The Welder Shortage: What the Public Data Actually Shows

"Welder shortage" is one of the most repeated phrases in manufacturing, and one of the easiest to exaggerate. This piece sticks to a few published figures from authoritative sources, explains what they mean and is careful about what they do not prove.

What a shipbuilding welder earns

The U.S. Bureau of Labor Statistics publishes the Occupational Employment and Wage Statistics (OEWS) survey, which estimates employment and wages by occupation and industry. In the May 2023 data, the mean hourly wage for welders, cutters, solderers and brazers in the ship and boat building industry was $27.82.

Across all industries, BLS reports a May 2025 median of $25.84 an hour for the same occupation. Both figures are wages, not fully loaded labor costs. Employers also pay benefits, payroll taxes, overtime premiums, training and supervision, so the loaded rate companies use to cost an hour of welding is higher than the base wage. The Catfish payback estimator asks for your own fully burdened hourly cost instead of supplying a default.

Robots are being installed

The International Federation of Robotics (IFR) tracks industrial robot installations worldwide. Its preliminary results put U.S. installations at about 38,000 in 2025, up 11 percent year over year (IFR, June 18, 2026), across every industry, with automotive and electronics historically the largest buyers.

The figure shows that industrial robots are a mature, widely bought technology in the U.S. For shipbuilding the open question is different: whether robots can handle high-mix, large-structure welding, where most of today's installed robots are not used.

What the numbers do not say

Neither dataset measures a welder shortage. A shortage is the gap between the skilled people employers need and the number available at prevailing wages. Measuring it means projecting demand, retirements and new entrants, and different organizations make those projections with different assumptions. We would rather be precise than dramatic, so we are not citing a single headline shortage number.

What shipyards report is simpler and more local. Experienced welders are hard to hire and keep, training new ones takes time, and the most experienced people often spend hours on repetitive joints instead of the difficult work only they can do.

Capacity matters more than headcount

Automation is often framed as replacing workers. In shipbuilding the realistic goal is capacity. When a yard has more steel to weld than skilled hours available, every hour of repetitive work moved to a robot is an hour a welder can spend on fit-up, complex joints, repair or training.

That changes the evaluation. Instead of asking how many people a cell replaces, ask how many skilled hours it frees and what that capacity is worth. That is the calculation behind a payback estimate.

Measure your own situation

Industry-wide numbers are context. A yard's decision should rest on its own data:

• How many welding hours go into each weld family?
• Which families are repetitive enough to automate?
• What is the loaded hourly cost of that labor?
• What utilization can a cell realistically reach on that work?

A weld-family study measures exactly these inputs, so the payback estimate is built on the yard's numbers rather than industry averages.

We do not publish productivity claims for Catfish systems that have not been measured in production. When we have measured results from installations, we will publish them with their context.

Skilled welding labor is valuable, and industrial robots are widely adopted. Whether automation makes sense for a particular yard is a question for that yard's own data.