Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We use theory and experiment to explore the performance of multi-round, price-guided, combinatorial auctions. We define efficiency-relevant and core-relevant packages and show that if bidders bid aggressively on these and losing bidders bid to their limits, then the auction leads to efficient or core allocations. We study the theoretically relevant behaviors and hypothesize that subjects will make only a few significant bids, and that certain simulations with auto-bidders will predict variations in performance across different environments. Testing the combinatorial clock auction (CCA) design, we find experimental support for these two hypotheses. We also compare the CCA to a simultaneous ascending auction. (JEL D44)