Ed Lazowska — Technology and Inequality

(Image from https://medium.com/@TechforLifeUK/pay-inequality-in-tech-how-close-are-we-to-closing-the-gap-e74c00309da9)

LC: I wanted to get into one more big concern with the advancement of technology, which is that it causes and sustains wealth inequality. At some global forum in November of 2017 Bill Gates said that technology has a massive ability to accentuate the divide between the rich and the poor if we let it, and I was sort of shocked—hasn’t it done that already? How big is that man’s bank account? So what do you think is the key? What is tech’s role in wealth disparity and how do we tackle this cycle?

EL: I’ll say a couple of completely obvious things. One is that technology can also moderate these differences. For example, online education makes top level education available to a lot of people for whom it was previously inaccessible. The bad news is it doesn’t help much in this country. What it really does is maintain the educational disparities and level the playing field globally because most people who aren’t well educated in the US, it’s not due to geographical accessibility of it. Whereas people in the developing world may not have access to education, and this gives them access. Nonetheless, it does do that. We have a very vibrant effort here in technology for development, that is technology for the developing world. Whether it’s low cost community cellular base stations or financial tools for the developing world or medical tools or things like that, there’s a lot of ways in which technology can help. 

It’s also worth noting that technology is certainly not the only thing to blame for homelessness in this city and county; it has played a role certainly but probably not the predominant role, but the question is what technology can do to alleviate homelessness. Can it give us better data on why people enter homelessness and what interventions are best designing to get them out?

I was at a meeting earlier this week with the city and the county about what tech can do to help address the homelessness issue here. We have a whole bunch of busted public policies, and we need a boogieman to blame it on, so now we’ll just blame amazon for that. The truth is it’s us, and more importantly these people are us. To go down this route for a minute, it turns out that only seventeen percent of the homeless in King County essentially arrived in King County homeless. Eighty-three percent had housing in King County before becoming homeless. It turns out that thirty-four percent of eighteen-year-old kids who exit foster care become homeless within twelve months because there’s no support system after you’re out of foster care. These are problems we can fix. We got to dig in and help fix them, because we are going to get blamed for them.

We’ve done a bunch of studies with the Gates foundation analyzing Pierce, Snohomish, and King County homelessness data trying to determine which interventions and which sequences of interventions are most likely to lead to permanent, non-subsidized housing as an output, but it seems to me that there are a set of social issues related to education and other aspects that disproportionately afflict economically disadvantaged people that are really the source of the problem. It’s a set of people who are progressively disadvantaged in this society and that’s increasing. It’s not the people at the top making vast amounts of money who are causing that to happen—it’s our elimination of the social safety net that this country used to have and the ever-increasing pace of change. In the 1800s something like ninety-five percent of employees in America worked on the farm and now it’s less than two percent. That change occurred over a long period of time, so those individuals and their successors were able to educate themselves for other jobs. Now the change occurs so much faster, and the social safety nets are no longer there. 

In countries that have government supported universal health care, which we for some reasons don’t have and can’t manage to vote for, there is a social safety net that prevents people from entering poverty. But back to this homelessness thing: major causes of homelessness are job loss or medical emergency, and these are not things for which tech is responsible, for which tech necessarily has a cure. The price of housing is one of probably a dozen things that factor into the homelessness equation, and so I think we’ve got to solve those societal problems. 

It sounds like you just made a pretty good argument for universal basic income. One of the positive of technology, in my eyes, is that it allows fewer people to have to work for the same tasks to be achieved. Automation is just a more efficient way of getting things done—the only drawback is that it makes people lose their jobs. Why can’t we just pass the profits of automation onto these people who our society no longer needs to perform tasks now being automated via Universal Basic Income? 

Well, steering clear of… okay, you can’t steer clear of universal basic income. Yeah, I don’t know what’s bad about that. You know, people hopefully will be employed fewer hours in more interesting lines of work but they have to be prepared for that, and the problem right now is we’re not preparing people for that. An interesting statistic from a few years ago from a guy at MIT named David Autor is that the introduction of the automated teller machine caused the number of tellers to increase, and the reason is bank branches proliferated because opening them became vastly cheaper. So there were fewer tellers per branch, and they were doing presumably more interesting things like not counting dirty bills. Presumably, these people have more interesting jobs, and, in the aggregate, there were more of them. There’s been a turnaround in recent years in the number of bank branches, I don’t know why that is. But again that points out the counterintuitive nature of this. 

It seems like at some point all of the jobs that can be performed by a machine will be performed by a machine. Do you think there’s a way to assimilate everyone into these more complicated, non-automatable jobs, or maybe there’s some reason there aren’t as many of these jobs as there are people who can fill them?

That’s a very hard question that I’m not equipped to answer, but my feeling is the goal has to be to educate people so they can make productive contributions and feel good about what they’re doing. What I don’t believe is that a majority of under-employed people want to be underemployed. I think the challenge is for people to recognize what you clearly recognize is that those of us at the top of the pyramid were born at third base for the most part. We have an obligation to assist people who were not born on third base to get around the diamond. That’s what so many people have lost. 

So I think this notion that we have a set of advantages and many of them weren’t even visible to us that they were advantages. And it doesn’t mean we haven’t worked hard or that we weren’t smart. It means we had a set of advantages that we benefited from. UW has for example this STARS program, and these are economically disadvantaged and minority students who are given a fifth year at UW. The first year is a set of on-ramp courses because they just didn’t go to the high schools that you and I went to and they don’t have the learning skills and the study skills or the technical background, but many of them are every bit as smart as you and I. There’s nothing really different about these people except the circumstances in which they were born.  

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s