What I do
My research focuses on understanding the context of socioeconomic systems and analyzing the impact of context on individual decision-making. Contexts in social systems encompass social norms, economic regimes, infrastructure, and government policies. These social contexts have a significant impact on individual decision-making. Next, individuals in socioeconomic systems have heterogeneous characteristics such as demographic information, economic conditions, and social networks. Heterogeneous characteristics of individuals also affect their decision-making along with the context. To better understand socioeconomic systems, I believe it is essential to integrate contexts and heterogeneous individuals into a single model. For this reason, I consider an agent-based modeling framework to be a suitable methodology for my research.
Using an agent-based modeling framework, I replicate abstract socioeconomic systems and their contexts in computational models. Also, I apply real-world micro-level data to implement heterogeneous agents in the model. In my Ph.D. research, I modeled and implemented several socioeconomic systems, including macroeconomic systems, housing markets, and urban transport systems. By simulating the implemented agent-based model, I demonstrate the influence of context on individual decision-making, which further leads to aggregated statistics.
1. Agent-based Computational Economics
The first line of research interest is to understand the context of the current economic system and examine the impact of that context on individual inequality. Among the economic contexts, I focused on a new financial context called quantitative easing (QE). My research investigates the impact of financial quantitative easing on individuals' economic decision-making and further on economic inequality.
In my paper titled “More Finance Makes It More Equal?”, we model the Korean housing market with an agent-based simulation. Through simulation, we confirmed that quantitative easing of housing finance gives different benefits according to household income quintiles. Furthermore, we verified that excessive quantitative easing of housing finance aggravates household wealth inequality. As a policy implication, we identified that appropriate macro-prudential regulation lowers asset prices and improves wealth inequality.
Next, the study titled “Finance and Decreasing Business Dynamics: Using Macroeconomic Agent-Based Simulation” models the macroeconomic situation of Korea based on the Keynes meeting Schumpeter (K+S) model framework. From the simulation, we confirmed that the quantitative easing of corporate finance contributes to more favorable market conditions for large enterprises than for SMEs. The results also imply that exaggerated corporate finance increases market concentration and lowers the labor income share. In both studies, we found that the impact of financial quantitative easing on individuals differs according to their characteristics, and this difference leads to economic inequality.
2. Urban Digital-Twin Model
The second line of research interest is urban digital twin modeling. Going beyond digitizing and visualizing urban infrastructure, we implemented the resident agent in a digital twin model and try to understand the behavior and movement patterns of individuals within the infrastructure. Through this model, we obtain policy implications for urban administration, such as forecasting demand for urban infrastructures.
In the paper titled “Agent-Based model for Urban Administration: Focusing on Traffic”, we implement Sejong city infrastructure on a real scale, including buildings, roads, and public transport systems. We also use full-scale demographic information of citizens to create a residential agent in the model. We generated the behavioral patterns of these agents based on time-use survey data. Next, by constraining the type of building required for each behavior, we model the movement demands of residential agents. Using the model, we test the traffic dispersion effect according to the new bridge construction locations. We also conduct the research titled “Agent-based modeling and simulation on residential population movement patterns: the case of Sejong city” to predict public transport demand using the same city model.
3. Financial Network Analysis
The third line of my research interest is to understand the financial risk of individual institutions in the context of the financial network. As modern finance becomes intricate, counterparty information in the financial network materializes as a new risk factor. My paper titled "Too-central-to-fail systemic risk with PageRank algorithm" suggest a new metric that specifies systemically important financial institutions (SIFIs) in consideration of network information. Since network information between financial institutions is not subject to disclosure, we estimate it from stock market data using the granger causality methodology.
Another paper titled “Counterparty network Influence: It Matters where the capital comes from” shows that the capital flow network information has significance in explaining extreme capital flow episodes in emerging countries. This result implies that the position of counterparties in the capital flow network should be managed as a financial stability indicator in emerging countries.
A unifying thread in all of my research is that microscopically expands the understanding of socioeconomic systems. This expanded understanding enables us to generate microscopic indicators and conduct more realistic counterfactual policy experiments. It also creates a demand for micro-level data in the recent trend of increasing data in various forms. I believe that this expansion should also occur in other socioeconomic domains and that there is a contribution to be made.