The idea of universal basic income (UBI) is near the peak of the hype cycle. Democrat Andrew Yang made it the flagship issue of his presidential campaign. A small industry of advocates tirelessly push arguments in its favor. I will address two in this piece. The first: the claim of permanent elimination of jobs. The second: the resulting need for income to compensate for the fall in purchasing power from the lack of work. Both rely on long-discarded economic fallacies.
No one doubts that robots, software, and automation eliminate some need for human labor where adopted. But the automation doomers' scenario assumes that when jobs are eliminated by automation in one place, that number of jobs are permanently gone. For this to be true, there would have to be no compensating growth in the need for labor elsewhere.
The purchasing power argument says that the economy will suffer from an overall loss of demand due to the reduction in income when people are out of work. Martin Ford (futurist and New York Times bestselling author of Rise of the Robots), thinks that UBI is “the answer to job automation” because it will “ensure that consumers have money to spend—because the market economy requires that there be adequate demand for products and services.”
These two arguments turn out to be related through Say’s law. This is the name we give to the observation that when a producer supplies a good, their action constitutes a demand for a different noncompeting good. It is correct that if the workers remained permanently unemployed, the economy would experience a lack of demand from the reduction in supply. The formerly productive workers, who are no longer producing, no longer contribute to the supply of goods. By not supplying, they remove their contribution to demand as well. However, if the workers who were made redundant in one industry can find gainful employment doing something else, then they may continue to supply, and therefore to demand. And then, there would be no systematic deficiency of demand.
The UBI advocates are correct that some jobs are replaced when capital goods do the work that was done by labor. Robots are a capital good. If the same amount of output can be produced by a mix of more robots and fewer people, an industry will not offer as much employment as before. Does it follow that when one industry uses fewer workers there is no need for their services anywhere else? How would the advocates of this view explain the enormous growth in the labor force since the Industrial Revolution two centuries ago—a period characterized by increasing capital intensity?
The answer is that fall in demand for labor in the more capital-intensive industries is only the start, not the end, of the story. The economist does not stop with the immediate effect of a change. The true economist analyzes how the entire system adjusts to a change. Hayek saw “the increasing concentration on short-run effects … not only as a serious and dangerous intellectual error, but as a betrayal of the main duty of the economist and a grave menace to our civilization.”
Where did the robots come from? Was there a huge warehouse of unused robots dropped by aliens? Someone had to produce the robots. The process of creating the new capital goods starts with less consumption to fund additional saving and investment. This initially creates more demand for labor in the capital goods sectors, to build the robots. This demand to some extent compensates for the loss of employment in consumer goods from spending shifted from consumption to capital goods. The creation of robots requires engineers, manufacturing, sales, marketing, and all of the other services that form a full supply chain.
Jesús Huerta de Soto in Money, Bank Credit, and Economic Cycles shows how the increase in the supply of capital goods relative to labor changes the most profitable combinations of labor and capital for that industry. The lowest-cost mix then consists of more capital and less labor. More output in that industry can be created at lower cost by fewer people:
This increase in real wages, which arises from the growth in voluntary saving, means that, relatively speaking, it is in the interest of entrepreneurs of all stages in the production process to replace labor with capital goods.
Jesús Huerta de Soto, Money, Bank Credit, and Economic Cycles, trans. Melinda A. Stroup, 4th ed. (Auburn, AL: Ludwig von Mises, 2020), p. 329.
The adoption of more productive capital goods raises the productivity of labor. This means that less labor is required in that industry to produce the same quantity of goods at lower cost. Mises explains what happens following the substitution of capital for labor:
What happens is that labor is rendered more efficient by the aid of machinery. The same input of labor leads to a greater quantity or a better quality of products. The employment of machinery itself does not directly result in a reduction of the number of hands employed in the production of the [product].
Ludwig von Mises, Human Action: A Treatise on Economics, scholar's ed. (Auburn, AL: Ludwig von Mises Institute, 1998), p. 768.
The lower costs are through the resulting price competition passed on to consumers. This is another way of saying that consumers now have a higher real wage. Consumers can buy the same amount of products—let’s say shoes—as before and still have some money left over to buy something else that they previously could not afford.
All increases in voluntary saving exert a particularly important, immediate effect on the level of real wages…. increases in saving are generally followed by decreases in the prices of final consumer goods. If, as generally occurs, the wages or rents of the original factor labor are initially held constant in nominal terms, a decline in the prices of final consumer goods will be followed by a rise in the real wages of workers employed in all stages of the productive structure. With the same money income in nominal terms, workers will be able to acquire a greater quantity and quality of final consumer goods and services at consumer goods’ new, more reduced prices.
Huerta de Soto, Money, Bank Credit, and Economic Cycles, p. 329.
This is another way of saying that when one industry becomes more productive, everyone else gets a raise. Contrary to Martin Ford, there is no need for income to generate purchasing power disconnected from production. More capital-intensive production, through lower prices, generates the purchasing power to buy the final products.
But that is still not the end of the story. The ability of the workers in other industries to buy something new creates the opportunity for some businesses to expand production, or for new businesses to make new products. Without the drop in the output prices of the more capital-intensive industries, these new products would not have been affordable for these workers. The labor needed to produce those products would not have been available at a wage that would have made it profitable to produce them, because the labor would have been more urgently needed doing something else. When the labor was replaced by robots, things changed.
Mises describes the process using a hypothetical product called “A” to make the point that the release of labor from some uses makes other uses economically viable:
The technological improvement in the production of A makes it possible to realize certain projects which could not be executed before because the workers required were employed for the production of A for which consumers’ demand was more urgent. The reduction of the number of workers in the A industry is caused by the increased demand of these other branches to which the opportunity to expand is offered.
Mises, Human Action, p. 768.
Incidentally, this insight explodes all talk about "technological unemployment.”
The British Austrian school economist William H. Hutt addressed the unemployment of labor extensively in several of his books. He emphasized the importance of price flexibility. Any productive service has a value, and therefore a price, somewhere, doing something useful. Price flexibility and open labor markets are necessary in order for workers to be matched up with work that consumers value the most. When labor productivity increases in one industry, on average more goods are available for everyone to buy, but wages can still rise, or fall, in each industry or geographic area, depending on the skills of the workers in the labor market, the types of goods and services in demand, and the quantity and quality of preexisting capital goods. If the job losses occur in the increasingly capital-intensive industry, those workers must be free to offer their services in other areas where there is demand.
Problems with the price system can appear to manifest as chronic unemployment. What appeared to be permanent unemployment in the British labor market of the 1930s, according to Hutt, was in reality excessively rigid and inflexible pricing in British labor markets. He blamed this primarily on labor unions, who, with government encouragement or inaction, were demanding above-market-clearing wages in many industries.
A UBI scheme does not replace the demand from the unemployed workers. According to Say’s law, demand originates with production. Paying people not to produce destroys their ability to demand. Hutt cites the nineteenth-century economist Frederick Lavington’s observation that the consumers’ proper name is “other producers.”
Handing out money can only transfer demand created by the people who remain in the labor force. If the program is paid for with money printing then the resulting inflation only transfers purchasing power from those workers who had sold their services for money and not yet spent it. The unemployed can only create demand by returning to work, by becoming producers again. The traditional means for people to obtain income—by earning it—is still the best way, even with robots.