Winners and losers from remote work: How firm capabilities impact productivity

Work from home is now a permanent feature of how firms organise production, yet its effects on productivity remain unclear. Using new data covering all remote workers in Italy, this column shows that the COVID-19 pandemic-driven shift to work from home initially reduced productivity, but these losses disappeared as firms adapted. Larger and digitally prepared firms weathered the transition better, while firms with highly qualified workers achieved productivity gains over time. Overall, the productivity effects of work from home depend on complementary investments in digital technologies, organisational capacity, and skills.

Work from home (WFH) has become a standard arrangement for employees in advanced economies (Aksoy et al. 2022, Barrero et al. 2023, Crescenzi et al. 2026, Hansen et al. 2023). Yet the policy and managerial debate remains unusually polarised, reflecting a sharp divergence between employees’ preferences and employers’ organisational concerns. While studies show that most employees value the flexibility and benefits associated with working from home, such as improved work-life balance and reduced commuting time, some executives are imposing a return to a five-day in-office schedule, citing concerns about collaboration, innovation, and company culture. Firms are now at a crossroads, weighing the potential advantages of flexible work arrangements against perceived challenges in order to decide whether to retain their remote work policies.

This tension is mirrored in the academic literature. The impact of work from home on firm performance and productivity remains unclear, with existing studies finding mixed evidence depending on the context. Some studies find benefits for both firm productivity and worker satisfaction (Angelici and Profeta 2023, Bloom et al. 2015, 2024, Choudhury et al. 2024). Others report negative effects on firm productivity (Gibbs et al. 2023) and on workers’ career progression (Emanuel et al. 2023). This mixed evidence leaves firms and policymakers without clear guidance.

Evidence from firm-level data

Studying the implications of work from home for firm productivity has proved particularly difficult due to several empirical challenges. First, comprehensive data on the adoption of work from home are scarce. Most existing studies rely on surveys or focus on individual companies. In a recent paper (Boeri et al. 2025), we use previously unexplored administrative data for Italy covering the entire population of remote workers. Italian law requires firms to declare workers contractually authorised to work from home to ensure insurance coverage, generating detailed and reliable measures of  WFH adoption at the firm level across the entire economy.

Second, identifying causal effects is challenging because the adoption of work from home is not random. Firms with better organisational capital, digital infrastructure, and managerial capabilities are more likely to adopt work from home, making it difficult to separate the effects of work from home itself from pre-existing firm characteristics. To address this selection problem, we exploit a sharp, pre-pandemic constraint on firms’ ability to adopt work from home: access to fibre broadband. Unlike the previous ADSL technology available in Italy, fibre broadband enables the high-bandwidth, synchronous tasks, and scalability essential for effective work from home, such as videoconferencing, large data transfers, and multiple simultaneous connections.

We instrument firm-level WFH adoption with pre-pandemic local fibre availability, measured at a highly disaggregated geographical level. Leveraging the pandemic as a common shock, we compare outcomes before and after its onset between firms with and without fibre access, operating in the same local labour market and narrow industry.

A potential concern with this approach is that infrastructure deployment is often targeted towards more dynamic firms or growing areas, which could bias our results. However, we show that, at the local level, the fibre rollout in Italy was not influenced by pre-pandemic firm characteristics or local economic trends. This is because we exploit a massive public investment implemented by the Italian government in 2015, the National Ultra-Broadband Plan, under which operators were required to cover the entire Italian population within a short period. This left no scope to target specific areas based on local economic conditions.

Instead, geography played the central role. Conditional on a broader area being served, the choice of which neighbourhood to connect next was driven by cost minimisation, specifically, physical proximity to the existing network, rather than by firm performance or neighbourhood economic characteristics. Differences in fibre availability across neighbourhoods are therefore as good as random, providing a credible source of variation in firms’ ability to adopt work from home when the pandemic hit.

Work from home and firm productivity

We first examine the impact of work from home on firm productivity. We find a large short-term negative effect of  WFH adoption on both labour productivity and total factor productivity growth during 2020 (see Figure 1). However, by 2021 and 2022, the estimated average effect becomes statistically indistinguishable from zero. This suggests that Italian firms adjusted to the new working arrangements over time, likely aided by accumulated experience with remote-work practices and the widespread uptake of digital tools and collaboration software introduced during the pandemic.

Figure 1 Impact of work-from-home adoption on firm productivity (IV results)

Notes: The figure illustrates the effect of  WFH adoption on firm productivity, estimated with instrumental variable regressions. The figure shows coefficient estimates and their 95% confidence intervals.

The role of ICT, size, and skills

We next document systematic heterogeneity across firms, revealing which firm characteristics determine success or failure with work from home.

First, the initial productivity losses were concentrated among firms that combine remote and on-site tasks, such as manufacturing firms. This pattern indicates that the productivity effects of work from home are not simply the sum of individual remote workers’ performance. Rather, they also reflect coordination and communication frictions between remote and in-person workers operating across different functional areas within the firm, for example, between office staff working remotely and production workers on-site.

Second, pre-pandemic investments in information and communication technologies (ICTs), including investments in laptops and server infrastructure, helped firms mitigate these early disruptions (see Figure 2). Firm size also mattered. Larger firms, typically endowed with stronger organisational and managerial capabilities, were better able to absorb the initial shock and did not experience significant productivity losses.

Figure 2 Heterogeneous effects of work-from-home adoption on firm productivity (IV results)

Notes: The figure shows heterogeneous productivity effects of work-from-home adoption, measured by the interaction between work from home and firm characteristics in instrumental variable regressions. Coefficient estimates and 95% confidence intervals are reported. ‘ICT’ denotes the pre-pandemic investments in servers per employee; ‘Size’ the number of employees (in log); and ‘Skills’ the share of employees with a postgraduate degree.

Finally, human capital becomes more salient over time. Firms with a higher share of highly qualified workers, those holding a master’s degree or an MBA, recorded significant productivity gains from work from home by 2022 (Figure 2). This suggests that advanced skills and managerial capabilities facilitate the organisational adjustments required to exploit the productivity benefits of work from home.

Conclusions

This column speaks to a central post-pandemic question: should the work-from-home practices adopted during the pandemic be retained as part of firms’ longer-run production model?

Using population-level firm data for Italy, we show that work from home is neither inherently productivity enhancing nor productivity reducing. Crucially, its impact is highly heterogeneous, depending on complementary investments in ICT, organisational capacity, and human capital. This pattern echoes a broader lesson from the technology adoption literature: new technologies deliver uneven effects across firms (Bloom et al. 2012).

For managers and policymakers, the implication is to avoid one-size-fits-all approaches, whether in the form of blanket return-to-office mandates or unconditional promotion of remote work. A more productive and pragmatic approach is to focus on the constraints that prevent some firms, particularly smaller and less digitally prepared ones, from operating hybrid models effectively and to support capacity-building through investments in digital infrastructure, training in hybrid management practices, and workforce upskilling.

Important questions remain open for future research. In particular, little is known about the long-term implications of work from home for innovation, learning, and corporate culture, all of which are central to firm growth. Understanding whether the capability gaps we document are widening or narrowing over time – and what this means for aggregate productivity and market concentration – also represents an important priority for future work.

Source : VOXeu