Thread by @Caffar3Cristina

Platform regulation #DMA: “ink is almost dry” on Art 5/6 various “thou shall not’s” for “gatekeepers”. BUT rules need translating into relevant prescriptions for each business model. We focused on e-commerce last week for a great discussion @cepr_org #Competition_RPN/ @MaCCI 🧵1/

Video recording can be found here 2/

Aside from public narrative (universally adverse), a ton of economic research underway looking at the behavior of hybrid marketplaces, esp. Amazon. What does it show? Does anyone care? Should regulators/enforcers care & acknowledge the tradeoffs? Or is it all beside the point? 3/

Economists have trouble being heard now given view many have lent their tools to consolidation & creation of power, incl. in digital.  Still, there’s a growing body of good papers studying the conduct of a dominant marketplace & collecting evidence. What do we do with that? 4/

Papers look at welfare effects of hybrid marketplaces, self preferencing to favour own products or logistics, raising rivals’ costs through commissions on rivals, using seller data to enter or copy products & then self-preference one’s own…(see list of paper at end of 🧵) 5/

What does it all say? Theory is mixed. Marketplace enters w own products where it has competitive advantage, creates more short term competition/variety, self preferencing may be fine if consumers like the product, but various longer term harms. General call for more scrutiny 6/

Discussion of contrasting models: @BedreOzlem predicts “insidious steering” as hybrid marketplace raises commissions to sellers in product lines it competes in; Etro @uni_firenze finds marketplace may lower commissions instead to attract more sellers & consumers. “Depends”. 7/

@theplatformguy: 1. Hybrid marketplaces are wide-spread: e-commerce companies like Amazon, Walmart, Target, JD in China, Flipkart in India, game consoles w marketplace for 3P developers & selling own games, any SaaS company of a certain size opens up a marketplace.  8/

cont/ Small ones too. Why?Efficiency reasons, can be pro-competitive: company is good at providing some products (cost/quality advantages) but maybe 3P better at providing others. Not necessarily nefarious. Also one-stop shopping is good for consumers. 9/

@theplatformguy: 2. So a blanket ban of hybrid model makes no sense. Even w bad conduct like steering & imitation, better do behavioral fixes: monitor algos making APIs available to regulators & trusted researchers, audit recommendations; robust data firewalls against copycat 10/

Policy discussion: What are legislators and decision makers thinking of? Do any of these insights matter? 11/

@AlexandraGeese:Legislators are concerned abt speed w which large actors have acquired disproportionate influence & have access to data of sellers & consumers. DMA Art 5/6 are now settled on self-preferencing, use of data, & MFNs. Implementation nebulous, yes, needs resources 12/

@sarahmillerdc: 1. Tech monopolies were created in a regulatory vacuum, but we have strong US tradition of skepticism of large companies & that muscle memory has returned. We see workers organizing in warehouses,small businesses speaking up, impact on postal service, defence…13/

@sarahmillerdc: 2. New antitrust leadership thinks v differently. Issue is power, not just behavior, as any attempt to rein in abusive behavior at a large institution will fail bcs of its political power. “Fallibility of the regulatory state” is recent history of antitrust 14/

@sarahmillerdc: 3. We need to look at conflicts of interest & say we don't need that, there aren't any real efficiencies to be gained & there are downstream problems we haven't taken seriously before. Quite optimistic at level of sophistication of the discourse in the US 15/

@steve_tadelis: But externalities are v real for a platform, when a seller disappoints & loses buyers’ trust, they are more likely to leave. Platform needs to police sellers’ behaviour. Problem even more severe for hybrid marketplace than for pure platform like Etsy or eBay 16/

@steve_tadelis: 2. Same is true w logistics, if 3P delivery experience is poor it hurts the marketplace. If 3P seller has a more efficient distribution it can deliver effectively. Platform needs to generate & maintain trust w consumer, “Italian decision is problematic” 17/

@steve_tadelis: 3. “When we talk about power, I see it for Facebook. But to say FB and Amazon live in the same space & we could have regulation to adequately address both, don't see it - v different business models, cannot have a single coherent regulatory framework for both” 18/

@GregorySCrawfor : 1. We heard theory has identified salient trade-offs that are first-order relevant to regulation, some in the short run, but also short-run vs long-run. Yet the empirical literature has very little to say on this yet – data has not been there to test 18/

@GregorySCrawfor : 2. Literature on self preferencing is in its infancy, what's out there in last year shows short-run consumer benefits & possible long-run trade-offs. Literature on effects of entry is even thinner, old. Regulators need to request data & study this properly 19/

@GregorySCrawfor: 3. BUT shld we bother? Yes there's trade-offs we should look at if we want to understand downsides of forbidding perhaps efficient practices, but may lead to failure to enforce, endemic for last 20 years. OR we just go for simple rules & live w consequences? 20/

/cont. There’s a big trade-off here too:  loss function between digging in deep trying to understand, vs just okay we're not going to bother for these big companies, don’t care for efficiency loss. 21/

@GregorySCrawfor: 4. Maybe depends on business practice? Conduct more obviously concerning that can protect/extend market power (like MFNs/foreclosing strategies) needs bright line rules, other things (like commissions) that are not so obvious can be looked at in detail? 22/

Overall, fantastic exposure to today’s two competing frameworks: one rooted in economics & efficiency, but a question on the degree to which our fun economic toys can be reliable & administrable by institutions w v limited ability to understand what giant corporations do 23/

Another that says the politics of all that is acute, the economics/efficiency framework is no longer apt, we need clear rules even if blunt, and structural separations to take apart conflicts of interest. There will be efficiency loss but we are prepared to bear them. /24

More to say on others panelists’ contributions @ameliafletecon, Thorsten Kaeseberg @BWB_WETTBEWERB, Antonio Butta’ @AGCM but will stop here! 25/

PS some relevant literature, all mentioned at the event.
Anderson & Bedre-Defolie (2021), Hybrid Platform Model, papers.ssrn.com/sol3/papers.cf… 
Cabral (2021). Seller reputation and price gouging. onlinelibrary.wiley.com/doi/10.1111/ec…

Etro (2021). Product selection in online marketplaces, onlinelibrary.wiley.com/doi/full/10.11… 
Etro (2021). Hybrid marketplaces with free entry of sellers, disei.unifi.it/upload/sub/pub…  disei.unifi.it/upload/sub/pub…

Hagiu, Teh, Wright (2021) Should platforms be allowed to sell on their own marketplaces? wrighteconomics.com/post/should-am…
Hervas-Drane & Shelegia (2021). Retailer-led marketplaces tse-fr.eu/publications/r…

Lee & Musolff (2021). Entry in two-sided markets shaped by platform-guided search, lmusolff.github.io/papers/Entry_a… 
Masden, E., Vellodi, N. (2021). Insider imitation, mimeo, New York University. papers.ssrn.com/sol3/papers.cf…

Shopova (2021), Private labels in marketplaces papers.ssrn.com/sol3/papers.cf… 
Zennyo (2021) Platform encroachment and own-content bias papers.ssrn.com/sol3/papers.cf…

Btw, RPN condition is everyone doing consulting for any side must disclose, & everyone did. Check video.