TL;DR
A 2024 study of the U.S. telephone industry found that automatic switching reduced opportunities for young operators and harmed some women already in the job. Young women’s overall employment did not fall; many moved into clerical and service work. The history offers a case study for AI’s potential effects on clerical jobs, though it cannot predict exactly how today’s labor market will change.
Research on the introduction of automatic telephone exchanges in the United States found that the technology sharply reduced employment opportunities for young telephone operators, while young women’s overall employment did not decline. The findings, reported by economists James Feigenbaum and Daniel Gross in a 2024 study, offer a historical comparison as generative AI draws attention to clerical work, which the International Labour Organization ranks among the occupations most exposed to the technology.
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Telephone operators connected calls by plugging cords into switchboard jacks. By 1910, the job was done almost entirely by women, and in 1920 operators made up roughly 2% of working women in the United States. Feigenbaum and Gross estimate that about one in seven white, U.S.-born women born around 1900 worked as an operator at some point.
Automatic switching was adopted gradually, city by city. After a city converted to dial service, the number of young operators fell by 50% to 80%, according to the study’s findings as described in the source material. Women already working as operators were less likely to be employed a decade later; those who remained in work were more likely to have lower-paid jobs.
The change did not produce a comparable fall in young women’s employment overall. The source says many found work as typists, secretaries and in food service, including in industries that had not previously hired many women in those roles. Telephone operator numbers continued to rise nationally for decades as telephone use and other types of switching work expanded, reaching about 420,000 in the 1970 census before the decline accelerated.
What telephone operators teach us about AI and jobs
Automatic switching cut opportunities in one occupation, but young women’s overall employment did not fall. That distinction matters as generative AI reaches clerical work.
One technology, different outcomes
The study distinguishes job losses within an occupation from changes in overall employment.
- Some were less likely to be employed a decade later.
- Those still working were more likely to have lower-paid jobs.
- Employment did not decline.
- Many moved into typing, secretarial and food service work.
Automation spread over decades
Dial service arrived city by city, while other kinds of switching work continued to expand.
What this means for AI
The comparison offers questions to track, not a forecast of what generative AI will do.
Potential contact
Clerical work is highly exposed to generative AI, according to the source. Exposure alone does not show that a job will disappear.
Workers may fare differently
People already in exposed jobs may face different outcomes from those entering the labor market.
Tasks, hiring and pay
Track which tasks change, how hiring and pay shift, and whether affected workers move into other roles.
Lessons From a Switchboard’s Decline
The history separates two effects that can be blurred in debates about automation: a technology can sharply reduce demand for a particular occupation while workers’ overall employment follows a different path. In this case, younger women entered other kinds of work, while many incumbent operators faced worse prospects. That distinction matters for AI because a shift in tasks or job openings can impose costs unevenly, even if economy-wide employment does not fall.
The source article says the International Labour Organization identifies clerical work as the occupational group most exposed to generative AI, with exposure in richer countries falling more heavily on women’s jobs than men’s. Exposure does not establish that a job will disappear: it describes potential contact with the technology, not a measured employment outcome. The operators’ history makes the distribution of gains and losses a central question for current policy and workplace decisions.
Products Worth Considering

DJI Mini 4K Drone Combo with 4K Camera, 3-Axis Gimbal, and Two Batteries

DJI Mini 5 Pro Fly More Combo with DJI RC 2, 1-Inch CMOS

Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Bluetooth 5.3, USB-C Charging, and IPX4 Water Resistance
As an affiliate, we earn on qualifying purchases.
How Dial Service Spread
The Bell telephone company opened its first exchange in New Haven, Connecticut, in 1878. Manual exchanges required operators to connect each call, and the work force expanded alongside the network. Almon Strowger patented an automatic exchange in 1891, but adoption took many decades. A third of Bell telephones were on dial service by 1930 and 60% by 1940; the final conversions in the network cited by the source came in 1978.
The national employment count did not immediately track the technology’s local effects. The source gives about 182,000 telephone-industry operators in 1930 and around 420,000 in 1970, as long-distance calling and office switchboards also grew. It reports that the Bureau of Labor Statistics recorded a loss of 75,000 telephone operator jobs from 1972 to 1980, all held by women. Current counts use different occupational definitions: for May 2025, the BLS listed about 3,400 telephone operators and 34,300 switchboard operators, including answering services. These figures show broad scale, but the differing definitions limit direct comparisons.
What the Comparison Cannot Show
The telephone study describes a specific industry and period. It does not establish that generative AI will produce the same pattern, or that workers displaced from exposed tasks will find comparable jobs. The available source material also does not specify the study’s full methods or provide the underlying estimates, so the reported findings should be read as a summary rather than a substitute for the paper.
It remains unclear how quickly employers will adopt AI for clerical tasks, which tasks they will automate, and whether adoption will reduce headcount, change job duties or create other work. The 2025 operator figures use categories that differ from earlier census counts, and the source cautions that they are suitable for comparing orders of magnitude only.
Tracking AI’s Effects on Work
The next evidence will come from workplace adoption and labor-market data: which clerical tasks change, how hiring and pay shift, and whether affected workers move into other roles. Researchers and policymakers will need to distinguish workers already in exposed jobs from people entering the labor market, since the telephone history suggests those groups can experience different outcomes. The cited source identifies no single upcoming study or policy milestone that will settle those questions.
Key Questions
What happened to telephone operators when dial service arrived?
After a city switched to automatic service, the number of young operators fell by 50% to 80%, according to the study’s findings summarized in the source. Some incumbent operators also had worse employment prospects.
Did women’s employment overall fall?
The source says young women’s overall employment did not decline. Many moved into jobs such as typing, secretarial work and food service, though the article notes that women already working as operators could face losses.
Does the history predict what AI will do to jobs?
No. It is a historical comparison, not a forecast. It shows that occupational disruption and overall employment can move differently, but the effects of AI will depend on how employers adopt it and what work becomes available.
What does AI exposure mean for clerical jobs?
The source reports that the International Labour Organization ranks clerical work as highly exposed to generative AI. Exposure indicates potential interaction with the technology; it does not by itself show that a job will be eliminated.
Source: Thorsten Meyer AI
