Yeah, I got that, but why when you can do it without pairing? This method would get exponentially more complex as the number of creditors increases, especially for a people who still had nothing but an abacus to work with.

He just explains the pairwise equal division of the contested sum as a social custom.

In some ways it makes sense if you want to maintain the best pairwise relationships in a small community.

If you did it without pairing, the person who is owed the least might resent the people who received more.

I guess it does sort of force a minimum on repayment. Those sorts of things tend to create much resentment. I'd be curious to see if this would be a better solution than the one we currently use.

I have seen the argument that this system is more fair to small creditors. Under the current bankruptcy, if I extend $1000 of credit to a financially shaky company, and then some bank extends a million-dollar credit line, and then the company becomes insolvent, we all have to line up for our X cents on the dollar; the bank's easy credit effectively reduces the amount I can recover. Under the system the OP described, small creditors are far more likely to get repaid in full or mostly in full, regardless of how much the debtor also owes to large creditors.

But I'm not sure how this system could be extended to a world where, for example, a company takes out loans from Bank A and Bank B, both of which are actually subsidiaries of a bank holding company that owns 75% of A and 30% of B. Talmudic law doesn't recognize corporate personhood, so the Talmudic discussion doesn't have to consider that problem.