Does working from home (WFH) or hybrid work causally reduce employee innovation, and if so, through which mechanisms and under which organizational designs? The only firm-level study with directly measured innovation output — an event study of over 48,000 employees at a large Indian IT-services firm — found that idea quality fell during full WFH and idea quantity fell during hybrid work, with the largest hybrid losses in teams whose office attendance was poorly coordinated; but its observational design cannot rule out confounding or reverse causation. Both landmark randomized trials of hybrid work measure retention, satisfaction, and performance ratings, not innovation. The open problem is to establish, with credible causal variation, how work arrangements and within-hybrid team co-attendance coordination affect the quantity and quality of employee-generated ideas, to disentangle the mechanisms (reduced collaborative feedback, lower refinement effort, loss of weak ties and network bridging, and selection in the idea-review process), and to test whether interventions such as synchronized office days restore innovation under hybrid regimes.

Background

Three terms frame the problem. Work from home (WFH) means performing one's job away from the employer's premises; hybrid work splits the week between home and office (WFO). An employee suggestion system (also called an idea-management system) is a firm infrastructure through which employees submit improvement ideas that are reviewed and rewarded; it yields rare, directly measured innovation outcomes — the number of submitted ideas (quantity) and independently rated idea scores (quality) — rather than survey self-reports. Co-attendance denotes teammates being physically present at the office at the same time, which hybrid schedules can coordinate well or badly.

After the Covid-19 pandemic, firms worried that remote and hybrid arrangements would damage innovation, the intangible output that depends on unplanned interaction. The evidence base is thin and fragmented. Gibbs, Mengel and Siemroth (Journal of Political Economy Microeconomics, 2023) showed, using personnel and analytics data at a large IT-services firm, that productivity fell during full WFH, with more time in coordination activities and less uninterrupted focus time. The same authors' follow-up (Scientific Reports, 2024) — the source of this problem — used an event-study design around the firm's transitions from office work to full WFH (2020) and then to a hybrid regime (2022), tracking idea submissions of over 48,000 employees. Idea quantity did not change under WFH but idea quality fell; under hybrid, quantity fell — by about 22% in raw submission rates (0.009 vs. 0.007 ideas per employee-month) and, in adjusted regression estimates, to roughly one-third of the WFO level — without any compensating quality gain, and the decline was concentrated in teams whose members' office days were poorly synchronized. The design is observational: the work-mode transitions coincided with the pandemic and its aftermath, and although the analysis controls for many employee, team, and time-varying characteristics, the authors themselves caution that causality cannot be inferred — for example, teams may have reduced co-attendance in response to changing work content rather than the reverse.

Adjacent strands corroborate parts of the story without settling the causal question. Yang et al. (Nature Human Behaviour, 2022) found that remote work at Microsoft made collaboration networks more static and siloed, with fewer bridging ties; weak ties — casual acquaintanceship links that carry novel information across groups, in the sense of Granovetter — are considered especially important for innovation, and their erosion under remote work is a leading candidate mechanism. Brucks and Levav (Nature, 2022) showed in laboratory experiments that videoconferencing narrows visual attention and curbs creative idea generation relative to face-to-face interaction. On the policy side, the two randomized field experiments of hybrid work — Bloom, Han and Liang (Nature, 2024), a six-month trial at a Chinese technology company, and Choudhury, Makridis, Khanna and Schirmann (Review of Economics and Statistics, 2026) — find that hybrid schedules improve retention and job satisfaction without damaging performance grades or ratings; neither experiment measures idea generation or idea quality, so the innovation-specific question is untouched. Survey-based follow-ups (e.g. Crowley et al., International Journal of Innovation Management, 2026, using the European Working Conditions Survey) associate physical proximity with self-reported involvement in idea generation, but cannot identify causation. A 2025 working paper by the same authors of the source study ('Innovator Networks within the Firm and the Quality of Innovation') shows that remote work is associated with lower network degree and less bridging among 28,000 innovators, which 'may explain' the quality decline — evidence that is still correlational.

What remains missing is threefold: (i) causal identification of work-mode and co-attendance effects on measured innovation output, beyond one firm's observational event study; (ii) disentangling which mechanism — reduced collaborative feedback, lower refinement effort, loss of weak ties and network bridging, or selection in the review process — drives the losses; and (iii) any field-experimental test of the organizational remedies firms are deploying in practice (mandated synchronized office days, team-level attendance scheduling, structured virtual interaction channels).

Problem Statement

Establish, with credible causal variation, how work arrangements affect employee innovation output, and whether organizational coordination can prevent the losses. The problem has three jointly linked components serving this single objective: (1) estimate the causal effects of full office work, full remote work, and hybrid work — including within-hybrid team co-attendance coordination (how synchronized teammates' office days are) — on the quantity and independently rated quality of employee-generated ideas, in settings with directly measured innovation outcomes such as firm suggestion systems, going beyond the single-firm observational event study that currently motivates the question; (2) identify which mechanisms drive any estimated effects — the leading candidates being reduced collaborative feedback, lower refinement effort, loss of weak ties and network bridging, and selection in the idea-review process, though other mechanisms are not excluded — rather than leaving them as competing speculations; and (3) test whether coordination-focused interventions such as mandated synchronized office days, team-level attendance scheduling, or structured virtual interaction channels restore idea quantity and quality under hybrid or remote regimes. The question is general across firm types, occupations, and countries, not restricted to IT services; the single-firm Indian evidence is the motivating instance, not the target population. A resolution need not cover every setting: component (1) is met by at least one further, independently identified setting — a different firm type, occupation, or country with directly measured innovation output — that supports the generality claim, and a single such new-setting estimate does not by itself resolve components (2) and (3). Component (3) requires an actual test of a coordination practice, not merely a design proposal.

References

  1. Gibbs, Michael; Mengel, Friederike; Siemroth, Christoph (2024). Employee innovation during office work, work from home and hybrid work. Scientific Reports 14:17117. DOI 10.1038/s41598-024-67122-6. https://doi.org/10.1038/s41598-024-67122-6
  2. Gibbs, Michael; Mengel, Friederike; Siemroth, Christoph (2023). Work from Home and Productivity: Evidence from Personnel and Analytics Data on Information Technology Professionals. Journal of Political Economy Microeconomics 1(1):7-41. DOI 10.1086/721803. https://doi.org/10.1086/721803
  3. Gibbs, Michael; Mengel, Friederike; Siemroth, Christoph (2025). Innovator Networks within the Firm and the Quality of Innovation. SSRN working paper. DOI 10.2139/ssrn.5326116. https://doi.org/10.2139/ssrn.5326116
  4. Bloom, Nicholas; Han, Ruobing; Liang, James (2024). Hybrid working from home improves retention without damaging performance. Nature 630:920-925. DOI 10.1038/s41586-024-07500-2. https://doi.org/10.1038/s41586-024-07500-2
  5. Choudhury, Prithwiraj; Makridis, Christos A.; Khanna, Tarun; Schirmann, Kyle (2026). Is Hybrid Work the Best of Both Worlds? Evidence from a Field Experiment. Review of Economics and Statistics 108(4):1134-1140. DOI 10.1162/rest_a_01428. https://doi.org/10.1162/rest_a_01428
  6. Yang, Longqi; Holtz, David; Jaffe, Sonia; Suri, Siddharth; Sinha, Shilpi; Weston, Jeffrey; Joyce, Connor; Shah, Neha; Sherman, Kevin; Hecht, Brent; Teevan, Jaime (2022). The effects of remote work on collaboration among information workers. Nature Human Behaviour 6:43-54. DOI 10.1038/s41562-021-01196-4. https://doi.org/10.1038/s41562-021-01196-4
  7. Brucks, Melanie S.; Levav, Jonathan (2022). Virtual communication curbs creative idea generation. Nature 605:108-112. DOI 10.1038/s41586-022-04643-y. https://doi.org/10.1038/s41586-022-04643-y
  8. Kim, Min Han; Kim, Kyungdong; Park, Jeongmin (2026). Does hybrid work heterogeneously affect innovative work behavior? A null result and permutation-based safeguards for uplift modeling. European Journal of Innovation Management (online first, 1-27). DOI 10.1108/ejim-11-2025-1607. https://doi.org/10.1108/ejim-11-2025-1607
  9. Crowley, Frank; Doran, Justin; McCarthy, Nóirín; O' Connor, Mari (2026). The impact of working from home on employee innovative work behaviour. International Journal of Innovation Management (online first). DOI 10.1142/s1363919626500283. https://doi.org/10.1142/s1363919626500283
  10. Ding, Yuye; Ma, Mark (Shuai) (2024). Return-to-Office Mandates. SSRN working paper. DOI 10.2139/ssrn.4675401. https://doi.org/10.2139/ssrn.4675401