By D. James Greiner
“Harry suddenly realized that the tape measure, which was measuring between his nostrils, was doing this on its own. Mr. Ollivander was flitting around the shelves, taking down boxes. ‘That will do,’ he said, and the tape measure crumpled into a heap on the floor.” —Harry Potter and the Sorcerer’s Stone
I’m a statistician. My job is to measure things. But I don’t want to be like Mr. Ollivander’s tape measure. I don’t want to end up crumpled into a heap on the floor. I want the things I measure to be useful. Even more, I want the measurements I make to be used, which is a lot harder.
What does that mean?
It means that every time I think about measuring some thing, particularly if I’m trying to find out whether that thing “works,” the first thing I have to do is to ask, “What is this thing for?”. Without deciding first what the purpose of a thing is, there’s no way to know what I should measure to see if that thing is working.
Take legal services to low-income folks. Some lawyers and law students feel a strong desire to provide legal services to the poor. I’m all for it. But unless the only reason we’re providing the services is to make ourselves feel better, we ought to care about whether we’re accomplishing anything by providing the services. And we ought to care about whether there might be a better or more efficient way to accomplish whatever it is that we’re trying to do. That might mean less money for lawyers. But unless the point of legal services to the poor is full(er) employment for lawyers, that shouldn’t bother us too much.
Here’s an example: a few legal services providers have recently begun to provide services for defendants in small claims court debt collection cases. Incidentally, my guess is that this is a great idea. My speculation is that legal services providers have focused for too long on the “iron triangle” of government benefits, eviction defense, and family law.
So, great. A focus on debt collection defendants. Super. Now, I want to measure whether debt collection litigation defense works. So how do I do that?
I first have to ask what debt collection defense is for.
Suppose I decide: debt collection litigation defense is about preventing the defendant from having to pay the debt. After all, a dismissal of the lawsuit with prejudice is the best possible legal outcome that a litigator can achieve through litigation defense of a debt collection defendant. OK, great, I can measure that. I can randomize a group of debt collection defendants to either an offer of representation or to no assistance, look at court records, and compare judgments in the offer-of-representation to the no-offer group. Super! Off I go to measure (as soon as I can find a legal aid office to do the study with me).
Wait a minute.
If the purpose of providing litigation defense to debt collection defendants is to keep the defendant from having to pay the debt, there’s a far, far, far cheaper way to do that: Just buy the debt the plaintiff is suing on. Buy the debt on the open market, and then forgive it by telling the alleged debtor that she’ll never have to pay. You can probably do that for, say, five cents on the dollar. So you can buy (and then forgive) a $2,000 debt for $100. Is that more efficient than litigation defense? It almost has to be. When you think about what a litigator needs to function (computer, printer, IT support, office, transportation, salary, benefits, administrative support, etc.), it’s hard to think of how one can produce a functioning litigator for less than (this is a blind guess) $60,000 per year. And how many $2,000 debt collection cases can that lawyer realistically defend in a year? 300? 400? For $60,000, a legal aid office could buy and forgive 600 $2,000 debts. So, the best thing to do for the legal aid office is not to hire a lawyer; rather, the best thing to do is to buy debts on the open market and forgive them. You get twice the bang for the buck.
The part about just buying debts feels wrong somehow. More to come in the next post.