Poverty reduction policies persist in playing an essential role in the context of a developing country like India. The role of social assistance programs in poverty reduction is recognised in the Sustainable Development Goals (1.3). In this thesis, I study about Indira Gandhi National Old Age Pension program, a social assistance program targeted at low-income elderly in India. Given that there are very few institutionalised social security arrangement focussed on the poor, this thesis examines about a program that pioneered the cause of providing a social security arrangement for the poor elderly. In this thesis, I explore different aspects of the program, and I also examine the impact of this program on improving welfare. There are three essays in this thesis. The first essay unpacks the role of political and social connections in providing access to the scheme and also targeting errors in the program. This essay underscores the importance of connections in the distribution of the welfare programs in India. I have used an individual level panel data from the Indian Human Development Survey, and have used a panel probit and recursive bivariate probit method for this essay. There are two significant findings from this paper. The first findings from this essay suggest that the political connection of the household increases their chances of receiving the program. The second finding is that political association perpetuates type-1 targeting error in the program. In the second essay, I evaluate the impact of the program on short-term and long-term household welfare. I have used a household level panel data from the Indian Human Development Survey for this purpose. I have employed propensity score matching to build a valid counterfactual group, and have applied a fixed effects regression in the propensity score match setting. In the last stage, I have also used an instrumental variable to validate the robustness of the findings further. The results from this paper provide evidence that the program improves several dimensions of household welfare that includes monthly per capita spending, food, and non-food expenditures (short-term). The program also has a positive effect on improving the long-term household welfare (household assets); the results also indicate that the scheme has a small negative impact on household labour supply. The last essay analyses the impact of the program when women constitute as program recipients. I have used the same household panel data, and have adopted the same empirical framework used in the second essay for this paper. The results from this essay confirm that the program has a positive impact on the consumption expenditure incurred on nutrient-rich food and increases education and medical spending of the household when women receive the welfare program. The overall finding from this thesis provides conclusive evidence on the role of social assistance programs in improving various welfare dimensions, and thereby reducing household poverty. The success of a welfare program is conditioned on several factors that include the implementation strategy, impact of the program, and how the payment from the program is spent. In the first essay, I explore the implementation strategy by probing into the issue of targeting errors in the program. The second and the third essay investigates the impact of the program by assessing the use of pension payments.