How (not) to design procurement mechanisms: A laboratory experiment

In this paper, we examine the relative performance of three commonly used procurement mechanisms: price-only auctions, scoring auctions, and benchmarking. We do so both in theory and in a laboratory experiment. We find that the auctions yield the same level of welfare, and welfare dominate benchmarking. In theory, the "pie" is shared the same in both auctions between buyer and suppliers, and both the buyer and suppliers obtain higher utility than in benchmarking. In contrast, in our experiment, we observe that the price-only auction generates higher supplier utility than the scoring auction, while the scoring auction dominates the price-only auction in terms of expected profit for the buyer. We do find support for the underperformance of benchmarking versus the auctions.