The Commonplace
Home Papers Evidence Explore Trends Syntheses Digests References Docs 🎲 Workforce Futures
← Papers
Direction, evidence grade, and study type are AI-generated labels (gpt-5-mini), not human-verified. Syntheses are LLM-written. "Tensions" are machine-detected candidates, not confirmed contradictions. A research-acceleration tool, not peer review. How this is built →

AI adoption in Turkey’s logistics sector has cut informal work — informal employment fell from 27.6% in 2019 to 19.2% in 2023 — but has not delivered broad productivity or employment gains; wage rises are concentrated among male technical, AI-complementary roles while unemployment remains high and gender gaps widen.

The Impact of Artificial Intelligence on Workforce Displacement and Transformation in the Logistics Sector: The Case of Türkiye
Mustafa Ergün · December 15, 2025 · İnsan ve Toplum Bilimleri Araştırmaları Dergisi
openalex descriptive low evidence 7/10 relevance Full text usable extracted full text DOI Source PDF

Structured author observations

Linked only from stored provider relations; the raw author line above is never matched by name.

OpenAlex

Latest observation:

  1. Mustafa Ergün exact ORCID
In Turkey's logistics sector (2015–2025), AI-intensive digital adoption coincided with falling informal employment and concentrated wage gains for male, technical/AI-complementary roles, while productivity improvements were modest and structural unemployment and gender wage gaps persisted.

Citation observations

Cumulative provider counts captured on specific dates; providers are never combined.

The rapid integration of artificial intelligence (AI) technologies into the logistics sector is triggering a radical transformation process in the sector, particularly by replacing human labor through automation in operational tasks that require routine and medium skills. This study empirically examines the dual dynamics of workforce displacement and transformation in the Turkish logistics sector, focusing on the period 2015–2025. The main hypothesis of the study is that the use of AI deepens the productivity-employment paradox, encourages skill-based wage increases, and contributes to persistent labor market inequalities. Longitudinal labor market data disaggregated by gender and occupational groups were used in the analysis process, and four key employment indicators were evaluated through descriptive statistics and trend analyses: informal employment, unemployment rates, wage increases, and output per worker (productivity). The findings reveal a structural decrease in informal employment rates (e.g., a decrease from $27.6 in 2019 to $19.2 in 2023); However, it shows that productivity gains remain limited during periods of AI-intensive applications, and structural unemployment persists. Wage increases are observed to be concentrated in technical and AI-complementary roles, which are predominantly male-dominated. The most striking finding is that AI-enabled digital innovations have not translated into inclusive employment growth, as high unemployment rates persist and gender wage gaps widen. These results support the "productivity-employment paradox" and the skills-based technological change (SBTC) thesis. The study points to the urgency of reskilling strategies for low- and medium-skilled workers and emphasizes the need for equity-based policy initiatives to prevent the digital transformation process from reproducing socioeconomic inequalities.

Summary

Main Finding

AI adoption in Türkiye’s logistics sector (2015–2025) has produced a clear dual dynamic: automation has displaced routine and medium‑skill operational jobs while simultaneously creating and rewarding AI‑complementary technical roles. The net outcome is limited productivity gains during AI‑intensive periods coupled with persistent structural unemployment and widening gendered wage gaps. The results support the productivity–employment paradox and the skill‑biased technological change (SBTC) thesis.

Key Points

  • AI substitutes routine warehousing, sorting, forklift and administrative tasks, raising displacement risk for mid‑skill workers.
  • Concurrently, demand and wage premiums rise for AI‑complementary occupations (systems operators, data analysts, robotics maintenance), which are predominantly male.
  • Informal employment rates fell substantially (example reported: from 27.6% in 2019 to 19.2% in 2023), but formalization has not translated into inclusive job growth.
  • Productivity (output per worker, used as an AI proxy) shows limited gains during AI‑intensive years; productivity increases did not proportionally reduce unemployment.
  • Wage growth is concentrated among technical roles; lower and medium‑skill workers saw little real wage improvement, increasing income polarization and gender wage gaps.
  • Findings align with SBTC and the productivity‑employment paradox: technology complements skilled labor while substituting routine tasks, reinforcing inequality.
  • Policy urgency: targeted reskilling/upskilling for low/medium‑skill workers and equity‑focused interventions are needed to prevent reproduction of socioeconomic inequalities.

Data & Methods

  • Design: Quantitative descriptive‑analytical, longitudinal trend analysis (primary focus period 2015–2025).
  • Data sources: ILO Modelled Estimates and Projections; Türkiye Household Labour Force Survey (HLFS). Dataset includes ~820,000 records spanning 2010–2024 (analysis emphasizes 2015–2025 trends).
  • Sectoral scope: Combined NACE Rev.2 sections G & H (Trade, Transport and Storage) used as the logistics proxy due to data availability.
  • Indicators (mapped to SDG markers):
    • SDG 8.2.1 — Annual growth rate of output per worker (%) used as a high‑level proxy for AI/automation adoption and productivity.
    • SDG 8.5.2 — Unemployment rate (%) to track displacement.
    • SDG 8.5.1 — Average hourly earnings to assess SBTC / wage polarization (disaggregated by occupation and gender).
    • SDG 8.3.1 — Informal employment rate (%) to measure vulnerability and inclusion.
  • Analytical approach: Descriptive statistics and trend analyses disaggregated by gender and occupational groups to test hypotheses derived from SBTC and routine‑biased technological change frameworks.
  • Limitations acknowledged by the authors: lack of firm‑level AI adoption/robotics data (necessitating use of productivity as a proxy), aggregation of trade and logistics in sectoral classification, and inability of descriptive trends to establish causal inference.

Implications for AI Economics

  • Empirical support for SBTC and productivity‑employment paradox in an emerging economy logistics context: AI can raise productivity for capital/skill‑owners without automatic job creation for displaced routine workers.
  • Importance of accounting for informality: models and policy analyses that omit informal employment will understate labor market vulnerability in emerging economies and misjudge distributional outcomes of automation.
  • Measurement recommendation: relying on output per worker as an AI proxy is practical but imperfect—future work should prioritize firm‑level measures (robot density, AI software adoption, capital expenditure) to identify causal channels.
  • Policy levers implied:
    • Invest in widespread, accessible reskilling/upskilling focused on technical and digital literacies for low/medium‑skilled logistics workers.
    • Design gender‑sensitive training and hiring incentives to avoid exacerbating male domination of high‑paid AI roles.
    • Promote formalization pathways and social protection for displaced informal workers (to improve take‑up of training and buffer transitions).
    • Encourage technology deployment incentives conditioned on workforce development and inclusive hiring.
  • Research agenda: prioritize causal analyses using microdata on firm technology adoption; evaluate the effectiveness and cost‑benefit of reskilling programs; incorporate heterogeneity (gender, informality, firm size) into structural models of technological change; comparative studies across emerging economies to generalize findings.

Assessment

Paper Typedescriptive Evidence Strengthlow — The study relies on descriptive longitudinal trends without causal identification (no counterfactual, instrumental variables, difference-in-differences, or other strategies). AI intensity is not measured with a clear exogenous source of variation, so observed co-movements may reflect confounding macro, policy, or sectoral shocks rather than causal effects of AI. Methods Rigorlow — Analysis appears limited to descriptive statistics and trend analysis of four indicators; there is no discussion of robustness checks, formal econometric controls for confounders, or micro-level causal inference methods. This constrains internal validity and the ability to rule out alternative explanations. SampleLongitudinal labor-market data for the Turkish logistics sector covering 2015–2025, disaggregated by gender and occupational group; indicators analyzed include informal employment rates, unemployment rates, wage changes, and output per worker (productivity). (Data source, sample size, and measurement details not specified.) Themeslabor_markets productivity inequality skills_training GeneralizabilitySingle country (Turkey) — results may not generalize to other institutional or labor-market contexts, Single sector (logistics) — sector-specific automation dynamics limit applicability to other industries, Descriptive, observational design — inability to isolate causal effects reduces external validity, AI intensity not directly or consistently measured — findings may conflate AI with other technologies or broader automation, Potential aggregation bias — sectoral averages may mask firm-level or regional heterogeneity, Time-bound (2015–2025) — specific technologies and labor market conditions may change rapidly

Claims (10)

ClaimDirectionOutcomeConfidence & EvidenceDetails
AI technologies in the logistics sector are replacing human labor in operational tasks that require routine and medium skills. Job Displacement negative replacement of human labor in routine and medium-skill operational tasks
Reading fidelity high
Study strength medium
not reported
0.18
The study uses longitudinal labor market data (2015–2025) disaggregated by gender and occupational groups and evaluates four employment indicators via descriptive statistics and trend analyses. Other null_result data and analytic approach (longitudinal, gender- and occupation-disaggregated)
Reading fidelity high
Study strength high
not reported
0.3
Informal employment rates in the Turkish logistics sector decreased structurally (e.g., from 27.6 in 2019 to 19.2 in 2023). Employment positive informal employment rate
Reading fidelity high
Study strength medium
decrease from 27.6 in 2019 to 19.2 in 2023
0.18
Productivity gains remain limited during periods of AI-intensive applications in the logistics sector (output per worker shows limited improvement). Firm Productivity null_result output per worker (productivity)
Reading fidelity high
Study strength medium
not reported
0.18
Structural (persistent) unemployment persists in the Turkish logistics sector despite AI adoption. Employment negative unemployment rate
Reading fidelity high
Study strength medium
not reported
0.18
Wage increases are concentrated in technical and AI-complementary roles, which are predominantly male-dominated. Wages mixed wage growth by occupation and gender
Reading fidelity high
Study strength medium
not reported
0.18
AI-enabled digital innovations have not translated into inclusive employment growth; high unemployment rates persist and gender wage gaps widen. Inequality negative inclusive employment growth; unemployment; gender wage gap
Reading fidelity high
Study strength medium
not reported
0.18
The empirical results support the productivity-employment paradox (productivity gains not accompanied by employment growth) and the skill-biased technological change (SBTC) thesis (skill-based wage increases). Research Productivity positive support for theoretical frameworks (productivity-employment paradox and SBTC)
Reading fidelity high
Study strength medium
not reported
0.18
There is an urgent need for reskilling strategies for low- and medium-skilled workers to address AI-driven labor market changes. Training Effectiveness positive need for reskilling strategies
Reading fidelity high
Study strength speculative
not reported
0.03
Equity-based policy initiatives are needed to prevent the digital transformation from reproducing socioeconomic inequalities. Governance And Regulation positive call for equity-based policy initiatives
Reading fidelity high
Study strength speculative
not reported
0.03

Notes