Federal Reserve Bank of Dallas

Contemporary JavaScript-based web applications generate huge volumes of interactivity and connectedness data that is very sensitive and vulnerable to tight privacy and governance policies. To establish secure and responsible data management in such distributed environments, the approach requires a combination of decentralized control, privacy safeguards and effective collaborative learning. In this research, a hybrid federated learning/blockchain architecture, PSOAdam Federated Blockchain Governance (PA-FedGov) is proposed and aimed at meeting verifiable and privacyprotecting data governance. The new hybrid optimization approach is a combination of Particle Swarm Optimization (PSO) algorithm of adaptive global aggregation and the Adam optimizer of local model optimization which leads to faster convergence and better generalization on non-independent and identically distributed data sources. To achieve immutable, auditable and transparent governance mechanisms, smart contracts are used by the blockchain layer to store model provenance, different privacy budgets, and compliance metadata. To conduct experimental evaluations, the system utilized Benchmark datasets such as FEMNIST to identify user specific handwritings, CICIDS 2017 to detect intrusion in web systems and the Edge-IIoTset to analyze and govern edges, and IoT. These datasets are used in the publicly available CSV or JSON formats. The findings show that the suggested PAFedGov model is more accurate and more stable by up to 12 percent when compared to the conventional federated learning approaches like FedAvg and FedProx, with the same set of differential privacy restrictions. The architecture that ensures traceability and decentralized policy enforcement is blockchain based, and it provides a scalable and reliable foundation of safe data control within federated and privacy-sensitive JavaScript web environments.

We use emergency outages of coal generators as an exogenous source of variation in the power generation stack to study how changes in marginal fuel affect real-time price volatility. Contrary to anecdotal evidence, we find that wholesale prices are less volatile when natural gas is on the margin more often. JEL Classification: Q4

Abstract Prior studies show that women are more willing to accept lower wages for shorter commutes than men. We show that gender differences in commuting preferences lead to a gender wage gap only if there is a wage penalty for shortening commutes, determined by the geography of jobs. We demonstrate this by showing that the commuting and wage gaps are considerably smaller among workers living near city centers, especially for occupations with a high geographic concentration of high-wage jobs. We highlight the geography of jobs as a key force that amplifies the impact of commuting preferences on the gender wage gap.

We study how random variation in the availability of highly educated, foreign-born workers impacts firm performance and recruitment behavior. We combine two rich data sources: 1) administrative employer-employee matched data from the US Census Bureau; and 2) firmlevel information on the first large-scale H-1B visa lottery in 2007. Using an event-study approach, we find that lottery wins lead to increases in firm hiring of college-educated, immigrant labor along with increases in scale and survival. These effects are stronger for small, skill-intensive, and high-productivity firms that participate in the lottery. We do not find evidence for displacement of native-born, college-educated workers at the firm level, on net. However, this result masks dynamics among more specific subgroups of incumbents that we further elucidate.