Investment in digital infrastructure can improve employment outcomes? Evidence from Türkiye

Do Investments in Digital InfrastructureImprove Employment Outcomes? Evidence from Türkiye Banu DemirArti Grover Policy Research Working Paper11314 Abstract This paper examines the impact of improvements in digitalinfrastructure on labor market performance, focusing onemployment and productivity, measured by average wages.The empirical setting exploits the staggered expansion ofhigh-speed fiber broadband across provinces in Türkiye,using linked employer-employee administrative data andcomplementary Labor Force Surveys. Across specifica- Detailed occupational data reveal that these effects aredriven by within-province switches from non-teleworkableto teleworkable jobs, consistent with the relaxation ofmobility constraints and the expansion of work-from-homeopportunities as a key channel. Wage gains are concentratedamong high-skilled workers, although employment effectsalso extend to lower-skilled women in teleworkable roles. Do Investments in Digital Infrastructure ImproveEmployment Outcomes? Evidence from Türkiye* Banu Demir Keywords:digital infrastructure; jobs; gender, high-speed Internet; work-from-homeJEL Codes:J16, J82, L96, O14, O33 1Introduction Digital infrastructure plays a pivotal role in driving economic growth, creating jobs, andlifting populations out of poverty. It serves as a catalyst for economic development. Forhouseholds and individuals, digital technology improves access to timely informationand reduces transaction costs, increases educational outcomes, participation in the laborforce, income, consumption, and welfare (Jensen, 2007; Aker and Mbiti, 2010; Hjort and To bridge this gap, our paper investigates how investments in digital infrastructureshape labor market outcomes, with particular attention to employment opportunities and productivity, as measured by average wages.Using spatial and temporal variation infiber optic roll-out in Türkiye during the 2010s and combining this information with richadministrative linked employer-employee data, we show three main results.First, im- and increased job flexibility, alongside demand-side productivity gains. Thus, we showthat the high-speed Internet increases relative female employment and wages, thus nar- We conduct several robustness checks to test the sensitivity of our results on improvedlabor market outcomes for women.First, we include an interaction term between thefiber network and the intensity of female employment to address the potential selectionof more women into teleworkable occupations. The estimated coefficient on the interac-tion with teleworkability remains unchanged, suggesting that the relative increase in fe- Considering the transmission mechanisms underlying the impact of digital connectiv-ity we find that high-speed Internet availability, encourages women to switch occupationswithin a province, particularly from non-teleworkable to teleworkable jobs. The effectson labor market outcomes for females are more pronounced when the move is between able occupations. Thus, women in high-skill occupations also experience relative wage Lastly, we compare the effects of digital infrastructure investments with investmentsin road infrastructure in reducing gender disparities.We exploit the large-scale publicinvestment in roads in Türkiye during the 2000s, which aimed to improve safety andtravel times across the country (Cosar et al., 2021). Our findings suggest that unlike the We contribute to the vast and growing literature investigating the effects of Internet ac-cess on various economic outcomes Hjort and Tian (2021). These include employment cre-ation Hjort and Poulsen (2019a), skill composition of employment Akerman et al. (2015), and household earnings Zuo (2021). For businesses, digital technology can improve deci-sion making, increase efficiency, facilitate innovation, and expand markets (Bloom et al., From a broader perspective, investments in digital infrastructure can support policy-makers’ overarching objectives of job creation and economic development, with poten-tial implications for reducing gender disparities where connectivity enables occupational rity—may enhance these benefits (Grover, 2025). In addition, our research suggests thatto harness the full potential of digital connectivity forinclusiveeconomic growth, policy-makers could prioritize facilitating flexible work-from-home arrangements through, for The remainder of the paper proceeds as follows. Section 2 describes the data sourcesand the background of fiber cable roll-out in Türkiye.Section 3 sets up the empirical 2Data Our analysis combines spatially disaggregated data on investment in fiber cable with two 2.1Micro-level Data on Employment and Wages For most of our analysis, we use linked employer-employee data, an administrative datasetcollected by the Social Security Institution. It covers the universe of formal workers andallows tracking workers over time and across establishments (plants) over the 2012-2019 Activites economiqu