The slide that gets a flattening decision approved looks almost the same everywhere it runs. Two org charts, side by side. The one on the left is tall, boxes stacked five and six deep between the front line and the C-suite. The one on the right is wide and clean, half the boxes gone, arrows running straight from individual contributors to a handful of remaining directors. Underneath sits a number: percent of the management layer removed, dollars saved a year, sometimes a projected gain in decision speed. The slide was built to answer a cost question. The question it rarely answers is who does the coaching now.

What the data actually shows

Gusto, a payroll company, looked at its own data on 8,500 small and mid-size business customers and reported in June 2025 that by late 2024, hiring for managers was running 40 percent below its January 2022 level, against about 11 percent for individual contributors, while layoffs had risen three times as much for managers.1 Gusto is careful to note that January 2022 sat in the middle of an unusual hiring boom, so some slowdown was expected. The number of individual contributors per manager went from roughly three in 2019 to nearly six by the fall of 2024, and between 2022 and 2024 it grew 44 percent at companies with 100 to 499 employees, against 24 percent at the smallest ones.1 Across the group, Gusto counts 14 percent of managerial roles cut.1 The case made for cuts like these is efficiency: fewer layers, faster decisions, lower cost.

The same report complicates that case. Pairing its data with Bureau of Labor Statistics figures on output per hour, Gusto found that more productive industries tend to employ more managers. Utilities and real estate, two of the most productive industries by that measure, had managers making up 13 and 9 percent of their workforces, while food service and administrative services, near the bottom, sat around 3 percent. Gusto says plainly that it is unclear why.1 An industry-level pattern like that cannot tell you management density causes productivity. It does mean the evidence behind a flattening decision is thinner than the slide suggests. It also sits comfortably beside the resume study I wrote about last week, where public companies with more layers earned higher returns on assets and steadier numbers.

What a manager is actually buying you

A management layer catches a problem before it becomes six problems. It is the person with time to notice a new hire is drowning before the new hire quits, and cutting it by headcount ratio removes the part of the system that was turning raw effort into results.

An earlier piece here made the case that a widened span erodes the coaching one manager can still give each person on the team. That argument stops at the individual manager. This one is about a different arithmetic: what happens when the number of managers in the system falls at the same time the number of people who might grow into the role also falls.

The other end of the ladder

The manager pipeline does not start with a promotion. It starts years earlier, in the entry-level job where someone first learns to plan a piece of work, own a mistake, and explain a decision to someone above them. Stanford's Digital Economy Lab, using payroll records from ADP, reported in August that employment among workers ages 22 to 25 in the occupations most exposed to generative AI now stands about 19 percent below where it would be had it kept pace with similar-aged workers in less-exposed jobs, measured from November 2022.2 The researchers call these descriptive patterns, not causal estimates. The gap shrinks once education is accounted for, and some of the divergence shows up before generative AI arrived. The ADP sample also shows bigger gaps than national surveys do.2 Caveats attached, this is roughly the group that would otherwise grow into the manager roles a company opens five years from now.

Put the two findings side by side and the flattening story gets harder to tell as pure discipline. They come from different companies, so treat this as a question to ask of your own building. A company that cuts managers while leaning on tools in place of junior hires is thinning the pool of people who would have grown into its managers by 2031. Neither decision looks reckless by itself, since a cost review that removes a layer reads as discipline and a hiring plan built around tools reads as the same discipline one rung down. Together, the company is eating its bench from both ends, and the annual report will not show it until the bench is empty.

None of this argues for keeping every manager who happens to hold the title today. Some of those layers earned the cut. What it argues for is a different question before the slide gets approved: which function those boxes were performing, and who inherits it. If the honest answer is nobody, the savings are borrowed against a bench nobody has checked in years. What would your own org chart tell you if you asked that question about the last layer you removed, and who would you have to go ask to find out?