Winners, Losers, and Losers Who Might Be Winners
Date Issued
May 2026
Author(s)
Verville, Thomas J
The University of Tennessee
Advisor(s)
Don Bruce
Additional Advisor(s)
Matt Van Essen
Celeste Carruthers
Michael Ryan
Abstract
This dissertation examines the equitable allocation of resources across both public policy and private disputes. Chapter 1 studies firm-specific subsidy competitions among local governments, while Chapters 2 and 3 analyze a historically significant fair-division procedure practiced in Wales under gavelkind inheritance law. Across these settings, the dissertation highlights how institutional details—information, timing, transparency, and the feasibility of side transfers—shape whether resource allocation achieves fairness and efficiency.
Chapter 1 evaluates firm-specific subsidy competitions among local governments. Using county-level data constructed from Site Selection magazine and Good Jobs First, I estimate difference-in-differences models comparing auction winners and runners-up. Governments frequently offer large incentives to attract firms with the expectation of increased employment and local economic growth, yet the consequences of these policies remain contested. The estimates suggest limited evidence that winning subsidies increases employment, while local business activity effects are heterogeneous across winners and runners-up, highlighting potential unintended consequences of place-based bidding wars.
Chapter 2 turns to fair division theory through a formal analysis of a Welsh inheritance procedure known as gavelkind. Under this mechanism, the youngest heir divides an estate into
n
n portions and the heirs select sequentially by seniority, leaving the divider with the final parcel. I model the game induced by this generalized divide-and-choose mechanism and characterize its fairness and efficiency properties under alternative behavioral and informational assumptions. When preferences are private, maxmin (and maxmin-perfect) reasoning yields an equal-split division that is envy-free but not generally Pareto efficient unless preferences are homogeneous. Under perfect information with heterogeneous preferences, equilibrium allocations satisfy a directional (lexicographic) form of envy-freeness but may be inefficient. The mechanism is also vulnerable to collusion, as its sequential structure allows the divider and first chooser to coordinate on partitions that benefit them at the expense of later choosers.
Chapter 3 provides the first laboratory test of this generalized divide-and-choose procedure in three-person groups. Treatments vary the feasibility of coalition formation by allowing (i) no communication or transfers, (ii) pre-division private communication, or (iii) communication combined with voluntary ex-post transfers. The experimental evidence indicates that introducing transfers alters outcomes in a manner consistent with coalition behavior: efficiency declines and payoff distributions shift toward earlier movers, while the second chooser is disproportionately harmed. Although collusive outcomes do arise, they are not ubiquitous, suggesting that trust and coordination frictions limit the full implementation of side agreements even when transfers are available.
Collectively, these chapters show how institutional design influences whether resource allocation mechanisms produce equitable and efficient outcomes across both public policy and strategic private settings.
Degree
Doctor of Philosophy
Major
Economics
File(s)![Thumbnail Image]()
Name
VERVILLE-PRIMARY-2026.pdf
Size
4.94 MB
Format
Adobe PDF
Checksum (MD5)
d822f904108bd66c6ba762d7e17ffb61