Premium natural and organic bulls**t

Geoffrey Manne —  6 June 2007

It appears that the FTC is moving to stop the proposed Whole Foods/Wild Oats merger.  Says the FTC:

If Whole Foods is allowed to devour Wild Oats, it will mean higher prices, reduced quality, and fewer choices for consumers [in the premium natural and organic supermarkets market]. That is a deal consumers should not be required to swallow.

Very punny.  I’d be laughing hysterically if I didn’t find this conception of market definition ludicrous almost beyond belief.  I haven’t seen anything so absurd since the allegation (by the reliably-interventionist AAI, at least; not the DOJ) that front- and top-loading washing machines are in different markets.

I would think that the triumphant entry of Wal-Mart alone into the natural and organic foods market (ooh–but not premium natural and organic! Some people just won’t buy their pesticide-free bok choy at Wal-Mart, you know) would save this merger.

But notice something important.  The FTC doesn’t claim that the relevant market is the market for natural and organic food.  The market is for natural and organic supermarkets.  The agencies have been down this road before, mistaking channels of distribution for relevant markets.  On the same grounds it stopped the Libbey/Anchor-Hocking merger and the Staples/Office Depot merger, to say nothing of a number of other grocery store mergers.  Now, in some cases, it may be that the market is defined by the merging distributors (as many would say was true in the Staples case).  But because economically-relevant market definition turns on demand elasticity among consumers who are often free to purchase products from multiple distribution channels, a myopic focus on a single channel of distribution to the exclusion of others is dangerous.

In other words, there is a serious risk of conflating a “market” for business purposes with an actual antitrust-relevant market.  Whole Foods and Wild Oats may view themselves as operating in a different world than Wal-Mart.  But their self-characterization is largely irrelevant.  What matters is whether customers who shop at Whole Foods would shop elsewhere for substitute products if Whole Food’s prices rose too much.  The implicit notion that the availability of organic foods at Wal-Mart (to say nothing of pretty much every other grocery store in the US today!) exerts little or no competitive pressure on prices at Whole Foods seems facially silly.  

Market definition is thorny, and it has (unfortunately) become the end point, rather than the starting point, of merger analysis at the agencies.  I don’t know what a hard look at the data would show in this case, but my strong suspicion is that the FTC is choking on its organic, locally-grown edamame with this one.   

(By the way–for an extended meditation on these issues, see my article with Marc Williamson, Hot Docs vs. Cold Economics).

Geoffrey Manne


27 responses to Premium natural and organic bulls**t


    Very nice. This has been the best reading entertainment I’ve had this month. Sit back, won’t be long now…


    How ironic is it that the VERY SAME Wall Street Journal issue announcing the FTC’s bone-headed move also included this article:

    After years of decline brought on by fighting Wal-Mart Stores Inc. on price, the nation’s grocery chains are on the mend.

    The supermarkets are winning back shoppers by sharpening their differences with Wal-Mart’s price-obsessed supercenters, stressing less-hectic stores with exotic or difficult-to-match products and greater convenience. …

    Earlier this decade, the hidebound supermarket business was expected to fall before Wal-Mart’s aggressive supercenter rollout and the rise of membership clubs like Costco Wholesale Corp. and high-end specialty chains like Whole Foods Market Inc. Many chains did collapse — 26 filed for bankruptcy earlier this decade, unable to match the falling prices of their better-run rivals — and a wave of consolidation swept the business. But the survivors rallied by redesigning stores, introducing a more relaxed shopping experience and marrying low-priced staples with higher-margin breads, meats and wine. Now, the stronger chains like Kroger Co. and SuperValu Inc. are taking market share from weaker, often regional, grocers. …

    Many of the chains are still learning to sidestep Wal-Mart. They are cutting back on drugs and health and beauty products, which are Wal-Mart strengths, to stress fresh produce, higher-quality meat and easy-to-prepare foods. Subdued lighting and high-end selections buttress the nonsupercenter experience. Instead of the rows of aisles with commonplace brands, the supermarkets are adding tables providing ingredients for planned meals, luring the kind of customer who shops for dinner instead of stocking up on groceries once a week, says Paul Weitzel, managing partner at grocery consultants Willard Bishop LLC. Mr. Frondorf says he was pleasantly surprised recently to find Kroger carried the walnut oil he needed for a gourmet recipe.

    Safeway Inc. has converted about half of its 1,755 stores into “Lifestyle” markets with wood floors, on-site bakeries and high-end private-label brands. The third largest food retailer after Wal-Mart and Kroger, it expects to convert all its stores by 2009.

    Safeway has also invested in precise temperature controls for its produce and other perishable foods as they move from suppliers to stores. And it strives to find food its competitors don’t offer, says Steven A. Burd, Safeway’s chairman. For instance, it worked with growers to get individual-sized watermelons two years before others. It also works with a single meat supplier to offer its own brand of tenderness-tested beef. The business picked up, says Mr. Burd, when “we started behaving more like a consumer packaged-goods company.”

    Supermarkets “have come to the understanding they can’t put cookie-cutter stores out there anymore,” says Sandra J. Skrovan, a senior vice president at TNS Retail Forward.

    This article suggests two important things that are relevant to the FTC’s decision. First, the idea that other grocery stores don’t provide the Whole Foods/Wild Oats experience (assuming “the experience” is indeed a relevant consideration) is just silly. The trend for many national chains is in the WF/WO direction, as they realize they can’t beat Wal-Mart on price. Second, there are huge economies of scale in the grocery business. The chains that have failed have been regional chains that have had a more difficult time taking advantage of economies of scale.

    In opposing this merger, the FTC is likely squandering real productive efficiencies in order to avoid allocative inefficiencies that are almost certain not to materialize. Long live Vons Grocery. Sigh.


    I first and last and in between suggested that both supply and demand-side effects were relevant to the relevant market determination; never just a single factor. In the absence of econometric data, I think the fact that two products are functionally identical is good evidence that the cross elasticity between them will be very high. I also think (not seperately, but in addition) that the fact that manufacturers of one type of machine could easily begin producing and selling the other suggests that the market should include the productive capacity of manufacturers of both types. Frankly, I would probably add a lot of other manufacturers (“uncommited entrants”) into the market, as well.

    I backtracked not at all from my original position. What I did was to try to address your questions on their face. You asked about relevant market analysis and I answered. My willingness to answer does not mean I endorse the relevant market analysis; only that I recognize that it is what we have and that I am able to converse about it even though I have problems with it. Let me add quickly–I don’t have a very good replacement for it; it may even be the best we have. I happen to think our best isn’t good enough. In the immortal words of Ronald Coase (paraphrasing), “sometimes ‘do nothing’ is the right answer.”

    Bottom line on Whole Foods: If Whole Foods is able magically to raise the price post-merger and earn supra-competitive profits on naturally-filtered flaxseed oil, one cannot assume the competitive response of the rest of the retail world (meaning those in the “market” and folks outside the market now who could come in) will be to sit on their hands. Rivals will reposition their own images and brands to get a piece of the monopoly action. Our unilateral effects models do a horrible job accounting for repositioning. Plus, there is not a great deal of evidence that these models actually predict competitive effects well. I believe that the tools we have to assess the economic consequences of mergers ex ante are weak, and, on the basis of sound intuition and some armchair economics (I confess, I haven’t done the econometrics; I believe I don’t have to, most especially because this is a blog, not a refereed journal), I think it’s a mistake to bring just about any any retail merger challenge, and most especially this one. But, hey, if I’m wrong, I’ll eat my hat. With a nice fair-trade quinoa and hydroponically-grown nettle salad.


    No, I didn’t change factors. I was responding to your point. I wasn’t suggesting a new reason. You first suggested that they were in the same “relevant product market” because they had the same end use. And then you said they were because a manufacturer could re-allocate resources to produce different ones quite easily. Either way, the fact that they have the same end use and/or that a manufacturer could shift its output does not place those products in the same relevant product market. You suggested that those factors independently would be an appropriate proxy for determining whether two products are within the same relevant product market. (I acknowledge, however, that those might be relevant factors in determining whether products are within the same relevant product market.) In any case, there was nothing curious about my responses.

    You also seem to backtrack from your initial position. It appears that you acknowledge that the relevant product (as well as geographic market) are important components of an antitrust analysis but fault the FTC for improperly using those analyses in deciding whether to challenge a particualr merger.


    Alex: You have twice now suggested that I make my relevant market determination on the basis of a single factor. But each time you have cited a different factor. Curious.

    What I did say is that supply and demand elasticities are both essential to determining the relevant market (despite what the Merger Guidelines might have you believe). Although Mr. Market Failure adduces an alluring set of facts to describe a situation where top and front loaders might not be perfectly interchangeable, his example is worthless. The two types need not be interchangeable in every situation and for every person ex post for the two to be in the same economically-relevant market. Ex ante, even in his example, the type of available washing machine could factor into the apartment purchase decision, even if ex post only one type would be available in a given apartment.

    But regardless, and to answer your question: Yes, a Cadillac and a Cavalier are probably in the same market. That they are does help to demonstrate the limitations of market share analysis. In that particular case, while Cavalier probably exerts little demand-side pressure on the price of a Cadlillac (although the opposite is much more likely), the potential availability of Cavalier manufacturing resources exerts significant pressure.

    On the point about throwing up our hands–yes, that’s my point. Just because you want to intervene doesn’t mean you should. If we don’t have enough information to get the analysis right, then we shouldn’t intervene. Why is this so hard to accept? Maybe someday we’ll have the werewithal to make accurate and efficient assessments of merger effects. Then by all means, let’s release the hounds. But if we don’t have the tools available now, we should refrain. See my comments on Type I errors, above.

    I do wish that in our current situation the relevant market analysis were really the starting point. Unfortunately, as a practical matter (and this is especially true in the case of the peculiar mechanics of FTC merger enforcement), the product and geographic market analyses are often enough on their own to end the analysis and the proposed merger. I would be a lot less concerned about the limitations of our ability to do relevant market analysis if this weren’t true. See, again, my point about Type I errors, above.

    market failure, right here 7 June 2007 at 2:46 pm

    Interesting. Well, thanks for the reply GM, even if we clearly have different perspectives.

    Just to put your cards on the table a little more explicitly, would you care to confirm that, in your opinion, the world would be better off if we: 1) abolished the FTC and removed antitrust enforcement authority from DOJ; 2) decriminalized price fixing and bid rigging; and 3) repealed the Sherman Act, Clayton Act, and FTC Act?

    Is it that you think the market always takes care of itself, or just that we lack sufficient information to adequately distinguish anticompetitive from procompetitive behavior?


    Market failure: I don’t think I ever said that a tomato from Wal-Mart is identical to a tomato from Whole Foods. Actually, it doesn’t even require any pop psychology to appreciate the difference. I recognize there are such things as brands and that price is not the only thing that differentiates products. Yes, there may be a cognizable difference between shopping at Wal-Mart and shopping at Whole Foods. My response to all that is–so what? The question for relevant market analysis is, effectively, does the availability of one tomato exert competitive pressure on the seller of the other tomato? That doesn’t require that the products be identical, nor does it require Cass Sunstein to tell us that people are boundedly rational. But at the end of the day, demand curves slope downward, and if there’s profit in it, Wal-Mart will look more like Whole Foods and Whole Foods will look more like Wal-Mart and many more millions of people will be added to the ranks of the many, many millions who currently shop at both stores. It may just be a matter of my faith, but I’m pretty sure that if the merged Whole Foods/Wild Oats store starts acting like it has a monopoly on organic tomatoes, Wal-Mart will become a more significant competitor in the organic tomato market — even the one with all those snooty, boundedly rational Whole Foods shoppers. And it doesn’t require one iota of faith to know that all those other stores that are not Wal-Mart will definitely step into the breach.

    With me at the helm, all FTC emplyees would have permanent time off; yes they would have lots of free time (even Josh–sorry, Josh). I guess you meant that as a hyperbolic criticism, but it’s probably the one characterization of my position that you actually got correct.

    Oh–the point about Type II errors: If you think the background rule is that most behavior is anticompetitive, then perhaps you are right. But that is an unrealistic and unwarranted assumption. The reality, as even the staunchest of interventionists would probably tell you, is that most behavior is procompetitive. Given that background and imperfect information, I’m quite confident that Type II errors are more likely than Type I errors. Now, I also think that because of the systematic exclusion of dynamic effects from most merger and other antitrust analysis, and because of high litigation and enforcement costs, the costs of Type II errors are also much greater than the costs of Type I errors. So we should worry a lot more about the one than the other.

    market failure, right here 7 June 2007 at 2:23 pm

    Responding to Alex’s post, which I think hits on some of the same issues raised in mine, isn’t it possible that: (1) front-loaders are the only types that can be used by people living in apartments, condos, and other condensed living quarters, and (2) there are significant barriers to entry preventing manufacturers of top-loaders from shifting production to front-loaders?

    If those two things hold, wouldn’t it tend to look like they’re distinct markets? I have no idea whether they are in fact true, it’s just that I think it’s your simple reliance on functionality that’s “crude,” not carefully considered market delination in the merger guidelines.


    None of that answers my question. You simply assume that top-loaders and front-loaders are substitutes for one another because manufacturers can “cheaply” produce the the top (or front) loaders. Again, are you saying that a Cadiallac and a Cavalier are substitutes for one other because they are both cars and that GM could presumably “cheaply” switch from one to the other? (Or other manufacturers could just switch production from one set of cars to another in response to market conditions.)

    Assuming that the relevant market analysis is “too crude,” your response seems to be throwing up your hands and saying we don’t know what to measure or how to measure it. How would you propose we measure the competitive effects of a (would-be) merger.

    But that said, I am also not entirely clear what you are referring to when you say that relevant market analysis is “too crude.” Too crude for what? The relevant market analysis is not the beginning and the end of determing whether a would-be merger is anti-competitive. Rather, the relevant market is a starting point for the analysis. It gives us an idea (hopefully a clear one) of which products are actually substitutes for the products produced by the would-be merged entity. Indeed, determining the relevant market is important in order to do the other things you suggest, for example if it is possible to consider entry effects.

    market failure, right here 7 June 2007 at 11:58 am

    errata– I meant to say in issue 1, “Whether economics is informed by . . .”

    market failure, right here 7 June 2007 at 11:56 am

    As I see it, there are three basic assumptions that do not share:

    1) Whether is informed by psychology, anthropology, sociology, and many other, related social sciences, or is an entirely autonomous discipline, uncovering basic metaphysical truths about the universe.

    2) Whether Type I errors are always perferable in merger regulation to Type II errors.

    3) Whether business documents, market participant opinions, economic data, and natural experiments have useful contributions to make to merger analysis.

    As far as issue 1 goes, I think the case has been made pretty convincingly by Sunstein, Jolls, Ayres, and others that economics always contains certain basic assumptions about human behavior. Why not make those assumptions as realistic as posisible? Why not utilize insights from psychology and other related social sciences?

    With respect to Whole Foods, we would try to analyze the psychology of the Whole Foods customer base to see if any of them consider WalMart to be a reasonable substitute. An orgainic tomato from WalMart probably isn’t the same to them as one from Whole Foods. To many people these days, shopping and eating as much a political statement as voting. (pace Omnivore’s Dilemma, Peter Singer)

    Issue 2–I know it’s an article of faith in the Chicago School that Type II errors are always more harmful than Type I errors. But I haven’t actually seen any empirical support for this proposition. I simply seems to be such a central dogma that so undergirds Chicago school analysis that it’s no longer even questions it.

    But even if true, one also needs to consider the relative probabilities at stake. If a particular decision rule virtually eliminates Type II errors, but dramatically increases Type I errors, shouldn’t we account for that fact?

    Issue 3–For the people that actually *do* merger and antitrust analysis, business docs, interviews with market participant (customers, competitors, suppliers), natural experiments, and market data are all essential tools of the trade.

    But for you, a little armchair a priori economics is sufficient to take the place of all this.

    The FTC would certainly have a lot more free time with you at the helm. 30 seconds of armchair analysis would dispose of just about every merger, without any of that nasty doc review, market interviews, data analysis, or consideration of natural experiments. But if you don’t care about Type I errors, then maybe that’s okay.


    Relevant markets are determined by supply and demand elasticity. If two products are almost-perfect substitutes for each other, the cross elasticities will put them in the same market. This would be bolstered by the fact that suppliers of one type of washer could simply and cheaply begin manufacturing the other type if there were more profit in it. But this all begs the question–why do we care about relevant markets anyway?The only reason to ask the question about relevant markets in the first place is because our tools of analysis are too crude to assess likely competitive effects of a merger in the contect of the entire economy across time. But technically, that is what is relevant. The relevant market analysis is a relic of market share or HHI analysis, which in turn we use because we don’t know how to measure dynamic, market-wide efficiency effects. Even if a crude HHI analysis seems to suggest that the Maytag-Whirlpool merger will permit the merged company to raise prices more than 10%, why should we be confident that that measure captures everything of relevance? What about incentives to innovate? What about multi-part pricing? Do we really know what entry will look like, and how it will respond to increased prices? Etc. These problems are not merely trifles on the fringes–they are core, essential problems with our mode of analysis. The hubris lies not in claiming that intervention is unwarranted, but in believing that we have enough information to be sure that it is.


    I meant “relevant market” — sorry for the typo


    Because an item accomplishes the same task cannot be the sole basis for concluding that they are in the same relevant product. Would you say that “cars” is the relevant product market if one were to consider a merger between say GM and Ford?


    Because they accomplish precisely the same task. Customers view them as interchangeable. Manufacturers that make one type can easily make the other. Prices for one type of washer are surely a constraint on prices for the other type. Etc.

    I believe all of this in much the same way that I also believe that Deloreans (look here, for you youngsters out there) are in the same market as cars with regular, vertically-hinged doors.


    Why do you believe that the top and front load washers are not separate markets?


    Gosh, I didn’t know that views on the economic propriety of merger interventions mapped to the political left-right spectrum. I certainly wouldn’t claim to be to the “right” of Majoras, but then I don’t really know her politics, and, if you think mine are far right, you obviously don’t know mine.I find it interesting that out here in ideologue land, where we can’t converse rationally, we’re having a covnersation about the systematic problems of merger review at the agencies, but over there in objectively-rational land, disagreement with the infallible FTC is prima facie evidence of hubris and heresy.Your brief stab at anything resembling a logical argument (about Whole Foods shoppers defecting and Whole Food’s pricing structure) makes, more or less, precisely my point: The FTC (and, apparently, its apologists) has a tendency to view markets in terms of readily-approachable indicia like channels of distribution which may or may not actually map onto economically-relevant reality. I know, because I have seen it, that FTC staff support enforcement recommendations with reassuring heuristics like: “well, I wouldn’t shop at Wal-Mart, so it must be in a separate market.” As I said, I have not seen the data, but my intuition is pretty strong on this one: FTC staffers and you, “market failure, right here,” are not the marginal consumers. And even leaving direct competition between Whole Foods and Wal-Mart aside (although one should not, because, your quaint class consciousness notwithstanding, I’m confident that there is plenty of cross-over between Wal-Mart and Whole Foods shoppers), there is surely strong competition between Whole Foods and [Safeway, QFC, Costco, Harris Teeter, Wegman’s, Kroger, Giant, etc.], to say nothing of competition between all of these stores and Wal-Mart.More to the point, grocery store prices are determined by a fairly complex array of factors, many of which (including demand elasticity) almost certainly put all grocery stores (to say nothing of the potential grocery stores that haven’t yet been built) in the same market. Organic foods compete with non-organic; prepared foods compete with non-prepared foods (and restaurants); manufacturers (not just retailers) have lots of say in pricing; the list goes on. Compartmentalizing for purposes of easier antitrust analysis is like looking for your keys where the light is better, not where you lost them. I understand the desire to make complex markets accessible; I just don’t think it tends to yield the right results.But that’s just the right-wing, fascist ideologue in me talking.

    market failure, right here 7 June 2007 at 9:05 am

    I’d like to add that the Commission’s vote was 5-0 in favor of blocking the merger. That includes some pretty conservative votes, including the Chairman.

    Manne–if you’re positioning yourself to the right of Majoras, you’re so far outside the mainstream that you probably can’t even converse rationally about these issues. That’s ideologue-land.

    market failure, right here 7 June 2007 at 8:43 am

    Are you actually suggesting that Staples-Office Depot was incorrectly decided? That’s a rather incredible position, one that puts you squarely outside the mainstream of contemporary antitrust thought.

    Leaving that aside, do you seriously think that a significant number of Whole Foods consumers would defect to WalMart in response to a Small But Significant Non-Transitory Increase in Price (SSNIP)? The demographics are completely different. Do you think Whole Foods accounts for WalMart in their pricing structure? Their strategic decision-making? That’s laughable.

    Why not show a modicum of respect for the FTC staff who have actually reviewed the documents, deposed the corporate witnesses, and crunched the economic data. Your hubris is breathtaking.


    I’m with you on this one, Geoff. The barriers to entry here are virtually nonexistent. With just a phone call, any “non-premium” supermarket could directly compete with Whole Foods on any product it wishes to. And needless to say, there are already many food places besides Whole Foods at which premium organic products are already available.

    You put it very well by suggesting that “[t]he agencies have been down this road before.” One would have thought that by now the lessons of Von’s Grocery had been well learned!

Trackbacks and Pingbacks:

  1. The Volokh Conspiracy - August 16, 2007

    Judge Exhibits Common Sense in Whole Foods/Wild Oats Case:…

    A federal judge has denied the FTC’s request for an injunction to block the proposed merger between Whole Foods and Wild Oats. I was just telling co-c……

  2. TRUTH ON THE MARKET » Manne Vindicated! - August 16, 2007

    […] Oldwestrider on Premium natural and organic bulls**t.TRUTH ON THE MARKET » Chemerinksy’s Theory of the Roberts’ Court’s Antitrust Jurisprudence on My Take on Credit Suisse . . ..TRUTH ON THE MARKET » Chemerinksy’s Theory of the Roberts’ Court’s Antitrust Jurisprudence on Credit Suisse and “Sector Regulation”: SCOTUS Picks the Right Poison.TRUTH ON THE MARKET » Chemerinksy’s Theory of the Roberts’ Court’s Antitrust Jurisprudence on Evaluating Leegin.Max Huffman on How to Survive A Motion to Dismiss After Twombly.Michael R. Bernstein on Backdating stock options is a crime? Go figure..Elizabeth Nowicki on Backdating stock options is a crime? Go figure..Jim Lebeau on Backdating stock options is a crime? Go figure..David Zaring on Backdating stock options is a crime? Go failure, right here on Junk Social Science in the Medical Bankruptcy Debate. […]

  3. TRUTH ON THE MARKET » Newsflash! AAI supports merger review! - July 10, 2007

    […] guess I am among the pundits who have already condemned this merger.  See here, here and here.  Thom also weighed in against the FTC with a great piece […]

  4. TRUTH ON THE MARKET » FTC’s Whole Foods complaint: still bulls**t - June 19, 2007

    […] Oats.Organic Baby Formula | Will Taft . com on NYT on Preserving the Purity of the Organic Club.TRUTH ON THE MARKET » More Thoughts on Whole Foods/Wild Oats on Premium natural and organic bu…TRUTH ON THE MARKET » PIPEs on FTC Rethinks Bad Pun?.Diana Bixler on FTC Rethinks Bad […]

  5. TRUTH ON THE MARKET » More Thoughts on Whole Foods/Wild Oats - June 14, 2007

    […] ON THE MARKET » PIPEs on FTC Rethinks Bad Pun?.Diana Bixler on FTC Rethinks Bad Pun?.~Efficient Happiness on Premium natural and organic bulls**t.Non-Sequiturs: 06.11.07 | Fair or Unfair on FTC Rethinks Bad Pun?.TRUTH ON THE MARKET » FTC […]

  6. ~Efficient Happiness - June 12, 2007

    Organic Economics: Regulatory Environmentalism or Market Loving Hippies?…

    Lord Melchett points out for example that the artificial fertiliser used in conventional farming is made using natural gas, which is “completely unsustainable”….

  7. TRUTH ON THE MARKET » FTC Rethinks Bad Pun? - June 11, 2007

    […] on New antitrust blawg.tim zinnecker on Coffeemaker Found. Time to Move On..Thom on Premium natural and organic bulls**t.Geoffrey Manne on Premium natural and organic bulls**t.Alex on Premium natural and organic […]