This dissertation consists of three separate papers (as Chapters 1, 2 and 3). The first chapter deals with a spatial hedonic price analysis of used excavators. Chapter 2 deals with empirical estimation of functions for auction house revenue. Chapter 3 deals with empirical analysis of price relationships in a multi-item, multi-type auction.

The objective of the first paper is to specify and estimate a spatial hedonic price function for used excavators sold by auction in North America from 1996 to 2005. The results indicate that prices of used excavators differ significantly for various reasons. Not surprisingly, prices vary by region of sale, physical condition, and brand of the equipment. We also find that prices are different between auction houses themselves. Furthermore, there exist 'within sale' spatial effects on the selling price that are statistically significant. Auction houses, besides providing auctioneering service to sellers and bidders, can influence selling prices in the auctions for used construction equipment.

The objective of the second paper is to quantify the effects of multiple units of items and to identify the influence of different types of items on auction revenue. The results indicate that auction revenues increase with an increasing rate for some items (graders) and with a decreasing rate for others (excavators and wheel loaders). Magnitude of effects varies from auction house to auction house, which, might be due to heterogeneity in the groups of bidders across auction houses. Cross quantity spillover effects also vary in magnitude and direction from auction house to auction house.

Auction houses stage events where sellers offer for sale multiple units of an item, items of different types, and items in various physical conditions. Objective of the third paper is to specify and estimate theoretically consistent inverse supply functions for multi-unit and multi-type auctions. It is hypothesized that price complementarities exist between different items being offered at an event, offering opportunity for auction houses to exploit price relationships by preselecting the number, type and condition of equipment. Normalized quadratic inverse supply functions are estimated for different types of equipment, and hypothesis are tested using transaction level, time series cross section data on equipment sales from 1996 to 2006.