Latent problems with Universal Basic Income
Universal Basic Income (UBI), in a nutshell, is the idea that everyone should receive a guaranteed income—and in the case of "technological unemployment," is critical to offset the awesome and earth-shattering crater that artificial intelligence and job automation will likely create in the upcoming decades.
If you aren’t already familiar with the concept of UBI, then you’re already behind. Similarly, if you don’t understand the looming threat to labor from artificial intelligence, then you are even more behind. Where I live, these concepts are being hotly debated, because startups want to innovate in the creation of AI technologies. The inverse, which is ensuring that people are taken care of during this intense disruption, is also appealing since it is a source for revenue generation via the extraction of value from an individual’s UBI.
Will AI replace human labor in a variety of markets? Absolutely. We are watching this happen through the disruption caused in the transportation sector, which is working around the clock to replace human drivers altogether (and as someone who has ridden with dangerous ride-sharing drivers lately, who can blame them?) Nay, just transportation being disrupted by implementing autonomous driving systems is going to change the entire game, and this is happening right now—it’s not fictional, or far off. Again, living in the Valley, this is rather obvious as we shuttle about in self-driving Teslas—which by the way—are amazing.
The primary argument against UBI is simply that we can’t pay for it. As we saw during the presidential campaign of Bernie Sanders, many of his opponents and naysayers would really twist the proverbial knife and work to extract a meaningful plan to pay for grandiose ideas like free education and/or healthcare. Various economists working on the numerical tradeoffs came to the conclusion that this is extremely difficult from a financial viability standpoint, since many important social programs must be eliminated to even dream of having the budget—and as for demanding the monies from Wall Street—that sounds nice in theory, but you do realize that this is one of the most heavily protected groups on the planet, right? The legal fees alone from attacking this white-shoe beehive could sink the budget.
The point here is again one of where the money actually comes from. In the case of UBI, paying everyone a salary of just $10K/year (which equates to ~$3 trillion/year) is enough to cause serious palpitations in the American economy. It’s incredibly difficult to support the entirety of the American populous, let alone while leaving various social programs untouched. Once we consider larger salary ranges per individual, it becomes even less realistic.
With the primary argument against UBI being a fairly simple macroeconomic one, what’s the response from proponents of UBI? Simple, AI will generate enough wealth and prosperity through efficiency, optimization, and new services that we will easily be able to account for the difference in potential salary discrepancies, and people can finally live a life of leisure and pursue their creative dreams! What an ad for utopia. . . .
We know for a fact that human behavior doesn’t work this way. I wrote a research paper that focused on the behavioral economics of choice pertaining to food consumption, and again, in a nutshell—people regularly make poor decisions about the type of food they eat even when presented with healthier alternatives. What brought me into this research was the overwhelming number of children developing diabetes, and the unbelievable statistics related to the current obesity epidemic. Most people in the country aren’t eating a healthy balanced diet, which is a system-level problem with various moving parts, some including access to food, education around diet, overriding the satiating nature of processed foods relative to their less delicious (but often healthier) options, and parents feeding their children diets that are primarily composed of processed foods—which a growing body simply can’t sustain (insofar as not developing diabetes, that is). I was fascinated with the fact that people’s better judgment can be manipulated to such a degree that one could argue that they simply don’t have a choice, even if they are presented with multiple options.
One of the facets of behavioral economics that resonates with me so deeply is the idea of choice, and when and where choice is completely overwhelmed. Marketers and large corporations constantly use research in this domain to provide people with an overall experience that is so well designed that they aren’t in control of their own choices anymore. This doesn’t even begin to touch on the federal or state regulations that help fight back against research backed problems in this domain, but with any new shock to the system, we must consider all the players: government, corporate, and the individual.
I regularly hear people here discussing UBI and how wonderful it would be not to have to work, and to finally achieve autonomy—to be able to spend the day pursuing your own artistic endeavors—which could mean playing video games all day, creating a startup, doing drugs until you drop, or traveling. (Individual preferences will vary.) We know that when people are given money, whether through federal programs or holiday bonuses, that it’s hard to predict how it will be used. Some argue that it will stimulate the economy through consumer spending, but never before have we tested a concept as revolutionary as UBI—in terms of wide spread salaries for everyone.
A local startup incubator, Y Combinator, is currently testing this theory in Oakland. They are giving 100 residents $1,000-2,000 dollars per month for six months to track how it is spent. There are a variety of problems with this methodology, including: small sample size, it’s not geographically (or demographically) representative, and worst of all, participants know their purchases are being tracked. This is completely different than doing a study across the nation with random candidates who have their habits monitored via a secure one-off credit card—or some less obvious system for tracking their spending habits. I do appreciate their testing the theory at all, but these results aren’t necessarily generalizable or repeatable.
Let’s assume for the sake of brevity, that AI generates an abundance of wealth that can be distributed across the nation to support UBI in practice.
I won’t dive into the obvious problem of who controls the AI and how they decide to distribute corporate wealth from the private to the public sector (something we know many large corporations historically fight tooth-and-nail against). Let’s assume that Alphabet, Facebook, Baidu, Tesla, and Nvidia create new AI-based products that replace a large subset of American jobs, yet lower the cost of labor to such a degree that a large chunk of said profits will be taxed with a new American tax that is passed to account for when companies choose to upshore (a term I’m going to use for replacing American workers with software/hardware solutions), which would estimate the amount of money that would have gone to a human worker and earmarks a portion of the profits for the federal government to use in their UBI package.
The basic supply and demand problem here is simple. If, let’s say, 20% of the American labor force is replaced by upshoring (AI), then you have 20% less of a local consumer market to buy your new highly efficient services. Upshoring will literally be destroying its own consumer market, since they won’t have jobs and therefore can’t buy these new products without UBI. This is where the plot gets even more sinister. Many of the most profitable AI solutions aren’t being developed for the consumer market, and are instead harnessed for applications like algorithmic trading (50-70% of all stocks are now traded by high-frequency trading algorithms). This brings our wonderful friends at Wall Street back into the picture, because whoever controls the AI controls the new world economy.
If by some stroke of luck, we are able to convince the companies with the most powerful AI programs to share a portion of their profits with the federal government for UBI, then we finally get to my initial problem—what will people ACTUALLY do with their time?
If everyone in America suddenly got a location-adjusted guaranteed salary, what then? This is where the research should be focused, because we know so much about how difficult it is to change behavior and build new habits. Will people hang out with friends more, travel more, and consume media with the majority of their time? So on and so forth. This is one of the single biggest unaddressed research questions relative to the UBI argument, and until we have a decent enough handle on it, we can’t make any big moves—though AI research is quickly progressing either way.