Jobs Lost Automation PwC: Facts & Figures
- Time horizon (years)
- 10 years
- You entered
- Annual automation adoption rate
- 3 %
- You entered
PwC’s research on automation and jobs does not predict a specific net loss of jobs to automation by a fixed year. Instead, it projects that around 30% of jobs in the UK and up to 38% in the US are at potential risk of automation by the early 2030s, with the actual impact depending on how quickly businesses adopt the technology and how workers adapt. This article explains what PwC’s studies actually say, how the numbers are calculated, and what they mean for workers and policymakers.
Formula
The calculator above estimates the percentage of jobs that could be at risk of automation over a given time horizon, assuming a constant annual adoption rate. The formula is:
Jobs at risk (%) = 100 × (1 − (1 − r/100)^t)
- r = annual automation adoption rate (percentage of tasks automated per year)
- t = time horizon in years
The result is the cumulative share of jobs that may be affected, assuming the adoption rate stays constant. It is a simplified model that does not account for new job creation or task-level variation.
Worked Example
Using the default values: r = 3% per year, t = 10 years.
- Convert r to a decimal: 3/100 = 0.03
- Calculate (1 − 0.03) = 0.97
- Raise to the power of t: 0.97^10 ≈ 0.7374
- Subtract from 1: 1 − 0.7374 = 0.2626
- Multiply by 100: 26.3%
So with a 3% annual adoption rate over 10 years, about 26.3% of jobs could be at risk. This matches the calculator’s output.
What PwC’s Research Actually Says
PwC’s widely cited reports, such as the UK Economic Outlook (2017) and the Will robots really steal our jobs? (2018), used an AI exposure index based on tasks that are automatable with current or near-future technology. They found that:
- In the UK, about 30% of jobs (around 10 million) could be at high risk of automation by the early 2030s.
- In the US, the figure is slightly higher at 38%.
- In Germany and Japan, the estimates were 35% and 21% respectively.
These percentages refer to the share of jobs with a high probability of being automated, not a guaranteed loss. PwC emphasized that automation will change tasks within jobs more than eliminate entire roles.
How the Estimates Are Calculated
PwC’s methodology divides jobs into three categories based on task composition: routine, non-routine, and manual. They then apply an “automation potential” score to each task, drawing on data from the OECD and national labor surveys. The key inputs are:
- Task-level data from surveys like the O*NET database in the US.
- Technical feasibility of automating each task with current AI and robotics.
- Sectoral employment data to weight the results.
The result is a percentage of tasks in each job that could be automated, which is then translated into a job-level risk score. Jobs with more than 70% of automatable tasks are considered “high risk.”
Which Sectors and Jobs Are Most Affected
PwC’s analysis consistently shows that certain sectors have higher automation potential:
- Transportation and storage – driving and logistics tasks are highly automatable.
- Manufacturing – repetitive assembly and quality control.
- Wholesale and retail – cashier and stock management tasks.
- Administrative and support services – data entry and scheduling.
In contrast, education, health care, and professional services have lower automation potential because they require complex human interaction and judgment.
Common Misconceptions
A frequent mistake is to interpret “at risk” as “will be lost.” PwC’s reports are clear that automation can also create new jobs and increase productivity. Another misconception is that automation will happen overnight; the adoption rate depends on cost, regulation, and social acceptance. Finally, the 30% figure is not a prediction of unemployment – it is a measure of potential exposure, not actual job losses.
How to Use This Calculator
This calculator is a simple tool to explore how different adoption rates and time horizons affect the cumulative share of jobs at risk. It is not a forecast. Use it to understand the sensitivity of the results to assumptions. For example, if you think automation will be adopted faster (say 5% per year), the share rises to about 40% after 10 years. If adoption is slower (1% per year), it falls to about 9.6%.
Remember that the real world is more complex: new jobs are created, tasks change, and humans adapt. The calculator is for educational purposes only.
Frequently Asked Questions
What is the exact number of jobs PwC says will be lost?
PwC does not provide an exact number of job losses. They estimate that around 30% of UK jobs (about 10 million) are at high risk of automation by the early 2030s, but “at risk” does not mean “will be lost.” Many of those jobs will change rather than disappear.
Is PwC’s 30% figure the same as the Oxford University study?
No. Oxford’s 2013 study by Frey and Osborne suggested 47% of US jobs were at risk, but PwC’s methodology is different and more conservative. PwC’s task-based approach yields lower numbers for most countries.
Which jobs are safest from automation according to PwC?
Jobs that require high levels of social intelligence, creativity, and physical dexterity in unpredictable environments are safest. Examples include nurses, teachers, and managers.
How can workers prepare for automation?
Focus on developing skills that complement automation, such as critical thinking, emotional intelligence, and technical literacy. Lifelong learning and adaptability are key.
Does the calculator account for job creation?
No. The calculator only estimates the cumulative share of jobs at risk under a constant adoption rate. It does not model new jobs that automation might create.
Related
Formula used
in t | Time horizon (years) | years | 10 | 0 in r | Annual automation adoption rate | % | 3 | 0 out jobs | Jobs at risk (per 100) | % | 100 * (1 - (1 - r/100)^t) | 1
Each line is one input (in, sel, list) or one result (out: label, unit, formula, decimals). How the calculators work
Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.