Will they hire us back? What the 1990s downsizing wave teaches

In the 1990s, about two-thirds of firms that cut staff in one year cut again the next year. That is the first thing history says about the question every layoff round leaves in the corridor: will they hire us back when this is over? It is not the answer people hope for. About a third of the firms that laid people off later rehired some of them, often as contractors. And when the cuts came from a real downturn, getting back to the old headcount took 4 to 6 years.

Contents · 11 sections

TL;DR #

• In the 1990s, about two-thirds of firms that cut staff in one year cut again the next year.
• About a third of firms that laid people off later rehired some of them, often as contractors.
• After the dot-com bust, tech employment took about 4 years to recover; after 2008, US payrolls took about 6.
• Rehiring starts when demand recovers, not when the layoffs stop. For the 2025–26 cohort, I expect selective rehiring, mostly on contract.

I want to walk through that history, because 2026 is rhyming with it, and because the rhyme tells us something useful about 2027 and 2028.

Where we are, as of 13 November #

Microsoft has offered voluntary buyouts to about 60,000 people, roughly a quarter of its staff, with a deadline in early December. LSN News, my news engine, grouped the coverage of the staff open letter that followed into one story: seven outlets in three languages. Across the US, though, the picture is calmer than the headlines. According to Challenger, Gray & Christmas, announced US job cuts in October were 43,994, a three-year low. “Market and economic conditions” has overtaken AI as the most-cited reason this year.

So we are in the phase after a wave, not in the middle of one. That is exactly when the “will they hire us back” question gets asked. It is also when the 1990s can help.

The decade of re-engineering #

The 1990s downsizing wave had a manifesto. Michael Hammer’s 1990 Harvard Business Review article, “Reengineering Work: Don’t Automate, Obliterate”, and the 1993 book with James Champy told managers to redesign work from scratch. By 1993, as many as 60% of the Fortune 500 said they had started re-engineering or planned to, according to a US General Accounting Office guide. In practice the label often became a polite word for layoffs. The cuts ran for about a decade and, unusually, through an economic expansion, not only in recessions.

The best long-run test of whether this worked is in a book on my shelf. In my library, Resizing the Organization, edited by Kenneth De Meuse and Mitchell Marks, carries a chapter by Wayne Cascio and Clifford Young. They followed S&P 500 companies from 1982 to 2000. Their conclusion is blunt:

“It was not possible for firms to ‘save’ or ‘shrink’ their way to prosperity.”

The numbers behind it are small and sobering. Firms that cut employment shrank by an average of 5.66% over two years. Their return on assets rose by about half a percentage point, but their profitability never exceeded that of stable employers who did not cut. A dollar invested in a downsizer grew to $1.72 after two years; a dollar in a stable employer grew to $1.69, about the same. The firms that grew their asset base turned the dollar into $2.42.

Why they cut again #

If cutting had solved the problem, firms would have cut once. They didn’t. Cascio, in his 2002 book Responsible Restructuring, put it plainly: on average two-thirds of firms that lay people off in a given year do so again the following year.

The American Management Association’s surveys of the period, as summarised in Newsweek, add the rest of the story:

• only about half of the firms that downsized said operating profits went up;
• only about a third reported better productivity;
• 88% said morale had declined;
• and about one-third later rehired some of the people they had laid off, as contractors, because they still needed the skills.

That last point is the honest answer to “will they hire us back”. Some of you, yes. But often on a contract, without the benefits, and through a staffing firm that takes a margin.

A 2001 Society for Human Resource Management survey, reprinted in Cascio’s book, found a similar split: among firms that had laid people off since 2000, 45% had rehired some full time and 17% had brought some back as consultants.

Cascio also explains why firms end up rehiring. In the same chapter of Resizing the Organization he writes that flexibility needs some slack, in people as much as in plant:

“It may be advantageous to maintain human resources even in slower periods in order to support flexibility.”

Firms that cut to the bone discover, when demand returns, that they have no one to run the new work. So they buy the skills back at market price.

How long the way back took #

Two recent episodes give us a clock.

The dot-com bust. US tech employment fell by about 415,000, roughly 20%, between January 2001 and December 2002, according to the industry’s own trade association figures reported at the time. It took about 4 years, until around mid-2005, for headline tech employment to recover. Even then, reported programmer employment was still down by about a quarter. The total came back; the same jobs did not.

2008–09. This was a demand collapse, not a technology shift. US payrolls took about 6 years, until mid-2014, to regain their January 2008 peak, according to the Bureau of Labor Statistics. Matthew Desmond notes in Poverty, by America, also in my library, that incomes for families in the bottom half took nearly a decade to recover.

Two lessons from those clocks. First, the rehiring clock starts when demand recovers, not when the layoffs stop. Second, what comes back is a different mix of jobs, at different terms.

There is one bright example of the other path. In Cascio’s 2009 guide for the SHRM Foundation, TSMC cut about 3% of its staff in early 2009 when revenue fell. When revenue jumped 80% the next quarter, it rehired 700 of the people it had let go and paid extra to those who chose not to return. It cut little, cut late and kept the door open. That is rare.

What 2026 already shows #

Some of the 1990s pattern is already visible.

• “AI rebounding”. A survey by the assessment firm SHL, reported by Fortune, found that 19% of US respondents said they had been replaced by AI and then asked back. Young workers were the most affected. It is a vendor survey, so treat it as reported, not documented.
• Quiet rehiring. Meta is reported to be hiring “hundreds” of staff in Ireland months after its AI-push cuts, according to the Business Post. Again, reported.
• Voluntary first. Microsoft, Jaguar Land Rover and Michelin are all running voluntary schemes this autumn. In the 1990s, voluntary rounds were usually followed by compulsory ones.

What to watch: what history predicts for 2027–28 #

Here is my reading. These are inferences from the pattern, not facts, and I give rough odds so you can hold me to them.

1. Repeat cuts are more likely than rehiring. The two-thirds rule suggests that many firms that cut in 2025–26 will cut again in 2027. I think it is likely (about 65%) that Microsoft follows its December buyout window with compulsory cuts within six months, as it did between April and July 2026.
2. No broad rehiring of the 2025–26 cohort. I put it at about 70% that rehiring stays selective: customer support, quality and compliance, where AI disappointed, and often on contract or offshore.
3. If a downturn comes, it will be called AI. A strong US labour market plus a possible Federal Reserve rate rise is the mix that tends to come before the next cost-cutting round. If a slowdown arrives in 2027, many firms will label cyclical cuts as AI. That would restart the clock: on the dot-com and 2008 timetables, the broad way back would then run to 2030–2032.
4. The way back will be through contracts. The AMA’s one-in-three contractor figure is the most durable number in this history.

Checklist #

If you have been cut, or expect to be, the history suggests three practical moves.

• Keep the door open. The first people rehired are the ones the remaining managers still remember. Leave well, and stay in touch with the people who know what you did.
• Price the contract properly. If you are asked back as a contractor, you are being asked because your skills are scarce. Charge for the benefits and the risk you now carry.
• Watch demand, not announcements. Rehiring follows orders and revenue, not press releases. If your old employer’s sales are recovering, that is when to call.

What I don’t know #

The 1990s figures come from surveys of large US firms, and survey answers are not audited. The AMA and SHRM numbers are cited second-hand, from Cascio’s books and Newsweek, not from the original reports. The 2026 rehiring signals are early and mostly reported, not documented. And my own news data only goes back to 7 November, so it can show what is being covered now, not the shape of the whole wave.

Frequently asked questions #

Do companies rehire people after layoffs? #

Some do, but rarely at scale. In the 1990s, about a third of the firms in the American Management Association’s surveys rehired some laid-off staff, often as contractors. A 2001 SHRM survey found 45% had rehired some people full time and 17% as consultants.

How long does it take for jobs to come back after a wave of cuts? #

It depends on demand. After the dot-com bust, US tech employment took about four years to recover. After 2008, total US payrolls took about six years. In both cases the jobs that came back were not the same jobs.

Did the 1990s layoffs make companies more profitable? #

Not on average. Cascio and Young’s study of S&P 500 firms from 1982 to 2000 found that firms that cut staff never beat the profitability of stable employers. Firms that invested and grew did best.

Will people cut for AI in 2025–26 be hired back? #

Some will, selectively. Early signs, such as SHL’s “AI rebounding” survey and Meta’s reported rehiring in Ireland, point to targeted rehiring in areas where AI disappointed, often on contract. I don’t expect a broad return of the whole cohort.

The lesson #

Shrinking alone rarely made firms richer. They kept cutting, and when demand came back they bought skills back on worse terms for the people concerned. If 2026 follows that pattern, the honest answer to the corridor question is: some people, later than they hope, and probably on a contract. The useful move is to watch demand, keep the door open and price the contract properly.