Investments in police and punitive sanctions can reduce crime, but they are costly solutions and can lead to harmful consequences, mainly for historically marginalized communities. The recent protests against police use of force and racism, compounded with the pervasiveness of unfading concentrated disadvantage, have pressed policymakers to explore alternative crime prevention strategies. Community investments have long been considered promising alternatives. Still, there is limited research on which are most effective in improving public safety, particularly on scalable interventions that leverage private investments. This dissertation contributes to closing this knowledge gap. The first chapter examines whether a prominent place-based capital investment policy –Opportunity Zones– influences public safety using two quasi-experimental designs. Four years after its implementation, there are no neighborhood changes –urban development, property prices, poverty, employment, and income–nor have not impacted public safety –calls for service, police stops, crimes, and arrests. Investing in disadvantaged areas is crucial, but it should be well-targeted to the neighborhoods' needs and physical design. The second chapter focuses on how expanding residential lending can reduce neighborhood crime. By leveraging differential exposure to banks' local market share and common national mortgage shocks, the research finds that when banks make more home loans, communities experience a public safety improvement. Black, Hispanic, and poor communities benefit more, and there is no evidence of gentrification. Overall, private investments can provide benefits beyond their intended recipients and be a promising alternative to reducing crime. The third chapter studies the long-term public safety impacts of the once-legal 1930s racially discriminatory maps that limited residential loans to racial-minority creditworthy individuals. Comparing areas near different color-grade boundaries around a small bandwidth reveals significant crime increases in redlined areas. Changes in arrests seem to concentrate on property and low-level offenses. Areas labeled green experienced public safety improvements and fewer arrests for violent and non-serious offenses and police stops. Long-term structural disinvestment is a driving factor preventing safer neighborhoods. In summary, this dissertation provides evidence that some place-based investments can improve public safety while others may be inadequate, at least in the short term, to foster neighborhood changes.