Socio-economic perceptions and signaling could still be at play here. Does someone not buy the cards because the person is black, or are they less likely to buy

Socio-economic perceptions and signaling could still be at play here. Does someone not buy the cards because the person is black, or are they less likely to buy the card because they worry that it might not get delivered, or that it might be stolen. I wonder if you'd see the same effect for a women's hand as a man's hand. I also wonder how many fewer bidders you would need to show this effect. In small illiquid markets the loss of a couple of bidders could put more downward pressure on the top bid. The part that was most convincing that the authors identified the correct effect though was that the price was also affected by the race of the person on the card. Looks like there is an opportunity for racial arbitrage.