Examination of New Jersey alcohol and tobacco outlet density in relation to socio-demographics, unemployment and poverty rates, using geospatial and random effects models

This study investigates whether New Jersey alcohol and tobacco retail outlets per zip code are more highly prevalent in areas with persons who have lower educational attainment, are Black or minority status, and earn less than median income. Estimated poverty and unemployment rates from the US Census are also mapped by zip code to each region with tobacco and alcohol vendors. Using t-test, alcohol packaged goods stores are compared to bars/taverns for differences among socio-demographic characteristics. Geospatial analysis exhibits a correlation among lower educational attainment, high estimated poverty and unemployment percentages in areas with disproportionately high alcohol and tobacco dispensaries. Thus, are these determinants of social disadvantage an artifact of population density or the result of targeted ploys to entice less educated and poor into alcohol and tobacco use? Two-proportion z-test demonstrates that alcohol and tobacco outlets are located in more densely inhabited areas, with both high and low minority population, and unemployment and poverty rates near State averages. Separate regression analyses performed on the two alcohol outlet density groups reveal that lower education and income are predictors of purchasing liquor in stores versus bars. A random effects model is also tested to ascertain if poverty level for each zip code, in conjunction with a ratio of geo-coded poverty rate compared to the State level, affects the distribution type of alcohol and tobacco outlets by zip code. Thus, this research addresses policy question if reducing alcohol and tobacco outlets in areas with higher poverty and unemployment levels will decrease consumption.