作者: Noli Brazil , David S. Kirk
DOI: 10.1093/AJE/KWW062
关键词:
摘要: Uber and similar rideshare services are rapidly dispersing in cities across the United States beyond. Given convenience low cost, has been characterized as a potential countermeasure for reducing estimated 121 million episodes of drunk driving 10,000 resulting traffic fatalities that occur annually States. We exploited differences timing deployment US metropolitan counties from 2005 to 2014 test association between availability Uber's total, driving-related, weekend- holiday-specific 100 most populated areas using negative binomial Poisson regression models. found given county had no with number subsequent fatalities, whether measured aggregate or specific drunk-driving during weekends holidays.