INternational Journal of Criminal Justice Sciences

Vol 3 Issue 2 July - december 2008

Copyright © 2008 International Journal of Criminal Justice Sciences (IJCJS)   ISSN: 0973-5089 Vol 3 (2): 138–157

This is an Open Access article distributed under the terms of the Creative Commons Attribution-Non-Commercial-Share Alike License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. This license does not permit commercial exploitation or the creation of derivative works without specific permission.

The Etiology of Crime in India[1]


Arvind Verma[2]

Indiana University, Bloomington, USA

 Manish Kumar[3]

Research Associate, New Delhi, India


The reasons why people commit crime willfully has been a matter of great concern. However, most theories of criminal behavior are grounded in western society and culture. Attempts to explore the causes of crime in the Indian context and situation have been sporadic and generally non-empirical in nature. This research explores the causes for criminal motivations in India. The paper describes and analyzes the major reasons of criminality amongst convicted offenders. Results suggest that growing urbanization, breakdown of extended family system and lure of consumerism and influence of criminal associates are some of the main reasons for these people to fall into a life of criminality.


Key Words: Crime; India; Convicted Offenders; Prison.




    The scientific inquiry into the causes of criminal behavior is now more than hundred years old. The discipline traces its roots to the growth of the Classical School in Europe around the period of the French Revolution. A series of legal reforms were propounded by social reformers such as Beccaria and Rousseau that established the foundations of modern legal justice system. The reforms that led to the codification of the new laws, creation of the modern police forces and the prison systems were radical innovations that also changed fundamental assumptions about criminal behavior (Brantingham & Brantingham, 1991; Hall, 1960). The belief that criminals are different from ‘normal’ changed and criminal behavior came to be explained in terms of social, economic, political, demographic and psychological factors.

    This scientific search for an etiology of crime has contributed an impressive literature to the subject. Criminology has now grown into a major subject of inquiry that is characterized by an inter-disciplinary approach and is taught on most university campuses. Explanations of criminal behavior draw concepts from a vast array of social and physical sciences. The reasons why people are motivated to commit crime range from an Ecological perspective (Shaw & McKay, 1942) to Differential Association (Sutherland, 1992), Cultural Conflict (Sellin, 1938), Sub-Cultural influences (Cohen, 1955; Matza, 1969), Anomie and Strain (Durkheim, 1951; Merton, 1968), Lack of Opportunities (Cloward & Ohlin, 1960) to Lack of Self Control (Gottfredson & Hirschi, 1990). However, all these explanations are strongly rooted in the structure, culture and history of the western [read American] society. The empirical support for most of these theories of crime and deviance has come from the studies of urban growth and social disorganization witnessed in western societies (Baldwin & Bottoms, 1976; Brantingham & Brantingham, 1991; Fattah, 1991; Voss & Petersen, 1971; Fetchenhauer & Simon, 1998; Indermaur, 1999; Paternoster & Brame, 1998; Wikstrom, 1991).

    There has not been a similar development of criminological theorization in India for criminology has yet to emerge as a strong discipline in Indian research institutes. Only a handful of universities in India have an independent criminology department and few institutes offer doctoral programs in this field. Moreover, the application of some of these theories, such as Social Disorganization, to the Indian situation poses considerable challenges as the Indian cities hardly resemble the western urban centers. Most Indian cities are hundreds of years old where Mughal period structures, Victorian Villas, modern high rise apartment complexes and thatched huts exist side by side. Almost every residential neighborhood is surrounded by sprawling slums. There are few exclusive commercial layouts since workers and the Indian business class has traditionally favored living close to their work places (Thakur, 1980; Weinstein, 1974). In most areas the residential areas near the centers of the cities have high economic values in contrast to the North American ones. Furthermore, there is little evidence of zonal decay in Indian cities as found in Chicago by Shaw and McKay (1942).

    Moreover, the tradition of extended families still persists despite growing urbanization and scarcity of space. An obvious implication of having a large family is that supervision over the young is ubiquitous, social control is extensive and strict. It is no surprise that a careful study found juvenile delinquency to be a minor problem in a country of one billion people (Hartjen & Priyadarshini, 1996). Social Control Theory may provide an explanation but the wide spread prevalence of child labor in many industries and limited supervision during day by working parents in metropolitan cities will need additional explanations. Similarly, real and perceived social and economic inequalities have been a constant feature of the Indian society for hundreds of years. The caste system continues and people seem to accept their position in the social hierarchy without strong dissent. These disparities too, do not seem to promote strain and be a factor for criminal behavior. But social conflict and violence against scheduled castes appear to suggest that loss of traditional power amongst the so called ‘higher castes’ is causing strain, particularly from positive discrimination policies of the government for the ‘backward castes’.

    Clearly, the nature of the society, its culture and history are significant variables in understanding criminal behavior. It follows that the nature of criminality in India must be explained as a consequence of its peculiar social-cultural, political and historical conditions. The texts Manu Smriti and Arthashastra describe various forms of criminality in ancient Indian society and prescribe various ways of dealing with them (Shamashastry, 1929). However, these explanations were largely rooted in religious beliefs of re-incarnation. Criminal motivation was suggested to be a product of previous life’s Karma and influences (Sukla, 1987; Somerville, 2000).  During the British period a large section of the people were designated as ‘criminal tribes’ and were subject to extremely harsh treatment by the regime (Yang, 1985). Nevertheless, studies of these criminal tribes have been limited to the descriptions of social organization, family relationships and economic hardships suffered by these people (Simhadri, 1991). Some other literature books dealing with the concept of criminal behavior in India (e.g., Nayar, 1975; Upadhyaya, 1978; Shukla, 1982; Thilagaraj, 1987; Mohanty, 1990; Phulia, 1992), have largely focused upon officials’ descriptions of criminality or dealt with the psychological dimensions, comparing ‘delinquents’ with samples of ‘normal.’ A large number of police officers have written their memoirs or about policing in which references have been made about certain types of offenders (Lobo, 1992; Shah, 1992). Yet, these efforts have been impressionist rather than based on some careful scientific inquiry. The genesis of dacoity [armed robbery by a gang] in Chambal valley [north-central India] has been a popular theme too (Bhaduri, 1972; Khan, 1981; Jatar, 1980; Katare, 1972). Again, most explanations have simply suggested feudal exploitation as the cause that provoked many people of this region to take to arms. The fact that many gangs operating in this valley were composed of higher castes and wealthy people appears to suggest that feudalism may only be a partial explanation of dacoity in Chambal valley. Clearly, the accounts of criminal behavior in India remain unsatisfactory. Hartjen and Ketheneni (1996, p. 7) in a review of Indian literature on juvenile delinquency lament that “in contrast to the voluminous literature found for other countries, a coherent body of research and thought relating to delinquency in India is simply not available.”

    An empirical examination and a well formulated theory that provides testable hypotheses about the reasons why people commit crime in India are the objectives of this research. The circumstances that motivate, encourage deviance in the developing state, the economic and social disparities, the rapid urbanization and exposure to new cultural ethos brought about by the new communication systems are factors that may be construed to influence criminality. This paper makes an attempt to explore the causes of criminality amongst a specific set of people and thereby add to the knowledge of criminality in India.


The Research Framework


    This exploration of the etiology of crime in the Indian context is based upon an empirical examination of background of those people who have been recognized to be deviant by the laws and procedures of the country. The empirical reality that has been sought in this research work is an understanding of the circumstances and motivations that have forced some people to commit acts defined as crime by the law based upon the Positive School of Criminology perspective. For this purpose, the data has been collected from people convicted for their offenses through a well established due process of the law. There is thus little dispute about their legal innocence although one still cannot discount the absence of any kind of false accusations or defective procedures. Clearly, this approach is an unsatisfactory basis of criminality since the notion of criminal behavior is controversial. Treating officially incarcerated people as criminals and then exploring their reasons for criminality is tautological in one sense. Yet, this process appears to be a good starting point for an exploratory research. Understanding the reasons why people convicted by a due legal process got involved in criminal behavior is likely to provide some insight about the cause of deviant behavior in Indian society.

    It is nevertheless clear that results of this study cannot be extended to the general population for this method does not differentiate between criminals and non-criminals. People who have been incarcerated through a legal process are an unrepresentative and particular subset of the general population. Their behavior patterns cannot be ascribed to the people of India. Indeed, the population of interest in this research will simply be the set of convicted offenders- those labeled as ‘criminals’ by the processes of the criminal justice system. It also needs to be kept in mind that the criminal justice processes and procedures are themselves defective and biased (Verma, 2005). The nature of this study portend that any conclusion be treated with caution. Yet, we believe that this study is important for it does inform about some of the situations and circumstances that have driven a particular set of people towards criminality.


Criminogenic factors in the Indian context

    India is a heterogeneous society with an ancient culture where diversities stand out. There is a small affluent class that is as rich as any segment of people around the world. On the other hand the vast majority is impoverished. According to the annual report by the United Nations Development Program, the percentage of population living below $1 per day is an estimated 34.3 in 2005 (UNDP, 2009). The nation has recently acquired nuclear weapons and its Information Technology sector is catching the attention of the world. Yet, illiteracy is widespread and basic schooling is not available to a majority of the children. Women have reached top positions in every sector- politics, business, bureaucracy and academia and still the society remains strongly patriarchal where incidents of female infanticide and the problems of dowry are not uncommon (Madhok, 1998). India has the fourth largest economy in the world and there is rapid industrialization and urban growth. Nevertheless, rural population still accounts for 74.3% of the population (Census of India, 2001). Despite constitutional provisions of equality and one-person one vote, caste remains a dominant factor in social relations. A minority of so-called ‘forward castes’ continues to occupy the commanding heights of economic and political power in the society. In such a situation, the sociological, economic, political and psychological factors peculiar to the Indian society are likely to play a major role in determining the etiology of crime in the country.

    This research project therefore sought information from the convicted offenders about their i) Sociological factors such as family upbringing, social association, educational and work environment influences, ii) Economic variables such as income, number of family dependents and unemployment, iii) Political factors such as group conflict or labeling effect by the official agencies, iv) Psychological reasons like mental history of the family etc. Based upon the above-mentioned factors we developed the following hypotheses for criminal behavior:

1) Large, economically deprived families where parents did not have sufficient education created circumstances that encouraged deviant behavior

2) Neglected childhood was a major factor for falling into criminal behavior

3) Criminal association developed more for those growing up in low socio-economic status families

4) Criminal association led to a labeling effect that promoted subsequent criminal behavior

5) Failure of the justice system to provide speedy justice was responsible for forcing people to take to criminal acts

6) Growing consumerism in Indian society tempted the people to adopt illegal short cuts for making money and led them into criminal activities.

7) The loosening of social bondage due to rapid urbanization, induction of new cultural norms and an independent life style led some people towards deviant behavior.


Data Collection

     Data for this study was collected from Tihar Jail, the largest prison in the country, situated at Delhi, the capital of the country. This city, due to its political and economic importance attracts people (and offenders) from all corners of the nation. Accordingly, the possibility of finding a representative Indian sample at Tihar prison appeared to be high. There are some valid grounds for supporting this expectation.  Being the capital of the country and also the biggest business center of Northern India, Delhi attracts a large number of persons from all over India in search of employment and better opportunities. All the important institutions are situated here and there is a convenient connecting train from every major city of the country. It may also be pointed out that India has a uniform criminal justice system. Although, some acts are governed by state and municipal laws but the vast majority of crimes against property and body are enshrined in the Indian Penal Code. The procedures of investigation and prosecution are governed by the Code of Criminal Procedure 1973; again applicable to every region. The police and judicial systems are also uniform across the nation. It was hoped that the Tihar prison would be a good starting place for an exploratory study, which may be extended to other parts of the country once the viability and utility of the research design is established.       Initially, the plan was to have a sampling frame consisting of the roster or the official records of the prison that contain a list of all the convicted prisoners incarcerated there. However, despite several attempts, the permission to interview prisoners from within the prison premises was denied by the authorities. Accordingly, an arrangement was made to interview offenders who were released after serving their term. There is a turnover of almost 50-75 people every day from this prison. However, majority are under-trials who are arrested as suspects by the police and released on bail by the courts. The proportion of convicted prisoners in this lot would barely be 3-5%. Since our concern was with people whose legal criminality was not in doubt, we attempted to contact convicted prisoners only.

    A team of researchers with assistance by the prison guards identified these prisoners as soon as they came out of the prison. These people were then informed about the purpose of the study and assured of their anonymity. The procedure was to contact every second such convicted prisoner walking out of the prison. If he[i] did not agree then the fourth such prisoner was contacted and so on. Through this procedure an attempt was made to be systematic but this was a sample of convenience. Those who agreed were interviewed in one of the teashops scattered around the prison complex. The interview would normally take 25-30 minutes. Some relatives who came to receive the prisoners were also included in the interview to verify or add comments given by the convicted prisoner. A set of survey questions [in Hindi language] was provided to each of the researchers to ensure that the interview was uniform and the variables could be operationalized for statistical analysis.

There were many cases in which the respondents refused participation. Furthermore, the strictness about selecting every second person coming out of prison could not be maintained for various reasons. In some cases, the respondents initially agreed and then refused to continue. All these reasons did vitiate the procedure of selecting a probabilistic sample but nevertheless a quasi-random selection was followed.  Considering that the period of data collection was spread over a twenty month period [1999-2000] we feel confident that the collected sample is fairly representative of convicted prisoners released from Tihar prison during the study period.


Data Description

    As on 31st December 1999, the population of all the prisoners lodged at Tihar was reported to be 10625 and the number of convicted male prisoners was 1835 (Tihar prison, 2001). We collected data from 424 respondents from Tihar prison but two of the cases were found to be incomplete on some key questions. Accordingly, our sample consists of 422 respondents that roughly form 22.9% of the convict male population. This proportion is a probable estimate only since the numbers of convicted prisoner changes daily.

    The following are the descriptive statistics of these respondents. The mean age in the sample was 30.12 with a standard deviation of 12.09. These ages have been recorded on the basis of statements made by the respondents and some of this information seems doubtful. 11 respondents stated their age to be over 65 years and one said he was 11 years old. There is a possibility that some of them did not know their correct age and spoke in terms of approximate years. This also appears to be supported by the fact that detention above the age of 65 is rare and a child under the age of 12 who has not attained sufficient maturity will also not be detained for criminal behavior. Either there have been gross violations of these provisions [the possibility cannot be discounted] or else the ages recorded by the officials are different from the ones informed to us. It is also worth pointing out that these respondents were interviewed at the time of their release. Thus, their age at the time of committing the crime and being arrested would have been smaller. Clearly, in at least one of the cases the age appears inaccurate. Since we were denied access to official records we could not verify these statements about age and have retained the information given by the respondents. We aggregated the ages into three groups ranging from <22 years [frequency 114]; from 23-30 [frequency 180]; and >30 years [frequency 128]. Age is everywhere correlated with delinquency and research in the US suggests that the peak age is from 15-25 after which there is a gradual decline in criminal behavior (Hirschi & Gottfredson, 2000). The Indian data appears to conform to this pattern.

    In terms of their residential place, there seems to be a mixed situation: there is somewhat a larger proportion from the city but the number from the rural areas is not insignificant either. There were 64% from urban areas that do not seem to be in accordance with the 25.7% of urban population in the country (Census of India, 2001). Perhaps this is reflective of the adverse impact of industrialization and urban growth being witnessed in the country. The choice of living in an urban or a rural area is likely to offer different opportunities and set of associates. In view of the fact that most educational institutions and employment opportunities lie in urban areas, seems to supplement the sample finding that urban living is being preferred by growing proportion of the population.

    In terms of their educational background, an overwhelming majority had less than a High School certificate. A good number of the subjects were illiterate and lacked any formal education. Even for those who had attended some years in school or were beginners, it was clear from the interview that their ability to read and write was severely limited. Education of course plays an important role in criminal behavior (Crutchfield, Bridges, Weis, & Kubrin, 1999). Education is not only a socializing effect but opens new avenues and imparts skills that can assist in finding jobs and building a stake in conformity. Schooling creates social bonds that may affect delinquent behavior (Jenkins, 1997). Accordingly, we decided to consider the years of schooling in terms of the likelihood of imparting some skills and opening new avenues. The response category was thus sub-divided into four levels. The first category consists of those who had no schooling and described themselves as illiterates [n=116]. The next level consists of those who had some formal schooling, till grade VI on the basis that they had some exposure to formal training but it was not complete [n=61]. The third category is of people who have more schooling, from grade VII to High School [n=106]. The last category consists of those who have studied beyond High School and have done either their Intermediate [senior high school] or some form of College [n=139]. This group had a greater exposure to learning, was likely to pick some skills and knowledge and would have undergone greater socialization into conformity. Schooling or education did not appear to have a strong correlation with criminal behavior.

    Although criminologists have long debated about the effects of economic conditions on criminal behavior but an enduring consensus appears to be that unemployment has positive correlation with crime. Moreover, it has been suggested that juvenile delinquency has correlation with later adult unemployment (Robin, 1966). It is a common observation that marginalized groups are over-represented in jobs with low pay, poor benefits and limited prospects for future. Crutchfield and Pittchford (1997) suggest that those who are not confident about retaining their jobs are more likely to have committed crimes. Accordingly, we were interested in determining the employment status of the respondents. From the job descriptions provided by the respondents we determined that 97 were clearly unemployed and jobless, 263 were engaged in jobs that require little formal education and were mostly labor oriented. There were 37 respondents whose jobs required some formal education and training. Unfortunately, not everyone gave sufficient details to determine their employment status and hence the remaining was designated as ‘unemployed.’

    We decided to classify the job variable into that of Unemployed [n=110], Manual Worker [n=144], Skilled Worker [n=145] and White-Collar Worker [n=23] since these classes indicate better the nature of work and opportunities associated with it.

    The marital status is an important variable in the study of criminal behavior. An unattached youth is likely to have fewer stakes in the society while a married person is expected to realize greater responsibility towards his family. Marital bonds are also expected to strengthen conformity towards established behavior and restrict movement in both place and time. Accordingly, for us knowing this factor was important. The status was reported to be as Married [n=234]; Unmarried [n=185]. Widower [n=2] and one did not declare anything. There were no divorcees in the sample.

    Not only were most respondents from low socio-economic background but they also came from families where the parents had little or no formal education. We asked questions about the educational qualifications of their parents to probe this matter. It is expected that educated parents are more likely to ensure some education to their children and as we have mentioned above, greater education helps a person develop stake in conformity. Furthermore, educated parents may be more effective in the supervision of their children and perhaps more assertive in controlling their behavior. Accordingly, the educational level of the parents becomes an important factor in this exploration. The results are as follows:


Educational Qualifications of Parents


                                                Father                                      Mother

            Illiterate                       206                                          287

            Primary                        61                                            59

            High School                77                                            48

            College                        78                                            28


       We asked one more question to ascertain their socio-economic status. This question concerned the approximate monthly income of the family and whether their imprisonment resulted in some decline in this family income. Clearly, family income plays a decisive role in maintaining the family, in providing educational opportunities, in meeting the basic needs and other wants of the members and even determining the association with other people. We expect income variable to be an important factor in explaining the etiology of crime and categorized it in the following four ordinal levels:


Monthly Family Income


                                    Income                                                            No. of Respondents               

                               Less than Rs. 1000 /month                                                     119

            Between 1000 and Less than Rs. 3000/month                        118

            Between 3000 and Less than Rs. 5000/month                        76

            More than Rs. 5000/month                                                    109


    Religious belief and associated practices plays an important role in determining the behavior of the person in India. It not only places one into a specific mode of conduct but also determines who the associates and acquaintances will be. Furthermore, religion plays an important function in adhering to a moral conduct and in restraining behavior for fear of divine consequences. 443 respondents indicated that they follow the Hindu religion; there were 42 Muslims, 2 Sikhs, 1 Christian and 12 refused to answer. Considering that in India around 85% of the population follows the Hindu religion and Muslims constitute 12%, Sikhs and Christians perhaps 1% each, the above break-up appears to conform to the general population distribution without any one religion having unusually high or low representation. Of course religion may not be an important variable for what is important is the degree of adherence to religious beliefs and practices. However, in such a limited study it was difficult to measure this factor.

    We asked one more question regarding the size of the family. This seems to be an important consideration in the etiology of criminal behavior since large family size may reduce family income and control exercised on the respondents. 222 respondents stated their family size to be greater than 9 members including themselves. It is not only the family size but also the influence of family members that perhaps pushes or pulls a person to or away from crime. We were particularly interested to know if there were any family members who had themselves been arrested and prosecuted. We wanted to assess the influence that they may have had upon the criminal activities of the respondents. Criminality of any family member or friend is an important factor in drawing a person into the life of crime. Sutherland’s proposition that criminal behavior is learned within intimate personal groups and that this learning takes place in close association, where not only the tricks of the trade but also the justification for criminal actions are learnt (Sutherland, Cressey & Luckenbill, 1992), was likely to come from close family members and friends. Altogether 278 respondents indicated that they had criminal associates [of which 68 stated they had some family member arrested in a criminal case one time or the other]. This variable was named ‘Criminal Associate’ and dichotomized in Yes and No categories.

    Our last query was about the nature of crime committed by the respondent for which he had been imprisoned. According to the criminal statutes, offenses are categorized as those against the state, the body and property. Indian Penal Code of 1860 is the basic statute covering a majority of the offenses but there are many special laws too that cover areas like demanding dowry, bearing arms and social offenses against the weaker sections. In order to summarize these legal definitions of crime types in more understandable terms, we aggregated them into three classes: of property, physical and miscellaneous crimes. As the labels suggest, property crimes include offenses like theft [all kinds], robbery, extortion, dacoity [or its preparation], pick-pocketing, forgery, cheating and dealing in stolen property. The physical offenses include murder [and its attempt], assaults, rape, kidnapping, affray, rioting etc. The rest, like the offenses under the arms act, prostitution, smuggling, drug related and such were classified as miscellaneous types. Under such a categorization the frequency of offenders for these three crime types is summarized in table given below:

                        Property Offenses                    =         138

                        Physical Body Offenses            =         251

                        Miscellaneous Offenses            =         33


Reasons for Criminal Behavior

    The reasons for committing crimes were not answered directly in a straightforward manner and in many cases the respondents gave more than one explanation. The various reasons mentioned were enmity and association with ‘bad-companions’; disputes, domestic problems, anger, feeling of revenge or rivalry; greed, lust or need for money; influence of alcohol, drugs and intoxicants; influence of associates, family members, dispute over land, property, girl, politics, to being dragged into crime by various circumstances. Some acknowledged that an opportunity to commit a crime or such behavior becoming almost like a profession, or the ‘fun’ of action had been their reasons. A few also stated the desire for money; involvement with a girl, business interests, and poverty and unemployment problems initiated them into crime. Many stated that they were ‘dragged’ into the crime. This may have come from situational factors like old enmity or association with disreputable characters that initiated police suspicion towards these respondents. However, in this we need to acknowledge that a good proportion [11.4%] of the respondents resolutely indicated that they were falsely implicated in the incident or charged for a major crime, which was not the true incident. There seems little reason to disbelieve these assertions. The reputation of the police in India to act with probity is not very high (Verma, 2000) and false implications are possible. We noted every reason mentioned by each of the respondents and then determined a main reason under a broader classification that best characterized the situation.

    Quite clearly, a desire to gain money or property directly or by availing of an opportunity seems a common denominator of some of these statements. Similarly, dispute is another important factor that describes reasons such as enmity, rivalry, jealousy, and conflict over property, land, caste and arising from involvement with a girl and even politics. Another common denominator appears to be lack-of-control that is underlying the influence of alcohol, drug, peer-pressure, mental problem, and involvement with a gang, fun etc. There were some statements that suggest bad association to be a major reason. This variable covers reasons such as bad company, family influence, gang membership and so on. We also included a factor described as dragged that describes the situation wherein the respondent felt he was unintentionally drawn into the act due to fear, friendship, family matters or even false accusation. Clearly, this drag may have come due to one or more of the reasons described above such as bad-company, enmity or rivalry. Nevertheless, the respondents themselves asserted that they were not motivated to commit the act and did not feel guilty of crime. We believe that their belief needs to be mentioned since the Indian criminal justice system has serious defects (Verma 2005). There is a fair possibility of false accusations and other situations exist that can get an innocent person involved in a criminal case. We decided to honor this situational causality through a separate category. Accordingly, we developed five distinct types of reasons given by the respondents and categorized them based upon the main reason stated by them. These categories and associated numbers are summarized below:


Reasons                           Number of Respondents


                        Bad Association                                                 77

                        Dispute                                                            147

                        Dragged                                                             50

                        Gain                                                                  79

                        Lack-Of-Control                                                 69


Since the objective of this research is to explore the reasons why people commit crime this variable ‘Reason’ is our dependent variable- nominal in five categories.




    Our interest was in developing a model that could explain the dependent variable in terms of the social-economic and demographic variables described earlier. But before developing our model we examined different forms of association and tested bi-variate relationships to determine and select the most important independent variables.


Inter-variable Correlations

       We examined the bi-variate correlations amongst the variables to determine meaningful associations. The analysis suggested that respondents’ education, parent’s education, marital status and number of family members in the house are not significantly correlated with the ‘Reasons’ for criminal behavior. However, father’s education has relationship with employment status and family income, which have significant relations with the reasons for crime and hence we decided to retain these variables in the model.

We also explored the bi-variate relations through cross-tabulations. Our interest was in exploring if there is any inter-dependent relationship between the independent variables like age group, income and others with the dependent variable of reasons for criminal behavior. The following are the statistically significant results of cross-tabulating the variable ‘Reasons’ with five other variables:

Contingency Table Results

Cross-tab with ‘Reasons’

F Educ.        Place  Employment Age    Income        Crim. Assoc. Crime-Type

Chi.sq:             21.97               32.64   33.15   16.56   34.83   59.68               73.964

df                     12                    4          12        8          12        4                      8

p-value            0.038               0.000   0.001   0.035   0.000   0.000               0.000


    The above results suggest a close relationship between the ‘Reasons’ for criminal behavior and the respondents’ family income, employment status, place of residence [urban or rural], age group and quite obviously bad association. We also found that father’s educational level is a much better predictor than mother’s education level or even a combined variable created to reflect parental educational status. Sarvate (2000) argues that Indian families are strongly patriarchal and the father plays a dominant role in the family. Accordingly, we decided to use this as an independent variable. Naturally, the type of crime committed [body, property or other kinds] adds to the explanation of the dependent variable. These variables can therefore be used as explanatory factors in the multinomial model that we now describe below. Religion and family size [not reported above] did not provide a significant relationship and were dropped from the model.

    Accordingly, we decided to model the relationship with age [A], employment [EM], father’s education [FE], family income [I], bad association [CA], place of residence [P] and types of crimes committed [T] as the independent variables. Considering that the dependent variable of ‘reason for criminal behavior’ [R] is a nominal variable and some of the independent variables are also of nominal category, the best equation to model this relationship is the Multinomial Logit function. Our equation [and model] is of the form

R = f [A, EM, FE, I, CA, P, T].


Multinomial Logit function [MLF]

MLF is the most generalized probability model to handle explanatory [independent] variables that can range from the highest data type [continuous or interval category] to different categories of the lower types like nominal or ordinal types. The mathematical model is derived as follows:

Let pi (xi) denote the probability of response j, j = 1, ..., J, at the ith setting of values of k explanatory variables xi = (1, xi1, ..., xik). In terms of the response probabilities, the generalized Multinomial Logit model is given by


pi (xi) = [exp(bj xi)] / [S1J { exp(bj xi)}]                                                           A

            This reduces to log {[pj (xi)]/[ pJ (xi)]} = bj xi for j = 1, ..., J-1.


    Here a separate parameter vector applies for each of the J-1 non-redundant Logit equations. To obtain the ML estimates [the bis] the independent multinomial likelihood is maximized subject to the constraints of equation ‘A’. ‘The parameter estimates have large sample normal distributions with asymptotic standard errors that are obtained from the inverse of the data matrix. An iterative process like the Newton-Raphson’s method is used to determine these estimates’ (Agresti 1990: 313). The computations are quite complex and a computer software program; SPSS was used to estimate them. The results of this Multinomial Logit Regression modeling are as follows:



Model                                      -2 Log Likelihood       Chi-Square      df         Sig.

Intercept Only                         1131.309                                            

Final                                         887.159                       244.150           60        .000


    The Pearson value of this goodness of fit measure [1122.033] is also statistically significant at a p-value of 0.002. Cox and Snell’s value indicate a variance explanation of 43.9% while Nagelkerke unit suggests it is almost 46.0%. This also indicates that the set of independent variables are able to explain a substantial proportion of the variance of the dependent variable.


Likelihood Ratio Tests

Effect   -2 Log Likelihood of Reduced Model                        Chi-Square      df         Sig.


Intercept                                  897.977                                   .000                 0         

Employment                           902.361                                   15.202             12        .231

F’ Education                            905.055                                   17.896            12        .119

F’ Income                                910.300                                   21.141             12        .027*

Bad Association                       928.787                                   41.628              4          .000**

Place of residence                    908.648                                   21.489              4          .000**

Age                                          916.149                                   28.982              8          .000**

Types of Crime                       928.166                                   41.007                8          .000**

* p < 5% and ** p < 1%


    The Chi-Square statistic is the difference in -2 log-likelihood between the final model and a reduced model. The reduced model is formed by omitting an effect from the final model. The null hypothesis is that all parameters of that effect are 0. From the above table it is clear that except for the variables of ‘Employment’ and ‘Father’s Education’ the rest are statistically significant. However, as seen in table I these two variables do provide statistically significant results for some of the sub-categories of the dependent variable.



The derived parameters suggest a clear relationship between the levels of explanatory variables and those of the dependent variable. Table 1 provides the summary of variables, parameter estimates and their significance.


                                                  Table 1

Parameter Estimates

N = 422                                  B          Std. Error         Wald                df         Sig.      Exp(B)




Bad Association Intercept        -19.357            1.293   224.152           1          .000    

                        [EM=1]           17.801             .659     729.473           1          .000** 5.3E6

                        [EM=2]           17.132             .556     948.474           1          .000** 2.7E6

                        [EM=3]           16.742             .000                 .           1          .           1.8E6

                        [EM=4]           0                      .           .                       0          .           .

                        [FE=1]            1.650               .736     5.031               1          .025*   5.209

                        [FE=2]             -.299                .868     .119                 1          .731     .742

                        [FE=3]              .655                 .761     .740                 1          .390     1.925

                        [FE=4]                   0                      .           .                   0          .           .

                        [I=1]                -.728                .699     1.083               1          .298     .483

                        [I=2]                -.999                .661     2.282               1          .131     .368

                        [I=3]                -.215                .785     .075                 1          .784     .807

                        [I=4]                0                      .           .                       0          .           .

                        [CA=1]           3.018               .558     29.307             1          .000** 20.458

                        [CA=2]           0                      .           .                       0          .           .

                        [P=1]               .381                 .481     .626                 1          .429     1.464

                        [P=2]               0                      .           .                       0          .           .

                        [A=1.00]         1.549               .590     6.896               1          .009** 4.706

                        [A=2.00]         2.052               .601     11.647             1          .001** 7.780

                        [A=3.00]         0                      .           .                       0          .           .

                        [T=1.00]         -1.233              .959     1.655               1          .198     .291

                        [T=2.00]         -.923                .992     .867                 1          .352     .397

                        [T=3.00]         0                      .           .                       0          .           .

Gain                 Intercept          -1.134              1.275   .792                 1          .374    

                        [EM=1]           1.596               1.083   2.174               1          .140     4.935

                        [EM=2]           1.027               1.017   1.021               1          .312     2.794

                        [EM=3]           .768                 1.014   .573                 1          .449     2.155

                        [EM=4]           0                      .           .                       0          .           .

                        [FE=1]           1.017               .686     2.201               1          .138     2.765

                        [FE=2]            -.499                .809     .380                 1          .538     .607

                        [FE=3]            -7.506E-02      .743     .010                 1          .920     .928

                        [FE=4]                        0                      .           .                       0          .           .

                        [I=1]                -.466                .710     .431                 1          .511     .627

                        [I=2]                -.452                .654     .478                 1          .489     .636

                        [I=3]                .492                 .757     .422                 1          .516     1.636

                        [I=4]                0                      .           .                       0          .           .

                        [CA=1]           1.606               .457     12.350             1          .000** 4.985

                        [CA=2]           0                      .           .                       0          .           .

                        [P=1]               -1.141              .496     5.295               1          .021*   .319

                        [P=2]               0                      .           .                       0          .           .

                        [A=1.00]         .389                 .574     .459                 1          .498     1.475

                        [A=2.00]         1.670               .584     8.171               1          .004** 5.312

                        [A=3.00]         0                      .           .                       0          .           .

                        [T=1.00]         -1.489              .931     2.555               1          .110     .226

                        [T=2.00]         -.197                .957     .042                 1          .837     .821

                        [T=3.00]         0                      .           .                       0          .           .

Lack Control   Intercept          -.392                1.178   .111                 1          .739    

                        [EM=1]           .679                 .941     .520                 1          .471     1.971

                        [EM=2]           .580                 .832     .486                 1          .486     1.787

                        [EM=3]           .608                 .834     .531                 1          .466     1.836

                        [EM=4]           0                      .           .                       0          .           .

                        [FE=1]           1.363               .694     3.852               1          .050*   3.907

                        [FE=2]             .453                 .765     .350                 1          .554     1.573

                        [FE=3]             .698                 .706     .980                 1          .322     2.010

                        [FE=4]                        0                      .           .                       0          .           .

                        [I=1]                -1.987              .693     8.224               1          .004** .137

                        [I=2]                -2.015              .647     9.716               1          .002** .133

                        [I=3]                -.421                .725     .338                 1          .561     .656

                        [I=4]                0                      .           .                       0          .           .

                        [CA=1]           1.206               .443     7.413               1          .006** 3.341

                        [CA=2]           0                      .           .                       0          .           .

                        [P=1]               -1.101              .487     5.108               1          .024*   .332

                        [P=2]               0                      .           .                       0          .           .

                        [A=1.00]         -4.570E-02      .570     .006                 1          .936     .955

                        [A=2.00]         1.341               .567     5.587               1          .018*   3.824

                        [A=3.00]         0                      .           .                       0          .           .

                        [T=1.00]         -.252                .967     .068                 1          .795     .778

                        [T=2.00]         -3.289E-02      1.012   .001                 1          .974     .968

                        [T=3.00]         0                      .           .                       0          .           .

Dispute            Intercept          -.280                1.084   .067                 1          .797    

                        [EM=1]           1.189               .767     2.404               1          .121     3.283

                        [EM=2]           .302                 .629     .231                 1          .631     1.353

                        [EM=3]           9.934E-02       .675     .022                 1          .883     1.104

                        [EM=4]           0                      .           .                       0          .           .

                        [FE=1]             .907                 .614     2.183               1          .140     2.476

                        [FE=2]             .610                 .650     .880                 1          .348     1.841

                        [FE=3]             .485                 .641     .574                 1          .449     1.625

                        [FE=4]                        0                      .           .                       0          .           .

                        [I=1]                -.342                .604     .321                 1          .571     .710

                        [I=2]                -.938                .567     2.735               1          .098     .391

                        [I=3]                -6.548E-03      .689     .000                 1          .992     .993

                        [I=4]                0                      .           .                       0          .           .

                        [CA=1]           .911                 .379     5.790               1          .016*   2.486

                        [CA=2]           0                      .           .                       0          .           .

                        [P=1]               3.969E-02       .384     .011                 1          .918     1.040

                        [P=2]               0          .           .           0                      .           .

                        [A=1.00]         1.155E-03       .500     .000                 1          .998     1.001

                        [A=2.00]         1.603               .506     10.024             1          .002*   4.970

                        [A=3.00]         0                      .           .                       0          .           .

                        [T=1.00]         3.132E-02       .911     .001                 1          .973     1.032

                        [T=2.00]         -.846                .966     .767                 1          .381     .429

                        [T=3.00]         0                      .           .                       0          .           .

Base: ‘Dragged’           

 * p < 5% and ** p < 1%


    Considering the multinomial Logit model involves categorical data analysis, the results have to be interpreted in terms of probability values. The table uses the last category of each variable as the base for comparison purposes. For instance, in the case of the dependent variable ‘Reasons for committing the crime’ the category ‘Dragged’ is used as the base while for the explanatory variables, the highest level is the base with which other levels are compared. We have highlighted the statistically significant results in the table and we will interpret the results for these cases only.

    For the category of Bad Association, level 1 of father’s education [that he is illiterate] is a significant result in comparison to level 4 [that he is college educated]. This implies that the log odds of a person committing crime due to a bad associate in comparison to being dragged into criminal behavior is greater by a value of 1.65 for the case when the father is illiterate than when he is college educated. This suggests that a more educated father is perhaps able to control his son and prevent him falling into a life of crime even if he has some relatives or friends who are criminals. The other significant results for this category are that of employment, bad association and age. Clearly, being unemployed or engaged only in manual, low paying jobs entices people to commit crimes much more in comparison to those having some kind of white collar job [the base category]. It makes sense to also observe that those having bad association are more likely to get involved in criminal behavior. The age factor also suggests that bad association is more important for those between the ages of 22-30 than the younger ones. Perhaps with growing age a person learns to develop more lasting associations.

    For the reason of ‘Gain’ living in urban areas show significant results. It appears that growing urbanization creates opportunities for crimes that are intensified with criminal association especially for those in the age group of 22-30 years. Most young people of this age are forced to seek jobs in urban areas in view of declining economic opportunities in rural areas. These young men are generally by themselves and face declining family supervision. It is likely that they get involved with criminal associates who induce them to seek illegal gains. All these factors show significant odds for the reason of ‘Gain’ for committing crimes.

    The reason ‘Lack Control’ has slight significant result for father’s low educational status [p-value 0.050]. This seems to make sense since an illiterate father [level 1] is unlikely to understand the activities of his children. This lack of supervision may encourage the children to experiment and give vent to their desires. Experimentation with alcohol, drugs and sexual pursuits are some of the more specific reasons within this category and it seems to fit in with the notion that in a family, where the father is unable to provide discipline, children are unlikely to control themselves and will fall into criminal ways. In addition, the significant results of low income [levels 1 & 2] and bad association both point towards the situation where social control is less, the family is in dire circumstances and criminal friends are inducements towards illegal and ill-advised experimentation. That the results are significant for young people between the age group of 22-30, living in urban area as compared to rural regions is again in conformity to the expectations.

    Finally, the reason of ‘dispute’ has largely to do with conflicts of different kinds. The respondents have described these as arising from family problems, land ownership, caste and clan feuds and enmity with some people, especially due to gang affiliations. The significant results for the variables of young age [level 2], bad association and (slightly) for family income [level 2, p-value <1%] appear to draw the respondents into this situation. Disputes are common in any situation, but young age and bad association can get a young person into serious conflicts.

Discussion and Conclusion

    Based upon the analysis of the data as described above we are now in a position to empirically comment upon the objectives of the research. Our first hypothesis was that large economically deprived families where parents do not have sufficient education create circumstances that encourage deviant behavior. Our second hypothesis was that neglected childhood is a major factor in falling into criminal behavior. This is similar to the assertion that criminal association generally develops for those growing up in low socio-economic status families. Although, the analysis suggests that family size is not a significant factor but income is undoubtedly a major factor in criminal behavior. Since income is expected to be related to the attainment of higher education it follows that the underlying cause of criminal behavior of the respondents is definitely linked to their poor educational status. Considering that 283 respondents [67%] had at most, only a few years of schooling, suggests that it pays for the Indian society to educate its children since it provides marketable skills and helps wean them away from a life of crime.

    We could also explore these assertions by examining the educational standards of the father of the respondents. Our belief is that uneducated guardians are unlikely to discipline and supervise their children especially if their economic conditions are desperate too. The poverty and absence of any learning initiative would have caused neglect in the upbringing of the children and making them easy prey to delinquent behavior. This hypothesis appears supported by the fact that the percentage whose parents had at most primary education and whose income was less than Rs. 1000 per month form 16.8% of the sample. In comparison, the group of respondents whose parents were well educated [college level] and had high family income [> Rs. 5000 per month] were only 11, a mere 2.2% of the sample. It is significant that of these 11 respondents, 9 were involved in body type crimes [6 murders] and 5 of them cited dispute or being dragged in the case as their reason for criminal behavior. Their criminality appears to be related to old standing disputes and questions of family honor. This suggests that better-educated parents are more likely to supervise and prevent their children from getting involved in serious property crimes though not in family related feuds.

    Our next set of hypotheses suggested the failure of the criminal justice system in pushing the people into crime. The hypotheses suggest that criminal association leads to labeling effect and the failure of the justice system to provide speedy justice makes it difficult to get out of criminal behavior and perhaps forces people to take to criminal acts. The survey amply supports these hypotheses since a very large number of respondents have complained about being ‘dragged’ into the case. Most of these respondents refused to consider themselves as criminal so perhaps the labeling effect had not taken permanent roots but they were vociferous in their assertions that they had not committed the crimes for which they were charged. Moreover, the fact that ‘dispute’ was cited as a major reason seems to suggest that failures of the justice system in resolving conflicts pushed many to indulge in criminal acts. In our discussions the respondents argued that they had to commit the crime [attack or even kill their opponent] or otherwise they [or their family] were themselves at risk of harm. These conflicts appear to have been going on for many years and were related to land, property, rivalries and other kinds of mutual antagonism. The failure of the system to quickly resolve them within the institutional framework appears to have pushed many to find their own solutions with disastrous effects.

    It is also noteworthy that only 36% [154 out of 422] reported not having any criminal association. The large majority reported links with family and friends who were engaged in criminal activities and this clearly affected the respondents. Many stated that they were dragged into the case because of their links with these criminal associates. Accordingly, having association with those who act criminally is also a strong reason for criminal behavior.

    Our next hypothesis is suggestive of the weakening of the traditional bonds that hold a person to the norms of the society. The rapid socio-economic and demographic changes that are occurring in the country, especially the culture of consumerism to us appeared to be an important factor in tempting people to act illegally. The artificial desires created by advertisements and the inability of a large number of people to satisfy these desires seemed to drive many towards illegal acquisition of property. The statements about desires for material goods and lack of family supervision suggest to us some of the ill effects of India’s economic policies. The ubiquitous advertisements for a consumerist life style, loosening of bonds due to a developing economy appear to have an adverse effect on young impressionable people. Many respondents, especially younger ones expressed clear desires for quick gains. The fact that many were living away from immediate family and developed association with those already engaged in criminal activities seemed to support this conclusion. The desire to acquire money, property or avail an opportunity was frequently mentioned as one the reasons for their crimes. Even in cases where criminal association or being dragged into a case was given as the first reason, the respondents did agree that lure of money was still one of the underlying reasons. The ill effects of alcohol, drugs and sexual drives, especially in combination appear to have been quite devastating. Not surprisingly, in our sample the percentage of respondents whose primary reasons were ‘gain’ or ‘lack-of-control’ was seen to be around 40%.

    Our final observation from this study has been that severe socio-economic circumstances and failures of the justice system to control social conflicts are important causes that stand beneath a large proportion of deviant behaviors in the country. The acute poverty and the lack of education not only deny entry into legitimate pursuits but it also implies a lack of supervision and discipline during the formative years. Most of the respondents mentioned anger, jealousy, feeling of revenge and enmity as the driving force in their criminal activities. The lack of any institutional system that could deal with these at the initial stage and resolve conflicts appear to be responsible for driving many towards adopting illegal actions. Most of the respondents also displayed extreme anger towards the police and justice officials. A majority did not even contemplate going to the justice officials to seek redress. It seemed clear, and many indeed stated in no uncertain terms that after coming out of the prison they were again going back to a life of crime since their enemies would not let them live otherwise.

    The fact that an over-whelming majority of the respondents came from poor, economically disadvantaged background also seems to support the perception that the Indian Criminal Justice System targets only those at the bottom rung of the ladder. The fact that there were only 5 respondents with a post graduate degree, only 15 who had well established jobs appear to confirm that the vast number of people being processed by the police and judiciary come from low socio-economic background. Obviously, criminal behavior is not an innate trait of that class of people who find themselves at the bottom. Only 17 respondents [4%] mentioned poverty as one of their reasons for committing the crimes. Clearly, despite the fact that vast majorities of Indian people are poor, they are not contemplating crime as a means of sustenance. Most do believe that this is undesirable and should not be done. Yet, if the jails are filled with people from low socio-economic strata, then it reflects more about official practices than reasons for breaking the law. The Indian authorities need to pay attention to these findings and examine the actions of the police and criminal justice agencies more carefully. 



Agresti, A. (1990). Categorical Data Analysis. New York: John Wiley & Sons.

Baldwin, J., & Bottoms, A. E. (1976). The Urban Criminal. London: Tavistock.

Bhaduri, T. C. (1972). Chambal: The Valley of Terror. Delhi Vikas Pub. House.

Brantingham, P. J., & Brantingham, P. L. (1991). Environmental Criminology. Prospect Heights, IL: Waveland Press.

Census of India. (2001).

Cloward, R. A., & Ohlin, L. E. (1960). Delinquency and Opportunity: A Theory of Delinquent Gangs. New York: Free Press.

Cohen, A. K. (1955). Delinquent Gangs: The Culture of the Gang. Glencoe Il. Free Press.

Crutchfield, R. D., & Pittchford, S. (1997). Work and Crime: The Effects of Labor Stratification. Social Forces, 76, 93-118.

Crutchfield, R. D., Bridges, G. S., Weis, J. G., & Kubrin, C. (Eds.), (2000). Crime Readings. Thousands Oaks, Ca.: Pine Fore Press.

Durkheim, E. (1951). Suicide. Translation by John A. Spaulding and George Simpson edited by George Simpson. New York: The Free Press.

Fetchenhauer, D., & Simon, J. (1998). An experimental test of Gottfredson and Hirschi's ‘general theory of crime’. Monatsschrift fuer Kriminologie und Strafrechtsreform, 81(5), 301-315.

Hall, J. (1960). General Principles of Criminal Law (2nd Ed.) Indianapolis: Bobbs Merrill.

Hartjen, C. A., & Kethineni, S. (1996). Comparative Delinquency: India and the United States New York: Garland Publishing Inc.

Hirschi, T., & Gottfredson, M. (2000). Age and the Explanation of Crime. In R. D. Crutchfield., G. S. Bridges., J. G. Weis., & C. Kubrin. (Eds.), Crime Readings (pp. 138-141). Thousands Oaks, Ca.: Pine Fore Press.

_______. (1969). Causes of Delinquency. Berkeley: University of California Press.

Indermaur, D. (1999). Situational prevention of violent crime: theory and practice in Australia. Studies on Crime and Crime Prevention, 8(1), 71-87.

Jatar, D. P. (1980). The Problem of Dacoity in Bundelkhand and the Chambal Valley. New Delhi  S. Chand.

Jenkins, P. H. (1997). School Delinquency and the School Social Bond. Journal of Research in Crime and Delinquency, 34(3), 337-367.

Katare, S. S. (1972). Patterns of Dacoity in India: A Case Study of Madhya Pradesh New Delhi: S. Chand.

Khan, M. Z. (1981). Dacoity in Chambal Valley. New Delhi: National Publishing House.

Lobo, J. (1992). Leaves From a Policeman’s Diary. New Delhi: Allied Publishers.

Madhok, S. (1998). Gender bias, infanticide bane of Indian Women. The Tribune, June 19. Retrieved on 28th May 2008 from

Matza, D. (1969). Becoming Deviant. Englewoods Cliffs, NJ: Prentice Hall.

Merton, R. K. (1968). Social Theory and Social Structure. Gleencoe, Illinois: The Free Press.

Mohanty, S. (1990). Crimes and Criminals in India: A Socio-Economic Survey. New Delhi: Ashish Publishing House.

Nayar, B. R. (1975). Violence and Crime in India: A Quantitative Study. Delhi: Macmillan.

Paternoster, R., & Brame, R. (1998). The structural similarity of processes generating criminal and analogous behaviors. Criminology, 36(3), 633-669.

Phulia, S. S. (1992). Criminology: A Psychological and Anthropological Analysis. Delhi: Friends Publications.

Robin, L. N. (1966). Deviant Children grown Up: A Sociological and Psychiatric study of Sociopathic Personality. Baltimore: Williams & Wilkins.

Sarvate, S. (2000). Patriarchy and Landlordism. Asian Week, February 24, 21(26). Retrieved on 28th May 2008 from

Sellin, T. (1938). Culture, Conflict and Crime. New York: Social Science Research Council.

Shah, G. (1992). The Indian Police: A Retrospect. Bombay: Himalayan Publishing House.

Shamashastry, R.  (1929). Kautilya’s Arthashastra (Tr.)  Mysore: Wesleyn Press.

Shaw, C., & McKay, H. D. (1942). Delinquency and Urban Areas. Chicago: University of Chicago Press.

Shukla, K. S. (1982). Juvenile Delinquency in India: Research Trends and Priorities. Indian Journal of Criminology and Criminalistics, 2, 103-13.

Simhadri, Y. C. (1991). Denotified Tribes: A Sociological Analysis. Delhi: Classical Publishing House.

Somerville, A. (2000). Crime and Religious Beliefs in India. London: Laurier Books.

Sukla, S. (1987). Pracina Bharata mem Aparadha aura Danda [In Hindi]. Kanapur: Praja Prakasana.

Sutherland, E., Cressey, D. R., & Luckenbill, D. F. (1992). Principles of Criminology (11th Ed.) New York: General Hall Inc.

Thakur, B. (1980). Urban Settlements in Eastern India: Entropy Changes and Pattern Analysis. New Delhi: Concept.

Tihar Prison. (2001). Prisoner’s Profile. Retrieved on 28th May 2008 from

Thilagaraj, R. (1987). Goal Perceptions of Delinquents and Non-Delinquents: A Comparative Study. Indian Journal of Criminology, 2, 145-48.

United Nations Development Fund. (2008). Human Development Report 2008. Retrieved on 28th May 2008 from

Upadhyaya, V. (1978). A Study of Hindu Criminology. Delhi: Chaukhambha Orientalia.

Verma, A. (2000). Lies, Damn Lies and Police Statistics. Indian Police Journal, XLVI (2-3), 29-36.

________. (2005). Indian Police: A Critical Evaluation. Delhi: Regency Publications.

Voss, H. L., & Petersen, D. M. (1971). Ecology, Crime and Delinquency. New York: Appleton.

Weinstein, J. A. (1974). Madras: An Analysis of Urban Ecological Structure in India. Beverly Hills, Calif.: Sage Publications.

Wikstrom, P-O. H. (1991). Urban Crime, Criminals, and Victims: The Swedish Experience in an Anglo-American Comparative Perspective. New York: Springer-Verlag.

Yang, A. A. (Ed.), (1985). Crime and Criminality in British India. Tucson, Arizona: University of Arizona Press.

[1] This research project was supported by a grant from Indian Council of Social Science Research (ICSSR), New Delhi. 

[2] Associate Professor, Department of Criminal Justice, Indiana University, Bloomington, IN, USA. Email: (Corresponding Author).

[3] Research Associate, New Delhi, India.

[i] We have collected data on female offenders also but it is a small sample and its analysis is being done separately.


INternational Journal of Criminal Justice Sciences

Vol 3 Issue 2 july- december 2008

2008.Creative Commons BY-NC-SA International Journal of Criminal Justice Sciences unless otherwise noted.

© 2008  International Journal of Criminal Justice Sciences. All rights reserved

Web Journal created, published and maintained by Dr. K. Jaishankar Last updated 02/10/2009