Is Return-to-Office Really Accelerating?

Is Return-to-Office Really Accelerating?

What LY Corporation’s “Three Days a Week” Means in the Five-Year Post-COVID Rebuild of Work

On April 1, 2026, LY Corporation announced a phased shift to an in-office principle of three days per week, timed with the opening of its Akasaka office. The statement framed the move not as tighter surveillance, but as a redesign for higher performance, faster decision-making, and more cross-organization co-creation. In April 2025, the company had still been operating with much lighter attendance expectations in several divisions, so this is a meaningful policy swing.

This should not be reduced to a simple “remote work is over” narrative. What is happening globally is more granular: companies are reclassifying tasks by where they work best, not restoring a single pre-pandemic standard. Amazon moved toward five days in office from 2025, Tesla required approximately forty in-office hours back in 2022, and office mandates at X became part of legal disputes. Yet none of this proves that full-time office work has universally “won.”

---

Reading the Data: Japan Is Seeing Dilution, Not Disappearance

National numbers, private surveys, and sector gaps tell the same story

Japanese national data also does not support the claim that telework has disappeared. Government survey results for fiscal year 2025 show that telework use had fallen after pandemic peaks but then turned upward again. When you track policy adoption and actual utilization over time, levels remain structurally above pre-2020 baselines even after normalization.

Private-sector surveys point in the same direction. The overall implementation rate among regular employees appears relatively flat year over year, but usage frequency has shifted downward. More workers are now in the “less than once a week” or “about once a week” segment. In other words, Japan is seeing “telework remains, but gets thinner,” not “telework is gone.”

Industry structure is critical. Information and communications remains an outlier with far higher remote-capable utilization than national averages, while many physical or site-dependent industries cannot operate by the same model. A single national average therefore hides two different labor markets: one where remote and hybrid still dominate core workflows, and one where physical presence is inherently central.

---

Why RTO Is Advancing: The Four Rationales Companies Repeat

Why are firms increasing office attendance? The key is to read what companies themselves claim. The recurring themes are collaboration quality, faster alignment, mentoring, innovation speed, and cultural cohesion. LY Corporation, Amazon leadership statements, and other executive communications converge on these points even when their exact policies differ.

Counterpoint: Hybrid Is Not Empirically “Low Productivity” by Default

But this does not imply that remote work was inherently unproductive. High-quality evidence from randomized and large-scale studies suggests a more nuanced reality. Hybrid designs can preserve or improve retention and avoid clear performance penalties in many job families. Blanket claims that “hybrid lowers output” are not strongly supported by the best empirical work.

Caveat: Full-Remote-at-Scale Also Has Structural Costs

At the same time, full-remote-at-scale is not costless. Research using large organizational communication datasets found that fully remote settings can make collaboration networks more siloed and less cross-bridging. Other work suggests video-heavy interaction can narrow cognitive exploration during idea generation. The implication is task-specific: individual focus work can thrive remotely, while discovery, boundary-spanning, onboarding, and creative divergence often benefit from in-person density.

---

What High-Scale Implementation Looks Like in Practice

From this angle, the strategic question is not whether to “go back to the old days,” but which moments require intentional co-location. Several major Japanese firms illustrate that different stable designs are possible: telework-first with targeted in-person intensives, high remote participation rates maintained at scale, or explicit hybrid defaults. These are not fringe cases; they show large-enterprise feasibility.

---

Retention Risk and Hiring Competitiveness

A separate issue is whether workers accustomed to flexibility will actually leave when office requirements rise. US attitude data consistently shows a significant share of remote-capable employees view flexibility as a core job attribute, not a perk. That said, attitude data is not identical to realized quits, and Japan still lacks a clean, unified public statistic that directly measures quit rates caused by return-to-office mandates.

So the careful claim for Japan is: preference for flexibility is demonstrably high, but deterministic “X% will resign” statements are not yet evidence-backed at national scale. For employers, that still creates a practical risk channel through hiring competitiveness and mid-career retention, especially in remote-competitive occupations.

---

Household and Care Dynamics: Time Reallocation Is Material

Family and care dynamics matter as well. Japanese policy research has shown that when commuting time declines, a portion of that time reallocates to childcare, housework, and discretionary recovery time. Because unpaid labor remains unevenly distributed by gender, changes in attendance policy can have asymmetric household impacts. For many dual-income families, this is an operational constraint issue, not simply a comfort preference.

---

Remote-Side Risks Also Need Explicit Design

There are also real downsides on the remote side: blurred boundaries, potential overwork, reduced physical activity, and managerial visibility concerns. These are observable operational problems, not merely cultural rhetoric. Any credible hybrid policy must address both sides rather than romanticizing either full remote or full office.

---

Near-Term Outlook: Stabilized Hybrid, Not an Extreme Reversal

Looking ahead, medium-term forecasts from major US survey programs suggest limited additional decline in overall work-from-home share, even amid high-profile RTO headlines. International comparative research similarly indicates stabilization around persistent hybrid patterns rather than collapse to zero remote work.

Japan Will Likely Stay Segmented by Industry and Job Type

The likely Japanese trajectory is comparable but segmented. Sectors with high telework suitability and intense talent competition have weak incentives to drive remote work to zero. Meanwhile, organizations facing heavy needs in security control, hardware-linked development, cross-functional invention, or apprenticeship-style training may continue to increase in-office days.

That is why LY Corporation’s shift is best read as a selective reweighting toward in-person collaboration density, not as abandonment of hybrid logic. The policy signals a design choice about where value is created, not a universal verdict that remote work failed.

Manufacturing Cases Should Not Be Over-Generalized to IT

Cases such as manufacturers returning to in-office norms should also be interpreted in context. What is rational for production-linked or lab-heavy environments does not automatically translate to software, cloud, data, or digital service organizations.

---

Conclusion: The Winners Will Be Firms That Design Work Intentionally

Over the five years since 2020, the pendulum swung from emergency remote expansion to partial office restoration. By 2026, however, the evidence does not support either extreme endpoint: neither a full return to five-day office work everywhere nor permanent full-remote dominance.

The strongest practical conclusion is division of labor by work type. Flexibility supports concentration work, labor-market reach, and retention; in-person presence supports creative friction, cross-team trust formation, and apprenticeship effects. Competitive advantage will come from firms that can design this mix explicitly by role, task, and objective.

In that sense, what is becoming weaker is “default full remote without design” and “default daily office without rationale.” What is becoming stronger is evidence-based work architecture. For workers too, the advantage shifts to those who can articulate which parts of their job improve with co-location and which improve with autonomy.

So can we “go back to pre-COVID”? Not really. But that does not mean remote work simply won. It means office and home have been reassigned new functions in a hybrid operating system. LY Corporation’s three-days-a-week move is a clear marker that this redesign has entered its next phase.