Research

The opioid crisis has had adverse public health effects, with the White population disproportionately affected in its initial phase, yet our understanding of the effects of the crisis on neighborhood inequality remains underdeveloped. This study examines how opioid-related policies transform neighborhood poverty by analyzing the 2010 reformulation of OxyContin---a supply-side intervention to reduce prescription opioid abuse. It argues that the reformulation inadvertently prompts impoverished opioid users to switch to heroin and relocate to neighborhoods with heroin availability, leading to the emergence of White high-poverty neighborhoods. Using a continuous difference-in-differences design, it finds that following the reformulation, White high-poverty neighborhoods emerge at higher rates in counties with initially high opioid dispensing rates compared to lower-rate counties. This effect is absent in non-White neighborhoods and amplified in neighborhoods with initially high housing vacancy. Within high-vacancy areas in high-opioid counties, heroin availability increases post-reformulation, and the resulting high-poverty neighborhoods stem from the relocation of the poor rather than increasing the overall poor population. Taken together, the post-reformulation emergence of White high-poverty neighborhoods is due to a sorting process into areas where heroin becomes available. The study highlights the unintended consequences of supply-side drug policy on the geographic distribution of poverty along racial lines.

This study investigates how permanent shocks to legal sectors where specific criminal groups operate affect the evolution of organized crime as a whole, both in the short and long run. It focuses on Japanese organized crime, the yakuza, and the 1960s energy transition from domestic coal to imported oil. The transition reduced coal mining jobs, negatively impacting both coal miners and yakuza groups involved in the coal mining industry. The study shows that as coal mining declined, mining-associated yakuza members substituted to income-generating crime and joined non-mining yakuza groups rather than coal miners joining the yakuza. New business opportunities expanded in nearby non-mining municipalities over time, which yakuza groups exploited to expand their territories. Coal-rich areas experienced persistent violent yakuza conflicts lasting over fifty years. Today, areas with greater mining job losses have more yakuza members and groups. However, yakuza exclusion ordinances enacted in the 2010s, which reduce demand for yakuza services and restrict the substitution of their economic activities, help partially break this persistent cycle.

In the wake of demographic shifts in the United States, understanding how changing racial and ethnic compositions affect residential segregation patterns is paramount. We examine a critical yet overlooked dimension of increased diversity: the relative size ranking of Black, Hispanic, and Asian populations within counties. Using U.S. census data from 1990, 2000, and 2010, we demonstrate that segregation---whether in levels or trends---changes at thresholds where minority group population rankings shift. Controlling for absolute population size, we find that both between-group rank position and within-group rank changes over time increase residential racial segregation. These rank effects are more pronounced when the gap between White and minority population sizes narrows, consistent with group threat theory. Our findings explain the persistence of segregation across different minority groups and temporal variation in segregation patterns within the same locations.

Despite being the least suburbanized group, Black suburbanization doubled from the 1970s to the 2010s. This study traces a portion of the rise of Black suburbanization to the emergence of crack cocaine in mid-1980s and early 1990s. It argues that the outbreak of the crack epidemic---a race- and location-specific negative shock that severely affected inner-city Black communities---triggered migration capacity, thereby driving the long-term Black flight to the suburbs. Analyses using metropolitan, individual, and census tract level data collectively produce the following results. Two decades post-crack emergence, the trend of Black suburbanization still escalated at a greater rate in high crack exposure areas. Black suburbanization increased particularly in areas with a high concentration of initial migration flows, where Black migrants from their origin cities relocated to a select few suburban destinations. Black suburbanization following crack exposure was not limited to inner-ring suburbs or disadvantaged neighborhoods. Quantifying the magnitude, the study reveals that the crack epidemic led to the suburbanization of nearly 1 million Black people. The study demonstrates the understudied role of the crack epidemic in shaping the racial geography of the United States in the long run.

This study investigates the effects of an indirect economic sanction on organized crime. While research traditionally focuses on direct enforcement, where law enforcement agencies actively suppress criminal organizations, we study the Yakuza Exclusion Ordinances (YEOs), which adopt an indirect strategy by prohibiting non-yakuza citizens from providing material benefits to Japanese organized crime, the yakuza. We investigate the collateral consequences of the YEOs on organized fraud, a white-collar crime that accounts for nearly half of all property crime damage in Japan. We find that the YEOs increase the financial damage associated with organized fraud, with the effects being greater in regions with lower levels of yakuza competition, but do not significantly affect the number of fraud cases. Additional evidence suggests that former yakuza members, who resign in response to the YEOs, as well as current members, whose traditional income sources are restricted under the YEOs, engage in organized fraud. Our model rationalizes the findings, showing that larger-scale fraud schemes increase financial damages without necessarily raising the number of cases.