Digital tools such as BIM and AI do more than boost efficiency — they can alter workplace dynamics to make UK construction more inclusive and improve women's retention by increasing perceptions of respect, support and value.

Main Finding

Digital technologies (e.g., BIM, AI, cloud collaboration, VR, online mentoring) act as active non‑human agents that can reshape socio‑technical networks in UK construction to improve women’s retention by enhancing perceptions and experiences of Respect, Support, and Value (RSV). Using Actor‑Network Theory (ANT), the study shows digitalisation can do more than improve efficiency — it can increase transparency, enable equitable decision‑making, surface harassment/bias, and provide psychologically safe training/mentoring environments that mitigate structural barriers driving women out of the sector.

Key Points

• Problem context
  • Women comprise <15% of the UK construction workforce and only ~2% on‑site; many leave within ~5 years.
  • Persistent retention barriers are structural and cultural (masculine norms, harassment, inflexible work, weak mentoring, pay gaps, maternity penalties).
  • Prior retention research largely focused on social interventions; digitalisation’s role was underexplored.
• Theoretical approach
  • Actor‑Network Theory (ANT) treats digital technologies as non‑human actors that interact with human actors to produce social outcomes.
  • The paper develops a socio‑technical framework linking digital functions to the RSV dimensions (Respect, Support, Value).
• Empirical findings (from qualitative study)
  • Sample: 23 qualitative interviews with women involved in digitally enabled construction projects.
  • Analysis: thematic coding using NVivo 13.
  • Identified retention issues: rigid/inflexible work practices, hegemonic masculinity, bullying/sexual harassment, insufficient training/mentoring, opaque promotion/pay decisions.
  • Digital roles observed:
    • Transparency & accountability (e.g., digital performance management, data analytics) — reduces opaque informal networks, can reduce bias in promotions/pay.
    • Monitoring & early detection (AI analytics) — can flag patterns of harassment or discriminatory behaviours for proactive intervention.
    • Access to training and psychologically safe skill development (VR, e‑learning) — supports confidence and career progression.
    • New mentoring/collaboration channels (online platforms) — widen networks and role models beyond local masculine networks.
    • Work flexibility enabled by cloud tools — supports work–life balance and remote/hybrid roles.
• Contribution
  • First study to explicitly conceptualise and empirically link digitalisation with gender equity/retention in construction through ANT.
  • Offers practical strategies for firms to target RSV via digital interventions.

Data & Methods

• Literature review
  • Database: Scopus search with keywords “women”, “retention”, “construction sector”.
  • Initial yield ~30 papers; refined to 25 after exclusions (e.g., education sector).
  • Synthesised barriers into RSV themes and summarised current (largely non‑digital) retention initiatives and policy context.
• Primary data
  • 23 semi‑structured qualitative interviews with women working on digitally enabled construction projects (UK context).
  • Participants selected to capture experiences where digital tools were in active use.
• Analysis
  • Thematic analysis using NVivo 13 to identify retention barriers and how women interact with digital technologies.
  • Analytical frame: Actor‑Network Theory to map human (women, managers, peers) and non‑human (BIM, AI, platforms) actors and their interactions producing retention outcomes.
• Outputs
  • Development of a socio‑technical framework connecting specific digital functions to RSV mechanisms.
• Limitations noted by authors
  • Qualitative, UK‑focused sample limits generalisability.
  • Digital maturity and practice heterogeneity across firms mean effects will vary by context.
  • Study is exploratory/pioneering; causal estimates and quantitative magnitudes are not provided.

Implications for AI Economics

• Labor market and retention economics
  • AI and other digital tools can change effective labor supply by improving retention of underrepresented workers (women), altering the composition and size of the workforce — relevant for models of labor supply, skill shortages, and human capital investment.
  • Wage transparency enabled by digital systems can reduce information asymmetries, potentially compress gender pay gaps and affect wage bargaining dynamics and compensation models.
• Productivity vs inclusion trade‑offs
  • Digitalisation often pitched for productivity gains; this study shows inclusion gains are a distinct and measurable benefit. Economic evaluations (cost–benefit, ROI) of AI adoption should include retention and diversity externalities, not just efficiency savings.
• AI governance, fairness, and incentives
  • AI systems that detect harassment or bias can create private/public externalities: reduced turnover and litigation risk for firms, but also privacy and surveillance concerns (potentially negative utility effects). Economic policy design should balance incentives for deploying fairness‑oriented AI with regulation on privacy and due process.
  • Procurement and regulatory incentives (e.g., requiring auditability, fairness metrics) can accelerate adoption of inclusive AI, changing industry equilibrium and firm adoption thresholds.
• Empirical and modelling research opportunities
  • Quantify the causal effect of digital interventions (AI monitoring, transparent HR analytics, digital mentoring) on female retention rates using quasi‑experimental or experimental designs.
  • Integrate retention gains into macro or sectoral models of labor shortages (e.g., projection scenarios where digital retention reduces required recruitment).
  • Firm‑level heterogeneity: model adoption decisions in response to expected retention benefits, productivity gains, and compliance costs; explore complementarities between digital capital and human capital (training).
  • Wage dynamics: evaluate how increased transparency and bias detection affects wage dispersion, promotion rates, and gender wage gaps over time.
  • Risk analysis: assess potential negative impacts (surveillance disutility, algorithmic bias) and welfare implications; design mechanisms for auditing and stakeholder participation.
• Policy implications for economists and policymakers
  • When estimating returns to AI/digital investments, include retention and diversity outcomes in benefit streams.
  • Support pilot programmes that test AI tools aimed at detecting harassment/bias and improving transparency, accompanied by rigorous impact evaluation.
  • Design regulatory frameworks that mandate fairness audits, transparency of HR algorithms, and protections against intrusive surveillance while encouraging tools that demonstrably improve RSV outcomes.
  • Invest in digital infrastructure and training (especially for SMEs) to enable sector‑wide inclusion benefits rather than concentrating gains in digitally mature firms.

Suggested next steps for researchers/policymakers interested in AI economics: - Conduct randomized controlled trials or difference‑in‑differences studies of specific digital tools (e.g., AI harassment detection, algorithmic promotion scoring, digital mentoring) and measure retention, promotion, and wage outcomes. - Build structural or agent‑based models incorporating ANT‑like socio‑technical interactions to simulate long‑run labour market impacts of digital adoption on gender composition and productivity. - Cost‑benefit analyses at firm and sector levels that monetize retention and diversity gains from digitalisation alongside traditional efficiency savings.

(Reference: Blay et al., 2026. Exploring Digital’s Role in Retaining Women in Construction. Construction Economics and Building, 26:2. DOI: https://doi.org/10.5130/tej2v289)