Miho Hong (홍미호)

We study how the organization of trade affects efficiency and surplus division in South Korea's agricultural wholesale markets, where farmers transport and sell produce through auction houses to buyers. Integrating fragmented trading pools can improve match quality and limit market power on both sides. However, it can also discourage participation and strengthen auction houses' fee-setting power. Consolidating trade in fewer locations can also raise transportation costs. We develop a model of farmer sorting, bidder participation, spot-market trade, downstream sales, and auction house fee setting. We use detailed fresh-produce auction records from 2016-2023 to estimate the model. On the seller side, we construct a panel of farmers' shipment choices under varying competitive environments shaped by exogenous harvest patterns, which identifies their sensitivity to distance and price. On the buyer side, optimal participation conditions identify bidder markdowns. We combine these markdowns with auction prices to identify match values created from trade as a flexible function of supply and bidder counts. Estimates show that match quality increases substantially with the number of participating bidders, especially for crops with more within-crop varieties. Yet neither side of the market fully internalizes these gains because of bidder market power and matching benefits that accrue to other participants; counterfactuals show that removing these distortions, as well as auction house market power, would raise total surplus by 27%. Nevertheless, we find that even with such distortions, integrating co-located auction houses---under the existing cap on auction house fees---would have a larger welfare effect (+40%), with farmers capturing roughly 41% of the increase. Finally, we show that consolidating niche market locations into regional hubs can raise total surplus despite higher transport costs. 

Allocation of differentiated goods to buyers with heterogeneous tastes introduces a tradeoff between efficient product-buyer matching and competition between buyers. Separate auctions improve matching by letting buyers target preferred items but thin competition, while pooling differentiated objects increases competition at the expense of match quality. We quantify this tradeoff in wholesale peach auctions in South Korea, where hundreds of lots are auctioned sequentially each day. We document bidder-variety sorting in auction participation and estimate a model of sequential auctions with multi-unit bidder demand, heterogeneous latent tastes for varieties, and selective entry. The estimates reveal heterogeneity in both relative valuations and willingness to bid across varieties. Revenue-maximizing schemes pool varieties that attract different bidders while preserving the horizontal differentiation most valuable for matching. These schemes would raise revenue by 4.8-7.5%, with effects on efficiency ranging from a 2.2% loss to a 0.5% gain. The status quo is typically dominated by pooling schemes that would improve both revenue and match value. Supply composition changes the balance between participation gains and matching losses, and an index of these forces closely predicts the resulting revenue ranking of pooling schemes. 

We study patient welfare, capacity investment, and wait times in a universal healthcare system where patients can either wait for publicly funded care or pay for faster private treatment. Allowing private-pay care can stimulate capacity investment and relieve public queues, but market power can distort prices and the allocation of treatment. We build a model of patient choice, wait times, pricing, and capacity investment. Estimation combines administrative data on publicly funded care and scraped private prices and appointment availability. We find that patients have heterogeneous preferences for timely care. Moreover, pricing and capacity choices affect equilibrium wait times, but profit-maximizing providers do not internalize the resulting congestion externality. The socially efficient allocation would lower private prices and expand capacity. Disallowing private-pay care would reduce patient welfare through the loss of sorting on willingness to pay for timeliness, increased congestion in public queues, and reduced capacity investment. Attaining the status quo welfare under public-only provision would require a 35% higher public budget to fund additional publicly provided capacity. Private-pay care benefits patients in nearly all urgency, age, and income groups, including those unlikely to use it themselves, who gain from reduced congestion in public queues.