TL;DR
Automatic telephone exchanges sharply reduced opportunities for young women entering operator work, while employment among young women overall did not fall. A 2024 study estimates that about one in seven white, U.S.-born women born around 1900 worked as an operator at some point; the effects on women already in the job were harder to absorb.
Automatic telephone exchanges replaced many manual switchboard jobs over several decades, sharply reducing opportunities for young women entering the occupation while young women’s employment overall continued. A 2024 study by economists James Feigenbaum and Daniel Gross estimates that about one in seven white, U.S.-born women born around 1900 worked as a telephone operator at some point, making the change a large labor-market shock with different effects on new entrants and women already in the job.
Get everyday essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
The shift began city by city after Almon Strowger patented an automatic exchange in 1891. The first dial systems did not immediately end manual switching: about one-third of Bell telephones were on dial by 1930, 60% by 1940, and the last parts of the U.S. network were converted in 1978. The number of operators continued to rise for decades as telephone use, long-distance calling and office switchboards expanded.
The telephone-industry census count peaked at about 182,000 operators in 1930, before the dial system took over. A later census counted about 420,000 in 1970, including workers at private switchboards. The figures use differing definitions and should be read as measures of scale, not as a directly comparable series. The source also reports that, after cities adopted dial service, the number of young operators fell by 50% to 80%.
Feigenbaum and Gross’s analysis finds that women already working as operators faced lasting costs: they were less likely to be working ten years after a city’s conversion, and those still employed were more likely to have lower-paid jobs. Younger women’s employment did not fall in the same way. Many took work as typists, secretaries and food-service employees, including in industries that had not previously hired such workers in large numbers.
What Happened to Telephone Operators in the Age of AI?
Automatic exchanges transformed telephone work over decades. Young women found other occupations, while many experienced operators faced lasting employment and pay costs.
A gradual shift
Dial systems spread city by city; operator jobs continued growing as phone use and switchboards expanded.
Operators connected calls by linking cords on switchboards.
Automatic exchanges reduced the need for manual connections over several decades.
Young women moved into typing, secretarial and food-service roles.
Different costs for workers
The study distinguishes the experience of new entrants from that of women already in operator jobs.
- Young women’s employment overall did not fall in the same way.
- Many entered other kinds of work.
- Less likely to be working ten years after conversion.
- Those still employed were more likely to have lower-paid jobs.
What the counts show
Historical totals use differing definitions, so they show scale rather than a directly comparable trend.
| Year | Reported count | Context |
|---|---|---|
| 1930 | About 182,000 | Telephone-industry census count |
| 1970 | About 420,000 | Broader census count including private switchboards |
What this means for AI
The operator case offers a historical example, not a forecast for generative AI.
Clerical work
The ILO ranks clerical work as the occupational category most exposed to generative AI.
Employment effects
Exposure does not predict job losses; outcomes depend on adoption, changing tasks and workers’ ability to move into other roles.
How Automation Changed Women’s Work
The history shows that an occupation can shrink without a matching decline in overall employment among the group most associated with it. In this case, new entrants found other work, while women with experience in the operator role were less likely to remain employed or to retain comparable pay. That distinction matters when assessing who bears the cost of a technology-driven change.
The comparison is relevant to current debates about generative AI, but it does not establish that AI will produce the same outcomes. The International Labour Organization ranks clerical work as the occupational category most exposed to generative AI and reports that exposure in higher-income countries falls more heavily on women’s jobs than on men’s. Exposure is not a forecast of job losses: the eventual effects depend on how employers adopt the tools, what tasks change and whether workers can move into other roles.
For readers weighing AI’s employment effects, the operator case offers a specific historical record: the job changed faster than the labor market could protect everyone already in it, even as later workers entered different occupations.
Products Worth Considering

Anker SOLIX C300 Portable Power Station, 288Wh

Bose QuietComfort Ultra Earbuds (2nd Gen), Wireless Noise-Cancelling Earbuds, Black

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite
As an affiliate, we earn on qualifying purchases.
From Switchboards to Dial Service
Bell opened its first telephone exchange in New Haven, Connecticut, in 1878. Calls initially required an operator to connect cords between jacks on a switchboard. The first operators were mostly boys and young men; telephone companies soon turned to young women, and by 1910 operators were almost all women.
Operator work became a major source of employment. In 1920, operators accounted for roughly 2% of working women in the United States. They were also about 4% of the nearly three million young, white, U.S.-born women who worked that year. AT&T was the country’s largest employer of women in the 1910s and its largest employer overall by the early 1920s; operators made up about half its workforce. High annual turnover meant many more women passed through the occupation than were employed at any one moment.
Companies described the job as demanding. A 1946 U.S. Women’s Bureau report listed expectations including a high-school diploma, good eyesight and hearing, a “pleasing voice” and a “stable disposition not easily ruffled by irritable customers.” Former operators described it as better paid and more social than factory work. Those accounts provide context for what was lost when the work contracted, though they do not describe every operator’s experience.
“About one in seven white, American-born women born around 1900 worked as a telephone operator at some point, according to their estimate.”
— James Feigenbaum and Daniel Gross, economists
What the Historical Counts Show
The employment figures come from different sources and use different definitions. Census counts include workers at private switchboards, while the May 2025 Bureau of Labor Statistics figures are wage-and-salary estimates that separate telephone operators from switchboard operators, including answering services. The source gives about 3,430 telephone operators and 34,280 switchboard operators in May 2025; these counts are useful for scale, but they are not directly comparable with earlier census totals.
The historical study describes employment patterns after dial conversion, but the source does not provide the full underlying estimates or enough detail to assess every difference among cities and workers. It also does not show how the arrival of generative AI will affect specific occupations. The ILO’s exposure ranking concerns which jobs may be affected by AI, not how many jobs will be eliminated or what workers will earn afterward.
The Next Test for AI at Work
The next developments will depend on how employers deploy generative AI in clerical and other exposed work. Relevant evidence will include changes in hiring, hours, wages and job tasks, alongside whether workers already in affected roles remain employed or move into other occupations. The historical operator case also suggests that counts of new openings alone may miss costs borne by incumbent workers.
The source materials provide no specific future milestone or forecast for AI-related job losses. Until comparable employment and wage data emerge, the scale and distribution of those effects remain unsettled.
Key Questions
How many women worked as telephone operators?
The 1930 census counted about 182,000 operators in the telephone industry, with another 23,000 at private switchboards in places such as hotels, hospitals and large firms. A 1970 census counted about 420,000, using a broader count that included private switchboards.
Did dial telephones immediately eliminate operator jobs?
No. Dial service spread gradually, and the total number of operators continued to grow as telephone use and switchboards expanded. The source says the number peaked in the 1970 census, even though conversions had begun decades earlier.
What happened to women who were already operators?
The study reports that incumbent operators were less likely to be working ten years after a city switched to dial service. Women still working were more likely to have lower-paid jobs, according to the study’s findings.
Did the automation reduce young women’s employment overall?
The source says young women’s employment did not fall overall. Many moved into roles such as typing, secretarial work and food service, although that does not mean every displaced operator found equivalent work or pay.
Does this history prove AI will eliminate clerical jobs?
No. The historical study concerns telephone switching, while the ILO ranking cited in the source measures occupational exposure to generative AI. Exposure does not by itself establish how many jobs will disappear or how workers’ pay and tasks will change.
Source: Thorsten Meyer AI
