The advancement of information and communication technology (ICT) has significantly transformed how people communicate and obtain information in both their personal and professional lives. In the realm of digitally-facilitated social interaction, this dissertation calls for attention to the importance of information technology artifacts (IT artifacts) on social welfare by showcasing that the careful design and usage of IT artifacts have the potential to enhance the effectiveness, efficiency, and fairness of social interactions. The first study centers around addressing the cold-start issue that often arises when new products are introduced. Specifically, I investigate how machine-generated content can enhance the equity of new products. Analyzing data from Kaggle.com, my research demonstrates that the use of machine-generated content is effective in tackling the cold-start problem by increasing the adoption of the product in the initial phase. Additionally, my findings reveal that machine-generated content can also reduce information asymmetry for users regarding the datasets or associated providers. As a result, these outcomes provide strong evidence supporting the use of machine-generated content to enhance equity in online communities. The second research investigates the impact of a platform’s decision to impose application fees on enhancing the quality of matching results in an online labor market. Based on data obtained from Freelancer.com, my analysis demonstrates that the implementation of application costs serves as a motivator for workers to submit fewer but more selective bids. This, in turn, increases the likelihood of employers offering contracts, as workers are less likely to apply casually or without much thought. Overall, these results indicate that application costs can enhance the efficiency of the matching process. In the third study, I examine whether gender differences exist in telework adjustment as a response to disasters and to what extent such adjustments can help reduce gender inequality, using the COVID-19 pandemic as an example. The study's findings reveal the following: 1) Female workers exhibit a higher rate of telework adjustment than their male counterparts by 7% after accounting for differences in job sorting, and female workers are more responsive to external constraints. 2) Telework adjustment can help mitigate gender inequality in the labor market.

Biases in online platforms pose a threat to social inclusion. I examine the influence of social biases on online platforms. In my dissertation, I conduct empirical studies on online crowdfunding platforms (prosocial lending and educational crowdfunding) to investigate the influence of funders' or recipients' social backgrounds on the funding dynamics. In the first study, I examine the influence of a novel source of bias in online philanthropic lending, namely that associated with religious differences. I further propose a set of contextual moderators that characterize individuals’ offline (local) and online social contexts, which I argue combine to determine the influence of religion distance on lending activity. In the second study, I theoretically and empirically explore the role of value homophily in shifting lending priorities in online pro-social platforms. Considering the full spectrum of cultural influences, I develop the concept of “culturalist choice homophily,” where value-based similarities emerge based on the culturally-motivated behaviors and “historicist choice homophily,” where value-based similarities emerge based on similarities in historical-cultural barriers. Further, I introduce a novel content-context value congruence perspective for crisis fundraising, where the synergy between a borrowers’ request reasoning and the optimal crisis outcome determines the volume of lending received by crisis victims. I utilize the Arab Spring crisis in a Difference-in-Difference (DID) setting to test my hypotheses. Finally, in the third study, I add to the recent literature on the impact of the design of educational crowdfunding in alleviating inequality for public schools' fundraising. I particularly explore the effects of the platform intervention in terms of signaling students’ need to alleviate biases toward racially and economically disadvantaged students. Utilizing data from DonorsChoose.org, I first show that the online platform cannot automatically make up for all biases, especially toward classrooms with students with a higher level of poverty or racially marginalized communities. Further, I show that labeling projects as equity-focus can alleviate biases. However, the results are heterogeneous across different sources of identity. In particular, I discuss that equity-focus labeling has a greater impact on improving inequality toward hard-to-observe identities, e.g., economically disadvantaged students, than easy-to-observe identities such as racially underprivileged communities.