When I was a child, I remember when the migrant farm workers would arrive every summer to work in the nearby bean fields. Those fields stayed, but the migrant workers stopped coming by the late 1960s, which I now realize was about the time that minimum wages and the Fair Labor Standards Act started being applied to farm labor.
This brings me to an interesting story by Miriam Jordan in the New York Times about new technology that is being introduced to farms, especially for those growing labor-intensive crops that have to be picked by hand. One of the most ingenious examples of this is a machine for cutting lettuce, a crop that is quite delicate and easily crushed by wheels and torn by blades. It cuts the lettuce with jets of water instead. Another is a robotic weeder that uses cameras and software to differentiate the crops being raised from plants that need to be ripped out.
It’s easy to attribute the rise of these machines to the genius of some tech wizard in their labs. It certainly looks whiz-bang compared to hand picking and weeding. But in fact, this equipment is not the result of any fundamental breakthrough in technology. It is progress in engineering, though: applying existing technology to a new and different problem, such as using image recognition to identify plants.
Why is this coming now, and not sooner? Because it’s gotten more difficult to get agricultural laborers to do the unskilled work of picking and harvesting. These are not attractive jobs. The work is hard and is seasonal. And while minimum-wage requirements do apply to farm workers, there are lots of exemptions to wage-and-hour requirements. Overtime provisions do not apply, since farms that use less than 500 days of farm labor in a calendar quarter in the previous year are exempt. Most of the “hand labor” who pick crops can be paid piece rates instead.
Evidence suggests that the vast majority of agricultural labor has been–and still is–performed by undocumented workers, who don’t have many other options and are unlikely to report violations in the laws. It has gotten much more difficult to find those workers, while people with other options are not likely to be interested in this work.
It is not surprising that the agriculture industry has long resisted efforts to restrict immigration and crack down on illegal immigration. When that doesn’t work, necessity is the mother of invention and the market for solutions in the form of technology grows large enough that it is worth it for companies to figure out how to do the work with machines. These solutions are not perfect, but they are already pretty good, and they will get cheaper and better.
What’s the general lesson from this? It is a reminder that what drives technological change in the world of business has more to do with economics than anything else. We automate–or as economists would put it, we substitute capital for labor–when the need is big enough to draw investors and engineers into the problem. The reason the work of picking crops remained largely unchanged for generations has been because labor was cheap enough not to merit seeking out alternatives.
The arguments we hear from pundits about which jobs could be automated are irrelevant. We could automate almost any work if we put enough effort and investment into it. We just won’t do it unless it is worth it. If you go to the developing world where unskilled labor is incredibly cheap, you will see construction work being done with picks and shovels. It is not because they do not know about backhoes; it’s because labor is so cheap that it is cost-effective not to use them.
Here’s one last thing to keep in mind whenever we hear that the economy needs cheap labor: Substituting capital for labor–or automation if we want to use that term–is the main force driving increases in productivity, which, in turn, is the primary factor in making countries wealthy. Cheap labor reduces that incentive to automate.