extraordinary. and number of courts per province. we were able to establish a linear relationship between crime and the factors population density. This may be due to the fact that as the number of courts increases.

Separating the regression lines would allow for classification of factors among different types of crime. This may lead to a better model in terms of the coefficient of determination. and crimes against property and person) as different factors may affect each category of crime. As a means of improving the study. this particular study focused on crime rate in the Philippines from provincial data for the year 2000 only. The group also recommends formulating separate regression lines for the different classifications of crimes (index and non-index crimes. .A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 30 Recommendations As previously mentioned. the group recommends considering data from other time periods as well as data from municipalities or cities.

org/ Ehrlich. Makati City. poverty. (2004. United States of America: National Bureau of Economic Research. In F. I.gov.A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 31 References The Philippine countryside in figures. University of Washington). 76(2).asp Becker. crime. The current situation of crime associated with urbanization: problems experienced and countermeasures initiated in the Philippines. Reynolds.jp/english/pdf/PDF_rms/no68/09_Leones-1_p133150. On the relation between education on crime. & Cruz.. . (n. (2003). G. 47(1).org/ Ehrlich. T.. 2010. The Journal of Political Economy.ph/ncs/9thncs/papers/publicOrder_PanelData. Pakistan Economic and Social Review.. 397-417. Panel data estimation of crime rates in the Philippines. M. F. The deterrent effect of capital punishment: a question of life and death. M. The American Economic Review. Washington. and crime nexus: cointegration and causality analysis of Pakistan..pdf National Statistical Coordination Board. Retrieved from http://www.jstor. Determinants of crime rates in Latin America and the world: an empirical assessment. Retrieved from www.pdf. Crime and punishment in Texas in the 1990s. c. I. 65(3). October 4-5). and context: a multi-level analysis of the relationship between work and crime (Doctoral Dissertation. Juster (Ed). Sanidad-leones.nscb.ncpa. Wadsworth. Retrieved from http://www.T. (1975). (1975). Yasir.gov. Fajnzylber. Crime and punishment: an economic approach. Available from the National Criminal Justice Reference Service (NCJRS) website http://www. T.org/pub/st237?pg=6. and human behavior (pp. The Philippine countryside in figures (2003 edition).d. Gillado. Employment.or. 169-217. (2001). (2009). D. inflation.).jstor. income. (2000). Unemployment. Retrieved May 18.gov/pdffiles1/nij/grants/198118.unafei. Retrieved from http://www. (1998). Education. et al.pdf. 79-98.ncjrs. (1968). Retrieved from http://www.T. S. 313–338).nscb.C.ph/countryside/default. United States of America: The World Bank. et al. from http://www. Philippines: Author. pp. (2010). P.

Durbin-Watson Table .A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces 32 Appendices Appendix A. Results (SAS Outputs) Appendix B.

- Thesis Criminology 2
- NSO Philippines in Figures 2010
- EViews 7 Users Guide I
- 144464779 STATA Commands
- Ch18 Multiple Regression
- Regression Exercise IAP 2013
- Chapter 8
- Chapter4 Solutions
- EviewsTutorial2011-2012v7
- e Views Guide
- chap14
- Points for Session 4_updated
- 372 Report
- Simple Regression With SPSS
- Forecasting and Regression Berg
- Quantitative Analysis
- Regression Using R
- www_statisticshowto_com_excel_regression_analysis_output_exp.pdf
- Analysis of Mva on Eva
- CH12
- MLR with Dummy.pdf
- Poisson Regression
- Regression
- Chapter03 Solutions Hansen Mowen Cost Behavior Sixth Edition
- Regression
- Getting Started With Regression
- Ads
- ETC1000 2008 S1 Exam
- 1c778c8180655ba623a6996f16714668_business-stat-module-2.pdf
- Correction
- A Regression Analysis on the Determinants of Crime Rates Across Philippine Provinces - Revised

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd