In the Criminogenic Family Fold: Investigating the Predictive Power of Family Management Practices in Understanding Youth Crime in Nairobi and Mombasa Counties, Kenya

Abstract

There has been a rise in juvenile delinquency worldwide, particularly in developing countries such as Kenya. Failures in interventions point to the possible inefficacy of the measures, misdiagnosis, or scanty analysis of the security problem. This was precipitated by mixed results and insufficient research. There has been therefore a need to comprehensively confirm, analyze, and document the relationships between various criminogenic situational predictors and juvenile delinquency in Kenya. The purpose of this study thus was to examine various criminogenic situational predictors and juvenile delinquency in Nairobi and Mombasa counties in Kenya. For this purpose, the researcher evaluated the hypothesized predictors on both non-delinquents and delinquents to establish clear relations. The study was anchored on Social Learning and Social Disorganization theories. The study adopted a Concurrent Nested mixed-method Research approach with a qualitative component being embedded in a quantitative Causal-Comparative Research Design. The target population of the study was 2,908,950 juveniles in the two counties. The accessible population was 235,861 respondents, out of which a sample of 400 was drawn, from which the researcher got 360, a response rate of 90%. The researcher drew the sample using a disproportionate stratified random sampling thus ensuring representation of both delinquents and non-delinquents in each county. The delinquent population entailed both delinquents who had committed minor violations and delinquents who had committed serious violations. Delinquents who had committed serious violations were drawn from borstal institutions, while those who had committed minor violations were drawn from the Probation Department. The non-delinquents were drawn from county secondary schools in each of the counties. Random sampling was done using the Excel data analysis tool pack. In addition, 12 respondents were selected purposively from the authorities dealing with children matters, 6 from each county. The total sample size attained was thus 372 respondents. A face-to-face interview questionnaire and an in-depth key informant interview schedule were the main instruments of data collection. Binary logistic regression was conducted on the quantitative data at a 95% confidence interval (CI) and a p-value < 0.05 considered significant with the help of Statistical Package for Social Sciences (SPSS) version 26.0 software. Thematic content analysis was conducted on the qualitative data with the help of Max. Qda software. The odds ratio indicates that when holding all other variables constant, a child is 7.2 times more likely to turn out delinquent with poor parental characteristics, 2.3 times more likely to turn out delinquent with poor family management practices, 3 times more likely to turn out delinquent with poor parent-child attachment and 2.2 times more likely to turn out delinquent with poor neighborhood characteristics than turn out non-delinquent. The odds ratio for combined situational predictors indicates that when holding all other variables constant, a child is 67.5 times more likely to turn out delinquent than a non-delinquent with poor situational conditions. Therefore, employing a .05 criterion of statistical significance, the null hypothesis was thus rejected because the findings show a statistically significant predictive relationship between combined situational predictors and juvenile delinquency. The findings will be useful to the government, academia, policy actors as well as parents in developing crime prevention policies, contributing to theory and literature, informing on better family management practices, and informing effective tailoring of relevant social policies respectively.