Silicon Valley and Basic Income

Last week, Y Combinator announced an intent to fund a basic income study. As my name turns up when you Google "basic income," I gave a few interviews on the subject, but did not have the time to lay out my take on the blog. So I do that here.

Who Owns The Robots

The recent Silicon Valley focus on the basic income seems a bit misplaced. As I understand it, tech people are reasonably concerned that increased robotization could eventually create a world where tons of working-age adults simply cannot find work. So they think a basic income could fix the capitalist thing where you starve if you can't find work.

The reason I think the focus is misplaced is that the bigger issue in the robot world is going to be who owns the robots. In a world with tons of robots and few workers, most of the nation's factor income is going to flow to the owners of the robots. You could try to tax the robot capital income to fund the UBI, but that risks capital flight in a world where capital flight could be extremely devastating.

So, in an automated world where robot ownership is as concentrated as current capital ownership, you risk extreme inequality and destitution if you don't engage in massive transfer programs but you risk mass capital flight if you do. The only feasible way out of this conundrum is collective ownership of robot capital. How to get such collective ownership is a more pressing question that Silicon Valley types should focus their attention on, rather than a UBI.

The Problem With Private UBI Studies

When it comes to the UBI study itself, I must register skepticism in how effective a private study can be.

First, the resources necessary to do this sort of study on a large enough scale to be worth it are immense. I don't know what kind of money Y Combinator has, but I am skeptical that it's enough.

Second, without coordination with the existing welfare state, private UBI studies could be incredibly messy. How would it interact with Medicaid and Food Stamps and Disability Insurance? If it counted toward income, you could cause people to lose some of these benefits, which would be very bad in some cases. If it didn't count as income, then people could get both the UBI and other welfare benefits, which could muddle any conclusions you draw from it.

These two issues are solved by having an actual government run UBI trials. Actual governments have the resources and have the ability to coordinate it with the existing welfare state in order to properly test what they are trying to test. Obviously Y Combinator cannot snap its fingers and make such a thing happen in the US. But Finland has a test on the horizon that looks very promising in this regard (even more so than some of the stuff that's gone on in the Netherlands).

Do A Time Use Survey

When it comes to what to measure in a UBI study, I think comprehensive time use is very important. There are 24 hours in a day. People use those hours in a specific way right now. And it would be interesting to see, in a very granular way, how the use of those hours change under the UBI.

Do people sleep more? Do they spend more time in school? Do they work less? Do they spend more time with their family?

Because people are concerned that a UBI could create a problematic decline in the labor supply, any study on a UBI would naturally try to determine what the labor supply effects would be. But if you don't couple an analysis of the labor supply effects with a comprehensive time use survey, you have no idea where those lost work hours (if any) went. They just appear to go into a bad black hole when in fact they might be used for something very constructive and worth sacrificing some work hours for, such as spending time with friends and family or getting education.

So, if I had to give advice on how to do a UBI study, I'd say include a very comprehensive time use survey if possible.