Mathis Preti*, Kla Kouadio**

*Between 1954 and 1998, the tobacco industry spent $355 million on academic research. I show that the industry's funding decisions delayed research on smoking’s harms, despite strict protections against industry interference. Using large language models to process internal tobacco-industry documents, I build the first comprehensive dataset of academic researchers who applied for tobacco-industry grants and link them to bibliometric records. I first show that tobacco-health research projects were systematically less likely to be funded than otherwise similar projects. Using difference-in-differences, I then compare the publications of researchers who received a grant with those of researchers whose applications were rejected or nearly approved. Only grants supporting tobacco-health research increased tobacco-health publications. Moreover, during the period when the industry penalized these projects in funding decisions, the publications they induced mostly established smoking’s harms. By reallocating funding away from tobacco-health research, the industry therefore delayed the production of this knowledge. These findings provide the first empirical evidence that an industry can use research funding to delay the production of inconvenient knowledge.

**Between 1970 and 2000, predominantly Black urban neighborhoods in the United States experienced sharp socioeconomic decline, even as Black incomes rose in the country as a whole. Over the same period, more than six million African Americans moved from urban areas to the suburbs, a process of Black suburbanization disproportionately driven by higher-income Black households. This selective out-migration plausibly reduced economic resources and exposure to higher-income peers in the neighborhoods left behind. This paper provides evidence on the long-run consequences of this deterioration for the children who grew up in urban areas. To identify the effect of Black suburbanization on children's outcomes, I use an instrumental variable based on the location of land unavailable for new construction near preexisting suburban Black communities, which shaped where departing households predominantly moved. I find that Black suburbanization reduced upward mobility, employment, and marriage rates among children raised in these urban neighborhoods. I further provide evidence consistent with these effects operating in part through rising violent crime, declining house values, and the departure of experienced teachers in the neighborhoods left behind. These findings support the idea that Black suburbanization, while reflecting positive individual economic mobility, worsened the economic prospects of children whose parents were unable to leave.