Tipped Off

I would have never thought that work as a hotel valet could be so exciting. I once heard wild stories about the night shift at a metropolitan hotel. “The cab ring is run by an Somalian named Ali,” one valet said. “His index finger was shot off in the Somali Civil War.”

The valets and the cab drivers have an arrangement. Whenever a hotel guest checks out and needs a ride, the valet calls Ali. Ali sends “one of his Somalian taxi drivers” to pick up the guest. In exchange for the referral, the driver slips the valet 10$. “It’s the most illegal thing ever,” the valet said. “We get death threats from the unions.”

I started wondering: What’s really going on in this transaction?

Unions try to retain finders’ fees

My first evaluation was the following.

The valets, by connecting riders to drivers, are acting as “market-makers”. This is a genuine and valuable service. The drivers (and counterfactually the riders too, were the valet’s services not already funded by the hotel) rightfully agree to pay for this service. After all, the benefit of finding more riders is (apparently) well worth the small fee the valet charges.

In fact, the passengers probably do pay for this service, albeit indirectly. If we assume here that – as is the case in just about any market situation – “economic profits are competed away in the long run”, it follows that cab drivers are just barely breaking even as it stands. The only way that they could get away with paying finders’ fees is by passing the resulting cost down to consumers. Of course, riders are free to choose other cabs. They don’t, though, indicating that for the passengers these higher fares are worth the convenience of a procured ride.

The cab drivers (and indirectly, the passengers) can, and surely already do, bid freely to set, in a market-directed manner, the precise level of this “finder’s fee”. In particular, cab drivers can abstain from paying this fee, though then they are left to find riders on their own.

The union has other plans. The union realizes that if it manages to coerce its drivers to band together to stiff the valets, many would-passengers would find taxi rides anyway. The union thus desires to shuttle resources away from the valets (and the passengers) and into the taxi industry.

Why is this practice prohibited not just by the union, but also by the law? Perhaps union leaders have convinced lawmakers to enact these laws on their behalf, in a classic case of rent-seeking.

Valet tips are bribes in a zero-sum game

If we introduce a bit of additional complexity, this view seems to give way to a much more sinister one.

Sure, riders enjoy convenience, and they benefit when valets find them rides. Yet these guests have very little knowledge of cab fares ahead of time, and they’re likely to blindly accept whichever rides the valets procure, even if fares rise inappropriately high. They have asymmetric information. Drivers observe that those who tip valets earn rides. Though they have to pass the resulting costs onto their passengers, they are not penalized for doing so, and market forces fail to stop the practice. Drivers begin aggressively outcompeting each other, forced to hand over more and more to access the customers the valet controls. The unsuspecting passengers, who ultimately bear the cost, are never the wiser. The process continues indefinitely. The passengers lose while the cab drivers, and most of all the valets, win.

Ideally, cab drivers – with the sort of organizing help that a union, or the law, is designed to provide – would stand together to forbid paying the valets, keeping fares low. Symmetrically, hotel managers, who desire to protect their guests, forbid these collusive arrangements. Even a few drivers and valets, however, can agree to conspire, generating immense gains, while riders barely notice. The optimal state is hard to preserve.

Things get even more scary if we consider the possibility that the colluding drivers are unskilled, unlicensed, or uninsured. Add to this the fact that the hotel guests perhaps defer to the valet not just out of convenience, but also out of trust. The valets are abusing the trust and endangering the safety of the guests and the hotel for their own personal gain.

The valets are, in effect, exploiting their privileged position as agents for the hotel and its guests by accepting bribes, from the drivers, for exclusive access to the guests whom they’re paid to assist and protect. Strikingly, the drivers too could soon become trapped. Drivers might find themselves competing with other drivers, ceding higher and higher bribes for access to the same fixed pool of riders. They have no choice but to play along.

Somewhere in between?

The real situation might lie somewhere still more complex.

Granted that this activity occurs – and that it’s both damaging and unethical – the full extent of the potential harm remains unclear. For one, hotel guests might well still benefit from the services of the valet (compared to no service at all) even if the services have been compromised. And while hotel guests’ information is asymmetric, it’s not nonexistent. If prices rose and quality fell too much, passengers would begin stepping out of cabs and walking to the light-rail (or to other cabs). The guests can’t be exploited unlimitedly. Finally, hotels could catch and fire their rogue valets, or the valets could decide to act properly.

Even a degree of information asymmetry, though, would permit the process to continue too far, allowing fares to rise and quality to fall beyond what better informed drivers would otherwise have demanded. Valets stand to gain and passengers stand to lose. This is a subtler form of the above argument.

Asymmetric information probably does exist.

There are other countervailing factors. The valets don’t have complete control over ride allocation, and nothing stops non-participating drivers from waiting for passengers at the hotel. Valets do have some of this ability, though, and it’s more economically efficient for drivers to wait for the call than to wait at the curb. Even so, though, it’s difficult to drop everything when the call comes in, and yet this is why Ali’s dispatching organization is so effective. Ali could presumably do the same without paying bribes, though. Finally, it’s not clear how much Ali’s group benefits from the practice. For all we know, rides are a dime a dozen on the street, and information asymmetry is equally present there too.

Valets as impartial agents

Hotels expect valets to act as impartial agents. This is indeed a service, and it might seem strange to demand that they do it for free. We don’t, though; hotels, presumably by asking higher room fees, subsidize the valet’s wages (as well as the hotel’s compliance enforcement budget). This neglects the broader question, though, of why hotels don’t simply relinquish control of the “taxi agent market” to actors closer to the market forces. Hotels could excise the payment corresponding to this service from valets’ wages, perhaps lowering room fees in the process. Drivers and riders could contract directly with valets, or remove them from the process altogether.

In fact, these are two different proposals. In the latter, agents cease to exist, and passengers and drivers negotiate directly. In the former, passengers and cabs choose freely among valets in a competitive “agent market”.

That both of these ideas are poor is due precisely to information asymmetry. Guests are ill-equipped to operate in the cab market, or in the agent market, single-handedly. Value is gained when hotel guests contract with their hotels for the provision of an impartial agent. In fact, the superiority of this strategy is manifested in the empirical fact that market forces have shaped the hotel industry to do just such a thing.

This approach is harder to enforce and preserve. Yet collective societies – featuring cooperation, punishment, and trust – exist precisely so that we can turn to collective solutions to collective problems. To simply abandon the collective is certainly not always the answer.

When and why does bribery happen?

Sometimes, though, it just might be. Instead of wringing our hands in exasperation as the immorality of the valets, we should soberly examine the incentives they face.

If valets cooperated — or, what amounts to the same thing, if it were possible to force them to cooperate — then passengers would surely benefit from the hotels’ arrangement for their services. This cooperation, though, could be an unreasonable goal. Guests might prefer to handle the thing themselves and save on the room fee. Perhaps a persistent market failure explains the hotels’ insistence on handling the service.

Bribery tends to occur, or become possible, in the presence of encumbering social structures in which individual incentives deviate from the social good. It’s only because the valets are entrusted to their positions — and because, in these positions, good behavior is effected via promise and not reward  — that bribery begins to carry benefit.

Provided that valets have signed employment contracts which demand their cooperation, their conduct does assume moral import. But a larger, and purely descriptive, question looms: given the difficulty of enforcing their cooperation, would the collective or the market solution bring better a outcome?

Though collective approaches can help, they often have costs. The ideal choice might depend on empirical matters of fact.

Economic assessment and policymaking

This investigation has already begun to exhibit the problem, classic in economics, whereby the ground-level realities of a situation make it very difficult to analyze economically. We can construct small formalized models, and discuss them. As the model accrues layers of complexity, though, things can get out of hand.

Normatively evaluating economic situations like this one, though, is crucial for policy.

To form this judgment, it’s not sufficient to observe the behavior on the ground. Sure, this behavior exists. But is it good or bad, the result of constructive or destructive economic forces? Similarly, it’s not sufficient to observe the policy response. The union and government response seems overwhelmingly negative. But is this response good or bad, the result of constructive or destructive political forces?

Complication aside, the valets here seem wrong to me.

Advertisements

6 comments on “Tipped Off

  1. Ben says:

    Here’s another confusing transaction, more hilarious than nefarious.

    At our mom’s new house in Virginia, our mom was thinking of getting the house’s gravel driveway paved. She called over a few construction and paving contractors to get the job done. These guys rolled over in a big truck with a container full of gravel in the back. The driveway is steep and curving, and the gravel is loose in parts; they struggled to make it up to the house.

    Though this is unrelated to the transaction, I can’t help but also mention how goofy these guys were. One of them was a hardy, straight speaking farm-worker type, young, with a bit of a belly and a UVa shirt. The other must have been around 50. He was extremely tall and lanky. He wore a dirty yellow polo shirt and dirty blue slacks, both of which were slightly too small for him. He had the posture of a gumby slightly bent at the waste. His hair came down to his ears in two thick shaggy curtains more-or-less parted in the middle. His skin seemed smooth and yet had deep wrinkles, like someone ironed his face without straightening out the creases first. He couldn’t stand still and had mannerisms like he was craving a cigarette. The young one, standing on the entry walkway, did most of the talking, while the older one seemed to linger around his truck, piping in with a comment only once in a while.

    The workers informed us that since were were planning to undergo remodeling, and to build a garage, it would be better to wait to pave the road until after the cement trucks had done their job on the garage. “They’re so heavy, they’ll crack the asphalt,” the young one said. They started getting ready to go, offering to come back when we were ready.

    They had come out for free, and we felt a little guilty for taking their time for nothing. But here, things started getting even weirder.

    “It won’t be easy to haul all this gravel back down your driveway,” one of them told us. “Could we just go ahead and dump it here?” They pointed to a perfect flat little landing space next to a bend in the descending driveway. “Once you go ahead with the paving, we’ll come back and it’ll be waiting.

    “We won’t charge you,” he said. “Though I suppose if 6 months or so go by, we’ll probably send you a bill,” he added hesitantly, with a wave of his hand. He didn’t sound determined. “Can we go ahead?” the old one asked, climbing into his truck.

    Trained by habit to perform miniature economic analyses on all nontrivial transactions, this one left me temporarily dumbfounded.

    One of my first thoughts was that we were getting a bit of a bad deal. For various reasons, a buyer (resp. seller) typically prefers to delay (resp. advance) as much as possible the purchase (resp. sale) of a fixed good for a fixed price. In effect — and ignoring the apparent uncertainty of the eventual bill — we were undertaking the transaction much earlier than we needed or were ready for.

    One of those reasons is interest accrual on cash. Only before (resp. after) the transaction is made can the buyer (resp. seller) earn interest on the cash. But wait. We weren’t even paying early, just receiving the goods early. Arguments about interest couldn’t apply. In fact, we were getting the goods early and earning the extra interest.

    Another reason for this is uncertainty. In six months, the sellers were certain still to gain from the transaction, while we were only likely to, and anything could change by then. It was in the sellers’ interest, and against ours, to lock in the transaction early. (Of course, this still assumes the eventual bill comes.) On the other hand, assuming our right to return the gravel with a refund, the company, in doing this, might risk the cost of loading the gravel back up.

    Another factor is storage, which presumably has a cost, and which effectively fell on us for the intervening 6 months. On the other hand, it hardly hurts us, though it might have hurt them. Perhaps the relationship is commensalistic.

    We still have the gravel, and the bill hasn’t come yet.

  2. Josh says:

    One interesting corollary I would bring up is travel agencies. They’ve gone largely obsolete in today’s internet age. But they used to fulfill a role similar to that which valets fulfill now: they helped us find the best deals on transportation.

    The fact that these at least used to exist demonstrates that, if the valet’s job were left to the market, the result might not be all bad. You said that “guests are ill-equipped to operate in agent market”. But the existence of travel agencies indicates that this might not be true. Of course, the idea of a valet agency might be unrealistic in the first place. It makes more sense to offer professional assistance for a $1000 expenditure than for a $10 one. If 5% of the costs go to the agent, then the valet’s cut wouldn’t be worth his time, while the travel agent’s cut would be. So, the valet market might be more difficult to enter, which might explain why hotels take up the burden.

    Another note, though, is that the fact that travel agents have largely disappeared today might indicate that valets are equally unnecessary. If we can find the best flight deals on expedia, we should just as easily be able to find the best taxi deals on the internet (or on Uber). This would suggest, at least empirically, that a hotel without valets wouldn’t be a bad one.

    • Ben says:

      I think you’re correct to worry that the factors which make, or made, travel agents useful would fail to apply in the valet market case. The transactions are much smaller, and the presence of agents would be somewhat cumbersome.

      You should also worry, however, that the factors you state which made travel agents cease to be useful would also fail to do so in the case of valets. Information is now very prevalent on the internet, while it remains sparse on the hotel curb.

      In other words, there has always been a small need for valet agents, and this need has largely failed to be addressed by the rise of the internet.

      I think Uber is the key idea here. Now, and only now, that Uber exists, the valets’ job might become truly unnecessary.

  3. Ben says:

    A related case happened at LaGuardia airport in New York. At LaGuardia, taxis pick up airport passengers via a taxi dispatching system. The airport hires a third-party company which oversees the dispatching process. In this process, taxis wait to pick up passengers until they are called by the dispatcher. They wait for a whopping 2 to 3 hours, which they spend gridlocked in a massive auxiliary lot. Dispatchers were caught taking bribes from drivers to bring them to the front of the pack. The participating dispatchers were sent to jail.

    How do the cases compare? They have quite a lot in common. In both cases, an actor hired to act as an impartial allocator takes bribes from those providing the service he is allocating in exchange for allocating improperly.

    Unlike the case of the valets, it was perhaps more clear what the dispatchers’ expectations were. Furthermore, the scale of the whole thing was much larger, and the legal penalties more severe.

    More is common to the two cases. As we wondered whether the hotels should cut the valets loose altogether, here we might wonder whether the whole arrangement might be wrong if cabs are being made to sit useless and idle for hours in a lot. Is this a market, or a political, failure, or somewhere strange and in between? Perhaps the dispatching situation should be eradicated and left to the market. It’s conceivable that a mob of dozens of shouting and competing cab drivers doesn’t serve the consumer, and that LaGuardia thought as much when they decided to hire the dispatcher. But perhaps the market arrangement would help too, albeit in more indirect ways. Perhaps fares would be lower for aiport passengers. (What kind of fares are the drivers charging, in the face of which they can get away with paying bribes?) Perhaps fares would also be lower on the street corners of New York city, to which a number of LaGuardia’s drivers would undoubtedly turn once their 3-hour wait began to take on opportunity cost. Perhaps the cost of goods and labor would fall, imperceptibly, all across the city and nation, as a few cab drivers realized, finally, that the whole thing wasn’t a good idea.

    For the time being, though, those dispatchers are still in jail.

    As a final note, I’ve read that corruption among cab dispatchers is extremely widespread, and not just at airports. (I’m not sure the role of dispatchers in cabs’ everyday lives.) Makes you really appreciate Uber, doesn’t it?

    There’s something sad about the fierce struggle of the taxi industry — already bloated, inefficient, and corrupt — against the impending rise of Uber. These cab drivers expended massive amounts of resources training for and complying with the cab industry’s standards; more cynically, they had to buy their way into the industry’s rent-seeking guild. Now they stand to lose it all. Such are the travails of creative destruction, which has been fought every time it has ever appeared in history.

    • Josh says:

      Alright, I did a bit of thinking about this.

      So, the bribery was clearly illegal. But why? A good question to ask, when asking why something is illegal, is who’s being hurt?

      Well, in this scheme, the taxi drivers who bribe are making off excellently. Note that cabbies are sorted into different lots based on the length of the ride they’re prepared to give. So, those who wait 3 hours end up giving a ride, say, all the way to New Haven, and so they make the money lost during those 3 hours back, and then some. So, once again, they do great. But, of course, the other taxi drivers lose. They provide the same ride, but they have to wait longer in line.

      Well, why not just let all taxi drivers pay?

      If all taxi drivers could slip the valet 10 bucks, and get to the front of the line, then they would all succeed enormously. However, it wouldn’t quite work like that, since it’s impossible for everyone to get to the front of the line at once. So, more people would bribe, but the bribe would become less valuable. At the end of the day, the price of a bribe would fall until it’s more of a take it or leave it type of deal. A driver could pay $5, to get a bit of a head start in line, or he could just pay nothing and wait. And $5 would be the break-even point, where the driver isn’t quite sure whether he’d rather pay $5, save 20 minutes, and spend those 20 minutes on the street-corner offering rides, or just save the $5.

      But is this ideal? It seems like a lose-lose system for the drivers. The status quo allowed drivers to pay nothing and still get to the front of the line in 3 hours. Now, if they pay nothing, they get to the front of the line in 3 hours, 10 minutes, because they’re being cut by drivers who bit the bullet. Of course, they could pay, and cut that time to 2 hours 50 minutes (recall, it’s a 20 minute spread. And, wait time is a zero sum game, as there’s no change in the amount of rides being offered). But we already agreed that 20 minutes is worth $5. So, in effect, they’re spending 3 hours 10 minutes in line, whether they pay or not.

      So, the taxi drivers lose as a whole in this situation. And, of course, the valets (or whoever ends up getting the money) wins.

      Now: is this ideal? Well, no, it’s not. Because, once again, we ask, who’s really being hurt? In this scenario, not just the cabbies, but also the riders, are hurt. Some cabbies might go out of business under this new strain. And, the ones that are in business would have to charge more. So, customers would pay more for their rides.

      We can imagine a town run by a mafia, who charges every business for “protection”. Under this burden, half the bakeries in town might have to close. And the ones remaining would have to charge more for bread to make ends meet. Thus, the customers of the town end up losing. Back to La Guardia, the valets would be the mafia, casting upon the cabbies the burden of the bribe, which, ironically, used to be helpful.

      So, in sum: when bribery is allowed, cabbies lose; customers lose; and valets win. I think this conclusion is pretty strong and stands for Ali’s ring as well. And, so, La Guardia was right in jailing the valets. Justice!

      • Ben says:

        Yes, the bribery is harmful. The exact nature of the harm depends on the nature of the bribery.

        (1) In one situation, there are two lanes: a fast track and the regular lot, and a fixed-price bribe grants one access to the fast lane. Quickly, the fast lane will become somewhat full, and at the break-even point, which depends on the price, the amount of drivers in the respective two lanes will stabilize. This is the situation you describe.
        (2) Alternatively, a fixed number of spots at the front of the line are granted to the highest payers. In this situation, the price could get very high, but not arbitrarily high.
        (3) In yet another situation, any paying customer gets an instant pass to the front of the line. We could think of it this way. Between two successive passengers, the privilege of giving the impending ride is granted to the highest bidder in a miniature auction. This process repeats for each call.

        I’m not sure the specifics of what happened in LaGuardia. In all cases, the equality and fairness of the dispatch process is compromised, the passengers suffer through higher fares, and the dispatchers rake in cash. All are harmful. You offer one description of why. One way I like to think of it is that the valets are extracting money from certain cab drivers in exchange for granting a boon to them which is zero-sum across the total pool of cab drivers. (The cost is effectively extracted from the passengers.) I don’t think much more needs to be said than this. The details will differ according to which of the three above “models” we choose.

        Following Thomas Sowell (Knowledge and Decisions, 1980), we should ask “not what decision to make but who is to make it–through what processes and under what incentives and constraints, and with what feedback mechanisms to correct the decision if it proves to be wrong.” I think that the big problem here is that among the alternatives available to us — which include free-for-all competition among cabs for passengers and contracting of a third-party dispatching company — all face deficits both in decision incentives and constraints and in feedback mechanisms. Particularly, the former faces shortfalls of knowledge and the latter faces problems in incentives; both lack adequate feedback mechanisms.

        That the dispatchers were caught and punished is a promising sign that knowledge is being attained and applied properly by the government. Interestingly, if I recall correctly the investigation began upon complaints by the other drivers. It should be no surprise to us that the ultimate source of the key information was a local one close to the matter at hand.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s