[nemild]

If useful, I wrote my own thoughts on ethics in software, after reflecting on certain experiences over the years:

> A serial tech entrepreneur in Silicon Valley once asked me to design a “social stockade” for his financial services customers. It would lock people out of their social media accounts and tweet out/FB share to their friends when they hadn’t paid a loan. He pitched it to prospective employees as meaningful work that would reduce the cost of loans for the needy.

> I was horrified that his product was being built and that many others would likely take the role I was turning down. And he was hardly the first to pitch his “innovation” as providing only good.

https://www.nemil.com/musings/software-engineers-and-ethics....

If anyone ever wants to discuss something, feel free to reach out (see HN profile).

[nostrademons]

Interestingly, many microfinance programs that are widely heralded as having lifted many people out of poverty (eg. Grameen Bank, CARE) rely heavily on peer pressure to boost their repayment rates. They lend out to groups within a village, and then if any one member of the group fails to repay the loan, the group can't access more capital. This creates a strong incentive for other members of the group to exert social pressure to make sure everyone pays back their loan.

...which goes to illustrate the complexity of most ethical issues that arise out of social systems. Oftentimes, something that seems cruel to an individual within the system is actually in the best interests of the participants of the system as a whole, and sometimes can even be in the long-term best interest of the person themselves. And then whether you view such features as cruel & unethical or necessary & beneficial depends on your perspective & role within the system.

(This could also be taken as a synecdoche for capitalism itself, which on a micro level is about as cruel as you can get - individuals compete in a race to the bottom to do things more cheaply, and nobody will help you unless it serves their interests too - but on a macro level is the most effective system we know of for satisfying consumer wants.)

[jclos]

> They lend out to groups within a village, and then if any one member of the group fails to repay the loan, the group can't access more capital.

That reminds me of those loan systems where you need to send naked pictures of yourself as a collateral, and that seemed to become popular in China (at least according to Western media).

https://qz.com/707770/chinas-college-students-are-using-nude...

[nemild]

Absolutely, and I say that having worked in microfinance before receiving this request.

But to me, that doesn't mean we engineers can't still draw a line somewhere, especially if we are called on to participate. Just because peer pressure works in some context, it doesn't mean that it is always the right choice, and we need to debate the tradeoffs in different contexts (much like engineers debate tradeoffs in any technical decision).

For example, the easiest way to cut the price of loans down would be to kill anyone if they didn't pay; defaults — and loan costs — would fall dramatically. This would immediately provide loans to many people who are priced out. While that may be useful in some scenarios, it's not a system I personally believe in.

[justin66]

> For example, the easiest way to cut the price of loans down would be to kill anyone if they didn't pay; defaults — and loan costs — would fall dramatically.

You state that as a hypothetical, but it's not like this has never been tried. The loans handled by lenders who include the threat of violence in the repayment plan are absolutely not characterized by low costs.

[nemild]

But loans for these activities (such as a loanshark) have their own set of risks that have to be factored in and affect the interest rate:

- You may have no legal recourse and no collateral to seize

- The loan may be funding risky or illegal activity with a high likelihood of failure, which demands a higher interest rate

- There may be no competition that drives the price down

Ceteris paribus, increasing the cost of non-payment should reduce interest rates. If you relax the "ceteris paribus", then all bets are off.

Another way to see this is this question: if the lender had to forsake the threat of violence, would the loan price go up or down?

[justin66]

Introducing the threat of violence isn't smoothly adjusting a variable in a formula. It's introducing a gating factor that's going to keep not-desperate people from dealing with you.

[nemild]

I'm happy to discuss with you offline (see my profile). The point I'm trying to make is that increasing the ability for greater enforcement mechanisms, should — on average — reduce the cost of loans. As I point out, there are real debates about where to draw the line about what is appropriate lender enforcement that I've personally struggled with.

I apologize that my example isn't perfect, and you're absolutely right, there is selection bias, unless there is little recourse for other products.

[fvdessen]

I know of a microfinance company that hires former convicts to reclaim payments and receives government subsidies for it as part of a social reinsertion program.