Abstract

The growing prevalence of longitudinal research in the social sciences, coupled with technological advances, provide new opportunities for researchers to more readily find participants from earlier studies. In addition, these advances necessitate the development of new strategies and methodologies for locating and building rapport with respondents. Drawing on previous methods of subject identification and location, this paper examines techniques for locating and interviewing former gang members from the 1954-1957 Special Youth Program in Roxbury, MA. In contrast to most longitudinal studies, (1) more than 50 years have passed since the original study with no contact with respondents in the interim, (2) we build  on a study not  intended to include long-term follow-ups, and (3) much of the location and identification of participants occurred using digital resources. Implications for harnessing digital technologies and methods of developing relationships with study participants otherwise wary of engaging researchers are discussed.

Introduction

Criminology has substantially benefitted from the growing prevalence  of life course research, for this has provided insights into the causes and consequences of criminal behavior over the lifespan (e.g., Farrington, 1995; Giordano, 2010; Glueck & Glueck, 1950; Soothill, Humpreys, & Francis, 2012; Thornberry & Krohn, 2002; Tracy & Kempf-Leonard, 1996; Wolfgang, Thornberry, & Figlio, 1987). Concurrent with this research, a well established body of literature details the organization of prospective longitudinal examinations of youth and adolescents (e.g. Loeber, Farrington, & Stallings, 2011; Schubert et al., 2004), with studies in which contact with respondents has been maintained for many years (Farrington, 1995; Glueck & Glueck, 1950; Vaillant, 2002). Taken together, this body of literature highlights the importance of studying human development over time. As Kleck, Tark, and Bellows (2006) recently demonstrated, longitudinal research accounts for oneeighth of empirical criminological research. A smaller but equally substantive body of research has emphasized fieldwork with both active and former offenders (Decker & van Winkle, 1996; Densley, 2012; Hagedorn, 1990; Jacobs, 1999; Jacobs & Wright, 2006; Polsky, 1969; Shover, 1983, 1985; Wright, Decker, Redfern, & Smith, 1992). Falling somewhere between each of these methods is the practice of using archival, historical, and other older records to understand phenomena related to both crime and criminal justice (e.g. Bright, Decker, & Burch, 2007; Hughes & Short, 2005; Papachristos & Smith, 2012; Watkins & Moule, 2013).

Bosworth (2001) suggested a decade ago that historical, archival, and similar forms of retrospective research fit uneasily into the mainstream criminological discourse.1 Despite the growth of longitudinal research and theoretical frameworks emphasizing this type of research, and perhaps due to this ill fit, little has been written on the process of integrating historical records and long-term retrospective research in criminology (but see Dempster-McClain & Moen, 1998; Laub & Sampson, 2003; Sampson & Laub, 1993), primarily because of the difficulties associated with locating offenders or members of otherwise hidden populations (Biernacki & Waldorf, 1981; Coen, Patrick, & Shern, 1996; Hampson et al., 2001; Hansen, Tobler, & Graham, 1990; Laub & Sampson, 2003; Wright et al., 1992). Indeed, some researchers have described such research as more “craft than science” (e.g. Call, Otto, & Spenner, 1982). Nonetheless, this style of research offers substantive benefits, including lower costs compared to prospective research. Further, such thinking stands in sharp contrast to the perceptions of the Internet and the digital age, where a wealth of information is increasingly accessible online and of great use for research purposes (Borgman, 2007; Featherstone, 2000; Ortiz & Ballon, 2007). Although criminologists have not made a more systematic attempt to integrate the Internet into their academic toolbox, the utility of the web should not be overlooked.

To that end, the current paper details the intersection of sample reconstruction, retrospective research, and subject recruitment techniques employed to locate and interview participants from the Boston Special Youth Program (SYP), a 1950s gang intervention program. We provide a “how to” of conducting this style of research, focusing on three specific areas: reconstructing an incomplete dataset using a variety of sources; locating respondents; and conducting interviews with former study participants. In doing so, we note three key improvements over past research. First, in contrast to prospective longitudinal studies and similar past research (Ortiz & Ballon, 2007; Root, Smith, Whelan, Sandler, & Voda, 1994; Sampson & Laub, 1993), we were able to construct a substantial portion of our participant rosters from limited information obtained roughly 60 years ago. Second, given technological advances that have occurred in the past decade, we detail the utility of the Internet in facilitating elements of retrospective research, including participant location. Third, we describe early encounters with subjects and illustrate rapport-building methods used in attempts to overcome respondent concerns about privacy and potential scams. Overall, we demonstrate the feasibility of this style of research among a sample of former delinquents who have been out of contact with researchers for a half-century. As criminology continues to examine crime over the life course, this research provides a potentially useful alternative to traditional prospective longitudinal research practices. We begin by detailing the original project and the steps taken to facilitate the current project. 

The Boston Special Youth Program

Following the high profile murder of a local Rabbi on New Year’s Eve of 1952, the citizens of Roxbury, MA demanded action (Miller, 1957). News reports indicated that the murder was the result of a mugging that had quickly escalated and involved gang members from a local housing project. Boston’s United Community Services responded by organizing a committee to address the local gang problem. From this committee emerged Walter Miller’s study of juvenile delinquency, the Boston Special Youth Program.

The study began in the summer of 1954 to evaluate the effectiveness of a concentrated delinquency control program (Miller, 1957). The program was to be multi-pronged, using detached case workers to reorient gangs away from delinquency, provide social services to the gang members’ families, and give the community the necessary tools to engage in self-help once the study ended (Miller, 1957, 1963). The study received funding from United Community Services and the National Institute of Mental Health; it focused predominantly on the use of outreach workers, documented their interactions with gang members, and consulted the workers to find out more about the gangs with which they would associate (Miller, 1957). Using graduate students from major Boston universities, Miller set out to locate gangs in Roxbury. The research project was based in the area where the Rabbi had been killed and located nearby street corner groups. By engaging community stakeholders and groups of teenagers, seven intense study gangs were identified and given an outreach worker.2 Of these gangs, five were Caucasian (four male, one female) and two were African-American (one male, one female) (Miller, 1973).3 Miller (1962) also compiled data on control groups who did not receive an outreach worker but shared similar demographic features and neighborhoods as the intense groups.

Data collection lasted through the spring of 1957 and consisted of four sets of records. First, outreach workers maintained contact cards detailing the time, place, and nature of each interaction between a group member and outreach worker. Second, Miller kept a journal, detailing  his  interactions and interviews with outreach workers regarding their gangs (Miller, 1957). Third, he also engaged gang members, taking handwritten notes and tape recordings of these interactions and group meetings.4 Finally, secondary data sources were compiled, including gang members’ juvenile criminal histories, aggregate criminal statistics for the neighborhoods in which the gangs resided, as well as census tract data on these neighborhoods. Early examinations of these data were promising, though subsequent analyses revealed the program to have few positive results. Case workers had minimal effect on gang members’ behavior, and engaging the families of members only occurred near the end of the project (Miller, 1957; 1962). There was, however, moderate success in providing the community with the ability to help itself, mainly through increased inter-agency cooperation. More findings from the Special Youth Program were published, but by the 1970s, criminological research shifted away from gangs (Bookin-Weiner & Horowitz, 1983). Miller moved on to direct the National Youth Gang Survey and later helped found the National Youth Gang Center (Klein & Maxson, 2006).

Rediscovering Miller and the Special Youth Program

After Miller’s death in 2004, his professional papers and effects were given to Hedy Bookin-Weiner, a graduate student when Miller was at Harvard. The second author received these papers from Dr. Bookin-Weiner in 2006. Among this collection, two specific pieces drew immediate scrutiny; there were boxes containing the outreach worker contact cards and the manuscript of a book, City Gangs. This book was Miller’s unpublished monograph about the gangs and gang members of the project (Miller, 1963). When a publisher told him in the 1960s that the book could be published only if the manuscript was shortened considerably, Miller’s response was to add several hundred pages to the manuscript, instead. The publisher subsequently declined the opportunity to publish the book. Between 2006 and 2011, Miller’s works were inventoried, and the manuscript was digitized (see Miller, 2011). In July of 2011, gang rosters from the study were discovered among the remaining papers.5 Rosters of the intense study groups were very thorough, often including the full name, date of birth, gender, race/ethnicity, gang, home address, parish/church, high school, and father’s name for many of the 231 intensive study participants.6 Rosters of the control groups, comprised of 258 members, contained similar information, but were less thorough due to infrequent contact with project workers. Remaining groups were comprised of 43 female auxiliary and young male members of the study groups.

In total, 532 study participants were identified from the Miller records. Table 1 provides the demographic characteristics of the sample, including gang, group type, race, gender, and age range from the study.

Table 1. Characteristics of gangs

Difficulties of retrospective research

Past research highlights a number of difficulties in the collection of prospective longitudinal data (Call et al., 1982). Many of these same obstacles are present when trying to reconstruct, locate, and interview study participants using historical data. With respect to the Special Youth Program sample, the youngest participants would be in their late 60s, with many beyond the average life expectancy for their birth cohort. Coupled with the fifty year gap between the study period and our efforts, many pieces of information (1950s addresses, father’s name, school) seemed to be of limited use for locating participants; the neighborhood of Roxbury had undergone many changes since the 1950s—deindustrialization, the closing of schools and churches—and as a result, many participants were unlikely to still reside there. These two points are magnified by the inclusion of minorities and women; such data is a rarity in retrospective criminological research. Indeed, most research typically contains only Caucasian male participants (e.g. Laub & Sampson, 2003; Vaillant, 2002; see also Giordano, 2010). New strategies must be used to account for name changes among female participants because of marriage (Ortiz & Ballon, 2007; Root et al., 1994). Lastly, Miller’s data indicated that some individuals were chronic offenders. For those who continued their criminal career into adulthood, there may be limited incentive for participation. For individuals who did not continue offending, there remain incentives to hide their pasts from friends and loved ones (Farrington, Gallagher, Morley, St. Ledger, & West, 1990).

These obstacles were problematic for multiple reasons. Our human subjects review board was equally concerned about subjects’ privacy7, especially divulging SYP participants’ criminal behaviors to spouses and family members, their recent offending behavior, and our ability to acquire participant social security numbers. We addressed these concerns by first noting that we would not reveal the nature of the SYP to family members. Instead, we referred to very generic information—individuals we were contacting had grown up in Roxbury and had participated in a study in the 1950s. Second, when interviewing participants, we did not inquire about recent criminal behavior such as substance abuse. Lastly, in Massachusetts, the deceased do not have privacy rights. This meant opportunities for access to school records. Social security numbers are available online through the Social Security Death Index on both Ancestry.com and Legacy.com (see also Acquisti & Gross, 2009). Our organizing efforts thus focused on three goals, (1) reconstructing the sample by verifying information on program participants, (2) locating living participants and verifying the identity of dead participants, and (3) conducting interviews with SYP participants regarding their experiences in the program, their gang, and growing up in Roxbury. To accomplish these goals, we first reviewed the gang records in our possession, particularly those which contained birth information and began reconstructing the sample using Internet searches. 

Sample reconstruction

Past research illustrates the construction of samples from official criminal records (e.g. Blokland & Nieuwbeerta, 2005). In the absence of nationally centralized databases for various personal records such as marriage or criminal offending, fewer in-depth treatments of sample reconstruction exist. Bosworth highlighted a number of reasons why this remains so, including “trails running cold, illegible documents, restricted opening hours [at libraries, archives, and state facilities], cool and even resistant librarians and archivists, incomplete, inconsistent catalogues, and the task of sifting through infrequent, useful documents from the mass” (2001, pp. 434-435), Sampson and Laub (1993) provided one account of sample reconstruction for their original, six year (1987-1993) examination of the Glueck data. Elder, Pavalko, and Clipp (1993) similarly detailed working with and recoding archival data. In contrast to the thorough records of those studies, records from the Special Youth Program were more consistent with Bosworth’s (2001) description of historical data.

The Special Youth Program consisted of interactions between participants and outreach workers between 1954 and 1957, with limited followups with two of the Caucasian male gangs conducted in 1962. Because of the poor nature of records on the other groups, we focused our efforts on the 489 intense and control group members. Miller’s gang-based study design provided general parameters for participant characteristics. All gangs were age-graded, with members typically born within a twoor three-year time span. In addition, while Roxbury was racially integrated, the racially divisive nature of Boston during the pre-Civil Rights Era meant the gangs were not integrated. For example, notes from a housing project meeting from the late 1940s indicated residents were unhappy about Negroes moving into the neighborhood. Lastly, gangs were only composed of boys or girls, with auxiliary groups being exclusively female. These general parameters were believed to make verifying subjects’ personal information easier.

We began reconstructing the sample in autumn of 2011, conducting public records searches on Internet sites such Ancestry.com, a genealogy website (see Appendix A for a useful list of online resources). Ancestry draws from nationally available indices—locally reported sources, federal public record indices, and the Social Security Death Index. In the spring of 2012, Ancestry. com also provided access to the recently released 1940 Census. Searches on Ancestry were originally conducted using participant name and demographics, specifying Boston as a place of birth, with up to a five-year age range. This range was narrowed for individuals for whom we had more specific information, such as a full birth date, a birth year, or just a general age-range within a gang. Such an approach casts a wide net and may result in a number of individuals being identified. As such, searches were supplemented with official state documents.

Prior to the release of the 1940 Census in 2012, birth records were available through the Massachusetts State Archive and the Massachusetts Registry of Vital Records and Statistics. The State Archives house a number of historical records, including birth record rolls. Covering five-year increments, these microfilm rolls indicate the full name, location, and year of all births occurring in the state, as well the volume and page number where official birth records can be obtained. While photographs may be taken of these rolls, records were instead transcribed by hand with pencil and paper to save time. This recorded information was then taken to the Massachusetts Registry of Vital Records and Statistics, where official birth, marriage, and death volumes were available. These volumes were reviewed, recorded by hand using pencil and paper, and compared to gang roster information. Individuals were matched through a combination name, date and place of birth, parents’ names, and home addresses.

While most birth records are public information in Massachusetts, some records are restricted, typically for children born out of wedlock, adopted, or when one parent was not named on the birth certificate. When dealing with restricted records, only the year of birth was available. Birth records were also unavailable when individuals were not born in-state (though as we illustrate in Table 2, this does not mean that information on them was entirely unavailable or that these individuals are unable to be located). When birth records were unavailable or inaccessible, birthdates may be triangulated using a number of sources. For example, the 1940 Census data contained the age of participants when forms were filled out, providing a smaller window of possible birth years for individuals born prior to the administration of the Census. Pay websites, such as Intelius.com, often contain limited birth and other personal information. Potential participants were matched on demographic characteristics as well as their addresses from the 1950s. When marriage records can be located, certificates may provide demographic features on which individuals can be matched (race, gender, name, parents’ names and addresses) and the age when the individual got married, which will provide a narrower range for possible birthdates. Marriage records for individuals born out-of-state also identify location of birth, which may be used to acquire more complete birth information.

A number of other resources may prove useful when trying to confirm participant identities. For example, we were able to view yearbooks from Roxbury and Boston area high schools that SYP participants attended. Located in the Boston City Archives, these yearbooks provided no specific birth information, but depending on grade and year, we were able to narrow down potential birth years for some individuals. We also received permission from the Massachusetts Executive Office of Public Safety to use their digital Criminal Offender Record Information (CORI) system. This database contains all known adult offenders in Massachusetts. Individual records are accessible by submitting an offender’s name and date of birth into the database. While the database does not make juvenile records accessible, CORI provided another way of verifying birth records in the Miller files. Criminal records for individuals who were over the age of 18 during the course of the program were matched with CORI offending profiles, thus illustrating that the individuals we had identified were the same as those in the original sample.

Table 2 shows the breakdown of available birthdates by gang status, race, and gender, with remaining birthdates to still be found.

Table 2. Available birthdates by gang status, race, and gender (n=489)

Hidden populations or just hidden in plain sight? 

In addition to finding current information on SYP participants, we were interested in communicating with those participants who were still living. This was not an easy task, and prior research highlights the difficulties of locating hidden populations—those individuals whose activities fall outside of the law (Heckathorn, 1997; Watters & Biernacki, 1989). With respect to offenders, these difficulties manifest in three ways that are problematic for both longitudinal and retrospective research. First, offenders’ lives are often chaotic, involving transience, periods of incarceration, and failed social relations. As a consequence, locating these individuals can prove quite challenging. Second, past research has documented that chronic offenders are also the most likely to experience early death; offenders who may be of the most interest to criminologists may also be the least likely to be living (Piquero, Farrington, Shepherd, & Auty, 2011). Third, even if subjects can be located, gaining the trust of current or former offenders can be a difficult and time consuming process (Laub & Sampson, 2003; Wright et al., 1992).

Two general methods have typically been used to overcome these obstacles. Snowball sampling techniques have been employed as a way to prospectively identify potential study participants (Biernacki & Waldorf, 1981; Coleman, 1958; Wright et al., 1992). This strategy uses the first responding participants to start referral chains which branch out over time and provide researchers with a degree of credibility when engaging new respondents. Targeted sampling strategies have also been used to find hidden populations (Watters & Biernacki, 1989). This approach draws from snowball and ethnographic sampling methods to provide a systematic sample of a specific population such as drug users. If historical records are drawn from a specific geographic region, the latter approach may prove more fruitful for retrospective research.

These challenges coincide with the notion that while most individuals engage in some level of deviance over the life course, few individuals continue to do so beyond adolescence (e.g. Gottfredson & Hirschi, 1990; Moffitt, 1993). The problems of locating a hidden population along with the recognition that ex-offenders undergo a series of cognitive and social changes which color how they view their past indiscretions creates further hurdles in locating former offenders (Ebaugh, 1988; Giordano, Cernkovich, & Rudolph, 2002; Laub & Sampson, 2003; Maruna, 2001). Going straight means having much to lose if friends, colleagues, and family learn of this past (Farrington et al., 1990; Maruna, 2001). Locating former offenders may cause personal issues for the participants; while they may have settled down and thus may be more easily found, they remain cognizant of the consequences surrounding their past. Former offenders then represent something akin to a population hidden in plain sight, for this population is widespread and operates within legitimate institutions and enterprises. 

Locating participants

Official records formed the foundation for being able to locate former study participants. Marriage records are a necessary starting point for participant location, particularly female study members. Without these records, it is unlikely that researchers will be able to locate female participants, whether they are living or dead. Indeed, prior research suggests that marriage records are the key piece of information for locating female participants since historically, many women change their name or relocate after marriage (Klebanoff, Zemel, Buka, & Zierler, 1998; Ortiz & Ballon, 2007; Root et al., 1994).

The same processes for locating birth records were applied to finding marriage records of SYP participants. Given the time period of the original study, we assumed that all marriages would begin shortly after participants turned 18; researchers focused on records covering the period between 1951 and 1972, at which time records were digitized. Marriage rolls in the State Archives were available on microfilm for five-year intervals, with full marriage certificates available at the Registry of Vital Statistics. Restrictions were also placed on marriage records where either spouse had birth restrictions. Prior to 1960, only marriage years and locations were available for restricted records. After 1960, restricted records provided the year and location of the marriage and spouse’s first and last name. Marriage records provided additional information useful for locating participants: occupation and address at the time of marriage. These addresses, especially for individuals still residing on well known streets in Roxbury, reinforced our belief that we had accurately verified the correct people. Likewise, the listing of occupations provided indicators that subjects might have relocated, such as the case with military service.

In conducting retrospective research that drew on records from the 1950s, we found a number of the original study participants were deceased. This necessitated research using death records. For participant death records, even those for whom birth information was unavailable, we began with Ancestry.com, which provided access to the Social Security Death Index and the Massachusetts Death Index 1970-2003. First, the Massachusetts Death Index was searched for each participant, based on name and any known birth information. Information provided here included decedent’s full name, death date, death place (town and/or county), birth date, and birth place. The Social Security Death Index provides similar information, as well as social security numbers for those have been deceased more than ten years, and the state where the number was issued. This provided a greater degree of accuracy when searching for individuals without full birth information, because all would be expected to have been issued their social security numbers from Massachusetts. In conjunction with these main sources, we used Legacy.com and the obituary page of the Boston Globe to search for potentially deceased participants or their spouses. Online obituaries are an excellent source of information even if participants are not able to be located. In some instances, we came across the obituaries of parents, siblings, and spouses, all of which provided clues as to whether the individual we were seeking was alive, as well as where they might currently be residing. Death records are also available in the Massachusetts Registry of Vital Statistics, although much of our research there concentrated on birth and marriage records, given the greater prevalence of death information on the web.

While searching for information on deceased individuals, we simultaneously searched for information on those participants who were still alive. Preliminary searches were conducted on Google, and while many of our participants’ names are common and ethnic, general Internet searches provided information on respondents with uncommon names, as well as occasional links to participant web pages, including professional sites, and Twitter and Facebook profiles. For example, one notorious delinquent from the SYP was located because his unique nickname was part of his Twitter username. Pictures that he posted were geotagged, and this address was matched to the White Pages. Another individual’s SYP record noted that he had planned to attend a prestigious Ivy League medical school; he was located by combining this information in a Google search. Classmates.com was also used, along with its collection of digital yearbooks. From this, we discovered that one participant played college football in southern California. U.S Public Records Indices from the late ‘80s and early ‘90s were available on Ancestry.com, although many of the addresses provided by these indices were out of date. For more up to date contact information, we drew from multiple sources. Intelius.com provided background information on individuals matched by name and age, as well as other personal records, such as relatives’ names, prior addresses, and phone numbers. Phone numbers and addresses were run through Whitepages.com. Phone numbers frequently changed as our participants relocated, so sometimes it was necessary to run multiple numbers or addresses. With few exceptions, the living individuals who have been located have been so through Intelius and Ancestry, while death records were obtained from the aforementioned death indices. Table 3 shows the results of these efforts to date.

Table 3. Participants found living, dead, and unknown across race, gender, and gang affiliation (n=489)

Contacting and building rapport with SYP participants

Once longitudinal, secondary data sources are acquired and statistical analyses are conducted, a research project usually comes to a close (but see Giordano et al., 2002; Laub & Sampson, 2003). Such an approach to retrospective research leaves key questions unanswered. What was it about a particular experience that mattered? Why made a given experience more salient than another? Because we were able to reconstruct Miller’s original sample by identifying, locating, and acquiring criminal histories for SYP participants, we were interested in taking the next step, contacting and interviewing them. In doing so, we drew on both traditional techniques used to engage participants in longitudinal research and dealt with problems unique to engaging elderly populations.

Based on the records we had collected through Spring, 2012, we sent one dozen individuals who were believed to be former participants a one-page recruitment letter as well as a self-addressed stamped envelope and a ten dollar bill. This strategy was previously shown to increase response rates (Biner & Kidd, 1994). Letters described the project and provided contact information for the principal investigator. In some instances, letters were returned unopened, and we were forced to reexamine respondent addresses. In other instances, former participants left the researchers voicemails wishing to know more about the project. Respondents were initially wary, unsure of how we had located them or why we were interested in their lives (see Laub & Sampson, 2003). For men especially, their wives or children were gatekeepers with whom they would discuss participation, and one of our respondents admitted that he would proceed with us only with the blessings of his family. Because a key component of finding former members would be respondents’ willingness to help (Call et al., 1982), these immediate concerns by respondents highlighted the need for meaningful rapport building. For research involving the elderly and other marginalized populations, rapport building is especially critical (Tewksbury & Gagne, 1996; Wright et al., 1992). Therefore, we adopted numerous strategies to build rapport and maximize the cooperation with respondents.

First, prior research highlights the fears that the elderly have in regard to fraud and other scams (Johnson, 2003;  Sundeen, 1977), and longitudi nal research may be incomplete because elderly subjects do not participate based on these fear (Mody et al., 2008). To rectify these concerns and establish relationships with participants, we began by reinforcing our association with our university, an affiliation originally reflected in the recruitment letters. We explained that our records indicated that the individual had participated in a study during the 1950s. Even if the individuals did not remember the project exactly, we asked that they still speak with us. We explained our willingness to meet or speak with participants at a convenient time and place, including over the telephone, and promised confidentiality regarding information they shared. Further, we framed our research as being part of an historical account of Roxbury and of how these people had grown up. Indeed, sharing your story is an appealing prospect for criminals and non-criminals alike (Letkemann, 1973; Wright & Decker, 1994; 1997).

In addition to the money enclosed in the recruitment letter, our second strategy invoked the “norms governing the exchange of money and other kinds of favors” (Berk & Adams, 1970: 112). We were asking strangers to cooperate with a study examining their lives, regardless of the quality of their life experiences, so it was necessary that we be forthright and honest in answering questions related to the project. We explained how individuals had been found through phone directories as well as the Internet. For example, one participant was a faculty member at an East Coast university; he had been located through his faculty web page. We also explained why participants were contacted in the first place. While most of the subjects did not initially remember participating in the study, key details emerged when they were asked about topics such as their sports clubs and which outreach workers had assisted in getting equipment. The aforementioned faculty member, an African-American, distinctly remembered his group being given secondhand sports equipment, while the Caucasian gangs were given new equipment. While minute, this detail was substantial enough to be remembered more than 50 years after the fact. Often, participants wished to know what would be done with these interviews, and we explained that they would be incorporated into books, dissertations, and academic articles, but their names or identities would not be divulged.

Third, past research highlights the role that word meaning and syntax play in conducting field work (Irwin, 1972; Wright & Bennett, 1990; Wright et al., 1992). We experienced firsthand how one word can inhibit rapport building. Each individual we interviewed was adamant that he/ had not been a member of a gang, but rather part of a social or athletic club as well as a less organized corner or neighborhood group (e.g. Jansyn, 1966; Suttles, 1968; Whyte, 1943). For these individuals, gang evoked the imagery of serious, violent crime and involvement with drugs, neither of which were prevalent in the 1950s. Following the negative feedback we received when mentioning the gang, we immediately began referring to our participants as group members and inquired about the nature of the group. Likewise, some participants were unwilling to discuss certain topics in specific terms. One Caucasian male, for example, did not want to provide information about his wife or the quality of his marriage. In such instances, we did not investigate further, for fear that the participant would refuse to answer other questions or rebuke any future contacts we wished to make.

Though the individuals involved in the Boston Special Youth Program were now in their 70s, they were willing to discuss their experiences in the program, their formative years, and their lives in the decades following participation in the project. In doing so, they provided evidence of the breadth and effectiveness of rapport building techniques, as well as vindication regarding the identification and location of research subjects over an extended period of time.

Discussion

The advent of the life course perspective has spurred a substantial body of research examining offending over time; a variety of research methods have been employed to understand the experiences of current and former offenders alike. The present study illustrated the broader utility of the Internet for criminological research, including locating and conducting follow-up research, even among study subjects who have been out of the research spotlight for a number of decades. Despite such a powerful tool being at the hands of capable researchers, little has been written on Internet use as a tool for conducting such research (Ortiz & Ballon, 2007). Drawing from previous research on both longitudinal and retrospective studies, our research demonstrates efforts to identify, locate, and build rapport with an elderly population of former gang members over a one-year period. Given the substantial obstacles presented by this line of research, we relied heavily on the Internet to verify identities and locate participants. After working on this for a year, we have acquired birth information for 68.3% of the sample and located 64.2% of the sample. We are in the beginning stages of conducting interviews with more members of the original program and continue to search for and locate information on remaining individuals. From our experiences, three implications for future research emerge.

First, the absence of research integrating various information sources and methods of data collection is troubling. To be sure, such research is by no means easy but is necessary to continue pushing the discipline forward. As we have demonstrated, such research is possible. There are many opportunities for conducting retrospective longitudinal research using historical or archival data, and we encourage researchers to look more closely for such opportunities. Papachristos’ recent work with the Chicago Crime Commission data from the 1920s, Ward’s research (2012) using the Memphis Church Records, and the investigations of Bright et al. (2007) with St. Louis Juvenile Court data from the early 1900s are recent, noteworthy examples. In the event that historical studies or data contain participants that may be living, these individuals can be located through various sources, especially with the advent of the Internet, and criminology benefits from their input (Giordano, 2010; Laub & Sampson, 2003). The Internet and the information it contains can be an asset for locating such individuals and supplementing other research methodologies such as referral chains for snowball sampling (Hampson et al., 2001; Laub & Sampson, 2003; Ortiz & Ballon, 2007). As criminology becomes more focused on longitudinal research, the life-course framework, and the process of desistance, the Internet is an increasingly valuable resource that should not be overlooked.

Second, prior research has used the Internet, but only to a limited extent (e.g. Ortiz and Ballon, 2007). The Internet is a viable source of several important kinds of information, including addresses and phone numbers, and because much of the information online is drawn from public records,   it may be especially useful for locating older individuals, particularly those with substantial community ties. Records of home ownership, a stable mailing address, or engagement in civic functions such as voting or owning a driver’s license all increase the likelihood of someone having an online presence. Younger individuals may not have these established ties, but may be more easily accessible through a growing number of social networking sites such as Facebook or Twitter, or where parental records are available. With the growing use of these websites, their use as sources of information for future research will likely increase, and scholars should be cognizant of this utility. The convenience of the Internet, however, is tempered by its ability  to present limited or obsolete information (Eysenbach & Diepgen, 1998). This may especially be the case when trying to locate marginalized populations who move frequently, live with relatives, reside in correctional or other monitoring facilities. Because of this, it is necessary to triangulate information between multiple online and offline sources (see Appendix A for a list  of potential sources). Given this caveat, the best manner to use the Internet for follow-up research is likely as a supplement to other data sources. While we were fortunate with the recent release of the 1940 Census, most birth and marriage records do not appear to be online at this time, thus occasionally necessitating travel. We have relied on the Web to match known birth information as closely as possible, since travel between Phoenix and Boston is expensive and time consuming.

Focusing on the Internet only as a tool for conducting retrospective research provides an underwhelming account of its potential for social science in general. A growing body of criminological literature details the use of the Internet for other research purposes, such as providing descriptive accounts of numerous antisocial groups online and the recruitment and marketing strategies they employ (e.g. Burris, Smith, & Strahm, 2000; Reid & Chen, 2002; Schafer, 2002; Selepak, 2010; Zhou, Reid, Qin, Chen, & Lai, 2005), qualitative accounts of offenders’ web use (e.g. Holt & Blevins, 2007; Holt, Blevins, & Burkert, 2010; Holt, Blevins, & Kuhns, 2008), and the use of online survey administration for both criminological (e.g. Nobles and Fox, 2013; Wells, Cavanaugh, Bouffard, and Nobles, 2012) and general research (e.g. Edelman, 2012). Recent research similarly details the importance of the Web for criminal justice agencies (Heverin and Zach, 2010; Police Executive Research Forum, 2013); the web will become an increasingly valuable tool for researchers in the years to come (see also Pyrooz, Decker, and Moule, 2013).

Third, to the extent that personal information is becoming more commonplace online and locating people using this information is becoming more possible, potential respondents may not be comfortable knowing how little online privacy they have. Using the Internet to locate individuals may necessitate explaining limited digital privacy to them. We were cognizant of this and sought to put study members at ease early on to facilitate the relationship building process. Steps taken in this process included the initiation of reciprocal exchange by including money in our recruitment letter, emphasizing confidentiality of communication, and respecting generational differences in perceptions of gangs and corner groups. Once we explained the project in greater detail, including how subjects were located, participants were more willing to speak with us. Perceptions of Internet privacy may be age-graded. Younger generations may be more comfortable maintaining an online presence and thus being located through the Web (but see Zimmer, 2010). Older individuals, on the other hand, are aware of and are concerned about scams. Indeed, in more than one case, a potential subject referred us to an adult child who screened us to see if we were a legitimate research group. We often provided Miller’s earlier publication as evidence of our legitimacy, as well as a university story about the research. As more individuals are located, both in the context of this and other research projects, privacy is an issue to keep in mind.

There are, however, caveats surrounding the nature of this type of retrospective research, particularly regarding both generalizability and research ethics. The circumstances of the data described are by no means typical:  the primary was data collected for three years and later acquired purely by chance by the authors; it was collected before the existence of Institutional Review Boards; and the data contained enough personal information to attempt follow-up research. Although the number of respondents and pre-IRB studies suitable for this type of research within criminology are likely limited, we cannot be sure of the extent to which this is so. This limitation notwithstanding, besides the seminal work of Laub and Sampson, the medical literature has its own distinct body of retrospective research, some of which spans generations (e.g. Bargagli, Sperati, Davoli, Forastiere, & Perucci, 2001; Blane, Berney, Smith, Gunnell, & Holland, 1999; Hser, Hoffman, Grella, & Anglin, 2001; Klebanoff, Zemel, Buka, & Zierler, 1998). Beyond retrospective research, the Internet can also be a useful tool for both cross-sectional and prospective longitudinal research, as these examples illustrate.

But just as we encourage this manner of research, the ethical considerations of delving into the private pasts of research participants must  also  be acknowledged. Contemporary, secondary datasets, including the NLSY and Add Health, are ill-suited for the methods described in this paper. Besides containing anonymous personal identifiers, these datasets often include terms of use prohibiting attempted identification of respondents. On the other hand, individuals conducting longitudinal research or cross-sectional research with the potential to become longitudinal research may also find the methods described in this paper useful. This is especially the case if there have been extended time periods between study waves. Contemporary studies should also clearly present the nature and scope of data collection effects, and researchers should work closely with IRBs and study participants to have proper procedures for data collection in place and to make clear if and when those collection efforts will end. These procedures are important in respecting the privacy of study participants as well as offering the ability to refuse future participation.

In the end, we encourage researchers to continue expanding their horizons when it comes to conducting research. Different forms and sources of data lend themselves to a variety of applicable methods and theory, and we illustrate the means through which quantitative and qualitative data can be successfully intertwined. Likewise, we suggest that researchers consider the new opportunities presented by evolving technologies, such as the Internet, for conducting research. These technologies are unlikely to fade away, and the sooner they are embraced, the better. Combined approaches leave the field well-positioned to simultaneously revisit its past and confront its future in a productive fashion.

Appendix

Appendix A. Sources of Information in the Search Process for SYP Participants

References

Acquisti, A., & Gross, R. (2009). Predicting Social Security numbers from public data. Proceedings of the National Academy of Sciences, 106, 10975-10980.

Bargagli, A. M., Sperati, A., Davoli, M., Forastiere, F., & Perucci, C. A. (2001). Mortality among problem drug users in Rome: An 18‐year follow‐up study, 1980–1997. Addiction, 96, 1455-1463.Berk, R., & Adams, J. (1970). Establishing rapport with deviant groups. Social Problems, 18, 102-117.

Biernacki, P., & Waldorf, D. (1981). Snowball sampling: Problems and techniques of chain referral sampling. Sociological Methods and Research, 10, 141-163.

Biner, P. M., & Kidd, H. J. (1994). The interactive effects of monetary incentive justification and questionnaire length on mail survey response rates. Psychology and Marketing, 11, 483-492.

Blane, D., Berney, L., Smith, G. D., Gunnell, D. J., & Holland, P. (1999). Reconstructing the life course: Health during early old age in a follow-up study based on the Boyd Orr cohort. Public Health, 113, 117-124.

Blokland, A. A., & Nieuwbeerta, P. (2005). The effects of life circumstances on longitudinal trajectories of offending. Criminology, 43, 1203-1240.

Blumstein, A., & Cohen, J. (1987). Characterizing criminal careers. Science, 237, 985-991.

Bookin-Weiner, H., & Horowitz, R. (1983). The end of the youth gang. Criminology, 21, 585-602.

Borgman, C. L. (2007). Scholarship in the digital age: Information, infrastructure, and the Internet. Cambridge, MA: MIT Press.

Bosworth, M. (2001). The past as a foreign country?: Some methodological implications of doing historical criminology. British Journal of Criminology, 41, 431-442.

Bright, C. L., Decker, S. H., & Burch, A. M. (2007). Gender and justice in the Progressive Era: An investigation of St. Louis Juvenile Court cases, 1909-1912. Justice Quarterly, 24, 657-678.

Burris, V., Smith, E., & Strahm, A. (2000). White supremacist networks on the Internet. Sociological Focus, 33, 215-234.

Call, V. R. A., Otto, L. B., & Spenner, K. I. (1982). Tracking respondents: A multi-method approach. Lexington, MA: Lexington Books.

Coen, A. S., Patrick, D. C., & Shern, D. L. (1996). Minimizing attrition in longitudinal studies of special populations: An integrated management approach. Evaluation and Program Planning, 19, 309– 311.

Coleman, J. S. (1958). Relational analysis: The study of social organizations with survey methods. Human Organization, 17, 28-36.

Decker, S. H., & Van Winkle, B. (1996). Life in the gang: Family, friends, and violence. Cambridge: Cambridge University Press.

Dempster-McClain, D., & Moen, P. (1998). Finding respondents in a followup study. In J. Z. Giele & G. H. Elder Jr. (Eds.), Methods of life course research: Qualitative and quantitative approaches (pp. 128-151). Thousand Oaks, CA: Sage.

Densley, J. A. (2012). Street gang recruitment. Social Problems, 59, 301-321. Ebaugh, H. R. F. (1988). Becoming an ex: The process of role exit. Chicago: University of Chicago Press.

Edelman, B. (2012). Using Internet data for economic research. Journal of Economic Perspectives, 26, 189–206.

Elder, G. H., Pavalko, E. K., & Clipp, E. C. (1993). Working with archival data: Studying lives. Newbury Park: Sage.

Eysenbach, G., & Diepgen, T. L. (1998). Towards quality management of medical information on the internet: Evaluation, labeling, and filtering of information. British Medical Journal, 317, 1496-1500.

Farrington, D. P. (1995). The development of offending and antisocial behaviour from childhood: Key findings from the Cambridge Study in Delinquent Development. Journal of Child Psychology and Psychiatry, 36, 929-964.

Farrington, D. P., Gallagher, B., Morley, L., St. Ledger, R. J., & West, D. J. (1990). Minimizing attrition in longitudinal research: Methods of tracing and securing cooperation in a 24-year follow-up study. In D. Magnusson & L. R. Bergman (Eds.), Data quality in longitudinal research (pp. 122–147). New York, NY: Cambridge University Press.

Featherstone, M. (2000). Archiving cultures. British Journal of Sociology, 51, 161-184.

Giordano, P. C. (2010). Legacies of crime: A follow-up of the children of highly delinquent girls and boys. Cambridge: Cambridge University Press.

Giordano, P. C., Cernkovich, S. A., & Rudolph, J. L. (2002). Gender, crime, and desistance: Toward a theory of cognitive transformation. American Journal of Sociology, 107, 990-1064.

Glueck, S., & Glueck, E. (1950). Unraveling juvenile delinquency. New York, NY: The Commonwealth Fund.

Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press.

Hagedorn, J. (1990). Back in the field again: gang research in the nineties. In C. Ronald Huff (Ed.), Gangs in America (pp.240-259). Newbury Park, CA: Sage.

Hampson, S. E., Dubanoski, H. P., Hamada, W., Marsella, A. J., Matsukawa, J., Suarez, E., & Goldberg, L. R. (2001). Where are they now?: Locating former elementary-school students after nearly 40 years for a longitudinal study of personality and health. Journal of Research in Personality, 35, 375–387.

Hansen, W. B., Tobler, N. S., & Graham, J. W. (1990). Attrition in substance abuse prevention research: A meta-analysis of 85 longitudinally followed cohorts. Evaluation Review, 14, 677–685.

Heckathorn, D. D. (1997). Respondent-driven sampling: A new approach to the study of hidden populations. Social Problems, 44, 174-199.

Heverin, T., & Zach, L. (2010). Twitter for city police department information sharing. Proceedings of the American Society for Information Science and Technology, 47, 1-7.

Holt, T. J., & Blevins, K. R. (2007). Examining sex work from the client’s perspective: Assessing johns using on-line data. Deviant Behavior, 28, 333-354.

Holt, T. J., Blevins, K. R., & Burkert, N. (2010). Considering the pedophile subculture online. Sexual Abuse, 22, 3-24.

Holt, T. J., Blevins, K. R., & Kuhns, J. B. (2008). Examining the displacement practices of johns with on-line data. Journal of Criminal Justice, 36, 522-528.

Hser, Y. I., Hoffman, V., Grella, C. E., & Anglin, M. D. (2001). A 33-year followup of narcotics addicts. Archives of General Psychiatry, 58, 503-508.

Hughes, L. A., & Short, J. F. (2005). Disputes involving youth street gang members: Micro-social contexts. Criminology, 43, 43-76.

Irwin, J. (1972). Participant observation of criminals. In J. Douglas (Ed.), Research on deviance (pp. 117-137). New York, NY: Random House. Jacobs, B. A. (1999). Dealing crack: The social world of streetcorner selling. Boston: Northeastern University Press.

Jacobs, B. A., & Wright, R. (2006). Street justice: Retaliation in the criminal underworld. New York: Cambridge University Press.

Jansyn, L. R. (1966). Solidarity and delinquency in a street corner group. American Sociological Review, 31, 600-614.

Johnson, K. D. (2003). Financial crimes against the elderly. Center for Problem-Oriented Policing. Retrieved from http://www.popcenter. org/problems/crimes_against_elderly

Klebanoff., M. A., Zemel, B. S., Buka, S., & Zierler, S. (1998). Long-term followup of participants in the Collaborative Perinatal Project: Tracking the next generation. Paediatric and Perinatal Epidemiology, 12, 334346.

Kleck, G., Tark, J., & Bellows, J. J. (2006). What methods are most frequently used in research in criminology and criminal justice? Journal of Criminal Justice, 34, 147-152.

Klein, M. W., & Maxson, C. L. (2006). Street gang patterns and policies. Oxford: Oxford University Press.

Laub, J. H., & Sampson, R. J. (2003). Shared beginnings, divergent lives: Delinquent boys to age 70. Cambridge, MA: Harvard University Press.

Letkemann, P. (1973). Crime as work. Englewood Cliffs, NJ: Prentice-Hall. Loeber, R., Farrington, D. P., & Stallings, R. (2011). The Pittsburgh Youth

Study. In R. Loeber & D. Farrington (Eds.), Young homicide offenders and victims: Risk factors, prediction, and prevention from childhood (pp. 19-36). New York: Springer.

Maruna, S. (2001). Making good: How ex-convicts reform and rebuild their lives. Washington, DC: American Psychological Association.

Miller, W. B. (1957). The impact of a community group work program on delinquent corner groups. Social Service Review, 31, 390-406.

Miller, W. B. (1962). The impact of a “total-community” delinquency control project. Social Problems, 10, 168-191.

Miller, W. B. (1963). City gangs: A report on the Midcity Delinquency Study. Boston University School of Social Work Journal, 1, 24-29.

Miller, W. B. (1973). The Molls. Society, 11, 32-35.

Miller, W. B. (2011). City gangs. Retrieved from http://gangresearch.asu.edu/walter_ miller_library/walter-b.-miller-book/city-gangs-book

Mody, L., Miller, D. K., McGloin, J. M., Freeman, M., Marcantonio, E. R., Magaziner, J., & Studenski, S. (2008). Recruitment and retention of older adults in aging research. Journal of the American Geriatric Society, 56, 2340–2348.

Moffitt, T. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674701.

Nobles, M. R., & Fox, K. A. (2013). Assessing stalking behaviors in a control balance theory framework. Criminal Justice & Behavior, 40, 737-762.

Ortiz, V., and Ballon, E. G. (2007). Longitudinal research at the turn of the century: Searching for the Mexican American people. Sociological Methods and Research, 36, 112-137.

Papachristos, A. V., & Smith, C. (2012). The small world of Al Capone: The embedded and multiplex nature of organized crime (Working Paper). Retrieved from http://ssrn.com/abstract=2159899

Piquero, A. R., Farrington, D. P., Shepherd, J. P., & Auty, K. (2011). Offending and early death in the Cambridge Study in Delinquent Development. Justice Quarterly.

Police Executive Research Forum. (2013). Social media and  tactical considerations for law enforcement. Retrieved from http://policeforum.org/library/technology/SocialMediaandTacticalConsiderationsforLawEnforcement.pdf

Polsky, N. (1969). Hustlers, beats, and others. Garden City, NJ: Anchor.

Pyrooz, D. C., Decker, S. H., and Moule Jr., R. K. (2013). Criminal and routine activities in online settings: Gangs, offenders, and the Internet. Justice Quarterly.

Reid, E., & Chen, H. (2002). Internet-savvy US and Middle Eastern extremist groups. Mobilization: An International Quarterly Review, 12, 177-192.

Root, J., Smith, K. R., Whelan, E. A., Sandler, D., & Voda, A. M. (1994). Tracing women over half a century: Strategies to locate subjects lost to follow-up in a longitudinal health study. Research on Aging, 16, 375388.

Sampson, R. J., & Laub, J. H. (1993). Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press.

Schafer, J. (2002). Spinning the web of hate: Web-based hate propaganda by extremist organizations. Journal of Criminal Justice and Popular Culture, 9, 69-88.

Schubert, C., Mulvey, E., Cauffman, E., Steinberg, L., Losoya, S., Hecker, T., Chassin, L., & Knight, G. (2004). Operational lessons from the Pathways to Desistance Study. Youth Violence Juvenile Justice, 2, 237–225.

Selepak, A. (2010). Skinhead Super Mario brothers: An examination of racist and violent games on White supremacist web sites. Journal of Criminal Justice and Popular Culture, 17, 1-47.

Shover, N. (1983). The later stages of ordinary property offender careers. Social Problems, 31, 208-218.

Shover, N. (1985). Aging criminals. Beverly Hills, CA: Sage

Soothill, K., Humphreys, L., & Francis, B. (2012). Middle-class offenders: A 35-year follow-up. British Journal of Criminology, 52, 765-785.

Sundeen, R. A. (1977). The fear of crime and urban elderly. In M. A. Young-Rifai (Ed.), Justice and older Americans. Lexington, MA: Lexington.

Suttles, G. D. (1968). The social order of the slum: Ethnicity and territory in the inner city. Chicago: University of Chicago Press.

Tewksbury, R., & Gagne, P. (1997). Assumed and presumed identities: Problems of self-presentation in field research. Sociological Spectrum, 17, 127-155.

Thornberry, T. P., & Krohn, M. D. (2002). Taking stock of delinquency: An overview of findings from contemporary longitudinal studies. New York, NY: Kluwer.

Tracy, P. E., & Kempf-Leonard, K. (1996). Continuity and discontinuity in criminal careers. New York: Plenum.

Vaillant, G. E. (2002). Aging well. Boston, MA: Little Brown.

Ward, G. (2012). The Black child savers: Racial democracy and American juvenile justice. Chicago: Chicago University Press.

Watkins, A. M., & Moule, R. K. (2013). Older, wiser, and a bit more badass?: Exploring differences in juvenile and adult gang members’ gang-related attitudes and behaviors. Youth Violence and Juvenile Justice. Forthcoming.

Watters, J. K., & Biernacki, P. (1989). Targeted sampling: Options for the study of hidden populations. Social Problems, 36, 416-430.

Wells, W., Cavanaugh, M. R., Bouffard, J. A., & Nobles, M. R. (2012). Nonresponse bias with a web-based survey of college students: Differences from a classroom survey about carrying concealed handguns. Journal of Quantitative Criminology, 28, 455-476.

Whyte, W. F. (1943). Street corner society. Chicago, IL: University of Chicago Press.

Wolfgang, M.E., Thornberry, T. P., & Figlio, R. M. (1987). From boy to man, from delinquency to crime. Chicago, IL: University of Chicago Press.

Wright, R., & Bennett, T. (1990). Exploring the offender’s perspective: Observing and interviewing criminals. In K. Kempf (Ed.), Measurement issues in criminology (pp. 138-151). New York, NY: Springer-Verlag.

Wright, R., Decker, S. H., Redfern, A. K., & Smith, D. L. 1992. Snowball’s chance in hell: Doing fieldwork with active residential burglars. Journal of Research in Crime and Delinquency, 29, 148-161.

Wright, R. & Decker, S. H. (1994). Burglars on the job. Boston: Northeastern University Press.

Wright, R. & Decker, S. H. (1997). Armed robbers in action. Boston: Northeastern University Press.

Zhou, Y., Reid, E., Qin, J., Chen, H., & Lai, G. (2005). US domestic extremist groups on the web: Link and content analysis. IEEE Intelligent Systems, 20, 44-51.

Zimmer, M. (2010). “But the data is already public”: On the ethics of research in Facebook. Ethics and Information Technology, 12, 313-325.

Contributors

Richard k. Moule Jr. is a doctoral student in the School of Criminology and Criminal Justice at Arizona State University. He received his BS in Criminology from The College of New Jersey and his MS in Criminology and Criminal Justice from Arizona State University. His research interests include gangs and deviant networks, life course criminology, and the intersection of technology and criminological theory. He has recently co-authored articles appearing in Social Science Research, Youth Violence and Juvenile Justice, and Justice Quarterly. He is currently the archivist for the Walter B. Miller Library, housed in the School of Criminology and Criminal Justice.

Scott H. Decker is foundation professor of the School of Criminology and Criminal Justice at Arizona State University. He received his BA in social justice from DePauw University and MA and PhD in criminology from Florida State University. His main research interests are in the areas of gangs, criminal justice policy, and the offender’s perspective. His most recent books include European Street Gangs and Troublesome Youth Groups (Winner of the American Society of Criminology, Division of International Criminology Outstanding Distinguished Book Award, 2006), Drug Smugglers on Drug Smuggling: Lessons from the Inside (2008, CHOICE Academic Press Book of the Year), and Criminology and Public Policy: Putting Theory to Work (2010, with Hugh Barlow).

Acknowledgments

This material is based upon work supported by the National Science Foundation under Grant No. SES-1228472. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. The authors thank Gary Sweeten for providing useful references to life-course research, and to the editor and anonymous reviewers who provided feedback on the manuscript.