Reputation Systems as Regulators

Tyler Cowen and Alex Tabarrok write,

In recent times, information technology has made it easier to observe a seller’s reputation and to contribute to the formation of a seller’s reputation at low cost. Yelp, Angie’s List, and Amazon Reviews all make it easy for past buyers to report their observations on seller quality and for future buyers to observe a seller’s accumulated reputation.

This might mean that we need less regulation. Other possibilities:

1. We have always had the means to use reputation systems as an alternative to regulation. But regulation is the preferred outcome of the political system, perhaps for bootleggers and baptists’ reasons. The advent of better information and technology actually does not change the relative economic and political advantages of regulation.

2. One issue with reputation systems is that there is an ongoing temptation to game them. Consider “search engine optimization” or “social media marketing.” Surely there are folks out there offering to get your business top rankings on Yelp. Given that gaming will work at least sometimes, is an unregulated equilibrium necessarily the best?

3. It becomes harder for new entrants to break into a market governed by reputation than one governed by regulation. Obtaining a license from a regulatory agency might be easier than obtaining visibility in a rating system.

This entry was posted in Information Goods, regulation. Bookmark the permalink.

16 Responses to Reputation Systems as Regulators

  1. Matt says:

    Mr. Kling-

    Are you familiar with “Dewey Defeats Truman”? Angie’s List and Yelp are nothing more than that mistake multiplied millions of times over.

  2. Wouldn’t it be arguably easier to break into the reputation based system? Assuming all sellers must attract buyers, a seller must build reputation regardless, which is usually done through word of mouth. However, immediately a buyer faces an information asymmetry problem. Review system announce the reputation of a buyer reducing this information problem and the cost to obtain reputation. Still, in a license system, the seller must incur an additional fixed cost. In total, license system face the cost of regulation and the cost of information.

  3. On point 2, remember that Yelp et al. have an incentive to maintain the integrity of their system (for their own reputation’s sake). Google has done a pretty good job of eliminating SEO tricks beyond the regular production of good content. And of course, the question isn’t whether an unregulated equilibrium is “the best”; it’s whether it’s better than what we will have otherwise.

    • James says:

      I’m not sure that the companies see it the same way. Yelp! for example, has a practice of charging participating business for advertising and then removing negative reviews if they pay up. Worse still, if they didn’t pay, some positive reviews would vanish!

      “The small business owners had all enjoyed positive reviews and high ratings on Yelp!’s five-star ranking system.
      But that changed, they said, after they rejected solicitations to pay for advertising on the Yelp! website. Some of their positive reviews were taken off the site and their overall rankings fell.”!-engaged-in-hard-bargaining-not-extortion-appeals-court-says

  4. R Richard Schweitzer says:

    The topic is “Asymmetry” which is basically an issue of “balance” in the access to, and capacity to use, information.

    Bringing in the concept of reputation as a form of information (opinions, evaluations, experiences) involves additional “participants” in an event (transaction, cooperation, association, etc.) which is also the effect of regulation (an outside participant with its information) to be added to the “balance” issue of asymmetry.

    In a complex society, events do not involve simply the interactions of say, two participants and their respective balances of information, but the effects of others who may be direct or indirect participants in an event. That observation goes as far back as the effects of the Rules of the Bazaar on free traders.

    It may not be just so simple as comparing knowledge OF reputation and OF regulation, but rather the effects of those participants (and what passes for their knowledge) on the actions of the participants directly concerned with the event.

  5. Mike says:

    An evolution in the flow of information is not the same thing as a shift in widely-held priors. Regulation is not only the politically preferred option but the socially accepted default, but is corrosive because it erodes trust or impedes its growth. It is easy to put the scarlet letter (U, unregulated) on people and groups, and for most people it is simply assume that stringency, review, gatekeeping, etc are all done legitimately and with good and observable outcomes.

    Libertarians are as guilty as liberals of making optimistic just-so stories about recent developments. When progressives take credit (some of it deserved) for the shift in attitudes on gays and gay marriage, the narrative does not work for the current condition of blacks, prisoners, or poor students. Similarly, I doubt that widespread, plausible, and accurate information about goods and services is likely to affect the Commanding Heights, political candidates, etc.

  6. Jackson says:

    Amazon has is working to overhaul its review system which, like Yelp, often serves as a bludgeon against books/media that don’t speak to the mainstream line of propaganda. I think the idea of going to a “verified purchase” condition as a rule for writing a review is a good start.
    Yelp pretty much blew its credibility with the pizzeria fiasco, you know, the pizzeria that didn’t refuse or discriminate against anyone, but was destroyed anyway. Yelpers who didn’t eat there were happy to review it anyway.
    The wisdom of crowds is often the shrill voice of the most belligerent shouting down and drowning out all others.

    • Andrew' says:

      I may spend more time in Amazon reviews, on net, pardon the pun than anything else on the internet.

      That said, I spend more time there partly because it such a bad interface!

      It is one of those weird things that is both a wonder of the world and a crime against humanity.

    • Dain says:

      I think it’s interesting the way recent developments of the “outrage machine” variety might get conservatives to rethink the wisdom of civil society, which they often tout. The reputation thing being an aspect of that. Also see the way the undermining of free expression isn’t lately coming from any actual undermining of the 1st Amendment but instead the “spirit” of the 1st Amendment, which will be found (or not) in the heads of everyone who makes up what we call society.

  7. Tom DeMeo says:

    I disagree with the initial quote almost completely. The information that Cowen and Tabarrok speak of is of a lower quality than “reputation”. It is easy enough to look at and better than nothing, so it is worth looking at, but it aint reputation. Some of the other commenters have addressed why. If you make it more important, it will be become even lower in quality.

    Curation is something that makes sense, but so far, the market isn’t paying for it.

    The way I see it, reputation and regulation have very different purposes, and if they overlap, that means the regulation is badly conceived.

  8. Bryan Willman says:

    1. Why does anybody really think that regulation as a whole can be justified by protecting consumers in accord with consumer wishes? Some part of it is. But much more of it is about various social arm twisting exercises – observe the ongoing CAFE versus what people actually buy wars, the ongoing efforts to raise automobile safety being undercut by people just driving harder, etc.
    Look at building codes – it seems very clear to me that while some of them are indeed about protecting me from shoddy contractor behavoir, others are designed force down demand for water, electricity, heating fuel – presumably to keep prices for those things low.
    Likewise regulations related to zoning – it’s not about protecting the consumer of the new house, it’s about protecting the owners of the existing neighboring houses.

    2. We’re completely leaving out regulatory capture, various sorts of cronyism, attempts to regulate people into confirming to a particular group’s image of social correctness.

    3. Reputation alone is useless for anything more complicated than dinner.

  9. Chris S says:

    #3 is irrelevant unless you also outlaw the reputation systems.

  10. Slocum says:

    @2. The owners of reputation systems have strong incentives to fight gaming in order to preserve the value of the system in the face of competition. But regulators are a government monopoly with no competitors and only weak incentives against regulatory capture but strong incentives (lucrative private employment after their government ‘service’) to acquiesce to such capture.

    @3. I’d say it’s the reverse. Getting a government license is a pretty low bar that doesn’t provide much of a guarantee of excellence to prospective customers (especially given endemic regulatory capture). Reputation systems, in this area, are really a substitute for brand names rather than regulation. A product or service with no reviews is at at disadvantage, but it does not take very many reviews to get a statistically meaningful measure. Excellent products from unknown manufacturers are much better able to out-compete mediocre products from established brands now that reputation systems are generally available.

  11. Bryan Willman says:

    I did not mean to argue that reputation systems couldn’t give regulations a run for their money in many contexts.

    My point was that an argument that goes “information asymmetry is the real reason for regulation, correction of this assymetry will cause regulation to be less necessary and go away” is naive. There are lots of regulations that are arm twisting, rent seeking, and so on – they’re not about “correct for information asymmetry” they’re about attempts to force people to behave in particular ways.

  12. Shayne Cook says:

    Isn’t an advanced college degree a form of reputation system?

    For example, Both Alex and Tyler have earned Phd’s in Economics – an earned “reputation” (backed, I might add, by a “Certificate of Authenticity”, ostensibly issued by a reputable educational institution, which is in turn has it’s own “Certificate of Authenticity” issued by a regional accreditation board.) So it’s even a multi-layered reputation system.

    And yet ….

    While reading the Cato Unbound (a reputable publication by a reputable organization) article authored by Tyler and Alex, I saw this statement in their section on CarFax information:

    “In the future, states or the private sector could provide this information online for free.” [emphasis, mine]

    Don’t get me wrong. I like free stuff as much as the next guy. I’ve just never actually found any.

    Kidding aside, I just wrote that off as a “slip-of-the-pen” on their part, as it were. No reputable economist could actually believe, or would actually state, that anything is actually free. That’s not information, it’s mis-information – a lie. And no reputable economist would lie. So I moved on with the article – where I stumbled across this part of their health care/health insurance discussion:

    “For instance if you have a cancer of a given kind, this is verifiable to the outside world, and if the treatment costs are $200,000, the cost of an insurance policy will in turn be about $200,000.”

    So it seems Tyler and Alex DO actually believe in the existence of free stuff. In this case, the health care cost is “verifiably” $200,000, but the health insurance cost is “free”, or at least “about free”.

    In actual fact, per PPACA stipulation (black letter law), if the verifiable health care cost is $200,000, then the legally mandated health insurance cost is $50,000 – making the total cost $250,000.
    (PPACA stipulates that “health insurance companies must spend 80 cents of every premium dollar collected on actual [verifiable] health care costs.”)

    On reputation and reputation systems in general ….

    As several commenters here have already noted, “reputations” are a relatively fragile thing. They take a great deal of time and effort to build, but they can be negated in the twinkling of an eye. That phenomena is reflected in the old saying, “It only takes one ‘Ah Darn’ to negate 10,000 ‘Atta Boys'”. And that, completely irrespective of the “Certificates of Authenticity” attesting to the previously hard-won ‘Atta Boys’.

    With respect to this topic/article in particular, once I got to that “health care” part, I just stopped reading their article. I can get this “quality” of information from any politician, advertisement, most news articles, all SPAM e-mails, or even the “National Enquirer”. They all talk about free stuff. I’m always looking for real information, from reputable sources. Mis-information I can get almost anywhere.

    Do I actually have to go to sources other than Cato, Cato Unbound, Alex Tabarok and Tyler Cowen to find real information?

    (Frankly, they’ve put their reputations in dire jeopardy here. Someone should maybe tell them, before they completely negate it.)

  13. Daublin says:

    I’ve never seen a well-functioning reputation system, but have participated in several. I agree with other commenters that the best way to do “reputation” is still the old fashioned mechanism of soliciting references and then reading and considering carefully what they say. You can usually spot a faker, e.g. a product review by someone who works for the company selling the product, because the details of their story don’t sound quite right.

    Rather than focussing on aggregation of reputation data, I would turn attention to *filtering*. There have been some big, helpful advances in filtering reputation data to something you can find useful. An example I’ve found helpful is that the Google Play store will show you reviews by people you know. That’s great; I know those people and I know how to weight their decisions on various kinds of products.

Comments are closed.