Do As I Say and As I Do: Revisiting Price and Time-on-Market for Agent-Owned Properties

Abstract

Prior research documents that real estate agents sell residential properties they personally own faster and/or for a higher price than their client-owned listings. This finding is often explained in one of two ways: a) agents engage in shirking behavior when the principal cannot observe them directly (i.e., when selling a client-owned property) or b) agents listen to their own advice about how to best sell a property (i.e., when selling a property owned by the agent). To uncover the better explanation, the current study analyzes a sample of income-producing residential real estate. Sellers of income-producing real estate are likely more experienced and should recognize (and act on) good advice from an agent when it is given. That is, there should be less information asymmetry between agents and owner/clients of income producing residential properties. Thus, the agent-owned income-producing properties in our sample should not differ from client-owned income-producing properties along the dimensions of price or marketing time if agents are simply better at taking their own advice relative to the general public populating samples in related prior works. The absence of such differences in the present study suggests the price and marketing time effects observed in prior studies may be best attributed to information asymmetry rather than shirking.

Keywords:

• Hedonic modeling
• hazard modeling
• two-stage least squares
• time-on-market
• duration
• incentive compatibility
• agency problem

JEL Codes:

• C36
• C41
• G29
• K25
• L15
• L85
• R31

Notes

1 This is the essence of the argument put forth in Bian, Turnbull, and Waller (2017). They find evidence that suggests agents shirk their responsibilities to non-agent owned, concurrently listed properties in favor of focusing on their own properties listed for sale.

2 See Benefield, Cain, and Johnson (2014).

3 Z_ij includes ATYP, DOP, and NOMKT, which were excluded from X_ij, and excludes the variables related to construction materials that are included in Xij. These latter variables serve as instruments for the 2SLS estimation.

4 See, for example, Schuetz, Been, and Ellen (2008); Clauretie and Daneshvary (2009); Lin, Rosenblatt, and Yao (2009); Leonard and Murdock (2009); Harding, Rosenblatt, and Yao (2009); Rogers and Winters (2009); Daneshvary, Clauretie, and Kader (2011); Aroul and Hansz (2014); and Depken, Hollans, and Swidler (2015).