[1] Tying in Two-sided Markets with Below-Cost or Negative Pricing [PDF]
-Formerly, Tying in Two-sided Markets with Below-Cost or Negative Prices
-Co-authored with Jong-Hee Hahn (Professor at Yonsei Univ.)
-Presented at Kyoto-Yonsei Joint Workshop on Economic Theory; Yonsei-Ewha-Sogang Young Economists Seminars; 6th Student Workshop in School of Economics
<Abstract>
We show that tying can be profitable in two-sided markets even if below-cost or negative pricing is allowed. With the coexistence of two consumer groups (one treating tying and tied goods as perfect complements and the other as independent products), a tying-good monopolist may face difficulties in extracting rent and wish to use tying to directly capture the large advertising revenue created in the complementary segment. Such tying normally reduces consumer surplus and total welfare. Our theory of tying can be applied to the practice of self-preferencing or requiring pre-installation as in the Google Android and Shopping cases.
[2] A Simple Correction of Gini Estimates Using Tax Return Data, The Review of Social & Economic Studies [PDF]
Originally written in Korean. The translated version available on request.
-Formerly, Corrected Gini coefficient Based on Tax Data
-Co-authored with Taehwan Kim (Professor at Yonsei Univ.) and Juhyung Shin (Ph.D. student at Columbia Univ.)
-Presented at 2019 Summer Academic Congress of the Korean Econometric Society; 5th Junior Scholar Club Annual Conference; 5th Student Workshop in School of Economics
<Abstract>
The Gini coefficient is typically estimated using income data collected by surveying a group of randomly selected people. It is well known that people tend to underreport their income when asked about their income. As a result, the Gini coefficient is likely to be underestimated. This paper investigates how to correct such a prevalent underestimation problem occurring when estimating the Gini coefficient using survey data. We calculate top 1% income shares using tax return data, and use these results to correct the problem. Using our correction method, we have found the following: (i) the official Gini estimates announced by the National Statistical Office are underestimated, (ii) the degree of income inequality might have become worsened over time than described by the official Gini estimates, and (iii) the extent of underestimation is likely to be larger for South Korea than other OECD countries such as the US.
[3] Which rule can survive? Self-stability of a constitution [PDF]
-Formerly, Which rule can survive?: Self-stability of the constitution
-Co-authored with Semin Kim (Professor at Yonsei Univ.)
-Presented at 7th Student Workshop in School of Economics
<Abstract>
Society has two layers of rules. The first layer decides whether to implement a new policy, while the second maintains or revises the first layer. In a self-stable constitution, the first layer will not be replaced by any alternative rule under the second layer. Allowing heterogeneity between the first and the second layer, we discover the conditions for a self-stable constitution. Our results indicate that self-stable constitutions require that the two layers share the coalition with sufficiently large voting power. In reality, it may explain why those that are influential in reforming the law, also prevail when implementing policy.
[1] Searching for Friends
-Co-authored with Euncheol Shin (Professor at KAIST)
Without any addition assumption except the low-rank assumption, our paper imputes missing data in adjacency matrices of a network. As interpolating absent respondents' data can be viewed as structured matrix completion, we apply the methodology of Cai, T (2016, JASA) into the completion of a binary random matrix.
[2] Should firms regulate ID-sharing? The case of platform markets with network structure
In this paper, I investigate the subscription business model where people can share their subscription through ID-sharing. In the presence of a strong peer effect, this paper analyzes the dynamic optimization problem of a monopolist (through solving Bellman equation) in the subscription business. My results indicate that steady-state price and demand are less than the model precluding ID-sharing.
[3] Is network formation altruistic?
In this paper, I found robust evidence, from Networks and Employment Transitions Study (NETS) released at World Bank Open Data site, that becoming a friend is adversely affected by the negative income shock in most cases. Moreover, I observe the gender effect on the extent of altruism in network formation, which shows that women form more relationships that can help each other in the financial crisis.