4 ms·
Amazon Is a Ripoff
- Finnucane 3y ago"There's a cheat-code in US antitrust law, one that's been increasingly used since the Reagan administration" AKA The Bork Doctrine, for the originator, Robert Bork.
- 1vuio0pswjnm7 3y ago"There's a cheat-code in US antitrust law, one that's been increasingly used since the Reagan administration, when the "consumer welfare" theory ("monopolies are fine, so long as the lower prices") shoved aside the long-established idea that antitrust law existed to prevent monopolies from forming at all." Antitrust law governs anti-competitive conduct, not the formation of monopolies (cf. "trusts", as they were known at the time the Sherman Act was passed). There is is nothing illegal about forming a monopoly. The question is whether the defendant used their monopoly position to engage in anti-competitive practices.
- camdenlock 3y agoThis is far too much nuance for a hack like Doctorow. He just wants to fume and stamp his feet.
- mvdtnz 3y agoI am by no means an expert on US anti-trust law but the first sentence of the Wikipedia page[0] on the topic says > In the United States, antitrust law is a collection of mostly federal laws that regulate the conduct and organization of businesses to promote competition and prevent unjustified monopolies. The first paragraph finishes with > Section 2 of the Sherman Act prohibits monopolization. And in fact the word "monopoly" (or its derivatives) appear 76 times, so it seems to be at least an important component of the law. [0] https://en.wikipedia.org/wiki/United_States_antitrust_law https://en.wikipedia.org/wiki/United_States_antitrust_law
- aidenn0 3y ago> Antitrust law governs anti-competitive conduct, not the formation of monopolies (cf. "trusts", as they were known at the time the Sherman Act was passed). There is is nothing illegal about forming a monopoly. The question is whether the defendant used their monopoly position to engage in anti-competitive practices. It's more complicated than that; "Innocent monopolies" (i.e. those that occur and are maintained without any anticompetitive practices) are allowed, but the list of things you can't do to form or maintain them (particularly after the Clayton act) makes it rather hard. For example, Mergers and Acquisitions that are likely to "substantially reduce market competition" are not allowed, so forming a monopoly by buying out your competitors would seem to violate the intent of the law.
- 1vuio0pswjnm7 3y ago"Holding a dominant position or a monopoly in a market is often not illegal in itself; however, certain categories of behavior can be considered abusive and therefore incur legal sanctions when business is dominant." https://en.wikipedia.org/wiki/Monopoly https://en.wikipedia.org/wiki/Monopoly
- Gunax 3y agoI am generally skeptical towards anti trust laws--though i do recognize that they are necessary. They give the state too much power to decide who is a market winner. Too often the definition of anti-competitive comes down to political favoratism. I would caution the author to consider the implications of what he seems to be asking for. Most businesses run at a loss at first. You really cannot start a business--nevermind an unproven startup--if you're required to charge at a profitable level initially. So I think the argument about Uber is accurate. Clearly the investors do plan on re-couping their losses. Were they planning on doing it anti-competitively? Maybe. The line between 'taking losses initially to get started' and 'taking losses so we can clear the field of competition and then reign as a monopoly' is dubious and difficult to quantify. Most of the tactics mentioned re: Amazon seem like regular retail tactics (staples in the back, high-margin items prominently displayed). All retailers control the price of items (usually they don't want a reputation for being expensive, a few don't want a reputation for being cheap). And 40% profit margins are not unusual. And consider even after this, Amazon basically breaks even in it's retail.
- 1letterunixname 3y agoSpeaking of scraping: One fundamental problem is the lack of standardized presentation of discoverable semantic web data cough RDF cough ontologies cough. If it's standardized and provided, then it's less of a burden to the scrapee and more useful for others. For example, finding the precise version history of open source software inevitably requires scraping a web page, download hosting server, or some source control. This is costly for the servers hosting it and fragile. Circling back to the case of price data. When an item is not in-stock, merchants have a habit of displaying disingenuous prices that are very high or low. Another dark pattern of gamification: displaying deceptive prices based on User-Agent, such that crawlers see one price while potential buyers see another.