Why vertically integrate?

Lest ye think that I am not a fan of vertical integration, given how firm I am in my arguments for the dis-integration of the vertically-integrated, regulated electric utility, I recommend this post from Seth Goldin about reasons why firms vertically integrate. In large part it’s a concise summary of Klein, Crawford, & Alchian (1978), one of the seminal papers in transaction cost economics. A good reminder of the extent to which the transactional boundaries of firms depend on the relative cost of pursuing similar transactions in markets, and that all of those costs can change over time and with technological change.

Gasoline supply chain stories in post-Sandy New Jersey and New York

Michael Giberson

Edward McAllister and David Sheppard, with Reuters, have a great story on the connection between disaster preparedness and the nature of retailer ownership.

They report that corporately-owned retailers, such as convenience-store chain WaWa and vertically-integrated gasoline refiner/retailer Hess, drew on corporate resources in advance of the storm to be ready to return stores to service quickly. Meanwhile, most of the big gasoline brands in the area–Exxon, BP, and Shell, for example–are small franchise operations lacking deep pockets, geographical scope, and logistical sophistication. Perhaps unsurprisingly, it looks like WaWas and Hess and other large corporate retailers were able to help most of their stores to get back in business and keep them supplied. The smaller franchise operations, on the other hand, had more trouble.

Also probably not too surprising, none of the seven gasoline retailers charged with price gouging in New Jersey on Friday were large corporately-owned stores. (The brands of the stations charged: Lukoil (2), Gulf, Delta, Exxon, BP, and Sunoco.)

As I’ve noted in the past on the topic of industrial organization and price gouging, large geographically-diversified companies are much better positioned to respond to disasters. A company with distribution centers across the region already employing constant-contact computer-coordinated restocking technologies are well position to re-organize shipping patters in the wake (or even in advance of) damaging storms and natural disaster. If nothing else, companies with national reputations can enjoy much broader payback from positive public relations stories than can a two- or three-station chain of gasoline stations.

Of course, one implication of these differences is that anti-price gouging laws will fall heaviest on very small retailers. Surely not the intended result, but nonetheless.

ALSO NOTE: WSJ Q&A on the localized gasoline crises.

Virginia Postrel on Delta’s refinery purchase

Lynne Kiesling

Just a quick note to accompany the discussion in the comments on Mike’s post about Southwest Airlines, Delta Airlines, and fuel price hedging: a couple of weeks ago Virginia Postrel had a very good analysis of the reasons why the Delta-Conoco transaction is not a good idea, in her regular column at Bloomberg View. Virginia’s analysis emphasizes the extent to which vertical integration is only profitable when transaction costs make markets and contracting more expensive ways to accomplish the transaction. In this case, markets do not have substantial transaction costs.

But what about fuel price risk? Here Virginia quotes friend of Knowledge Problem Craig Pirrong:

The proposed purchase “doesn’t make a huge amount of economic sense — in fact quite the opposite,” says Craig Pirrong, a finance professor and director of the Global Energy Management Institute at the University of Houston’s Bauer College of Business.

You might think that owning a refinery would at least protect the airline from price fluctuations. But, Pirrong notes, crude oil prices affect the profits of airlines and oil refineries exactly the same way. When oil prices go up, their profits go down. Owning a refinery would simply magnify the effect. “If anything,” he says, “it increases the risk exposure that has bedeviled the airline industry for years.” …

Delta simply seems to be falling for the great fallacy of vertical integration: the belief that the inputs you get from an in-house supplier are cheaper than those you buy in the open market. There’s no markup. You’ve cut out the middle man!

But this story misses the real cost of those inputs.

Basically, if fuel prices are high, Delta will still not fly those costly half-full flights, but will instead sell their fuel in the low-transaction-cost markets. So what’s the point of owning the refinery when it’s not their comparative advantage and refining is such a low-margin business?

Google, Motorola, and the effects of vertical integration

Lynne Kiesling

Yesterday Google announced its purchase of Motorola Mobility, the device manufacturing half of the former Motorola. Today’s Wall Street Journal has a front page full of stories about this move, including “Bid Comes Amid Tougher Scrutiny” (You know how to read this even though it’s subscriber-only, right? Do a search at a news aggregator for the article title.) Despite being a “new industry” technology merger, this transaction actually is grounded in a long history of the economics of vertical integration, and it may provide another issue for FTC and EU regulators to analyze as they pursue their respective antitrust investigations of Google.

According to the Wall Street Journal article,

Motorola doesn’t have a dominant position in handsets. Five years ago, it sold nearly one in five cell phones worldwide; today it has just a 2.4% market share, says researcher Gartner.

Google doesn’t directly compete with Motorola. Legally, it’s more difficult for the government to successfully challenge such a “vertical” deal between two companies that aren’t direct rivals.

In vertical mergers the dominant competition issue is called vertical foreclosure — when an upstream firm (Google in this case) merges with a downstream firm (Motorola Mobility in this case), does that merger enable the upstream firm to reinforce or extend its market power and market share? And if so, does that increased market power and market share benefit consumers or not? If there are significant economies of scope, then a vertically-integrated firm may enjoy substantially lower production costs than separate upstream and downstream firms (a related issue here is reducing “double marginalization” when the demand curve is downward-sloping in both upstream and downstream markets). Also, does the vertical integration economize on transaction costs — is there a Coase “Nature of the Firm” impetus for vertical integration to lower costs? If those are the merger drivers, then the merger can improve product quality and diversity and reduce prices in ways to benefit consumers, particularly if the downstream market is a rivalrous one. I would certainly describe the mobile device market as rivalrous!

But the vertical merger story is not all rosy and welfare-enhancing. What if the merger enables the upstream firm to leverage its upstream market power into higher downstream market power by reducing availability of its product to its downstream rivals? A more concrete example may be iron ore and steel; what if a dominant iron ore producer merges with a downstream steel firm, and reduces its sales of iron ore/raises prices to the other non-related steel mills? The vertically integrated iron ore-steel firm could increase its steel production by restricting its sales of iron ore to others, and the price of iron ore to independents may increase, which would increase the vertically-integrated firm’s market share in the downstream steel market, and may lead to higher steel prices to consumers. This is called vertical foreclosure.

The concern about vertical foreclosure with Google-Motorola Mobility is access to the Android operating system and updates to it. From the Wall Street Journal article:

On Monday, several handset makers that use Android publicly welcomed the deal. But privately, some may end up questioning whether Google will grant Motorola preferential treatment. For example, Google could theoretically give Motorola its latest versions of Android exclusively, placing them at a competitive disadvantage.

One way for the Justice Department to hold Google to its promise not to discriminate against Motorola rivals would be to seek to implement a so-called “firewall” between Google and its Motorola unit.

In exchange for not challenging the deal, the Justice Department might also ask Google to sign a consent decree legally compelling it to license Android to any comers on fair, non-discriminatory terms.

This issue is one reason that the FTC and the EU may incorporate this merger into their antitrust investigations of Google. For example, as quoted in a Huffington Post article on the merger:

Others were more concerned. Gary Reback, a prominent Silicon Valley antitrust litigator, who has become a vocal critic of Google’s business practices, said that the deal deserved to be closely scrutinized.

“You’re dealing with a company that is already a monopolist, that is already under investigation for allegedly anti-competitive behavior,” he said. “By buying this, they get a huge additional dose of market power.”

At the very least, the Motorola takeover bid provides federal regulators probing Google yet another reason to subject the company to scrutiny.

Mr. Reback clearly sees this as a move that would result in vertical foreclosure, but other economists and lawyers are not as worried about such foreclosure.

I don’t see the vertical foreclosure issue as being particularly problematic here. The mobile device market is very rivalrous, with hundreds of devices and at least four different platforms (Apple iOS, Android, Blackberry, Windows, etc.). It’s also an information-rich retail market, with a lot of consumers who choose Android over Apple because they appreciate the more open Android platform to Apple’s “walled garden” approach to OS and application development; if Google tries to restrict updates preferentially to Motorola devices, that restriction is likely to trigger such “anti-walled-garden” complaints in a way that will rebound negatively on Google. Furthermore, the role of the mobile device carriers (Verizon, AT&T, Sprint, Vodafone, etc.) changes the dynamic between the operating system owner and the device manufacturer, and may add some impetus to ensuring that non-Motorola devices get the same OS updates at the same time.

Rather, I tend to agree with the analysts who say that this move is Schumpterian platform competition on a different front: patents and the threat of patent lawsuits. Last week Apple managed to block release of the Samsung Galaxy tablet in Europe because of their lawsuit alleging Samsung’s violation of Apple patents. Google’s acquisition of a large set of Android device-related patents gives them a stronger bargaining position against Apple and Microsoft specifically:

The amount of money being spent now on patent armaments is staggering. Google will spend $12.5 billion to acquire Motorola. Earlier this summer, a consortium including Apple, Microsoft, and Research In Motion, bought 6,000 Nortel patents for $4.5 billion.

For the emerging patent superpowers–Apple, Microsoft, Google–the likely end result of this nuclear build-up could be a legal stalemate that may in fact free up the giants to maneuver on the product side. While every launch may carry with it a back-room patent agreement (I’ll agree to not sue you on patent X today if you agree to not sue me on Patent Y when I launch my new thing tomorrow), the balance of power could in, in a limited way, work in favor of innovation at the large companies.

This move, too, has a long pedigree in economic history; more on that later.

Retail electric power market shakeout in Texas, II

Michael Giberson

Following up on a Monday post, in the news another report of a Texas electric power retailer seeking to acquire generation as a natural hedge.  From Platts:

Direct Energy said Wednesday that it plans to acquire and/or develop new generating capacity in the US to support its electricity retailing business.

Direct, a subsidiary of UK-based energy giant Centrica, owns gas-fired plants totaling about 1,300 MW in Texas, where it serves about 800,000 retail electric customers, Phil Tonge, president of Direct’s North American mass market business, said in an interview…

“One thing we’ve found–and it’s not specific to Texas–is that there are obvious advantages to owning generation” to back up retail load that Direct serves, Tonge said, noting that access to generation “takes some of the volatility out.” …

Tonge said that it some instances it can make sense for a retailer like Direct to own generation equal to as much as 40% to 50% of its peak needs. Direct said that in Texas it can currently meet 27% of its peak demand from its three gas plants and five wind PPAs.

Cheryl Morgan summarized my post as asking “whether vertical integration might have been a better option for Texas,” but I wouldn’t say it quite that way.  “Vertical integration” in the electric power industry is typically conceived as bundling retail, local distribution, transmission, and generation.  As I recall Sally Hunt’s point (in her book, Making Competition Work in Electricity, and I too don’t have the book handy so I’m relying on memory), she argued that it make sense to unbundle the wires from the non-wires portions of the business, but it wasn’t inherently desirable from a policy standpoint to unbundle retailing from generation.

Once unbundled from the wires business, the retailers’ decision to own or contract for power supplies seems to be just another “make or buy” decision that any business must consider.  Some gasoline retailers are vertically integrated with refiners and crude oil production companies, others are content to buy on the wholesale market.

Morgan worries about the loss of transparency that comes when retailers acquire generation (rather than buying power through long-term contracts and short-term markets).  So long as consumers can switch retailers and have a reasonable choice of alternatives, we don’t need to worry about the loss of transparency.  Consumer switching can discipline inefficient power supply arrangements, whether through contract or ownership of generation.

Which brings us to Morgan’s second point, that the shakeout will eventually result in relatively few large retailers.  If you think that customer choices will dwindle to a few large companies, then transparency may become an issue.

But I’m not ready to lump the California and Texas retail experiences together.  The approach taken in the two states differed and the differences have seemed to matter.

RELATED: an exchange of views in the February 2009 issue of Energy Policy between Christophe Defeuilley and Stephen Littlechild. (Subscription may be required, check with your local library if you don’t have direct access).

  • Defeuilley says experiences with retail electric power “have proven less than stellar” and wants to blame the disappointing results on an inadequate view of competition arising from the Austrian School of economics, which Defeuilley says served as foundation for the reforms.
  • Littlechild disputes both the claim that retail power competition has been disappointing and the claim that lackluster retail competition is somehow connected to Austrian School views on competition. Littlechild adds that the behavioral economics ideas that Defeuilley cites are more consistent with Austrian ideas than with the concept of compeition offered by traditional neo-classical economics.
  • Finally, Defeuilley elaborates on why he characterizes retail competition as falling below expectations and on the limits he sees in the Austrian School conception of competition and why it matters to the case of retail electric power.

The retail electric power market shakeout in Texas

Michael Giberson

One last Texas electric power post for the day, this one on the upheaval among some electric power retailers in Texas, via Platts:

The shrinking field of retail marketers is not an indictment against restructured power markets but an expected result of the credit crisis and its impact on retailers, Jim Burke, CEO of TXU Energy said Monday morning.

Burke’s comments at the KEMA Executive Forum in Houston came after NRG announced plans Monday to buy Reliant Energy’s retail business in Texas for $287.5 million. Integrys Energy Services’ parent last week said it would wind down or sell IES, which is one of the larger retail marketers operating in competitive retail markets…

Integration of generation assets and retail market assets, as evidenced in the NRG-Reliant deal and in the business strategy of marketers such as TXU and others, also makes sense and may be a future trend, Burke added.

In many states that restructured, unbundling of vertically integrated utilities often simply produced the spin-off of retail and generation units into separate entities still under the same corporate parent, sometimes with limited sales of utility-owned generation to independent power companies. The point wasn’t to break up companies, per se, but to separate the power generation and retail ends of the business from the transmission and distribution wires companies. In Texas, they pushed a little harder than most states on developing a competitive retail sector separate from the former regulated utilities.

Unbundling of electric utilities doesn’t require divorcement between generation and retailing, however, and as the comments above suggest, the combination makes some sense. The combined company finds itself naturally hedged against price risk, and it may be cheaper to hedge the risk in this manner than through contracts.

I recall that Sally Hunt makes this point well in her book Making Competition Work in Electricity.

While many restructured states are not experiencing a similar “shakeout”, it is mostly because they never saw much of a “shake in” in the first place.