5 ms·
let's say you want to permit gambling, because people should be free to gamble. Even in the case of permitting it, there are still legitimate reasons to have re
by zemo 6y ago
let's say you want to permit gambling, because people should be free to gamble. Even in the case of permitting it, there are still legitimate reasons to have regulatory systems. For example, if you go to a casino to play craps, you as a consumer have an expectation that the casino is giving you fair dice. Now we could say that such a thing shouldn't be regulated, let the market decide or whatever, but now every time you go to a craps table you have to test the fairness of the dice yourself. So every time someone new comes to the table, they're testing the dice. It would be ridiculous. Even if you ran an entirely fair organization, everyone would be coming to the table and testing the dice because they don't trust the ecosystem, and your ability to have a fun and care-free gameplay experience becomes hampered by bad actors creating an ecosystem of poisoned consumer expectations; it would not only harm players, but it would also harm proprietors that wish to host games with fair dice. Or you can just say "you know, maybe casinos should not have the freedom to give people unfair dice".
- freedomben 6y agoAbsent government regulation doesn't necessarily lead to the scenario you described of everyone having to test the dice all the time. Third party certifiers could (and many believe would) emerge as trusted testers. For example, if you own a casino you can request certification from the "Fair Dice Association." They can then certify that your dice are fair. I as a consumer know they're reputation, and I trust their due diligence. This also has the benefit of utilizing free-market competition to drive down costs and ensure quality. It also avoids granting an authorization of violence, which some people believe is wrong whether it's a "government" or not. This isn't purely theoretical. It already happens. SOC 2 certification from the AICPA is a great example.
- bioipbiop 6y agoThe problem here is that somebody has to pay the Fair Dice Association and generally it ends up being the Casino. It also generally pays better to be a dealer than a dice certifier, so certifiers are often weighing the possibility of future employment when inspecting the dice. Finally, the risk of repetitional harm has proven, in practice, ineffective at preventing malpractice. This isn’t purely theoretical. It has happened repeated in recent history. Arthur Andersen signed off on obvious accounting fraud at Enron due to conflicts of interest. During the GFC, credit ratings companies graded low quality MBS derived securities as ‘triple A’ without bothering perform anything beyond a cursory examination. Now imagine what would happen if every rando could start playing in the private markets. Well you don’t have to imagine too hard, we did a natural experiment not long ago with crypto currencies. What did we get? Shameless pump and dump schemes organised openly on social media, pyramid schemes, vapourware ICOs, and all manner of other dodgy dealings. One reason these scams were so hard to stop was that so many unsophisticated scammers and scammees people were involved. The wealth requirement at least reduces the numbers of people involved to a somewhat manageable level. I do agree though, it’d be nice to have a more equitable way to do this. I think this action from the SEC is an important step towards that as it now opens the doors to qualified people that might otherwise fail the wealth test.
- __s 6y agoThey aren't arguing against regulation They're arguing that you shouldn't have fineprint that says "if you only let millionaires sit at your table, you can use loaded dice"