A large share of public expenditure goes towards building public infrastructure. Often, the infrastructure is targeted particularly to groups that have previously been under-served. Sometimes, there is no targeting but infrastructure improvements end up benefiting these groups more because of pre-existing inequalities. Indian society is deeply unequal along several such group lines: caste, religion, gender, language, and more. Therefore, group identity can be a pivotal determinant of access to goods, services, and infrastructure. Through this dissertation, I contribute to the big picture question: What is the relationship between group identity, and access to goods and infrastructure? We know from existing causal evidence that group identity plays an important role in determining economic outcomee, improving schooling infrastructure can lead to long-lasting impacts on educational attainment and workforce participation and household water access can have important consequences for health and overall welfare. My thesis contributes to both strands of the literature by evaluating two (previously unevaluated) nationwide infrastructure projects and measuring their effect on under-served groups (religious minorities and women). Many studies in the past have been constrained by a lack of granular nationwide data on many measures in India. There have been recent efforts to alleviate this problem. My thesis also contributes to this effort by creating two new granular nationwide datasets for public use. I collect and digitize data on infrastructure spending and construction in India from the last decade, that has not been used before. I also put together nationwide consumer data at a granular level to study identity-based goods consumption. In Chapter 1, I evaluate an infrastructure policy in India that invests in regions with a high share of religious minorities. I use an event study design to measure the medium-term effect of these investments on schooling infrastructure and schooling outcomes. I find that there is an improvement in classroom infrastructure (as targeted by the policy), but this improvement doesn't lead to an improvement in schooling outcomes. I also use an alternate approach to study the effect of the policy on regions with close to 25\% share of religious minorities using an RD design. In Chapter 2, I study the effect of a rollout of household water access points (pipes, wells, tanks, pumps, etc.) on educational outcomes for children, especially girls. I use an event study design to measure the medium-term effect of receiving access to a substantial volume of water on student enrollment. I find that there is a large increase in student enrollment and retainment over time. I see a slight increase in share of girls enrolling in school but only 3 years after treatment; I discuss reasons as to why this could be the case. In Chapter 3, I ask if religious minority groups make different consumption choices based on the religious identity of their neighbors. I use National Sample Survey data to measure these patterns. I find that Muslims who live amongst Hindu neighbors silence their Islamic identity more than those who live amongst Muslim neighbors. I don't see this difference for Hindus.