vi21maobk9vp comments on Does Hyperbolic Discounting Really Exist? - Less Wrong

19 Post author: gwern 03 December 2011 03:07AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (12)

You are viewing a single comment's thread.

Comment author: vi21maobk9vp 03 December 2011 08:55:57AM 0 points [-]

Preface: I am explaining why this argument doesn't work that well; it doesn't say anything about HD existing.

Well, there are some shady short-term high-interest loan operations in many cases. And it has to be shady, which limits the amount pumped.

Imagine you want to open hyperbolic-discounting operation for money pumping. What can you do? For it to be profitable you have to offer people to exchange little money now for more money later. Looks like banking, so you need to be in the banking business - or be a shady operation. Some part of what banks do fits the bill here, but they can't overdo it without huge and obvious colusion between all players: because short-term interbank loans are relatively cheap, if you try to profit too much there is a huge incentive to undercut you.

Also, with short-term loans (in money or any other form) it simply looks too suspicious...

Comment author: Unnamed 03 December 2011 06:54:08PM 1 point [-]

There are a lot of short-term high-interest loans, which take the form of credit card debt, payday loans, refund anticipation loans, installment plans, and so on. Refund anticipation loans are probably the closest to pure temporal discounting - that's where a tax preparer like H&R Block offers a customer who is due a $X refund from the government a choice between getting that $X later or getting $Y right now instead (in the form of a loan for $Y, which will be repaid by handing over the $X tax refund once it comes in). These examples at least suggest a high discount rate, although they don't definitively point to a hyperbolic shape.

There are a few reasons why these kinds of high-interest loans aren't more common. One is that it's hard to sell a short-term loan to someone with money in the bank, so they mostly target poor people. A second is the point that vi21maobk9vp makes: there's a somewhat competitive market so businesses need to compete by lowering their interest rate (a person who is willing to borrow at a 100% rate will still choose the loan with a 20% rate if they have that option). Also, governments try to put a stop to lending that they see as predatory; for instance, there has been a crackdown on refund anticipation loans over the past few years.

These examples only show high discount rates, not the preference reversal that is characteristic of hyperbolic discounting. But a business model that relies on getting your customers to change their minds (and explicitly reverse their prior decision) seems hard to pull off, and the profit comes from the customers' high-discounting decisions so the incentives aren't really there to try to induce a visible preference reversal.

Comment author: Khoth 03 December 2011 08:08:17PM 3 points [-]

Subscriptions to services that are cheap for an initial period then get expensive might be a case of preference reversal. People often sign up for them intending to re-evaluate after the introductory period, but when the time comes start procrastinating about it.