بازدارنده های رسمی و غیر رسمی جرم و جنایت در ژاپن : نقش پلیس و بازنگری سرمایه اجتماعی

Using Japanese prefecture-level panel data, this paper examines how the crime rate is affected by formal and informal deterrents as reflected by police and social capital, respectively. Both, however, suffer from the endogeneity problem and therefore the estimation results are biased when regression analysis is conducted. Hence, the fixed effects 2SLS method is employed to control for the endogeneity bias as well as for unobservable fixed effects. As well, the relationship between inequality and crime is examined. The main findings are: (1) police and social capital reduce the crime rate and their effects increase when the endogeneity bias is controlled for through fixed effects 2SLS estimation. (2) The effects of social capital, which is smaller than that of police, is however is reinforced by police through the complementary relationship existing between them.

Since the seminal work of Becker (1968) provided the first theory of crime based upon rational behavior, many works have been devoted to the study of the economics of criminal behavior (e.g., Ehrlich, 1973, Fajnzylber et al., 2002a, Glaeser and Sacerdote, 1999, Levitt, 1997 and Carneiro et al., 2005). The central tenet of Becker's theory is that crime will decline if police presence rises. This prediction, however, is not consistent with the findings of the ensuing empirical studies that tested it (e.g., Cameron, 1988 and Eck and Maguire, 2000). As recent works point out, failing to find empirical support of Becker's prediction is in large part due to a serious endogeneity problem stemming from simultaneous determination of crime and police presence (Levitt, 1997, Di Tella and Schargrodsky, 2004 and Kelly, 2000). Hence, researchers are essentially required to control for this endogeneity problem to present unbiased estimation results. The influence of the attitude and conduct of others on a person's behavior seems apparent among neighbors and colleagues in schools and workplaces, and such interactive mechanism above also applies to a criminal behavior (Funk, 2005a, Funk, 2005b and Glaeser et al., 1996). The cost of committing a crime is not only attributable to formal sanctions applied by police but also informal sanctions such as stigma probably reinforced by a strongly bonded community (Funk, 2005b and Rasmusen, 1996). That is, borrowing an expression of Coleman (1990, p. 557), “the external sanctions by which persons’ actions are made responsible range from informal ones, such as expressions of ostracism, to legal sanctions for breaking laws or regulations”. Social science researchers have shed new light on social capital (e.g., Coleman, 1990 and Putnam, 1993), thereby provoking discussions concerning economic development and growth (e.g., Bloch et al., 2007, Guiso et al., 2004, Hall and Jones, 1999 and Knack and Keefer, 1997). Social capital appears to play a critical role in economic development through various channels such as deterring criminal behavior; however, surprisingly few attempts have so far been made to investigate the association between crime and social capital. Lederman et al. (2002) made the first attempt to examine the influence of social capital on the incidence of crime using cross-country data. As pointed out by Lederman et al. (2002), estimation results might suffer from an endogeneity bias when the effects of social capital on crime are examined through regression estimations. To control for this bias, they employed 2SLS estimation and, as a result finds no clear link between them. On reason is probably the omitted variable biases arising from a scarcity of data and unobservable country specific effects.1,2 With the aim of not only controlling for endogeneity bias but also omitted variable bias, panel data containing various important variables such as the presence of police, inequalities, and education level should be used when employing fixed effect 2SLS estimation. According to Benedict (1946), Japanese culture is characterized as a ‘culture of shame’ in contrast to the ‘culture of sin’ in the west. While the sin that Westerners feel they should avoid is identified by a person as he contrasts his deeds with the commandments of an absolute being, shame is felt by a person when his behavior is seen and gossiped about by other people. This implies that Japanese people are more likely to be affected by neighbors. The culture of shame, a feature of Japan, is relevant to the psychological cost of committing a crime through local interaction (Funk, 2005a and Funk, 2005b). Hence, compared with the Western situation, it seems more appropriate, as well as interesting to examine the effects of social capital on crime in Japan. In this research, I consider the police presence a formal deterrence to crime, whereas social capital is regarded as an informal deterrence.3 In the field of law and economics, the association between formal and informal agreements has already drawn much attention.4 Although it is essential to consider them both when factors determinant of crime are ascertained, little attention, with the exception of theoretical work of Funk, 2005a and Funk, 2005b, has been given to capturing them at the same time. Further, whether the relationship between formal and informal deterrents is a substitute or complementary one is profoundly relevant to this topic from the standpoint of institutional analysis (Hayami, 2001 and Hayami and Aoki, 1998). The circumstance of Japan characterized by a ‘culture of shame’ is thought to generate more sensitive responses to informal punishment and stigma, making it interesting to examine criminal behavior in light of this.5 The aim of this report is to examine not only the effects of formal and informal deterrents, such as police presence and social capital, but also their cross-product effect upon criminal behavior after controlling for the endogeneity bias as well as unobservable individual fixed effects. To this end, I use the comprehensively complied panel from 46/47 Japanese prefectures for the years 1994–2001 to conduct fixed effect 2SLS estimations. The major findings provided through estimations conducted in this paper make it evident that police and social capital pronouncedly reduce crime, and, furthermore, that police and social capital are complementary influences reducing crime. The organization of this paper is as follows: Section 2 surveys features of Japanese society and social capital in Japan. Section 3 explores the simple relationships between crime rate and various key variables, and presents a simple econometric framework. Section 4 discusses the results of the estimations. The final section offers concluding observations.

Human society is controlled by both formal rules as law and by informal ones as norms. Under such conditions, the presence of police and the existence of social capital are expected to play critical roles in deterring people from committing crimes, thereby promoting economic development. If this is the case, the question arises about whether the relationship between the police presence and social capital is a complementary or a substitute one? This question, which has hitherto not been empirically investigated, is important to ascertain the association between formal and informal institutions. This paper focuses on this question, and conducts an empirical examination. Using Japanese prefecture-level panel data, I have examined how the crime rate is affected by the presence of police and the existence of social capital; considered to reflect formal and informal deterrents, respectively. Both however, suffer from the endogeneity problem and therefore the estimation results are biased when regression analysis is conducted. I employed the fixed effects 2SLS method to control for the endogeneity bias as well as for unobservable fixed effects. What is more, I examined the relationship between inequality and crime. The key findings are as follows: (1) Police presence and social capital reduce crime rates and their effects become larger when the endogeneity bias is controlled for. (2) The relationship between police presence and social capital is complementary in their reduction of the crime rate. This empirical study provides evidence that the informal deterrent sustained by a tightly bonded community is complementary to the formal deterrent provided by the police so that the informal deterrent makes the formal deterrence more effective. This is in part presumably thanks to characteristics of Japan where society is homogenous and local communities highly value interdependency among their members, which through local interactions is inevitably relevant to the psychological cost of committing a crime. Therefore, compared with that in Western countries, informal deterrence is effective in Japan. This research is based upon data aggregated at the prefectural level and therefore individual rational behavior at a micro level is not explored. Disaggregated micro level data would allow an investigation of individual criminal behavior in response to the features of surrounding people and the feature of the group to which that individual belongs. This would better clarify how a rational decision regarding crime is affected by society-wide and group-specific features. In addition to the link between social capital and police presence considered in this paper, the association between social capital and inequality should also be taken into account when the likelihood that income redistribution through social capital to reduce crimes is considered. It will be worthwhile to explore whether social capital decreases the impact of inequality on crime so that economic policies concerning deterrence of crime can be presented. These are major issues remaining to be addressed in future research.