This dissertation explores how digital technologies help mental health outcomes at both community and individual levels, offering a multi-level perspective on digital mental health support. The first essay investigates the impact of hospital telehealth adoption on county-level mental health outcomes across the United States. Guided by social capital theory and employing causal forest modeling on a panel dataset of over 2,800 counties, the study reveals that telehealth's effectiveness varies significantly by community context. Counties with limited broadband access but high economic or cultural capital see the greatest improvements, highlighting the need for complementary community resources to realize the full benefits of digital infrastructure. The second essay examines how anonymity affects user engagement in online mental health discussions. Leveraging a natural policy change on Quora and grounded in signaling theory, it demonstrates that anonymity leads to substantial increases in user participation, content specificity, and emotional expression—particularly among users with fewer disclosed credentials—underscoring the role of platform design in empowering marginalized voices. Together, these essays demonstrate how digital technologies can improve mental health outcome and foster inclusive mental health discourse.