Question:

What are your instinctive "first reflections" on this piece? What have you learned from Labor and Capital, Substitutes or Compliments that you believe, businesses should take into account in adapting new machinery to pair with their existing labor force? Finally, what kinds of companies do you think are more likely to achieve complementarity between capital and labor? Which ones are not as likely to succeed? How so?

Reading:

Automatic Teller Machines (ATMs) Have Complemented Some Types of Labor While Substituting for Other Types of Labor

As you have learned, a firm achieves its least-cost combination of inputs when the last dollar it spends on each input makes the same contribution to total output. This raises an interesting real-world question: What happens when technological advance makes available a new, highly productive capital good for which MP/P is greater than it is for other inputs, say, a particular type of labor?

The answer is that the least-cost mix of resources abruptly changes, and the firm responds accordingly. If the new capital is a substitute for a particular type of labor, the firm will replace that particular type of labor with the new capital (substitution effect). But if the new capital complements a particular type of labor, the firm will add additional amounts of that type of labor (output effect).

Consider bank tellers. One of their core functions, before ATMs became common in the 1980s, was to handle deposits and withdrawals of cash. Bank tellers had several other tasks that they needed to handle, but one particular type of labor they supplied was handling cash transactions. The task of handling cash transactions can, however, be equally well-managed by ATMsbut at one-quarter the cost. Naturally, banks responded to that cost advantage by installing more ATMs.

That might make you think that ATMs displaced tellers because it would seem natural in this scenario that ATMs were a substitute for bank tellers. But the data tells a different story! The number of human bank tellers actually increased, along with the ATMs. In 1985 there were 485,000 tellers, and by 2015, there were 526,000 tellers. This means that ATMs must have been a complement to the labor provided by the bank tellers.

The process was not instantaneous. Some 80,000 teller jobs were indeed lost during the 1990s because bank managers at first did think of ATMs and bank tellers as substitutes. But by the early 2000s, bank managers realized that the cost savings offered by ATMs had given them the chance to operate in new ways that would actually require more tellers rather than fewer tellers.

Before ATMs, the average branch needed 20 employees. After ATMs, the average branch needed only 13 employees. That major increase in efficiency gave banks the chance to compete against one another by opening more branchesand more branches meant having to hire more human tellers. Thus, the efficiencies generated by having ATMs handle cash transactions at one-fourth the cost caused the demand for bank tellers to increase.

In addition, the banks also realized that bank tellers could be trained in more complex tasks like selling financial products and helping to issue home mortgages. Once banks figured out these new ways to employ tellers, ATMs turned from a substitute for bank teller labor into a complement for bank teller labor. By freeing human beings from having to handle cash transactions, ATMs acted as a complement for other types of human labor, such as selling financial products.

We can generalize from the history of ATMs. Capital is, overall, a complement to, rather than a substitute for, human labor. Certain types of human labor are substituted away as new technologies arrive, but humans end up being complemented by capital as they perform other tasks. Some types of work will disappear and the government may need to help displaced workers train for new jobs. But the newly deployed capital will end up increasing wages because it will, as a complement, increase the overall demand for human labor.