Innovation and Diversification Policies for Natural R

Innovation and Diversification Policies
for Natural Resource Rich Countries
Mueid Al Raee
UNU MERIT, UM MGSoG
Supervisors
Professor Jo Ritzen
Dr. Denis de Crombrugghe

2
3
Contents
1. Introduction
...............................................................................................
..................... 7
Appendix 1-A
...............................................................................................
.................... 16
2. Productivity and Innovation Policy
.............................................................................. 19
2.1. Introduction
...............................................................................................
............ 20
2.2. Innovation policies and the path towards successful
innovation ............................ 24
2.3. Identification Strategy
...........................................................................................
29

2.4. Data
...............................................................................................
....................... 32
2.5. Results
...............................................................................................
.................... 36
2.5.1. Global
...............................................................................................
.............. 36
2.5.2. Arabian Gulf countries - A special case?
........................................................ 42
2.6. Conclusions and Discussion
................................................................................... 45
Appendix 2-A
...............................................................................................
.................... 49
Appendix 2-B
...............................................................................................
.................... 50
3. Policy and Economy in the GCC
.................................................................................. 53
3.1. Introduction
...............................................................................................
............ 55
3.2. Perspectives on innovation
.................................................................................... 59

3.2.1. General
...............................................................................................
............ 59
3.2.2. The literature on GCC countries
.................................................................... 61
3.3. The Case of GCC – Policies and Enablers
............................................................ 68
3.3.1. Section Summary
............................................................................................
68
3.3.2. Development of education systems
................................................................. 71
3.3.3. Literacy, primary education, secondary education,
reforms and performance 72
3.3.4. Tertiary education and vocational education
................................................. 77
3.3.5. R&D
...............................................................................................
................ 80
4
3.3.6. Business and
Entrepreneurship.....................................................................
.. 85

3.3.7. Governance and Infrastructure
....................................................................... 87
3.4. The Outputs of GCC – Indicators of Innovation and
Diversification .................... 89
3.4.1. Section Summary
............................................................................................
89
3.4.2. Patents, Trademarks and Industrial designs
.................................................. 91
3.4.3. Non-traditional sector - share in the economy and labour
productivity ......... 94
3.5. Connecting Policies, Enablers and Outcomes
........................................................ 96
3.6. Summary, Discussion and Conclusion
.................................................................... 98
Appendix 3-A
...............................................................................................
.................. 103
4. Natural Resource Abundance: No Evidence of an Oil Curse
...................................... 107
4.1. Introduction
...............................................................................................
.......... 109
4.2. Literature Review
...............................................................................................

. 110
4.3. Modelling the natural resource extraction and capital
investment relationship ... 113
4.4. Empirical Model
...............................................................................................
... 113
4.5. Data
...............................................................................................
..................... 121
4.6. Data Reliability
...............................................................................................
.... 124
4.7. Results
...............................................................................................
.................. 127
4.8. Postestimation tests and robustness
.................................................................... 133
4.9. Discussion and Conclusion
................................................................................... 134
5. “Stars in their Eyes?”
...............................................................................................
... 137
5.1. Introduction
...............................................................................................
.......... 138
5.2. Background and Literature

.................................................................................. 140
5.2.1. Diversification
.................................................................................... ..........
140
5.2.2. Evaluation of Diversification Strategies
........................................................ 140
5.2.3. Methodologies for Evaluation
....................................................................... 143
5
5.2.4. Oman and Saudi Arabia Evaluations
........................................................... 144
5.3. The Predictive Model
..........................................................................................
146
5.4. Review of the Economic Plans of Oman and Saudi Arabia
................................. 147
5.4.1. Oman
...............................................................................................
............ 150
5.4.2. Saudi Arabia
...............................................................................................
. 151
5.4.3. Reference Condition and Scenarios
............................................................... 152

5.5. Results and Discussion
........................................................................................ 154
5.6. Summary and Conclusion
.................................................................................... 162
6. Conclusion
....................................................................................... ........
................... 165
6.1. Background of the dissertation
............................................................................ 165
6.2. Summary
.......................................................................................... .....
.............. 166
6.3. Limitations and Suggestions for Future Research
................................................ 171
6.4. Integrated insights from the dissertation and policy
implications ........................ 174
6.4.1. Institutional Effectiveness, Productive Efficiency, Human
Capital, Education
and R&D
...............................................................................................
..................... 174
6.4.2. Natural Resources, Oil, Productivity and Investment
.................................. 176
6.4.3. Regional Infrastructure, International Trade and Peace
.............................. 178

7. References
...............................................................................................
.................... 181
6
7
1. Introduction
A challenge that many countries face today is the sustenance of
their economic growth in the
face of increasing reliance on natural resources. The number of
natural resource-driven
economies has increased from 58 in 1995 to 82 in 2017. 1
Among these countries, 57% were
low- and lower-middle-income countries, while only 13% were
high-income countries. The
natural resource-driven countries were home to more than two-
thirds of all people living in

extreme poverty. If low-income natural resource-driven
economies engaged in sound policy
for effective and efficient utilisation of natural resources, aimed
at broad economic
development, it is expected that almost half of the world’s poor
could be lifted out of poverty
by 2030. 2 This number is more than the number of poor people
lifted out of poverty due to
China’s rapid economic development from 1996 to 2015. In the
face of the fast-evolving global
demand for natural resources, diversification of the economy
offers a path for economic
development in low- and middle-income natural resource-driven
economies and for sustaining
economic growth in the high-income ones.
The central aim of this dissertation is to examine the challenge
faced by natural resource-
driven countries, in particular, the countries in the Arabian
Peninsula, to diversify their
economies. We investigate what policies can help stimulate
innovation and diversification in
natural resource-driven economies to ensure sustained
development. The research carried out
in this dissertation draws upon the evidence of the policies for

development in the global
context. This part of the dissertation is complemented by
research on the state of
1 Natural resource-driven economies are defined as those that
qualify under at least one of the criteria:
1. Natural resource rents are higher than 10% of the economic
output of the country, and/or,
2. Natural resource rents amount to more than 20% of the fiscal
revenue of the country, and/or,
3. Natural resource rents represent more than 20% of the total
exports of the country.
The natural resource-driven classification is used by various
sources (IMF, 2012; Dobbs, et al., 2013; Addison &
Roe, 2018). The 2017 classification of natural resource-driven
countries is based on the author’s calculations.
More details are presented in Appendix 1-A, Table 1.1.
2 See Dobbs, et al. (2013, pp. 5, 135-136) for the methodology
used in determining the estimated reduction of
poverty due to the effective and efficient utilisation of natural
resource rents. The poverty line of 1.90 USD 2011
PPP a day is used to determine the number of people living
under extreme poverty.
8
diversification and its relation to the policy measures
undertaken in the Gulf Cooperation

Council (GCC) 3 and its member countries.
The complexity of the problem at hand can be understood by
bringing together multiple
strands of academic literature. Some of the distinct yet
interconnected areas that this
dissertation relies upon include economics and policy of
innovation, growth economics, studies
of natural resource-driven economies and economic
diversification.
A review of the innovation policy literature at the outset of the
research for this dissertation
identified a need for more empirical and theoretical research in
the area. 4 Addressing this
need is expected to push the field further in its current
transition from pre-paradigmatic phase
towards a defined set of theories of innovation poli cy. It was
observed that research from the
broad innovation policy perspective was limited, and more
attention was needed for the
interaction of policy instruments with systemic conditions and
institutional settings.5 Lastly
and most importantly, discussions of suitable methodologies
were necessary to advance the

field and facilitate research that can close the gaps in the
innovation policy literature
mentioned earlier. One of the propositions has been to
understand the policy realm as having
two parallel spaces. The first space covers macro-level,
systemic and institutional enablers
and determinants of innovation, such as governance, education,
research and development
investment, business environment, fiscal policies, and
infrastructure. The second space
encompasses the dynamics of the innovation process itself such
as knowledge and skills
required, the creation of products and services, intellectual
property protection, incentives for
innovation, production factors, value chains and feedback. The
enablers and determinants in
3 The Gulf Cooperation Council (GCC) is the colloquial term
used to refer to the Cooperation Council of the
Arab States of the Gulf (GCC). We use the abbreviation GCC to
refer to the member countries as of 2017 –
Bahrain, Kuwait, Oman, Qatar, Saudi Arabia and the United
Arab Emirates. These are synonymously referred
to as the countries of the Arabian Gulf, countries of the Arabian
Peninsula, countries of the Persian Gulf and
Arab countries of the Gulf. Other countries on the
Arabian/Persian Gulf such as Iran and Iraq, and countries
in accession talks such as Yemen, Jordan and Morocco are not

included.
4 For more information on the evolution of the innovation
policy literature landscape refer to Radosevic (2012),
Fagerberg (2016) and Borras and Edquist (2019).
5 According to Lundvall (2007) systems of innovation in a
narrow sense “leave significant elements of innovation-
based economic performance unexplained”. In the “broad” sense
the core knowledge-producing and disseminating
institutions (such as educational institutes and research units)
are embedded in a wider socio-economic system
and the relative success of innovation policies is a function of
influences and linkages beyond these core
institutions (Freeman, 2002).
9
both these spaces form the focus of the innovation policy
studies. Therefore, it is frequently
recommended to carry out more research that considers the
collective effect of these
determinants of innovation (Freeman, 2002; Lundvall, 2007;
Fagerberg, 2016).
The literature of systems of innovation from a growth theory
perspective has been criticised
for having a narrow conceptual scope. 6 Additionally, the policy
relevance of the results in this

line of research varies depending on the measures of growth and
innovation used.7 The
research for innovation policy through the growth theory lens
has been carried out using
various productivity measures, most prominently Multi-Factor
Productivity (MFP).
However, its use is challenged because of the imperfect
imputation for developing countries
and a lack of direct policy implications that can be derived from
research using MFP. Other
measures also face challenges and their use is context and
target-driven. Given this, the use
of labour productivity-based measures (such as labour
productivity in the modern sector) 8
generally increases the policy relevance of the results and
widens the geographical and
conceptual scope of the innovation policy studies. In addition to
this, innovation studies have
also been criticized as being too centred on rich countries. This
critique is made plain in the
question, “whether innovation systems and policies are only for
the rich” (Perez, 2013, p. 90) 9.
The substantial focus of innovation studies on innovation in the
developed or rich countries

has partially contributed to the lack of understanding about the
determinants of innovation
in a broader country income level context.
6 For more details on the literature of systems of innovation
from growth theory perspective see Fagerberg, et
al. (2013) and (Soete, et al., 2010).
7 The relationship between productivity and innovation, and a
detailed discussion on the pros and cons of the
alternative measures of innovation are presented in Mohnen and
Hall (2013).
8 In this dissertation the terms non-traditional sector and
modern sector exclude the natural resource and
agricultural sectors, unless otherwise mentioned. The sectors
where the productive efficiency is a function of the
knowledge and skills of the labour force is often called the
productive sector or modern sector. In the empirical
literature, the modern sector has varyingly been defined to
exclude either the natural resource sector, or the
agricultural sector, or both sectors. This usage is derived from
the assumption that the natural resource and
agricultural sectors are predominantly non-productive sectors.
9 The question is posed by the editors of “Innovation Studies:
Evolution and Future Challenges” Jan Fagerberg,
Ben R. Martin and Esben Sloth Andersen to Carlota Perez as a
discussion point for her contribution in the book
(Fagerberg, et al., 2013).

10
One of the main limitations of the growth literature, in general,
is to ignore the importance
of natural resources in the economy (Sachs & Warner, 1995;
Auty, 2007). The same limitation
is also observed in the literature on innovation policy from the
growth perspective.
Additionally, the innovation policy mix for countries that are
close to the technological
frontier and for those that are far behind in the catching-up
process is likely to be different
(Szirmai, et al., 2011; Aghion, et al., 2014). Countries with
natural resources form a unique
subset of world economies. Among other reasons, the
uniqueness originates in the use of
natural resource revenues as a facilitator of broader economic
growth. Thus, investigations
that focus on the policy mix that has been used and can be used
in these countries for catch-
up, diversification and stimulation of innovation are warranted.
The natural resource
economics and policy literature for a long time has been
focussed on the so-called natural

resource curse and the search for its determinants. However,
some argue that the outcome of
the mismanagement of national revenues for a natural resource-
driven economy does not differ
from the consequence of the mismanagement of revenues in
countries that are not dependent
on natural resource revenues (Maloney, 2002). Even without the
precise assignment of blame
for the economic woes of the countries rich in natural resources,
the literature appears to lead
towards a probable solution. 10 The general inference is that the
diversification of natural
resource-driven economies is key for ensuring their long-term
economic development. This
concept is not dissimilar from the more generalised perspective
that countries that provide
more diverse products and services are more likely to have
higher output growth stability
(Krishnaa & Levchenko, 2013; Content & Frenken, 2016).
Given the reservations of the neo-classical literature in
considering the role of natural
resources as a contributor to economic growth, the theoretical
contributions in this area have
been limited. Also, the determinants of economic growth in

developed countries have
frequently been used to try to explain the same in low income
economies without
contextualising the research. Diversification policy has long
been a part of the national
discourse of natural resource-driven economies. However,
policy instruments in some countries
10 For more information on the natural resource economics
literature and the natural resource curse debate see
Badeeb, et al. (2017) and Papyrakis (2017).
11
have been limited to monetary and fiscal policy instruments,
and ineptly chosen targets for
economic policy programs make it hard to evaluate the
attainment of a positive net stimulus
or the lack thereof (World Bank Institue, 2010). Such
complications in policy design,
evaluation and implementation are symptomatic of the absence
of research-based innovation
and diversification policymaking. There is a lack of
comprehensive theoretical background
that adequately applies to a broad set of economies including

natural resource-driven and
developing economies. This limitation indicates that the
complacency of the academics and
economic policy consultants is as likely to be blamed as that of
the national governments and
policymakers for the failure of many countries to have an
effective innovation policy. Among
the natural resource-driven economies, the GCC countries have
attracted the general curiosity
of economists. However, the scientific publications aimed at
trying to understand the
economic structure, systems of innovation and the state of
diversification in the GCC
countries have been limited. The most frequently given reason
is the lack of data. In view of
the academic context described in the preceding paragraphs, we
have an opportunity to look
for the key where it is dark 11.
This dissertation presents a conceptual framework that can be
used to analyse the interplay
of the determinants of innovation and diversification. As a first
step, this is used in an
empirical exercise to discover the effective relationship between
selected determinants and

labour productivity growth in the modern sector. A discrepancy
among the regions is observed
and explored further by mapping the enablers and inputs of
innovation and diversification
for three GCC countries. The relationship of these determinants
is discussed in light of the
innovation and labour productivity outputs. These undertakings
support theorising in the
area of innovation policy studies and ensure that such theorising
covers the economic reality
and systemic conditions of all countries regardless of their
income levels.
11. The streetlight effect, or the drunkard's search principle, is a
type of observational bias that occurs when
people only search for something where it is easiest to look. A
parable featuring the Seljuk Sufi mystic Nasrudin
Hodja is considered as the earliest form of the story (Shah,
1964, p. 70). The parable relates him as looking for
his key outside his house because there was more light outside,
while he knew that the key was lost inside the
house. It was popularised in social sciences by Abraham Kaplan
in his book “The Conduct of Inquiry:
Methodology for Behavioural Science” (Kaplan, 1964).
12

The introduction of the natural resource sector in this
dissertation as an important
contributor to economic growth as well as to the development of
the modern sector follows
the same logic as mentioned above. The aim is to ensure that the
research contribution has
broad academic and policy applicability in terms of economic
structure and country income
classification. The research places a special emphasis on the
GCC countries to illustrate the
thesis that important economic questions can be addressed even
in a state of data scarcity.
Economists should never be wary of research on a seldom
studied region. Rather, the opposite
should be the norm, and, purposefully and promisingly, it has
often been so for the last ten
years. The frameworks, tools and models developed in this
dissertation can guide policymakers
to not only fix realistic targets for economic inputs and outputs
but also embed an ex ante
and ex post evaluation of the achievements into their programs.
Accordingly, the use of an
empirical model is presented to highlight the importance of
research-based policymaking.

The findings in the various chapters of this dissertation
illustrate that all the determinants
of the innovation system need to be working well in order to
stimulate innovation and
productivity growth in the modern sector. The highlights, along
with the structure of the
dissertation, are summarised in the following.
Chapter 2, titled “Productivity and Innovation Policy” presents
a conceptual framework of
innovation policy and its empirical application in developed and
developing countries. The
condition of the GCC region in terms of labour productivity
growth in the modern sector is
explored within the context of this model. The results of
Chapter 2 underscore the importance
of investment into enablers of innovation such as tertiary
education, and research and
development expenditures. It is observed that not only the level
of investment matters but
also the effectiveness of the system in which the policy is
executed. This observation highlights
the importance of improving governance and investing in the
development of institutions.

Chapter 2 also reveals that the modern sector in the GCC region
is performing relatively
poorly compared to other regions in terms of labour
productivity growth in the modern sector.
This is followed by a comprehensive review and comparative
analysis of the broad innovation
system in three of the six GCC countries in Chapter 3, titled
“Policy and Economy in the
13
GCC” in the form of an eclectic comparative case-study. We
present a snapshot of the
development of enablers and policies for diversification and
innovation in Oman, Saudi Arabia
and the United Arab Emirates and highlight the limitations of
their systems. The low
innovation and diversification output in these countries is a
consequence of the observed
limitations. This observation leads to the notion that the lowest-
performing policy area limits
the system performance and that it is essential to ensure the
robust performance of all the
determinants of the system.

Chapter 4 of the dissertation, “Natural Resource Abundance: No
Evidence of an Oil Curse”,
outlines a theoretical model in order to examine the possibility
of using natural resource
revenues to fund fixed capital investments and develop the
modern sector. It also examines
empirically, to what extent oil wealth has been used for
diversification by the six GCC
countries. Chapter 4 shows that the lower performance in the
GCC region in terms of labour
productivity growth in the modern sector is not due to natural
resource rents. The GCC
countries have been rather successful in investing their natural
resource revenues into fixed
capital. The chapter highlights that natural resource rents can be
used for the development
of the modern sector.
This is followed by an analysis of the stated economic
diversification policies of Oman and
Saudi Arabia contrasted against the predicted outcomes in
Chapter 5, titled “Stars in their
Eyes?”. It is an ex ante evaluation of national policy
programmes based on the empirical
model developed in Chapter 2. The challenges and prospects for

meeting the desired and
declared diversification targets are also discussed. The results
of Chapter 5 project that Oman
and Saudi Arabia are not likely to meet their 2030 targets for
diversification. A research-
backed discussion of the policy limitations and possible
pathways to meet future targets is
thus presented. It is observed that the several enablers that may
be improved through
investments must all be understood in the national and regional
context in order to achieve
successful diversification and improvements in the innovation
system of the country. In the
final Chapter 6, a short review of the background and results of
the dissertation are presented.
It summarises the chapters of the dissertation, formulates
academic research
14
recommendations, acknowledges some of the limitations of this
dissertation, proposes an
integrated view of the dissertation and outlines the policy
implications of the research carried

out. The synopsis of this dissertation is that there is no natural
resource curse and that there
is no one policy or area of investment that drives growth and
productivity in the modern
sector. A broad and holistic approach to academic enquiry in
innovation policy and
productivity growth in the modern sector is justified and
recommended. The policy relevance
of this dissertation, already discussed in the subsection
“Integrated insights from the
dissertation and policy implications,” is exposed a second time
in the “Valorisation
Addendum”.
The successes and limitations in the policy actions of the GCC
countries brought to light in
this dissertation support a broad approach to policymaking for
diversification. The GCC
countries started using their oil revenues for economic and
social development while in a
position of relative poverty in comparison to the rest of the
world (Khalaf & Hammoud, 1987;
Pamuk, 2006). 12 They invested in human capital, fixed capital
and the improvement of the

standards of living of their population. As a result, the GCC
countries successfully eradicated
extreme poverty in their countries, and by 2017 all the six GCC
countries were included in
the list of “very high human development countries” based on
their Human Development
Index (HDI) (UNDP, 2018; GCC-STAT, 2019). Along with
using their natural resources for
human development, the GCC countries have attempted to
stimulate diversification of their
economies and ensure sustained economic development. They
have been successful to varying
degrees and are undertaking policy actions to deepen
diversification aimed at securing and
increasing the prosperity gains of the last fifty years. The policy
lessons from the GCC
countries are critical for other natural resource-driven
economies. The empirical evidence from
this dissertation, on innovation policy, diversification and
natural resource-based
development, helps illustrate the importance of innovation and
diversification policy research.
The results demonstrate that low- and low-middle-income
natural resource-driven countries

12 By 1970s almost all GCC countries had control over a
substantial portion of the oil revenue generated from
oil extraction in their countries.
15
can utilise their natural resources effectively and efficiently and
aim their policies at diversified
production and broad economic development. Such
policymaking is expected to support
improvements in the state of their human development as it has
been accomplished by the
GCC countries, and also help eradicate poverty in all forms.
16
Appendix 1-A
Table 1.1 – Classification of countries as natural resource-
driven economies
Natural resource-driven economies are defined as those that
qualify under at least one of the criteria:
Export: Natural resource rents represent more than 20% of the

total exports of the country.
Revenue: Natural resource rents amount to more than 20% of
the fiscal revenue of the country.
Output: Natural resource rents are higher than 10% of the
economic output of the country.
Based on Addison and Roe’s (2018) definition in the book
“Extractive Industries” which is adapted from
McKinsey & Company’s report “Reversing the curse” (Dobbs,
et al., 2013). The resource dependence
classification, country income levels and poverty dynamics were
inspired by the mentioned references and are
based on the author’s calculations using the World Bank (2019)
data published under Creative Commons
Attribution 4.0 International License (CC-BY 4.0).
Country Name Exports Revenue Output World Bank Income
Level
Algeria ● ● ● Upper Middle Income
Angola ● ● ● Upper Middle Income
Armenia ● Upper Middle Income
Australia ● ● High Income
Azerbaijan ● ● ● Upper Middle Income
Bahrain ● High Income
Benin ● ● Low Income
Bhutan ● Upper Middle Income
Bolivia ● Upper Middle Income

Brazil ● Upper Middle Income
Brunei Darussalam ● ● ● High Income
Burkina Faso ● ● ● Low Income
Burundi ● Low Income
Cameroon ● ● Lower Middle Income
Central African Republic ● ● ● Low Income
Chad ● ● ● Low Income
Chile ● ● ● High Income
Colombia ● ● Upper Middle Income
Comoros ● Lower Middle Income
Congo, Dem. Rep. ● ● ● Low Income
Congo, Rep. ● ● ● Lower Middle Income
Cote d'Ivoire ● Lower Middle Income
Ecuador ● ● Upper Middle Income
Egypt, Arab Rep. ● ● Lower Middle Income
Equatorial Guinea ● ● ● Upper Middle Income
17

Level
Eritrea ● Low Income
Ethiopia ● ● ● Low Income
Gabon ● ● ● Upper Middle Income
Gambia, The ● Low Income
Ghana ● ● ● Lower Middle Income
Guinea ● ● ● Low Income
Guinea-Bissau ● ● ● Low Income
Guyana ● ● ● Upper Middle Income
Indonesia ● ● Upper Middle Income
Iran, Islamic Rep. ● ● ● Upper Middle Income
Iraq ● ● ● Upper Middle Income
Kazakhstan ● ● ● Upper Middle Income
Korea, Dem. People’s Rep. ● Low Income
Kuwait ● ● ● High Income
Kyrgyz Republic ● ● Lower Middle Income
Lao PDR ● ● Lower Middle Income

Liberia ● ● Low Income
Libya ● ● ● Upper Middle Income
Madagascar ● ● ● Low Income
Malawi ● ● Low Income
Malaysia ● Upper Middle Income
Mali ● ● ● Low Income
Mauritania ● ● ● Lower Middle Income
Mongolia ● ● ● Upper Middle Income
Mozambique ● ● ● Low Income
Myanmar ● ● Lower Middle Income
New Caledonia ● High Income
Niger ● ● Low Income
Nigeria ● ● Lower Middle Income
Oman ● ● ● High Income
Papua New Guinea ● ● ● Lower Middle Income
Peru ● ● Upper Middle Income
Qatar ● ● ● High Income

18
Level
Russian Federation ● ● ● Upper Middle Income
Rwanda ● ● Low Income
Saudi Arabia ● ● ● High Income
Senegal ● ● Lower Middle Income
Sierra Leone ● ● Low Income
Solomon Islands ● ● ● Lower Middle Income
Somalia ● ● ● Low Income
South Africa ● ● Upper Middle Income
Sudan ● ● Lower Middle Income
Suriname ● ● ● Upper Middle Income
Syrian Arab Republic ● ● Low Income
Tajikistan ● Low Income
Tanzania ● ● Low Income
Timor-Leste ● ● ● Lower Middle Income
Togo ● ● ● Low Income

Trinidad and Tobago ● ● ● High Income
Turkmenistan ● ● ● Upper Middle Income
Uganda ● ● ● Low Income
United Arab Emirates ● ● ● High Income
Uzbekistan ● ● ● Lower Middle Income
Venezuela, RB ● ● ● Low Income
Yemen, Rep. ● ● Low Income
Zambia ● ● ● Lower Middle Income
Zimbabwe ● ● Lower Middle Income
19
2. Productivity and Innovation Policy
Education, Research & Development, Governance, Business,
and Productivity
Abstract
In this chapter, we examine the relationship between
“innovation policy” and labour
productivity growth in non-traditional sectors, for a cross-
section of more than 95 developed

and developing countries. We consider that labour productivity
growth in non-traditional
sectors is in part explained by innovation and catch-up. In
developing countries, catch-up is
a substantial contributor to productivity increases in addition to
new-to-the-world
innovations. The ability to catch-up is considered to be
dependent on the absorptive capacities
of the countries. We term the policies that contribute to
improvements in the absorptive
capacity as innovation policies. In this chapter, we include
investments in tertiary education
as a percentage of gross domestic product (GDP), investments
in research and development
(R&D) as a percentage of GDP, the freedom in the business
environment, as well as overall
government effectiveness. Our results confirm the convergence
of non-traditional sector labour
productivity amongst the countries. We could show a significant
positive effect of the
interaction between, government effectiveness, and, the
government expenditures in tertiary
education as a percentage of GDP, on labour productivity
growth in non-traditional sectors.

Also, for developing countries, a positive and significant
relationship between the growth
variable and effective R&D expenditures was observed. We
could not uncover a relationship
between other policies considered in this chapter and labour
productivity growth in non-
traditional sectors. Non-traditional sector labour productivity
growth in the oil-rich Arabian
Gulf countries was observed to be consistently slower than
western countries. We propose
that there is the likelihood of higher oil prices crowding-out
innovation in oil-rich countries
while stimulating innovation in oil-importing countries.
Keywords: Innovation policy, labour productivity, catch-up,
structural change, government
effectiveness, developing countries, Arabian Gulf countries.
JEL Classification: O2, O3, O38, O43, O47
20
2.1. Introduction
In this chapter, we analyse how individual innovation policies

and their interactions influence
labour productivity in non-traditional sectors and innovation
globally (both in developed as
well as in developing countries). We consider the different
strategies that may be required to
innovate under various conditions of development. We also
explore the relationship between
labour productivity growth and innovation policies in Arabian
Gulf countries, that are
characterised by a high share of natural resource rents in the
economy.
Knowledge, technological change, and innovation have been
introduced as drivers of growth
in the growth economics literature as it has moved beyond only
considering capital and labour
as sources of growth 13. It discusses healthy institutions as
necessary for technological change
and points towards innovation policy to nurture the institutions
that promote knowledge
production and technological progress. The lumping together of
the factors that contribute
to human capital, physical capital, and institutional capabilities
has been considered in the
1960s and 1970s as a common deficit in the literature. The need

for a more in-depth enquiry
of complementarities in policies that affect economic activities,
capabilities and institutional
arrangements has been emphasized by Easterly & Levine
(2001), Freeman (2002), Aghion et
al. (2009) and notably the winner of the 2018 Nobel Memorial
Prize in Economics, Paul
Romer (1994). As such, “innovation policy” including,
education policy, R&D policy, business
policy, and governance is considered in this chapter.
The role of education and R&D policy for innovation and
modern sector labour productivity
growth has been explored in the context of developed and
developing countries. This has
promoted the need for enquiries on whether higher education
and R&D expenditures have
dissimilar returns for developed and developing countries
(Krueger & Lindahl, 2001; Keller,
2006; Aghion & Durlauf, 2009). An important insight is that
countries could be more or less
13 Solow’s works, and studies by Denison, showed that
something other than labour and capital was responsible
for increasing growth rates in the US (Solow, 1957; Denison,
1963). Romer (1986) incorporated technology as

an endogenous factor in constructing a model of increasing
returns of technology and knowledge for long-run
growth.
21
efficient in the translation of innovation policies to productivity
growth and innovation
(Griffith, et al., 2004; Ritzen & Soete, 2011; Loukil, 2014).
Therefore, we consider the
effectiveness of the government in analysing the relationship
between labour productivity
growth in non-traditional sectors and innovation policy.
Considering that innovation-based
growth is less likely to affect traditional industries (Becheikh,
et al., 2006), we consider the
response of innovation policies, for both developed and
developing countries, to exhibit in the
non-traditional sector labour productivity growth. That is labour
productivity growth
exclusive of natural resource rents and agricultural value added.
Productivity can increase in three ways: through economies of
scale, through the utilisation

of unused capacity and by innovation. Innovation encompasses
creating new products and
services, adopting technology that is not available anywhere
else, adapting technology that
has been available elsewhere and as such moving towards the
best practice, improving
management techniques, and marketing novelty. Sectors where
capital investment is more
likely to increase economies of scale in the short-run include
traditional natural resource
extraction and agricultural sectors (Behrens, et al., 2007; Kislev
& Peterson, 1996). As such,
labour productivity excluding these traditional sectors is
expected to be more closely
associated with innovation. We term this part of the economy as
the modern sector
(interchangeably, non-traditional sector). In the modern sector
increasing output by utilising
the same input generally requires improving management
techniques (for enhancing
production), better supply chain (for delivering increased inputs
and production outputs),
and marketing innovation (for ensuring that production is
consumed in the market).

Productivity is defined as the index of output over the index of
input. Labour productivity is
one of the ways to measure productivity. It is the total output
over the total amount of labour
employed. As such, it does not account for inputs of capital in
the denominator. Another
approximate measure is the Total Factor Productivity also
termed as Multi-Factor
Productivity (MFP). It is measured as the Solow residual or
using the output of the economy
divided by weighted inputs of capital investment and labour.
The standard weighting of 0.7
for labour and 0.3 for capital is often employed. A concern with
the output measure used for
22
the MFP is that quality improvement is not accounted for within
this indicator as price
changes do not necessarily reflect the changes in quality. Using
output and input deflator is
also inaccurate as we may be undervaluing output deflators or
overvaluing input deflators
and by doing this, we might be reflecting price effects and not

necessarily innovation effects.
In their work “Innovation and Productivity: An Update”,
Mohnen and Hall (2013) provide a
brief review of the empirical literature and conclude that
innovation leads to better
productivity performance. They note that innovation affects
productivity increases through
both real output and the price. Innovators take advantage of the
two effects in order to
increase their profitability. Firstly, they are able to sell their
products or services at above
competitive prices, as in monopolistic competition at least for
some time. Secondly, the
product or service itself is higher quality and delivers higher
value to the consumer – this is
called the real innovation output. In terms of the limitation of
the productivity measure,
Mohnen and Hall find that the innovation price effect and the
innovation product effect (real
output) are hard to disentangle in the absence of good
individual price measures. It is
important to note that in this chapter, we are interested in
capturing catch-up effects as well
as new-to-the-world innovation effects. The innovation price

effect and innovation product
effect are both important in this context and the
disentanglement is not necessitated. Product,
process, organisational and marketing innovations all improve
productivity through either or
both of these effects, in other words, innovation leads to better
output per employee
performance.
Innovation counts are widely used to represent innovation,
specifically to account for the
innovation product effect. However, innovation counts are not
without limitations. In fact,
the reason that researchers are unable to find a stronger
association between productivity and
innovation may be related to the use of imperfect measures to
represent innovation (Mohnen
& Hall, 2013). It is thus suggested that recording innovation
counts in a continuous measure
including quality of innovation instead of a binary “yes or no”
format would lead to more
meaningful and robust estimates on the relationship between
productivity and innovation.
Patents while promising are controversial as they represent only
a small part of innovation.

23
Some patents may never be converted to finished and marketed
products or services. At the
same time, some innovation may never be patented. Also, new -
to-the-country and new-to-
the-region innovations may not be reflected in patent data. This
type of innovation is vital
in developing countries and accounts for the majority of
increases in labour productivity
through innovation.
In this chapter, we are interested in increases in labour
productivity associated with
innovation policy. As discussed above, we ensure that labour
productivity is closely related
to innovation by using labour productivity in the modern sector.
The issue of selecting the
most appropriate measure is however complicated further when
considering the availability
of data for a broad set of countries and, in particular, for the
Arab countries of the Gulf that
are the central focus of this dissertation. One commonly quoted
disadvantage of productivity

data is that of varying recording standards among countries. We
find that labour productivity
data is most widely available across countries and time. In order
to account for the probable
variation in recording standards, we use standardised data from
a reliable source.
Additionally, it is also important for us to use a measure that
has direct policy implications.
In the construction of MFP, which is derived as the residual of
the Cobb-Douglas equation,
the unit of MFP does not have a simple economic interpretation
and appears to be a modelling
artefact. The use of labour productivity in the modern sector
ensures the ability to derive
simple economic interpretation of the results that can inform
policymakers on innovation as
well as diversification-concerns.
The catch-up process in developing countries is dependent on
absorptive capacity. The
countries as such can induce catch-up and increases in labour
productivity by improving
education, R&D, business environment and governance.
Innovation policy is a broad term

and successful innovation is related to many systemic processes
working in sync. As such, we
acknowledge that defining the ideal combination of institutions
and policies that are
important for innovation is a complex task. We consider
innovation policies as those directed
at improving the innovative and absorptive capacity of the
countries. In this chapter
innovation policy is characterised by the interaction of
expenditures on education with the
24
effectiveness of government, the interaction of expenditures on
R&D with the effectiveness of
government and business policy. Our identification strategy
comprises five-year labour
productivity growth rates regressed as a function of initial
labour productivity, relevant lagged
innovation policy variables, control variables and dummy
variables.
The literature on innovation, growth and productivity,
innovation policies, education, R&D,
business, governance and the associated relationships is

discussed in Section 2.2. Our
identification strategy is drawn out in Section 2.3. The data are
presented in Section 2.4. The
results related to the effect of innovation policies in the global
context, as well as for developed
and developing countries are presented in Section 2.5. Finally,
we discuss the results, their
policy implications, and present our concluding remarks in
Section 2.6.
2.2. Innovation policies and the path towards successful
innovation
The innovation literature is distributed between “narrow” and
“broad” focus on innovation
policies. In the “narrow” sense, only formal R&D systems and
organizations systematically
active in knowledge generation and diffusion are the focus. An
example of the application of
the systems of innovation framework in the former sense is the
World Bank Knowledge
Assessment Methodology (Chen & Dahlman, 2005). However,
systems of innovation in a
narrow sense “leave significant elements of innovation-based
economic performance
unexplained” (Lundvall, 2007). In the “broad” sense the core

knowledge-producing and
disseminating institutions are embedded in a wider socio-
economic system, and the relative
success of innovation policies is a function of influences and
linkages beyond these core
institutions (Freeman, 2002; Soete, et al., 2010). Among the
works that discuss new-to-the-
world innovation in the latter sense, Furman, Porter and Stern
(2002) integrate ideas-driven
growth theory, microeconomics-based models of national
competitiveness and industrial
clusters theory. They consider R&D manpower, knowledge and
technology base as important
sources of innovation. Archibugi & Coco (2004) define
innovation system through patents,
publications, ICT, electricity consumption, and education. To
formulate a comprehensive
narrative, we draw from literature sources focused on individual
policy variables, governance,
25
policy interaction, growth, labour productivity, and innovation.
This conceptualisation is

carried out with the frame of reference that growth is primarily
driven by technological change
and innovation. The three main factors of growth are generally
considered to include MFP,
labour input, physical capital and human capital. Technology
growth and efficiency are
considered as the sub-sections of MFP. Efficiency improvement,
in the very least, need
management and process innovation. Together we term these
components as innovation. This
is the origin of the consideration that innovation and
technological change is the primary
contributor to economic growth (Solow, 1957; Denison, 1963;
Romer, 1986). At the same
time, Nelson and Winter’s (1982) work on the evolutionary
approach views the free-market
economic structure as continuously evolving with emphasis on
the influence of institutions
and government policies on economic activity.
Education is one of the policy variables that affect innovation.
Aghion et. al (2009) set out
to estimate causality of the effect of education on growth, using
actual measures of investment
in education. They question the adequacy of using lagged

spending – in their previous work
(Vandenbussche, et al., 2004) – as an instrument to overcomes
biases caused by omitted
variables such as institutions, especially in the case of sma ll
low variation data. They show
that there are positive effects of exogenous increases in
education expenditure related to four
years’ tertiary education in the states of the United States of
America (US) that are close to
the world frontier (Aghion, et al., 2009). Krueger and Lindahl
(2001) find societal returns to
schooling in terms of increased growth in cross-country
analysis; the relationship is
statistically significant and positively associated with
subsequent growth for countries with
the lowest level of initial education.
Faster growing countries in Asia have had higher expenditures
on primary education. Keller
(2006) does not obtain positive significant results consistently
for the effect of such
investments on growth. However, it is suggested in the paper
that inefficiencies in resource
allocation of secondary and tertiary education expenditures may
be the reason behind. In

other studies, government expenditures on education relate
positively to growth in developing
countries (Bose, et al., 2007). Also, the inclusion of other
policy variables in the studies, such
26
as openness, public spending, and health variables results in a
lower estimated impact of
education on growth (Benos & Zotou, 2014). From these
studies, we conclude that education
should be considered as a part of “broad” policies for
innovation-based growth.
Loayza et al. (2005) find that regulatory burden reduces growth.
However, a higher quality
of institutional framework leads to the negative effects of
excessive regulation on growth to
be lessened. In a simple model Djankov et al. (2006) observe
the effect of business regulations
as represented by the doing business indicators while
considering the effect of initial level of
growth, and control variables and other determinants of growth
that include corruption, law

and order, the political system, primary and secondary school
enrolment, and civil conflict.
They find that going from the worst to the best quartile of
business regulation shows a 2.3%
increase in annual growth rate. They also observe that the
effects of improvement in primary
and secondary education from worse to better quartiles of policy
or output are significantly
lower than the effects of business regulation on growth rate.
Hanusch (2012) suggests that
regulations related to credit, contract enforcement, costs, time,
starting a business, registering
property, and protection of investors within the realm of
business policies are statistically
significantly related to economic growth.
Griffith and team (2004) find that R&D, as represented by
BERD, is statistically and
economically important in the catch-up process as well as for
stimulating innovation directly
and suggest that the social rate of return of R&D has been
underestimated in the literature
as many studies only focus on the US. A look into cross-country
labour productivity
differences due to investment in R&D reveals that R&D

investment has a significant positive
impact on productivity (Lichtenberg, 1993). Nadiri and Kim
(1996) find rates of returns of
domestic R&D expenditures to be in the range of 14 to 16% and
adding the effect of spill-
overs of international R&D spending for six (6) advanced
economies showed the returns to
be 23 to 26% varying amongst the countries. Hall, Mairesse,
and Mohnen (2010) in their
review of the econometric literature measuring the private and
social returns to R&D find
that the literature identifies private returns to R&D as strongly
positive, social returns to be
greater than private returns and public-funded R&D to be less
productive in terms of private
27
returns. In many research avenues, the incentives to invest in
R&D is determined on the basis
of private returns and not social returns. It is also observed that
despite having higher social
returns to R&D investments developing countries are not able to
achieve the maximum

potential in R&D. This may be due to inappropriate or
inadequate social policies (Griffith,
et al., 2004).
Jalilian, Kirkpatrick and Parker (2007) find that there is a
strong causal link between
government regulation, regulatory quality indices, and economic
performance. Other cross-
sectional studies also report causal effects of governance on
long-run income per capita, using
instrumental variables (Kaufmann & Kraay, 2002). Also, the
mechanism behind this causal
link has been examined and it has been pointed out that one
path through which government
effectiveness improves economic performance is by creating a
better investment environment
(Kirkpatrick, et al., 2006). It is likely that government
effectiveness translates into high
economic growth not only through the path of providing a good
investment environment but
also by creating a good environment for innovation policies to
be effective.
The common theme that emerges from the literature is that
policies work in coherence with
each other and have a combined and complementary effect on

growth and productivity. The
translation of policy to increased labour productivity growth
must go through the
governments’ ability to effectively convert inputs of policy into
innovative products and
services, as well as innovative management, production, and
marketing practices. In this
respect, non-traditional sector labour productivity is closely
associated with innovation, across
developed as well as developing countries. Most of the
literature referenced here is aimed at
determining the relationship between various policies and
growth measures such as GDP,
Income per Capita and Labour Productivity. In this chapter, we
try to study the relationship
between selected policies and innovation proxied by labour
productivity growth in the modern
sector. As such we are delving deeper into the relationship
between one of the proximate
causes of growth – technological change or innovation and some
of the fundamental causes of
growth – the policies and institutional settings that drive
growth. In this sense in the literature
we have discussed is relevant in two ways – one, in pointing out

the relationship of
28
fundamental causes of growth to growth itself; and the second,
by pointing out that there is
a gap in trying to understand the link between the fundamental
and proximate causes of
growth.
Figure 2.1 – Innovation Policy Framework Conditions
Figure 2.1 above represents our interpretation of how the flows
of knowledge enable an
increase in innovation. The innovation eco-system is thus
arranged into conditions, linkages,
the firms and the market itself. The increase in innovation is
linked to increasing labour
productivity through innovative management, design,
production, and marketing techniques.
The change in the state of these conditions is determined
through natural transformation and
29

policy. The education condition is affected by government
policies, such as government
financing of the tertiary education system, policies determining
graduate ratios in science and
technology fields, alignment to labour demand from the market,
university autonomy, and
others. Similarly, research and development conditions are
impacted by the government
expenditure on research and development, type of research
grants, targeted scientific field
grants, the competitiveness of grants, intellectual property
regime, and private-sector research
funding, and so forth. Business conditions are related to
industrial policies, competition policy,
entrepreneurship policy, taxation policy, financial policy, the
health of the financial sector,
availability of finance, and market access for firms that create
new products or services.
Infrastructure conditions include the availability of ICTs,
Transport, Energy, Standard-
Setting, Metrology, Security, among others. Finally, it is
considered that without efficient
and effective linkages the production of knowledge, as well as
transfer of knowledge for the

creation of new products and services, would be hampered. For
innovation to thrive in the
production space it is important that the innovation environment
conditions are healthy,
governed by sound policy, with effective linkages across
various conditions, the production
space and the market for consumption of the innovations.
2.3. Identification Strategy
We use the framework in Figure 2.1 above to understand policy
factors that promote
innovation. Consequently, we explore how individual
innovation policies and their interactions
influence innovation globally, in developed, and developing
countries. As such, labour
productivity growth in the non-traditional sector is modelled as
a function of innovation
policy and the effectiveness of innovation policy.
We assume the drivers of innovation to be: the initial level of
labour productivity, the
interaction of government effectiveness with educational
expenditures, the interaction of
government effectiveness with research and development and a
facilitative business

environment. We estimate the relationship with an Ordinary
Least Squares regression with
exogenous variation in explanatory variables of policy.
30
The response variable is defined as the natural log of the ratio
of final to initial labour
productivity in non-traditional sectors, where final labour
productivity is taken to be five
years after the initial measure. We use four years averages
starting from the initial year to
smoothen out one-off effects for the countries. The explanatory
variables are lagged one year
and include the natural logarithm of initial labour productivity,
interaction of government
effectiveness and government expenditures in tertiary education
as a percentage of GDP
(alternatively called effective tertiary education expenditures in
this chapter), interaction of
government effectiveness and gross expenditures on R&D as a
percentage of GDP
(alternatively called effective R&D expenditures in this chapter)

and index of economic
freedom. Innovation is a medium to long term phenomenon,
and innovation policies typically
take a long time to bear fruit. Using lagged average variables
accommodates for long term
nature of innovation and also provides a way to exclude reverse
causality.
The literature provides evidence that initial level of labour
productivity is a determinant of
labour productivity growth. As such, we account for the initial
level of labour productivity in
the estimation equation (Barro, 1991). Also, the initial level of
education has an impact on
how innovation policies influence the role of tertiary education
expenditures on innovation
itself (Keller, 2006). Natural resources and agricultural
endowments also influence the growth
path of a country (Lederman & Maloney, 2007). Finally, a
country’s regional situation
influences its growth trajectory as well (Moreno & Trehan,
1997). We introduce regional
dummies, educational attainment in terms of years of education
from primary to tertiary
level, natural resource rents as a percent of GDP, and

agricultural value added as a percent
of GDP as additional control variables. Total natural resource
rents as a percent of GDP is
defined as the sum of oil rents, natural gas rents, coal rents,
mineral rents, and forest rents
(World Bank, 2014). Agriculture in Agricultural value added
corresponds to International
Standard Industrial Classification (ISIC) divisions 1-5 and
includes forestry, hunting, and
fishing, as well as cultivation of crops and livestock production.
The estimation equation thus takes the form;
31
Equation 2.1: Δ tN-t1ln(labprod) = α0 + βo • labprod t1 + β1 •
goveff t1 • edu t1 + β2 • goveff t1 • r&d t1 + β3 •
econfreedom t1 + natresrents t1 + agrirents t1 + eduattain t1 +
regional dummies +
Table 2.1 – Variable Definitions
Variable Definition
ΔtN-t1 ln(labprod)
Productivity Growth
Natural log of the ratio of final to initial labour productivity in

non-
traditional sectors (alternatively including traditional; natural
resource and
agriculture sectors – see discussion in the main text)
labprod t1
Initial Productivity
Natural log of initial labour productivity
goveff t1 • edu t1
Effective Tertiary Education
Interaction of government effectiveness and government
expenditures on
tertiary education as a percent of GDP - initial
goveff t1 • r&d t1
Effective R&D
Interaction of government effectiveness and gross expenditures
on research
and development as a percent of GDP - initial
econfreedom t1
Economic Freedom
Index of economic freedom - initial
natresrents t1

Natural Resource Rents
Natural resource rents as a percentage of GDP - initial
agrirents t1
Agricultural Value Added
Agricultural value added as a percentage of GDP - initial
eduattain t1
Educational Attainment
Number of years of schooling from primary to tertiary level -
initial
Note: The subscript “t1” in Equation 2.1 and the reference
“initial” in Table 2.1 specifies the magnitude of the
variable during the initial year(s) considered. The subscript
“tN” in Equation 2.1 and the reference “final” in
Table 2.1 specifies the final year. As such the growth is
considered between t1 and tN. In this chapter, this
period of growth is 5 years and the policy variables are the
average of four years, lagged by one year from the
final year for which a five-year growth rate is considered.
In addition, we evaluate the same equation for labour
productivity growth including natural
resource rents and agricultural value added. This helps us
identify differences in the influence
of innovation policies on innovation-based growth versus mixed
innovation and traditional

sector growth and confirm the robustness of our results. We also
estimate the model for
developed and developing country groups separately to
understand the differences in the
influence of innovation policies and analyse the need for
varying policies for both the groups.
The 15-year data from 1998 to 2013 is regressed in three groups
of five years. The results for
32
each period are observed in order to understand period-specific
differences. These period-
specific differences are controlled-for through a period dummy
in the pooled dataset regression
that is aimed at generating a larger data set leading to
significant and robust coefficients.
2.4. Data
Labour productivity is calculated in terms of real GDP per
labour force. The labour
productivity indicator is constructed by using the GDP from
World Bank Development
Indicators (World Bank, 2014) and the number of employees’

data from Penn Worlds Table
Version 8.1 (Feenstra, et al., 2015). The use of Purchasing
Power Parity (PPP) Constant
2010 USD GDP ensures that the data is comparable across time
and countries in level and
growth rate. Literature suggests that innovation-based growth is
less likely to reflect in
traditional industries such as those in natural resource and
agricultural sectors (Becheikh, et
al., 2006). As such the labour productivity growth measure
excludes natural resource rents
and agricultural value added.
Data for government expenditure on tertiary education as a
percentage of GDP is acquired
from the subset of the UNESCO Institute of Statistics education
dataset that is related to
financial resources (UIS.STAT, 2016). UIS.STAT receives data
on education expenditure from
country governments responding to UIS's annual survey on
formal education. Tertiary
education is considered as one of the most important
contributors to innovation. When
interpreting this indicator, however, we should keep in mind
that in some countries, the

private sector and/or households may fund a higher proportion
of total funding for education,
thus making government expenditure appear lower than in other
countries. Educational
attainment is based on years of school life expectancy primary
to tertiary.
The gross domestic expenditure on research and development
(GERD) as a percentage of
GDP is the total intramural expenditure on R&D performed in a
country or region during a
given year, expressed as a percentage of GDP of the country or
region (UIS.STAT, 2016).
The data is used as an indication of research and development
policy. The ideal case would
be to use GovERD, that is government expenditure on research
and development as a
33
percentage of GDP. We use the GERD measure because it
captures wider geographical space
and time, and is a good representative of what similar higher
expenditures can achieve.
We use the Index of Economic Freedom as an indicator of

government policy towards
business. The Index of Economic Freedom is an annual index
and ranking created by The
Heritage Foundation and The Wall Street Journal in 1995 to
measure the degree of economic
freedom in the world's nations. The creators of the index took
an approach similar to Adam
Smith's in The Wealth of Nations, that “basic institutions that
protect the liberty of
individuals to pursue their own economic interests result in
greater prosperity for the larger
society" (Heritage Foundation & Wall Street Journal, 2016).
The index of economic freedom
is based on ten quantitative and qualitative factors. These ten
factors are property rights,
freedom from corruption, fiscal freedom, government spending,
business freedom, labour
freedom, monetary freedom, trade freedom, investment freedom,
and financial freedom. Each
factor is graded on a scale of 0 to 100. A country's overall score
is derived by averaging these
ten economic freedoms, with equal weight being given to each.
It would have been ideal to use Ease of Doing Business data
from the World Bank Doing

Business Indicators. The ten constituti ve measures used in the
composite ease of doing
business indicator are, starting a business, dealing with
construction permits, getting
electricity, registering property, getting credit, protecting
minority investors, paying taxes,
trading across borders, enforcing contracts and resolving
insolvency (World Bank, 2015). As
such ease of doing business accounts for objective as well as
subjective measures that are
directly related to business policy in the country. However, due
to limited time-period
availability, we resort to using the Index of Economic Freedom
that relates to the business
environment in a relative bird-eye manner.
Government effectiveness captures, “perceptions of the quality
of public services, the quality
of the civil service and the degree of its independence from
political pressures, the quality of
policy formulation and implementation, and the credibility of
the government's commitment
to such policies” (World Bank, 2015). Notably, the indicator is
a mix of quality and perception

of infrastructure, bureaucratic, state, and policy stability. As
such, it is used as a measure of
34
expected effectiveness of innovation policies as related to the
enabling conditions that affect
linkages amongst various policy conditions and knowledge flow
necessary for innovation (See
Figure 2.1 – Innovation Policy Framework Conditions).
Governance is difficult to account for using any kind of
measure. We find it important to
touch upon the topic of the selection of Government
Effectiveness as an interaction term for
the policy measures of expenditures in tertiary education and
research and development in
more detail. The representative sources for constructing this
indicator include quality of
bureaucracy, instituti onal effectiveness, excessive bureaucracy
or red tape, infrastructure,
quality of primary education, satisfaction with public
transportation system, satisfaction with
roads and highways, satisfaction with education system, basic
health services, drinking water

and sanitation, electricity grid, transport infrastructure,
maintenance and waste disposal,
infrastructure disruption, state failure, and policy instability.
The composite is constructed
from a weighted average of the individual indicators obtain
through an Unobserved
Components Model (UCM). The UCM assigns greater weight to
data sources that tend to be
more strongly correlated with each other. This weighting
improves the statistical precision
of the aggregate indicators and typically does not affect the
ranking of countries much on the
aggregate indicators. There are two rationales for using
Government Effectiveness. First, it is
indicative of the governments’ ability to implement their
policies and as such the interactive
term represents the efficiency of each dollar spent. Second, the
interaction of Government
Effectiveness with the expenditures can be looked at with much
simpler view that is of
representing the policies as related to the governance
environment. Both explanations relate
well to the definition of Government Effectiveness indicator
and its use in the context of this

paper and the framework represented graphically in Figure 2.1.
The variable, effective government expenditures on tertiary
education, is constructed by
interacting the index of government effectiveness with the
government expenditures on
tertiary education as a percent of GDP. The same approach is
taken to construct the variable
effective GERD as a percent of GDP. The prefix “effective”
signifies an interaction with the
measure of government effectiveness. Effective expenditure is
obtained by the interaction of
35
government effectiveness that runs from 0 to 1 by actual percent
expenditures per GDP in
the relevant policy areas. As such government effectiveness is
translated as the percentage of
effectiveness of each dollar spent or simply the interaction of
the governance environment
with the policy measures.
Table 2.2 - Summary Statistics
Variable Countries Years Mean Std Dev Min Max

Δ yN-y1 ln(labprod)
Productivity Growth
150 1998-2013 0.559 0.364 -0.452 1.540
labprod y1
Initial Productivity
157 1998-2013 9.614 1.183 6.884 12.084
goveff.edu
Effective Tertiary Education
164 1998-2013 0.483 0.427 0.024 2.198
goveff.r&d
Effective R&D
129 1998-2013 0.485 0.716 0.003 3.427
econfreedom
Economic Freedom
177 1998-2013 57.740 12.190 8.900 89.060
natresrents
Natural Resource Rents
187 1998-2013 6.290 10.560 0 86.170
agrirents
Agricultural Value Added
164 1998-2013 16.830 14.510 0 61.800

eduattain
Educational Attainment
182 1998-2013 11.790 3.210 3.100 20.230
Note: Description of abbreviations is provided in Table 2.1
The indicator used to represent innovation-based growth is the
natural log of the ratio of final
to initial labour productivity excluding natural resource rents
and agricultural value added.
It is noteworthy that the number of countries for which these
data points are available varies
from 129 for effective GERD as a percent of GDP to 177 for the
index of economic freedom
for the year between 1998 to 2013. However, in our regression
between 95 to 106 countries
are represented depending on the time period and extent of the
data available. The correlation
coefficient of effective tertiary education expenditures and
effective research and development
expenditures is 0.57. The same for Economic Freedom with
effective tertiary education
36

expenditures is 0.47 and with the effective research and
development expenditures is 0.51.
The pairwise correlation between our explanatory variables of
concern is considered moderate
and is not expected to have an effect on the coefficients of the
estimation. In order to make
sure that this is the case, we also regress excluding two of the
explanatory variables and
compare the results with the original estimation.
2.5. Results
2.5.1. Global
Here we present the observed influences of the explanatory
variables of concern, on the
response variable i.e. labour productivity growth excluding
natural resource and agricultural
rents. Second, we present the results separately for developed
and developing countries.
Finally, we glance at how labour productivity growth in the
Arabian Gulf countries compares
with western countries (See footnote 18 associated with
Appendix 2-B).
The first result that is observed and presented in Table 2.3
below is that of “beta-convergence”.

This term implies that the partial correlation between growth in
income or productivity over
time, and its initial level is negative. It refers to a process in
which poorer regions grow faster
than richer ones and therefore catch-up on them. We observe
that the initial labour
productivity is negatively and statistically significantly
correlated to labour productivity
growth for pooled data for three periods. An increase of 1% in
the country’s initial labour
productivity results in the ratio of final to initial labour
productivity to be lower by 0.045%.
Countries with relatively lower labour productivity are able to
grow faster and hence converge
to the frontier. In this chapter, we make the observation for
innovation represented by labour
productivity excluding natural resource rents and agricultural
value added. In this context,
the results confirm the convergence of labour productivity
between countries on the
innovative frontier and those away from it.

37
Table 2.3 – Labour Productivity Growth and Policy Variables
Response Variable Productivity Growth
(Net of natural resource rents and agricultural value added)
Period 1 Period 2 Period 3 Pooled Pooled Pooled
1998-2003 2003-2008 2008-2013 Period
1 & 2
Period
2 & 3
Period
1, 2 & 3
Initial Productivity -0.111*** -0.051 0.039 -0.075*** -0.022 -
0.045**
(0.038) (0.037) (0.028) (0.025) (0.027) (0.021)
Effective Tertiary Education 0.024 0.035 0.041* 0.029 0.048*
0.041*

(0.040) (0.041) (0.023) (0.028) (0.025) (0.021)
Effective R&D -0.005 -0.037 -0.01 -0.023 -0.022 -0.018
(0.029) (0.030) (0.019) (0.020) (0.020) (0.016)
Economic Freedom 0.003 0 -0.003 0.002 -0.002 -0.001
(0.002) (0.002) (0.002) (0.001) (0.001) (0.001)
Arabian Gulf Dummy -0.118 -0.268** -0.148 -0.215** -
0.278*** -0.269***
(0.142) (0.125) (0.114) (0.087) (0.097) (0.078)
Period 1
-0.068*** 0.103***
(0.017) (0.019)
Period 2
0.149*** 0.147***
(0.018) (0.017)
Root Mean Squared Error 0.107 0.126 0.090 0.113 0.126 0.120

Adj. R-squared 0.404 0.446 0.138 0.463 0.342 0.383
N 91 96 100 187 196 287
* p<0.10, ** p<0.05, *** p<0.01
Note: Regional dummies, educational attainment, natural
resource rents, and agricultural value added included
as control variables
We observe that effective expenditures on tertiary education as
a percent of GDP have a
positive relationship in all periods with the explanatory
variables. The pooled data for the
three periods shows that there is a statistically significant
positive relationship between
effective tertiary education spending as a percent of GDP of the
country and labour
productivity growth. This is statistically significant at the 10%
level for two sets of pooled
data and for the third period. In this case, the magnitude of the
increase is considerable i.e.
38

an increase of 1% in average effective tertiary education
expenditure as a percentage of GDP
would result in an increase of 4.2% in labour productivity
growth. To simplify, a country
effectively investing 1% of their GDP in tertiary education will
improve their growth rate by
4.2% if they invest an equivalent of 2% of their GDP in tertiary
education. Since this variable
is represented by an interaction of government effectiveness and
tertiary education
expenditure as a percent of GDP it is useful to break down this
result. A hypothetical country
with 1% effective expenditure on tertiary education as a percent
of GDP driven by 0.45
government effectiveness, investing 2.2% of GDP as tertiary
education expenditure and
having an annual labour productivity growth rate of 3% can
improve its productivity growth
rate to 3.13% (that is an increase by 4.2%) by increasing its
effective expenditure to 2% that
could be accomplished either by improving government
effectiveness to 0.9 or tertiary
education expenditure as a percent of GDP to 4.4%. Note that
non-traditional sector labour

productivity is being discussed here, as it is a measure
representing innovation that covers
developing countries as well as developed countries.
We do not observe positive results for effective R&D
expenditure percent of GDP for the
complete set of countries. We observe no statistically
significant results for the Index of
Economic freedom. The magnitudes are small and the pooled
data for the three periods show
inconsistent correlation for the business policy and labour
productivity growth. The
estimation is robust to the inclusion of interaction variable’s
constituents – government
effectiveness, tertiary education expenditure as a percent of
GDP, and GERD as a percent of
GDP – in the estimation in addition to the interaction variables
named effective tertiary
education and effective R&D. Also, the signs of the coefficients
for effective tertiary education,
effective R&D and Economic Freedom do not vary and the
magnitudes do not vary by a
considerable extent when included individually in the
estimation, that is, when the remaining
two explanatory policy variables are excluded. This confirms

that the moderate pairwise
correlation for our explanatory variables discussed in Section 4
has no influence on the results.
Table 2.4 shows the results for developed and developing
countries. It provides a perspective
into differences in the relation of innovation policy to labour
productivity between developing
39
countries and developed countries. We observe in Table 2.4 a
negative relationship between
effective research and development expenditures and labour
productivity in the modern sector
for developed countries. The results may seem precarious at
first. However, the difficulty in
finding a relationship between productivity and innovation in
the developed countries is well
known and termed as the “productivity puzzle” and “Solow
paradox”. We can observe for
developing countries in Table 2.4 that the effective R&D
expenditures variable has a positive
effect on labour productivity growth for developing countries14
and the relationship is

statistically significant at 10% level for the pooled data for
three periods. An increase of 1%
in the effective R&D expenditures as a percent of GDP in
developing countries would result
in the labour productivity growth rate to increase by 27.5% for
pooled data of periods 1, 2
and 3. However, we do not find the same for developed
economies. The differences between
developed countries and developing countries are indicative of
different stages of development.
The developed countries may be more prominently engaged in
new-to-the-world type
innovation. At the same time, the developing countries are
benefiting from catch-up type
innovation and associated R&D. The productivity output of such
research is considered to be
higher and have lower lag. The lag is higher in the case of new -
to-the-world innovation as
was observed in the case of the acceleration in productivity
growth that started in the
technology sector and spread to the overall economy only many
years later leading to the
rapid productivity growth period of 1995 to 2004. A recession
followed this period, as such

overall productivity growth was decreasing in the developed
countries in our sample set. We
discuss these result along with the “productivity puzzle” further
in Section 2.6.
14 Note that when resource and agricultural dependency
dummies are used instead of actual resource rents and
agricultural value added the pooled data for three periods shows
significant results at 10% for both effective
tertiary education and R&D expenditures.
40
Table 2.4 – Labour Productivity Growth and Policy Variables –
High Income OECD and
Developing Countries Separately
Developed Countries
Productivity Growth
Developing Countries
Pooled Pooled Pooled Pooled Pooled Pooled

Period
1 & 2
Period
2 & 3
Period
1, 2 & 3
Period
1 & 2
Period
2 & 3
Period
1, 2 & 3
Initial Productivity -0.001 0.057 0.032 -0.076** -0.024 -0.043*
(0.088) (0.083) (0.066) (0.030) (0.033) (0.026)
Effective Tertiary Education 0.022 0.023 0.020 0.024 0.038
0.036
(0.038) (0.035) (0.028) (0.041) (0.033) (0.028)
Effective R&D -0.006 -0.009 -0.009 0.164 0.165* 0.166*

(0.020) (0.019) (0.015) (0.103) (0.097) (0.079)
Economic Freedom -0.002 0.002 0.000 0.002 -0.004 -0.002
(0.002) (0.002) (0.002) (0.002) (0.002) (0.002)
Educational Attainment 0.001 0.001 -0.002 0 0.004 0.001
(0.006) (0.007) (0.006) (0.004) (0.007) (0.005)
Adj. R-squared 0.172 0.632 0.491 0.349 0.326 0.321
N 44 45 67 127 135 196
* p<0.10, ** p<0.05, *** p<0.01,
Note: Regional dummies, time dummies, educational
attainment, natural resource rents, and agricultural value
added included as control variables
This result is consistent with Nadiri and Kim (1996) who find
the rate of return for domestic
R&D spending to be between 23% and 26% varying amongst
different countries. The
breakdown of government effectiveness and R&D expenditures
can be explained in similar

terms as effective education expenditures as a percent of GDP.
A hypothetical country with
1% effective expenditure on research and development as a
percent of GDP driven by 0.45
government effectiveness, investing 2.2% of GDP as research
and development expenditure
and having an annual growth rate of 3 percent can improve its
growth rate to 3.825% (that
is an increase by 27.5%) by increasing its effective expenditure
to 2%. This increase in effective
41
expenditure could be accomplished either through improving
government effectiveness to 0.9
or research and development expenditure as a percent of GDP to
4.4%. As in the case of the
regression where all countries are included, we find that the
signs of the coefficients for
effective tertiary education, effective R&D and Economic
Freedom do not vary and the
magnitudes do not vary by considerable extent when included
individually in the estimation.
We find a positive effect of effective tertiary education on
labour productivity growth for both

developed and developing countries. However, the coefficients
are not significant as it was
observed in the case of the pooled data for all countries. We
observe that effective tertiary
education expenditures are important for labour productivity
growth in the non-traditional
sector in addition to initial educational attainment represented
by average years of schooling.
Also, the exclusion of the educational attainment represented by
average years of schooling
does not affect the results. Robustness tests show that the
inclusion of natural resource rents
and agricultural value added in the regression equation (instead
of the resource dependency
dummy) does not change the results.
We have excluded the possibility of reverse causality. In this
chapter, we have accounted for
the initial economic state of the country, the initial level of
educational attainment in the
country region-specific differences, and time-specific
differences. Also, the labour productivity
growth variable is lagged by a period of five years in order to
exclude the possibility of reverse

causality. In terms of omitted variable bias, we acknowledge
that not accounting for capital
investment in the estimation equation may lead to a bias in the
estimation.
The important question is whether we can assume plausible
causality in the case where we
observe statistically significant relationships in our empirical
outcome or not. We account for
lagged labour productivity. The lagged capital investment itself
is expected to be associated
with lagged labour productivity, which in turn is expected to be
associated with labour
productivity growth. In this situation, if we are interested in
determining the exact magnitude
of the effect of lagged labour productivity on labour
productivity growth in the modern sector
then the omitted variable bias is of serious concern. However,
we are trying to determine
whether policies such as tertiary education expenditures,
42
expenditures and business environment affect labour

productivity in the modern sector.
Considering lagged labour productivity in the estimation model
we are able to account for
much of the omitted variable bias concerning policies as the
missing factors would be
associated with the lagged labour productivity. As such, it is
plausible that the interaction of
higher government effectiveness and higher investment in
tertiary education as a percent of
GDP leads to higher labour productivity growth excluding
natural resource and agricultural
rents. We also consider that tertiary education expenditures,
expenditures and government effectiveness may be associated
with other omitted variables.
For example, educational attainment can be correlated to
tertiary education expenditures and
research and development expenditures. Considering that
governments with higher
educational attainment may be able to invest more in both
education and R&D simply
because of the availability of educated populace. This leads to a
simultaneity problem.
Investing in tertiary education and R&D would most probably

not be the first choice of the
government of a country that has overall lower educational
attainment. In this sense, we note
that our estimation suffers for the omitted variable bias related
to simultaneity concerns. This
might be leading to an over-estimation or under-estimation of
the relationship between the
explanatory variables and our variable of interest labour
productivity in the non-traditional
sectors. However, we have a choice to make in terms of
selecting our variables of interest and
are restricted by the coverage of data that we intend to keep
geographically wide.
2.5.2. Arabian Gulf countries - A special case?
We also present results for Arabian Gulf country dummies in
contrast to the reference region
(includes North America, Western Europe and Nordic countries
- See Appendix 2-B for more
details) in Table 2.5 below and compare them to those already
seen in Table 2.3 above. We
observe that the growth in labour productivity in the non-
traditional sector in the Arabain
Gulf region is much lower in comparison to the reference group.
With rising oil prices from

2003 onwards most of the growth in Arabian Gulf economies
appears to have been mostly
based on resource rents (Ftiti, et al., 2016). We observe in Table
2.5 below, that, the same
regression without excluding natural resource rents and
agricultural value added, results in
43
diminished statistical significance for the Arabian Gulf
countries’ dummy variable for the
pooled sets. This result indicates that the growth in the non-
natural resource sector has been
slower in comparison with the reference group. It is noteworthy
that the coefficient of the
Arabian Gulf Dummy is significant for Period 2 in both cases
where labour productivity
growth excludes and includes natural resources rents. Periods 1
and 3 also corresponds with
low oil prices.
Table 2.5 – Total Labour Productivity Growth and Policy
Variables

(Inclusive of natural resource rents and agricultural value
added)
Period 1 Period 2 Period 3 Pooled Pooled Pooled
1998-2003 2003-2008 2008-2013 Period
1 & 2
Period
2 & 3
Period
1, 2 & 3
Initial Productivity -0.084*** -0.053** 0.009 -0.074*** -0.041
-0.065***
(0.026) (0.026) (0.022) (0.018) (0.027) (0.020)
Effective Tertiary Education 0.013 0.036 0.011 0.022 0.013
0.007
(0.028) (0.029) (0.018) (0.020) (0.025) (0.020)
Effective R&D -0.014 -0.050** -0.006 -0.030** -0.027 -0.019
(0.020) (0.021) (0.015) (0.015) (0.020) (0.016)
Economic Freedom 0.001 -0.002 -0.001 -0.001 -0.003 -0.002*

(0.001) (0.002) (0.001) (0.001) (0.002) (0.001)
Arabian Gulf Dummy -0.098 -0.164* -0.08 -0.112* -0.081 -
0.065
(0.098) (0.088) (0.092) (0.064) (0.097) (0.074)
Adjusted R-squared 0.617 0.381 0.049 0.521 0.112 0.228
N 91 96 100 187 196 287
* p<0.10, ** p<0.05, *** p<0.01
Note: Regional dummies, time dummies, educational
attainment, natural resource rents, and agricultural value
added included as control variables
As such in the following, we attempt to substantiate the effect
of oil price on non-traditional
sector labour productivity growth. In Figure 2.2 the predicted
labour productivity growth
excluding natural resource rents and agricultural value added
for two Arabian Gulf countries
44

(Oman and Saudi Arabia) and two reference group countries
(Netherlands and Norway) is
plotted against the annual growth rate of crude oil price. The
predicted labour productivity
growth function is computed for each country by using their
respective data points and
estimation results of pooled data for periods 1, 2 and 3 as shown
in Table 2.3. In Figure 2.2
it is observed that lower non-traditional sector labour
productivity growth in the Arabian
Gulf countries Oman and Saudi Arabia is associated with higher
oil prices and vice versa ,, 15
but not for the two countries from the reference group Norway
and Netherlands. This provides
confirmation that oil prices partly drive the non-traditional
sector labour productivity growth
and innovative development in the Arabian Gulf countries.
Figure 2.2 – Predicted labour productivity growth as a function
of the annual growth rate of
crude oil prices
15 The years 1999 and 2000 witnessed strong oil price recovery
after the oil price crash related to Asian Financial
Crises. Excluding 1999 and 2000 would results in an even

stronger correlation of oil price growth with labour
productivity growth.
45
2.6. Conclusions and Discussion
This chapter presents the analyses of the relationship between
innovation policy and
productivity growth related to innovation and catch-up. It
establishes the correlation and
plausible causality between innovation policies and labour
productivity growth in non-
traditional sectors in a cross-sectional evaluation among
countries. A selection of innovation
policies was chosen based on the literature review and the state-
of-the-art “broad” innovation
policy approach. Innovation policy in this chapter is represented
by indicators of education,
research and development, and business. The policy
implementation capability and potential
of the governments are also analysed.
In our results, we observe the convergence between countries
with lower labour productivity

and those at the innovative frontier. This result is in line with
earlier findings of convergence
in labour productivity between richer and poorer countries –
beta-convergence (Barro, 1991;
Barro, 2012). Also, a study by Verspagen (1991) confirms the
catching-up of relatively
backward countries through technological spill-overs. Further,
we observe that there is a
significant and positive relationship between the interaction of
government effectiveness and
government expenditures in tertiary education, and labour
productivity in the modern sector.
This observation answers one of the questions raised in Keller
(2006), where the returns to
tertiary education are not found to be consistently positive.
Keller (2006) hypothesizes that
tertiary education expenditures might be inefficiently allocated.
We consider the
multiplicative term of government efficiency and tertiary
education investment while
including tertiary education investment. We found that the
interaction of government
efficiency and tertiary education expenditures as a percent of
GDP were positively and

significantly related to labour productivity growth in non-
traditional sectors. We could also
challenge the notion that primary and secondary investment has
priority over tertiary
education investment, on the basis of economic returns, by
including the initial educational
attainment in the form of years of schooling in the explanatory
variables. The initial level of
educational attainment in the country turns out to be not
significantly correlated with labour
46
productivity growth in the modern sector, while tertiary
education is. This is important for
policymakers as it demonstrates substantial societal returns to
tertiary education.
When separating developing countries and developed countries,
we do not observe a
significant effect of effective tertiary education. At the same
time, for developing countries,
the coefficients of the effective R&D expenditures show a
consistently positive and statistically
significant effect on labour productivity growth. This

relationship contrasts with findings
elsewhere, which often highlight the importance of research and
development expenditures
for developed countries, speculating the opposite for developing
countries. For example,
Griffith et al. (2004) point out that developing countries are not
able to achieve the maximum
potential in R&D. They see this as a consequence of
inappropriate social policies. Our results
indeed highlight that the influence of the interaction between
the government effectiveness
and R&D expenditures is positive. Through these results, the
importance of looking at
innovation policies as a complete set within an innovation eco-
system rather than only looking
at them individually is highlighted further. These results are
unique and to the best of our
knowledge first of their kind in confirming the interaction of
sound governance and innovation
policy measures such as expenditures in tertiary education and
R&D.
In line with the academic literature, we find that near-term
lagged effective investments in
R&D do not result in increased productivity in developed

countries. Following the moderate
growth of the 1980s, the developed countries witnessed high
productivity growth in the years
from 1995 to 2004. This productivity growth episode was
associated with the maturity of the
technological revolution. The rapid growth in the application of
technological advances in
productivity-enhancing innovations, and semi-conductor and
computer manufacturing lead to
rapid labour productivity increases from the mid-1990s
(Manyika, et al., 2001). This period
was followed by a recession in 2008. There are thus two main
reasons we do not find positive
returns of effective R&D expenditures for developed countries.
Firstly, as observed during the
technological revolution, the effects of R&D investment and
new-to-the-world innovation take
more time to yield productivity increases than we have
considered in this chapter. The
47
productivity increases in the first two periods were associated
with R&D expenditures that

were mainly carried out in the last 10 to 20 years and not during
the preceding five years.
The findings suggest evidence on the “Solow paradox” or the
“productivity paradox” that is
found in the manufacturing outside technology-producing
sectors (Acemoglu, et al., 2014).
There are three other possible explanations of the “Solow
Paradox”. One argument is that the
current innovations are not as impactful as those of the first and
second industrial revolutions,
such as the steam engine, electricity, piped water and sanitation,
and antimicrobial drugs
(Gordon, 2012). The second argument is of secular stagnation,
that is, the decline of growth
due to the ageing population and lower investments in capital,
despite the productivity-
inducing innovations (Eggertsson, et al., 2016). Finally, the
third argument is related to the
mismeasurement of productivity, such as the difficulties in
measuring the output of cheaper
software and accounting for the benefits of internet-based
services (Mokyr, 2013).
We do not find any relationship between labour productivity
growth and the index of

economic freedom. In other words, we cannot demonstrate that
a good business environment
as defined by the index of economic freedom is conducive to the
transformation of knowledge
and research into marketed goods and services. The results may
be a consequence of the type
of indicator we have selected to represent the quality of the
business environment. The
variable used presents only a bird-eye view of the business
environment. Also, our specification
fails to catch a potentially shorter response time to business
policies. It would be ideal in
future research to work with different time lags for business
conditions and to work with
indicators that objectively represent the business policy and
environment in the countries.
Overall, Arabian Gulf countries experience lower labour
productivity growth in the non-
traditional sector as the oil prices increase. For these countries,
a crucial policy implication is
to devote resources towards tertiary education and R&D, while
improving government
effectiveness, if they want to grow independent of oil and gas
resources.

Innovation and Diversification Policies for Natural R

Innovation and Diversification Policies for Natural R

Recommended

Recommended

More Related Content

Similar to Innovation and Diversification Policies for Natural R

Similar to Innovation and Diversification Policies for Natural R (20)

More from LaticiaGrissomzz

More from LaticiaGrissomzz (20)

Recently uploaded

Recently uploaded (20)

Innovation and Diversification Policies for Natural R