Professor Bainbridge...on law, politics, religion, culture & food

Why Starbucks Doesn’t Franchise

Keith Sharfman posts at TOTM:

Starbucks’ willingness to sell its product widely rather than reserve it exclusively for its full-service franchisees suggests to me that the firm is competing aggressively in an “output enhancing,” pro-consumer way, rather than seeking to find ways to reduce output and raise price, as some opportunistic antitrust plaintiffs have erroneously alleged.

All therefore seems well for the Starbucks franchise–even if, as Jackie Mason has quipped, it is a bit much to ask customers both to clean up after themselves and also to leave a tip!

One starting problem with his analysis is that Starbucks does not franchise. As the company’s FAQ explains:

Starbucks does not franchise operations and has no plans to franchise in the foreseeable future.

In North America, the majority of our stores are Company-operated. As an exception, Starbucks may enter into licensing arrangements with companies who provide access to real estate which would otherwise be unavailable such as airport locations, national grocery chains, major food services corporations, college and university campuses and hospitals.

I addressed this issue in a post back in 2003, which got mangled in the process of which switching fromn TypePad to Expression Engine. So I’m going to repost the complete text here. (BTW, if you’re a blogger contemplating switching services, always keep a text copy of your old posts!)

My day job is studying the legal rules that facilitate and define governance relationships with the sub-set of economic institutions known as business associations. I decided a long time ago that economic analysis was the only way to make sense of those rules. A bit later I decided that I got the most bang for my analytical buck from transaction cost economics a.k.a. new institutional economics. Most of the time, if I work at it long enough, I can come up with a transaction cost story that explains the particular governance structure I’m studying (at least to my own satisfaction). I’ve done it for things like insider trading, participatory management, the existence of boards of directors, the business judgment rule, and limited liability. There is one governance problem that vexes me, however; namely, why are Subway stores owned by franchisees, while Starbucks stores are owned by the corporation. (Links to WSJ; sub. req’d.)

The usual suspects when it comes to transaction costs include such things as search costs, uncertainty, complexity, bounded rationality, opportunism and shirking, collective action problems, bilateral monopolies, and, especially, asset specificity. When we observe differing governance structures, it is usually because the two institutions in question face differing transaction cost schedules in one of these areas. A classic example is Klein, Crawford and Alchian‘s explanation for why we observe vertical integration of the printing process in newspapers but not in book publishers. The risk of ex post opportunism in the former case coupled with the difficulty of forming complete contracts makes vertical integration the transaction cost minimizing solution for the newspaper.

What bugs me about the Subway v. Starbucks problem is that I can’t see any reason to believe that Subway’s transaction cost schedule is going to differ from that of Starbucks. The businesses are essentially identical; yet, the business model is quite different. A fairly standard transaction cost explanation for franchising is based on the cost of monitoring employees to prevent shirking. In some settings, monitoring must be quite intrusive. One reason Subway is successful is that the experience in one Subway restaurant will be largely identical to that in another restaurant. (When you pull of the Interstate in a strange city, you know what you’ll get.) Replicability requires close monitoring. In turn, this leads to the problem of incentives. How do you make sure that the employees work hard? You hire a supervisor? But how do you make sure the supervisor works hard? And so on. Alchian and Demsetz‘s famous solution to this problem was to give the ultimate monitor the residual claim, so as to provide incentives that monitor to work hard. Franchising gives a residual claimant-like status to the local franchisee, while the franchise contract gives the franchisee incentives to ensure that the local employees comply with brand requirements. Franchising thus can be understood as an adaptive response to the problem of monitoring numerous employees in countless locations.

If this analysis is correct, one would expect to see corporate ownership in settings where monitoring via a vertically integrated management structure can be effected at low cost (relative to situations in which franchising dominates). But does Starbucks really face lower monitoring costs than Subway? If not, did Starbucks make an economic error by not going the franchise route?

Or has Starbucks discovered that franchising wasn’t all that efficient to begin with? In other words, the transaction cost story for why franchising is efficient may turn out to be wrong. The franchiser may not be better off giving the franchisee a residual claim. Instead, vertical integration may be the efficient solution for fast food restaurants.

If so, the decision to go the franchising route might be driven not by monitoring efficiencies but by initial access to startup capital and availability of financing on an ongoing basis (as a smart reader suggested). If franchising requires less startup capital on the franchisor part, since much of the cost of opening new restaurants is borne by the franchisee, a franchisor denied access to capital may have to accept an inefficient monitoring system.

An alternative version of this story depends not on availability of capital but on the consequences of raising capital. A startup franchisor might well be able to raise capital through a public offering of stock, but public ownership entails a potential loss of control for the initial entrepreneur. Where the entrepreneur places a high value on maintaining control, franchising may allow it to raise the necessary startup capital without having to risk the loss of control that might follow from public ownership.

It’s also possible, of course, that the Subway story may have more to do with behavioral economic concepts like path dependence and/or herd behavior than with rational choice among competing governance systems.

Perhaps now you can see why I find this problem both interesting and vexing. (If not, don’t worry; I’ll get back to wineblogging and poliblogging soon enough.) Personally, I remain puzzled and fascinated. It’s fun to spin out the various possibilities.

UPDATE: Tyler Cowen of Marginal Revolution sent along this email:

Starbucks has pursued the unique strategy of not worrying about cross-store cannnibalization, saturating an area with stores, very close to each other, in the hope of building up a dominant brand name. This is harder to do when you don’t own all the outlets, the franchisees fear confiscation of the value of the outlet,etc.

His partner Alex Tabarrok followed up with:

This strategy would probably not have been possible with franchisees since they demand exclusive terrorities and don’t take into account the brand externality.

Joe Carter proved he can do econoblogging in addition to his many other talents, by offering up a more detailed version of these arguments. Finally, Econ blogger Arnold Kling blogged a similar analysis:

Starbucks has an unusual real estate strategy that involves "flooding the zone." They are willing to have several stores near one another. That is the opposite of traditional franchise fast food, in which territory is carefully carved up.

The theory of "flood the zone" is that it is better for the corporation to have a lot of franchises in one territory, even though they appear to compete with one another. You need a corporate ownership to implement such a strategy--you could never convince individual franchise owners to do it.

Well, yes, but. What about the monitoring problem? Has Starbucks come up with a unique monitoring mechanism that allows them to achieve efficient levels of monitoring within a vertically integrated structure? Or have they accepted shirking as a cost of doing business? Actually measuring agency costs is very difficult, of course, but it would be interesting to know whether agency costs are higher at Starbucks than at Subway.

UPDATE2: It often surprises me which posts get picked up in the blogosphere and which sink into oblivion unnoticed. I thought it the puzzle discussed here was a nifty problem, but I didn’t expect it to attract much attention. I was wrong:

Joe Carter: Starbucks vs. Subway: How Strategy Determines Structure
Wayne Eastman: Starbucks and Subway
WR Harper: Starbucks vs. Subway
Paul Jaminet: Subway v. Starbucks: Explaining Vertical Integration with Transaction Cost Economics
Arnold Kling: Starbucks vs. Subway
CE Petit: Joining the Coffee Achievers
Lawrence Solum: Subways, Starbucks & Transaction Costs Economics

These links worked in 2003. No guarantees.