How IBM once paid people to *not* work

In 1993, IBM faced a crisis. The computing giant had bloated to 406,000 employees while profits cratered, posting an $8 billion loss—the largest corporate loss in American history at the time. Layoffs were the obvious move, but severance packages, outplacement services, and potential lawsuits made the math ugly.

So IBM's finance team found a loophole in their own logic. They created the 'resource rebalancing' program: identify workers whose skills were obsolete or redundant, pull them off projects entirely, keep them on full salary, and park them in a holding pattern while the company figured out what to do with them. Thousands of employees reported to empty offices, attended pointless meetings, or simply stayed home. Some waited months. A few waited over a year.

The scheme worked because of accounting arcana. Active employees were expensive—benefits, office space, management overhead. But workers on the bench? They became a line item that could be quietly whittled down through attrition, early retirement packages, and eventually, small voluntary buyouts that cost far less than mass layoffs. IBM's stock recovered. By 1994, the company was profitable again.

The program became a Silicon Valley whisper, a case study in how creative accounting could turn a humanitarian nightmare into a fiscal strategy. Other companies took notes. The 1990s dotcom boom saw dozens of firms adopt similar 'redeployment' schemes, though none as brazen or as large as IBM's.

Today, it's mostly forgotten outside HR departments and business school ethics modules. But the math still works, which is why every few years, some CFO rediscovers it.