Skip to main content

Table 4 Pseudo-panel FE estimations: subsamples

From: Does ICT affect the demand for vocationally educated workers?

 

(1)

(2)

(3)

(4)

(5)

(6)

Untrained

VET

High school

PET

UAS

University

Full sample

 ICT

− 0.0983***

(0.0252)

− 0.188***

(0.0307)

0.0299**

(0.0126)

0.0569**

(0.0221)

0.120***

(0.0178)

0.0797***

(0.0254)

 N

48386

48386

48386

48386

48386

48386

Subsamples by sector

      

Manufacturing sector

 ICT

− 0.153**

(0.0683)

− 0.264***

(0.0543)

0.00239

(0.0175)

0.0821***

(0.0301)

0.203***

(0.0356)

0.129***

(0.0287)

 N

22473

22473

22473

22473

22473

22473

Service sector

 ICT

− 0.0890***

(0.0251)

− 0.155***

(0.0344)

0.0386***

(0.0148)

0.0521*

(0.0267)

0.0914***

(0.0204)

0.0614**

(0.0310)

 N

26538

26538

26538

26538

26538

26538

Subsamples by firm size

Small-sized firms

 ICT

− 0.0828***

(0.0296)

− 0.189***

(0.0369)

0.0247**

(0.0117)

0.0506*

(0.0273)

0.133***

(0.0227)

0.0637**

(0.0306)

 N

34173

34173

34173

34173

34173

34173

Medium-sized firms

 ICT

− 0.137**

(0.0558)

− 0.103*

(0.0608)

0.00771

(0.0157)

0.0974**

(0.0394)

0.0442*

(0.0257)

0.0909*

(0.0516)

 N

10422

10422

10422

10422

10422

10422

Large firms

 ICT

− 0.130

(0.0947)

− 0.283**

(0.111)

0.155

(0.103)

0.0426

(0.0390)

0.0898*

(0.0517)

0.126

(0.0921)

 N

3791

3791

3791

3791

3791

3791

  1. The table shows OLS coefficients of fixed effects estimates of a firm pseudo-panel with cells defined based on MS-region, industry, and firm size. Standard errors in parentheses are clustered at the cell level. N reports the number of pseudo-panel cells that contain at least one firm. The main explanatory variable measures the cell percentage of ICT workers according to ICT tasks (1996–2010) or ICT occupations (2012–2018). Estimates further control for time*MS-region fixed effects and cell averages of total employment, tenure, age, part-time status, and gender. Each column shows the results for a different dependent variable measuring the percentage of workers with a particular education within the cell
  2. *P < 0.10, ** P < 0.05, ***P < 0.01