Every question answered in the Atlas, on one page: the short answer, the finding behind it, the study it comes from, and a link to the full answer with every study it uses.
In the studies we mapped, what gets called resistance usually has a reason behind it.
The finding Over three years of a radical change, middle managers went from supporting their top leaders to mutiny as they reassessed whether those leaders were legitimate (Huy et al., 2014).
Study: Huy et al. (2014) · qualitative case · Medium. Based on 8 studies from the Atlas of the Middle: 0 rated high strength, 3 medium, 5 low. Designs: qualitative case (6), qualitative interviews (1), survey (1).
Because it gets stuck between levels, and the studies show several ways that happens.
The finding In a European consumer-goods firm, each extra manager at the second-highest level was associated with roughly 10% fewer employee ideas passed upward (Reitzig & Maciejovsky, 2015).
Study: Reitzig & Maciejovsky (2015) · mixed · Medium. Based on 10 studies from the Atlas of the Middle: 0 rated high strength, 7 medium, 3 low. Designs: qualitative case (5), survey (2), mixed (1), conceptual (1), other (1).
The research does not settle it: about half the studies support starting change close to the work, and nearly as many find it depends.
The finding In one survey, employees backed change most when middle managers started it and top managers carried it out, yet that pattern was the rarest, about 3% of cases (Heyden et al., 2017).
Study: Heyden et al. (2017) · survey · Medium. Based on 9 studies from the Atlas of the Middle: 0 rated high strength, 6 medium, 3 low. Designs: qualitative case (4), survey (3), experiment (1), qualitative interviews (1).
They translate: in the studies we mapped, middle managers turn leaders' ideas into everyday work, negotiate what vague mandates mean, recast the case for a change in terms senior leaders accept, and, with senior managers, build shared understanding in informal conversations around formal meetings.
The finding In an Italian court house modernisation, middle-manager 'mediators' translated leaders' ideas in three distinct ways, and the approaches differed in how well they turned those ideas into everyday work (Agostino et al., 2013).
It loads them up: more work, closer monitoring and fewer chances of promotion.
The finding In 64 interviews across five UK organizations going through large-scale restructuring, including delayering and downsizing, middle managers reported longer and more intense hours, closer monitoring and fewer promotion prospects, even with more autonomy (McCann et al., 2008).
Study: McCann et al. (2008) · qualitative interviews · Medium. Based on 8 studies from the Atlas of the Middle: 0 rated high strength, 6 medium, 2 low. Designs: qualitative interviews (3), qualitative case (2), survey (1), experiment (1), mixed (1).
The research we mapped doesn't say: only six of the 173 studies in the Atlas (as of 30 October 2026) speak to reviewing a change or holding it steady, and none of them tests how long a change should settle.
The finding Across 331 Australian firms, budgeting, forecasting and reporting routines in which top and middle managers interact were tied to stronger business-unit performance (Peters et al., 2019).
Study: Peters et al. (2019) · survey · Medium. Based on 6 studies from the Atlas of the Middle: 0 rated high strength, 5 medium, 1 low. Designs: qualitative case (3), survey (1), mixed (1), qualitative interviews (1).
The studies we mapped don't test whether cutting layers pays off, but they show two things worth weighing first.
The finding In 64 interviews across five UK organizations going through large-scale restructuring, including delayering and downsizing, the middle managers who stayed reported more autonomy but longer, more intense hours, closer monitoring and fewer promotion prospects (McCann et al., 2008).
Study: McCann et al. (2008) · qualitative interviews · Medium. Based on 7 studies from the Atlas of the Middle: 0 rated high strength, 5 medium, 2 low. Designs: qualitative case (2), qualitative interviews (2), other (1), conceptual (1), archival (1).
Yes, in several kinds of studies, though most show association rather than cause.
The finding At a large trucking company, a randomized empowerment program for unit managers (38 treated and 30 control managers) lifted their units' measured performance (Logan & Ganster, 2007).
Study: Logan & Ganster (2007) · experiment · Medium. Based on 7 studies from the Atlas of the Middle: 0 rated high strength, 7 medium, 0 low. Designs: survey (5), experiment (1), archival (1).
Frame the idea in senior leaders' terms and time it well: that is where the studies we mapped point, along with the standing of the person pitching it, though much of this comes from cases, theory and practitioner advice.
The finding At IBM, after a plasma display bet failed, middle-level champions recast the case for LCDs around business issues rather than technology, and IBM went on to lead LCDs (Eggers, 2016).
Study: Eggers (2016) · qualitative case · Medium. Based on 8 studies from the Atlas of the Middle: 0 rated high strength, 5 medium, 3 low. Designs: qualitative case (3), survey (2), conceptual (1), other (1), mixed (1).
The finding Across 107 hotels run by one management company, in a survey not tied to any particular change, middle managers who saw their senior bosses as keeping their word were in turn rated by frontline staff as more true to their own word (Simons et al., 2007).
Study: Simons et al. (2007) · survey · Medium. Based on 7 studies from the Atlas of the Middle: 0 rated high strength, 5 medium, 2 low. Designs: survey (3), qualitative case (2), qualitative interviews (1), mixed (1).
In the cases we mapped, changes faded or failed once under way when old hierarchies came back, middle managers were left powerless, or bad news was filtered on its way up, though no study here measures stalling directly.
The finding In one of two hospitals, where nurse middle managers coped less well with pressure to meet regulations and save money, a change that worked at first faded as old hierarchies reasserted themselves (Spyridonidis & Currie, 2016).
Study: Spyridonidis & Currie (2016) · qualitative case · Medium. Based on 9 studies from the Atlas of the Middle: 0 rated high strength, 8 medium, 1 low. Designs: qualitative case (6), qualitative interviews (2), survey (1).
Look for gaps between what you are told and what the middle sees.
The finding In a hospital facing downsizing and a move, executives talked more about participation and communication, while supervisors and frontline staff talked more about conflict and division (Jones et al., 2008).
Study: Jones et al. (2008) · qualitative interviews · Low. Based on 9 studies from the Atlas of the Middle: 0 rated high strength, 7 medium, 2 low. Designs: qualitative case (5), qualitative interviews (2), mixed (2).
No study we mapped compares ways of communicating a change head to head, but in one, interviewed employees preferred their supervisors for practical information about a change and saw senior leaders as the source of strategy (Allen et al., 2007).
The finding In a government department facing restructure, a follow-up field study in the same paper found that those who rated change communication highly were more open to change, partly through feeling less uncertain.
Study: Allen et al. (2007) · mixed · Medium. Based on 6 studies from the Atlas of the Middle: 0 rated high strength, 2 medium, 4 low. Designs: qualitative case (2), mixed (1), survey (1), other (1), qualitative interviews (1).
A real one: middle managers help shape strategy, though the research describes their role more than it measures it.
The finding Drawing on 262 interviews, one study argues the same role expectations can both free and limit middle managers' strategic action (Mantere, 2008).
Study: Mantere (2008) · qualitative interviews · Medium. Based on 9 studies from the Atlas of the Middle: 0 rated high strength, 3 medium, 6 low. Designs: qualitative case (4), conceptual (2), review/meta-analysis (1), qualitative interviews (1), archival (1).
Because they sit on both sides of control at once: that is one answer from the studies we mapped.
The finding In focus groups, middle managers shifted between going along with and pushing back on the standard manager role, both controlled and controlling (Harding et al., 2014).
Study: Harding et al. (2014) · qualitative interviews · Low. Based on 7 studies from the Atlas of the Middle: 0 rated high strength, 1 medium, 6 low. Designs: conceptual (3), qualitative interviews (2), qualitative case (2).
In the studies we mapped, mostly by finding and connecting what already exists.
The finding At one ICT multinational, middle managers ran intense, cross-silo searches that pieced together answers to unusual problems (Tippmann et al., 2014), and another study found such searches sometimes modified or created routines (Tippmann et al., 2013).
The studies we mapped tie it to the setting around middle managers, such as leadership, HR systems and how they compare with peers, though all of them show association, not cause.
The finding Across 55 units in 27 firms, units with more transformational leadership showed more corporate entrepreneurship, partly through members' shared belief that they could succeed together, and how strong that indirect link was depended on the firm's empowerment climate (Chang et al., 2017).
Study: Chang et al. (2017) · survey · Medium. Based on 5 studies from the Atlas of the Middle: 0 rated high strength, 4 medium, 1 low. Designs: survey (4), mixed (1).
In the studies we mapped, they shaped whether managers backed, ignored or resisted a change, and none of the six studies here is rated high strength.
The finding In two papers, each based on three years of fieldwork, emotions tied to managers' group identities, and to how they judged their leaders' legitimacy, saw them back an initiative, quietly brush it off, or move toward resistance (Huy, 2011; Huy et al., 2014).
Two things come up in the studies we mapped: backing from their own bosses, and leaders who attend to relationships as well as tasks.
The finding At a large trucking company, a randomized empowerment program (38 treated and 30 control managers) raised unit managers' sense of control only where they felt their own bosses backed them (Logan & Ganster, 2007).
Study: Logan & Ganster (2007) · experiment · Medium. Based on 6 studies from the Atlas of the Middle: 0 rated high strength, 4 medium, 2 low. Designs: mixed (2), survey (2), experiment (1), conceptual (1).
The evidence is thin: four of the five studies here are rated low strength.
The finding In one small year-long study at a Norwegian company the authors describe as a Fortune 500 high-tech firm, executives given outside coaching reported more satisfied needs at work than uncoached executives, and middle managers in a coaching-based leadership program reported more than uncoached managers, by a smaller margin (Moen & Federici, 2012).
Study: Moen & Federici (2012) · experiment · Low. Based on 5 studies from the Atlas of the Middle: 0 rated high strength, 1 medium, 4 low. Designs: experiment (2), qualitative case (2), conceptual (1).
Yes, they appear to in the studies we mapped, though most of the evidence is low strength.
The finding Among 89 mid-level managers at one U.S. urban hospital, those who spanned boundaries were more strategically active (Pappas & Wooldridge, 2007).
Study: Pappas & Wooldridge (2007) · survey · Low. Based on 6 studies from the Atlas of the Middle: 0 rated high strength, 2 medium, 4 low. Designs: survey (2), qualitative case (2), qualitative interviews (1), conceptual (1).
Invest in middle managers more than in trying to keep them: that is where the studies we mapped point, though none tests cause.
The finding In 463 firms operating in China, HR practices aimed at middle managers' performance went with better firm results through managers' emotional attachment to the firm, while practices aimed at keeping them did not (Gong et al., 2009).
Study: Gong et al. (2009) · survey · Medium. Based on 5 studies from the Atlas of the Middle: 0 rated high strength, 1 medium, 4 low. Designs: survey (2), conceptual (1), qualitative interviews (1), qualitative case (1).
Partly: in the studies we mapped, managers shape culture but cannot simply reset it.
The finding At 11 plants in nine industries, each successful plant manager actively shaped how employees saw the organization and its values (Smith et al., 2010).
Study: Smith et al. (2010) · qualitative interviews · Low. Based on 5 studies from the Atlas of the Middle: 0 rated high strength, 2 medium, 3 low. Designs: qualitative case (2), qualitative interviews (1), archival (1), survey (1).
Transformational leadership comes out best in the studies we mapped, but every study here is a low- or medium-strength survey showing association.
The finding Using 360-degree ratings of 367 managers from 38 UK manufacturers, 215 of them with an agreed hierarchical level, one study found transformational leadership effective at every level, transactional leadership only lower down, and hands-off leadership nowhere (Edwards & Gill, 2012).
Study: Edwards & Gill (2012) · survey · Low. Based on 6 studies from the Atlas of the Middle: 0 rated high strength, 1 medium, 5 low. Designs: survey (6).
Every summary on this page was written by Unfrozen from the study's abstract or full text and checked twice against the source. Loop coding is exploratory and provisional until an independent human coder checks it. Spot an error? Email [email protected] and it gets fixed.