A.I. and the Labor Force

A few weeks ago, not wanting to sacrifice class time, I offered a short history of corporations in the U.S. Now that we are reading Kate Crawford’s Atlas of AI, I complement that essay with one on labor.

The labor movement in the U.S. is an outgrowth of industrial capitalism. Free labor had long been unshackled from the constraints of feudalism and further separated from tradition by immigration (indentured servitude or free labor) or forced migration (people sold into slavery).

Most immigrants prior to the civil war engaged in agriculture. Those outside of farming worked as servants, in trades, in construction (think: buildings, canals, and early railroads), or as artisans. In the northeast, early factory systems emerged in textiles in the 1840s. Factories in Massachusetts, for example, were first populated by local farm girls supplementing their family’s finances. Later, immigrant families, mostly from Ireland, took those positions as working conditions worsened.

The rise of corporations after the Civil War launched the rise of the labor movement. It is important to remember that no government regulation existed. Contract law defined the relationship between workers and capitalists. Unequal bargaining power gave owners the power to determine conditions, wages and hours. The only means of counteracting exploitative circumstances were for workers to organize and use tactics such as strikes to protest. Early organizations such as the Knights of Labor formed in 1869. This union brought together both skilled and unskilled workers, who, for a time, were successful in advocating for reforms such as improved working conditions and the eight-hour day.

But as protests became more widespread and violent, capitalists used property, contract and criminal law to call in local or state police to quash rebellions, sometimes at the expense of workers’ lives. The failure of the American political and legal system to respond to workers’ needs fueled the rise of more extreme working and social movements such as the International Workers of the World (IWW, or Wobblies), as well as political organizations such as anarchists, socialists and communists. The Haymarket Riot in 1886 Chicago was a high-water mark for such conflicts. Government reacted to it by siding with capitalists and further suppressing the burgeoning labor movement.

In the early 20th century, American law remained recalcitrant in the face of even conservative labor organizations such as the skilled worker’s American Federation of Labor. In 1905, for example, the U.S. Supreme Court reinforced contract law to overrule state labor laws, such as the New York State law limiting to working hours for bakers to 10. In the 1908 Muller v. Oregon case, the court created a single exception to this rule: women. Gendered interpretations of the law grounded the decision. No such protection was offered to children until the New Deal.

For those interested in the history of Constitutional law, the 1937 “stitch in time that saved nine” represented a transitional moment. Threatened by Roosevelt’s “court packing plan,” the Court, which had been finding his New Deal legislation unconstitutional on due process grounds, finally backed down and allowed those laws to go forward. The 1935 National Labor Relations Act was one of them. Importantly, it set wage, hour, child and conditions laws. It recognized labor organizations. It also created the Department of Labor and the National Labor Relations Board to arbitrate labor disputes. While over ensuing decades organized labor became co-opted by corporate influence and corrupted through organized crime, the Labor Department remains the most powerful entity in American society to work with unions and to represent the interests of wage-earning people.

For understandable reasons, Crawford does not go into this history. First, she is taking a global view of labor in keeping with her atlas theme. Second, her concerns transcend the particular focus that labor protection has taken in the U.S. At the risk of stating the obvious, capitalism under a democratic republic won. Extreme left labor efforts intertwined with socialism or communism lost.

Franklin Delano Roosevelt can largely be thanked for realizing the ideal of his cousin Theodore Roosevelt who inveighed against the “malefactors on the right, and the malefactors on the left.” Third, Crawford’s atlas theme aligns with her chapter on mining. Finally, she is excellent on the subject scientific management, or Taylorism, that increasingly defined the factory experience for workers over the 20th century and unto today.

Remember Amy Zegart’s quote, “tech policy is public policy”? On the question of A.I. and labor, we have a prime example. Disruption to the labor market is a common thread among observers of the social impact of A.I. These observers are correct to raise this issue, not merely because of the potential of A.I. to affect knowledge and white-collar workers especially, but because the United States has unsuccessfully confronted this kind of problem. Globalization, outsourcing and mechanization displaced a significant percentage of blue collar, industrial workings from the 1970s to 2000.

We are still living with its effects. Anger and resentment for having been “left behind” informs right-of-center politics generally and speaks to the popularity of an “American First” campaign. President Biden alluded to these issues in his State of the Union address the other night, recognizing that both major political parties in the U.S. failed to help workers displaced by these large-scale changes. As the most union-aligned president in U.S. history, he also is a strong supporter of tax breaks and other incentives for bringing manufacturing back to our shores. Meanwhile, A.I. observers looking to the near future have political outcomes in mind. One can only imagine how much more inflamed our politics are likely to become with the disruption of a college-educated (and often college-indebted) labor force.

What lies at the root of this intractable problem in American society? Contradictions are built into our economy and politics. As a free-labor society driven by the market and framed around a negative rights constitution, government is not expected to take care of workers. But just as we observed in our discussion on corporations, government nonetheless takes the hit because who else or what other sector in society is visible to even hear these concerns? Presidents in particular seem to have to shoulder popular opinion about responsibility for an economy over which they have little control. The one area of the administrative state that touches the economy directly, the Federal Reserve, is also severely limited. It sets interest rates. It does not create jobs or government work programs or provide for workers disenfranchised from employment.

What a conundrum. The “invisible hand” of the economy is an amorphous metaphor. American law incentivizes corporations to act on behalf of shareholders, not the population at large. Socially, we have become increasingly partisan in large part due to labor force disruption. Politically, we have become so divided as to cripple Congress. Just about every observer of A.I. who has rung the bell about a potential for significant disruption in the labor market cries out in the wilderness. We live in a society largely allergic to proactive government planning.

To whom else might observers effectively speak? What changes in our existing political economy could be enacted that would make a meaningful difference? That was the public policy note upon which we ended the last class. I am eager to hear your thoughts about it when we resume on Tuesday. And we will also begin review of Crawford’s chapter on data.