Mobile and More Productive? Firm-Level Evidence on the Productivity Effects of Mobile Internet Use

TL;DR: In this article, the authors analyse whether employees' use of mobile internet access improves firms' labour productivity and find that mobile internet use does cause higher labour productivity in German firms, compared to the early stages of diffusion within firms.

Abstract: Mobile internet access allows for flexibility with respect to working time and working place. We analyse whether employees’ use of mobile internet access improves firms’ labour productivity. Our data set contains 2143 German firms and refers to the year 2014, when high-speed mobile internet was still at a relatively early stage of diffusion within firms. The econometric analysis shows that firms’ labour productivity significantly increases with the share of employees with mobile internet access. Our instrumental variables approach reveals that mobile internet use does cause higher labour productivity.

Summary

1 Introduction

• Computers and the internet are well-established working tools.
• The technological prerequisites for mobile internet, which is diffusing rapidly through the economy, are advances in high-speed wireless connections and mobile devices such as laptops, tablets and smartphones.
• Thus, based on their sample of 2143 German firms for the year 2014, the authors can claim that mobile internet access has a causal impact on firms’ labour productivity.
• The authors control for computer use and fixed line internet access as further types of ICT use at the workplace.

2 Related Literature

• To the best of their knowledge, there is no microeconometric study referring to the relationship between mobile internet use by employees and firms’ labour productivity.
• Furthermore, there exist various studies in information systems research and psychology about the implications of communication technologies and especially mobile internet use (i.e. constant connectivity) for individual employees.
• 2 with the strength of relationships varying considerably across countries and industries.
• Based on a cross section of firms from New Zealand collected in 2006, Grimes et al. (2012) find that firms using broadband internet have a 7 to 10 percent higher labour productivity.
• The analysis by Ward and Zheng reveals positive effects of mobile telephone services on growth in China.

3 Estimation Strategy

• In order to analyse the impact of mobile internet access on labour productivity the authors apply a production function framework as a standard approach to empirically analyse the relationship between productivity and technology.
• Personal computers diffused to firms mainly in the 80s and 90s, whereas the internet diffused to firms in the 90s and early 2000s.
• The authors assume that the use of mobile internet is restricted to employees that have used computers and the internet 4 before, and that the share of employees using computers or the internet is not affected by mobile internet access.
• In the first stage estimation, two variables are used as instruments for mobile internet access: the average mobile internet use at the level of 51 industries and the number of years the interviewee owns a smartphone.
• By contrast, it does not directly explain a single firm’s labour productivity.

4 Data and Measures

• The authors analysis is based on the ZEW ICT survey which is a survey of manufacturing and businessrelated services firms that are located in Germany and have at least five employees.
• The data were collected via computer-aided telephone interviews (CATI) based on a sample stratified with respect to industry and firm size.
• To capture mobile internet use, firms were asked about the percentage of employees with mobile internet access.
• Table 4.1 shows descriptive statistics of the sample.
• Since investment is taken in logs, zero investment is replaced by the 10th percentile of investment per employee observed in the corresponding industry multiplied by the number of employees.

5 Results

• Table 5.1 presents results from simple OLS estimations.
• 7 of years the interviewee owns a smartphone.
• The Kleibergen-Paap LM-test is a heteroscedasticity-robust test for underidentification.
• The estimated coefficient of mobile internet use of the second stage estimations of specifications (2) to (3) are still significant, implying that mobile internet use has a causal positive effect on labour productivity.
• The coefficient of their preferred specification (3) of Table 5.2 is, however, 6In an earlier version of this paper, based on data from 2010, the authors could not find any significant causal effect.

6 Conclusions and Future Research

• According to their empirical results, mobile internet use is positively and significantly related with firms’ labour productivity.
• The authors instrumental variables estimates suggest that this relationship is indeed causal.
• There are various avenues for future research.
• First, in order to learn more about how mobile internet can improve labour productivity, it is important to know which resources employees have access to when working remotely.

Figures

Table A.3: Distribution of Firms across Industries

Table 4.1: Summary Statistics: Estimation Sample

Figure A.4: Industry Means of Mobile Internet Use: Estimation Sample

Figure A.3: Share of Employees Using Mobile Internet: Estimation Sample

Figure A.5: Industry Means of the Number of Years the Interviewee owns a Smartphone: Estimation Sample

Figure A.1: Mobile Data Volumes and 3G/4G Users in Germany 2005-2014

Figure A.2: Share of Employees Using (Mobile) Internet/Predominately Work with Computers: Full Sample

Table 5.2: Dependent Variable: Log Labour Productivity - 2SLS Regression - Second Stage

Table A.1: Dependent Variable: Percentage of Employees using Mobile Internet - 2SLS Regression - First Stage of Table 5.2

Table 5.1: Dependent Variable: Log Labour Productivity - OLS Regression

Table A.2: Summary Statistics by Mobile Internet Use of Firms: Estimation Sample

Full text

Dis cus si on Paper No. 15-090
Mobile and More Productive?
Firm-Level Evidence on the Productivity
Effects of Mobile Internet Use
Irene Bertschek and Thomas Niebel

Dis cus si on Paper No. 15-090
Mobile and More Productive?
Firm-Level Evidence on the Productivity
Effects of Mobile Internet Use
Irene Bertschek and Thomas Niebel
Download this ZEW Discussion Paper from our ftp server:
http://ftp.zew.de/pub/zew-docs/dp/dp15090.pdf
Die Dis cus si on Pape rs die nen einer mög lichst schnel len Ver brei tung von
neue ren For schungs arbei ten des ZEW. Die Bei trä ge lie gen in allei ni ger Ver ant wor tung
der Auto ren und stel len nicht not wen di ger wei se die Mei nung des ZEW dar.
Dis cus si on Papers are inten ded to make results of ZEW research prompt ly avai la ble to other
eco no mists in order to encou ra ge dis cus si on and sug gesti ons for revi si ons. The aut hors are sole ly
respon si ble for the con tents which do not neces sa ri ly repre sent the opi ni on of the ZEW.

Mobile and More Productive? Firm-Level Evidence on the
Productivity Effects of Mobile Internet Use
Irene Bertschek
∗
Thomas Niebel
†
This Version December 2015
First Version December 2013
Abstract
Mobile internet access allows for flexibility with respect to working time and working place.
We analyse whether employees’ use of mobile internet access improves firms’ labour produc-
tivity. Our data set contains 2143 German firms and refers to the year 2014, when high-speed
mobile internet was still at a relatively early stage of diffusion within firms. The econometric
analysis shows that firms’ labour productivity significantly increases with the share of em-
ployees with mobile internet access. Our instrumental variables approach reveals that mobile
internet use does cause higher labour productivity.
Keywords: Mobile Internet, Labour Productivity, Firm-Level Data.
JEL Classification Numbers: D22, L20, O33.
∗
Corresponding author: ZEW Mannheim and University of Mannheim, email: bertschek@zew.de, Centre for
European Economic Research (ZEW), ICT Research Department, P.O. Box 103443, 68034 Mannheim, Germany.
†
ZEW Mannheim, niebel@zew.de. For further information on projects of the authors see www.zew.de/staff_
ibe and www.zew.de/staff_tni as well as the ZEW annual report on www.zew.de/en. We thank Wolfgang
Briglauer, Konrad Stahl and the participants at the EARIE conference (Munich) for helpful comments and James
Binfield for research assistance. Financial support by the state of Baden-Württemberg is gratefully acknowledged.

1 Introduction
Computers and the internet are well-established working tools. They have changed workplaces
significantly, contributed to improving labour productivity and changed the demand for employee
skills and qualifications. The technological prerequisites for mobile internet, which is diffusing
rapidly through the economy, are advances in high-speed wireless connections and mobile devices
such as laptops, tablets and smartphones. McKinsey Global Institute (2013) considers mobile
internet as one of twelve disruptive technologies with a very high potential economic impact.
OECD (2012, p.22) motivates the transformation from the information economy to the internet
economy and points out that “Wireless internet connections are the key source of recent internet
growth, increasing rapidly since 2001 and overtaking fixed broadband subscriptions in 2009.” In
Germany, the number of regular high speed mobile internet users increased from 13.6 million in
2008 to 52.6 million in 2014. During the same period, mobile data volume increased even more
rapidly, from 11.5 to 394.8 Petabytes (see Figure A.1).
While the role of information and communication technologies (ICT) in determining labour
productivity is well studied,
1
there is, to the best of our knowledge, no empirical work on the
firm-level productivity effects of mobile internet so far. Why would we expect productivity effects
from mobile internet? One important result from the empirical analysis of ICT is that reduced
communication costs support the decentralisation of organisation, such as the reduction of hier-
archy levels and the implementation of autonomous working teams (see for example Bresnahan
et al., 2002). Mobile internet access can further improve information flows and communication
and reduce involved costs. Employees are now able to access their firms’ data and documents
anywhere, at any time. This supports decentralisation in terms of organisation and time. By
contrast, coordination costs might increase if physical meetings become more difficult to arrange
since everybody wants to be flexible. Moreover, monitoring might become more difficult if em-
ployees work geographically dispersed. Thus, the net contribution of mobile internet is a priori
not evident.
In our analysis, we take a firm-level perspective in order to analyse the role that employees’
mobile internet access plays for firms’ labour productivity. Based on a sample of 2143 firms
from the German manufacturing and services industry, we estimate classical production functions.
Mobile internet use as an input factor is measured by the percentage share of employees with
mobile internet access in each firm. We control for ICT use at the workplace other than mobile
internet access by including measures of the use of computers and access to fixed line internet.
Since the estimates of the effect of mobile internet access might be prone to reverse causality, i.e.
more productive firms have more resources to invest in new technologies, we apply an instrumental
variables approach. We instrument employees’ mobile internet use within firms by the average
mobile internet use of 51 industries and by the number of years the interviewee owns a smartphone.
1
See for instance the literature reviews by Draca et al. (2007), Van Reenen et al. (2010), Bertschek (2012), and
Cardona et al. (2013).
1

Our initial econometric analysis shows that a one percentage point higher share of employees
with mobile internet access is associated with a 0.2 percent higher labour productivity. When
controlling for potential endogeneity bias by instrumental variables estimation, we still find a
significant effect of mobile internet use on firm productivity. Thus, based on our sample of 2143
German firms for the year 2014, we can claim that mobile internet access has a causal impact on
firms’ labour productivity.
Our paper contributes to the literature in various respects: (i) We provide first microecono-
metric firm-level evidence on the labour productivity effects of employees’ mobile internet use.
(ii) We control for computer use and fixed line internet access as further types of ICT use at
the workplace. (iii) By applying an instrumental variables approach we take account of potential
reverse causality between labour productivity and mobile internet use.
2 Related Literature
To the best of our knowledge, there is no microeconometric study referring to the relationship
between mobile internet use by employees and firms’ labour productivity. There is, however, a
quite extensive literature on the economic impact of mobile phones for small and micro enterprises
in developing and emerging countries (e.g. Jensen 2007, Muto and Yamano 2009, Aker and Mbiti
2010, Tadesse and Bahiigwa 2015 and Paunov and Rollo 2016). In general, this literature suggests
that the use of mobile phones improves market outcomes. Furthermore, there exist various studies
in information systems research and psychology about the implications of communication tech-
nologies and especially mobile internet use (i.e. constant connectivity) for individual employees.
Middleton and Cukier (2006) provide a qualitative analysis on the mobile email usage patterns of
individual employees. Their participants report positive aspects like allowing them to be efficient
as well as negative aspects like the infringement on work-life boundaries. Diaz et al. (2012) con-
clude that the use of communication technologies is associated with increased work satisfaction,
but could also create work-life conflicts. The studies by Mazmanian et al. (2013), Mazmanian
(2013) and Dery et al. (2014) provide further descriptive evidence on the implications of constant
connectivity and always on work practices for employees.
Apart from that, there is a broad literature on productivity effects of ICT in general as well as for
fixed-line broadband internet. According to the survey by Cardona et al. (2013), the estimated
production elasticity of ICT ranges on average between 0.05 and 0.06 and has increased over the
period of observation. Some studies analyse the contribution of the internet to productivity. At the
macro level, Koutroumpis (2009) and Czernich et al. (2011) show that broadband internet has a
positive and statistically significant impact on both productivity and growth in OECD countries.
2
Using a novel data set at the meso-level, Hagsten (forthcoming) finds a positive and significant
relationship between labour productivity and the share of broadband-enabled employees in firms
2
See also the survey by Holt and Jamison (2009).
2

Frequently asked questions

Q1. What have the authors contributed in "Mobile and more productive? firm-level evidence on the productivity effects of mobile internet use" ?

The authors analyse whether employees ’ use of mobile internet access improves firms ’ labour productivity. Their instrumental variables approach reveals that mobile internet use does cause higher labour productivity. 

Q2. What are the future works in "Mobile and more productive? firm-level evidence on the productivity effects of mobile internet use" ?

There are various avenues for future research. Thus, further analysis should take account of workplace models that are flexible with respect to working time and working place and that are supported by mobile devices, such as home office, co-working, working while traveling, etc. Finally, improving work-life balance and creating the possibility to combine work and family i. e. working from home arrangements ) are important assets for firms aiming to acquire and retain highly qualified employees, in particular in times of demographic change. Such a multidimensional concept, however, renders a causal analysis non-trivial, thus posing several challenges for future research. 

Q3. What is the effect of mobile internet access on labour productivity?

Mobile internet access is expected to further decrease communication costs and thereby allows employees to autonomously decide on their most efficient working place and working time. 

Q4. What are the main reasons why the ZEW has been able to change the workplace?

They have changed workplaces significantly, contributed to improving labour productivity and changed the demand for employee skills and qualifications. 

Q5. What is the importance of knowing what resources employees have access to when working remotely?

in order to learn more about how mobile internet can improve labour productivity, it is important to know which resources employees have access to when working remotely. 

Q6. What is the effect of mobile internet on firm productivity?

When controlling for potential endogeneity bias by instrumental variables estimation, the authors still find a significant effect of mobile internet use on firm productivity. 

Q7. How do the authors control for ICT use at the workplace?

The authors control for ICT use at the workplace other than mobile internet access by including measures of the use of computers and access to fixed line internet. 

Q8. What is the effect of broadband on labour productivity?

Based on a cross section of firms from New Zealand collected in 2006, Grimes et al. (2012) find that firms using broadband internet have a 7 to 10 percent higher labour productivity. 

Q9. What is the average production elasticity of ICT?

According to the survey by Cardona et al. (2013), the estimated production elasticity of ICT ranges on average between 0.05 and 0.06 and has increased over the period of observation. 

Q10. What is the coefficient of mobile internet use in the second stage estimations?

The estimated coefficient of mobile internet use of the second stage estimations of specifications (2) to (3) are still significant, implying that mobile internet use has a causal positive effect on labour productivity.