DIREC TIONS IN DE VELOPMENT
Communication and Information Technologies
The Jobs of Tomorrow
Technology, Productivity, and Prosperity
in Latin America and the Caribbean
Mark A. Dutz, Rita K. Almeida, and Truman G. Packard
The Jobs of Tomorrow
DIREC TIONS IN DE VELOPMENT
Communication and Information Technologies
The Jobs of Tomorrow
Technology, Productivity, and Prosperity
in Latin America and the Caribbean
Mark A. Dutz, Rita K. Almeida, and Truman G. Packard
© 2018 International Bank for Reconstruction and Development / The World Bank
1818 H Street NW, Washington, DC 20433
Telephone: 202-473-1000; Internet: www.worldbank.org
Some rights reserved
1 2 3 4 21 20 19 18
This work is a product of the staff of The World Bank with external contributions. The findings, interpreta-
tions, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its
Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the
accuracy of the data included in this work. The boundaries, colors, denominations, and other information
shown on any map in this work do not imply any judgment on the part of The World Bank concerning the
legal status of any territory or the endorsement or acceptance of such boundaries.
Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and
immunities of The World Bank, all of which are specifically reserved.
Rights and Permissions
This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://
creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to
copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following
conditions:
Attribution—Please cite the work as follows: Dutz, Mark A., Rita K. Almeida, and Truman G. Packard.
2018. The Jobs of Tomorrow: Technology, Productivity, and Prosperity in Latin America and the Caribbean.
Directions in Development. Washington, DC: World Bank. doi:10.1596/978-1-4648-1222-4.
License: Creative Commons Attribution CC BY 3.0 IGO
Translations—If you create a translation of this work, please add the following disclaimer along with the
attribution: This translation was not created by The World Bank and should not be considered an official
World Bank translation. The World Bank shall not be liable for any content or error in this translation.
Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the
attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in
the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by
The World Bank.
Third-party content—The World Bank does not necessarily own each component of the content con-
tained within the work. The World Bank therefore does not warrant that the use of any third-party-
owned individual component or part contained in the work will not infringe on the rights of those
third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to
re-use a component of the work, it is your responsibility to determine whether permission is needed
for that re-use and to obtain permission from the copyright owner. Examples of components can
include, but are not limited to, tables, figures, or images.
All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group,
1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org.
ISBN (paper): 978-1-4648-1222-4
ISBN (electronic): 978-1-4648-1223-1
DOI: 10.1596/978-1-4648-1222-4
Cover photos: ©istockphoto.com. Photographers (left to right): Cyano66, danchooalex, Zapp2Photo,
skynesher, Leonid Eremeychuk, and Vertigo3d.
Cover image design: Florencia Micheltorena
Cover design: Debra Naylor / Naylor Design Inc.
Library of Congress Cataloging-in-Publication Data has been requested.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Contents
Foreword ix
Acknowledgments xi
About the Authors xiii
Executive Summary xv
Abbreviations xxi
Chapter 1 Introduction 1
Channels Linking Technology to More Inclusive Growth 2
Policies to Enable the Positive Impacts of Technology 6
Notes 9
References 9
Chapter 2 The Need for Productivity-Enhancing Technology
Adoption in Latin America and the Caribbean 11
References 16
Chapter 3 A Conceptual Framework 17
What Do We Know? 18
Predictions about the Diverse Impacts of
Technology Adoption 19
Notes 25
References 25
Chapter 4 New Lessons from the Region on the Impacts of
Technology Adoption 27
Impact on Firm Productivity and the Demand for
Jobs, Types of Skills, and Wages 29
Impacts on Job Dynamics and the Role of
Complementary Investments in Skills 32
The Role of Labor Market Regulations on Firms’
Decisions and Jobs Outcomes 33
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 v
vi Contents
Impacts of Technology on Firms and Workers through
Trade and Labor Mobility 35
Notes 36
References 37
Chapter 5 Improving the Environment for Technology Adoption
with Inclusion 39
Technology Diffusion Support Policies 41
Product Market Policies 42
Education, Skills, and Labor Market Policies 45
Notes 47
References 48
Chapter 6 Conclusions 51
Questions for Further Research 53
Reference 54
Appendix A Background Studies 55
Appendix B Detailed Literature Review 57
Boxes
1.1 Déjà vu—Preoccupations of and Responses to Perennial Luddites 3
3.1 A Model of Firm Heterogeneity with Predictions of the
Impacts of Technology Adoption 20
Figures
2.1 Unemployment and Productivity by Region 12
2.2 Unemployment and Productivity in Study Countries and
Comparators 12
2.3 Rates of Adoption of the Internet across Study Countries 13
2.4 Rates of Internet and Mobile Phone Use by Households across
Latin America and the Caribbean, Latest Year 13
2.5 Digital Adoption in Latin America and the Caribbean Is Still
Far from the East Asia and OECD Averages 14
B3.1.1 Substitution and Inclusive Output Expansion Effects from
Technology Adoption 20
5.1 LAC Holds Last Place in the Business Environment
Related to Digital Technologies 40
5.2 Indices of Competition 44
5.3 PISA Results and GDP per Capita 46
5.4 Scientific Production by Geographic Region 47
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Contents vii
Maps
2.1 Internet Service Provision across Brazilian Municipalities,
1999–2014 15
5.1 LAC Has Some of the Highest Total Tariffs and Taxes for
ICT Products 41
Tables
3.1 Predicted Impacts of Technology Adoption on Productivity,
Jobs, and Wages 23
4.1 Empirical Impacts of Technology Adoption on Jobs, Wages,
and Productivity 28
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Foreword
In the first decade of the 21st century, the Latin America and the Caribbean
region achieved strong growth with greater shared prosperity. Between 2000
and 2014, the region reduced poverty from 43 to 23 percent. For the first time,
more people were in the middle class than those living in poverty. Now, all
countries face the challenge of sustainably expanding these social achievements
based on productivity growth. That’s why increasing productivity needs to
become a top priority by adopting new-to-the-firm technologies in ways that
both improve the job prospects of lower-skilled workers and increase the
incomes of the poorest.
To design development policies, we need an understanding of the impact of
new technology adoption on inclusive growth—growth that improves the job
prospects of lower-skilled workers. This is even more important given the new
wave of digitalization and automation that is rapidly altering many economies
around the world. One of the key findings presented here is that lower-skilled
workers can, and often do, benefit from adoption of productivity-enhancing
technologies biased toward skilled workers. Concerns that lower-skilled workers
will be replaced by new technologies are often misplaced in practice. With a sup-
portive business environment and procompetitive enabling policies and institu-
tions, higher firm output based on increased productivity can expand sufficiently
to increase jobs across all tasks and skill types within adopting firms—as long as
low-skill occupations are not predominantly automated and displaced by the
new technologies. Cross-country studies highlight additional ways that digital
technology adoption can fuel inclusive growth, including lowering the fixed costs
of exporting through online trading platforms, reducing mobility costs for work-
ers in poorer countries, and increasing labor market efficiency through Internet-
enabled worker-firm job matches.
This research highlights the critical role of three types of policies supporting
growth and jobs from technology adoption. First, technology diffusion policies
should ensure that all businesses have access to the latest global technologies at
competitive prices. Second, product market policies should ensure that adopting
businesses have the incentives and opportunities to grow. Third, education, skill,
and labor market policies should ensure that workers are equipped with the right
skills and that businesses can flexibly deploy workers to meet changing business
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 ix
x Foreword
needs. Implementation of these policies will help ensure that technology
adoption has a positive impact on both productivity and workers in this new
technological age.
Jorge Familiar
Vice President
Latin America and the Caribbean Region
The World Bank
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Acknowledgments
This book was prepared by Mark A. Dutz, Lead Economist, Macroeconomics,
Trade and Investment Global Practice (task team leader [TTL]); Rita K. Almeida,
Senior Economist, Education Global Practice (co-TTL); and Truman G. Packard,
Lead Economist, Social Protection and Jobs Global Practice (co-TTL) of the
World Bank Group. Robert D. Willig (Professor Emeritus, Princeton University)
helped immensely in the formulation of the ideas underlying the regional study
proposal, this book, and the associated background studies. Solid research assis-
tance was provided by Jon Mallek.
The work was conducted under the general guidance of Carlos Vegh, Chief
Economist of the Latin America and the Caribbean Region of the World Bank
Group, and Daniel Lederman, Deputy Chief Economist. It benefited from fund-
ing from the World Bank’s Latin America and the Caribbean Chief Economist’s
Office, under the regional study, “Digital Technology Adoption, Skills, Productivity,
and Jobs in Latin America.”
The book builds on an extensive set of background research studies prepared
for this regional study (see appendix A). The team is grateful for helpful
comments from Marialisa Motta, Margaret Grosh, and Reema Nayar, Practice
Managers of the Finance, Competitiveness, and Innovation; the Social Protection
and Jobs; and the Education Global Practices, respectively, as well as from the
authors of all the background papers, who provided useful comments through-
out the book’s preparation process. Special thanks go to Omar Arias, Paulo
Bastos, and Mary Hallward-Driemeier, who provided useful guidance and advice
for the book. The team also thanks Erhan Artuç, Miriam Bruhn, Kumud Ghimire,
Siddhartha Raja, Rita Ramalho, Indhira Santos, Marc Schiffbauer, and Joana Silva.
We also thank participants in the World Bank Group’s authors’ workshop in
Washington in October 2016; the American Economic Association’s invited ses-
sion in January 2017; and the Latin American and Caribbean Economic
Association’s invited session in October 2017 for their comments and
suggestions.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 xi
About the Authors
Mark A. Dutz is Lead Economist in the Macroeconomics, Trade, and Investment
Global Practice of the World Bank. He is responsible for work on productivity
growth and its interaction with poverty reduction and shared prosperity. He has
authored policy papers on innovation, productivity, knowledge-based invest-
ment, biotechnology, information communication technologies, climate change,
competition, investment, and trade policies, and their linkages with growth and
inclusion. He is co-editor of Making Innovation Policy Work: Learning from
Experimentation (2014) and Promoting Inclusive Growth: Challenges and Policies
(2012), and he is lead author of Jobs and Growth: Brazil’s Productivity Agenda
(2018) and Unleashing India’s Innovation: Toward Sustainable and Inclusive
Growth (2007). Dutz has worked at the World Bank since 1990, and has experi-
ence in all regions and in the Office of the Chief Economist. He has also worked
with Compass Lexecon, Inc., the Turkish State Minister of Economic Affairs and
Treasury, and the European Bank for Reconstruction and Development. In addi-
tion, he has been a consultant for the Organisation for Economic Co-operation
and Development, the World Trade Organization, the World Intellectual
Property Organization, and Canada’s Networks of Centres of Excellence. He has
taught at Princeton University and has published articles in journals and mono-
graphs on applied microeconomics, including on international trade, competition
and innovation, and public policy toward network industries. Dutz holds a doc-
torate in economics from Princeton University and a master’s in public affairs
from Princeton’s Woodrow Wilson School of Public and International Affairs.
Rita K. Almeida has worked for the World Bank since 2002, where she has been
leading policy dialogue in Latin America and the Caribbean, Eastern Europe, and
the Middle East and North Africa. Before joining the Bank’s Education Global
Practice, she worked on the Global Knowledge team of the Social Protection and
Jobs Global Practice, and she was Research Economist in the Development
Economics Department. Almeida’s main areas of expertise include education
and skills development policies, labor-market analysis, activation and graduation
policies for the most vulnerable, labor market regulations and social protection
for workers, firm productivity and innovation policies, public expenditure
reviews, and the evaluation of social programs. Rita has served as lead on several
World Bank publications, including The Right Skills for the Job? Rethinking
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 xiii
xiv About the Authors
Training Policies for Workers and Toward More Efficient and Effective Public Social
Spending in Central America. Her work has been covered in the news media,
featured in leading world economic reports, and published in top general interest
and specialized economic journals, including The Economic Journal, American
Economic Journal: Applied Economics, Journal of International Economics, Journal of
Development Economics and Labour Economics. Almeida earned her doctorate in
economics from the Universitat Pompeu Fabra in Barcelona. Prior to joining
the Bank, she worked in an investment bank and taught at the Portuguese
Catholic University. She has been a Fellow of the Institute for the Study of Labor
since 2003.
Truman G. Packard is a Lead Economist in the World Bank’s Social Protection and
Jobs Global Practice. He has worked at the World Bank since 1997, providing
advisory assistance to governments in emerging markets on how to improve labor
regulation and social security to create jobs. Packard has worked with countries
in Latin America, Central Europe, and East Asia; he currently focuses mainly
on Brazil and Indonesia. Trained as a labor economist, he holds a doctorate
from the University of Oxford in the United Kingdom. His published work
focuses on how labor law and social insurance programs—retirement benefits,
unemployment insurance, and health coverage—affect peoples’ incentives to
work and save.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Executive Summary
Over the past decade, many countries in the Latin America and the Caribbean
(LAC) region have achieved strong growth and poverty reduction—but in an
unsustainable manner, through a commodity boom. Now that the commodity
tailwinds have receded, Latin American countries face the challenge of securing
and expanding their needed social achievements sustainably, through productiv-
ity growth enabled by new technologies.
Although the adoption of new technologies enhances long-term growth and
average per capita incomes, its impact on lower-skilled workers is more complex
and merits clarification. Concerns abound that new machines and other forms of
advanced technologies developed in high-income countries would, if adopted by
firms in the LAC region, inexorably lead to job losses for lower-skilled, less-well-
off workers and exacerbate poverty. Conversely, there are countervailing con-
cerns that policies intended to protect jobs from technology advancement would
themselves stultify progress and depress productivity.
This book squarely addresses both sets of concerns with new research showing
that adoption of information and communication technologies (ICT) offers a
pathway to more inclusive growth by increasing adopting firms’ output, with the
jobs-enhancing impact of technology adoption assisted by growth-enhancing
policies that foster sizable output expansion. “Inclusive growth” in this book is
growth that improves the job prospects of lower-skilled workers. The research
reported here uses economic theory and multicountry LAC data to demonstrate
that lower-skilled workers can, and do, benefit from adoption of productivity-
enhancing technologies biased toward skilled workers, such as ICT. The use of
the Internet allows firms to benefit by increasing productivity in areas ranging
from supplier and customer relations to recruiting and training, while use of
production, client management, and other software further supports production
planning and processes, product pricing, and related business tasks; and as infor-
mation becomes more available across the firm, workers can become more
sophisticated and make better decisions.
The inclusive jobs outcomes arise when the effects of increased productivity
and expanding output overcome the substitution of technology for workers. The
impacts on lower-skilled workers occur through both substitution and output
effects. Although the substitution effect replaces some lower-skilled workers
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 xv
xvi Executive Summary
with new technology and more highly skilled labor, the output effect can lead to
an increase in the total number of jobs for less-skilled workers. After all, as the
adopted technology increases the firm’s productivity, it enables reductions in
variable costs and product prices that generate an expansionary output effect.
Critically, output can increase sufficiently to increase jobs across all tasks and skill
types within adopting firms, including jobs for lower-skilled workers, as long as
lower-skill task content remains complementary to new technologies and related
occupations are not predominantly automated and replaced by machines. It is
this channel for inclusive growth that underlies the power of procompetitive
enabling policies and institutions—such as regulations encouraging firms to com-
pete, policies supporting the development of skills that technology augments
rather than replaces, and institutions that are capable and accountable—to
ensure that the positive impact of technology adoption on productivity and
lower-skilled workers is realized.
The size of the output expansion effect from the use of better technologies
and its impact on lower-skill jobs depend on the competitive market environment
in affected industries. Firms producing tradable output with effective distribution
channels and flexible input supplies predictably will expand vigorously in
response to the rise in productivity attributable to technology adoption. Firms in
competitive markets will be further impelled to reduce prices as their costs
decrease through the use of more productive technologies, thereby stimulating
additional demand and output. Firms operating in countries with education sys-
tems that produce more abundant and easily accessible skills complementary to
technology will also adapt and expand faster. These output expansion effects are
more apt to lead to greater demand for less-skilled workers if the production and
distribution tasks required for output expansion are not largely fixed costs, so that
the output expansion requires the performance of tasks and generates demand for
more workers. Positive economy-wide inclusive effects are also more likely, to the
extent that less-well-off workers can acquire throughout their lives skills that
are complementary to the adopted technologies. The inclusive effects are also
more likely to occur where expanding firms can flexibly hire and reallocate work-
ers in response to market opportunities, so that workers displaced within exiting
or contracting firms are able to move and find similar or better employment
opportunities in expanding firms within their existing or other industries.
Country studies on Argentina, Chile, Colombia, and Mexico find inclusive
growth due to the increased productivity impact of adoption of ICT and the
resulting positive output effects on lower-skilled jobs. In Argentina, manufactur-
ing firms that invested in ICT capital witnessed larger job increases for low-skilled
as well as high-skilled workers in high-growth firms through strong output expan-
sion effects that drive inclusive growth. In Colombia, manufacturing firms’ use of
high-speed broadband directly increases demand for laborers and lower-skilled
production workers, as well as higher-skilled professional workers. In Mexico,
a greater share of labor in manufacturing firms using the Internet results in an
increased number of blue-collar workers, even though the increase is larger for
white-collar workers. In Chile, the use of complex (production, client
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Executive Summary xvii
management, and other business) software increases the number of low-skilled
production workers, while no significant change is observed for skilled production
workers and managers. These studies are partial equilibrium analyses at the firm
level and do not address the impact on total employment of possible job losses or
exits among less-efficient nonadopting firms. However, the Brazil studies on stag-
gered Internet rollouts, examining effects at the municipal level, do reflect general
equilibrium effects within municipalities, including firm downsizing and exits
and their economy-wide feedback effects on employment in each directly
affected municipality. The sectoral impacts study with both contemporaneous
and lagged effects finds no positive economy-wide net effect on the total number
of formal jobs in the directly affected municipality, while the tasks and labor poli-
cies impact study finds an overall negative impact on employment in the short
term, with a larger negative impact for routine, manual tasks. This is to be
expected in a country such as Brazil, where opportunities for efficient global
output expansion have been more limited in light of its policy distortions, includ-
ing high trade and other product market expansion barriers. Importantly, the
sectoral impacts study with lagged effects finds that aggregate employment shifts
from sectors with limited expansion opportunities (wholesale and retail trade,
public administration, and publicly owned utilities, which jointly made up almost
half of the formal workforce in 2010) to sectors with more output expansion
opportunities (such as manufacturing, transport, and finance and insurance).
In Brazil’s manufacturing sector, Internet access with lagged effects induces posi-
tive job and wage effects, not only for high-skill occupations but also for medium-
skill jobs. And in Mexico, the positive effects from technology adoption on jobs
in manufacturing are much larger than the effects in the less-tradable commerce
sector. The country studies are able to show causal effects rather than correlations
by focusing on drivers that are exogenous to output and the demand for jobs,
skills, and labor earnings. The effects of technology adoption on productivity and
job-related outcomes are identified in a number of the studies from plausibly
exogenous changes in the availability of ICT or in its quality over time and space.
These exogenous variations are exploited as instruments for the otherwise pos-
sibly endogenous firm-level use of ICT.
Cross-country studies highlight two additional channels from ICT adoption to
inclusive growth: a market access effect that works in favor of smaller firms, and
a worker mobility effect that reduces the cost of information about job opportu-
nities. With respect to market access, an increase in the share of online exports
across countries is found to be associated with a decline in the wage skill pre-
mium, thereby reducing wage inequality. This effect is driven by a decrease in
the fixed costs of exporting due to online trading platforms that level the play-
ing field between small and large firms for access to international markets.
International transactions over the Internet disproportionately benefit smaller
firms that also tend to hire relatively more lower-skilled workers, allowing them
to reach new consumers across the world and reap the accompanying productiv-
ity gains. With respect to mobility, workers face higher mobility costs in poorer
countries. Access to the Internet is associated with lowering workers’ costs to
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
xviii Executive Summary
move across sectors and regions, increasing labor market efficiency by allowing
better employer-employee matches.
Policies that enable technology diffusion and product market competition are
critical to ensuring that the positive impact of technology adoption on inclusive
growth is realized. The first priority for firms in the LAC region are policies to
facilitate technology diffusion, adoption, and use, including policies to support
the rollout of faster Internet service at more affordable prices and reduce the
high tariffs and taxes on digital technology business tools to enable digital tech-
nologies generally. Current adoption of digital technologies across the LAC
region is highly heterogeneous and lags behind comparators in the Organisation
for Economic Co-operation and Development, showing that there is still much
potential for additional adoption in LAC and for the accompanying benefits of
productivity and inclusive growth. Second, product market policies should
enhance opportunities and sharpen incentives for output expansion in response
to the productivity increases that technology adoption yields. Enabling product
market policies include boosting the intensity of local market competition,
further opening external trade, improving access to finance, and supporting
management quality upgrading.
In addition, education, skills, and labor market policies should ensure that the
skills provided through the formal schooling system, and those acquired through-
out life in the labor market, are supporting the adoption and use of digital tech-
nologies and are available to firms when they need them. The Brazil tasks and
labor policies impact study shows that more-technology-intensive industries
reduce their relative reliance on employment to conduct more routine tasks,
thereby shifting the skills composition of the jobs they create toward nonroutine
and more cognitive and analytical tasks. Among the set of nonroutine tasks, com-
munication and interpersonal skills are in particularly high demand. After the
adoption of complex software by firms in Chile, firms also increase their invest-
ment in training of ICT-specific technical skills.
Evidence also shows that the stringency of labor market regulations matters
for the skills demanded in the labor market. Evidence from Brazil suggests
that more stringent enforcement of labor market regulations, contrary to policy
intentions, disproportionately benefits more-skilled workers—because firms
react to labor policies by substituting technologies and occupations with higher-
level cognitive and nonroutine tasks for occupations that previously performed
mostly routine tasks. A cross-country study for this book finds that a higher
statutory minimum wage is positively associated with higher digital technology
use by firms. This study also finds that business use of digital technologies is
lower in countries that require firms to follow more burdensome procedures to
dismiss workers.
Importantly, procompetitive technology diffusion, product-market, skills,
and labor policies that are often lacking in the LAC region also help create a
business environment in which firms have stronger incentives and capabilities
for investing in technology adoption. Firms invest in productivity upgrading
through technology adoption when faced with the market discipline, profit
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Executive Summary xix
rewards, and capabilities with which to do so. The same business environment
characteristics that support sizable output expansion and more inclusive
growth—including sufficient competition in investment in and delivery of
ICT services to adopt these technologies; product and input market competi-
tion; alignment of higher education offerings with labor market needs; and
high-quality management skills—also provide incentives for and enable firms to
invest in technology adoption. Finally, while the output expansion effect—
collectively enabled by technology adoption, product market, skills, and labor
policies—is clearly a desirable pathway for making productivity gains inclusive,
redistributive fiscal policies to support displaced workers and those unable to
find new jobs are a complementary pathway to inclusion facilitated by the
overall efficiency benefits of technology adoption.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Abbreviations
AI artificial intelligence
ATM automated teller machine
DAI Digital Adoption Index
ICT information and communication technology
IT information technology
LAC Latin America and the Caribbean
OECD Organisation for Economic Co-operation and Development
PISA Programme for International Student Assessment
PPP purchasing power parity
SBTC skill-biased technological change
TFP total factor productivity
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 xxi
CHAPTER 1
Introduction
Policy makers throughout Latin America and the Caribbean (LAC) would like
to understand how best to leverage recent and ongoing global, business-relevant
technologies to support productivity upgrading with inclusion. During the
first decade of the twenty-first century, growth and poverty reduction in most
LAC countries have been driven by an unsustainable commodity boom rather
than by sustainable productivity increases. Recently rising levels of unemploy-
ment and poverty in a number of countries raise the issue of whether more
inclusive growth can be promoted more sustainably through productivity
upgrading. A number of important questions need to be answered. Under what
conditions do firms have an incentive to adopt better technologies, and what are
the overall impacts of technology adoption on productivity, the number of jobs,
the types of skills demanded, and the wages offered to workers? Are the level and
pace of technological change having a significant impact on the jobs available to
the workforce and the types of skills demanded by businesses in LAC? Do labor
market regulations and social protection institutions help or constrain technology
adoption in the region?
This book discusses technology adoption and its impact on inclusive growth
through productivity, jobs, types of skills, and wages in Latin America. Although
it investigates impacts across all four of these outcomes, the book focuses par-
ticularly on two dimensions of inclusive economic growth: overall job growth,
and how less-skilled, less well-off workers can also benefit from technology
adoption. Although employment opportunities for less-skilled workers, sup-
ported by on-the-job training and continuous education, are clearly a desirable
pathway for making technological gains inclusive, complementary pathways—
such as redistributive fiscal policy to support displaced workers and those unable
to find a job, given that the overall efficiency benefits are positive—should also
be explored.
The penetration of digital technologies during recent years in several middle-
income LAC countries has created a learning opportunity that yields valuable
policy insights into how technology adoption can create a pathway to inclusive
growth. By “inclusive growth,” this book means growth that improves the job
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 1
2 Introduction
prospects of lower-skilled workers.1 The book and background technical studies
that form its analytical foundations are focused on the production side of the
economy. The impacts on job creation that the studies explore are additional to
positive impacts of technology adoption on consumer welfare through lower
prices and greater variety of product choices.2 Digital technologies in this book
encompass different types of information and communication technologies
(ICT) used by businesses, from basic Internet and high-speed broadband Internet
to the use of production, client management, and other business software pack-
ages and the use of “big data” to better understand consumer tastes and better
tailor goods and services to identified needs. Mechanisms through which increas-
ing use of ICT raises firm-level productivity include, among others, lowering
costs and adding value to process and production design, to relations with sup-
pliers and customers, to recruiting and training, and to the integrated manage-
ment of core business processes. The book’s focus on five middle-income
countries—Argentina, Brazil, Chile, Colombia, and Mexico—is dictated both by
the availability of high-quality data and by the differentially paced penetration of
digital technologies in the LAC region.
The evidence and conclusions presented in this book are relevant for lower-
income countries in the LAC region and for helping to understand the impacts
of the adoption of other types of technologies beyond ICT that reduce costs and
expand firms’ sales opportunities. The conceptual framework underpinning the
analysis models the impact of the adoption of ICT at the firm level—with ICT
representing any technology that, in combination with higher-skilled workers,
raises firm productivity and reduces variable costs of production. The assumption
that technology in combination with more skilled workers increases firm produc-
tivity is known as “skill-biased technological change.” It is related to the fact that
skilled labor is relatively more abundant in developed than in emerging econo-
mies, given that most technologies are still generated in developed economies.
The adoption of technology also requires firms to incur fixed costs to install the
technology, train their workers in its use, and realign the skill mix of their work-
forces. An important conclusion of this conceptual framework is that the impacts
that technology adoption will have on the demand for workers and types of skills
firms seek are largely empirical matters that can be informed by the framework
of economic theory.
Channels Linking Technology to More Inclusive Growth
This book investigates three channels linking technology adoption with more
inclusive growth: a sufficiently large firm-level output expansion effect, a market
access effect that increases smaller firms’ relative demand for lower-skilled
workers, and a worker mobility effect that lowers cross-sectoral and cross-
regional worker mobility costs through better Internet access.
The hypotheses about the effect of technology adoption on firm productiv-
ity and on the jobs and wages of lower-skilled workers associated with output
expansion are the most general because they are not restricted to the use of ICT.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Introduction 3
They apply to all technologies that raise business productivity and reduce vari-
able costs, including a range of applications of robots and artificial intelligence,
3D printing (additive manufacturing), nanotechnologies and biotechnologies,
and new material technologies—as long as lower-skill task content remains
complementary to new technologies and related occupations are not completely
automated and replaced by machines. All the country-specific studies explore
these hypotheses based on available data on ICT. The hypotheses about the
effect of technology on market access by firms that favors smaller firms that tend
to hire relatively more low-skilled workers and on worker mobility costs are
dependent on the adoption of ICT. They are respectively explored in only one
cross-country study each. All the studies jointly provide new empirical findings
on the extent to which these three channels can create a pathway to more
inclusive growth. These findings are important because they debunk the fre-
quent presumption that technology adoption necessarily kills jobs. Box 1.1 sum-
marizes the recurring concern about the impact of firms’ technology adoption
Box 1.1 Déjà vu—Preoccupations of and Responses to Perennial Luddites
Radical and rapid change can be uncomfortable. It challenges norms and disrupts routines.
Fears that new technologies will replace workers and destroy jobs are no different. Throughout
recent history, technological change has been met with pessimism and even catastrophizing.
Periods defined by rapid advances in technology tend to stoke fears of “technological unem-
ployment.” From the early advances of the Industrial Revolution to the advent of driverless cars
today, the idea that new technologies will put people out of work is as recurring as it is contro-
versial (Mokyr, Vickers, and Ziebarth 2015). Despite successive warnings of mass unemploy-
ment in the wake of technological change, however, the dire predictions have yet to come
about. Nonetheless, each new wave of technological advances is met with similarly dire pre-
dictions, insistent that “this time is different.”
One of the original and perhaps most dramatic responses to spurts of technological change
came during the first Industrial Revolution. Led by Ned Ludd, English weavers stormed facto-
ries and destroyed industrial textile machines they were sure would lead to the extinction of
work and society’s downfall. While those very machines did change industrial manufacturing
forever, they did not result in mass layoffs and high levels of long-term unemployment as
feared (Vivarelli 2014). Although 98 percent of the labor required to weave cloth was auto-
mated, the number of weaving jobs actually increased because demand for the lower-priced
cloth more than offset the labor-saving automation (Bessen 2015). Deloitte (2015) finds that
technology, in fact, has created more jobs than it has displaced in England and Wales since the
mid-nineteenth century. And Alexopoulos and Cohen (2016) find that commercialization of
new technologies raised productivity and employment and lowered unemployment between
1909 and 1949 in the United States.
In the United States in the 1950s and 1960s, fears of job displacement in the wake of new
technological advances arose again. An investigative piece called “The Automation Jobless”
box continues next page
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
4 Introduction
Box 1.1 Déjà vu—Preoccupations of and Responses to Perennial Luddites (continued)
detailed the fear of increasing unemployment, claiming that automation was the second most
important worry among Americans, surpassed only by the desire for peace. The fear at the
time was not so much that technology would eliminate jobs but rather that technological
innovations would not create any new employment (Time 1961). These fears grew to such a
fevered pitch that President Lyndon Johnson created a task force assigned to seek solutions to
the apparently looming crisis. After long and careful deliberation, the National Commission on
Technology, Automation, and Economic Progress found that automation, in fact, did not
threaten employment. It ultimately concluded that although new technologies might destroy
particular jobs, they would not eliminate the need for human work.
In most recent instances of technology adoption, more jobs have been created than were
expected. The impact of the automated teller machine (ATM) on the demand for bank tellers is
instructive. Even though the ATM replaced cash-handling tasks, the number of full-time-
equivalent bank tellers has grown substantially faster than the entire labor force since 2000.
Because ATMs allow banks to operate branch offices at lower cost, they have been prompted
to open many more branches, creating more jobs: there are more bank teller jobs in the United
States now than when the ATM was introduced. Similarly, the number of cashiers has grown
since barcode scanners were widely deployed during the 1980s, and the number of paralegals
has grown robustly since the introduction of electronic document discovery software for
legal proceedings in the late 1990s (Bessen 2015, 2016). More broadly, the employment-to-
population ratio in the United States increased remarkably in the second half of the
twentieth century, even as women joined the workforce en masse (Autor 2015).
Researching “automation” today—using a machine vilified in the 1961 Time article (a com-
puter) to operate an online search engine (Google) that employs tens of thousands of workers
in jobs that did not exist two decades ago—yields ominous, yet familiar, results: “Robots
will eliminate 6% of all US jobs by 2021, report says” (Solon 2016), and “Technology could kill
5 million jobs by 2020” (Kottasova 2016). The parallels to past periods of rapid technological
advancement are uncanny. At least no one is storming Google’s headquarters and destroying
its servers—yet.
on people’s labor market opportunities, and just how polarizing the ensuing
debates have been.
The first and most important channel from technology adoption to more
inclusive growth is through the additional jobs for less-skilled workers generated
by a sufficiently large output expansion effect. Conceptually, the impacts on
workers of firms’ ICT investments, and any other investments in technologies
that reduce variable costs in a skill-biased manner, happen through both substitu-
tion and output effects. The substitution effects often work against low-skilled
workers. This is what underlies the thinking of the “perennial Luddites” (box 1.1).
The most visible immediate effect of the introduction of a skill-biased technol-
ogy is the replacement of lower-skilled workers in those tasks that are substituted
for by the newly adopted technology. However, according to the predictions of
the conceptual framework that are tested in the country studies for this book,
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Introduction 5
this replacement does not have to be the case. First, the substitution effect
depends on the type of technology adopted and whether it complements skilled
or low-skilled labor. Second, the hypothesis underlying the adoption of technolo-
gies is that they increase the firm’s productivity, enabling a reduction in variable
costs and product prices, which—combined with possible increases in product
quality, the introduction of new products, and marketing outreach efforts—can
generate strong output expansion effects. These output effects can lead to an
increase in the total number of jobs created. Critically, output can increase to
such an extent that jobs are increased across all tasks and skill types within adopt-
ing firms, including not only the highest-skill but also lower-skill jobs. The size of
this effect and its impact on lower-skill jobs will depend on the sector and on its
business environment, including the extent of competition in the product
market. Firms in the tradable sectors are hypothesized to exhibit the greatest
output expansion effect on employment in response to the reduction in prices
enabled by adopted technologies. This effect occurs because these firms benefit
from higher demand in markets made possible by greater regional and interna-
tional export possibilities. And firms facing greater product market competition
are hypothesized to reduce prices more in response to variable cost reductions,
also increasing the output expansion effect. These theoretical predictions are the
hypotheses that are tested in the country studies.
The country studies underpinning this book show that job growth can be
inclusive in the wake of digital technology adoption when supported by procom-
petitive technology diffusion, product market, skills, and labor policies that
jointly create a business environment in which firms have strong incentives and
capabilities to expand output. Both productivity and lower-skill jobs increased
following digital technology adoption at the firm level in Argentina, Chile,
Colombia, and Mexico.3 The two studies on Brazil and separate studies on
Colombia and Mexico explore the impact of new or greater use by businesses of
the Internet or high-speed Internet. In Mexico, a greater share of labor using the
Internet in manufacturing and services results in an increased number of blue-
collar workers.4 In Colombia, manufacturing firms’ use of high-speed broadband
directly increases demand for laborers and lower-skilled production workers as
well as higher-skilled professional workers. The Argentina study finds that manu-
facturing firms that invest in ICT capital witness larger job increases for low-
skilled as well as skilled workers in high-growth firms, supporting the importance
of strong output expansion effects in driving inclusive growth. The Chile study
focuses on the impact on workers of firms’ use of complex software, specifically
production, client management, and other business software, which is quite dif-
ferent from the Internet. It finds that over a six-year horizon this technology
increases the number of low-skilled production workers, with no significant
change in the number of skilled production workers and managers. The adoption
of more complex software in Chile increases the use of routine and manual tasks
while decreasing the use of abstract tasks. Importantly, the Brazil studies provide
insights at the economy-wide level: while increased Internet access has no net
effect on aggregate employment, employment shifts from sectors with more
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
6 Introduction
limited expansion opportunities (wholesale and retail trade, public administra-
tion, and largely publicly owned utilities, which jointly made up almost half of
the formal workforce in 2010) to sectors with more output expansion opportu-
nities (such as manufacturing, transportation, and finance and insurance). In the
Brazil study of sectoral impacts, jobs increase for middle-skilled workers only in
the manufacturing sector, not in other nontradable sectors such as wholesale and
retail trade, where middle-skilled workers lose jobs. And in Mexico, much larger
positive effects on jobs are seen in manufacturing than in the less-tradable com-
merce sector.
The second channel from digital technology adoption to inclusive growth is
through a market access effect, working in favor of smaller firms—with lower-
skilled workers benefiting disproportionately as a result of online platforms that
facilitate smaller firms’ access to world markets. The lower fixed entry costs into
more distant national and foreign markets that digital technologies enable—
brought about, for example, by online trading platforms—allow all adopting
firms to benefit from lower trade connectivity costs. Critically, this technology
adoption allows smaller firms, which tend to hire relatively more low-skilled
workers, to disproportionately benefit from the reductions in fixed costs and
thereby access larger markets and reap the accompanying productivity gains.
The ability to increase revenues from existing inputs allows these firms to
increase wages relative to more skill-intensive firms, thereby reducing the wage
skill premium. A cross-country study for this book finds that a 1 percent increase
in the share of online exports leads to a 0.01 percent decline in the wage skill
premium, reducing wage inequality—with this relationship driven by countries
that have a large share of employment in small firms.
The third channel from digital technology adoption to inclusive growth is
through a cross-sectoral and cross-regional worker mobility effect, resulting in
increased labor market efficiency. Adoption by workers of digital technologies
such as better Internet access can reduce labor market frictions by reducing
workers’ mobility costs across sectors and subnational regions, allowing better
employer-employee matches and reducing frictional unemployment. A cross-
country study for this book finds that mobility costs are kept high by information
asymmetries: the average costs for workers to move across both sectors and sub-
national regions are about 1.8 times the average annual wage, and are higher in
lower-income countries. Countries with higher mobility costs also have higher
wage inequality. If workers have better access to the Internet, these costs can be
lowered across sectors and sector-regions.
Policies to Enable the Positive Impacts of Technology
Enabling policies are critical to ensuring that the potential positive impact of
technology adoption on inclusive growth is realized. The empirical studies
underpinning this book clarify how the channels that can, in principle, link tech-
nology adoption to inclusive growth in fact play out. Importantly, the studies
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Introduction 7
carried out for this book provide insights into the proactive role policies should
play for technology adoption to actually yield more inclusive growth. Three main
types of supportive policies, each linked to the business environment, are jointly
needed to increase firms’ incentives and capabilities to expand output in ways
that also improve the job prospects of lower-skilled workers.
First are policies to support technology diffusion, adoption, and use—
including digital technology policies to support high-quality and competi-
tively priced Internet rollouts. The Internet is the platform on which digital
technologies thrive. Countries can do much more to support low-cost, high-
speed Internet access, including procompetitive support of higher-speed
broadband rollout regimes. They also should consider reducing the high tariffs
and taxes on digital technology business tools to enable digital technolo-
gies generally. Current adoption of ICT across the LAC region is highly het-
erogeneous and lags behind comparators in the Organisation for Economic
Co-operation and Development, demonstrating that there is still much poten-
tial for significant additional adoption in LAC, and for the expected accompa-
nying productivity gains.
Second, product market policies should enhance opportunities and sharpen
incentives for output expansion in response to the productivity increases that
technology adoption yields. If total sales were more responsive to adopting firms’
price decreases (quality increases and market outreach efforts) due to greater
regional and international export possibilities, then output expansion would be
even more likely. This output expansion could add significantly to the standard
benefits of reducing logistics costs and other interregional and international trade
barriers. Additional supportive product market policies include boosting the
intensity of local market competition (including lowering entry and exit costs,
supported by adequate bankruptcy protection to guard investors’ interests if
output retraction is necessary), improving access to finance (to purchase required
inputs and fund promotion efforts), and upgrading management quality, along
with other factors affecting firms’ ability and know-how to enlarge production
and distribution in response to lower variable costs.
Third, education, skills, and labor market policies are critical to ensuring that
the available skills of individuals in the labor market support the adoption and
use of digital technologies. The Brazil tasks and labor policies impact study shows
that more technology-intensive industries with earlier access to the Internet,
across all sectors of the economy, reduce their relative reliance on routine and
manual tasks, thereby shifting the skill composition toward more cognitive and
nonroutine tasks. Furthermore, among the set of cognitive tasks, technology-
intensive industries also increase their use of communication and interpersonal
skills in the aftermath of digital technology adoption. The Chile study also shows
that the adoption of complex software is positively correlated with increased
investment in digital ICT skills for managers. These findings have critical implica-
tions for the education and training systems in the region: in the future, as digital
technologies further expand, the skills mix needed to succeed in the labor market
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
8 Introduction
will significantly change, but most of today’s education and training systems
are failing to keep up. In particular, the new research shows that workers will
need solid higher-order cognitive, technical (ICT), and interpersonal skills.
Furthermore, labor market policies also matter for inclusion. A cross-country
study for this book looks at the relationship between digital technology adoption
and de jure labor market regulations, which are typically more restrictive in the
LAC region compared with low- and middle-income countries in other regions.
It finds that a higher statutory minimum wage is positively associated with the
extent of digital technology use by firms. Conversely, it finds that business use of
digital technologies is lower in countries that require firms to follow more
burdensome procedures to dismiss workers. The Brazil tasks and labor policies
impact study has a related finding: stringent labor regulations particularly con-
strain the flexibility of firms in hiring lower-skilled workers who perform routine
and manual tasks. Importantly, employment protection regulations in Brazil have
adverse distributional consequences, with stringent enforcement of regulations
affecting lower-skilled workers in routine and manual tasks more negatively.
In contrast, appropriate policies should facilitate required cognitive and socio-
emotional skill availability to meet the needs of business. They should facilitate
firms’ reallocation of workers across tasks in response to opportunities offered by
technology adoption. They also should facilitate workers’ mobility across firms
and industries, particularly through support for job-search and business-relevant
continuing education and training programs.
This book synthesizes policy-relevant analytical findings to inform an
active debate across the LAC region about the impact of technology adop-
tion on jobs and skills. The remainder of this book is organized around the
following issues: Chapter 2 provides a succinct context underpinning the
importance of fostering productivity with more inclusive growth through
digital technology adoption. Chapter 3 lays out the core assumptions and
implications of a conceptual framework for technology adoption that realisti-
cally assumes that both firms and workers are heterogeneous agents. The
conceptual framework is preceded by a brief overview of the academic litera-
ture on the impacts of technology adoption for developed economies and the
LAC region, with a more extensive survey in appendix B. Chapter 4 discusses
new learning from the region on the impacts of technology adoption. The first
section of chapter 4 focuses on impacts on productivity, business demand for
jobs, types of skills, and wages. The second section of the chapter discusses the
impacts on job dynamics and the role of complementary investments in skills.
The third section discusses the role of labor market regulations on firms’ deci-
sions and jobs outcomes, while the fourth looks at the impacts of technology
on firms and workers through trade and labor mobility. Chapter 5 discusses
the main policy implications related to improving the broadly defined busi-
ness environment, including technology diffusion support policies, product
market policies, and education, skills, and labor market policies. Chapter 6
concludes by summarizing the main findings and outlining some questions for
further research.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Introduction 9
Notes
1. The book’s focus on the impact of technology adoption on workers, and especially
lower-skilled (lower-income) workers rather than on all individuals in the lower part
of the income distribution (for instance, all individuals in the bottom 40 percent of the
income distribution, according to the World Bank Group’s definition of shared pros-
perity), or on a specific measure of income inequality, is dictated by data availability.
2. For robust measures of the contribution of home broadband to consumer welfare
through lower prices and greater variety of product choices during the early years of
broadband adoption by U.S. households, see Dutz, Orszag, and Willig (2012).
Consumer surplus from the Internet is found to concentrate in broadband services,
with the net consumer benefits from home broadband in 2008 on the order of
US$32 billion per year.
3. The result of an increase in lower-skill jobs holds across different types of digital
technologies.
4. The increase is larger for white-collar workers in manufacturing, and for blue-collar
workers in services. There is no evidence of reductions in the number of blue-collar
workers in commerce, with the sector exhibiting very low and even nonsignificant
coefficients in some of the econometric specifications.
References
Alexopoulos, Michelle, and Jon Cohen. 2016. “The Medium Is the Measure: Technical
Change and Employment, 1909–1949.” Review of Economics and Statistics 98 (4):
792–810.
Autor, D. H. 2015. “Why Are There Still So Many Jobs? The History and Future of
Workplace Automation.” Journal of Economic Perspectives 29 (3): 3–30.
Bessen, James. 2015. Learning by Doing: The Real Connection between Innovation, Wages
and Wealth. New Haven, CT: Yale University Press.
———. 2016. “How Computer Automation Affects Occupations: Technology, Jobs and
Skills.” Boston University Law and Economics Working Paper No. 15-49, October,
Boston University, Boston.
Deloitte. 2015. “From Brawn to Brains: The Effect of Technology on Jobs in the UK.”
Deloitte.
Dutz, Mark, Jonathan Orszag, and Robert Willig. 2012. “The Liftoff of Consumer Benefits
from the Broadband Revolution.” Review of Network Economics 11 (4).
Kottasova, I. 2016. “Technology Could Kill 5 Million Jobs by 2020.” CNN Money,
January 18.
Mokyr, Joel, Chris Vickers, and Nicholas L. Ziebarth. 2015. “The History of Technological
Anxiety and the Future of Economic Growth: Is This Time Different?” Journal of
Economic Perspectives 29 (3): 31–50.
Solon, Olivia. 2016. “Robots Will Eliminate 6% of All US Jobs by 2021, Report Says.”
The Guardian, September 13.
Time. 1961. Business: “The Automation Jobless.” February 24. http://content.time.com
/time/subscriber/article/0,33009,828815-1,00.html.
Vivarelli, Marco. 2014. “Innovation, Employment and Skills in Advanced and Developing
Countries: A Survey of Economic Literature.” Journal of Economic Issues 48 (1):
123–54. https://doi.org/10.2753/JEI0021-3624480106.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
CHAPTER 2
The Need for Productivity-Enhancing
Technology Adoption in Latin
America and the Caribbean
Slower economic growth and rising unemployment in Latin America and the
Caribbean (LAC) are giving new urgency to increases in productivity that create
more jobs. Lackluster rates of productivity have long been a worry for policy
makers in LAC (Perry et al. 2007; Pagés 2010). However, years of rapid growth
fueled by the commodity “super cycle” provided a distraction from underlying
structural issues that have long held countries in the region back from realizing
their economic potential. The years of easy bounty are now over, pushing the
need for productivity gains and job creation to the top of governments’ lists of
priorities once again (Lederman and Porto 2014; de la Torre et al. 2015).
The rate of economic deceleration is by no means uniform across countries.
Yet, as concerns about productivity and jobs grow, most governments in the
region are facing a combination of monetary and fiscal policy constraints that
could affect their capacity for policy maneuvering just as the risks of job losses
and extended periods of unemployment are on the rise. It is also increasingly clear
that unlike in the mid- and late 1990s, the current deceleration is not a transitory
shock but a downward adjustment to a new, lower growth equilibrium absent
structural microeconomic reforms. The average increase in unemployment in
LAC and in selected study countries makes tapping new sources of productivity
growth all the more important, as highlighted in figures 2.1 and 2.2.
The speed of adoption and extent of Internet use varies enormously within
the LAC region. Figure 2.3 shows the varying rates of Internet adoption by
households across countries between 2000 and 2013, highlighting the large
divergence with the Organisation for Economic Co-operation and Development
(OECD) average. Figure 2.4 compares levels of Internet household penetra-
tion, together with mobile phone penetration as a comparator, ranking all
countries in the region. Although the extent of mobile phone use is higher than
80 percent in most LAC countries, the extent of Internet use still varies greatly,
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 11
12 The Need for Productivity-Enhancing Technology Adoption in Latin America and the Caribbean
Figure 2.1 Unemployment and Productivity by Region
a. Unemployment b. Labor productivity
14 70,000
GDP per worker, constant PPP 2011 US$
Annual unemployment rate (percent)
12 60,000
10 50,000
8 40,000
6 30,000
4 20,000
2 10,000
0 0
00
02
04
06
08
10
12
14
16
20 0
20 1
20 2
20 3
20 4
20 5
20 6
20 7
20 8
20 9
20 0
20 1
20 2
20 3
20 4
20 5
16
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
20
20
20
20
20
20
20
20
20
20
East Asia and Pacific Middle East and North Africa
Europe and Central Asia South Asia
Latin America and the Caribbean Sub-Saharan Africa
Source: World Bank World Development Indicators.
Note: PPP = purchasing power parity.
Figure 2.2 Unemployment and Productivity in Study Countries and Comparators
a. Unemployment, study countries b. Labor productivity, study countries
and comparators
25 120,000
GDP per worker, constant PPP 2011 US$
Annual unemployment rate (percent)
100,000
20
80,000
15
60,000
10 40,000
20,000
5
0
0
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
20 15
16
20
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
20 15
16
20
Argentina China Korea, Rep.
Argentina Chile Mexico Brazil Colombia Mexico
Brazil Colombia Chile Germany United States
Source: World Bank World Development Indicators.
Note: PPP = purchasing power parity.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
The Need for Productivity-Enhancing Technology Adoption in Latin America and the Caribbean 13
Figure 2.3 Rates of Adoption of the Internet across Study Countries
90
Percentage of households
80
with Internet access
70
60
50
40
30
20
10
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
Argentina Chile Mexico East Asia and Pacific
Brazil Colombia OECD
Source: International Telecommunication Union.
Note: OECD = Organisation for Economic Co-operation and Development.
Figure 2.4 Rates of Internet and Mobile Phone Use by Households across Latin America and
the Caribbean, Latest Year
100
80
60
Percent
40
20
0
Ho rag i
nd ua
ne cua y
el r
Gu ras
rin na
Bo e
El Be ia
Gu lva e
em r
n P la
pu u
M blic
ra co
Ja , RB
na a
a
lo zil
ge ia
id C C na
an ta ile
To ca
ug o
OE ay
CD
ca ait
zu do
at do
Ve E gua
Pa aic
m
am
Sa liz
Ur ag
Re er
liv
Ar b
a
Co Bra
d Ri
u
ad o h
Pa exi
Su ya
i
u
nt
Ni H
m
a
b
m
s
ica
in
m
in
Do
Tr
Mobile phones Internet
Source: International Telecommunication Union.
Note: OECD = Organisation for Economic Co-operation and Development.
from as low as 7 percent and 15 percent of households in Haiti and Nicaragua,
respectively, to as high as 68 percent in Uruguay. In OECD member countries,
the average rate is already 85 percent. Although closely correlated with wealth
and development (represented by GDP per capita), there are substantial
differences in digital technology penetration by people and businesses, even
across countries in LAC at very similar levels of development, as highlighted
in figure 2.5.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
14 The Need for Productivity-Enhancing Technology Adoption in Latin America and the Caribbean
Figure 2.5 Digital Adoption in Latin America and the Caribbean Is Still Far from the East Asia
and OECD Averages
a. Digital adoption by people b. Digital adoption by businesses
1.0 1.0
CHL
0.8 BRZ 0.8
COL ARG
Digital adoption index
Digital adoption index
0.6 0.6
MEX BRA CHL
ARG MEX
HTI BLZ
0.4 0.4
COL
0.2 0.2 HTI GUY LCA
0 0
100 1,000 10,000 100,000 100 1,000 10,000 100,000
GDP per capita (constant 2005 US$) GDP per capita (constant 2005 US$)
Latin America and the Caribbean Rest of world
World average OECD average
East Asia average LAC average
Source: World Bank World Development Report 2016.
Note: The Digital Adoption Index was compiled for World Development Report 2016 and is described in detail in chapter 5. LAC = Latin America and
the Caribbean; OECD = Organisation for Economic Co-operation and Development.
Even in the wealthiest, institutionally most advanced LAC countries, digital
technology adoption by households and businesses is well below that of peer
countries and members of the OECD. Figure 2.5 plots countries according to the
Digital Adoption Indices (DAIs) created for World Development Report 2016:
Digital Dividends (World Bank 2016). Countries in LAC are highlighted along
with three benchmarks: the LAC average, the average in East Asia and Pacific (a
region that many countries in LAC regard as peers and trading partners), and the
average level in the OECD. It is unsurprising that the highest levels of the DAI
in LAC are observed in Chile, Brazil, and Argentina, and that the lowest levels of
digital adoption are observed in lower-income Haiti. What is more intriguing are
the substantial differences in the DAI across countries with similar levels of eco-
nomic development, and conversely how a relatively poor country like Haiti has
about the same level of digital adoption by people as much-wealthier Mexico.
What should cause additional concern for policy makers across LAC is that even
the LAC countries where the levels of digital adoption are highest lie well below
the average level of adoption for countries in East Asia and for the OECD. For
countries in LAC, the room for improvement is starkly apparent.
The adoption of digital technologies, including Internet penetration, also var-
ies substantially within countries in the LAC region. The variations in access to
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
The Need for Productivity-Enhancing Technology Adoption in Latin America and the Caribbean 15
the Internet by firms within countries at the subnational level are exploited by
all the country studies. For instance, between 1999 and 2014 access varied sub-
stantially across the more than 5,000 Brazilian municipalities. Larger population
centers benefited from earlier access to the Internet. In 1999, only 15 percent of
all municipalities had local Internet service, accounting for roughly 60 percent
of the Brazilian population. Access to digital technologies has grown consider-
ably since then. By 2006, more than half of all municipalities had a local Internet
service provider, accounting for almost 90 percent of the population. Map 2.1
illustrates the spreading of the Internet across the country over time. The darker
areas of the map obtained Internet service earlier than the lighter areas.
The municipalities in the center of the country and remote Amazon region, rep-
resented by white areas, remained without Internet service as of 2014.
Map 2.1 Internet Service Provision across Brazilian Municipalities, 1999–2014
Adoption year
As of 2014
As of 2012
As of 2009
As of 2006
As of 2005
As of 2001
As of 1999
Without Internet
Source: Almeida, Corseuil, and Poole 2017.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
16 The Need for Productivity-Enhancing Technology Adoption in Latin America and the Caribbean
References
Almeida, Rita, Carlos Corseuil, and Jennifer Poole. 2017. “The Impact of Digital
Technologies on Routine Tasks: Do Labor Policies Matter?” Policy Research Working
Paper 8187, World Bank, Washington, DC.
de La Torre, A., A. Ize, G. Beylis, and D. Lederman. 2015. Jobs, Wages and the Latin
American Slowdown. Washington, DC: World Bank.
Lederman, Daniel, and Guido Porto. 2014. “The Price Is Not Always Right: On the
Impacts of (Commodity) Prices on Households (and Countries).” Policy Research
Working Paper 6858, World Bank, Washington, DC.
Pagés, C., ed. 2010. The Age of Productivity: Transforming Economies from the Bottom Up.
Washington, DC: Inter-American Development Bank.
Perry, Guillermo, William F. Maloney, Omar S. Arias, Pablo Fajnzylber, Andrew D. Mason,
and Jaime Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. World Bank Latin
American and Caribbean Regional Study. Washington, DC: World Bank.
World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, DC:
World Bank.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
CHAPTER 3
A Conceptual Framework
This chapter presents a new conceptual framework designed to address the
key research questions of this book regarding the challenges and opportunities
for productivity enhancement and job creation through technology adoption.
The conceptual framework is intended to provide testable predictions that are
explored in the subsequently presented country-based empirical studies.
A review of the relevant economic literature (appendix B) helps put the con-
ceptual framework and the later summary of this book’s empirical findings
into a broader context. The research questions can be grouped into three
broad categories:
• Productivity, jobs, and skills: How does technology adoption affect productivity,
overall levels of employment, and the demand for workers with different types
of skills? Does technology adoption generate a sufficiently large boost to out-
put, with sufficiently responsive product demand, to expand jobs for both
high-skilled and low-skilled workers? How do impacts differ depending on
technology, sector, and the regulatory environment within which firms
operate?
• Job dynamics, complementary investments (on-the-job training), and the role of
labor market regulations: How does technology adoption affect firms’ incen-
tives to dynamically adjust jobs or invest in workforce skills that may not be
readily available? How do labor market regulations, such as a statutory min-
imum wage and employment protection, affect firms’ decisions to adjust
the workforce to their new needs? Do labor regulations promote or hinder
the adoption of technologies?
• Trade and labor mobility: Under which conditions does access to digital tech-
nologies reduce wage inequality through trade between countries and raise
labor mobility across sectors and regions?
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 17
18 A Conceptual Framework
What Do We Know?
Productivity upgrading through creative destruction is the main driver of long-
term per capita income growth. This dynamic involves innovation (both the
commercialization of frontier technologies and technology adoption, namely,
the adoption of existing improvements in product, process, marketing, mana-
gerial, organizational, and related practices by firms) and the reallocation of
resources across industries and across firms within industries (Aghion and
Howitt 2009; Comin and Ferrer 2013; Cirera and Maloney 2017). Rising
country-level productivity has been found to be associated with growing
aggregate employment over time (Autor and Salomons 2017).
The distributional impact of technology adoption is more complex, particu-
larly as it affects productivity and firms’ demand for labor and human capital.
The existing literature shows that the impact of technology adoption on the
welfare of different groups is ambiguous. On the one hand, technology adop-
tion can lead to income inequality if the benefits are disproportionately appro-
priated by owners of physical capital, managers, and high-skilled workers able
to implement new technologies, without sufficient benefit to low-income or
less skilled workers. At the limit, specific types of technology adoption that
substitute machines for labor can make certain categories of workers redun-
dant and unemployable (see Acemoglu and Autor 2011; Acemoglu and
Restrepo 2016, 2017, 2018; and the related large literature on automation and
skill-biased technological change). For example, technology adoption has
recently been associated with a decline in mid-skill occupations relative to
low- and high-skill ones in the United States (Autor and Dorn 2013). On the
other hand, automation and other forms of technology adoption may comple-
ment labor, decreasing variable costs and increasing productivity, thereby rais-
ing output in ways that lead to higher demand for labor and increased earnings
(Autor 2015). Such types of technology adoption (or catch-up innovation)
generate higher productivity, sales, and employment when spurred by a com-
petitive business environment. Indeed, technology adoption has recently been
shown to be associated with a larger employment share of low-skilled workers
and women in an empirical study of more than 26,000 manufacturing estab-
lishments across 15 Organisation for Economic Co-operation and Development
and 56 developing countries (Dutz et al. 2012). This finding suggests that firm
growth from innovation can be more inclusive. The use of the Internet and the
presence of job-training programs are shown in these data to make significant
contributions at every stage of the flow from idea generation to inclusive
employment growth.1
Most of the literature exploring the impact of technology adoption focuses
on high-income countries, with some recent and interesting exceptions from
Latin America and the Caribbean. Appendix B provides a more detailed sum-
mary of the existing literature on the impacts of technology adoption on jobs
and the skill content of tasks, on trade and labor mobility, and on the impact of
labor market regulations. The literature review serves as context for the book’s
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
A Conceptual Framework 19
conceptual framework and for the nine empirical studies commissioned for this
book (see appendix A). A contribution of this book, which becomes apparent
through the differing frameworks adopted in the literature review, is the value
of empirical studies across different Latin American countries all linked to a
similar conceptual framework.
Predictions about the Diverse Impacts of Technology Adoption
The conceptual framework developed for this book is designed to illuminate
the effects of technology adoption on firm-level outcomes by allowing tech-
nologies and wages to vary across firms within each product market. In her
theoretical paper, “Digital Technology Adoption and Jobs: A Model of Firm
Heterogeneity,” Brambilla extends Acemoglu and Autor’s (2011) task-based
model of technical progress and labor markets by allowing firm heterogeneity
in technology (firms differ in their efficiency, with some able to produce
higher output per unit of the composite of production tasks than others) and
wage heterogeneity across firms (wage rates vary across firms within each
market because of assumed noncompetitive labor markets with rent-sharing
combined with firm heterogeneity in technology). These extensions allow the
outcomes of the book’s basic model to better approximate the high variance
in labor force composition and wages across firms highlighted by empirical
firm-level studies—so that the predictions will be more realistic when tech-
nology adoption (firm investment in and use of information and communica-
tion technologies [ICT]) is introduced.
The model, summarized in box 3.1, combines assumptions about firms’
production technology, labor markets, and individual preferences to generate
outcomes that are compatible with observed, firm-level outcomes. In line with
the task-content-of-jobs approach, production is assumed to have a two-tier
structure: production of final goods is carried out by a combination of produc-
tion tasks (which correspond to production stages such as design, development,
assembly, management, commercialization, and distribution), with each of the
underlying tasks in turn requiring differing numbers of high-skilled and low-
skilled workers. For more complex tasks, the comparative advantage of high-
skilled workers increases. In the assumed noncompetitive labor markets,
firm-level wages for high-skilled and low-skilled workers follow rent-sharing
schedules that are increasing in profits, with workers assumed to have fair-wage
demands (the wage schedules can result from an efficiency wages model or
from a bargaining solution after job search).
Greater productivity translates to higher demand, growth in output, and more
jobs. The implication of these assumptions is that more productive firms charge
a lower output price and thus have higher output, profits, jobs, skill intensity, and
task complexity, and pay higher wages to workers across all skill types. High-
skilled workers are paid more than low-skilled workers, and workers of the same
type are paid the same wage across tasks. The demand side uses a standard
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
20 A Conceptual Framework
Box 3.1 A Model of Firm Heterogeneity with Predictions of the Impacts of
Technology Adoption
The conceptual framework developed by Brambilla extends Acemoglu and Autor’s (2011)
task-based model of technical progress and labor markets by realistically allowing firms to dif-
fer in their efficiency (production at differing levels of output per unit of input) and allowing
workers’ wages to vary across firms (in addition to skilled workers being paid more than low-
skilled workers, wages increase with profits across firms because of the assumption of rent-
sharing wage schedules resulting from bargaining after job search or from an efficiency wages
model in which workers only exert effort if they perceive their wage to be fair).
In a given industry, each differentiated variety is produced by a single-product firm under
economies of scale. In a situation in which information and communication technologies (ICT)
are not available, a firm that decides to stay in the market (because its fixed cost is sufficiently
low relative to its technology parameter θ, with higher θ corresponding to higher output per
unit of the composite variable input) chooses to produce each task, for instance, less complex
task i and more complex task j, with either low-skilled or skilled labor according to which
option is less costly, where the parameters aH( j, θ) and aL(i, θ) are the inverse requirements of
high-skilled and low-skilled workers needed to produce one unit of task i, which vary across
firms with θ.
The isoquant graph in figure B3.1.1 represents a situation at A in which task i, on the
horizontal axis, is performed by low-skilled workers and task j, on the vertical axis, is pro-
duced by skilled workers, with the optimal combination A depending on the relative cost of
Figure B3.1.1 Substitution and Inclusive Output Expansion Effects from
Technology Adoption
wL (θ) aH (j, θ)
t (j, θ)
sT (θ) aL (i, θ)
C
B
wL (θ) aH (j, θ)
wH (θ) aL (i, θ)
A
t (i, θ)
box continues next page
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
A Conceptual Framework 21
Box 3.1 A Model of Firm Heterogeneity with Predictions of the Impacts of Technology
Adoption (continued)
each task (the quotient represents the ratio of the relative costs of the two tasks, with an
increase in the wage of low-skilled labor or in the unit requirements of low-skilled labor for
task i implying an increase in the production of task j, and vice versa).
Investment in ICT increases the participation of more complex tasks because the combi-
nation of ICT with skilled workers increases the firm’s productivity and reduces variable costs
of production. The cost of a combined unit of skilled labor and ICT is represented by sT(θ),
lower than both the skilled wage and the price of one unit of ICT capital. The dashed orange
line represents the changed relative cost or productivity of tasks due to ICT investment. The
movement from A to B is the substitution effect, that is, the shift in the combination of tasks
conditional on no change in output (with zero output expansion effect, the movement along
the initial isoquant by assumption has to result in a decrease in the production of noncom-
plex tasks requiring fewer low-skilled workers and an increase in the production of more
complex tasks, represented by the white arrows). However, total production of tasks depends
on total output produced. The movement from B to C is the output expansion effect, repre-
sented by the extent to which the dashed orange line shifts out and by the green arrows for
each task. The magnitude depends on the drop in the firm’s marginal cost due to the adop-
tion of ICT; the level of competitive rivalry in the firm’s market (which will affect how much
the firm is disciplined by the market to reduce output prices); the availability of added labor,
capital, materials, energy, and distribution inputs for expansion opportunities; and the elas-
ticity of market demand, that is, how much consumer demand expands in response to a
given drop in output price (which will depend on substitution possibilities from rival suppli-
ers and other industries as well as regional and international export possibilities).
The model provides predictions for within-firm variation due to skill-biased technology (for
example, ICT) adoption. Firms that invest in these technologies experience an increase in pro-
ductivity and a reduction in variable costs, have larger sales, pay higher wages to both skilled
and low-skilled workers, and perform a larger share of complex tasks with a combination of
skilled labor and ICT. However, the predicted impact on jobs is ambiguous. For skilled workers,
both the output and substitution effects (including two additional substitution effects not
depicted here, namely, a within-firm shift to new tasks using ICT and skilled workers and an
across-firm shift of industry output to firms that become relatively more productive because of
their relatively more intense employment of both ICT and skilled workers) work in their favor,
except for the possibility of an additional adverse substitution effect of ICT replacing skilled
labor (as figure B3.1.1 is drawn, skilled labor performing the more complex task j is comple-
mentary to ICT, with the movement from B to C resulting in a further increase in the more
complex task). Jobs for low-skilled workers will increase to the extent that the output expan-
sion effect dominates the substitution effects (as figure B3.1.1 is drawn, the movement from B
to C results in a sufficient increase in the less complex task i to increase production of this task
as well and require more low-skilled workers).
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
22 A Conceptual Framework
monopolistic competition set-up, with consumers having constant elasticity of
substitution–type preferences across product varieties. Industry expenditures
across varieties allow for an “outside good” that captures substitution with the
rest of the economy.
The model yields predictions of the impacts of firms’ technology adoption on
productivity, jobs, and wages, depending on the skill type of workers. When a
firm invests in ICT, it incurs a fixed cost, including all the costs associated with
installing the technology within the firm and adjusting the firm’s labor force—
costs such as training existing less-skilled and possibly also higher-skilled workers
to use the technology, training newly hired workers, and paying severance costs
to any displaced workers. By assumption, high-skilled workers can be combined
with the new-to-the-firm technology, which increases the firm’s productivity and
reduces variable costs of production. Low-skilled workers, however, are assumed
not to be combinable with the new technology, and only benefit to the extent
that the firm’s profits increase and therefore their wages also increase (the
empirical studies clarify that this is an extreme assumption, and that lower-
skilled workers also use these new technologies and become more productive as
reflected in higher wages over time).2
The model assumes that varying degrees of complementarity between ICT
and skilled workers are possible, including as perfect substitutes, whereby the
technology replaces skilled workers. The firm’s technology adoption decision
depends on whether the benefits of lower variable costs and higher profits out-
weigh the fixed adjustment costs of adoption. The model also assumes that
profits are proportionately higher for initially higher-productivity firms, given
that they are able to distribute the fixed cost over a larger output base. The
implication is that more productive larger firms invest in and adopt ICT, whereas
less productive smaller firms apply a less efficient combination of inputs. To the
extent that digital technology is a substitute for certain types of workers or tasks,
investment in ICT is associated with employment reallocation. Employment
protection regulation that makes labor adjustment costly is, therefore, expected
to negatively affect firms’ decisions to adopt ICT.
The model predicts that the impacts of ICT investments on productivity,
output growth, task complexity, and wages for both higher- and lower-skilled
workers can be positive. First, the productivity of technology-adopting firms
increases as the new-to-the-firm technology enables a more efficient combina-
tion of inputs, reducing variable costs. Second, and most important for jobs, firm
output increases because of a reduction in variable costs and prices. Although not
explored in this model, firm output can also expand through increases in product
quality and market outreach efforts, and the added quality and marketing
expenses are worthwhile because of the lower costs enabled by the ICT. In addi-
tion, the use of digital technologies such as the Internet allows the firm to acquire
better knowledge of existing and prospective customer tastes and to facilitate
faster product delivery and payment. Third, conditional on total output, the
participation of tasks carried out with the combination of high-skilled labor and
ICT increases, resulting in a greater share of more complex tasks. Finally, the
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
A Conceptual Framework 23
increase in profits that results from the ICT investment (because lower variable
costs outweigh the fixed adjustment costs) is shared with workers through the
fair-wage schedules, and therefore wages increase for both high-skilled and low-
skilled workers. A number of these predicted impacts are summarized with plus
signs in table 3.1.
The theoretical model also predicts ambiguous effects that can only be deter-
mined empirically. Importantly, table 3.1 also highlights that the impacts of
technology adoption on total jobs, high-skilled and low-skilled workers, firm-
level skill intensity (the relative share of high-skilled to low-skilled workers, or
“jobs gap”), and firm-level wage inequality (the ratio of high-skilled to low-skilled
worker wages, or “wage gap”) are ambiguous. Although all these impacts are not
clearly determined in one direction or the other, the model helpfully sheds light
on the factors on which the direction of these impacts depends. The model there-
fore significantly helps in delineating the relevant empirical and policy issues.
Whether firms’ technology adoption is inclusive in job outcomes depends on
the size of the output effects: lower-skilled workers will benefit if the output
effect dominates the substitution effects. The effect of ICT investments on both
low-skilled and high-skilled employment is ambiguous. As discussed above, when
a firm invests in ICT, output increases because of the reduction in variable costs,
working through a reduction in prices. The increase in total output increases the
output of all tasks, including both high-skilled and low-skilled tasks. However, at
the same time, three substitution effects operate against the employment of low-
skilled workers: fewer tasks are performed by low-skilled workers because they
are replaced by tasks that use ICT and high-skilled workers; the activity levels of
low-skilled (noncomplex) tasks are reduced in favor of the activity levels of the
now less expensive tasks that use ICT and high-skilled workers, conditional on
output; and output shifts to the firms that become relatively more productive
because of their relatively more intense use of both ICT and high-skilled workers.
Although these three substitution effects work in favor of high-skilled workers, a
fourth substitution effect—the possibility that ICT will replace high-skilled
labor—may work against them. On the other hand, ICT adoption may support
increased high-skilled employment insofar as they are net complements at the
level of task performance. So while the overall effects of ICT adoption on
employment are theoretically ambiguous, sufficiently strong output effects can
increase both high-skilled and low-skilled employment, while likely substitution
effects can increase the ratio of high-skilled to low-skilled labor.
Table 3.1 Predicted Impacts of Technology Adoption on Productivity, Jobs, and Wages
Employment Wages TFP
Total High-skill Low-skill Jobs gap Total High-skill Low-skill Wage gap (Labor productivity)
? ? ? ? + + + ? +
Note: “Jobs gap” refers to the predicted impact on the relative share of high-skilled to low-skilled workers. “Wage gap” refers to the predicted
impact on the ratio of high-skilled to low-skilled worker wages. TFP = total factor productivity, that is, how efficiently all measured inputs are being
used to produce a given level of output.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
24 A Conceptual Framework
The impact of technology adoption on the wages of high-skilled and low-
skilled workers will depend on their relative scarcity and bargaining power.
Whether firm-level wage inequality increases or decreases depends on the
extent to which the rent-sharing schedules are relatively more increasing in
profits for high-skilled or low-skilled workers, which, in turn, depends on fac-
tors such as the relative scarcity and bargaining power of higher- versus
lower-skilled workers. The extent to which the wages of higher- or lower-
skilled workers are more tied to firm performance will depend on factors
such as whether it is more costly for the firm to lose one or another type of
worker, which, in turn, could lead the firm to pay higher wages to secure the
workers’ commitment.
The size of output effects at the product market level depends on demand
elasticity and general equilibrium effects. Interfirm substitution effects arise
through the demand side, including parameters in the utility function that
govern output effects from substitution both from other competing firms
and from the rest of the economy. As part of product market–wide effects,
the output and employment of competitive rivals that do not adopt the new
technology will both decline. Although nonadopting rivals will lose output
and employment, the overall product market is expected to gain demand
since the assumed preferences imply substitution across categories.
Importantly, positive industry-wide output effects and associated inclusive
job effects are more likely the larger the elasticity of product market demand.
To the extent that total expenditures on the industry’s output are more elas-
tic because of substitution from other industries and regional or international
export possibilities, output expansion effects will be even larger. In addition,
positive employment spillover effects are possible from neighboring firms’
process innovation associated with ICT adoption.
The extent of firm-level ICT investments and the size of overall output
effects also depend on technology diffusion, product market, education, skills,
and labor market policies. Regarding product market policies, the responsiveness
of output expansion (as well as quality and market outreach efforts) to the low-
ered variable costs enabled by technology adoption depends on, among other
factors, the following:
• The intensity of domestic market contestability and competition (including entry,
expansion, and exit policies). For example, output will expand more if there is
a lower risk of losses from the possibility that any output expansion needs to
be retracted, which is influenced by, among other things, the ability of firms to
retrain or lay off workers at moderate cost, and by the ability of bankruptcy
protection to guard investors’ interests if retraction is necessary.
• Policies affecting product tradability across regions within the country and inter-
nationally for all potentially tradable products, including policies lowering
transportation and logistics costs, and critically including policies to lower
tariff and nontariff barriers to external trade.
• Access to finance to purchase required inputs and promotion efforts.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
A Conceptual Framework 25
• Policies supporting the upgrading of the quality of firm management and other fac-
tors affecting firms’ ability and know-how to enlarge production and distribu-
tion in response to lower variable costs.
Regarding education, skills, and labor market policies, firms will be better able to
pay the fixed costs of adjustment (technology setup costs, worker training, hiring
and firing costs) associated with ICT adoption with, among others, the
following:
• Education systems that deliver skills that are needed to adopt and use digital
technologies
• Worker training and job-search support to facilitate skills upgrading and higher
worker mobility
• Lower transactions costs for hiring and dismissing workers (flexible labor policies
to facilitate worker choices by firms).
Notes
1. Based on a sample of more than 26,000 manufacturing establishments across 71 coun-
tries (both Organisation for Economic Co-operation and Development countries and
developing countries) using World Bank Enterprise Survey data collected between
2002 and 2006, Dutz et al. (2012) find that technology-adopting innovating firms
have higher employment growth rates and employ a higher share of low-skilled and
female workers than noninnovating firms, with country and industry fixed effects, as
well as firm-level controls by ownership, level of organization and legal status, size
class, and age group. However, because of data limitations, that paper could not address
issues of causality nor study the implications of technology adoption for relative earn-
ings, income growth, and attainment of higher skills. Accordingly, that paper motivated
the search for the more complete data sets and their analyses that have allowed this
book to address the broader scope of the implications of technology adoption.
2. Acemoglu and Restrepo (2017) explore the impact of industrial robots on employ-
ment and wages in a model in which lower-skill tasks are completely automated and
replaced rather than partially displaced through substitution effects. Based on their
estimates from U.S. labor markets between 1990 and 2007, one more robot per thou-
sand workers reduces the employment-to-population ratio by about 0.18–0.34 per-
centage point and wages by 0.25–0.5 percent.
References
Acemoglu, Daron, and David Autor. 2011. “Skills, Tasks and Technologies: Implications for
Employment and Earnings.” Handbook of Labor Economics, volume 4, edited by David
Card and Orley Ashenfelter, 1043–171. Amsterdam: Elsevier.
Acemoglu, Daron, and Pascual Restrepo. 2016. “The Race between Machine and Man:
Implications of Technology for Growth, Factor Shares and Employment.” Working
Paper 222521, National Bureau of Economic Research, Cambridge, MA.
———. 2017. “Robots and Jobs: Evidence from US Labor Markets.” Working Paper 23285,
National Bureau of Economic Research, Cambridge, MA.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
26 A Conceptual Framework
———. 2018. “Artificial Intelligence, Automation and Work.” In Economics of Artificial
Intelligence, edited by Ajay Agarwal, Avi Goldfarb, and Joshua Gans. Forthcoming.
Aghion, Philippe, and Peter Howitt. 2009. The Economics of Growth. Cambridge, MA: MIT
Press.
Autor, D. H. 2015. “Why Are There Still So Many Jobs? The History and Future of
Workplace Automation.” Journal of Economic Perspectives 29 (3): 3–30.
Autor, D. H., and D. Dorn. 2013. “The Growth of Low-Skill Service Jobs and the
Polarization of the US Labor Market.” American Economic Review 103 (5): 1553–97.
Autor, D. H., and A. Salomons. 2017. “Robocalypse Now: Does Productivity Growth
Threaten Employment?” Conference Paper prepared for the European Central Bank
Sintra Forum on Central Banking, June 26–28.
Cirera, Xavier, and William Maloney. 2017. The Innovation Paradox: Developing-Country
Capabilities and the Unrealized Promise of Technological Catch-Up. Washington, DC:
World Bank.
Comin, Diego, and M. Mestieri Ferrer. 2013. “If Technology Has Arrived Everywhere, Why
Has Income Diverged?” Working Paper 19010, National Bureau of Economic
Research, Cambridge, MA.
Dutz, Mark, Ioannis Kessides, Stephen O’Connell, and Robert Willig. 2012. “Competition
and Innovation-Driven Inclusive Growth.” In Promoting Inclusive Growth: Challenges
and Policies, edited by Luiz de Mello and Mark Dutz. Paris: Organisation for Economic
Co-operation and Development.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
CHAPTER 4
New Lessons from the
Region on the Impacts of
Technology Adoption
Based on the predictions of the conceptual model summarized in the preceding
chapter, the impacts of technology adoption on the total number of jobs, on
the jobs of high-skilled and low-skilled workers, on firm-level skill intensity
(“jobs gap”), and on firm-level wage inequality (“wage gap”) are theoretically
ambiguous. These are, therefore, empirical issues that need to be addressed
by applying economic theory to the available data. The first section of this chap-
ter presents information about the impacts of technology adoption on skills,
types of jobs, wages, and productivity based on six new firm-level studies of
Argentina, Brazil (two studies), Chile, Colombia, and Mexico. The second sec-
tion presents information from these studies about the impacts of technology
adoption on job dynamics and the role of complementary training investments.
The third section presents information on the effect of labor market regulations
on firms’ decisions and jobs outcomes based on studies exploring both firm-level
data and cross-country data. Finally, the fourth section presents information on
the impacts of technology adoption on firms and workers through trade and
labor mobility based on two new cross-country studies exploring household
data. Across these sections, the nine new empirical studies clarify how the
effects of digital technology adoption by firms have played out across different
Latin American countries.1
The country studies identify causal effects rather than correlations by focus-
ing on drivers that are exogenous to output and the demand for jobs, skills, and
labor earnings. Firm fixed effects are used to control for unobservable
time-invariant firm heterogeneity. This is important because unobserved firm
characteristics, such as the quality of products, firms’ commercial ties, and the
professional background of top-tier managers, might simultaneously affect the
propensity to adopt new-to-the-firm technologies and the outcome variables.
The effects of adoption on productivity and jobs-related outcomes therefore are
identified using within-firm changes over time and not the cross-section varia-
tion of heterogeneous firms across industries. Technology adoption could also
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 27
28 New Lessons from the Region on the Impacts of Technology Adoption
be endogenous to unobservable time-variant factors such as technology, produc-
tivity, or labor market shocks. These shocks could simultaneously affect the
scale and skill mix of labor demand as well as information and communication
technologies (ICT) adoption among firms. A number of the studies exploit
plausibly exogenous subnational changes in the availability of ICT access or in
its quality over time and space. These exogenous variations are exploited as
instruments for the otherwise possibly endogenous firm-level use of ICT. For
example, the Brazil tasks and labor policies impact study exploits the arrival of
Internet access at the municipal level over time and compares skill outcomes in
municipalities with different degrees of exposure to digital technology due to
sectoral and technological characteristics. In Chile, the analysis allows the adop-
tion of complex software over time to affect firms differently depending on
preexisting subnational technological intensity (based on the subnational share
of households with access to computers). The Mexico study interacts average
sector ICT intensity in the United States with the average elevation of munici-
palities to reflect the geographical challenges of Internet availability in more
difficult-to-access areas.
Table 4.1 Empirical Impacts of Technology Adoption on Jobs, Wages, and Productivity
Employment Wages TFP
High- Low- Jobs High- Low- Wage
Country Years Industries Focus variable Total skill skill gap Total skill skill gap (LP)
Argentina 2010–12 Manufacturing Investment in + + + + + + + + (+)
ICT capital
Brazil 2000–14 Economy-wide Percent 0 − − − − 0 − +
sectoral Internet
impacts availability
Tradables Percent + + 0 0 + 0
Internet
availability
Nontradables Percent − − − − − 0
Internet
availability
Brazil tasks 1996–06 Economy-wide Internet 0 + − +
and availability ×
labor labor market
policies regional
enforcement
Chile 2007–13 Economy-wide Complex + 0 + −
software use
Colombia 2008–14 Manufacturing High-speed + + + +
Internet use
Mexico 2008–13 Manufacturing Internet use + + + + + + + − +
Services Internet use + + + 0 − − − − +
Commerce Internet use + + + 0 − 0 − + +
Note: ICT = information and communication technologies. “High” refers to high-skilled (or white-collar) workers, “Low” refers to low-skilled
(or blue-collar) workers. Productivity measures are based on TFP (total factor productivity) wherever possible; labor productivity is used elsewhere
(shown in parentheses). +, −, and 0 refer to adoption resulting in an increase, decrease, or no effect on the specified variable. Blank cells reflect the
inability to calculate the effect on the specified variable. All studies are at the firm level except for the Brazil studies, which are at the municipal
level (except for the reported Brazil TFP estimate, which is based on firm-level manufacturing data).
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
New Lessons from the Region on the Impacts of Technology Adoption 29
Impact on Firm Productivity and the Demand for Jobs,
Types of Skills, and Wages
Across all countries except Brazil, ICT adoption by firms is associated
with increases in total employment and in employment of low-skilled labor
(table 4.1). In line with the predictions of the theoretical model, the productivity
of technology-adopting firms increased in all country studies where data were
available, with the finding in Argentina based on labor productivity. The total
number of jobs grew in most countries as a result of ICT adoption, with the only
exception being Brazil. Although there was no positive economy-wide net effect
on jobs in the Brazil sectoral impacts study with both contemporaneous and
lagged effects, jobs increased in tradable sectors with output expansion opportu-
nities, in line with the book’s conceptual framework; and the Brazil tasks and
labor policies impact study found that though the total number of jobs was
reduced, it was reduced by less among firms in the tradable sectors. It is likely
that policy distortions, including high trade and other product market expansion
barriers, played a more important role in constraining opportunities for efficient
global output expansion in Brazil than in other countries. In Argentina, firm
employment increased by 60 percent, on average, because of investment in ICT.
Employment rose by 50 percent for low-growth firms and by 72 percent for
high-growth firms. This result is compatible with the predicted output effect in
which the increase in jobs is driven by the increased output. Importantly, jobs
growth was inclusive, in the sense that increases in lower-skilled jobs were
observed in Argentina, Chile, Colombia, and Mexico. This empirical finding
across all country studies except Brazil on the link between technology adoption
and increased employment of lower-skilled workers is an important empirical
result documented by this book, since it is theoretically ambiguous and relates to
the main question of whether technology adoption can drive inclusive growth.
The extent to which the impact of ICT adoption on jobs is inclusive depends
on whether the product market provides opportunities for significant output
expansion and on the nature of the technology adopted. Most of the findings
where the impact of technology adoption is inclusive—that is, it boosts lower-
skilled jobs—are in tradable manufacturing industries. In the Argentina study,
manufacturing firms that invested in increasing their stock of ICT capital
experienced job increases for low-skilled as well as high-skilled workers.
Importantly, job increases for the low-skilled were larger for high-growth firms,
supporting the importance of strong output expansion effects in driving inclusive
growth: the increase in employment of low-skilled workers is 10 percent larger for
firms in which revenue growth is above the median. In the Brazil sectoral impacts
study, where skills were split into thirds (high skill, middle skill, and low skill), the
jobs for middle-skilled workers increased only for the tradable manufacturing sec-
tor, even though there was no change in the bottom third, with middle-skilled
workers in other nontradable sectors such as wholesale and retail trade losing jobs.
In Colombia, firms’ use of high-speed broadband directly increased demand for
laborers and production (lower-skilled) workers as well as for professional
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
30 New Lessons from the Region on the Impacts of Technology Adoption
(higher-skilled) workers. In the Mexico study, a 10-percentage-point increase
in the share of labor using the Internet in manufacturing was associated with an
11 percent increase in white-collar workers together with a 6 percent increase in
blue-collar workers, with much smaller positive effects for the less tradable com-
merce industries.2 Finally, regarding the effect of the nature of the adopted tech-
nology, the Chile study found that the adoption and use of complex software by
firms increased the number of low-skilled production worker jobs in the medium
term. Over the six-year period analyzed, the levels of employment of high-skilled
production workers and managers did not change significantly.
Greater investment and use of the Internet are also associated with a relatively
larger increase in the use of high-skilled workers. The assumption of skill-biased
technological change included in the conceptual framework is borne out in most
of the country-level studies. Even though the number of low-skill jobs increases
across most studies, a substitution effect occurs in favor of more skilled workers
with increased use of the Internet. This increased job demand gap in favor of
more skilled workers is found in the Argentina, Brazil tasks and labor policies
impact, and Mexico (for manufacturing industries) studies. In the manufacturing
industries in Mexico, for instance, the number of workers increased for both
white- and blue-collar occupations, with the increase being larger for white-
collar, therefore increasing the ratio of white- to blue-collar workers. Consistent
with this pattern, evidence from studies beyond those commissioned for this
book shows that the distribution of jobs in some countries in Latin America and
the Caribbean (LAC), including Guatemala, Honduras, and Panama, has become
increasingly polarized with a rise in the number of very-high-skill and low-skill
occupations, but with fewer jobs in middle-skill occupations (World Bank 2016).
Technology adoption is also likely to affect the task content of occupations
and the degree to which they involve more- or less-routine or cognitive versus
manual tasks. Firms’ access to digital technologies may lead to greater demand
for occupations that perform less-routine and less-manual tasks because some
of these tasks can be more easily automated. However, whether and how digi-
tal technology adoption is actually affecting different types of jobs in LAC
through the automation of some tasks remains an open question, with critical
implications for education and training systems.
The Brazil tasks and labor policies study and the Chile study go beyond the
more traditional classification of workers and assess impacts of digital technology
adoption on the skills content of occupations. These two studies exploit panel
data sets at the firm level and match firm-level occupations with detailed mea-
sures of skills’ task content, that is, abstract, routine-cognitive, routine-manual,
and manual tasks. The Brazil study relies on a unique concordance between the
Brazilian Classification of Occupations and the U.S. Department of Labor’s
Occupational Information Network to assign a numerical index capturing the
importance of distinct activities in each occupation. For Chile, the concordance
relies on the 2014 Program for the International Assessment of Adult
Competencies survey, collected by the Organisation for Economic Co-operation
and Development. Based on these data, measures of cognitive, manual, routine,
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
New Lessons from the Region on the Impacts of Technology Adoption 31
and nonroutine tasks are constructed for each occupation. Drawing on Acemoglu
and Autor (2011) and Autor and Handel (2013), the task content of occupations
using routine and nonroutine tasks in each occupation is defined. The work fur-
ther distinguishes routine and nonroutine tasks into routine-manual, routine-
cognitive, nonroutine-manual, and nonroutine-cognitive tasks. The Brazil study
also disaggregates nonroutine, cognitive tasks into tasks requiring communica-
tion, analytical, and social and emotional skills, following Deming (2015).
In Chile, the adoption of advanced software at the firm level is associated, in
the medium term, with a reallocation of employment away from abstract tasks
and toward more routine and manual tasks. The study exploits the relationship
between each firm’s complex software adoption and a measure of workforce skill
composition. It instruments the firms’ adoption of complex software with a
measure of household use of computers at the subnational level and exploits the
fact that industries vary in their degree of reliance on ICT: industries that inten-
sively use technology and that are located in regions where there is a greater
availability of computers and related technologies are most likely to adopt com-
plex software. The findings suggest that the adoption of complex software
increases the use of routine and manual tasks and decreases the use of abstract
tasks. These results are driven by the composition of employment. Even though
firms expand their overall activity, the share of high-skilled labor decreases
following the adoption of complex software.
In Brazil, the subnational rollout of Internet access nationwide is strongly asso-
ciated with a shift away from more routine and manual tasks and toward nonrou-
tine and cognitive tasks. The Brazil study’s methodology exploits the variation in
firms’ use of different types of skills over time as Internet access is rolled out,
comparing localities with more- versus less-technology-intensive industries. The
results show that in the aftermath of the technology shock, the composition of the
workforce changes toward more nonroutine skills within each of the Brazilian
municipalities. In addition, technology adoption is found to be strongly associated
with a change in workforce composition toward more cognitive tasks and away
from manual tasks. This finding validates, for a large middle-income country, some
of the concerns that routine, manual tasks are increasingly being replaced by tech-
nology, displacing less-skilled workers (Autor, Levy, and Murnane 2003).
Whether the effect of ICT adoption on wages is inclusive also depends on
whether firms invest in human capital—that is, in workers. Inclusivity of wages
within technology-adopting firms can refer to increasing wages for lower-skilled
workers, a reduction in wage inequality (or the wage gap), or both. In line with
the conceptual framework’s prediction, the wages of low-skilled workers
increased in Argentina, in Brazil (where wages increased for the middle-skilled
group in manufacturing), and in Mexico (in the manufacturing sector). And
importantly, the wage gap actually decreased in Mexico in the manufacturing
and services industries, which addresses another question that was theoretically
ambiguous. In the Mexico study, a 10-percentage-point increase in Internet
use in manufacturing was associated with a 14 percent increase in white-collar
wages versus a 16 percent increase in blue-collar wages. So ICT adoption can
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
32 New Lessons from the Region on the Impacts of Technology Adoption
narrow wage inequality between types of workers. Based on additional analyses
of Mexico’s national survey of enterprise ICT use, the higher wage increase for
low-skilled workers appears to be explained by the increased sophistication of
these workers over time as they learn on the job: as information becomes more
available in adopting firms through the use of business management software for
specific activities or more integrated enterprise resource planning systems, it
likely becomes easier for low-skilled workers to make more informed and decen-
tralized decisions.3 Instead of being substituted for by ICT, these workers appear
to be growing into stronger complements of technology.
Finally, whether the impact of technology adoption on wages is inclusive also
depends on the sector. The Mexico study highlights the importance of sectoral
effects. Contrary to what is found for manufacturing, wages in services industries
decrease for both groups, but the reduction is higher for white-collar workers,
thus decreasing the wage gap for services as well. These differences could be
explained by the fact that jobs in services appear to be comparatively more at risk
of automation than jobs in manufacturing (World Bank 2016). Therefore, as rou-
tine and nonroutine tasks within the services sector are automated, the skill value
of both white- and blue-collar workers who perform these tasks is reduced, but
the effect is stronger on white-collar workers, thus reducing the wage gap. The
commerce sector also has a higher risk of automation, but the wage reduction
affects blue-collar workers while the wages of white-collar workers do not change
significantly. The wage reduction could indicate that these types of workers are
performing simpler, less valuable tasks as the use of technologies allows some
processes to be at least partially automated.
Impacts on Job Dynamics and the Role of Complementary
Investments in Skills
In addition to the impacts on employment and skills levels, access to ICT also
affects the dynamics of job creation. Latin America is characterized by relatively
low levels of job creation (Alaimo et al. 2015). Access to ICT can create pres-
sure for firms to adjust production processes, creating new occupations and
destroying others, and changing the skills content of occupations. In Argentina,
evidence shows that investment in ICT has had an impact on employment
turnover. Firms that invest in ICT capital report that in roughly 5 percent of
cases, it led to replacing workers; in 10 percent, it led to replacing occupations;
and in 32 percent, it led to the creation of new occupations. The most relevant
characteristic associated with all three forms of changes in employment is the
firm’s undertaking of operations through the Internet. The Brazil tasks and labor
policies impact study also provides evidence that this dynamism in job creation,
at least in the short term, is accompanied by job destruction through the exit of
firms from the relevant market. During the period from 1996 to 2014, access to
the Internet is correlated with a reduction in the number of firms. This finding
is consistent with reduced variable costs allowing more efficient firms to expand
while driving others out of the market.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
New Lessons from the Region on the Impacts of Technology Adoption 33
ICT adoption at the firm level is also strongly correlated with firms’ comple-
mentary investments in the human capital of the workforce. The Mexico study
finds that firms that make more intensive use of ICT also provide more training
for both white- and blue-collar workers, with blue-collar workers receiving sig-
nificantly higher levels of training. These blue-collar workers likely become more
sophisticated and increase their use of digital technologies because of increased
access to information as firms make organizational improvements by using enter-
prise resource planning systems more intensively and by increasing training for
workers in these technologies. In Chile, the adoption of complex software by
firms is also associated with increased investment in technical, on-the-job train-
ing for managers, with a focus on digital skills.
The human capital of managers is also an important factor driving firms’
investment in ICT in the region. Formal education and work experience serve as
a proxy for managers’ human capital. In Argentina, the propensity to invest in
ICT is higher for firms with younger, experienced, and more educated managers
(managers with college or graduate degrees are 15–26 percentage points more
likely to invest in ICT). In Chile, the evidence shows that younger managers with
more formal education and past relevant experience are also more likely to invest
in technology adoption.
The Role of Labor Market Regulations on Firms’ Decisions and
Jobs Outcomes
The constraints on firms’ human capital management decisions from de jure
labor regulations also influence the extent of digital technology use in the region.
Evidence from a country-level analysis of a large sample of very diverse countries,
including almost all the countries in LAC, shows that the prevailing labor market
regulatory instruments—and other social protection institutions—are signifi-
cantly associated with firms’ adoption of digital technologies.4 Importantly, the
significance and direction of the relationship vary across regulatory instruments
such as statutory minimum wages, restrictions on hours and hiring, dismissal
procedures, severance costs, and contributions for social insurance.
A higher statutory minimum wage is positively associated with the extent of
digital technology use by firms. Two obvious economic factors help explain this
observed relationship. First, a higher statutory minimum wage is more likely to
bind, particularly at the lower end of the labor force distribution, by skill and
productivity levels. Therefore, the higher minimum wage could increase firms’
incentives to invest in digital technology that substitutes for labor inputs and thus
saves on labor costs. Second, faced with a policy-mandated lower bound on
wages, firms will have an incentive to make investments that pair labor more
efficiently with ICT tools to raise the marginal product of labor well above the
mandated minimum wage.
However, more restrictions on firms’ employment decisions are associated
with lower levels of digital technology use. Several measures of labor market
restrictiveness were used to analyze the relationship, including indices
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
34 New Lessons from the Region on the Impacts of Technology Adoption
constructed by the International Labour Organization and the Organisation for
Economic Co-operation and Development (and expanded to cover most coun-
tries in the LAC region by the Inter-American Development Bank), along with
the World Bank’s Doing Business data. Where these indices were found to be
statistically significant, a more restrictive regulatory framework for firms’ human
resource decisions was strongly associated with lower use of digital technology.
As with the level of the statutory minimum wage, the possible economic expla-
nation for this statistical relationship is easy to understand. All other things being
equal, more onerous procedures for firms’ choices of productive inputs impede
businesses’ ability to adopt and adjust to new technologies. If firms are more
constrained in their human resource decisions, they will find it more difficult to
embed new technologies into their production models, to adopt the processes
that the new technologies entail, and to find the complementary labor and
human capital they require to reach a new optimal level of operation.
Important nuances become apparent when different forms of employment
protection measures (that is, restrictions on hours of work, restrictions on the use
of temporary employment, and procedures and monetary costs of dismissals) are
analyzed separately. Cumbersome dismissal procedures appear to have the stron-
gest significant association with firms’ use of digital technology: business use of
digital technology is lower in countries that require firms to follow more burden-
some procedures to dismiss workers. Again, there are intuitive economic expla-
nations for why different forms of employment protection regulations could
relate differently to firms’ digital technology use. On the one hand, restrictions
on hours and limits on the use of fixed-term and temporary workers can con-
strain a firm’s ability to experiment and adapt to new technology and changes to
its production function. On the other hand, firms and workers might welcome
the certainty of an up-front dismissal payment (severance) to workers displaced
by technology to speed the adjustment process. The more cumbersome the
bureaucratic procedures that the labor code requires firms to follow to dismiss
one or more than one worker, the greater firms’ uncertainty about what the total
adjustment costs of technology adoption will be. The uncertainty created by
cumbersome de jure dismissal procedures is likely to be a formidable deterrent
to adopting new technology.
Yet what is written in the labor code is not always what is enforced in practice
by inspectors and hence what firms and workers react to. The gap between de jure
and de facto labor regulations is particularly wide in LAC economies, where many
firms and workers operate beyond the reach of the government’s capacity to
enforce regulations. The so-called informal economy— unobserved, unregulated,
and thus untaxed—is extraordinarily diverse, and is made up of a wide array of
firms and people (Perry et al. 2007). The advent and widespread adoption of digital
technology can simultaneously create opportunities for avoidance and evasion of
restrictive regulations and increase the ability of government agencies to observe
economic activity and enforce compliance. But a large and diverse informal
economy— sometimes tightly integrated with the formal economy through the
production chains of even large, international, formal firms—also offers the sort of
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
New Lessons from the Region on the Impacts of Technology Adoption 35
variation in the actual incidence and costs of labor regulation that is critical to test-
ing its actual impact on workers’ prospects in the wake of a technology rollout. The
Brazil tasks and labor policies impact study exploits just such an opportunity.
The enforcement of labor regulations in Brazil limits the degree to which
companies shift away from labor as technology becomes available. In Brazil, as in
all the LAC countries, labor market regulations exist to protect workers from
unanticipated shocks. But often these labor market protections also have the
unintended consequence of increasing businesses’ labor costs. The trade-off
between job security for workers, on the one hand, and economy-wide productiv-
ity and growth, on the other hand, is arguably one of the most prominent public
policy debates in the region. Until reforms to its labor code passed in the first half
of 2017, Brazil was an outlier, particularly in the restrictions it imposes on the use
of fixed-term, temporary, and outsourced employment. The Brazil tasks and labor
policies impact study exploits administrative data—specifically, changes in the
incidence of labor market inspections led by the Ministry of Labor—that capture
within-country and time variation in the de facto enforcement of de jure labor
market regulations. The evidence shows that where they are enforced, stringent
labor regulations constrain Brazilian firms’ flexibility in the short term, as employ-
ment falls by more—both for routine and nonroutine tasks—in municipalities
with looser enforcement compared with those with stricter enforcement.
In contrast to policy intentions, the evidence shows that strict labor market
regulations in Brazil differentially benefit the more skilled workforce, particularly
those workers employed in nonroutine and more cognitive tasks. Aside from
affecting the level of employment, the Brazil tasks and labor policies impact
study also shows that enforcement of labor regulations is found to significantly
affect the composition of employment at the local level. Specifically, the results
show that in localities with more stringent enforcement of regulations, higher-
skilled workers who are performing nonroutine tasks are disproportionally pro-
tected following technological adoption. Hence, when adopting ICT, stricter
labor protections have the perverse effect of helping the nonroutine workers
more because they result in a greater increase in the cost of nonroutine tasks. This
evidence is consistent with evidence from studies of regulations in other coun-
tries showing redistributive effects of labor regulation that work against the
interests of certain groups of workers, such as young workers and women
(Montenegro and Pagés 2003; Betcherman 2012).These relatively greater
impacts on tasks with more nonroutine and cognitive content may indicate that,
in addition to affecting efficiency, employment protection regulations in Brazil
may have important redistributive consequences.
Impacts of Technology on Firms and Workers through Trade and
Labor Mobility
When firms access the Internet, they reach larger and more diverse markets at
lower cost. Gaining access to international trade through online platforms can
have an impact on the wages of their workers, and particularly on the wage gap
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
36 New Lessons from the Region on the Impacts of Technology Adoption
between high-skilled and low-skilled workers. Exporting through traditional
physical channels can be very costly and out of reach for many microenterprises
and small firms in developing countries that lack a suitable workforce. Online
platforms, which serve as an international marketplace matching buyers and sell-
ers, provide them with a quick and cost-efficient way of reaching a larger number
of new customers in foreign countries. Using a standard gravity model for bilat-
eral trade flows on the e-Bay platform, the background study on the impact of
online platforms and the skills premiums in wages shows that a 1 percent
increase in online exports lowers the wage gap by about 0.01 percent. This effect
on the wage gap is driven by countries with a larger proportion of small firms,
more precisely by countries above the median of 44 percent share of employ-
ment in small firms (those with fewer than 10 employees). The decrease in fixed
costs attributable to accessing online export platforms disproportionately bene-
fits smaller firms that also tend to be more low-skill intensive.5 As the region
looks for ways to spur productivity while ensuring social gains, increased access
to online trade in LAC can be an additional mechanism through which the
reduction in the wage gap observed since the early 2000s can be sustained.
Access to ICT in the LAC region has also supported greater labor mobility
and contributed to lower wage inequality. An additional way that digital tech-
nologies expand economic opportunities is by reducing information costs and
thereby allowing for a better allocation of resources. In the LAC region this is an
especially important link because the region has traditionally suffered from rela-
tively low labor mobility across sectors and regions—the typical Latin American
worker has had somewhat “sticky feet” (Hollweg et al. 2014). Exploiting house-
hold data across several LAC countries and other comparators, the background
study on labor mobility adjustment costs highlights several interesting findings.
First, on average, sector labor mobility costs are higher than regional mobility
costs. The average cost of moving across both sectors and regions (that is, internal
migration) is about 1.8 times the average annual real wage, which is higher than
the cost of moving only across sectors or across subnational regions. Second, in
poorer LAC countries, such as El Salvador, Guatemala, and Honduras, workers
face higher mobility costs than do workers in middle-income countries, such as
Brazil, Costa Rica, and Mexico. Exploiting cross-country variations, the study also
suggests that labor mobility costs are partially driven by information asymme-
tries and that improved access to the Internet in the region could substantively
mitigate these costs.
Notes
1. See appendix A for a list of these studies.
2. Interestingly, increases in the share of blue-collar workers in the services industries are
higher than those in the manufacturing industries, and higher than the increase in
the number of white-collar workers: for services, a 10-percentage-point increase in the
share of labor using the Internet was associated with a 7 percent increase in white-
collar workers, together with an 11 percent increase in blue-collar workers.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
New Lessons from the Region on the Impacts of Technology Adoption 37
3. Enterprise resource planning is the integrated management of core business processes,
often in real time and mediated by ICT. These business activities can include business
intelligence, product and production planning, procurement, manufacturing or service
delivery, distribution, marketing, sales and customer service, materials and inventory
management, human resource management, accounting, and finance.
4. In a background study for this book, Packard and Montenegro (2017) expand on
Alesina, Battisti, and Zeira’s (2015) analysis using country-level data. The authors
exploit greater economic and regulatory variation across a larger and more diverse
sample of countries. The analysis unpacks labor regulation into separate components
that are likely to shape firms’ decisions differently. Whereas Alesina, Battisti, and Zeira
(2015) analyze regulatory instruments such as statutory minimum wages, employ-
ment protection, and the power of unions separately, the background paper for this
book uses four different measures that separately capture (1) the rigidity of working
hours; (2) restrictions on the use of temporary, fixed-term contracts; (3) the proce-
dural difficulty of dismissing workers; and (4) mandated payments to workers upon
dismissal (such as employer-paid severance). The paper also captures other statutorily
mandated nonlabor costs, specifically contributions required of employers and
employees for social insurance (old age pensions; disability, survivor, and unemploy-
ment benefits; and health coverage).
5. On average, in the sample in the background paper, 25 percent of workers in small
firms with fewer than 10 employees are considered to be skilled (at least a complete
secondary education) versus 49 percent in larger firms with more than 50 employees.
The larger share of skilled workers in large relative to small firms is observed in all
countries in the sample.
References
Acemoglu, Daron, and David Autor. 2011. “Skills, Tasks and Technologies: Implications for
Employment and Earnings.” In Handbook of Labor Economics, volume 4, edited by
David Card and Orley Ashenfelter, 1043–171. Amsterdam: Elsevier.
Alaimo, Verónica, Mariano Bosch, David S. Kaplan, Carmen Pagés, and Laura Ripani.
2015. Jobs for Growth. Washington, DC: Inter-American Development Bank.
Alesina, A., M. Battisti, and J. Zeira. 2015. “Technology and Labor Regulations: Theory
and Evidence.” Working Paper 20841, National Bureau of Economic Research,
Cambridge, MA.
Autor, D. H., and M. J. Handel. 2013. “Putting Tasks to the Test: Human Capital, Job Tasks,
and Wages.” Journal of Labor Economics 31 (2): 59–96.
Autor, D. H., F. Levy, and R. J. Murnane. 2003. “The Skill Content of Recent Technological
Change: An Empirical Exploration.” Quarterly Journal of Economics 118 (4):
1279–333.
Betcherman, G. 2012. “Labor Market Institutions: A Review of the Literature.” Policy
Research Working Paper 6276, World Bank, Washington, DC.
Deming, D. 2015. “The Growing Importance of Social Skills in the Labor Market.”
Working Paper 21473, National Bureau of Economic Research, Cambridge, MA.
Hollweg, C. H., D. Lederman, D. Rojas, and E. R. Bulmer. 2014. Sticky Feet: How Labor
Market Frictions Shape the Impact of International Trade on Jobs and Wages. Directions
in Development. Washington, DC: World Bank.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
38 New Lessons from the Region on the Impacts of Technology Adoption
Montenegro, Claudio E., and Carmen Pagés. 2003. “Who Benefits from Labor Market
Regulations? Chile 1960–1998.” Working Paper 4126, Inter-American Development
Bank, Washington, DC.
Perry, Guillermo, William F. Maloney, Omar S. Arias, Pablo Fajnzylber, Andrew D. Mason,
and Jaime Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. World Bank Latin
American and Caribbean Regional Study. Washington, DC: World Bank.
World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, DC:
World Bank.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
CHAPTER 5
Improving the Environment
for Technology Adoption
with Inclusion
Countries in the Latin America and the Caribbean (LAC) region have much to
do to encourage greater adoption of digital technology. When assessed using
benchmarks of the readiness of the business environment, LAC countries appear
to be the least ready to support businesses’ taking greater advantage of digital
technology to boost productivity. Figure 5.1 benchmarks countries in LAC and
other regions according the World Economic Forum’s Networked Readiness
Index subindex of the business environment. This subindex is constructed using
18 indicators that cover product market policies (such as intensity of local com-
petition and ease of starting a business and enforcing contracts), skills and labor
market policies (including tertiary education and quality of management
schools), and technology generation and diffusion support policies (including
availability of the latest technologies; government procurement of advanced
technologies; laws related to information and communication technologies
[ICT], such as e-commerce and intellectual property protection; and venture
capital availability). Countries in the LAC region score, on average, lower than
any other of the World Bank Group’s operational regions, though the differences
between some of these regions is not statistically significant. Importantly, there is
considerable variation within the region: Chile is in first place according to this
index, but Brazil and Argentina are in 14th and 15th places of the 20 LAC coun-
tries for which data are available. This benchmarking suggests that there is enor-
mous potential for reforms to the business environment in LAC countries to
encourage greater adoption of digital technologies, along with the boost to pro-
ductivity and inclusive outcomes that doing so can generate.
Policies should be oriented toward facilitating the diffusion and adoption of
technologies and sharpening incentives for output expansion. The empirical
work commissioned for this book, along with previous studies, shows that countries
can take important policy steps to improve the business environment for tech-
nology adoption, with the aim of increasing productivity and inclusive growth.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 39
40 Improving the Environment for Technology Adoption with Inclusion
Figure 5.1 LAC Holds Last Place in the Business Environment Related to Digital Technologies
7
6
NRI Subindex, 1–7
5
4
3
2
1
0
Ni H B
ra ti
Bo ua
Ar agu ia
nt y
Gu B na
La E ma l
tin cu la
Ho In a
du ia
in d P as
n b u
pu o
ya c
hi a
an S e na
Ce th o
an C tral sia
No ta ia
W P Afr a
or a ica
e a
m e
u a
st y
Ko C sia
a, le
p.
El me or
Sa Co lvad a
ha lo or
n ia
e i
Gu bli
ge a
at raz
Ea gua
ric
n
c
av m
Ur aic
Sa ric
Ja rag
R
Re ag
d ou xic
ica T er
ca ai
Pa liv
d
d os As
ra mb
re hi
Re
A ad
r
rth Ri
g
i
n A
A
a,
ld na
Af
C
el
o
M
n
r
zu
ne
m an
Ve
Do ad
b-
id
Su
pe
st
in
Ea
Tr
ro
Eu
e
dl
id
M
Country score Regional average NRI score
Source: World Economic Forum 2017.
Note: NRI = Networked Readiness Index.
First, policies should support extensive and rapid Internet rollouts, including
enabling markets for the competitive and high-quality provision of Internet access
and the diffusion and adoption of other digital technologies—as well as other pro-
ductivity-enhancing technologies. Second, governments should implement product
market policies that enhance opportunities and sharpen incentives for output
expansion in response to the productivity increases that technology adoption yields.
Governments should support education and job skills structures that are more
likely to ensure that the available skills of individuals in the labor market are
complementary to adopted technologies and support workers with increased
mobility patterns. Beyond solid foundational skills, new evidence in this book
shows that digital technology adoption complements skills that are easily trans-
ferable across jobs and occupations, including higher-level cognitive and analyti-
cal skills, technical skills (for example, in ICT), and socioemotional skills (such as
interpersonal skills). Critically, the education and training systems for the 21st-
century worker need to include support for retraining throughout life and
for skills renewal, both within firms and within industries, which implies a
renewed focus on solid foundational skills. Rather than employment protection
designed to make transitions and adjustments difficult, labor policies should be
recrafted to assume that more frequent disruption and the need for change will
occur and to support the necessary changes.
More detailed policy recommendations emerge from analytical findings,
and are presented in the remainder of this section. Notwithstanding these over-
arching policy directions, the results of the more detailed analyses commissioned
for this book point the way for addressing more granular and perhaps more
manageable policy questions that can be the subject of further research efforts.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Improving the Environment for Technology Adoption with Inclusion 41
Technology Diffusion Support Policies
Policies to facilitate technology diffusion, adoption, and greater use—including
digital technology policies to support high-quality and competitively priced
Internet rollouts—are essential.1 The Internet is the oxygen on which digital
technologies thrive. Countries can do much more to support Internet rollouts,
including by providing procompetitive support for higher-speed broadband roll-
out regimes.2 Current adoption of ICT across the LAC region is highly heteroge-
neous and lags behind comparators in the Organisation for Economic Co-operation
and Development (OECD), implying that there is still much potential for
significant additional adoption in LAC and for the expected accompanying
advances in productivity.
Additional tariffs and taxes on ICT may be holding back per capita GDP
growth by 1 percentage point or more per year. Tariffs and taxes on ICT goods
and services do not just dissuade businesses from adopting digital technologies,
they also may prevent those businesses from existing at all. ICT tariffs and addi-
tional ICT taxes can add significantly to consumer and business costs of ICT goods
and services (Miller and Atkinson 2014) (map 5.1). Brazil, Argentina, the
Dominican Republic, Ecuador, and Jamaica are all in the top 20 of the 125 coun-
tries for which the extra costs that government imposes have been calculated,
with Brazil’s tariffs adding 16 percent and special taxes adding an additional 5
percent to the cost of a basket of ICT goods and services.3 The estimated increase
in annual ICT adoption if these tariffs and special taxes were to be removed in
Brazil is significant, with increases in end-user consumer demand ranging
between 17 and 37 percent for low- and high-elasticity estimates, respectively.
Map 5.1 LAC Has Some of the Highest Total Tariffs and Taxes for ICT Products
Countries by total tari s and taxes for consumer ICT goods and services
Less than 1 percent Between 1 and 5 percent Between 5 and 15 percent Between 15 and 25 percent Greater than 25 percent Data not available
Source: Miller and Atkinson 2014. © Information Technology and Innovation Foundation. Used with permission; further permission required for reuse.
Note: ICT = information and communication technologies.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
42 Improving the Environment for Technology Adoption with Inclusion
And middle-of-the-road estimates for the growth effects of removing these ICT
tariffs and taxes show that these costs may be holding back growth by 1.5 per-
centage points of GDP per capita per year for Brazil, 1.2 percentage points for
Argentina and the Dominican Republic, 1.0 percentage point for Ecuador, and
0.8 percentage point for Jamaica. An additional indicator of the high cost of digi-
tal technology business tools in the LAC region is given by the iPad and iPhone
indexes, where two of the four listed Latin American countries (Argentina and
Brazil) are in the top two (for the iPhone) and top three (for the iPad) most
expensive countries out of 57 countries; the US dollar product cost of the most
expensive country (Argentina for iPhones, Brazil for iPads) is more than twice
the cost for the least expensive countries.4
Product Market Policies
A range of product market policies shape the size of the output expansion effect
on which inclusive job outcomes depend. Output expansion effects hinge on
the responsiveness of output prices (as well as quality and market outreach
efforts) to the lowered variable costs enabled by technology adoption; the avail-
ability of added labor, physical capital, energy, distribution inputs, and finance;
and the responsiveness of consumer demand to the drop in output prices.
Among others, a menu of product market policy reforms could include the
following:
• Increasing the intensity of domestic market competition (including entry, expan-
sion, and exit policies). For example, firm owners are more likely to pursue
new customers using lower prices, higher-quality products, or both where they
feel the pressure of competition. Lower variable costs are more likely to lead
to larger output increases in contestable, competitive markets. Furthermore,
output will expand more if the risk of losses from the possibility that any out-
put expansion will need to be retracted is lower, which is influenced by, among
other factors, the ability of firms to retrain or lay off workers at moderate cost
and by the ability of bankruptcy protection to guard investors’ interests if
retraction is necessary.
• Adopting policies that increase product tradability across regions within the
country and internationally for all potentially tradable products, including con-
nectivity policies lowering transport and logistics costs, and critically including
policies to lower tariff and nontariff barriers to external trade.
• Facilitating access to finance for adopting firms to purchase required inputs and
pay for product promotion efforts.
• Adopting policies supporting the upgrading of the quality of firm management and
other factors affecting firms’ ability and know-how to enlarge production and
distribution in response to lower variable costs.
Lack of competition, especially in nontradables, is a strong candidate for
explaining the LAC region’s lackluster technology adoption and output growth
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Improving the Environment for Technology Adoption with Inclusion 43
performance. Although the share of entrepreneurs (as measured by the number
of firms per capita) is higher in LAC than in comparator countries and regions,
this statistic masks a lack of dynamism in the region. Entry by entrepreneurs into
the higher firm-size end of the formal sector remains low. New firms do not grow
as much as firms in other regions and tend to remain small. Firms in the LAC
region introduce new products less frequently than firms in otherwise-similar
economies. High-end entrepreneurs tend to be further away from global best
practices in the management of their enterprises, firms’ investment in research
and development is low, and patent activity is well below benchmark levels. Lack
of competition also helps explain LAC’s flat innovation and business growth
performance, including technological adoption and catch-up (Lederman et al.
2014). Panel a of figure 5.2 benchmarks LAC countries’ revealed market concen-
tration in industries that are arguably not subject to international competition.5
Most countries appear at the upper end of the market concentration index, and
all but two (Brazil and Colombia) exhibit levels of market concentration well
above their international benchmarks. Panel b of figure 5.2 benchmarks LAC’s
level of openness based on the ratio of international trade flows (imports plus
exports) to GDP. The two countries that performed somewhat better relative to
their international benchmarks with respect to nontradables are the worst per-
formers. Brazil, in particular, has the lowest level of openness across all available
countries, underperforming by about 3 to 5 percentage points of GDP. Of the
13 LAC countries with deficits in innovation, 10 have deficits in competition in
nontradables, and 6 have deficits in competition in tradables.
In addition to stimulating larger output expansion effects, market competi-
tion also provides firms with incentives to adopt and use more digital technolo-
gies. Manufacturing firms in Mexico are more likely to invest in digital
technologies and use them more productively when they sell products in mar-
kets where they face more intense rivalry (Iacovone, Pereira-Lopez, and
Schiffbauer 2016). Firms that faced the external shock of higher foreign com-
petition from China between 2000 and 2008, either in the domestic or the U.S.
(export) market, increased their number of computers per employee, their
share of labor using the Internet, and their share of online purchases in total
purchases in the subsequent four-year period 2008–12. As a result, the share of
labor using the Internet in 2012 was 11 percent higher for firms that faced more
Chinese competition, and the share of online purchases was 114 percent higher.
The more intensive use of digital technologies translated into productivity
growth among firms, driven by a combination of cost efficiencies and output
expansion. By contrast, ICT use had no impact on labor productivity growth
among Mexican firms that did not face import competition from China.
Similarly, manufacturing firms facing an increase in competition in Brazil were
found to be more likely to adopt and use e-commerce systems, and to adopt
more advanced e-commerce systems (Cirera, Lage, and de Oliveira 2015). More
generally, aggregate sector and country data highlight a negative correlation
between regulatory barriers to product market competition and firms’ adoption
of digital technologies.6
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Figure 5.2 Indices of Competition
a. Index of competition in nontradables b. Index of competition in tradables
United States Brazil
Bulgaria Japan
Romania United States
Poland India
Canada Argentina
Hungary Australia
Russian Federation Colombia
Lithuania Uzbekistan
Czech Republic Russian Federation
Norway Philippines
Colombia Greece
Latvia Indonesia
Korea, Rep. China
Japan Venezuela, RB
Portugal Peru
United Kingdom United Kingdom
Switzerland Dominican Republic
Macedonia, FYR Turkey
China Spain
Italy Italy
Ireland France
Germany Norway
Croatia Jamaica
Serbia Uruguay
Belarus Canada
Guatemala
Thailand Albania
Spain Kazakhstan
Sweden Azerbaijan
Austria Armenia
Finland Turkmenistan
Netherlands Suriname
Argentina Portugal
Greece Chile
Denmark Croatia
Brazil Ecuador
Bosnia and Herzegovina Finland
Singapore Denmark
Australia El Salvador
Belgium Sweden
Turkey Bolivia
Mexico Switzerland
France Mexico
Philippines Paraguay
Malaysia Haiti
Moldova Bosnia and Herzegovina
Israel Serbia
New Zealand Germany
Kuwait Korea, Rep.
Hong Kong SAR, China Poland
Ecuador Georgia
Kazakhstan Ukraine
Indonesia Honduras
Chile Ireland
Peru Austria
Albania Moldova
Trinidad and Tobago
Saudi Arabia
Nicaragua
India Malaysia
Uruguay Thailand
Dominican Republic Macedonia, FYR
Oman Latvia
Guatemala Costa Rica
Bolivia Guyana
Jamaica Bulgaria
United Arab Emirates Slovenia
Paraguay Belarus
Venezuela, RB Lithuania
Trinidad and Tobago Czech Republic
Nicaragua Netherlands
El Salvador Hungary
Costa Rica Slovak Republic
Honduras Belgium
0 0.2 0.4 0.6 0.8 0 50 100 150 200
Herfindahl index % of GDP
LAC countries LAC countries Poisson
Other countries or economies Other countries Negative binomial
Benchmark Zero-in ated Poisson
Source: Figures 6.4 and 6.5 in Lederman et al. (2014).
Note: LAC = Latin America and the Caribbean.
Improving the Environment for Technology Adoption with Inclusion 45
Education, Skills, and Labor Market Policies
Human capital plays a critical role for both the adoption and the use of technol-
ogy because the demand for different and often more sophisticated skills gener-
ally increases following digital technology adoption. The evidence discussed in
the previous chapters suggests substantial shifts in employment composition
following Internet rollouts, favoring more educated and high-skilled workers.
This presents risks for workers in the LAC region. In Brazil, for instance, in the
short term, the adoption of the Internet creates increased local demand for more
cognitive and nonroutine tasks, and within these, interpersonal and communica-
tion skills are at a premium. In addition, the evidence from Argentina and Chile
also suggests that the adoption of technology is heavily influenced by the quality
of the manager’s human capital (as captured by her or his level of education and
past relevant experience). Without a sufficient supply of workers equipped with
these skills, the full productivity potential of digital technology adoption will be
more difficult to achieve.
Investing in the skills of the workforce is necessary for individuals to benefit
from the productivity and welfare gains from technology adoption. Technology
adoption could simply increase inequality in labor markets. However, to lower
the likelihood and severity of inequitable outcomes, investments in digital
technologies should be accompanied by simultaneous investments in human
capital through education and training systems. Without 21st-century skills to
perform higher-level cognitive and analytical tasks, workers may find it much
harder to be employed in the future, which may eventually increase labor
market gaps. Furthermore, without managers who have adequate skills and
experience, which could lead to lower-than-expected Internet adoption, firms
and workers in the region could be forgoing significant productivity and wel-
fare benefits. Hence, public education and training policies should focus on
ensuring that today’s youth—the future workforce—are well equipped with
solid foundational skills, both cognitive and socioemotional, that allow them to
take full advantage of the opportunities created by the digital world. Digital
literacy and advanced ICT skills are also critical for the adoption and use of
these technologies.
The evidence presented in this book suggests that technology adoption is
associated with labor markets becoming more dynamic. To help people manage
a more dynamic labor market, it will be essential to ensure that workers have
solid foundational skills, both cognitive and socioemotional, that allow them to
keep learning and acquiring human capital throughout their professional lives.
In addition, assessing—and addressing—any constraints on greater investment by
employers in job training is also important. There are market failures that com-
bine to make it difficult for employers to appropriate returns to investments in
transferable skills of their workers. Institutional arrangements, such as those that
allow employers and employees to write labor contracts in which they agree on
training and any eventual reimbursements if the contract is not fulfilled, can help
mitigate risks related to the poaching of workers.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
46 Improving the Environment for Technology Adoption with Inclusion
The promise of greater productivity from digital technology adoption creates
new urgency for reforms to education and skills-training structures in LAC coun-
tries. The highest priority should be given to changes in the system that encour-
age the development of solid foundational skills, higher-order cognitive skills,
socioemotional skills, and digital skills (many of which are highly transferable
across occupations). Today, most education and training systems in LAC coun-
tries still fall far short of ensuring that all children complete lower- or upper-
secondary education. And learning outcomes continue to languish below the
performance of students in many peer countries. Even the lowest performers
among OECD-member countries score well above the top-performing students
in LAC (figure 5.3). Important gaps also remain in rates of completion and qual-
ity of higher education, as highlighted by the relatively low levels of scientific
production in LAC countries compared with countries in Asia, North America,
and Western Europe (figure 5.4). Furthermore, large inequalities in learning persist.
Children from households in lower quintiles of the income distribution signifi-
cantly underperform those from households in the upper quintiles. The timeframe
required for reforms to education and skills training systems to bear fruit implies
that this policy agenda should be given priority.
More broadly and more politically contentious, the underlying objectives of
labor market policies need to be reformulated. Prevailing labor market policies
Figure 5.3 PISA Results and GDP per Capita
595
565
535
PISA 2015 score in science
505
475
445
415
385
355
325
0
0
00
00
00
00
00
00
00
00
00
00
00
00
00
00
,0
,0
,0
,0
,0
,0
,0
,0
,0
,0
,0
,0
,0
5,
10
15
20
25
30
35
40
45
50
55
60
65
70
GDP per capita in 2015 or latest, PPP (constant 2011 US$)
East Asia and Pacific Latin America and the Caribbean OECD
Europe and Central Asia Middle East and North Africa
Sources: World Bank 2016a, 2016b.
Note: OECD = Organisation for Economic Co-operation and Development; PISA = Programme for International Student
Assessment; PPP = purchasing power parity.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Improving the Environment for Technology Adoption with Inclusion 47
Figure 5.4 Scientific Production by Geographic Region
35
30
Percentage of documents
25
20
15
10
5
0
Western Eastern Northern Northern Latin Middle Asia Paci c
Europe Europe Africa America America East region region
1996 2008 2012
Source: Ferreyra and others 2017.
were designed for a time when most people reasonably expected to take the
skills they learned in school and apply them in a single profession and in the
same way for the course of their working lives. Technology adoption has always
entailed changes to how people work. But what is different now is the speed
of these changes. Technological change has accelerated to an unprecedented
pace that would have been difficult to predict when most labor policies were
adopted. Rather than a single occupation and place of employment, career
disruption and transitions are becoming the new norms. The social norms and
expectations on which labor policies rest are shifting away from aspirations for
occupational stability and long-term employment relationships, toward occu-
pational dynamism and greater mobility across professions, skill levels, geogra-
phy, and forms of economic engagement (such as between employment,
self-employment, and entrepreneurship). Technological advances are just one
of several forces driving this shift. If policies were intended and designed to
support people’s management of the uncertainties and demands of a dynamic
labor market, rather than the more stable labor markets of the past, labor
policies—from worker protection and income support to intermediation—
would probably look very different from their current form in most countries
of the LAC region.
Notes
1. For a review of relevant technology diffusion and adoption support policies, see Cirera
and Maloney (2017). They highlight management practices as one particularly prom-
ising area for technology adoption and upgrading. Key factors determining adoption
that are prone to policy support include competition, human capital, and learning
through participation in international markets (through trade, foreign direct
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
48 Improving the Environment for Technology Adoption with Inclusion
investment, and participation in global value chains), in addition to well-designed
management extension programs.
2. Guatemala benefited from a faster reduction in prices for mobile services than the rest
of LAC by being an early pioneer in spectrum auctions following its 1996 telecom-
munications law. Other LAC countries that have adopted similar policies have now
overtaken Guatemala in the spectrum allocated to mobile communications (see
box 4.6 and figure B4.6.1 in World Bank 2016b). Prices of mobile and fixed broad-
band services remain, on average, higher in LAC than in OECD countries (see box 4.1
and map 4.1.1 in World Bank 2016b).
3. The basket of ICT goods and services on which tariffs and special taxes are levied
includes wired broadband, wireless phone services, and core ICT products, such as
basic mobile phones, smartphones, computers, and other digital products like digital
cameras and digital audio devices.
4. Differences in prices include tariffs and extra local consumption taxes, freight, and
different markups. For iPhones, the product is the Apple 7, 4.7-inch 32GB device,
costing US$1,630 in Argentina versus US$649 in California and US$640 in Saudi
Arabia. For iPads, the product is the Apple Pro 10.5-inch 64GB Wi-Fi device, costing
US$1,619 in Brazil versus US$703 in California and US$638 in Hong Kong SAR,
China. Chile is the 7th most expensive location for this iPhone and the 9th most
expensive location for this iPad; Mexico is 28th and 30th, respectively (Commonwealth
Bank of Australia 2017). Import tariffs on computers and laptops are even higher in
Cuba than in Argentina and Brazil (see figure 5.5 in World Bank 2016b).
5. The 17 nonfinancial services industries include electricity, building and special-
ized construction, civil engineering, wholesale and retail trade, transport and
warehousing, telecommunications, insurance, real estate, engineering, and tour-
ism. The index shows the average Herfindahl index of concentration of revenues
across a selection of two-digit industries for which data were available for more
than 80 countries, with a value of 1 representing a market that is captured by a
single firm; lower values indicate lesser market concentration. Revenues were
averaged across 2007–10.
6. More restrictive product market regulations on firm entry in services industries are
associated with lower digital adoption, with most LAC countries in the sample
(including Brazil, the Dominican Republic, Honduras, Mexico, Nicaragua, and Peru)
having higher barriers to entry and lower levels of digital adoption than most other
countries (World Bank 2016b).
References
Cirera, Xavier, Filipe Lage, and Joao Maria de Oliveira. 2015. “E-Commerce and
Productivity in Brazilian Firms.” Background paper for World Development Report
2016: Digital Dividends. World Bank, Washington, DC.
Cirera, Xavier, and William Maloney. 2017. The Innovation Paradox: Developing-Country
Capabilities and the Unrealized Promise of Technological Catch-Up. Washington, DC:
World Bank.
Commonwealth Bank of Australia. 2017. “Economic Insights: CommSec iPad/iPhone
Indexes, Australia No Longer the Cheapest.” September 11. Commonwealth Bank of
Australia.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Improving the Environment for Technology Adoption with Inclusion 49
Ferreyra, María Marta, Ciro Avitabile, Javier Botero Álvarez, Francisco Haimovich Paz, and
Sergio Urzúa. 2017. At a Crossroads: Higher Education in Latin America and the
Caribbean. Directions in Development. Washington, DC: World Bank.
Iacovone, Leonardo, Mariana Pereira-Lopez, and Marc Schiffbauer. 2016. “Competition
Makes IT Better: Evidence on When Firms Use IT More Effectively.” Policy Research
Working Paper 7638, World Bank, Washington, DC.
Lederman, Daniel, Julián Messina, Samuel Pienknagura, and Jamele Rigolini. 2014.
Latin American Entrepreneurs: Many Firms but Little Innovation. Washington, DC:
World Bank.
Miller, Ben, and Robert Atkinson. 2014. Digital Drag: Ranking 125 Nations by Taxes and
Tariffs on ICT Goods and Services. Washington, DC: Information Technology and
Innovation Foundation.
World Bank. 2016a. “PISA LAC Country Profiles.” Unpublished, Washington, DC.
———. 2016b. World Development Report 2016: Digital Dividends. Washington, DC:
World Bank.
World Economic Forum. 2017. The Global Information Technology Report 2016. Geneva:
World Economic Forum.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
CHAPTER 6
Conclusions
The principal conclusion of this book is that digital technology adoption offers a
pathway to higher productivity and inclusive growth, contingent on appropriate
enabling policies. This conclusion is supported by the following findings from the
background empirical studies commissioned for this book (appendix A):
• Impacts on workers’ labor market outcomes—skills, jobs, and earnings—differ
depending on the type of technology adopted. Some technologies (such as com-
plex software) appear to be inherently more “inclusive” than others (the
Internet, for instance). Within firms, Internet availability, adoption, and use
typically create a substitution effect away from low-skilled workers and toward
more skilled workers. Internet adoption into firms’ productive processes also
creates a substitution effect away from manual and routine tasks toward more
cognitive and analytical tasks. In contrast, use of complex software by firms in
Chile creates a substitution effect toward low-skilled workers, and is associated
with a reallocation of jobs away from abstract, sophisticated tasks and toward
more routine and manual tasks.
• Output expansion effects can overcome digital technologies’ adverse substitution
effects on low-skilled labor and result in more inclusive growth. Importantly, siz-
able output effects can lead to an overall positive expansion in the levels of
employment of both lower-skilled (less-well-off) and higher-skilled workers.
Evidence from Brazil and Mexico suggests that these output effects tend to
be larger in more tradable sectors that provide larger output expansion oppor-
tunities. Complementary evidence from Argentina suggests that these output
effects help lower-skilled workers more in firms experiencing faster output
growth.
• Adoption of information and communication technologies (ICT) changes the
demand for skills and shifts demand away from more routine, manual tasks
toward nonroutine analytical and cognitive tasks. The Brazil tasks and labor
policies impact study finds that this pattern is consistent nationwide but also
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 51
52 Conclusions
across all the largest sectors of the economy. This finding validates some of
the concerns that routine, manual tasks are increasingly being replaced by
technology, thereby displacing less-skilled workers (Autor, Levy, and
Murnane 2003). However, the overall impacts depend on the technology in
use. The Chile study exploits the impact of advanced software. Over a period
of more than five years, firms expanded following the adoption of complex
software, and also increased their use of low-skilled labor while relying more
on routine tasks.
• ICT is positively correlated with firm dynamism, as exemplified by job destruction
and job creation. Evidence from both Argentina and Brazil suggests that digital
technologies lead firms to replace workers, replace occupations, exit, or expand,
creating new occupations. Importantly, in Argentina the share of firms creating
new occupations is almost three times larger than the share of firms replacing
occupations.
• Adopting firms have more-skilled managers and invest more in upgrading the skills
of their workers. In Argentina and Chile, adoption of ICT and advanced soft-
ware, respectively, are highly correlated with the human capital of the man-
ager. More educated and experienced managers are more likely to invest in
new digital technologies. The Mexico study documents that firms that make
more intensive use of ICT provide more training for both higher- and lower-
skilled workers, with lower-skilled workers receiving significantly greater levels
of training. In Chile, the adoption of complex software by firms across all sec-
tors of the economy is also associated with increased investment in on-the-job
training for digital skills, especially at the managerial level.
• The adoption and the impacts of ICT on jobs and employment outcomes depend on
the stringency of labor market regulations and in what combination they are
deployed. High statutory minimum wages and the direct monetary costs of sup-
porting displaced workers appear less detrimental for the adoption of ICT than
burdensome worker dismissal procedures. At reasonable levels, statutory mini-
mum wages can encourage firms to invest in developing the skills of their work-
force to maintain and even grow specifically useful human capital. Anticipating
a rise in their productivity, firms eager to adopt new technologies and proce-
dures might welcome the chance to contribute to the relatively certain up-
front costs of meeting the needs of workers who require up-skilling, re-skilling,
or indeed financial support while they look for new jobs. However, keeping the
“hassle tax” of making changes to their workforces low (that is, less cumber-
some administrative procedures for dismissing workers and fewer restrictions
on how workers are employed) could speed the pace of adjustment, the
achievement of new levels of productivity, and the types of increases in output
that translate into job creation. Furthermore, the Brazil tasks and labor policies
impact study shows that enforcement of labor regulations limits the degree to
which companies shift away from labor as technology becomes available.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Conclusions 53
In contrast with policy intentions, they seem to protect more the more
skilled workforce, particularly those leading nonroutine and higher-level cogni-
tive tasks.
• Inclusive employment outcomes from technology adoption are fostered by a busi-
ness environment in which greater productivity leads to greater output. Greater
productivity and output are more likely where labor can be more easily rede-
ployed within firms or across industries, where there is better connectivity and
distribution infrastructure (increasing opportunities to effect sales), and in
contestable, competitive markets where firms have strong competitive pres-
sures that provide them with incentives to pass on the productivity gains to
customers in the form of lower prices, better-quality products, or both.
• ICT reduces firms’ cost of entering more distant national and foreign markets.
Lower entry costs enabled by online trading platforms and other such
connectivity-enhancing technologies allow smaller firms with relatively more
low-skilled workers to benefit from international trade by extending their
reach to larger, more diverse markets. A boost to their production can increase
the wages that firms are able to pay their workers relative to workers in
firms that use skilled labor more intensively. Easier and lower-cost access to
international trade, therefore, supports more inclusive growth through a
market access effect that favors smaller firms and less-well-off workers.
• ICT reduces individuals’ information costs, leading to lower sector and regional
mobility costs. Worker mobility costs across sectors and regions are kept high by,
among other factors, information asymmetries. Access to the Internet can
lower these mobility costs, thereby supporting more inclusive growth through
lower adjustment costs and greater labor market efficiency.
Questions for Further Research
The more detailed analyses underpinning this book point the way for addressing
more granular and perhaps manageable policy questions in future work. Two
overarching questions remain: First, why does digital technology adoption remain
so low in Latin America and the Caribbean? Second, which policies are more
likely to foster the kind of technology adoption that will elevate productivity and
increase inclusive growth through the output expansion mechanism? Regarding
the challenge of improving product market policies, follow-on questions include
the following: Can the enhanced availability of new technologies to firms be
linked to their participation in industries with potential for substantial output
expansion? What policy measures also enhance firms’ ability to expand output
through the availability of capital financing, and through enhanced avenues for
expanding areas of sales, including through managerial assistance for growth
strategies? Furthermore, in a region that has always struggled with protected,
segmented, oligopolistic markets, the role of domestic competition policy is a
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
54 Conclusions
potential issue for future research—to address insufficient competition in invest-
ments and delivery of ICT services as well as in output and input markets of
adopting firms, including high transport costs, import duties and additional taxes,
and other policy-related barriers that raise the costs of adoption of ICT and other
technologies and that dull adopting firms’ output expansion incentives. The role
of appropriate domestic competition policy includes the important issue of what
policy safeguards can be put in place to ensure contestability and competition,
particularly in the face of digital technologies’ network effects.
Additional questions arise regarding the challenge of improving education
systems and adjusting labor market policies. How can education and training
systems better prepare students today for jobs that are increasingly more skill-
biased, and oriented toward higher-level analytical and socioemotional skills?
How can the business environment better support employers in fostering such
investment in human capital throughout workers’ lives? What should policy’s
role be in fostering the use of ICT to reduce information costs to individuals and
to lower sector and regional mobility costs? If career disruption is more likely
than in the past, what form should labor market policies take to assist workers in
making occupational and labor market transitions? And how can technology
more effectively support enforcement of regulations as well as enable employers’
compliance while providing incentives for firms and workers to keep adopting
and adapting new technologies? These questions provide a rich menu of further
policy-relevant analytical work to better understand how to make productivity
growth more inclusive.
Reference
Autor, D. H., F. Levy, and R. J. Murnane. 2003. “The Skill Content of Recent Technological
Change: An Empirical Exploration.” Quarterly Journal of Economics 118 (4):
1279–333.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
APPENDIX A
Background Studies
Conceptual framework:
Brambilla, Irene. 2018. “Digital Technology Adoption and Jobs: A Model of Firm
Heterogeneity.” Policy Research Working Paper 8326, World Bank, Washington, DC.
Studies exploiting firm-level and municipal-level data with new learning on the
impacts of digital technology adoption on productivity, jobs, skills, and wages:
Almeida, Rita K., Carlos H. L. Corseuil, and Jennifer P. Poole. 2017. “The Impact of Digital
Technologies on Routine Tasks: Do Labor Policies Matter?” Policy Research Working
Paper 8187, World Bank, Washington, DC.
Almeida, Rita K., Ana M. Fernandes, and Mariana Viollaz. 2017. “Does the Adoption of
Complex Software Impact Employment Composition and the Skill Content of
Occupations? Evidence from Chilean Firms.” Policy Research Working Paper 8110,
World Bank, Washington, DC, and CEDLAS-FCE-UNLP Working Paper No. 214,
Argentina.
Brambilla, Irene, and Darío Tortarolo. 2018. “Investment in ICT, Productivity and Labor
Demand: The Case of Argentina.” Policy Research Working Paper 8325, World Bank,
Washington, DC.
Dutz, Mark A., Lucas Ferreira Mation, Stephen D. O’Connell, and Robert D. Willig. 2017.
“Economy-Wide and Sectoral Impacts on Workers of Brazil’s Internet Rollout.” Forum
for Social Economics 46 (2): 160–77.
Iacovone, Leonardo, and Mariana Pereira-López. 2018. “ICT Adoption and Wage Inequality:
Evidence from Mexican Firms.” Policy Research Working Paper 8298, World Bank,
Washington, DC.
Ospino, Carlos. 2018. “Broadband Internet, Labor Demand and Total Factor Productivity
in Colombia.” Policy Research Working Paper 8318, World Bank, Washington, DC.
Studies exploiting household data with impacts of digital technology adoption
linked to trade and labor mobility, and links between labor market regulations
and digital technology adoption:
Cruz, Marcio, Emmanuel Milet, and Marcelo Olarreaga. 2017. “Online Exports and the
Wage Gap.” Policy Research Working Paper 8160, World Bank Washington, DC, and
Discussion Paper 12092, Centre for Economic Policy Research, London.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 55
56 Background Studies
Cruz, Marcio, Emmanuel Milet, and Marcelo Olarreaga. 2017. “Labor Adjustment Costs
across Sectors and Regions.” Policy Research Working Paper 8233, World Bank,
Washington, DC.
Packard, Truman, and Claudio E. Montenegro. 2017. “Labor Regulation and Digital
Technology Use: Indicative Evidence from Cross-Country Correlations.” Policy
Research Working Paper 8221, World Bank, Washington, DC.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
APPENDIX B
Detailed Literature Review
Impacts of Technology Adoption on Jobs
The effects of technology adoption on labor demand predicted by economic the-
ory are ambiguous. For instance, process innovation can lead to the substitution
of capital for labor while also increasing productivity, lowering prices, and
increasing demand for firms’ output, thereby leading to higher employment.
Obversely, while product innovation usually creates more demand, it can also
increase the market power of innovators, allowing them to profitably raise their
prices and suppress the output of their products (Castillo et al. 2011). With
further ambiguity, Bender et al. (2016) show that the enhanced productivity of
better-managed firms is accompanied by fewer layoffs since these firms are more
likely to recruit higher-ability workers.
People with different skill levels are affected differently by the adoption of
technology, which can change the structure of the workforce. The introduction
of new technologies can make workers more productive, leading to higher
wages, but it can also be associated with employment turnover. A large body of
empirical work argues that technology adoption has favored the wage and
employment prospects of relatively high-skilled workers while simultaneously
dampening the wages and employment of the less skilled (see, for example,
Autor, Katz, and Krueger 1998; Bresnahan, Brynjolfsson and Hitt 2002; Caroli
and van Reenen 2001).
The key parameters determining the impact of technology on jobs include
skills demand, labor costs, productivity, and output demand (including changing
price and income elasticities of demand). Novick and Rotondo (2013) estimate
a panel model for the period 2007–10 and find that wages and employment
growth are higher as the information and communication technologies (ICT)
structure becomes more complex for firms in the same sector with the same size,
age, and productivity level. The authors argue that the results are contrary to
the thesis that posits that technological unemployment is a consequence of the
incorporation of labor-saving technologies, particularly affecting low-quality jobs.
Moreover, the results are consistent with the perceptions of employers in com-
panies with highly complex ICT adoption processes, who indicated in a set of
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4 57
58 Detailed Literature Review
in-depth interviews that ICT adoption has no obvious impact on staff turnover.
A separate group of studies relies on the estimation of partial correlations based
on firm-level data from developed countries to analyze the adjustment in
employment, skills, and wages due to ICT adoption. The studies use a wide vari-
ety of firm-level ICT adoption measures, such as information technology (IT)
capital stock, computer adoption, the number of computers, IT investment, and
the number of IT workers. Their results indicate that ICT adoption is associated
with higher relative demand for high-skilled workers and higher wages (Caroli
and van Reenen 2001; Greenan and Topiol-Bensaid 2001; Bresnahan, Brynjolfsson,
and Hitt 2002).
Based on country- and industry-level data for 19 high-income countries over
more than 35 years,1 Autor and Salomons (2017) find that rising country-level
productivity is associated (correlated) with growing aggregate employment and
a rising employment-to-population ratio. They show that this is the result of two
different dynamics: on the one hand, industries that experienced rapid produc-
tivity growth exhibited diminished internal (own-industry) employment growth
over time;2 on the other hand, other-industry aggregate productivity growth
(occurring outside each industry) has strong predictive power for employment
growth within each industry, with these cross-industry productivity spillovers
accruing through rising final demand (an income effect raising consumer pur-
chasing power) or through inter-industry demand (output) linkages—fueling
aggregate output expansion and jobs growth. The indirect positive effect of pro-
ductivity growth on employment across industries is found to dominate and
more than fully offset the direct negative effect on own-industry employment.3
Although their empirical approach allows economy-wide and industry-specific
job effects to be assessed, the authors concede that their approach to measuring
technological adoption based on labor productivity does not distinguish among
the possible different labor market consequences of different technologies, nor
other distinct sources of productivity growth, such as those arising from shifts in
infrastructure investments, trade, offshoring, and global production chains.
The findings of Autor and Salomons are perfectly consistent with the impact
on jobs of different types of technological advances. On the one hand, there are
technological advances like textile weaving machines and robots on the assembly
line that directly displace (substitute for) lower-skilled workers, and in substantial
quantities, due to the scale economies in their operations and support infrastruc-
ture. Even with positive output effects, this is likely to decrease lower-skill jobs in
those enterprises, unless the domain of application of this technology is narrow
compared to the scope of unskilled tasks in the enterprise; see Bessen (2015) on
textiles and Acemoglu and Restrepo (2017) on robots. On the other hand, there
are quite different technological advances (like ICT and broadband availability, on
which we focus) that do not much directly displace and substitute for lowest-skill
work, but might somewhat substitute for work that is part of the hollowed out
middle-skill work story-line (see the section on tasks further below). Here there is
bound to be productivity gains, and potential positive output effects. The overall
country-wide impact on jobs of these different types of technological advances,
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 59
as highlighted by Autor and Salomons, shows an overall positive country-level
association of productivity and employment: even where there is direct substitu-
tion for lower-skill work, there will be productivity gains, lower prices for the
output which serves as inputs into other industries, and thus output gains in other
industries utilizing those inputs and correspondingly needing more labor, without
the directly offsetting impacts on employment from substitution that occurred in
the upstream industries.
Although current thinking on technology, skills demands, and jobs is shaped
mainly by evidence from high-income countries, a recent surge of studies focuses
on innovation and labor markets in Latin American countries. The body of
empirical work examining the effects of different types of innovation on employ-
ment growth and skills composition in Latin America is growing quickly. This is
particularly apparent for firm-level studies applying similar analytical techniques
(see, for example, Alvarez et al. 2011 for Chile; de Elejalde, Giuliodori, and
Stucchi 2011 for Argentina; Monge et al. 2011 for Costa Rica; and Aboal et al.
2011 for Uruguay). And for a broader assessment on these and related questions
associated with how emerging trends in technology and globalization are likely
to shape the feasibility and desirability of manufacturing-led development as a
generator of productivity and jobs from the perspective of all low- and middle-
income countries, see Hallward-Driemeier and Nayyar (2017).
Technological innovation in products and processes in Latin American
countries appears to affect employment differently and, on balance, positively.
The first conclusion from these papers is that while product innovations increase
employment, process innovations do not affect jobs. The results do not change
when the studies account for firm size. A second result is that product innovation
increases both high-skill and low-skill jobs, with a higher proportion of high-skill
jobs. However, process innovation, in general, has a weakly negative or zero effect
on low-skill employment growth. The third result from these papers is that pro-
ducing technology internally (in house) has the biggest positive impact on
employment, followed by make-and-buy or buy-only strategies. Overall, these
papers indicate that innovation does not generally lead to job losses and that it
generates greater demand for a more qualified labor force.
Introduction of new products seems to boost employment. Crespi and Tacsir
(2012) present a comparative analysis of four Latin American countries that
shows product innovation to be an important source of firm-level employment
growth due to a boost to demand. Process innovation accounts for a small share
of the changes observed in employment, inducing small or zero displacement
effects. This last result can be explained by the absence of productivity gains that
would lead to a reduction in employment, or the combined effect of productivity
gains (displacement or substitution effect) that induce demand growth through
market competition (creation or output effect). The results are similar for small
and large firms. The researchers also find some evidence of skill bias of product
innovation for high-tech sectors. Some public policies related to innovation
promotion have also been evaluated through quasi-experimental techniques.
Castillo et al. (2011) evaluate the impact of the Argentine innovation support
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
60 Detailed Literature Review
program PRE (Enterprise Restructuring Program) on employment and wages.
They find that although support for both process and product innovation–related
activities leads to increased employment, the support for product innovation has
a higher effect on real wages, exporting, and survival probability.
Impacts of Technology Adoption on the Skill Content of Tasks
Concerns for the impact of technology on jobs has been heightened by evidence
of a “bias” that favors people with skills and that many argue is inherent in tech-
nological change. This concern is fueled mostly by greater inequality in earnings
from shifts in the demand for skills in the United States and other high-income
economies. The rise in inequality in the United States that was led by a rise in
the skill premium, and especially the college wage premium during the 1980s
and 1990s, motivated a large literature that pointed to skill-biased technological
change (SBTC) as a driving force behind this phenomenon. SBTC is based on the
idea that technology is complementary to skills and therefore its benefits are
biased toward more skilled workers, while technology can substitute for less
skilled workers. When new technology is adopted, the demand for high-skilled
relative to low-skilled workers increases and the wage gap widens, increasing
wage inequality. These effects might be avoided by an increasing supply of
human capital, in what Goldin and Katz (2009) regard as a race between tech-
nology and education. In this sense, if the supply of highly skilled individuals
increases at a faster pace than ICT adoption, wage inequality can even be
reduced (World Bank 2016).
The available evidence of a skills bias mostly comes from the United States.
For instance, Krueger (1993) finds a strong positive correlation between wages
and computer use by workers, and Doms, Dunne, and Troske (1997) show that
establishments that invested relatively more in computing equipment had larger
increases in the share of nonproduction labor. Furthermore, Machin and van
Reenen (1998) provide evidence that SBTC is an international phenomenon that
has increased the demand for high-skilled workers over demand for less-skilled
workers. More recent studies have focused more on the mechanisms through
which increases in technology could affect wages and hours worked by introduc-
ing a “tasks approach” to labor demand. Autor, Levy, and Murnane (2003) and
Acemoglu and Autor (2011) focus on the task content of different occupations.
They explore how technology—and ICT in particular—substitutes for routine
tasks but complements cognitive and nonroutine tasks that are performed by
individuals with more skills. Workers with a certain skill level can change the set
of tasks performed to respond to changes in technology, how it is applied in the
workplace, and broader labor market conditions.
People in the middle of the skills distribution are particularly affected by
advances and adoption of technology. A growing body of theory and empirical
work shows that middle-skilled individuals working on routine tasks could also
be vulnerable to replacement by ICT. This substitution would lead to job and
wage “polarization.” Michaels, Natraj, and van Reenen (2014) study whether
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 61
ICT has contributed to the rise in polarization. Based on a comprehensive
industry-level data set that covers 11 advanced countries from 1980 to 2004,
they find that the industries that invested more heavily in ICT demanded more
highly qualified workers. By analyzing how the different occupations and tasks
for the United States are correlated with education, they find that indeed more
highly educated individuals perform cognitive nonroutine tasks, while middle-
educated individuals are overrepresented in occupations that require routine but
more complex tasks than the noncognitive routine tasks that less-educated work-
ers perform. The estimated effect explains one-quarter of the college wage bill in
the economy as a whole. Autor, Katz, and Kearney (2006) present evidence of
rising employment in the highest- and lowest-paid occupations. In a later paper
the same authors show that wage inequality at the bottom half of the income
distribution has not increased since the 1980s and that, instead, wage inequality
has risen in the upper tail (Autor, Katz, and Kearney 2008). Finally, Autor and
Dorn (2013) examine the impact of computerization on the demand for low-
skilled labor. They use district-level data for the United States and show that
areas with high levels of routine tasks have experienced greater adoption of ICT,
greater reallocation of workers from routine tasks to the service sector, wage
polarization, and large inflows of high- and low-skilled workers. These recent
approaches based on tasks predict positive effects of ICT adoption on the
demand for more-educated individuals and reductions in the demand for
medium-skilled individuals, with effects for less-educated workers being less clear.
Different technological advances and types of new technology adopted by
firms have very different effects on skills demand and jobs. Looking beyond com-
puterization and adoption of basic ICT, the labor market impacts of more sophis-
ticated technologies, including advances in automation, robotics, and AI (artificial
intelligence), have captured the attention of researchers (Frey and Osborne
2013; Brynjolfsson and McAfee 2011, 2014; Graetz and Michaels 2015;
Acemoglu and Restrepo 2016, 2017, 2018). In a book that has garnered consid-
erable attention, Brynjolfsson and McAfee (2014) argue that more sophisticated
technological innovations are no longer confined to routine tasks, but increasingly
can be applied to nonroutine domains, even performing tasks typically per-
formed by higher-skilled workers. An example of this trend is the number of
tasks usually performed by lawyers and accountants that are now being under-
taken by sophisticated algorithms. Machine learning techniques are advancing in
the direction of being able to program a computer to autonomously master a
nonroutine task (Autor 2015). The ensuing concern is that such technological
innovations may in the future replace many types of jobs that have previously
been insulated from more routine-biased technological developments. World
Bank (2016) follows Frey and Osborne (2013) in investigating the feasibility of
automating existing jobs given current and potential technological advances,
based on the occupations of workers, and shows for OECD countries that over
the next couple of decades half of jobs could be automated. Following a task-based
approach, Arntz, Gregory, and Zierahn (2016) allow for heterogeneity of work-
ers’ tasks within occupations and demonstrate that the threat from technological
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
62 Detailed Literature Review
advances is less pronounced, with a much smaller percentage of jobs in OECD
member countries being candidates for automation in the future. Acemoglu and
Restrepo (2018) extend their task-based framework to study the implications of
automation and AI. Their framework emphasizes the substitution (which they
call “displacement”) effect that automation creates as machines and AI replace
labor in tasks that it used to perform, just as in the model in this book. This sub-
stitution effect tends to reduce the demand for labor and wages. It is counter-
acted by an output expansion (which they call “productivity”) effect, resulting
from the cost savings generated by the productivity gains from automation,
which increases the demand for labor in nonautomated tasks, again just as in the
model in this book. Their output expansion effect is complemented by additional
capital accumulation and the deepening of automation (improvements of existing
machinery), both of which further increase the demand for labor. However, they
emphasize that these effects, which as in the model in this book counterbalance
the substitution effect and therefore are called “countervailing effects” in their
paper, are incomplete. Even when they are strong, automation increases output
per worker more than wages and reduces the share of labor in national income—
which is different from the model in this book. The more powerful countervail-
ing force against automation in their model is the creation of new labor-intensive
tasks, which reinstates labor in new activities and tends to increase the labor
share to counterbalance the impact of automation. In contrast to their direction
of study, which focuses on digitized production equipment, including robots and
AI that can replace workers, the work in this book is focused more on the impact
of information-based, digitized, cognition-supporting technologies that are typi-
cally more complementary to workers.
The firm-level evidence on SBTC, occupations and tasks, and labor outcomes
is scarce, but points to an advantage for workers with more skills. Bloom et al.
(2014) examine the impacts of complex software on firms’ organizational deci-
sions, although they do not address the impact on demand for different types of
skills. Their conjecture is that the use of business software reduces the costs for
workers to access information, allowing workers to solve more problems and rely
less on the training of specialists. Their evidence for firms in the United States
and Europe confirms that indeed the use of business software increases decen-
tralization within the firm, leading to more autonomy and a wider span of con-
trol for local plant managers. Another firm-level example is in the work of
Akerman, Gaarder, and Mogstad (2015), who use Norwegian data and a quasi-
experiment to provide compelling causal evidence that suggests that employ-
ment and wages of high-skilled (low-skilled) workers increase (decrease) with
broadband Internet availability. On the firm side, increased availability of broad-
band Internet is associated with an increase (decrease) in the output elasticity of
low-skilled (high-skilled) labor. They argue that broadband adoption in firms
complements high-skilled workers in executing nonroutine, abstract tasks, and
substitutes for low-skilled workers in performing routine tasks. Other related
studies generally show that firms’ ICT adoption does not lead to changes in over-
all employment but tends to be associated with increased wages and better labor
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 63
outcomes of high-skilled workers—linked to nonroutine cognitive tasks—and
worse outcomes for low-skilled workers—linked to more routine tasks that are
automated with the use of ICT (Bartel, Ichniowski, and Shaw 2007; Böckerman
et al. 2016; De Stefano, Kneller, and Timmis 2014; Gaggl and Wright 2017).
Occupations that entail more ICT-intensive tasks are more likely to require
workers with more cognitive skills. Evidence from Latin America shows that
occupations that use ICT more intensively have high demand for cognitive skills
and low demand for routine and nonroutine manual skills in developing coun-
tries (Santos, Monroy, and Moreno 2015). Messina, Oviedo, and Pica (2016)
offer a descriptive perspective of this job polarization in emerging market econo-
mies by analyzing the task content of jobs in Bolivia, Colombia, and El Salvador.
The authors are only able to present a snapshot of possible job polarization
because no information was available with which to measure polarization. They
compare their results with data from the United States and find that although
the cognitive content of jobs is similar, the tasks performed under routine and
manual jobs in Latin America and the United States are different. They also use
Mexican and Chilean data to test whether the labor markets in these two coun-
tries show a pattern of polarization. According to their analysis, only Chile shows
possible job polarization. They find no evidence of wage polarization in any of
the other Latin American countries, perhaps suggesting that any impacts pro-
duced by ICT adoption could have been overcome by the strong commodity
boom experienced by most economies during the 2000s, which benefited pri-
marily low-skilled workers (Cruces et al. 2015; Maloney and Molina 2016).
In addition, there is evidence of a wage premium associated with the use of
computers at the workplace between 2000 and 2006 (Benavente, Bravo, and
Montero 2011).
Job polarization in the wake of technology adoption is far less apparent in
low- and middle-income countries. Looking at evidence from 21 countries in
Latin America, Asia, the Middle East, and Africa, Maloney and Molina (2016)
find little evidence of polarization patterns. Only in Brazil and Mexico did they
find a relative reduction of routine jobs, suggesting potentially polarizing forces
at work in the labor markets of those two countries. The results for Mexico coun-
ter the evidence presented by Messina, Oviedo, and Pica (2016). Maloney and
Molina (2016) additionally discuss the channels through which technology
affects the labor market, because the effects of automation and offshoring of
routine tasks may work in opposing directions in developing countries. They sug-
gest that technical change reduces routine tasks, but the same technical change
could promote the development of more complex versions of existing tasks,
increasing job opportunities. The routinization effect is similar to that observed
in developed countries. Conversely, routine tasks might increase in developing
countries because of the new opportunities arriving as a result of the offshoring
process in developed countries. In addition, the routinization effect of technical
change may be mitigated as the feasibility of automation within the country
plays a crucial role. The automation potential would be limited by the technical
absorptive capacity of the country and the availability of skills in the workforce.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
64 Detailed Literature Review
Consequently, the relationship between technical change and job polarization is
not as straightforward as in developed countries.
Impacts of Technology Adoption on Trade and Mobility
Broadband Internet and other digital technologies open access to large and
diverse new markets, mainly by substantially reducing entry and transactions
costs. Akerman, Gaarder, and Mogstad (2015) contend that the Internet is biased
against low-skilled workers. However, online trade specifically could challenge
this contention by making the Internet available to everyone, with a possible
decrease in wage skill premiums through reductions in the cost of reaching
consumers in foreign markets brought about by online platforms (Lendle et al.
2016). As the cost of exporting declines, smaller firms, which generally have a
more low-skilled workforce, are able to export. This suggests that as online
exports develop, a reduction in the wage skill premium should be expected.
Indeed, the Internet and digital technology are challenging constraints on
the mobility and reach of labor and human capital. The literature has recently
produced various estimates of mobility costs (Hollweg et al. 2014). Kennan
and Walker (2011) develop a model of individual migration in which
expected income is the main force influencing migration. They test their
model using detailed U.S. data on individual workers. They find that inter-
state migration is strongly influenced by the prospect of higher income in
other states, and estimate an elasticity of 0.5 between wages and the migra-
tion decision. They do not consider sector mobility costs and exclusively focus
on regional mobility costs.
Lower mobility costs increase the set of opportunities that the Internet and
digital technologies offer to working people. Using the same kind of theoretical
tools but in a context of trade shocks, Artuç, Chaudhuri, and McLaren (2010)
propose a structural estimation of the reallocation cost of workers across sectors.
Using panel data in which workers’ movements can be observed over time, they
estimate the structural parameters of their model on U.S. data and find an aver-
age moving cost of about 13 times the average worker’s annual wage. Workers
are homogeneous in their model, which may explain the large moving cost they
obtain. Dix-Carneiro (2014) develops a model that assesses the heterogeneity of
workers. Using panel data for Brazilian workers, he estimates an average moving
cost of about two times the average annual worker’s wage. Considering that het-
erogeneity across workers appears to greatly affect the magnitude of the moving
cost, Artuç, Lederman, and Porto (2015) estimate sector mobility costs in a large
number of countries by adapting the methodology in Artuç, Chaudhuri, and
McLaren (2010), implemented using repeated cross-sectional data on sectoral
employment in each country. They find sector mobility costs that are, on average,
three times annual wages. One important difference between all these papers
and the analysis conducted by Cruz, Milet, and Olarreaga (2017) in one of the
background studies for this book is that the new study simultaneously allows for
regional and sector mobility costs whereas the previous papers exclusively focus
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 65
on only one of these two components. The authors find that simultaneously
accounting for both is important to outcomes.
Impacts of Labor Market Policies on Firms’ Technology
Adoption Decisions
Labor policies are intended to help people manage the uncertainty of shocks to
their income, including changes in demand for labor and skills caused by wide-
spread adoption of new technology. Regulations such as a statutory minimum
wage and employment protection legislation; interventions such as social insur-
ance for unemployment, disability, or retirement; and institutions such as the
labor code and the rights and rules of collective bargaining are formulated
according to each country’s norms and policy-making processes. However, many
features are common across countries. Labor market policies and programs are
put in place in an attempt to address well-established labor market failures.
These include, but are by no means limited to, uneven power between those who
seek (firms) and those who sell (individuals) labor and human capital, informa-
tion failures on all sides, and limited or weak insurance markets for mitigating the
risks to household well-being from loss of work and other shocks to income
(Boeri and van Ours 2008; Kuddo, Robalino, and Weber 2015). The predictions
of how regulations such as a minimum wage and restrictions on dismissals create
a wedge between the cost of labor and what people take home are well known
and actively debated (Heckman and Pagés 2004; Pagés, Pierre, and Scarpetta
2009). Furthermore, a large literature has been produced that applies the text-
book models to countries where most people work beyond the reach of regula-
tion in the informal economy (Gill, Montenegro, and Domeland 2002; Perry
et al. 2007; Packard, Koettl, and Montenegro 2012).
Labor market policies can play an important part in shaping firms’ decisions
about adopting new technology. The theory and empirical literature show that
regulations on the form and duration of matches between firms and workers are
likely to have important effects on how quickly or intensively digital technology
is adopted and on how much of an impact that technology can have on jobs
(World Bank 2016). In their analysis of the extent of technology adoption across
developed countries, Alesina, Battisti, and Zeira (2015) find that where labor
regulation was more restrictive, firms’ take-up of technology was greatest in
sectors that used low-skilled labor intensively. They demonstrate theoretically
that regulation raises the cost of low-skilled labor and reduces the skill pre-
mium. Their model shows that more restrictive labor regulation will lead firms
to adopt more labor-saving technology, but in sectors that mainly employ lower-
skilled labor. They posit, conversely, that firms in sectors that use high-skilled
labor more intensively will adopt less technology. Thus, among the otherwise
relatively homogeneous OECD member countries, their model predicts rela-
tively higher levels of technology adoption in lower-skilled manufacturing in
countries such as Spain and Italy, where labor regulation is more restrictive on
firms’ decision than in the United Kingdom and United States, where firms’
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
66 Detailed Literature Review
choices are less constrained by labor regulation. The same prediction is made by
Acemoglu and Restrepo (2016) in their consideration of firms’ decisions to
automate production by investing in industrial robots. Empirically, Alesina,
Battisti, and Zeira (2015) show that more restrictive labor regulation lowers the
ratio of capital in high-skill sectors to capital in low-skill sectors, which then
lowers productivity (output per worker) in high-skill sectors and raises produc-
tivity in low-skill sectors. Furthermore, the authors find that countries with
more restrictive labor regulation tend to produce more patents in the low-skill
sectors. The authors conclude that in countries with more stringent labor regu-
lation, production in the low-skill sectors will become more capital intensive
and firms will be more likely to innovate than will firms in the high-skill sectors
(and vice versa).
Of the regulatory instruments typically in place, restrictions on firms’ hiring
and dismissal decisions appear particularly important to technology adoption.
Gust and Marquez (2004) present empirical evidence that across industrialized
countries, ICT investment is negatively correlated with employment protection
legislation: where firms’ human resources decisions are more constrained by
regulation, investment in digital technology is lower. They develop a dynamic
model of vintage capital and SBTC. In each period a firm decides whether to
upgrade technology, which, in turn, requires improving the skills of the labor
force. Dismissal costs delay or prevent firms’ decisions to adopt the latest
technology. Employers that are unable to change their workforces to keep up
with new technology or otherwise align their workers with changing needs and
new processes within the firm can soon find themselves at a competitive
disadvantage.
Employment protection legislation can discourage firms from undertaking any
risky activity, such as investments in innovation, including adoption of digital
technology. The literature shows the discouragement effect of employment pro-
tection legislation is particularly strong in ICT-intensive sectors (Bartelsman,
Gautier, and de Wind 2016; Saint-Paul 2002; Koeniger 2005; Bartelsman and
Hinloopen 2005; Samaniego 2006). Bartelsman, Gautier, and de Wind (2016)
argue that, because of the experimentation and changes required in organiza-
tional structure, the outcome of investment in ICT is highly uncertain. This
theory is supported by the empirical finding that productivity is more dispersed
in ICT-intensive sectors (Brynjolfsson and Hitt 2003). If a given firm investment
in ICT is unsuccessful, the firm might be forced to exit the market because it
cannot break even. Thus, incentives to invest in ICT depend on exit costs, with
higher exit costs being detrimental to investment in ICT. In this scenario,
employment protection measures such as complex dismissal procedures, direct
firing costs, and restrictions on the use of temporary and fixed-term workers are
a barrier to investment in ICT.
Industries that require a greater degree of risky innovation may have a smaller
footprint in countries with very restrictive labor regulation. Using a cross-
industry and cross-country panel data set of the United States and the European
Union (EU KLEMS), Bartelsman, Gautier, and de Wind (2016) show that
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 67
high-risk, ICT-intensive sectors are smaller in countries with stricter labor regula-
tions, measured by the number of people they employ. These empirical facts hold
when comparing the European Union with the United States, where labor regu-
lations are more flexible, and when comparing countries within the European
Union. Furthermore, the effect of labor regulation is increasing in the risk of the
investment. The paper shows that aggregate productivity in the United States
would be 10 percent lower as a result of lower investment in ICT if sever-
ance payments in the United States were similar to the average severance cost
in Europe.
However, some research shows that restrictions on dismissal can have an
enabling effect on firms’ technology adoption. Acharya, Baghai, and Subramanian
(2013) make the argument that stricter labor laws work as a commitment device
by preventing firms from dismissing workers after short-term failures and thus
encouraging employees to engage in risky, innovative activities that are profit
maximizing in the long term. Furthermore, by creating a “tax on dismissals,”
employment protection legislation may increase the incentives that firms have to
train workers to use new technology and make them more productive.
The dearth of research on labor regulation and technology adoption in emerg-
ing market economies reveals an opportunity for learning. The current literature
on the relationship between labor market regulation and technology adoption
appears only to cover OECD member countries and some higher-middle-income
countries in the process of becoming members. The restriction of analytical work
to these countries is a critical shortcoming in the existing body of evidence.
Considerable added insight could be gained from analyzing the much broader
variation in labor market institutions and key contextual factors (such as level of
development, economic stability, and degree of openness to trade) across high-,
middle-, and low-income countries that firms in emerging market economies
face when choosing whether to adopt new technologies.
Notes
1. Their analysis draws on the EU KLEMS data set, an industry-level panel covering
OECD countries since 1970, limited to nonfarm employment across 32 industries for
developed countries of the European Union, and excluding Eastern Europe plus
Australia, Japan, the Republic of Korea, and the United States.
2. Bessen (2017) analyzes the productivity and jobs growth dynamics in industries such
as cotton textiles, where labor productivity increased nearly 30-fold and consumption
increased 100-fold during the nineteenth century, supporting a rapid increase in
employment, followed by a decline in employment in later stages of maturity (the
broadwoven fabrics industry using cotton and manmade fibers declined from 300,000
to 16,000 production workers between 1958 and 2011). He interprets this pattern
through a model of heterogeneous final demand, in which price declines in the initial
stages of productivity growth make formerly prohibitively expensive products afford-
able for mass consumption, yielding a large positive demand response. Once large
unmet needs are saturated and demand becomes less elastic, further productivity gains
may bring reduced employment.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
68 Detailed Literature Review
3. These findings are consistent with the recent findings of Lawrence (2017), who shows
that for many decades, the relatively faster productivity growth interacting with unre-
sponsive demand, and not trade performance, has been the dominant force behind the
declining share of employment in manufacturing in the United States and other
developed OECD countries—with aggregate job gains driven by more responsive
(more income elastic) demand for services.
References
Acharya, V., R. Baghai, and K. Subramanian. 2013. “Labor Laws and Innovation.” Journal
of Law and Economics 56 (4): 997–1037.
Alvarez, R., J. M. Benavente, R. Campusano, and C. Cuevas. 2011. “Employment
Generation, Firm Size, and Innovation in Chile.” Technical Note IDB-TN-319, Inter-
American Development Bank, Washington, DC.
Aboal, D., P. Garda, B. Lanzilotta, and M. Perera. 2011. “Innovation, Firm Size,
Technology Intensity, and Employment Generation in Uruguay.The Microeconometric
Evidence.” Technical Note IDB-TN-314, Inter-American Development Bank,
Washington, DC.
Acemoglu, Daron, and David Autor. 2011. “Skills, Tasks and Technologies: Implications for
Employment and Earnings.” In Handbook of Labor Economics, volume 4, edited by
David Card and Orley Ashenfelter, 1043–171. Amsterdam: Elsevier.
Acemoglu, Daron, and Pascual Restrepo. 2016. “The Race between Machine and Man:
Implications of Technology for Growth, Factor Shares and Employment.” Working
Paper 22252, National Bureau of Economic Research, Cambridge, MA.
———. 2017. “Robots and Jobs: Evidence from US Labor Markets.” Working Paper 23285,
National Bureau of Economic Research, Cambridge, MA.
———. 2018. “Artificial Intelligence, Automation and Work.” In Economics of Artificial
Intelligence, edited by Ajay Agarwal, Avi Goldfarb, and Joshua Gans. Forthcoming.
Akerman, A., I. Gaarder, and M. Mogstad. 2015. “The Skill Complementarity of
Broadband Internet.” Quarterly Journal of Economics 130 (4): 1781–824.
Alesina, A., M. Battisti, and J. Zeira. 2015. “Technology and Labor Regulations: Theory
and Evidence.” Working Paper 20841, National Bureau of Economic Research,
Cambridge, MA.
Arntz, M., T. Gregory, and U. Zierahn. 2016. “The Risk of Automation for Jobs in
OECD Countries: A Comparative Analysis.” Social, Employment and Migration
Working Paper 189, Organisation for Economic Co-operation and Development,
Paris.
Artuç, E., S. Chaudhuri, and J. McLaren. 2010. “Trade Shocks and Labor Adjustment:
A Structural Empirical Approach.” American Economic Review 100 (3): 1008–45.
Artuç, E., D. Lederman, and G. Porto. 2015. “A Mapping of Labor Mobility Costs in the
Developing World.” Journal of International Economics 95 (1): 28–41.
Autor, D. H., and D. Dorn. 2013. “The Growth of Low-Skill Service Jobs and the
Polarization of the US Labor Market.” American Economic Review 103 (5): 1553–97.
Autor D. H., L. F. Katz, and M. S. Kearney. 2006. “The Polarization of the US Labor
Market.” American Economic Review 96 (2): 189–94.
———. 2008. “Trends in U.S. Wage Inequality: Revising the Revisionists.” Review of
Economics and Statistics 90 (2): 300–23.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 69
Autor, D. H., L. F. Katz, and A. B. Krueger. 1998. “Computing Inequality: Have Computers
Changed the Labor Market?” Quarterly Journal of Economics 113 (4): 1169–213.
Autor, D. H., F. Levy, and R. J. Murnane. 2003. “The Skill Content of Recent Technological
Change: An Empirical Exploration.” Quarterly Journal of Economics 118 (4):
1279–333.
Autor, D.H., and A. Salomons. 2017. “Robocalypse Now: Does Productivity Growth
Threaten Employment?” Conference Paper for European Central Bank Sintra Forum
on Central Banking, June 26–28.
Bartel, A. P., C. Ichniowski, and K. L. Shaw. 2007. “How Does Information Technology
Really Affect Productivity? Plant-Level Comparisons of Product Innovation,
Process Improvement and Worker Skills.” Quarterly Journal of Economics 122 (4):
1721–58.
Bartelsman, E. J., P. Gautier, and J. de Wind. 2016. “Employment Protection, Technology
Choice, and Worker Allocation.” International Economic Review 57 (3): 787–826.
Bartelsman, E. J., and J. Hinloopen. 2005. “Unleashing Animal Spirits: Investment in ICT
and Economic Growth.” In The Economics of the Digital Society, edited by L. Soete and
B. ter Weel, 272–304. Cheltenham, U.K.: Edward Elgar.
Benavente, J. M., D. Bravo, and R. Montero. 2011. “Wages and Workplace Computer Use
in Chile.” The Developing Economies 49 (4): 382–403.
Bender, S., N. Bloom, D. Card, J. van Reenen, and S. Wolter. 2016. “Management Practices,
Workforce Selection and Productivity.” Working Paper 22101, National Bureau of
Economic Research, Cambridge, MA.
Bessen, James. 2017. “Automation and Jobs: When Technology Boosts Employment.”
Boston University Law and Economics Working Paper No. 17-09, Boston University,
Boston, MA.
Bessen, James. 2015. Learning by Doing: The Real Connection between Inovation, Wages and
Wealth. New Haven, CT: Yale University Press.
Bloom, N., L. Garicano, R. Sadun, and J. van Reenen. 2014. “The Distinct Effects of
Information Technology and Communication Technology on Firm Organization.”
Management Science 60 (12): 2859–85.
Böckerman, P., J. Cawley, J. Viinikainen, T. Lehtimäki, S. Rovio, I. Seppälä, J. Pehkonen, and
O. Raitakari. 2016. “The Effect of Weight on Labor Market Outcomes: An Application
of Genetic Instrumental Variables.” Working Paper 22200, National Bureau of
Economic Research, Cambridge, MA.
Boeri, Tito, and Jan van Ours. 2008. The Economics of Imperfect Labor Markets. Princeton,
NJ: Princeton University Press.
Bresnahan, T., E. Brynjolfsson, and M. L. Hitt. 2002. “Information Technology, Workplace
Organization, and the Demand for Skilled Labor: Firm-Level Evidence.” Quarterly
Journal of Economics 117 (1): 339–76.
Brynjolfsson, E., and L. M. Hitt. 2003. “Computing Productivity: Firm-Level Evidence.”
Review of Economics and Statistics 85 (4): 793–808.
Brynjolfsson, E., and A. McAfee. 2011. Race against the Machine: How the Digital
Revolution Is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming
Employment and the Economy. Lexington, MA: Digital Frontier Press.
———. 2014. The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant
Technologies. New York: W. W. Norton & Company.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
70 Detailed Literature Review
Caroli, E., and J. van Reenen. 2001. “Skill-Biased Organizational Change? Evidence from
a Panel of British and French Establishments.” Quarterly Journal of Economics 116 (4):
1449–92.
Castillo, V., A. Maffioli, S. Rojo, and R. Stucchi. 2011. “Innovation Policy and Employment:
Evidence from an Impact Evaluation in Argentina.” Technical Note IDBTN-341,
Science and Technology Division, Social Sector, Inter-American Development Bank,
Washington, DC.
Crespi, G., and E. Tacsir. 2012. “Effects of Innovation on Employment in Latin America.”
Technical Note IDB-TN-496, Inter-American Development Bank, Washington, DC.
Cruces G., G. Fields, D. Jaume, and M. Viollaz. 2015. “The Growth-Employment-Poverty
Nexus in Latin America.” UNU-Wider, Helsinki.
Cruz, Marcio, Emmanuel Milet, and Marcelo Olarreaga. 2017. “Online Exports and the
Wage Gap.” Unpublished.
de Elejalde, R., D. Giuliodori, and R. Stucchi. 2011. “Employment Generation, Firm Size
and Innovation: Microeconometric Evidence from Argentina.” Technical Note
IDB-TN-313, Inter-American Development Bank. Washington, DC.
De Stefano, T., R. Kneller, and J. Timmis. 2014. “The (Fuzzy) Digital Divide: The Effect of
Broadband Internet Use on UK Firm Performance.” Discussion Paper, School of
Economics, University of Nottingham.
Dix-Carneiro, R. 2014. “Trade Liberalization and Labor Market Dynamics.” Econometrica
82 (3): 825–85.
Doms, M., T. Dunne, and K. R. Troske. 1997. “Workers, Wages, and Technology.” Quarterly
Journal of Economics 112 (1): 253–90.
Frey, C., and M. Osborne. 2013. “The Future of Employment: How Susceptible Are Jobs
to Computerisation?” Working Paper, Oxford Martin School, University of Oxford.
Gill, Indermit, Claudio Montenegro, and Dorte Domeland. 2002. Crafting Labor Policy:
Techniques and Lessons from Latin America. Washington, DC: World Bank.
Goldin, C. D., and L. F. Katz. 2009. The Race between Education and Technology. Cambridge,
MA: Harvard University Press.
Graetz, G., and G. Michaels. 2015. “Robots at Work.” Discussion Paper 1335, Centre for
Economic Performance, London.
Gust, C., and J. Marquez. 2004. “International Comparisons of Productivity Growth: The
Role of Information Technology and Regulatory Practices.” Labour Economics 11 (1):
33–58.
Hallward-Driemeier, Mary, and Gaurav Nayyar. 2017. Trouble in the Making? The Future
of Manufacturing-Led Development. Washington, DC: World Bank.
Heckman, James J., and Carmen Pagés. 2004. Law and Employment: Lessons from Latin
America and the Caribbean. Chicago: University of Chicago Press.
Hollweg, C. H., D. Lederman, D. Rojas, and E. R. Bulmer. 2014. Sticky Feet: How Labor
Market Frictions Shape the Impact of International Trade on Jobs and Wages. Directions
in Development. Washington, DC: World Bank.
Kennan, J., and J. R. Walker. 2011. “The Effect of Expected Income on Individual
Migration Decisions.” Econometrica 79 (1): 211–51.
Koeniger, W. 2005. “Dismissal Costs and Innovation.” Economics Letters 88 (1): 79–84.
Krueger, A. B. 1993. “How Computers Have Changed the Wage Structure: Evidence from
Microdata, 1984–1989.” Quarterly Journal of Economics 108 (1): 33–60.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Detailed Literature Review 71
Kuddo, Arvo, David A. Robalino, and Michael Weber. 2015. “Balancing Regulations to
Promote Jobs: From Employment Contracts to Unemployment Benefits.” World Bank,
Washington, DC.
Lawrence, Robert Z. 2017. “Recent Manufacturing Employment Growth: The Exception
that Proves the Rule.” Working Paper 24151, National Bureau of Economic Research,
Cambridge, MA.
Lendle, A., M. Olarreaga, S. Schropp, and P.-L. Vezina. 2016. “There Goes Gravity: eBay
and the Death of Distance.” Economic Journal 126 (591): 406–41.
Machin, S., and J. van Reenen. 1998. “Technology and Changes in Skill Structure: Evidence
from Seven OECD Countries.” Quarterly Journal of Economics 113 (4): 1215–44.
Maloney, W., and C. Molina. 2016. “Are Automation and Trade Polarizing Developing
Country Labor Markets, Too?” Policy Research Working Paper 7922, World Bank,
Washington, DC.
Messina, J., A. M. Oviedo, and G. Pica. 2016. “Job Polarization in Latin America.”
Unpublished, World Bank, Washington, DC.
Michaels, G., A. Natraj, and J. van Reenen. 2014. “Has ICT Polarized Skill Demand?
Evidence from Eleven Countries over Twenty-Five Years.” Review of Economics and
Statistics 96 (1): 60–77.
Monge-González, R., J. A. Rodríguez-Alvarez, J. Hewitt, J. Orozco, and K. Ruiz. 2011.
“Innovation and Employment Growth in Costa Rica: A Firm-Level Analysis.”
Technical Note IDB-TN-318, Inter-American Development Bank, Washington, DC.
Novick, M., and S. Rotondo, eds. 2013. El desafío de las TIC en Argentina: Crear capacid-
ades para la generación de empleo. Santiago, Chile: UN-ECLAC.
Packard, Truman, Johannes Koettl, and Claudio E. Montenegro. 2012. In from the Shadow:
Integrating Europe’s Informal Labor. Washington, DC: World Bank.
Pagés, C., G. Pierre, and S. Scarpetta. 2009. Job Creation in Latin America and the
Caribbean: Recent Trends and Policy Challenges. Washington, DC: World Bank and
Palgrave Macmillan.
Perry, Guillermo, William F. Maloney, Omar S. Arias, Pablo Fajnzylber, Andrew D. Mason,
and Jaime Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. World Bank Latin
American and Caribbean Regional Study. Washington, DC: World Bank.
Saint-Paul, G. 2002. “Employment Protection, International Specialization, and
Innovation.” European Economic Review 46 (2): 375–95.
Samaniego, R. 2006. “Employment Protection and High-Tech Aversion.” Review of
Economic Dynamics 9 (2): 224–41.
Santos, I., S. Monroy, and M. Moreno. 2015. “Technological Change and Labor Market
Disruptions: Evidence from the Developing World.” IZA conference mimeo, Institute
of Labor Economics, Bonn.
World Bank. 2016. World Development Report 2016: Digital Dividends. Washington, DC:
World Bank.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
Environmental Benefits Statement
The World Bank Group is committed to reducing its environmental footprint.
In support of this commitment, we leverage electronic publishing options and
print-on-demand technology, which is located in regional hubs worldwide.
Together, these initiatives enable print runs to be lowered and shipping dis-
consumption, chemical use, green-
tances decreased, resulting in reduced paper
house gas emissions, and waste.
We follow the recommended standards for paper use set by the Green
Press Initiative. The majority of our books are printed on Forest Stewardship
Council (FSC)–certified paper, with nearly all containing 50–100 percent
recycled
content. The recycled fiber in our book paper is either unbleached
rocessed chlorine-free (PCF), or
or bleached using totally chlorine-free (TCF), p
enhanced elemental chlorine-free (EECF) processes.
More information about the Bank’s environmental philosophy can be found
at http://www.worldbank.org/corporateresponsibility.
The Jobs of Tomorrow • http://dx.doi.org/10.1596/978-1-4648-1222-4
While adoption of new technologies is understood to enhance long-term growth and average per capita
incomes, its impact on low-skilled workers is more complex and merits clarification. Concerns abound that
advanced technologies developed in high-income countries could inexorably lead to job losses of
low-skilled, less-well-off workers and could exacerbate poverty. Conversely, there are countervailing
concerns that policies intended to protect jobs from technology advancement would themselves stultify
progress and depress productivity.
The Jobs of Tomorrow squarely addresses both sets of concerns with new research showing that adoption of
digital technologies offers a pathway to more inclusive growth by increasing adopting firms’ outputs, with
the job-enhancing impact of technology adoption assisted by growth-enhancing policies that foster sizable
output expansion. The research reported herein demonstrates, using economic theory and data from
Argentina, Brazil, Chile, Colombia, and Mexico, that low-skilled workers can benefit from adoption of
productivity-enhancing technologies that are biased toward skilled workers, and often do.
The inclusive job outcomes arise when the effects of increased productivity and expanding output overcome
the substitution of workers for technology. While the substitution effect replaces some low-skilled workers
with new technology and more highly skilled labor, the output effect can lead to an increase in the total
number of jobs for less-skilled workers. Critically, output can increase sufficiently to increase jobs across all
tasks and skill types within adopting firms, including jobs for low-skilled workers, as long as low-skilled task
content remains complementary to new technologies and related occupations are not completely
automated and replaced by machines. It is this channel for inclusive growth that underlies the power of
competition-enabling policies and institutions—such as regulations encouraging firms to compete and
policies supporting the development of skills that technology augments rather than replaces—to ensure
that the positive impact of technology adoption on productivity and low-skilled workers is realized.
Cross-country studies highlight additional channels from digital technology adoption to inclusive growth,
namely by lowering the fixed costs of exporting through online trading platforms, and by reducing the cost
of information about job opportunities through Internet-enabled worker-firm job matches.
ISBN 978-1-4648-1222-4
SKU 211222