Understanding Auto and Uber fare regulations using first principles
Let’s start with auto fares. Why should auto fares be standardized? Why can’t auto drivers decide their own price as per their wish?
Consider a world where auto fares aren’t standardized. For the sake of simplicity, assume that Uber/Ola aren’t yet there in that world. In that world, assume you get down in a town and request a trip to a certain place? The onus is on the driver to quote a price. How would a driver behave in such circumstance?
Economic Theory (psychology!) sheds light on this context. Consider this as an auto driver — customer bargain. What incentive does the driver have to quote a fair price? Why should the driver forego (higher price) and be altruistic to quote a fair price?
Literature in psychology — game theory suggests that in such contexts, the cooperation is enhanced by a) possible interactions in future; and ii) reputation.
“Possible interactions in future” means that if the driver is to encounter the customer definitely again in future, then there is incentive for driver to act fairly, so that he doesn’t lose his future business. In case of a typical auto — customer transaction, it is highly likely that the auto driver isn’t going to see the customer again. Hence, there’s no incentive for the auto driver to be fair.
“Reputation” means that if other people observe you carrying out the transaction and if it affects your business, then there is an incentive to be fair. For instance, if another customer watches you quote high price, then the other customer might avoid you, should the auto driver meet the customer in future. It’s similar to the case of feedback in restaurant listing sites like Zomato where other customers’ feedback affects behaviour of new customers. In our context, the auto driver doesn’t face any such consequences because the possibility of future interactions with the other ‘observing customers’ is uncertain. The other customers also may not be aware of the prices. More importantly, who would continuously eavesdrop on others bargaining?
The other important factor influencing auto driver’s behaviour is cartel formation. Fragmented markets mean that there’s high scope for cartel formation and all neighbouring autos may quote same price.
Information asymmetry is the final factor. Suppose you land up in a new city and request an auto ride. A driver knows more than you about the distance, regular fare and his costs. Hence, might quote you high price. Probably, you would check with your friend beforehand about regular fare as a remedy. What if you don’t have a friend?
In summary, low probability of future transaction, no influence of current behaviour on reputation and future transactions, information asymmetry, possibility of cartel formation mean that drivers have less incentive to be fair.
This is about auto drivers quoting the price. Let’s come to the customer. What factors make customer accept the price?
One, time available/Opportunity cost. If customers have ample time in hands, they can decline the transaction and initiate bargaining with other, until they reach a price which the customer feels comfortable with. Does everyone have such time? What if you are running late to office? You would probably accept a higher price in first transaction itself to save money. In this case again, auto driver is at an advantageous position as compared to the customer.
Two, information on supply of autos. Suppose, you know that there are no autos in vicinity and there’s no possibility of their availability in near future. May be it’s a rainy day and there is huge demand for autos. Higher quote may be acceptable to the customer on such occasions because the customer is aware of higher demand for autos. In other words, if customer has knowledge of the supply shortage of autos, then the divergence from fair price may be acceptable. Such information is easy to decipher in extreme cases like rainy day.
The question however is — do all customers have such information all the time? If an auto driver quotes a higher price, what’s the guarantee that it’s due to a genuine short of supply and not a deliberate act of over pricing? Even in this case, auto driver is placed at a better position in the bargain.
Together, the information asymmetry, uncertainty in future transactions with same auto-customer transaction, no influence of behaviour on auto drivers’ reputation, higher opportunity costs to customers, lack of reliable source of information on supply means that the bargain is highly skewed in favour of auto drivers and there’s no incentive for them to quote fair price. All of these make a case for standardizing auto fares.
Enter Uber. For a moment, let’s assume a market where Uber and auto aren’t competing. Only Uber exists. Uber addresses many issues observed in the case of auto driver — customer bargain. The possibility of future transactions is increased because you are dealing with Uber as a company here and not a particular driver. So, there’s incentive for them to not exploit. The reputation issue is addressed to an extent because of its influence on future transactions, given that the future transactions are more probable.
Issue of information on supply of cabs is addressed to an extent because the customer intuitively knows that a surge price is due to higher demand of cabs. Unlike the case of auto driver -customer transaction, where a higher quote needn’t necessarily mean higher demand, in case of Uber, people intuitively understand that higher price is due to higher demand and trust Uber regarding this. Technology helps Uber to gather information on local demand and take decisions on demand — price relation.
All well and good. But then two questions arise
1) What led to the demands from public for regulation of cab fares?
2) What led the governments to regulate the cab fares?
Let’s talk about the first question — demands from public to regulate cab fares.
Have you ever wondered that people aren’t ready to pay even 5 additional rupees to an auto driver (bargain really hard) but they are ready to travel on 2x surge price with Uber? (Now, we are taking a detour for a while and venturing into Uber — Auto comparison.)
One simple answer to above question could be that Uber is comfortable with an AC etc and hence people don’t mind paying more price to Uber. It probably explains higher level of Uber fares but doesn’t explain the acceptance of amount charged over and above the normal level Uber’s fares.
A fair argument could be that there’s no bargaining involved in case of Uber. It thus reduces transaction costs. Some might prefer paying 10 extra rupees to avoid a bargain! But, is that the whole story?
E. P. Thomson in his essay “The Moral Economy of the English Crowd” discussed the phenomenon of ‘moral economy’. His argument was that bread riots broke out in England not because of higher prices of bread but because people felt that the procedure through which such prices are arrived at is unfair. Feudal lords were setting prices those days.
Similarly, in our context, customers are concerned about fairness in price determination and not the level of the price (2x or 3x). The fairness of price is signalled through Uber’s pricing mechanism which relates supply shortage to the surge. Customers may not know the exact demand — price relation but they at least know intuitively that Uber decides fares as per demand. Essentially, they trust Uber regarding demand — fare relation. Note that a mere existence of a surge mechanism is not sufficient. Customers have to trust it.
Existence of such signalling mechanism distinguishes price increase due to supply shortage from deliberate attempts to exploit. In case of auto drivers, as discussed above, there’s no mechanism to distinguish the genuine supply shortage (increase in demand) and deliberate attempts to quote high price. Hence, customers can’t trust autos easily and therefore bargain.
In summary, people prefer Uber despite surge due to perception of fairness in price determination mechanism which distinguishes itself from deliberate attempts to increase price.
So far, well and good. What then explains the popular demand from customers to cap Uber prices?
As usual, some columnists have quickly jumped to a conclusion that middle class likes doles and are no different from the poor. Such explanations may give content to fill up space in a column but have little explanatory value.
Why did the public outrage occur then? It’s precisely because people lost trust in the fairness of algorithms that determine demand — price relation. Uber’s prices have also ended up being like demands of an auto driver, indistinguishable from deliberate attempts to quote higher price, with an underlying motive to take advantage of customers.
There were scores of tweets and pics on facebook where people shared pics of ‘what they feel as outrageous charges’. There were also some analyses using publicly available Uber data which illustrated instances where there wasn’t a supply shortage of cabs but still Uber surged the price. These may or may not be true. Remember that no one knows the exact demand — fare relation. Customers are only using a proxy, the surge multiple to decipher demand-fare relation. The perception of fairness is based on trust in Uber’s demand — price relation signalled through surge multiple. Once that trust in surge multiple breaks down and customers start feeling that the surge multiple doesn’t reflect the supply shortage, whatever the reason may be, the demand to regulate fares arise. It’s similar to a demand/need to standardize auto fares to counter unfairness.
In summary, the public demand to cap Uber’s fares is not necessarily due to dole seeking behaviour but it’s due to breakdown of trust on Uber’s fare determining mechanism, which distinguished it from auto drivers.
This is all about our first question — What led to the demands from public for regulation of cab fares? Let’s turn to our second question — What led the governments to regulate the cab fares?
A part of the answer could be related to political economy where unionized radio taxi drivers exert pressure on the government. Let’s understand other reasons.
As argued above, the key to legitimacy of Uber’s surge fares is its demand to price mapping system. A quick recap: Uber’s system signals that the surge is not a deliberate attempt to leverage information asymmetry or exploit the situation of customer, but it is a result of demand-supply mechanism. If the key to legitimacy of surge fares is the existence of demand-price system then such surge may not be acceptable in cases without such systems, like the case of usual taxis. In the absence of such systems in radio taxis, a cap on their fares was the solution.
The question then is — What happens in a situation where two types of services coexist — Uber, with a distinguishing mechanism between demand led surge and deliberate exploitation; and a normal radio taxi system without such distinguishing mechanism. There are three options in such scenario.
One, allow Uber’s fares to cross usual fare caps (applicable to radio taxis) while retaining fare caps on radio taxis. If it’s done, then taxis may object to the obvious differential treatment. Hence, it isn’t a possibility.
Two, remove fare caps on both Uber and radio taxis.
Three, put fare caps on both Uber and radio taxis.
Faced with a dilemma between second and third options, governments prefer to tread the ‘known territory’ of applying fare caps, the third option. So, they end up bringing Uber also under the ambit of fare caps. The alternative of removing fare caps on everyone (second option) is too radical for a government because it doesn’t know its consequences and it’s worried about potential fallouts of cap less fares on radio taxis. The fear is exacerbated by the governments’ usual approach of focusing on containing risks and not on enhancing gains — similar to focus on ‘letting not other team score a goal’ instead of ‘scoring a goal and winning’.
In summary, the demands to regulate Uber’s fares originate because of breakdown of people’s trust in Uber’s price to demand algorithm. Governments are constrained to regulate Uber’s fares because they can’t categorize Uber separately and display differential treatment through different rate structures for Uber and radio taxis.
Now that we understand the problem better, how does one deal with demands to cap surge? As discussed above, the problem is the breakdown of trust on Uber’s price determining mechanism. The solution is to rebuild the trust by making the pricing mechanism transparent. Since, sharing the proprietary algorithm publicly is not a possibility, one should look out for other ways.
It will be useful to think — what information does signal fairness to customer?
1. How much does the trip usually cost?
2. How much more am I being charged now? By what factor?
3. What’s the reason for the diversion? — A proxy to indicate that.
The answers to above questions follow that customers can be given information on usual cost of trip, current cost (after surge) and also the factor by which the price is surged. Regular fare can be determined from current cost and ‘multiple of surge’.
It is important to give information on both surged price and regular price to allow comparison and to let customers know what they are paying for. Uber has shifted from displaying a surge multiple to mere surged price but it’s not helpful because it doesn’t help in comparison with usual price. To enhance the trust and fairness, customers should be given information on — by what factor are they being charged extra?
In an ideal world, it would be better if the customer is also given information on supply shortage, in form of a proxy. Surge price multiple is a proxy for supply shortage but some form of ‘crowd index’ which reflects the density of cars in vicinity can also be displayed, if it doesn’t run into the risk of exposing Uber’s proprietary algorithms.
The final aspect is the government. Increasing prevalence of cab aggregator companies raise important question on role of public and private sector regarding provision of goods and services. Traditionally, public transport was thought to be a government’s duty, at least government was expected to have a larger role. The inability of government to provide satisfying service led way to private companies to fill up the space, performing a duty that’s perceived to be government’s.
The role of private sector in areas where government is expected to take an active role isn’t clear; not just in transport but also in other sectors like education. It raises three concerns.
One, private sector providing services in ‘government’s zone of duties’ are expected to fulfill social obligations and not focus on profits alone.
Two, in systems with low trust on governments, it is also easy to believe that governments don’t deliberately improve public provision of services so as to let private companies take advantage of it.
Three, the prevalence of private sector increases dependency of people on such services. It reduces the incentive for government to invest in public services because people are already getting service from elsewhere. Such balance works fine as long as people trust private systems. The moment trust breaks down, demands for regulation arise citing the public nature of the services. Government is caught off the guard in a catch -22 situation, in the wake of such demands to cap prices. It has neglected public provision of services till then by relying completely on private to provide them. It can’t enhance public services in short time to counter private sector wrongdoings if any. On the other hand, the decision to regulate private sector runs into danger zone of interfering with market mechanism and also evokes an outrage. Once such discretionary power is institutionalized, there is a danger of abuse of such power.
There are no straight forward answers to these but one lesson could be that the government shouldn’t completely rely on private sector mistaking the growth of private sector for people’s acceptance. Such acceptance may be fragile. Government should work towards building public systems so as to counter balance private. Not to mention this is only in context of public services — education, health, public transport. Not obviously applicable to watches, detergent soaps. Are there any other ways to address this conundrum? May be! It needs more thinking.