Should federal contracting be more like Airbnb?

Roger Cohen is a columnist for the New York Times. He mostly writes about foreign policy, but his most-recent column strays from the beaten path to focus on Airbnb.

The column begins with an account of a recent conversation Cohen had with Brian Chesky, Airbnb’s founder and CEO. “He told me about trying to raise $150,000 in 2008 for his idea of a peer-to-peer home and room rental company. Everyone called him crazy," Cohen writes. "They scoffed at the notion that people would trust one another enough to allow strangers into their homes. They derided the idea that those strangers would be nice enough, or honest enough, to respect properties.”

Yet it turned out that millions of people were indeed willing to open their homes to strangers and millions of those strangers were willing to treat the properties right. “Without fundamental human goodness,” Chesky noted to Cohen, Airbnb “would not work.”

Chesky’s basic message is an extraordinarily hopeful one about human nature. And when I read the column for the first time, that was the only message I got. As those who know me are aware, I am the kind of guy who assumes the best about people, and I felt very good seeing an affirmation of this assumption.

Yet as I thought more about the column, I began to see Cohen’s story as one that, while not contradicting his basic hopeful message, is slightly more complicated. And it has implications that extend well beyond Airbnb to two areas I care a lot about: managing government contracts and managing government employees.

The philosophies of federal contract and employee management do not give expression to the trust that Cohen claims is basic to Airbnb. The infrastructure of compliance-oriented process rules, requirements and prohibitions all reflect a real worry that contractors and employees will cheat and loaf without extensive oversight and constraints. But that infrastructure, aside from being quite costly to administer, also feeds on itself, making participants less inclined to behave in trustworthy ways without harsh oversight and constraints.

So what is going on? Are government contractors and employees so dramatically different from Airbnb guests and hosts?

Admittedly, Cohen slightly oversimplifies the nature of the trust relationship for Airbnb. Chesky states in his conversation with Cohen that Airbnb is “pushing people to be more understanding and accepting of each other.” 

The word “pushing” is important (maybe, to use jargon common these days among scholars and others looking for ways society can influence behavior in more gentle ways, we should refer to “nudging”). The key way the Airbnb system pushes trust is through ratings -- that guests dive to hosts detailing how well they were treated, and that hosts write for guests based on how well they behaved. That hosts rate guests is key to their being willing to open their homes to strangers, because they have greater assurance the guest will behave well. That guests rate hosts is key to providing hosts an incentive to treat guests well, and constitutes the basis for the willingness of guests to use these homes strangers provide.

Airbnb tries to create a robust system for an easy ability to rate, pinging guests and hosts on departure to rate each other. (My property manager for the apartment we own in Miami, which gets guests through Airbnb, explains to me that if the guest has been good and they have good communication, he sends in a review immediately. If the guest was ok but there wasn’t much communication, “I wait -- if they write a review, I'll respond with a good review.” If the “apartment is dirty and damaged, regardless of anything else, I write a 1 star review with appropriate comments to warn future hosts.” He adds ,“This has happened two times in three years,” reflecting the basic point that most people are trustworthy.)

This system displays a light touch, unlike the harsher oversight of contractors and employees common in government. However, if this simple, unobtrusive system were not present, Airbnb itself might well collapse. Even in a world of fundamental human goodness, participants would likely feel the need to protect themselves using oversight against bad actors, and the obtrusive oversight would create a­­­­­­­­­ downward spiral of trust.

Readers may recognize that what I am describing here for Airbnb closely resembles a system of past performance assessment for government contracting that I tried to introduce when I served in government 20 years ago. That effort has helped, but not nearly enough -- it is not enough of a priority for most contracting officials, and a series of paperwork requirements and “due-process” protections for contractors make it harder for the government to prepare honest reports on performance.

So is there anything the government can do to make its past performance system closer to Airbnb’s? It would help to up the priority given to evaluating contractor performance so it becomes a central part of the job description of contracting folks, rather than emphasizing compliance reviews. I would also propose a sort of past performance grand bargain between the government and contractors who wanted to sign up: In exchange for reducing compliance requirements on contractors, contractors would allow government to develop past performance reports with less documentation and less contractor opportunity to appeal their ratings.

My lesson about the implications of the Airbnb story for managing contractors and employees is that we can loosen up on our draconian systems of process-oriented rules, prohibitions and oversight. Most people are good, and will behave well without it. But Airbnb also tells us that we can’t do without any controls, because this would likely lead the system eventually to implode from defections among the venal or lazy minority. Let’s keep the controls more gentle than harsh, and let’s rely more, as Airbnb does, on past performance as our regulator.