Global Economics and Management

Retail Discounts

The Wall Street Journal recently had a great feature on retail discounts - just in time for Black Friday! The article points out that discounts are not really discounts because the price reductions are applied to an artificially inflated base price. You might as well say that the high price is the price for those willing to pay a lot, while the discount price is in fact the "normal" price. The WSJ calls this a "dirty secret", but of course all of us in Management 405 already knew that!

The article is interesting because it shows that many more companies nowadays have come to find discounts to be exactly what they are, which is a way to extract more surplus from consumers, to the benefit of the sellers. Money quote: "Retailers didn't always price this way. It used to be that most items were sold at full price, with a limited number of sales to clear unsold inventory. That began to change in the 1970s and 1980s, when a rash of store openings intensified competition and forced retailers to look for new ways to stand out. (...) But the floodgates have opened. In a 2012 presentation, Mr. Johnson, then still Penney's CEO, said the company was selling fewer than one out of every 500 items at full price. Customers were receiving an average discount of 60%, up from 38% a decade earlier. The twist is they weren't saving more. In fact, the average price paid by customers stayed about the same over that period. What changed was the initial price, which increased by 33%."

Game Theory and Soccer Penalty Kicks

In class yesterday we discussed how game theory can apply to the analysis of penalty kicks in soccer. I mentioned an academic paper by Chiappori, Groseclose and Levitt showing that people indeed pursue mixed strategies in soccer - a rare instance of people playing mixed (probabilistic) strategies in the real world. You can find the paper here. 

Your classmate Brian Bainum followed-up with the following observations:

"You mentioned briefly penalty kicks today during our discussion of game theory, and I wanted to bring to your attention a YouTube video from the Champions League final penalty kick shootout of 2008 between Chelsea and Manchester United. This episode was discussed at length during a chapter of the book “Soccernomics” but the main point was that a Spanish economist Ignacio Palacios-Huerta had observed penalty kicks at length and discovered that Manchester United goalie Edwin Van der Saar tended to dive to the kicker’s natural side (ie a right footed kicker would kick it into the left corner of the goal). He shared this information with the Chelsea manager prior to the game.

In the penalty shootout, Van der Saar lives up to this tendency 5 of 7 times. Chelsea kicks 5 of 7 to the non-natural side because they had obtained this information ahead of time. In the final kick, you can see the game theory playing out because Van der Saar points to his left, almost daring Nicolas Anelka to kick to his non-natural side like he might have intended to. Anelka then kicks to his natural side and Van der Saar makes the save.

You probably have seen this already, but here is the link."

Guest Blogging: A Follow-up on Stack Ranking

Yesterday in class we discussed a Wall Street Journal article about Microsoft's decision to abandon the practice of stack ranking its employees. Your classmate Leslie Kwoh, who used to write for the Wall Street Journal, sends this follow-up:

"I wrote this piece about forced ranking for WSJ last year. I must say that while the practice does have its downsides, namely creating a cutthroat and backstabbing environment, I can also understand Jack Welch's point when he says: "This is not some mean system—this is the kindest form of management," This probably would not apply, of course, if a company is so lean and efficient that there are very few underperformers.

Also, I do think there is a negative connotation associated with the phrases "forced" and "stacked ranking." After all, many consulting firms and technology firms use an "up-or-out" policy, which similarly involves quotas and evaluates employees relative to others, but they rarely get bad press for it. McKinsey, for example, requires that its consultants move "up" or move "out" every two years; those who don't make the cut are gently transitioned into another job with career counseling services. Yet, employees I've interviewed there have praised the process. So there is probably something to be said about how managers carry out these ranking practices (soft skills, I know) that can have an impact on the effectiveness."

I agree with Leslie that framing matters, and that the subtle details of the implementation matter as well. For instance, who does the evaluation and whether the process seems fair and transparent is of the essence. From an economic standpoint the idea of having evaluations and imposing rewards for strong performance and punishment for weak performance seems basically like a good idea. Incentives matter. But you have to ensure things don’t backfire, and that is a much more difficult thing to achieve than coming up with the basics of a system of extrinsic incentives.

Google's 2004 IPO: A Modified Dutch Auction

We all know that Google is a very sophisticated player when it comes to auctions. For instance the position of an advertiser in a search result list is determined through second-price auctions - more specifically a Generalized Second Price Auction (GSP - wonks and nerds please go here for details). Little known is the fact that Google conducted its IPO as a Dutch Auction - although once again it was not a traditional Dutch auction but a modified one. Your classmate Faizaan Ghauri sends this useful Forbes article. More details on the auction format that was employed in the Google IPO can be found here.

Basically, imagine Google wants to sell a million shares. The auctioneer rankes the bids and the number of shares demanded by each bidder from highest bid to lowest. The lowest bid above which there is demand for a million shares becomes the IPO price. For instance suppose there are 3 bidders. Bidder 1 bids $100 for 500,000 shares, bidder 2 bids $90 for 500,000 shares, and bidder 3 bids $80 for another 500,000 shares. The highest two bids clear the market for Google shares (500,000+500,000=1,000,000). The IPO price is the price quoted by the second highest bidder ($90). Everyone to whom shares are allocated pays this price. Bids at the top of the distribution determine the IPO price, approximately like in a Dutch auction.

California Versus Texas

The Washington Post recently ran a very interesting piece on the continued widening of income inequality in the United States in the wake of the Great Recession. The article includes a graphical display illustrating the change in the number of households at each income level between two periods - 2007-2009 and 2010-2012. Here are the pictures for California and for Texas:

 

The differences are striking. I don't know exactly to what they should be ascribed. But I note that California is often pointed to as a state with generous policies toward the poor, and Texas a hard-nosed, market oriented, fend-for-yourself state. Yet Texas has managed to grow the upper middle class without hollowing out the middle class, and has experienced only modest growth in the number of households earning less than $25,000. California has experienced a dramatic hollowing-out of the middle, and a very high growth of low-income households. Worth pondering...

Guest Blogging: Big Data at Disney

Following up on our discussion last Thursday of the use of big data in the airlines business, your classmate Greg Holcomb sends me these fascinating observations on Disney's similar attempts. I will let Greg speak in his own voice for the remainder of this blog post:

Our discussion on the expanded customer data available to airlines reminded me of a similar initiative already in-progress by Walt Disney World in Florida. Disney’s system is called MyMagic+ and involves theme park visitors providing demographic information and planning their vacations months in advance to get faster access to three rides per day and a more personalized experience via radio frequency wristbands. Disney characters would have the ability to address children by name, know whether it’s their birthday, and know what they have seen on their vacations already. In exchange, Disney gains volumes of information on demographics and information on how often visitors come to their parks and the order in which they experience different elements. Visitors could opt out of providing this information, but by doing so they would lose the opportunity to book experiences in advance and thereby be forced to wait in longer queues relative to their counterparts who provide the information.

Another aspect is the ability to speed up transactions by charging directly to the wristbands and keeping purchase history on individuals across visits. Eventually, special moments could be sold at extra cost or bundled with stays in some of the more expensive hotels on property. Since hotel and restaurant bookings occur online months in advance, it would be relatively easy to price discriminate between different customers without their realization of different prices in the market.

This article in the Orlando Sentinel provides more information. 

There is a similar “creepiness” factor and privacy concerns as experienced by the airlines, and Representative Ed Markey of Massachusetts wrote a letter expressing some of these to Disney’s Bob Iger in January: 

Also interesting is that the cost of this initiative is over $1 billion which is enough to build an entire theme park. This suggests that the potential new revenue from analyzing existing customers and monetizing new experiences within the existing properties is greater than the potential extended length of stays and ticket revenue that a physical expansion would bring.

Real Life Cartels

In class last time, we discussed the formation and stability of cartels. In principle, cartels are illegal, and price fixing is a per se illegal antitrust violation. However, in some instance industries have obtained exceptions to those legal standards. A particularly extreme example of such an exception was sent to me by your classmate Michael Peck. You can learn all the details (and review many of the course themes) by listening to the following 20-minute NPR Planet Money segment on the raisin industry in California. Protected by a special law dating from the Great Depression and WWII, that industry can legally manipulate raisin prices by controlling the supply of all the raisin growers. What is interesting is that, in this case, it is violating the cartel's rules that happens to be illegal! When cartels are formed they are inherently unstable because individual cartel members have an incentive to expand their production and free ride on the high price brought about by the cartel's supply restrictions (this is to the benefit of consumers, who experience lower prices). The NPR segment describes the experience of such a free-rider, Marvin Horne. Mr Horne was sued for damaged by the raisin cartel. Talk of a world upside down!

Other examples are found in international cartels, that are often beyond the reach of domestic antitrust enforcement. The most famous example is the OPEC oil cartel, which we often discussed in class. Another is the diamond cartel, headed and organized by the diamond giant De Beers. Recently in the news, you might have read about the collapse of the potash cartel. Even when antitrust enforcement is ineffective, you can count on the inherent instability of cartels to ultimately bring about their downfall. But in the meantime, cartel members can enjoy very hefty profits indeed.

Online versus Brick-and-Mortar Stores, continued

Following up on last Thursdays interesting class discussion on threats to brick and mortar stores from the emergence of online sellers, your classmate Leslie Kwoh sends me these interesting observations and weblinks:

"I wonder if you saw this USA Today article about Amazon's efforts to court brick-and-mortar retailers in order to sell more Kindle devices through physical retail locations. 

Target actually used to sell the Kindle, but stopped doing so because Amazon was undercutting business by offering promotions that encouraged showrooming (read about the fiasco here). Now it appears that Amazon is bending a little...It's an interesting acknowledgement of the limits of online retail, as I'm guessing that Amazon simply hasn't been able to attract enough Kindle buyers through its website. Perhaps when it comes to some products -- like expensive technological devices -- the "touch-and-feel" experience is still irreplaceable. (That may be a competitive edge for Barnes & Noble's Nook?) 

It's also interesting to see the lines increasingly blur between online and brick-and-mortar retailers. As this New York Times article pointed out last month, retailers like Wal-Mart are moving into Amazon's turf, while Amazon is expanding its physical presence with more warehouses. The two will probably meet somewhere in the middle, as the article notes, which I suppose also means there will be little difference in cost structure."

I think Leslie's observations on the convergence on online and offline merchants are spot on.

Online versus Brick-and-Mortar Stores

Tomorrow at the beginning of class we will discuss this WSJ article: Fear of 'Showrooming' Fades.

On that occasion I wanted to point to a pair of older GEM blog posts touching upon the same issue:

1) Price matching of online prices by brick-and-mortar retailers.

2) Online retailers like Bonobos are increasingly opening brick-and-mortar showrooms.

Uber's Peak Demand Pricing

Yesterday in class we discussed the taxi medallion system, which limits the entry of additional taxicabs when demand expands, resulting in steeper prices for consumers and rents accruing to cab company owners and politicians. I pointed out that such rents invite the entry of enterprising alternatives. Among those, SideCar, Lyft and Uber have received some recent attention, as well as resistance to their entry from the entrenched interests that benefit from the monopolistic system that politicians and cab companies have managed to put in place. 

In this context your classmate Neil Kapur sends me the following two interesting links on Uber's peak pricing schemes: Article #1; article #2. It turns out that Uber raises prices very steeply in periods of peak demand, such as NYE or Halloween, or even around natural disasters. This has prompted pushback from some customers, but Uber's CEO has pointed out that without peak pricing, there would be excess demand and lots of frustrated customers unable to find rides. 

Neil wondered whether this behavior could be interpreted as price discrimination. The answer is yes and no. I think what you have here is a company trying to manage its quantity demanded in the context of capacity constraints. Absent capacity constraints that firm may prefer to elect a different pricing scheme – maybe a flatter one, because it would be free to set the quantity it wishes. With capacity constraints Uber cannot optimize price versus quantity the way it wishes because it cannot serve high demand. Thus, the only choices are either to charge a price that is too low and face demand in excess of capacity (and unhappy customers who cannot find a ride at a price they would be willing to pay), or to raise price steeply to meet the capacity constraint exactly.

The result is similar to price discrimination because the price charged is high when demand is high and inelastic, and price is lower when demand is low and more elastic. So you can think of it as price discrimination, but what is different is that without the capacity constraint the firm would probably adopt a different pricing scheme.

Incidentally the capacity constraint at peak times shows that there really is a scarcity of cabs in New York City, a result of the barriers to entry that the industry imposed. As I think I mentioned in class, there are roughly as many cabs in New York City today as there were in 1937.