Economic Commission for Latin America and the Caribbean (ECLAC, United Nations)

ICT AND KNOWLEDGE COMPLEMENTARITIES A factor analysis on growth www.eclac.org/SocInfo Marco Capasso and Nelson Correa [email protected] [email protected]

ICT and Knowledge Complementarities Contents 1. Our thesis 2. Theoretical framework 3. Data 4. ICT and knowledge 5. ICT and economic growth 6. Comparative analysis

1. Our thesis

Our thesis •

Knowledge is important to explain economic growth.

•

ICT use also increases growth, through knowledge.

• ICT is complementary to other means of production of knowledge.

ICT

Knowledge

Growth

From ICT use to Growth Scientific Publications

ICT use Knowledge Enrollment

R&D

Economic Growth Capital

Labor

2. Theoretical Framework

Knowledge and Growth

“The models that treat knowledge as an input are in striking contrast. […] Constant returns in Eq. (3) must then be assumed to be a fortuitous accident. An alternative is to assume that α + β = 1 in which case there are increasing returns to scale when L, K and C are increased in proportion.” (Carlaw Lipsey 2002)

ICT and Knowledge Complementarities “Those countries which had an educational and training system capable of a rapid and effective response to the huge demands for new skills coming from the ICT technoeconomic regime had a big comparative advantage in the efficient implementation of ICT itself.” (Christopher Freeman 1995). “However, our aim here is not to engage in the rather sterile debate on where to put the exact borderline between different forms of knowledge. The point we would like emphasize is that for firms these are in most cases complementary (rather than substitutes).” (Fagerberg and Hildrum 2002)

Definition of Complementarity

The sum of the changes in the function when several arguments are increased separately is less than the change resulting from increasing all the arguments together

“The whole is more than the sum of parts”.

3. Data

Database Dimension •

73 countries, of which…

•

… 15 from Latin America and the Caribbean;

•

time-span from 1996 to 2001 included;

•

in order to catch also one-year lagged relations among the variables, we compute for each series a two-year average.

Knowledge Proxies 1. Percentage of inhabitants having completed tertiary education 2. Percentage of GDP devoted to investment in R&D 3. Number of scientific publications (weighted for each country’s population) 4. ICT use (Info-use index by Orbicom, network of UNESCO Chairs in communications)

4. ICT and Knowledge

Knowledge index: construction • The observed co-movements among the four knowledge proxies are used to build a Knowledge Index by means of the principal factor procedure.

• The Knowledge index defines a level of Knowledge for each country for each year.

Knowledge index: ranking

Knowledge index: composition

Optimal Complementarity Proportion • Knowledge proxies are expected to be linked one another in a given proportion for any given level of the knowledge index.

• This proportion represents the golden proportion for which complementarity works optimally.

• Therefore, complementarity implies a negative relation between the distance from the golden proportion and economic performance.

Distance from the Optimal Complementarity Proportion • For each country and each year, a measure of the Euclidean distance from the optimal complementarity proportion can be built.

• It represents a measure of inefficiency.

5. ICT and economic growth

Regression of GDP We operate the following regression:

in order to evaluate the impact on GDP of the knowledge index and of the distance from the optimal complementarity proportion.

All variables are in logarithms but the distance.

Empirical analysis: estimation

Comments on estimation results: Knowledge

= 0.3821 Knowledge has a significant strong positive impact on GDP.

Comments on estimation result: Capital and Labor

= 0.2544 = 1.0302 = 0.7758 There are constant returns to scale in the inputs Capital and Labor.

Comments on estimation results: All inputs

= 0.2544 = 0.7758

= 1.4123

= 0.3821 There are increasing returns to scale in the inputs Capital, Labor and Knowledge.

Comments on estimation results: Distance from complementarity

= -0.0245 Distance from the golden proportion has a significant small negative impact on GDP: complementarity works better when the golden proportion holds.

6. Comparative Analysis

Simulation of an ICT-use increase

Effect on GDP of a 10% increase in ICT use

Comparative analysis (1): Argentina and Chile

• In the initial 2001 position, Chile shows a higher ICT use than Argentina, but a lower enrollment ratio.

• GDP percentage increase will be higher for Argentina than for Chile.

Comparative analysis (2): The Caribbean

• Panama starts from levels of enrollment and R&D which are similar or higher than the levels of Costa Rica, Trinidad & Tobago and Jamaica, but has in 2001 a lower ICT use index. • The effect on GDP is expected to be much higher for Panama.

Comparative analysis (3): Mexico and Venezuela

• In 2001, the ICT use of Mexico seems to be high with respect to the other knowledge proxies.

• Mexico´s GDP Venezuela´s.

increase

is

much

smaller

than

Conclusions • Knowledge plays an important role in explaining GDP growth. • ICT acts as a complement to the knowledge base of a country. • Best results in terms of GDP growth are obtained by those countries for which a given “golden” proportion holds among knowledge proxies • When the current ICT use is below the golden proportion, an increase in ICT will perform better in terms of GDP growth.