Start with a worker, not a layoff headline. A software developer with four years at one firm loses the role in a restructuring. She finds full-time work again within the year. Her new paycheque is smaller than her old one, and the data says her case is the common one.
The direct answer: of Americans displaced from jobs they held at least three years, 66.1% were employed again by January 2026. Of those back in full-time work who reported prior earnings, only 49% earned at least as much as before. Software vacancies still sit 27.5% below pre-pandemic levels. Displaced workers re-enter the market at lower pay, in thinner numbers, and then spend less with firms like yours.
- The re-entry numbers
- A thinner door back in
- Where the loop reaches your till
- What the numbers do not show yet
- Hire like the loop is real
The re-entry numbers
The displacement figures cover 2023 to 2025 and count workers who lost jobs they had held for at least three years. Two of three were back at work by January 2026. That sounds like recovery until the wage line lands. Half of the full-time rehires earn less than they earned before. The other half matched or beat the old wage.
Experience does not protect the wage. These are not first-job workers. They held roles for three years or more, in a market where employers say skill is scarce. The discount lands anyway.
A thinner door back in
The door back is narrowest in software. US software development vacancies remained 27.5% below pre-pandemic levels as of June 2026, after a year of partial recovery. Fewer postings chase the same displaced pool. A senior developer competes for a mid-level seat, and the mid-level worker competes downward with her.
Canada reads the same shape in its own numbers. Before you blame AI for fewer junior hires, read StatCan’s new numbers shows youth hiring falling across every exposure group, with separations steady. Two studies disagree about junior hires, and the answer is in your task list works through what that means at the level of one firm’s task list. The pattern matches the US data. The squeeze sits at the front door, not the exit.
Where the loop reaches your till
Follow the smaller paycheque. A rehired worker earning 15% less postpones the truck repair, the kitchen job, the winter holiday. Households with thin cash reserves cut spending fastest after a job loss. Multiply one smaller paycheque by a restructuring wave and the demand side of your order book softens.
The wider US numbers show how narrow the spending base already is. Real disposable income fell 0.3% in September 2026. The volume of goods and services bought stayed almost flat. Spending growth since 2023 has come mostly from households earning over $125,000 a year. A demand base that rests on top earners has less slack when mid-pay workers take wage cuts on re-entry.
This is the loop owners miss. Each firm counts its own payroll saving as a win. The same workers were customers somewhere else, often within the same province. When many firms cut at once, each one’s saving becomes another one’s lost sale. No single firm causes it. Every firm feels it.
What the numbers do not show yet
Hold the claim at the size the data supports. Nobody has shown that AI-driven cuts are the force behind these spending figures. The displacement count covers every cause of job loss, not AI alone. The honest reading is narrower. A large displaced pool is re-entering at lower pay while the vacancy door stays thin. That pattern weakens demand whether AI caused it or not.
The cause still matters for timing. If the cuts are a one-time restructuring wave, the loop fades as rehiring completes. If AI keeps removing the same roles each year, the pool refills and the loop runs on. Watch the next displacement release for the direction. One more cycle of the same numbers turns a wave into a structure.
Hire like the loop is real
Canadian firms will meet these resumes this winter. Large US tech employers are restructuring with Canada named in their workforce plans, and the displaced pool includes people with three-plus years in real roles. Treat that pool as an asset class, not a discount bin.
Price the role, not the desperation. A hire made at a panic discount leaves at the first full-price offer. Ontario’s AI hiring rule is hiding inside your applicant tracking system covers the disclosure line your posting may already owe if software screens the pile. And before you choose between a rehire, a contractor and a tool, Before you hire, outsource or automate, fix the process first gives the order of operations.
By spring, pull your own re-entry count. Count how many of your last ten hires came from a restructuring, and what each earns against the role’s old pay band. If the discount is deep and the work is strong, you found the market’s error. Hire before it corrects.
