Hal Singer is a Principal at Economists Incorporated and an Adjunct professor at Georgetown University’s McDonough School of Business
When the Federal Trade Commission (FTC) waved through Amazon’s purchase of Whole Foods without batting an eye last August, some New Brandeisians predicted doom. But I was a bit more sanguine. And that’s still the case on the one-year anniversary of the deal. Here’s why.
The mere act of vertical integration by a dominant platform such as Amazon isn’t by itself cause for alarm. I don’t begrudge Amazon making a run at local groceries. And unless New Brandeisism is adopted at the FTC—or Congress embraces its arguments for managing markets through structural separation—there’s nothing to stop Amazon, Facebook or Google from invading “verticals” that reside adjacent to (and often depend on) their platforms.
It’s vertical integration combined with discrimination in favor of the platform’s affiliated products (or content, in the case of Google) that is problematic in my mind. And until I get Congress to enact a nondiscrimination regime that operates outside of antitrust, there’s nothing to stop the tech giants from dominating large portions of the Internet economy through such anti-competitive conduct.
(There are other potential merger-related effects outside of discrimination that should be monitored, such as the added market power that comes from accumulating buyer information across online and brick-and-mortal activities, or the squeezing of supplier margins or workers’ compensation. But those effects are outside the scope of this short piece.)
Amazon’s users likely don’t perceive its platform to fill the same purpose for every product available on Amazon. While I can envision Amazon exploiting its dominant e-commerce platform to dominate the online sale of (say) batteries, online apparel and footware (think Zappos), and diapers (think Diapers.com), it’s harder to imagine Amazon doing the same for local groceries. An Amazon user who gets steered by Alexa (or Amazon’s search algorithm) to Amazon-branded batteries without putting up a fight seems plausible. But how many users would ask Alexa for a half pound of ground lean beef?
(Another brief digression: It would be very hard to make an antitrust claim that Amazon is leveraging its platform to monopolize the market for batteries. Presumably, Amazon has no plans to raise battery prices or restrict battery sales, but instead is content to earn the narrow margins in that vertical and perhaps learn how its customers plan to use their batteries. The bigger threat is a harm to innovation, which admittedly in the not-very-innovative battery vertical could be less meaningful than short-term battery price declines from a consumer-welfare perspective. My simple point is that batteries represent a vertical that is vulnerable to an Amazon takeover given the natural synergies between its platform and the purchase of batteries. Groceries, not so much.)
Perhaps I’m an old-fashioned Gen Xer, but I place local groceries in the same category—a bubble really—as yoga studios or coffee houses that (unlike Starbucks) serve lattes with foam art. Things inside this bubble managed to survive the (partially Amazon-driven) retail apocalypse. These “experiential” products and services are hard to sell over the Internet. Consumers enjoy the physical interaction associated with shopping for groceries in much the same way they enjoy the social engagement of a yoga studio or a coffee shop. The sale of “Holiday in Italy” cheese is much easier if the customer can sample the merchandise.
When it comes to non-experiential items, by contrast, Amazon can be trusted to deliver the commodity to your front door. A battery is a battery is a battery. But can Amazon, even in conjunction with a high-end, organic grocery label such as Whole Foods, be trusted to replace your judgments when it comes to meats, vegetables, fish or other items that require touch, feel, and smell?
Even for those daring souls who are willing to purchase groceries online, it might be hard to dislodge a customer of Peapod (virtually shopping at Giant) or Instacart (virtually shopping at Wegmans) or FreshDirect through some combination of leveraging the Amazon platform—say, by elevating Whole Foods to the top of AmazonFresh searches—and offering subsidized delivery of Whole Foods’ produce. Customers of rival online produce providers didn’t wind up there by accident; instead, they likely found their favorite online grocer through trial and error (and comparison with brick-and-mortar options), and have now developed an affinity. Moreover, many of those customers enjoy subsidized food delivery from their preferred vendors, largely negating the Amazon delivery subsidy. And since the Amazon-Whole Foods merger, many brick-and-mortar grocers such as Kroger, Target and Walmart have partnered or acquired online grocery companies or same-day delivery companies, indicating that they plan to fight hard for this slice of the grocery marketplace.
That said, Amazon is still the most popular place to shop among those inclined to purchase groceries online. According to a Morgan Stanley July 2018 survey, 56 percent of such respondents said they would most likely order from Amazon, compared with 14 percent who would go to a mass merchandiser and 10 percent who would use their local supermarket. It is this segment of the marketplace that is most vulnerable to a leveraging strategy.
A straightforward way to test whether Amazon’s platform is capable of steering grocery purchases towards Whole Foods is to compare Whole Food’s sales (or in-region market shares) before and after the merger. This test is admittedly crude for at least three reasons. First, it might take some time for Amazon to hammer out the kinks of integration—Whole Foods has just begun to centralize purchasing its supplies and merchandizing—and for those who are willing to shop online for groceries to become a sizable share of buyers in the grocery market. According to Forrester Research, just three percent of the world’s grocery spending happened online last year, though that share is expected to reach six percent by 2022. Second, a simple before-after comparison implicitly makes Whole Foods’ 2017 sales the benchmark, when in fact Whole Foods’ sales in 2018 may have fallen relative to 2017 but for some merger-induced advantage; at the time of the deal, Whole Foods was struggling, as same-store sales declined by 1.5 percent in 2017. Third, Amazon is not as forthcoming as Whole Foods was with information on grocery data such as same-store sales growth. (That Whole Foods’ sales increased post-merger could be consistent with a merger-induced synergy unrelated to discrimination. Further inquiry would be needed to rule out that alternative hypothesis.)
Even with these caveats, it’s still worth looking at the recent trends. Whole Foods’s sales since 2015 have been flat, with only low single-digit growth, according to data from Second Measure. This suggests Whole Foods is not yet getting a lift from the relationship. However, the percentage of Whole Foods’ new customers who are Prime Members increased post-merger, from 34 percent in June 2017 to 41 percent in June 2018. This suggests that Amazon’s platform is delivering customers to Whole Foods; to be fair, some portion of that lift was due to the growth of Prime membership generally.
Maybe I’m wrong, and Amazon will soon obliterate the local grocery vertical the same way that it overtook diapers and (soon) batteries. The future grocery store might not sell these homogenous things. But based on the latest available sales data from Whole Foods, there isn’t any reason to panic. Yet.