THE IMPACT OF BROADBAND ON JOBS AND THE GERMAN ECONOMY

THE IMPACT OF BROADBAND ON
JOBS AND THE GERMAN ECONOMY

by

Prof. Dr. Raul L. Katz1,
Dr. Stephan Vaterlaus, Patrick Zenhäusern,
Dr. Stephan Suter and Dr. Philippe Mahler2

Abstract
This study quantifies the macroeconomic impact of investment
in broadband technology on employment and output of
Germany’s economy. Building on the “National Broadband
Strategy” announced by the German government, two
sequential investment scenarios (2014 and 2020) are defined.
The economic impact of broadband development over a ten
year period in Germany amounts to 968,000 additional jobs and
EUR 170.9 billion in incremental output.



Keywords
Telecommunications, Broadband, Infrastructure, Growth, Employment,
Input-Output Analysis, German Economics

1 Dr. Raul Katz is Adjunct Professor in the Division of Finance and Economics at Columbia
Business School, Director of Business Strategy Research at the Columbia Institute for Tele-
Information, and President of Telecom Advisory Services, LLC.
2 Dr. Stephan Vaterlaus is President of Polynomics AG. Patrick Zenhäusern, Dr. Stephan Suter and
Dr. Philippe Mahler are also members of Polynomics AG.

This study was funded by Deutsche Telekom AG. The authors are responsible for the views expressed in this
study. The study is being submitted for publication in an academic journal (05/2009).

---

EXECUTIVE SUMMARY

This study quantifies the macroeconomic impact of investment in broadband
technology on employment and output of Germany’s economy. In times of
economic crisis, national governments are looking for immediate policy actions to
deal with rising unemployment and declining output. Infrastructure investments
have been singled out in the economic and policy debate as being tools of
recovery. In this respect, broadband technology is key to providing one of the
needed economic stimuli, not only in terms of construction, but also resulting in
longer term externalities, such as accelerated innovation and new business
creation.

Two sequential investment scenarios are analyzed: the first is based on the
"National Broadband Strategy" announced by the German Government which
aims at ensuring that 75 percent of German households have access to a broadband
connection of at least 50Mbps by 2014. The second scenario (labeled "ultra-
broadband" covering 2015-2020) defines the investment required to provide to 50
percent of households with at least 100 Mbps and another 30 percent with 50
Mbps by 2020.

It is estimated that fulfilling these targets will require an investment of 20.2 billion
Euros up to 2014. This will allow covering all unserved households (730,000),
upgrading households currently having sub-par access to download speeds up to 1
Mbps (2,8 million), provide 50 percent of German households with VDSL service
and 25 percent of households with FTTH. Beyond these short-term objectives, the
deployment of ultra-broadband fiber infrastructures in subsequent years (2015-
2020) will entail making FTTH available to an additional 25 percent of households
(an ambitious but realistic scenario), which will require an incremental 15.7 billion
Euros of infrastructure investments.

These investments will generate both a substantial number of jobs and an
incremental increase of the GDP. Based on input-output tables from the German
Federal Statistical Office, and relying on input-output analysis, the network
construction has been estimated to create 304,000 jobs (or 61,000 per year
between 2010 and 2014), and additional 237,000 jobs (or 40,000 jobs annually
between 2015 and 2020). Further, once the broadband infrastructure is deployed,
additional job creation will be triggered by network externalities, such as enhanced
innovation resulting in new services, additional business growth, and the attraction
of jobs from other countries as a result of a recomposition of industrial value
chains. Based on regression-based forecasting, it is estimated that 427,000 jobs
will be created. As a result, the accumulated number of jobs over a ten year period
(2010-2020) will reach 968,000 in Germany.





2

---

EMPLOYMENT IMPACT FOR GERMANY
(in thousand)


Network Construction
Network
Total
Direct
Indirect
Induced
Total
externalities
2010-14
158
71
75
304
103
407
2015-20
123
55
59
237
324
561
Total
281
126
134
541
427
968

In output terms, the construction of the broadband network will result in 18.8
billion Euros in additional GDP (2010-2014) and 14.6 billion Euros, in the
subsequent phase of ultra-broadband rollout (2015-2020). The cumulative increase
of GDP of 33.4 billion Euros between 2010 and 2020 means that for each Euro
invested in constructing the broadband network, the value added amounts to 0.93
Euros. In addition, the resulting incremental broadband penetration will add 137.5
billion Euros to the GDP.3 Over a ten year period (2010-20), this results in 170.9
billion Euros of additional GDP (0.60 % annual GDP growth) in Germany. Of
these, 108.8 billion Euros will be generated between 2015 and 2020.

GDP IMPACT ON THE GERMAN ECONOMY
(in billion Euros)


Network
Network
Total
Percentage
Construction
Externalities
Change
2010-14
18.8
43.3
62.1
0.49 %
2015-20
14.6
94.2
108.8
0.70 %
Total
33.4
137.5
170.9
0.60 %

Based on these results, it becomes obvious that the National Broadband Strategy
and the subsequent ultra-broadband evolution significantly improve employment
and economic growth in Germany. Broadband represents a high priority stimulus
program the government needs to rely on to improve the current outlook and
create a solid foundation for future growth. Since roll-out of broadband
infrastructure is primarily based on private investment, the Government’s
commitment to a newly designed growth and innovation-geared regulation is
critical. In that sense, the formulation of a framework which allows for appropriate
risk diversification between investors and non-investors as well as improved a-
priori planning and legal certainty for investing companies represents a key
element of the needed investment-friendly environment.



3 This number results from applying the calculated incremental GDP annual growth rate due to
network externalities to Germany's 2008 estimated GDP of 2,489 billion Euros for the years
between 2013 and 2020.

3

---

1. INTRODUCTION

In times of economic crisis, national governments look for policy actions that can
rapidly deal with rising unemployment and declining output. Infrastructure
investments have been identified as key tools in the economic and policy debate to
fight the ongoing crisis. This is partly because of the direct and short-term labor
effects in the building and construction industries, sectors that have been greatly
affected by the economic crisis. On the other hand, construction of network based
infrastructure, such as transportation routes and broadband networks carry
substantial spillovers in terms of improving efficiencies and stimulating innovation
in the production sector of the national economy. It is for these reasons that several
governments (United States, Germany, Australia, Portugal, Singapore, and Ireland,
among others) have decided to increase their investment in infrastructure, such as
the construction of roads, bridges and telecommunications.

In particular, the economic benefits of deployment of broadband infrastructure are
significant. In information-intensive economies, such as Germany where 54% of
the economically active workforce is considered to be information-based (from IT
professionals to content producers to clerks) (Katz, 2009a), the construction of
infrastructure to facilitate the flow of information in the economy can have an
impact on productivity, innovation and business growth. In fact, studies point out
that rankings in national competitiveness and network readiness are directly
correlated. If network readiness is a result of network technology investment, one
could argue that an increase in broadband investment should impact Germany’s
overall competitiveness.

The objective of this study is to estimate the impact of broadband infrastructure
investments on the German economy, in particular on employment and output.
This study analyzes the potential effects of Germany’s "National Broadband
Strategy" which is expected to be completed by 2014 (BMWi 2009a). In addition,
the study assesses the economic impact of a second phase of ultra-broadband
evolution, which results in a more advanced broadband network and is assumed to
be completed by 2020. Along these timelines, the study tackles the costs of
deploying broadband due to the National Broadband Strategy and due to ultra-
broadband evolution in Germany. It estimates the number of jobs created during
the roll-out phase of the National Broadband Strategy and incremental
employment effects thereafter. Furthermore, the economic effects of broadband
deployment are estimated in terms of domestic value added resulting from network
construction and incremental GDP growth.

The study relies on a three-step-approach. First, the total costs for broadband
deployment in Germany by 2014 and 2020 are determined. Second, by relying on
input-output tables generated by the German Statistical Office and utilizing input-
output analysis, the workforce which will be created by the roll-out of modern
broadband is estimated. Third, regression-based forecasting to estimate the
externalities of broadband deployment is used.

4

---

The study begins with an assessment of the research literature on the economic
impact of broadband. Based on the identification of what it is concluded to be the
more rigorous theoretical frameworks and methodologies, it then presents the
methodology and supporting data (see appendices). In the following section, an
estimation of investment required to deploy broadband technology to meet the
targets outlined in the National Broadband Strategy and in the ultra-broadband
evolution long term view is made. With that basis, the estimates of economic
impact, both in terms of jobs and output (value added growth), of implementing
the Strategy are presented in the final section.

2. THE RESEARCH LITERATURE ON THE ECONOMIC IMPACT OF
BROADBAND:

The study of the relationship between broadband and employment creation has
produced few empirically driven pieces of research. Two types of studies have
been conducted so far: a) national cross-sectional research focused on identifying
employment and/or output effects on national economies, and b) regional studies
oriented to the assessment of broadband economic effects at the regional level.
Two methodologies are primarily used in these studies: input-output analysis and
multivariate regression modeling (see figure 1).

Figure 1. Studies of the Employment Impact of Broadband


National economies
Regional Economies
Input-Output
• Crandall et al. (2003)
• Strategic Networks Group
analysis
• Katz et al. (2008)
(2003)
• Atkinson et al. (2009)
• Katz et al. (2009)
Multivariate
• Lehr et al. (2006)
• Kelly (2003)
Regression
• Crandall et al. (2007)
• Ford and Koutsky (2005)
Modeling
• Thompson et al. (2008)

All studies conclude that, while it is difficult to measure and establish causality,
broadband technology contributes to the creation of employment. Furthermore,
some studies differentiate two types of employment impact of broadband:

• First impact: Jobs created to deploy the infrastructure, and
• Second impact: Employment generated as a result of network externalities
on other sectors of the economy. The results of the research to date in these
two areas will be reviewed in turn.

First, it is obvious that network construction will result in some level of job
creation, in terms of direct effects. The four national studies that attempted to
estimate this amount are Crandall et al. (2003), Katz et al. (2008), Atkinson et al.

5

---

(2009), and Katz et al. (2009). They all relied on input-output matrices4 and
assumed an amount for capital investment: US-Dollars 63 billion (needed to reach
ubiquitous broadband service) for Crandall et al. (2003), CHF 13 billion for Katz
et al. (to build a national multi-fiber network for Switzerland), US-Dollars 10
billion for Atkinson et al. (2009) (as a US broadband stimulus), and US-Dollars
7.2 billion for Katz et al. (US Broadband stimulus plan approved by the US
Congress in February 2009).

All studies that have relied on input-output analysis have calculated multipliers,
which measure the total employment change throughout the economy resulting
from the deployment of a broadband network. Beyond network construction
(direct employment effects), broadband construction has an employment effect at
two additional levels. Following the sector interrelationships of input-output
matrices, network deployment will result in indirect job creation (incremental
employment generated by businesses selling to those that are directly involved in
network construction) and induced job creation (additional employment induced
by household spending based on the income earned from the direct and indirect
effects).

The interrelationship of these three effects is measured through multipliers, which
quantify the total employment change throughout the economy from one unit
change on the input side. Type I multipliers measure the direct and indirect effects
(direct plus indirect divided by the direct effect), while Type II multipliers
measure Type I plus induced effects (direct plus indirect plus induced divided by
the direct effect). While multipliers from one economy cannot be applied to
another one, it is useful to observe the summary results of multipliers of the four
input-output studies:















4 From the Bureau of Economic Analysis for the US studies or the national statistics authorities of
Switzerland for the Swiss study. In addition, the Strategic Networks Group (2003) also relied on
input-output tables, although in this case they were the regional ones created by Canada's statistics
agency, Statistics Canada.

6

---

Figure 2. Breakdown of Employment Multipliers of Studies relying on Input-
Output Analysis


Geography
Direct
Indirect Induced
Total
Effects
Effects
Effects
Crandall et al.
US
1.00
1.17
2.17
(2003)
Strategic Analysis
Canadian
1.00
1.03
1.4
3.42
Group (2003)
county
Katz et al. (2008)
Switzerland
1.00
0.38
N.A.
1.38
(*)
Atkinson et al.
US
1.00
1.47
1.13
3.60
(2009) (**)
Katz et al. (2009)
US
1.00
0.83
1.59
3.42
(*) This study calculates only direct and indirect effects; induced effects were not calculated
(**)Atkinson at al. (2009) multipliers have been recalculated by relying on Bivens (2003), which
the authors cite as their source.

Although multipliers from one economy should not be extrapolated to another,
their comparison can be useful in terms of their wide variance. According to the
sector interrelationships depicted above, a European economy appears to have
lower indirect effects than the US economy. Furthermore, the decomposition also
indicates that a relatively important job creation effect occurs as a result of
household spending based on the income earned from the direct and indirect
effects.

Beyond the employment and output impact of network deployment, researchers
have been focusing on a set of network externalities variously categorized as
"innovation", or "network effects" (Atkinson et al., 2009). In general, studies
based on regression analysis do not differentiate between construction and spill-
over effects. However, after examining the conclusions of the regression studies,
the evidence regarding externalities appears to be quite conclusive. First,
broadband spill-over employment effects are not uniform, they tend to concentrate
in service industries (e.g., financial services, health care etc.), although Crandall et
al. (2007) identified an effect in manufacturing as well. Second, two studies (e. g.
Lehr et al., 2006; Thompson et al., 2008) point to the productivity impact of
broadband, which can result in a net reduction in employment resulting from
capital-labor substitution.

Beyond, what can be inferred as "network effects" from the regression studies, two
types of approaches have been utilized to isolate this impact: 1) top-down based on
"network effect" multipliers, and 2) bottom-up estimates based on extrapolating
findings of microeconomic analysis of impact of broadband on efficiency and
effectiveness at the firm level.

Within the first group, key studies are Pocsiak (2002) and Atkinson et al. (2009).
Both studies relied on an estimated "network effect" multiplier, which is applied to

7

---

the network construction employment estimates.5 While the top-down approach
allows to rapidly estimating a number, it does not have a strong theoretical
support. Network effects are not built on interrelationships between sectors. They
refer to the impact of the technology on productivity, employment and innovation
by industrial sector.

On the other hand, only one bottom-up study of network effects has been
identified (Fornefeld et al., 2008). This study identified three types of impact of
broadband on employment: first, the acceleration of innovation resulting from the
introduction of new applications and services (with the consequent creation of
employment); second, the improvement of productivity as a result of the adoption
of more efficient business processes enabled by broadband; and third, the
possibility of attracting employment from other regions as a result of the ability to
process information and provide services remotely. These three effects act
simultaneously, resulting in contradictory impact on employment. The increase in
broadband penetration can have a positive impact on productivity, contributing as
a consequence to a negative effect on employment.6

However, this negative effect is compensated by the increase in the rate of
innovation and services, thereby resulting in the creation of new jobs. Finally, the
third effect may be comprised by two countervailing trends. On the one hand, a
region that increases its broadband penetration can attract employment displaced
from other regions by leveraging the ability to relocate functions remotely. On the
other hand, by increasing broadband penetration, the same region can lose jobs by
virtue of the outsourcing effect. While a better understanding of these combined
"network effects" is being gained, the research is still at its initial stages of
quantifying the combined impact. The study by Fornefeld et al. (2008) is probably
the first attempt to build a causality chain applying ratios derived from micro-
economic research to estimate the combined impact of all effects.

While the research of broadband impact on employment is somewhat limited, the
literature regarding the impact on economic output of broadband (primarily of
Information and Communications Technology (ICT) is quite extensive. As in the
case of impact on employment, studies tend to point out the existence of a causal
relationship.

A comparison study of OECD countries relating the impact of ICT capital
accumulation on output growth reveals that ICT contributed between 0.2 and 0.5
percentage points between countries in the last two decades (Coleccia and

5 For example, Pocsiak relied on two multiplier estimates (an IT multiplier of 1.5 to 2.0 attributed
to a think tank and another multiplier of 6.7, attributed to Microsoft) and calculated an average of
4.1. Similarly, Atkinson et al. (2009) derived a multiplier of 1.17 from Crandall et al. (2003).
6 This effect was alluded to by Lehr et al. (2006) when they said that "broadband might facilitate
capital-labor substitution, resulting in slower job growth", and is also alluded to by Thompson et al.
(2008) as they mention that "there may be a substitution effect between broadband and
employment."


8

---

Schreyer, 2002). The authors show that in the second half of the 1990s the impact
of ICT investments on economic growth rose to 0.3 to 0.9 percentage points.
Among other results they find that not only in the US ICT investments have a
strong impact on economic growth but that the possible effect of ICT investments
depends largely on the industry framework in the respective country. In this study
a big gap between the US and EU countries in the contribution of ICT investments
to economic growth is identified. Based on this finding, a series of studies
investigate the ICT different impact on economic growth between the US and EU
countries. For example, a study relating labor productivity growth on ICT
investments on an industry level shows that the faster productivity growth in the
US compared to EU countries can be attributed to a larger employment share in
the ICT producing sectors and a faster growth in industries that intensively use
ICT (van Ark et al., 2003). The same authors show that the key difference between
the growth impact of ICT in the US and the EU are the sectors heavily using ICT
services. In the second half of the decade the US has a rapid increase in
productivity growth whereas the growth in the EU countries stalled (van Ark et al.,
2002).

To sum up, the literature on the economic impact of broadband provides solid
grounding on the framework and methodology to estimate the impact on
employment and output of network construction. Input-output analysis is
analytically rigorous and its application to the issue under consideration has been
sufficiently codified. With regard to the externalities estimations, it is necessary to
develop a methodology that goes beyond the top-down multipliers and relies on
econometric modeling. This approach will require handling relative large time
series with a high level of disaggregation in order to establish regional effects and
build bottom estimates of impact.

3. THEORETICAL FRAMEWORK, METHODOLOGY AND DATA

The study methodology comprises three modules: 1) an estimation of total
investment required to meet the broadband targets, 2) an assessment of economic
impact of construction of the broadband network required to meet those targets,
and 3) an estimation of the economic impact to be achieved once the network is
deployed.

3.1. Estimation of total investment for broadband deployment

The coverage and service targets established by the Broadband Strategy are used
to estimate the costs of deploying the targeted broadband infrastructure. However,
two sequential scenarios were defined: a 2014, built around the strategy targets
(BMWi 2009a), and a 2020, defined on the basis of longer-term goals outlined in
other government (see citation of the 2018 case in BMWi, 2009c)

Once defined, the targets were compared against the current situation of
broadband deployment. Data on current coverage was extracted from the National

9

---

Broadband Strategy and the Broadband Atlas produced by the Federal Ministry of
Economics and Technology (BMWi 2009b). The comparison between the current
situation and the targets allows estimating the deployment objectives in terms of
number of lines to: 1) cover the "white" spots (unserved areas), 2) upgrade the
"grey" spots (areas with inferior service measured by low access speeds), and 3)
deploy VDSL and FTTH. Lines were estimated for different type of platforms
(wireless, DSL, VDSL and FTTH).

Once the number of lines by service target is estimated, they are multiplied by the
costs per broadband line by type of platform. In order to determine the costs per
line, the costs from deployment experience in Europe and the United States were
relied upon, adjusted for factors such as urban density, economies of scale, and
experience curve7. This calculation yielded the total investment required for
wireless and wireline technologies. The total investment was split according to
three cost categories: 1) construction labor, 2) electronic equipment, and 3)
telecommunications labor. These splits were based on cost allocations based on
"real life" deployment data for NGAN (furnished by a European operator), and for
3G and Wimax for a US operator.

The resulting process yielded the amount of total investment by cost category (see
figure 3):

Figure 3. Module for Estimating Deployment Costs


Broadband strategy
Future broadband
Total investment
Investment required
targets
infrastructure

required
by input
• 2014-20
• Cover “white” spots
• Wireline
• Construction
• Source:
and upgrade “grey”
• Wireless
• Electronics

Breitbandstrategie
• Telecommunications
spots
der Bundesregierung
• Deploy Ftth and

VDSL

Current broadband
Cost per broadband
Cost structure per
infrastructure
line
broadband line

• “White” spots
•“Real life” costs
• Construction labor,
• “Grey” spots
(EU/US)
electronic equipment,
• DSL/VDSL/cable/

• Adjusted for density,
telecommunications
satellite
economies of scale,
labor, civil and
• Source:
experience curves
electronic engineering,
Breitbandstrategie der

• Sources: WIK; JP
etc.
Bundesregierung
Morgan; AD Little;
• Sources:

Alliance Bernstein;
Deployment data for
Analysis Mason
NGAN, 3G and Wimax

in EU/US

Legend:
Input Calculated


3.2. Assessment of economic impact of network construction


7 See, in particular, Elixman et al. (2008)

10

---