A workplace settlement, not a temporary compromise
For a period, arguments about the future of work were reduced to a narrow dispute over attendance: how many days in the office, which teams should come in, and whether flexibility weakens output. That framing now looks inadequate. The more consequential change is that work has entered a prolonged institutional transition. Employers are testing how to coordinate dispersed teams; workers are renegotiating autonomy, time and career progression; and cities, regulators and educators are adjusting to altered patterns of demand.
What is emerging is not a stable endpoint but a new workplace settlement. Digital communication tools made large-scale remote coordination feasible; the pandemic accelerated adoption; tight labour markets in some sectors strengthened worker expectations; and advances in data analytics have given employers more ways to observe, measure and standardise performance. The result is a labour market in which flexibility and surveillance can rise together.
Hybrid work is not chiefly a question of place. It is a test of whether organisations can replace informal supervision with intentional management.
The key managerial challenge is straightforward to describe and difficult to execute: how to preserve the benefits of proximity without treating physical presence as a substitute for leadership. Research from the Stanford Institute for Economic Policy Research suggests that hybrid arrangements can sustain productivity and improve retention when designed carefully, especially where jobs contain measurable outputs and workers have some autonomy. But the same research also underscores that gains are not automatic; the quality of management remains decisive.
This matters because the office used to carry hidden organisational functions. It enabled apprenticeship by osmosis, reinforced status hierarchies, and generated weak ties across teams. As those functions become less incidental, they must be rebuilt deliberately. The future of work is therefore less about choosing between remote and in-person routines than about redesigning the architecture of collaboration.
Productivity will depend on management more than attendance
The evidence on productivity is more nuanced than either enthusiasts or sceptics often admit. Studies differ by occupation, task design and workforce composition. The broad conclusion from recent years is that location alone explains less than many executives assumed. What matters more is whether organisations can clarify goals, document workflows, coordinate handovers and train managers to lead distributed teams.
The OECD has argued that teleworking can support productivity when combined with managerial practices that encourage autonomy, learning and well-being. Yet it also warns of risks: weaker knowledge spillovers, blurred boundaries between work and life, and reduced innovation if social interaction deteriorates. In other words, hybrid work is a complement to good organisational design, not a substitute for it.
This has significant implications for corporate governance. Senior leaders have often treated office mandates as a simple lever because they are visible and easy to enforce. But attendance rules can mask more difficult weaknesses: poorly defined roles, overreliance on synchronous meetings, and a lack of objective performance metrics. A company that cannot explain how work creates value is unlikely to solve the problem by increasing headcounts in meeting rooms.
That is one reason the debate has become so heated. Return-to-office policies are rarely only about space utilisation. They are often proxy battles over managerial control, trust and the distribution of bargaining power. Where leaders lack confidence in measurement systems or middle managers struggle to supervise remotely, physical presence regains appeal. But that should be understood as an organisational symptom, not a strategic answer.
The labour market is being redrawn around flexibility
Flexibility is becoming a sorting mechanism in labour markets. Workers with scarce digital skills or specialist expertise can increasingly bargain for location autonomy. Those in place-dependent work cannot. This is creating a sharper divide between high-agency knowledge workers and employees whose schedules remain tightly constrained by logistics, retail, manufacturing, healthcare or hospitality.
Hybrid work is not chiefly a question of place. It is a test of whether organisations can replace informal supervision with intentional management.
The International Labour Organization has noted that remote-capable work is concentrated in higher-income occupations and among workers with greater educational attainment. That pattern means flexibility can widen inequality even while it improves working conditions for some. One group gains time, geographical freedom and lower commuting costs; another continues to absorb rigid schedules, rising transport burdens and limited control over hours.
Policy discussion often misses this asymmetry. Hybrid work is sometimes presented as a broad social advance, but its distribution is uneven by class, sector, age and geography. Graduates in digital and professional roles have enjoyed the largest gains. Younger workers, meanwhile, may face hidden trade-offs if flexibility reduces mentoring, network formation or visibility. For carers and people with disabilities, remote options can expand participation dramatically, yet only if progression does not become biased towards those more often seen in person.
The future of work, then, is not simply more flexible. It is more segmented. Labour-market institutions will need to decide whether flexibility remains a privilege of particular occupations or becomes part of a wider agenda around predictable scheduling, worker voice and time sovereignty across the economy.
Cities will need a new economic logic
The reorganisation of work is also reshaping urban economies. Central business districts were built around dense weekday commuting patterns that supported office towers, transport systems, cafés, retail and commercial real estate valuations. Hybrid routines weaken that concentration. They do not eliminate the city, but they change its rhythm and economics.
Evidence from research by the McKinsey Global Institute and others suggests that lower office attendance can depress demand for city-centre services while shifting spending towards suburban neighbourhoods. Commercial property markets have already reflected this divergence in several global cities. The adjustment is not merely cyclical. If fewer workers commute five days a week, then transport planning, retail mix and land-use assumptions all need revision.
The real urban effect of hybrid work is not empty offices; it is the unbundling of economic activity that once clustered around the daily commute.
This does not mean dense cities are obsolete. Agglomeration still matters for high-value sectors that depend on talent pools, finance, universities and cultural infrastructure. But it does mean that local leaders can no longer assume office demand will anchor municipal revenues and surrounding services in the same way. Some districts may become more mixed-use; others may struggle with stranded assets and fiscal stress.
There is also a geopolitical dimension. If a larger share of knowledge work can be performed outside traditional hubs, secondary cities and smaller regions may capture talent and spending. Yet such redistribution will favour places with strong digital infrastructure, attractive housing and credible public services. Geography still matters; it is simply being reweighted.
Artificial intelligence will change jobs before it eliminates them
The most significant medium-term force acting on work is not location policy but the spread of artificial intelligence across cognitive tasks. The strongest current evidence suggests that AI will first recompose jobs rather than erase them wholesale. It can draft, classify, summarise, search and assist decision-making, but its impact depends on how tasks are bundled, how errors are managed and who retains accountability.
The IMF, the OECD and the World Economic Forum each argue that a substantial share of jobs in advanced economies are exposed to AI in some form, with highly skilled occupations often more exposed than lower-skilled manual work. Exposure, however, is not the same as replacement. In many cases, AI will augment workers by automating parts of a role while increasing the importance of judgment, client handling, compliance and exception management.
This creates a familiar pattern in labour economics: technology can complement some workers, substitute for others, and alter wage structures depending on skills and institutional protections. Generative AI may raise productivity for experienced employees who can verify outputs and integrate them into workflows. It may also compress entry-level tasks that once served as training grounds for junior staff in law, consulting, marketing or software development.
The real urban effect of hybrid work is not empty offices; it is the unbundling of economic activity that once clustered around the daily commute.
That last point is especially important. Organisations often focus on immediate efficiency gains while underestimating long-term capability risks. If junior workers perform fewer basic tasks because software handles first drafts, where will they acquire tacit knowledge? The future of work may therefore feature a paradox: firms become more efficient in the present while weakening their apprenticeship pipeline for the future.
The middle manager is being reinvented
Few roles are under greater pressure than middle management. For decades, many managers relied on proximity, ad hoc check-ins and visible busyness as cues for coordination. Distributed work and digital systems make those habits less effective. Managers are being asked to do harder things: set clearer priorities, provide better feedback, codify processes, protect attention, and support employees whose needs differ sharply.
Research from Microsoft’s Work Trend Index and academic studies on digital collaboration point to a consistent problem: workers are overwhelmed by fragmented communication, excessive meetings and constant interruption. In such an environment, the manager’s job is no longer to transmit information down a hierarchy. It is to curate focus, remove friction and ensure accountability without generating surveillance fatigue.
This is an underappreciated institutional challenge. Many organisations promote strong individual contributors into management without training them in coaching, conflict resolution or workflow design. Hybrid work exposes that weakness. So does AI, which can automate some reporting and administrative tasks while raising expectations that managers will add human value through judgment and development.
The future manager will be judged less by oversight and more by an ability to create clarity, trust and learning across dispersed teams.
If that transition succeeds, management could become more demanding but also more meaningful. If it fails, organisations may drift towards two undesirable poles: brittle bureaucracy, with every action documented and monitored, or chaotic informality, in which responsibility is vague and burnout rises. Neither is a sustainable model for high-performance work.
Surveillance and measurement are becoming the new frontier of power
As work becomes more digital, the capacity to observe workers expands. Software can track log-ins, keystrokes, message volumes, system activity and, in some contexts, location or biometrics. The technical ability to measure work is advancing faster than the institutional norms governing when such monitoring is legitimate.
The European Commission, trade unions and privacy regulators have increasingly warned that intrusive surveillance can undermine dignity, trust and data protection rights. Yet employers face genuine pressures to improve security, compliance and productivity. The future of work will therefore be shaped by a contest over proportionality: what should be measured, for what purpose, with whose consent, and with what safeguards?
There is a practical reason this matters beyond ethics. Excessive monitoring can distort behaviour. Workers adapt to metrics, often in counterproductive ways. If responsiveness is rewarded, messages proliferate. If screen time is rewarded, presenteeism goes digital. If dashboards become substitutes for judgment, organisations may produce abundant data while learning less about actual performance.
Better measurement is possible, but it requires discipline. The most useful indicators are usually tied to outcomes, quality, risk and team effectiveness rather than crude proxies of activity. That in turn demands a sharper understanding of work itself. In this sense, surveillance debates are really governance debates. They ask whether technology will make organisations more thoughtful about performance or merely more intrusive.
Skills policy is moving from education to continuous adaptation
The future manager will be judged less by oversight and more by an ability to create clarity, trust and learning across dispersed teams.
The future of work is often discussed as if skills were a stock acquired early in life and updated occasionally. That model is becoming obsolete. Longer careers, faster technological change and shifting task composition mean workers will need repeated episodes of reskilling and role transition. The challenge is not simply training more people in advanced tools; it is building institutions that make learning continuous and economically viable.
The World Economic Forum’s recent analyses of jobs and skills highlight that employers expect growing demand for analytical thinking, technological literacy, resilience and leadership, alongside occupation-specific capabilities. But headline skills lists can be misleading if they imply that broad competencies alone are enough. Workers still need routes into sectors, recognised credentials and time to learn without taking untenable income risks.
Public policy has lagged here. Many systems still separate higher education, vocational training and adult learning into disconnected silos. Employers meanwhile underinvest in transferable skills because they fear poaching or focus on short-term returns. The result is a collective-action problem: everyone agrees adaptation matters, but incentives to finance it remain weak.
A more credible settlement would combine portable credentials, modular training, stronger employer participation and targeted public support for mid-career transitions. Without that, technological change will not simply reward the adaptable; it will reward those already positioned to absorb risk.
Worker power is changing form, not disappearing
Predictions that digital work would atomise labour and weaken worker voice have proved only partly correct. Traditional unions remain stronger in some sectors than others, but new forms of coordination are emerging through professional networks, online communities and issue-based organising around pay transparency, surveillance, scheduling and fairness in algorithmic management.
The significance of this shift lies in its form. Worker power is becoming less uniformly tied to a single workplace and more connected to labour-market visibility, skill scarcity and reputation. In some fields, individuals with sought-after expertise can exercise bargaining power simply by switching employers or publicising poor practices. In others, especially platform-mediated and lower-wage work, collective action remains difficult but not impossible, particularly where regulators intervene.
This changing terrain complicates old assumptions about employment relations. Formal flexibility may coexist with weak security. A worker may enjoy geographic autonomy but have little say over performance systems or advancement criteria. Another may be protected by a contract yet subject to algorithmic scheduling that reduces practical control over time. The future of work will therefore turn not only on flexibility but on voice: who gets to shape the rules by which work is organised and evaluated.
What a durable settlement will require
If there is a single lesson from the past few years, it is that work is not merely a technical system for allocating tasks. It is also a social institution that distributes income, status, time, autonomy and belonging. That is why debates about attendance became so emotionally charged: they touched far more than logistics.
A durable settlement will require several shifts. Organisations will need to define performance more clearly and train managers more seriously. Governments will need to modernise labour protections for a world of digital monitoring, AI-assisted work and uneven flexibility. Cities will need to adjust land use and transport for altered commuting patterns. Education systems will need to treat adult learning as core infrastructure rather than a residual policy category.
None of this points to a simple triumph of remote work, nor to a full restoration of the pre-pandemic office. The likelier outcome is a more plural labour market in which different sectors settle on different models, shaped by task requirements, bargaining power and regulation. Knowledge work will continue to disperse in some respects while clustering in others; AI will automate certain functions while elevating the value of judgment and interpersonal skill; and the best organisations will learn that flexibility works only when matched with coherence.
The office is not disappearing. But it is no longer the unquestioned centre of gravity around which work, management and urban life automatically revolve. That distinction matters. Once the office ceases to be the default answer, institutions are forced to confront the harder question: what, precisely, is work for, and how should it be organised to serve both productivity and human flourishing?
