How technological innovation can massively reduce the cost of living

At the Y Combinator blog, Sam Altman announces that they are hiring a researcher to study a basic income . I am all for more research, but skeptical about the basic income as a policy solution. It has great properties on the recipient side—no marginal work disincentive!—but the sheer expense of it means that it is terrible on the funding side. The deadweight loss of taxation, remember, is quadratic in the tax rate.
More interestingly, in the Hacker News discussion , Altman asks a different question: “What can startups do to increase prosperity for everyone? I think basic income is important to do but decreasing the cost of living is a critical component as well.”
This is something that I have been thinking about for some time, and it was reinforced this week when Kareem Abdul-Jabbar came to Mercatus for a Conversation with Tyler . Kareem has an interesting perspective on poverty and opportunity: handouts aren’t the answer, he says, but the solutions posed by the political right haven’t worked and also aren’t the answer.
As Altman suggests, using technology to reduce poverty is an appealing solution that hasn’t really been tried. Technology can increase human welfare in two ways. First, it can create whole new goods and services, like the Internet, or smartphones, or Bitcoin. Second, it can make existing goods and services cheaper.
While the creation of whole new goods and services is extremely important, making existing goods and services cheaper has an advantage: we can know exactly where to look to make the most headway. The Bureau of Labor Statistics’s Consumer Expenditure Survey provides an annual snapshot of how Americans spend their money. More than half of all consumer spending occurs in three categories: housing, transportation, and food. If we want to dramatically reduce the cost of living, we have to make headway on these three categories.

Housing

Housing represents the largest share of consumer expenditures in BLS data. The BLS breaks it down into multiple categories, including shelter, utilities, household operations, supplies, and furnishings. Of these, no surprise, shelter is the biggest expense, representing 19.6 percent of consumer expenditures in 2014. Among the lowest income quintile, the expense share is even higher: 24.8 percent.
Part of the cost of shelter is the land, and part is the cost of the building. Putting aside, at least for purposes of this article, fanciful schemes to create more land, the amount of land available is approximately fixed. This means that there are three ways to reduce the cost of shelter: 1) build more densely (i.e., share expensive land between more households), 2) reduce the cost of construction, and 3) make currently cheap and underutilized land more accessible and useful.
The first, building more densely, is certainly worthwhile, but it is more of a political problem than a technological one. Cities should reduce zoning and height restrictions and eliminate ordinances, such as parking minimums that discourage density. Replacing existing municipal taxes with a land value tax would be helpful in encouraging better and denser use of land.
One not-entirely-political way to increase density might be to experiment with new housing modalities. For instance, what if we had urban dorms for 18–25 year olds who were not currently enrolled in college? That might be cheaper and denser than expecting new high school graduates to get their own full apartment.
The other two approaches to reducing the cost of housing lend themselves more to technological innovation. There has been substantial interest in pre-fab construction in recent years. The technology seems to be able to reduce construction costs by around 20–30 percent. And because pre-fab construction takes less time than traditional construction, it reduces some of the inconveniences that fuel NIMBYism .
Construction is also an area of interest for 3D printer aficionados, and increasingly, real developers. One developer in China used a single printer to produce 10 houses in a single day . 3D printing is significantly faster and cheaper than traditional construction. If the technique ever becomes the predominant method of building houses, that will put quite a dent in consumer expenditures.
Finally, technological innovation can make cheaper land more usable. For example, improved telepresence (perhaps VR) technology could make it less necessary to commute, and therefore allow more people to live away from cities. Autonomous vehicles could make long commutes more tolerable when they do have to occur, also making cheaper land more attractive.

Transportation

Speaking of autonomous vehicles, transportation is the next biggest bucket of consumer expenditures, amounting to 17.0 percent (15.0 percent in the lowest income quintile). The average household spends $3301 on vehicles (not counting leases), $2468 on gas, and $1112 on auto insurance per year. A broad “public and other transportation” accounts for $581 of spending per year.
Between Uber and Lyft, there has been a lot of innovation in transportation lately, but the big gains will come when more people give up owning a car entirely. Most cars spend 90+ percent of their lives sitting in a parking spot. If we could improve the convenience and cost of Uber-style services significantly, we could eliminate individual car ownership, a gigantic wasteful expense.
It’s no secret that Uber and Lyft (to say nothing of Google and Tesla) are investing in autonomous vehicles. One obvious way the future could play out would be for the Uber-style systems we have now to become fully autonomous. But an even cheaper system might be something like dynamically-routed autonomous buses. Think UberPool, but bigger and autonomous. We could have all the savings of autonomous driving with all of the benefits of public transportation with minimal infrastructure costs.

Food

Americans spend about 12.6 percent of their expenditures on food, and the lowest quintile spends around 15.5 percent. Most of that is “food at home,” as opposed to “food away from home.” (Interestingly, all income quintiles spend about 5 percent of their expenditures on food away from home).
What can be done to disrupt the cost of food? Of course Soylent is a famous startup producing relatively cheap and convenient meals. If you compare the cost of a bottle of Soylent to that of lunch in the workplace cafeteria, then Soylent has the potential to save people a lot of money—and the product continues to improve as the company iterates.
But as bullish as I am on the product personally (and I am), we won’t ever have a world where most people drink Soylent for most meals. If we want to reduce the cost of food for most people, we need to either reduce the cost of ingredients or the cost of preparation.
One promising approach for reducing the cost of food is to grow animal products without the animals. Modern Meadow is one startup experimenting with this, although they seem to be prioritizing leather over meat. By growing the parts of animals we want to consume without any of the parts that we don’t, there must be resource savings available.
Food preparation is something that can perhaps be automated through robotics. Most likely, this will happen quite incrementally—a general-purpose robot chef is probably many decades away. But what is the Roomba of cooking? Such a product would make it easier to shift some “food away from home” spending toward the generally more cost-effective “food at home” category, in addition to adding valuable time to peoples’ days.

What did I miss?

Reducing the cost of housing, transportation, and food is a surefire way to decrease the cost of living and to increase the welfare of all people, and especially the poor. I would love to see startups invest more resources in these areas. Almost certainly, some of them will succeed in ways that I have entirely overlooked, and I am eager to see what those are.
Eli Dourado is a research fellow at the Mercatus Center at George Mason University and director of its Technology Policy Program. Follow @elidourado on Twitter.