keywords

1. Introduction
Privatization of policing is not a new issue in criminal justice field, particularly in Western countries, such as England and the U.S. Based on the notion of self-help, private policing business traditionally has been widely accepted and practiced in private sector of civil and criminal justice systems in those countries (Becker, 1974). Even further, some argued that there would be no police tasks which could not be privatized (Chaiken and Chaiken, 1987). As results, private investigation business is well established in those countries (Gill and Hart, 1996). For example, a magazine in the field of private investigation claims that there are approximately 60,000 private investigators working in the U.S. (www.pimagazine.com). Also, by recent year, there are at least nine different national and regional investigator organizations in the U.S. (www.theprivateinvestigtordirectory.com).

In South Korea, private policing has been rapidly growing. More specifically, Lee (2004) showed that the growth rates of private policing were 55% from 1980 to 1985, 194% from 1985 to 1990, and 104% from 1995 to 2000 (Lee, 2004). According to Lee (2004), economic affluence increased availability of private policing. During the political development influenced by westernization, South Korean society gradually adopted litigation-based legal culture and decentralization of police authority to private sectors (Lee, 2004). He anticipated that needs of private policing would be even greater in a forthcoming new decade. Although his study did not differentiate private investigation (hereafter referred as PI) from private policing, it is certain that PI business has also been grown. Research on private investigation, then, is unavoidable prerequisite for structuring sustainable PI system for the future, particularly in South Korea, where PI business is at its early stage. Moreover, understanding factors potential clients may consider when they need to hire private investigators would be beneficial for PI business as well as public police.

Prior research on PI, however, has been limited.1 Among about 70 articles and books on PI published between 1971 and 2010, majority of the studies focused on historical development of PI (e.g., Gill and Hart, 1996, 1997a,c; Johnston, 1992; Kakalik and Wildhorn, 1971, 1972; Lee, 2004; Marling, 1995), philosophical and legal regulation issues (Becker, 1974; Gill and Hart, 1997c; O’Connor et al., 2008; Prenzler and King, 2002; Shearing, 1992), job specification (Prenzler and King, 2002), ways of increasing cooperation between police and PI (e.g., An, 2007; Choi, 2006; Cunningham and Taylor, 1985; LEAA, 1977; Shearing, 1992), and functions of PI in a society (e.g., Alarid and Montemayor, 2010; Reichman, 1987; Reiss, 1988; Shearing and Stenning, 1983). A few empirical research focused on attitudes of police toward private investigators (Kakalik and Wildhorn, 1971; Scott and McPherson, 1971; Shearing et al., 1985), and some examined private investigators’ perception toward police cooperation (Gill and Hart, 1997a; Kakalik and Wildhorn, 1971). None of prior studies, however, has examined citizens’ attitudes toward and preference for PI as an alternative to public policing. As noted by Gill and Hart (1999), demand of citizens for circumnavigating formal criminal justice systems of due process to achieve their personal goals is the most distinct matter regarding private investigation. Unfortunately, however, there is lack of empirical examination on how and why citizens choose to hire PI for resolving their civil or criminal matters.

Second, prior studies of private investigation are mostly outdated. They had been dominantly carried out between 1970s and 1990s. Only few research has been carried out during and after 2000s (e.g., Prenzler and King, 2002; Prenzler and Sarre, 2008), and all of those studies were conducted in Western countries. As noted (Lee, 2004), the growth of private policing business in recent era requires adequate understanding on political, legal, and economical circumstances depending on different cultural and social settings. Thus, a study of private investigation in South Korea setting will provide important information on how criminal justice system related to private policing should be arranged and operated (Lee, 2011).

To fill the void of literature, the current study of PI in South Korea investigate citizens’ characteristics, such as demographics, desire for personalized justice, their attitudes toward PI’s investigation skills, attitudes toward PI’s morality and confidentiality during investigation, fear of crime, and their attitudes toward public police. These factors are derived from relevant literature in various different countries, such as Australia, Canada, England, U.S., and South Korea. This study examines effects of the above mentioned factors on citizen’s preference for hiring PI in 12 criminal and civil cases. The cases include hit-and-run traffic incidents, violation of copy rights, criminal incidents in one’s own company/store, auto theft, fraud, defamation of character, incident where art works or jewelry were stolen, family members/friends are missing, hiring a new employee, background check for marriage/business contracts, fidelity case, and privacy invasion case.

2. Literature review

2.1. Private investigation in South Korea

Private investigation business has a long history in South Korea, but it was 1960s when the business was termed and recognized by a law (Heungshinso Act) as heungshinso activites, which means a company doing errand jobs. Heungshinso carried out many types of illegal activities, such as collecting debts and private information, and even using violence to investigate and solve personal civil cases (Na, 2007). During 1970s, the term heungshinso was abandoned and people began to use shimboorum centers or ‘errand centers’ (or centers for errand jobs) for the same types of activities to avoid the negative reputation previously attached to the business. In 1977, the Korean government discarded the Heungshinso Act, and enacted the Credit Investigation Act (Act 3039). This act banned private investigation activities, such as investigating and resolving civil cases (Na, 2007). As a result, private investigation business is currently illegal in South Korea.

Instead, errand business is legal. Initiation of the business is based on a reporting system. By 2005, 500 centers are registered to the governmental tax office, but unregistered centers are estimated to be 3000. Some of errand tasks are legalized. The following Table 1 shows what tasks are legal for shimboorum centers. Left side of the table shows tasks are currently banned, but considered to be legalized for PI in South Korea. Right side of the table shows types of tasks PI can carry out currently. Since establishing and opening shimboorum centers is legal but their works are not directly overseen by the police, the Korean National Police (KNP) does not have accurate statistics on their work activities. Even the KNP does not have rough estimate of number of all shimboorum centers. So it is hard to tell the volume of legal and illegal business activities conducted by the centers. But, it is known by a study surveyed shimboorum centers that the most clients of shimboorum centers are individual citizens (An, 2007).

To prevent illegal private investigation activities, some attempt to legalize private investigation in South Korea. Since legalization of PI will bring illegal activities into the umbrella of governmental control, it will have significant implications for the criminal justice system in South Korea, such as acquiring governmental control over illegal investigation activity and protecting civil rights (An, 2007). In addition, legalization of PI is expected to enhance efficiency and effectiveness of public policing as far as police actively utilize PI in terms of fraud investigation, product piracy investigations, and fidelity investigation (Kim, 2007; Lee, 2011, 2006; Lee and Lee, 2007). Therefore, in order to prepare the legalization, it is important task to carry out an empirical research to find out what factors affect citizen’s preference to hire private investigators to resolve their criminal and/or civil cases. Since PIs are tend to have more recent sophisticated technological skills (Gill et al., 1996) and higher success rates (Prenzler, 2001) and they solely work for clienteles , there is a strong possibility that citizens hire PIs instead of reporting their victimizations to the police for instance. Also, a study of citizen’s preference to hire PIs will guide legalization process in South Korea. Since there is no empirical study neither in Western countries nor in Eastern countries, empirical evidence derived from the current study will guide establishment of PI related public policy in countries, such as South Korea and Japan.

2.2. Individual characteristics of citizen

2.2.1. Demographics
Little studies investigated citizens’ demographic differences in terms of satisfaction with private security services. For example, Choi (2009) carried out a path analysis of customer satisfaction with private security services in South Korea. According to the findings of the study, female citizens who were married and employed were more likely to hire private security for their safety needs. Also, the study showed that age, educational and income levels were positively related to likelihood of hiring private security. Since there is no empirical study regarding the demographics of citizens (i.e., clients) hiring private investigators, the current study relies on the abovementioned findings from private security studies. Applying these findings to PI, this study hypothesizes that those who are older, educated, married with higher income are more likely to hire PI for resolving their civil or criminal matters.

2.2.2. Desire for personalized justice It is not a recent trend that police works are privatized. Becker (1974) argued that private policing is based on the idea of self-help and it relies on personal contracts with agencies that provide police services. The act of self-help or self-defense is viewed as an extension of individual’s responsibility to protect his property (see also, Kakalik and Wildhorn, 1972), and it reflects one’s desire for fulfilling his responsibility. Thus, PI should be considered as an asset for protecting individuals’ property (Shearing, 1992). Particularly, within adversarial legal system, hiring PI should be benevolent to defendants, because typically prosecutors have more options than defense lawyers do, e.g., offering reward for witnesses, compelling testimony, initiating plea bargaining, and keeping access to more document (Frank and Schwartz, 1996). By hiring PI, defendants in adversarial systems can achieve equal or at least competing control over information. So as to win the legal competition and to fulfill their desire to resolve cases, PI can be considered for citizens as an important asset to utilize. In this sense, ones who want to resolve legal issues in their favor, they will be more likely to hire PIs.

2.3. Citizens’ attitudes toward PI
Prior studies typically focused on examining police officers’ attitude toward PI, such as PI’s business intention (e.g., “goodwill”, Gill and Hart, 1997a, p. 554), investigation skills (Scott and McPherson, 1971), knowledge and behaviors (Gill and Hart, 1997b), or PI’s attitudes toward themselves, such as morality and confidentiality (Gill et al., 1996). The current study utilizes findings from those prior studies, and applies them to examination of citizen’s attitudes toward PI. The following aspects are included in this study; citizens’ attitudes toward PI’s investigation skills, PI’s morality and confidentiality.

2.3.1. Attitudes toward PI’s investigation and information skills
criminal cases, but they can choose PIs to investigate their personal matters. Particularly, they can select PIs depending on PI’s investigation and information skills. Gill and Hart (1997b) found that PI companies with bigger size and more specialized business or commercial specialists, compared to small business type PI companies, made greater income. Although this finding does not directly imply a relationship between PI’s investigation and information skills and citizens’ preference, the fact that PI companies which are more organized tend to make more income suggests that professionalized PI companies are more likely to be chosen by clients. Furthermore, Gill et al. (1996) found that citizens hired PIs due to their investigation skills and equipment that the public police might not have. These findings imply that citizen’s attitude toward PI’s investigation and information skills can influence their tendency to hire PIs.

2.3.2. Attitudes toward PI’s morality and confidentiality
Police officers’ attitudes toward PI’s morality had significant influence on cooperative tendency among the officers with PIs. In 1971, a nationwide study on cooperation between the public police and PIs found that the cooperation was influenced by officers’ negative attitudes toward PI’s quality, morality and financial intention (Kakalik and Wildhorn, 1971). Similar results were found in another study. Scott and McPherson (1971)found that most of the police officers in Minneapolis and St. Paul areas superficially showed positive attitudes toward PIs, but they possessed negative feelings for cooperation with PIs. They found that reasons for the negative tendency in cooperation included immoral and financially intended PIs business. Unfortunately, there has been no empirical study on citizens’ attitudes toward PI’s morality regarding their tendency to hire PIs. However, above mentioned findings suggest that if citizens possess negative attitudes on PI’s morality, they would not tend to hire PIs for their cases. In addition, citizens’ attitudes toward PI’s 192 C.-H. Lee et al. / International Journal of Law, Crime and Justice 42 (2014) 188e202 business confidentiality can increase their hiring tendency. Gill et al. (1996) found that people believed PIs more tightly protected personal information than the police do. Based on above mentioned findings, it can be assumed that positive attitudes toward PI’s morality and confidentiality can increase the probability of hiring PIs to find cases resolution.

2.4. Citizens’ fear of crime and satisfaction with police

2.4.1. Fear of crime
Consistent with the idea of self-help, citizens’ fear of crime can increase the tendency to hire PIs for their issues. In other words, as citizens who want to protect their property are more likely to hire PIs to resolve their cases, those who fear crime more will tend to hire PIs to have them investigate their cases. This is because the fact that they cannot manipulate public police works, but they can somehow influence investigation carried out by PIs on their behalf (Gill and Hart, 1999). Furthermore, it seems that citizens who fear crime victimization possess negative attitudes toward efficacy of police works (Dowler, 2003). According to Choi (2009), citizens who fear crime more tend to value service provided by private security. Thus, fear of crime stemming from the negative feeling about police work can influence the likelihood of hiring PIs for investigation of their cases.

2.4.2. Satisfaction with police
It seems that citizens’ satisfaction with policing can significantly affect the likelihood of hiring PIs. Especially, demands for PI can increase dramatically as the public cry more for privacy, personal freedom and security. This means that to satisfy, or at least fulfill, their needs, citizens who are not satisfied with police work will likely hire PIs to have them investigate their criminal or civil cases. As Gill and Hart (1999, p. 247) stated, hiring PIs will allow citizens to “circumnavigate formal systems of due process and ‘state justice’ that pervade private investigators’ activities.” Particularly, in adversarial criminal justice system in the U.S., defendants who need to compete with prosecutors who have more access to case information can utilize PIs as supports for them to correct imbalance in the system (Frank and Schwartz, 1996). Thus, it is logical to assume that citizens who are not satisfied with police works are more likely to hire PIs to have them investigate their cases in their favor 

3. Methodology

3.1. Sampling
This study utilized data collected for a grant research supported by the Police Science Institute, which is a governmental research institute in South Korea. The original research was carried out between July and November 2010. Two surveys were conducted; one for collecting data about citizens’ attitudes toward private investigation and another about police officers’ attitudes toward PI. The current study used the citizen data to analyze and identify factors affecting citizens’ preference to hiring PI. Three hundred citizens (sample elements) were selected through quota sampling design based on the probability proportionate to size, meaning that quota of the sample were determined in accord to the population sizes in three different geographical areas including Seoul (the capital city, 140 citizens), Incheon (80 citizens), and other satellite cities around Seoul (80 citizens). Fifteen data collectors were trained about anonymity and voluntarism in the survey and they visited houses in three research sites to conduct surveys and interviews. Among the 300 households, 225 citizens responded to the survey (response rate ¼ 75%).

3.2. Measurements
Survey items were developed based on the prior studies carried out in Western countries and based on field tests in the following steps. First, the grant study drew survey and interview items from the prior studies, and translated them into Korean. Two bilingual researchers backtranslated the Korean items into English to check the validity of the translation. On June 25, 2010, the study carried out field tests with a focus group of 20 citizens. Based on the crosscheck of reverse translations and field tests, items were revised. The final survey contained 141 items on demographics, personal characteristics, experiences and satisfaction with police, neighborhood traits, and preference for private investigation.

3.2.1. Dependent variables The dependent variable used for this study is preference for private investigation, i.e., the likelihood of hiring PIs. The variable was measured in 12 different situations with a question, “If you are involved with the following [12] cases/incidents, would you hire a private investigator to resolve the case?” Answers were measured with 5-point Likert scale (strongly agree ¼ 1, agree ¼ 2, neutral ¼ 3, disagree ¼ 4, and strongly disagree ¼ 5). The 12 cases/ incidents include 1) a hit-and-run traffic incident, 2) a copy right violation case, 3) a criminal incident in own company/store, 4) an incident where automobile was stolen, 5) a fraud case, 6) a defamation of character case, 7) an incident where art works or jewelry were stolen, 8) a case in which family members/friends are missing, 9) when hiring a new employee, 10) when background check for marriage or business contracts is needed, 11) a fidelity case, and 12) a privacy invasion case.

3.2.2. Independent variables
Individual demographics, e.g., age, sex, marital status, educational levels and monthly income, were measured to check the representativeness of sample to the general population. Desire for personalized justice variable (Cronbach a ¼ .82) was measured by three items; ‘Private investigators will investigate cases in my favor,’ ‘Private investigators will pay more attention to me than the police would do,’ and ‘Private investigators will work for me.’ Citizen’s attitude toward PI’s morality was measured by an item, ‘I think PIs and their works are morally cleaner than the police and their works.” Citizen’s attitude toward PI’s confidentiality was measured by an item, ‘I believe PIs will protect my privacy more than the police would do.” Citizen’s attitudes toward PI’s investigation and information skills (Cronbach a ¼ .78) were measured by two items, ‘Generally PIs have better investigation skills than the police do,’ and ‘I think PIs are better in collecting information than the police.’ Citizen’s fear of crime (Cronbach a ¼ .84) was measured by six items, ‘In my neighborhood, robbery is a serious issue,’ ‘In my neighborhood, violent crimes are serious problems,’ ‘Break and enter type of crimes is a serious issue in my neighborhood,’ ‘Sexual assault is a serious problem in my neighborhood,’ ‘I am afraid of walking in my neighborhood during day time,’ and ‘I am afraid of walking in my neighborhood during the night.’ Citizen’s satisfaction with police (Cronbach a ¼ .94) was measured by eight items, ‘Overall, I am satisfied with police work in my neighborhood,’ ‘I am satisfied with community-oriented policing in my neighborhood,’ ‘I think policing is effective in my community,’ ‘I believe the police are doing well in my community,’ ‘I am satisfied with police responding to crimes,’ ‘I am satisfied with preventive policing in my neighborhood,’ ‘I think the police are making arrests of criminals effectively,’ and ‘I am satisfied with police service.’ The response options for all of 194 C.-H. Lee et al. / International Journal of Law, Crime and Justice 42 (2014) 188e202 the above mentioned items were ‘strongly disagree (¼1),’ ‘disagree (¼2),’ ‘neutral (¼3),’ ‘agree (¼4),’ and ‘strongly agree (¼5).

4. Findings

4.1. Demographics of sample
As shown in Table 2, 52% of the sample was male and 48% of the sample was female citizens. The mean age was 35 years old, and the youngest respondent was 19, and the oldest was 85. About half of the sample respondents were single, and more than 56% of the sample had four year university level education. The average income per month was about $3200.00. However, when outliners were excluded (15 cases had extremely high incomes), the mean income was about $2180.00, indicating the sample mean approximates the population parameter (Table 2).

4.2. Exploratory factor analysis & reliability tests The exploratory factor analysis and reliability tests were conducted to develop scale factors. An initial exploratory factor analysis results showed that the 12 items measuring the dependent variable could not be loaded into one factor. The initial analysis found that there were two possible factors; one is about works that have been typically carried out by the police, and another is about works that have been generally conducted by PIs. Thus, the subsequent second factor analysis focused on identifying two different factors, and found that one item, ‘Family members or friends are missing,’ was not loaded neither of both factors. Consequently, excluding this item, remaining 11 items were grouped into two factors; 1) Likelihood of hiring PI for typical police works, and 2) Likelihood of hiring PI for typical PI works. For both factors, the lowest item loading was .766, and the Cronbach Alphas were .92 and .84 respectively (see Table 3). Factor loading scores for all other items fall in acceptable range, and the Cronbach Alpha ranged between .78 and .94.

4.3. Descriptive statistics and bivariate correlations Table 4 shows results of descriptive statistics of the variables. The mean of the likelihood of hiring PI for typical police works was 22.4, indicating slightly negative view (median is 24). However, the difference is smaller than the standard deviation, meaning that overall, citizen’s attitudes toward hiring PI is indifferent and neutral. Similarly, the mean of the likelihood of hiring PI for typical PI works was 11.6, indicating neutral and indifferent attitudes toward hiring PI (median is 12). Other variables seem to be normally distributed. The variable, citizen’sfear of crime in neighborhood, is slightly positively skewed (median ¼ 18.5, mean ¼ 22.3), but the difference between the median and the mean is smaller than the standard deviation (4.4). Thus, all of the variables seem to have normal distribution. Table 5 shows results from bivariate correlations among the variables. Overall, there seems no multiculinearity issue, since there is no extreme correlation between any variables. The highest correlation coefficient is .69 between two dependent variables. Although the correlation coefficient value is at the high end between the ranges normally accepted, it is not a concern because separate multivariate regression analyses will be conducted for two different dependent variables. With the likelihood of hiring PI for typical police works, desire for personalized justice (r ¼ .24), attitude toward PI’s confidentiality (r ¼ .18), and attitudes toward PI’s investigation and information skills (r ¼ .35) were significantly and positively correlated. However, citizen’s satisfaction with police work was significantly and negatively related to the variable. These results indicate that citizens who have positive attitudes toward PI’s work and who have negative attitude to police are more likely to hire PIs to have them investigate their cases, which could be typically investigated by the police.

With the likelihood of hiring PI for typical PI works, only two variables, desire for personalized justice (r ¼ .20) and attitudes toward PI’s investigation and information skills (r ¼ .25), were significantly and positively correlated. These findings suggest that, regardless of satisfaction with police, citizens who have positive attitudes toward PI’s work are more likely to hire PI for their cases, which are considered to be typical works done by PIs. There is somewhat strong positive correlation between attitude toward PI’s morality and attitude toward PI’s confidentiality (r ¼ .61). Also, the desire for personalized justice has significant and positive, but moderate correlations with attitude toward PI’s confidentiality (r ¼ .24) and attitudes toward PI’s investigation and information skills (r ¼ .28). These positive correlations indicate that citizens who have stronger desire for personalized justice may have more positive attitudes toward PIs and their works.

4.4. Multivariate regressions Due to the results of the exploratory factor analysis, the dependent variable was divided into two; the likelihood of hiring PI for typical police works and the likelihood of hiring PI for typical PI works. Thus, multivariate regression analysis was performed twice using each dependent variable independently. Table 6 shows results of the two separate multivariate regression analyses. Variables affecting the likelihood of hiring PI for typical police works include income, desire for personalized justice, attitudes on PI’s investigation and information skills, and citizen’s satisfaction with police work. Overall, the model was significant (F ¼ 4.36), and the model explained 15% of variance in the likelihood (Adjusted R2 ¼ .15). Specifically, among those above, citizen’s attitudes toward PI’s investigation and information skills was significantly and positively related to the dependent variable (b ¼ 1.15). This indicates that when citizen’s attitude increases one unit, their likelihood of hiring PI increases by 1.15 units. Also, this variable was the most influential factor affecting the likelihood (Beta ¼ .28). The next most significant factor was citizen’s satisfaction with police work, which was significantly and negatively related to the dependent variable (b ¼ .28, Beta ¼ .22). The analysis results show that desire for personalized justice was also significantly and positively related to the dependent variable (b ¼ .55, Beta ¼ .19). These findings indicate that citizens who have dissatisfaction with police works and who have desire for personalized justice are more likely to hire private investigators to resolve their cases, which are typically handled by police officers. Regarding the likelihood of hiring PI for typical PI works, the results show that desire for personalized justice and citizen’s satisfaction with police work were significant variables affecting the likelihood. Overall, the model was significant (F ¼ 3.16), and the model explained 10% of variance in the likelihood. Specifically, citizen’s satisfaction with police works was significantly and negatively related to their likelihood of hiring PI for typical PI works (b ¼ .15, Beta ¼ .22). Also, the likelihood was significantly related with citizen’s desire for personalized justice (b ¼ .32, Beta ¼ .21). These findings indicate that citizens who have negative attitude toward police service and who have desire for resolving their cases in their favor are more likely to hire PIs to resolve their cases, which are typically carried out by PIs in the past.

5. Discussion and conclusions

5.1. Summary of findings
Multivariate analyses of the current study found important findings regarding citizen’s attitudes toward private investigation. First, regardless of types of cases (for both criminal and civil cases), citizens who have low levels of satisfaction with police work were more likely to hire PIs to resolve their cases. This finding, in other words, implies that citizens who have positive attitudes toward public police are less likely to hire PIs for their matters. Furthermore, the bivariate correlations showed that citizen’s satisfaction with police was significantly and negatively correlated only with the dependent variables, i.e., the likelihood of hiring PIs. This finding indicates that citizen’s satisfaction with police was not related to any other independent factors, but it alone impacts likelihood of hiring PIs.
Second, regardless of types of cases, citizens who have desire for personalized justice are more likely to hire PIs to resolve their criminal and/or civil matters. And the bivariate correlations showed that this factor was significantly and positively correlated not with attitude toward PI’s morality but with attitudes toward PI’s confidentiality and investigation/information skills. These finding imply that citizens may not concern PI’s morality, but they want to hire PIs as far as PIs can keep their cases confidential and resolve their cases. This finding is particularly important because moral hazard may jeopardize criminal justice system in any country. If citizens who do not concern moral standards in a society but who want to deal with their criminal and/or civil matters with PI’s help, and if PIs provide services to the citizens regardless of their moral standards, private and personalized justice will be common in the society. Such capitalized private justice system will jeopardize public criminal justice system (Reiman, 2006). Third, the attitudes toward PI’s investigation and information skills were a significant factor for the likelihood of hiring PI for typical police works. This factor was the most significant influence on the hiring preference. However, it had an insignificant relationship with the likelihood of hiring PI for typical PI works. These two contradicting results show that, when citizens are seeking resolutions for their civil matters, PI’s reputation for expertise in information technology is not an important concern for them. There are possible two reasons for this: 1) tasks (e.g., fidelity case, finding missing person, background check, etc.) citizens want to have PI carry out do not require well-advanced technologies and skills, and 2) citizens may prefer PI to public police when they think that PIs have better technologies and skills to carry out investigation of criminal cases. These findings imply that, once private investigation is fully legalized like in Great Britain or the U.S., public police may need to decide to either compete or cooperate with private investigators in criminal investigation. Although regulations prescribe cooperative support of PIs to the police (Kakalik and Wildhorn, 1971; O’Connor et al., 2008; Prenzler and Sarre, 2008), cooperation between PI and the police is largely dependent upon police officers’ attitudes toward PI (Cunningham and Taylor, 1985; Kakalik and Wildhorn, 1971; Scott and McPherson, 1971). Furthermore, a study showed that the police officers do not have a high regard for private policing (Shearing et al., 1985). As advised, when there is competition, especially competition in criminal investigation, outcomes of the competition between the police and PIs would be unfruitful (Cunningham and Taylor, 1985). Thus, it is recommended to facilitate cooperation between the police and private investigators in criminal and/or civil cases.

5.2. Limitations and conclusion
Like other empirical studies, this study also contains several limitations. First, the data set utilized for the current study contains responses from only 225 citizens in South Korea. Furthermore, the study subjects were selected through quota sampling, which is a type of nonprobability samplings. Thus, readers are advised to be cautious in interpreting and applying the findings from the current study. Second, although the current study was intended to fill the void of literature on citizen’s attitudes toward private investigation and their hiring tendency, the models analyzed in this study explained small amount of the variance in citizen’s preference (Adjusted R2 s were .15 and .10). Future study needs to expand the models to investigate other factors relevant to PI hiring behavior. Finally, data collected for the study may reflect cultural and social specificity of South Korea. Thus, future study on this regard in Western countries, which have more advanced and well developed private investigation systems, is recommended. Furthermore, a comparative study on this regard across nations may produce useful information for private investigation business as well as public policing.
The current study attempted to understand how citizens decide to hire private investigators for their criminal and/or civil matters. Factors considered in this study include citizens’ characteristics, such as demographics, desire for personalized justice, their attitudes toward PI’s investigation skills, attitudes toward PI’s morality and confidentiality during investigation, fear of crime victimization, and their attitudes toward public police. Findings showed that citizens’ desire for personalized justice and their attitudes toward private investigation/police were important factors affecting PI hiring tendency.