Amanda Ripley

Nav

← Back to Posts

Random Ignorance

9th Dec 2010 posted in General

The New York Times has a troubling story on the front page about an experiment on homeless people. But what’s troubling is not the experiment, from what I can tell; it’s the newspaper’s fear-based spin on the story.

So here’s the situation: New York City is conducting a test to see if an expensive program to prevent homelessness actually…prevents homelessness. This is radical and admirable—because it almost never happens. We waste billions of dollars in education, welfare and other spending in this country because we don’t test programs in a careful, rigorous way to see if they work.

I’ve written about another such experiment in Time (one designed to test whether paying students led to more learning in schools), and what I learned is that these studies reveal how very differently humans behave from what we often expect.

Randomized studies are integral to any ethical social program. Because the only way to really know if an intervention is working, as medical researchers have understood for quite some time, is to randomly assign a group of people who will get the intervention—and a group of people who will not.

But you have to do it carefully! You don’t want to do harm to the people who don’t get the service. But this is not new ground. We know how to do this. We do it every day in other fields. And in fact, in this case, it seems the experiment was designed according to standard, humane practices. The study will monitor 400 households that sought help from the city this summer because they were behind on rent and in danger of being evicted. Two hundred families received the services of a program called Homebase, which offers job training, counseling services and, in some cases, emergency money to help people stay put. The other 200 families were given the names of other agencies that also help people in need.

The research firm, Abt Associates, approved the study as ethical because Homebase’s services were not already available to everyone to begin with, due to limited funds—and because the control group still had access to other, alternative services.

But the important question is, which is more unethical? Denying one specific service to some families to see if a program works? Or continuing to spend—and more importantly cut—money without any meaningful test of whether it works? The city, as the article notes much, much later, had to cut $20 million from its Homeless Services budget last month. And federal stimulus money (which went towards the $23 million bill for Homebase) will end in July of 2012. So cuts must happen. Shouldn’t they happen ethically?